HAEMOSTATIC AND THROMBOTIC DISORDERS: A JOURNEY FROM BENCH TO BEDSIDE SARAH MURIEL MEIRING Submitted in fulfillment of the requirements of the degree of Doctor in Medical Sciences (Haematology) in the Faculty of Health Sciences at the University of the Free State December 2019 Supervisor: Prof SC Brown Co-supervisor: Prof HF Kotzé 1 ACKNOWLEDGEMENTS I first want to thank our Heavenly Father for providing me with the opportunities and skills to make this possible. To Him my deepest gratitude. There are people, each in his or her unique way, who influenced my career: • My parents, who always loved, supported and encouraged me. • My science teacher, Mr Mike Olivier, who had the ability to open horizons to a young learner to think beyond just science. • Prof Pieter Pretorius, the Head of the Department of Physiology at the Potchefstroom Campus of the North-West University, who instilled in me a love for physiology. • Prof Philip Badenhorst, the former Head of the Department of Haematology and Cell Biology at the University of the Free State, who allowed me the freedom to follow my instincts and asked those questions that forced me to stop and reconsider the explanation extracted from results. • Prof Harry Kotze, my mentor, postgraduate supervisor and co- supervisor of this thesis, who believed in my abilities to become the best scientist I can – an NRF-rated scientist who has published widely in mainly international peer-reviewed scientific journals. • Prof Stephen Brown my study leader for his wisdom and initiatives. I appreciate the fact that he made this possible. The journey recounted in this compilation would not have been possible had I not been in contact with able and dedicated collaborators in South Africa, Belgium, Hungary and Australia. I extend my deepest appreciation to the following individuals and groups for their unique contributions to this research: • Prof. Hans Deckmyn from the Kortrijk Campus of the University of Leuven, Belgium, who allowed me to do a post-doctoral study in his 2 laboratory where he instilled the love for thrombosis and haemostasis research. • Prof. Jolan Harsfalvi from the Univeristy of Debrecen, Hungary, who collaborated with me on studies involving the collagen-binding property of von Willebrand factor; a topic close to our hearts. • Prof. Emmanuel Favaloro from the Royal College of Pathology in Sydney, Australia, who contributed to internationalise our reference centre for laboratory diagnosis of von Willebrand disease by participating in several international studies. • Prof. Alta Schutte and Prof. Leoné Malan from the North-West University, Potchefstroom Campus, who collaborated on several South African studies researching where increased von Willebrand factor levels might lead to thrombosis. I am deeply indebted to my colleagues – whom I would rather call my friends – who had been members of the research team and who contributed their time and expertise to research endeavours. They are Prof. Marius Coetzee, Jaco Joubert, Mareli Kelderman, Seb Lamprecht, Jan Roodt, Charmaine Conradie, Mmakgabu Khemisi, Rethabile Maleka, Anneke van Marle and Leriska Haupt. My scientific research was supported over a long period by the National Research Foundation, the Medical Research Council of South Africa and the Research Trust of the National Health Laboratory Services. Specialised equipment was funded by the University of the Free State. Without these funding opportunities, this research would not be possible. Finally, I cannot begin to thank my husband, Hermie, and our daughter, Marieke, who never complained about various avenues they had to travel with me to reach my goals, and who unconditionally love me. Sarah Muriel Meiring December 2019 3 DECLARATION Hereby I, Sarah Muriel Meiring, declare that the compilation in respect of the D Med Sc degree that I herewith submit at the University of the Free State is my independent work, and that I have not previously submitted it for a qualification at another institution of higher education. I declare that in all applicable publications, I had been involved in writing up the process of literature research, conceptualisation, and refining techniques on the bench; I had been deeply involved in the study design, data management, analysis and interpretation of data; condensing drafts, figure selection and approving the final versions of the manuscripts. I am aware that the copyright vests in the University of the Free State and that all royalties with regard to intellectual property that had been developed during the course of and/or in connection with the study at the University of the Free State, will accrue to the university. SIGNATURE: DATE: 6 December 2019 4 CERTIFICATE OF EDITING 5 TABLE OF CONTENTS SYNOPSIS ......................................................................................................... 7 1. INTRODUCTION ....................................................................................... 9 2. A JOURNEY FROM BENCH TO BEDSIDE ........................................... 12 2.1 Part 1: Testing of antithrombotics ....................................................... 12 2.2 Part 2: The use of phage display technology to develop antithrombotic peptides and antibodies and to develop cost-effective diagnostic tests ..................................................................................... 17 2.3 Part 3: Diagnosis of thrombotic and haemostatic disorders ............ 26 3. GENERAL SUMMARY ............................................................................ 36 4. REFERENCES ........................................................................................ 43 5. LIST OF PUBLICATIONS RELATED TO THIS COMPILATION (Supplements) ........................................................................................ 46 6 SYNOPSIS This compilation focuses on thrombotic and haemostatic disorders, illustrating my journey from basic research on thrombotic and haemostatic disorders to the differential diagnoses of these disorders. For purposes of clarity I shalll divide it in three parts. The first part includes the testing of antithrombotic agents. My scientific career started with the testing of antithrombotic drugs in a baboon model of arterial thrombosis. These antithrombotic drugs were mostly targeted at platelets and, to a lesser extent, coagulation. For my PhD, I clarified the catabolism, pharmacokinetics and exctretion of recombinant hirudin, an anti-thrombin drug. The second part includes the development of cost-effective diagnostic tests, mostly for von Willebrand disease (VWD), the most common congenital bleeding disorder. This comprises the largest part of the thesis. I developed four anti-thrombotic peptides by using Phage Display technology that I mastered during my post-doctoral study at the University of Leuven in Belgium. I was also part of the group of researchers that developed many new thrombosis models in baboons. For my M.Med Sc study, I developed a flow chamber model to study in vitro endothelial function, which was subsequently used to test the thrombogenicity of tissue-engineered small vessels. This study was the first where endothelial cells were successfully seeded onto decellularised baboon arteries. This study was undertaken in collaboration with the Department of Cardiothoracic Surgery at the University of the Free State, Bloemfontein. The third part is a spin-off from my research on VWD. I established the only Special Haemostasis laboratory of the NHLS, situated at the University of the Free State in Bloemfontein, South Africa, by developing, validating and implementing four diagnostic assays. As a result the laboratory now functions as a reference centre for von Willebrand disease; the most prevalent, but underdiagnosed bleeding disorder in South Africa. The developed diagnostic 7 assays resulted in nineteen peer-reviewed publications on the diagnosis of haemostatic and thrombotic disorders. As reference centre for VWD in South Africa, we published eight articles on its diagnosis, together with international leaders in the field. Additionally, we also published the first South African recommendations regarding the differential diagnoses of VWD. We published on other bleeding disorders and on a fatal thrombotic disorder, thrombotic thrombocytopenic purpura (TTP). Lastly, is it important to note that the research on VWD, haemophilia and TTP is ongoing. 8 1. INTRODUCTION Thrombosis is considered as a major cause of death worldwide, being responsible for heart attacks and strokes (Heidenreich et al., 2011). In addition, venous thromboembolism (VTE) is the major preventable cause of death in hospitals (Tsai et al., 2015). Strong evidence indicates that prophylaxis is still under-prescribed (Spyropoulos, 2010). Although thrombosis is the main cause of death in industrialized countries, thrombosis complications in patients infected with the human immuun difficiency virus (HIV) are on the rise in developing countries. HIV infection is now a known pro-thrombotic condition, because cases where HIV infection was linked to venous thrombo-embolisem (VTE) had been reported since the late 1980’s (Tsongo Vululi et al., 2018 ). A survey conducted by the Medical Research Council (MRC) in South Africa stated that, every day, 195 people died from some type of cardiovascular disease during the time period 1997 to 2004. Of these, 33 people died due to heart attack, 60 due to stroke and 37 people due to heart failure. Furthermore, the MRC projected that chronic disease (including heart disease) would increase significantly by 2010 and that this increase would continue thereafter. Before the age of 65, most people die due to chronic disease. These premature deaths have a huge effect on the workforce and a major economical impact and were expected to increase by 41% by 2030. It was also estimated that the cardiovascular disease (CVD) burden in South Africa would increase among all age groups to become the prime contributor to overall morbidity and mortality (Bradshaw et al., 2003). The American Heart Association further estimated that by 2030, the CVD death rate would be more than 23 million people world-wide (Heidenreich et al., 2011). Thus, the main challenge in cardiovascular research is to develop safe antithrombotics to prevent thrombosis, but cause less bleeding than the existing, commonly used clinical drugs (coumarin, heparin, direct thrombin inhibitors and factor Xa inhibitors). All these anticoagulants have limited success to maintain the haemostatic balance between thrombosis and bleeding (Hirsh & Weitz, 1999, Wolberg et al., 2012). 9 Bleeding disorders, that are on the other side of the haemostatic balance, largely influence the quality of life of patients and are also a burden on the economy. Haemophilia has a prevalence of 1:5000 boys in South Africa. Unfortunately, many patients with haemophilia or other bleeding disorders are either not diagnosed or diagnosed at a very late stage of the disease. Even more concerning, the South African Haemophilia Foundation (SAHF) estimates that many of those that are diagnosed remain without adequate care (Mahlangu, 2009). Von Willebrand disease (VWD) is the bleeding disorder with the highest prevalence of 1% in the general population. Despite this high prevalence, the diagnosis and classification remain a challenge (Sadler, 2005). The severity of a bleeding disorder usually depends on the actual amount of clotting factor that is missing or not functioning (Jacobson, 2013; Eyal & Veller, 2009). Our Specialised Haemostasis Laboratory is the reference centre for VWD in the country and is thus the only centre that diagnoses the sub-classes of VWD in sub-Saharan Africa. Figure 1 indicates the coagulation factors in haemostasis that I targeted in my research. Figure 1: The targeted thrombotic factors (red circles) that I researched in this compilation 10 I investigated endothelial cells and their function for my M.Med.Sc. Although I did not include my M.Med.Sc study in this compilation, my main research theme thereafter was concerned with the function of endothelial cells which include VWF and its cleaving protease ADAMTS13, and the important role that tissue factor (TF) plays in thrombosis and bleeding. I also developed and tested drugs that inhibit platelets and thrombin, both main players in thrombosis. These studies will be explained further in the compilation. For my PhD, I determined the sites of excretion and the pharmacokinetics of the recombinant thrombin inhibitor, recombinant (r)-hirudin, kindly supplied by Prof Fritz Markwardt from the Medical Academy in Erfurt, Germany (S-1). The sites of excretion and/or uptake of r-hirudin was previously unknown and I was the first to show that the half-life to r-hirudin, administrated as an intravenous bolus or 30 minute infusion, did not differ significantly. For this purpose, the r-hirudin was labelled with Iodine-131 and its in vivo distribution and sites of elimination were monitored with a scintillation camera, coupled with a computer-assisted image analysis program. The half-life of Iodine-131- r-hirudin was 21 ± 3 minutes. I also found that only 50% to 60% of lepirudin was excreted by the kidney. This was in contrast to the general assumption that r-hirudin is mainly excreted by the kidneys. I also showed that a sizeable portion of r-hirudin was found in the bile that indicated that it was eliminated by the liver. It is important to note that the method of delivery (bolus or infusion) did not influence the plasma half-life and clearance of r-hirudin. This compilation explains the journey through the different parts of my research. I shall discuss the contribution of my publications to the international body of knowledge in three distinct parts. For purposes of inclusiveness, the publications to which I refer is attached at the end of the document. The three parts of this compilation are outlined in Figure 2. S-1 Meiring, S.M., Lötter, M.G., Badenhorst, P.N., Bucha, E., Nowak, G.& Kotzé, H.F. (1999). Sites of elimination and pharmacokinetics of recombinant 131I-lepirudin in baboons. Journal of Pharmaceutical Sciences 88(5):523- 529. (Cited by 2) 11 Figure 2: The three parts of this compilation include testing of antithrombotics, development of cost-effective diagnostic methods and diagnosis of haemostatic and thrombotic disorders. 2. A JOURNEY FROM BENCH TO BEDSIDE My journey from bench to bedside may accurately be described as “Haemostatic disorders: towards a better understanding of mechanisms and treatment. It is important to note that unraveling the molecular mechanisms underlying haemostatic diseases does not only lead to a better understanding thereof, but also to the potential development of novel drugs for treatment and of tests to diagnose them (Bradshaw et al., 2003). 2.1 Part 1: Testing of antithrombotics Following my research for the M.Med.Sc and PhD (See section 1: Introduction), I started my research by testing anti-thrombotic agents in baboon thrombosis models. In collaboration with Prof. Hans Deckmyn from the Laboratory for Thrombosis Research of the Kortrijk Campus of the University of Leuven in Belgium, we tested the efficacy of platelet membrane receptor human monoclonal antibodies that inhibit arterial thrombosis. 12 My colleagues developed, in collaboration with Dr Stephen R. Hanson from the Scripps Clinic and Research Foundation in La jolla, California, USA, a baboon model of platelet dependent arterial-type thrombus formation (Hanson et al., 1985). This model consists of an arterial-venous shunt that is interposed between the femoral artery and vein and that contains a Dacron vascular graft insertion acting as an arterial-dependent thrombus generator. The first monoclonal antibody, MA-16N7C2 that we tested in our baboon model contains an echistatin-like RGD sequence that binds to the platelet receptor glycoprotein IIb/IIIa and prevents fibrinogen binding and so prevents platelet-platelet interaction (S-2). A bolus injection of 1mg/kg MA-16N7C2 occupied significantly more GPIIb/IIIa receptors for a longer period than a lower dose of 0.4 mg/kg, suggesting that the receptors might be internalised by the platelets. In the baboon, bolus injections of 1.0 and 0.4 mg/kg significantly inhibited ex vivo platelet-dependent deposition on the vascular graft material. It also prolonged the bleeding time and inhibited ex vivo platelet aggregation. These effects showed a dose-dependent response and lasted for several days. Therefore, MA-16N7C2 is regarded as a potent and long-acting GPIIb/IIIa inhibitor. Unfortunately, its effects were so substantial that it increased the risk of uncontrolled in vivo bleeding. In addition, because they are large proteins, monoclonal antibodies trigger the immune system to develop antibodies against them. It is therefore conceivable that MA-16N7C2 can only be used once in a patient. This study nevertheless showed that inhibition of the GPIIb/IIIa receptor effectively inhibits thrombosis. The platelet receptor glycoprotein Ibα interacts with VWF and mainly support platelet adhesion in arterial flow conditions at the site of injury to prevent platelet binding to subendothelial collagen. Prof. Hans Deckmyn developed a F(ab) fragment of the monoclonal antibody 6B4. This antibody inhibits the binding of platelet glycoprotein Ib (GpIb) to VWF. In addition, this F(ab) S-2 Kotzé, H.F., Badenhorst, P.N., Lamprecht, S., Meiring, M., Van Wyk, V., Nuyts, K., Stassen, J.M., Vermylen, J. & Deckmyn, H. (1995). Prolonged inhibition of acute arterial thrombosis by high dosing of a monoclonal anti- platelet glycoprotein IIb/IIIa antibody in a baboon model. Thrombosis and Haemostasis 74:751-757. 13 fragment does not trigger the immune response. We investigated the antithrombotic effect of 6B4 and its F(ab) and F(ab’)2 fragments in our baboon model (S-3). In this study we substituted the Dacron vascular graft material insertion (S-2) with fixed bovine pericardium to provide a collagen-rich surface. When we injected this antibody in baboons, both the intact IgG and its F(ab)2 fragments immediately caused thrombocytopenia and could not be investigated further. In contrast, the F(ab) fragments did not cause thrombocytopenia. Baboons were treated with bolus injections of 80 µg/kg, 160 µg/kg, 320 µg/kg and 640 µg/kg 6B4 F(ab) fragments. These doses remarkably reduced ex vivo platelet deposition on a collagen-rich surface by 43%, 53%, 56% and 65% respectively. The highest dose of 640 µg/kg significantly prolonged the bleeding time. With this study we were the first to show that an anti-human GpIb antibody is able to successfully prevent platelet adhesion and thrombus formation in vivo. This finding confirms the predominant role of GpIb in in vivo platelet adhesion. Another baboon model that we used to study the antithrombotic effect of the F(ab) fragment of 6b4 was the modified Folts model where we induced thrombus formation at a mechanically injured and stenosed site of the femoral artery. In this case, cyclic flow reductions (CFRs) measured on an extracorporeal femoral arterial-venous shunt was used as measure of the antithrombotic effect (S-4). We also measured receptor binding and the in vitro and ex vivo platelet aggregation response due to the F(ab) fragment. Increasing doses of 6B4 F(ab) decreased the platelet deposition onto the injured femoral artery devices in a dose-dependent manner. The F(ab) fragments prevented ex vivo ristocetin and botrocetin-induced platelet agglutination dose-dependency. The IC50 (dose that inhibit agglutination by 50%) was 1.8 ug/ml at a plasma F(ab) concentration of 36nmol/L and 2.5 ug/ml at 40 nmol/L. The F(ab) fragments bind to baboon platelets in a dose- S-3 Cauwenberghs, N., Meiring, S.M., Vautarin, S., Van Wyk, V., Lamprecht, S., Roodt, J.P., Novak, L., Harsfalvi, J., Deckmyn, H. & Kotzé, H.F. (2000). Antithrombotic effect of platelet glycoprotein Ib-blocking monoclonal antibody Fab fragments in nonhuman primates. Arteriosclerosis, Thrombosis and Vascular Biology 20:1347-1353. (Cited by 141) S-4 Fontayne, A., Meiring, M., Lamprecht, S., Roodt, J., Demarsin, E., Barbeaux, P. & Deckmyn, H. (2008). The humanized anti-glycoprotein 1b monoclonal antibody h6B4-Fab is a potent and safe antithrombotic in a high shear arterial thrombosis model in baboons. Thrombosis and Haemostasis 100:670-677. (Cited by 58) 14 dependent manner and reached a saturation level of 52%. It is important to note that this treatment had a small and insignificant influence on bleeding time. In addition, no blood loss, spontaneous bleeding or thrombocytopenia was observed. These results clearly indicated that the F(ab) fragments of monoclonal antibody 6B4 successfully prevent platelet adhesion and subsequent thrombus formation in vivo. These findings emphasized the important role that GPIbα plays in in vivo platelet adhesion and thrombosis. Based on the abovementioned results, we also predicted that the F(ab) fragments of the anti-GPIbα MoAb 6B4 may be used to constitute a useful approach to prevent arterial thrombosis in patients after balloon catherisation, vascular engraftment or endarterectomy. The modified Folts model was also used to compare, separately and in combination, the antithrombotic effect of the GPIb/IX MoAb fragment of 6B4 and the GPIIb/IIIa blocking MoAb 16N7C2 (S-5). A dose as high as 2 mg/kg 6B4-F(ab) completely prevented the CFRs without prolonging the bleeding time. MoAb 16N7C2 also abolished the CFRs at a high dose of 0.3 mg/kg, but with significant prolonging of the bleeding time. By combining a low dose of 0.6 mg/kg 6B4-F(ab) with a low dose of 0.1mg/kg MA-16N7C2, CFRs was inhibited again, but without prolonging the bleeding time. This study showed that partial inhibition of both GPIb and GPIIb/IIIa blocking is more effective and safe than inhibiting only one platelet receptor, since it abolished the CFRs in our model without any bleeding risks. We also evaluated the inhibitory effect on platelet adhesion by blocking the binding of VWF to collagen in the modified Folts model using murine antihuman VWF mAb 82D6A3 which had been developed in Belgium (S-6). This antibody binds to the A3 domain of VWF and interacts with collagen fibres I and III, but not to collagen fibre VI. This was the first study that S-5 Wu, D., Meiring, M., Kotzé, H.F., Deckmyn, H. & Cauwenberghs, N. (2002). Inhibition of platelet glycoprotein Ib, glycoprotein IIb/IIIa, or both by monoclonal antibodies prevents arterial thrombosis in baboons. Arteriosclerosis, Thrombosis and Vascular Biology 22:323-328. (Cited by 84) S-6 Wu, D., Vanhoorelbeke, K., Cauwenberghs, N., Meiring, M., Depraetere, H., Kotzé, H.F. & Deckmyn, H. (2002). Inhibition of the von Willebrand-collagen interaction by an antihuman VWF monoclonal antibody results in abolition of in vivo arterial platelet thrombus formation in baboons. Blood 99(10):3623-3628. (Cited by 139) 15 showed conclusively that the monoclonal antibody in vivo, even at a high dose of 600 µg/kg 82D6A3 abolished CFRs without significant prolonging the bleeding time. This clearly indicates a safe approach to target platelet adhesion inbibition of arterial thrombosis. As a follow-up to the study described in S-6, we decided to test the antithrombotic efficacy of this antibody in a baboon model of in-stent stenosis where we used standard treatment with heparin, aspirin and clopidogrel (S-7). This was the first reported study to investigate the possible antistenotic effect of an inhibitor of the VWF-collagen interaction following stent deployment. The studies were carried out 28 days after deployment. Our results did not supply sufficient evidence that additional inhibition of platelet adhesion reduces neointimal hyperplasia compared to the current routine clinical setting. This may indicate an overestimation of the role of vascular smooth muscle cells’ mitogenic and attractant factors that are released upon platelet activation following percutaneous transluminal coronary angioplasty and stenting. This study therefore underscored the importance of evaluating new antistenotic therapies in a clinically relevant model. This model of neointimal hyperplasia may become important in future studies assessing the prevention of restenosis. With the last study of this section, we tested the antithrombotic activity and pharmacodynamics of an anti-factor VIII human monoclonal antibody LE2E9Q in our baboon model of ex vivo thrombosis described in S1 (S-8). Mab- LE2E9Q was administered as a single intravenous dose of 1.25 and 5 mg/kg and thrombosis development was recorded in an expansion chamber (coagulation-dependent venous thrombosis) and on Dacron vascular material (platelet-dependent arterial thrombosis) in an extracorporal arteriovenous shunt that was implanted between the femoral artery and femoral vein S-7 De Meyer, S.F., Staelens, S., Badenhorst, P.N., Pieters, H., Lamprecht, S., Roodt, J., Janssens, S., Meiring, M., Vanhoorelbeke, K., Bruwer, A., Brown, S, & Deckmyn, H. (2007). Coronary artery in-stent stenosis persists despite inhibition of the von Willebrand factor – collagen interaction in baboons. Thrombosis and Haemostasis 98: 1343-1349. (Cited by 18) S-8 Jaquemin, M., Stassen, J-M., Saint-Remy, J-M., Verhamme, P., Lavend’Homme, R., VanderElst, L., Meiring, M., Pieters, H., Lamprecht, S., Roodt, J. & Badenhorst, P. (2009). A human monoclonal antibody inhibiting partially factor VIII activity reduces thrombus growth in baboons. Journal of Thrombosis and Haemostasis 7:429-437. 16 (Cadroy et al., 1989). The effect on thrombosis was measured at different time points (1 hour, 24 hours and 7 days) after administration of Mab- LE2E9Q. We found that a significantly lower number of platelets was deposited in both the arterial and venous chambers with both concentrations of Mab-LE2E9Q after 1 hour and 24 hours after administration, but not after 7 days. The antibody also did not lengthen the bleeding time. The results thus suggested that this antibody may be used as a novel type of long-acting antithrombotic agent and that it displayed an optimal safety/efficacy profile. 2.2 Part 2: The use of phage display technology to develop antithrombotic peptides and antibodies and to develop cost-effective diagnostic tests The successful testing of anti-platelet drugs in our baboon models prompted me to develop new anti-thrombotic drugs myself. These peptides and proteins were eventually used to develop diagnostic assays. I therefore applied to the South African National Research Foundation and received funding to become a post-doctoral fellow at the laboratory of our collaborator, Prof. Hans Deckmyn, head of the Laboratory for Thrombosis Research at the University of Leuven (Kortrijk Campus), Belgium. While there as a post-doctoral fellow, I mastered the technology of phage display. It is a powerful method to select proteins with a specific function where a protein or peptide is fused to a coat protein on the surface of a filamentous phage. These proteins or peptides then act as ligands or enzymes because they can bind to specific antigens of choice. During my post-doctoral study in Belgium, we identified a collagen-binding protein from the hematophagous human parasite, Necator americanus (S-9). We developed a cDNA-expressing phage display library from Necator americanus and cloned it as fusions into phagemids with the C-terminal part of the phage bound to coat protein pVI. Phages that bind to human collagen type I and III were enriched through four rounds of panning. All collagen- S-9 Viaene, A., Crab, A., Meiring, M., Pritchard, D. & Deckmyn, H. (2001). Identification of a collagen-binding protein from Necator americanus by using a cDNA-expression phage library. Journal of Parasitology 87(3):619-625. 17 binding phages were sequenced and one common 135-amino acid peptide sequence was identified that binds in a concentration-dependent manner to human type I and III collagen and rat type I collagen. This was my first experience of using phage display technology to select proteins based on their binding properties. On my return to South Africa, I published an invited review article on the application of phage display technology in thrombosis and haemostasis (S- 10). In this review I suggested that phage display technology was ideally suited to develop more effective anti-thrombotic agents. The first anti-thrombotic peptide that I identified by phage display technology was a thrombin inhibition peptide (TIP) (S-11). To achieve this, I used a cyclic heptapeptide phage display library to select phages that bind to human alpha- thrombin. One of the phage clones that bound the strongest to thrombin also competed with Phe-Pro-Arg chloromethyl ketone (PPACK) for binding to thrombin. The clone was selected and the aminoacid sequence determined. A peptide with the exact cyclic aminoacid sequence Cys-Asn-Arg-Pro-Phe-Ile- Pro-Thr-Cys was then synthesised. TIP competitively inhibited the function of thrombin with an inhibition constant (Ki) of 0.4974 mM. It prolonged the thrombin time and inhibited platelet activation by thrombin dose-dependently. It also prevented platelets from adhering onto a human micro-vascular endothelial matrix in a parallel plate flow chamber. This effect was seen under both high shear (arterial) and low shear (venous) conditions. The flow chamber studies were done in the laboratory of Prof Jolan Harsfalvi from the University of Debrecen, Hungary as described previously (Harsfalvi et al., 1995). S-10 Meiring, S.M., Kotzé, H.F., Pretorius, G.H.J. & Badenhorst, P.N. (1999). Die toepassing van peptiedblootlegging op fage in trombose en hemostase (The application of phage display technology in thrombosis and haemostasis). Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie (South African Journal of Science and Technology) 18 (3):76-81. S-11 Meiring, S.M., Littauer, D., Harsfalvi, J., Van Wyk, V., Badenhorst,P.N., Kotzé, H.F. (2002). In vitro effect of a thrombin inhibition peptide selected by phage display technology. Thrombosis Research 107:365-371. (Cited by 16) 18 By using the same cyclic heptapeptide library, I also selected a peptide with sequence Cys-Ala-Trp-Pro-His-Thr-Pro-Asp-Cys. This peptide competes with tissue factor for binding to coagulation factor VII (S-12). This peptide also lengthened the prothrombin time in a dose-dependent manner. It reduced platelet adhesion to both tissue factor and human endothelial cell matrices in a parallel plate flow chamber under high shear arterial flow conditions. Similarly, it also competes with tissue factor for binding factor VIIa with Ki 123.2 µM. In the early 2000’s, the development of antibodies and fragments thereof represented the fastest growing development in the biopharmaceutical market (Nicolaides et al., 2006). As a result, I decided to use a human single chain antibody fragment library to select single chain antibody fragments that inhibit tissue factor, the principal protein trigger of coagulation. These antibody fragments were only 26kD in size and are unlikely to be immunogenic in humans. I selected and purified a single chain variable antibody fragment that inhibited the action of human tissue factor. It prolonged the prothrombin time dose-dependently with an IC50 of 0.5 µM. Interestingly, this antibody fragment inhibited thrombogenesis more effectively in thrombophilic plasma from patients with thrombophilia than in normal plasma. Thus, TF inhibitors might achieve anti-coagulant activity in thrombophilic plasma without significantly interfering with haemostasis in normal plasma. This antibody fragment showed the strongest TF inhibitory activity currently described in the literature (S-13). Although I applied for an international patent cooperation treaty (PCT) , it was not granted because the results were published in Drug Development Research a few days before the patent application date. This intervention related to novel antibody fragments that inhibit the action of human tissue S-12 Meiring, S.M., Roets, C.E. & Badenhorst, P.N. (2006). Funksionele beskrywing van ’n faktor VIIa inhiberende peptied, IP-7, geselekteer deur faagblootleggingstegnologie. (Functional characterisation of a factor VIIa inhibiting peptide, IP-7, selected by phage display technology) Die Suid Afrikaanse Tydskrif vir Wetenskap en Tegnologie (South African Journal of Science and Technology) 25(4):209-220. S-13 Meiring, S.M., Vermeulen, J. & Badenhorst, P.N. (2009). Development of an inhibitory antibody fragment to human tissue factor using phage display technology. Drug Development Research 70:199-205. (Cited by 2) 19 factor. Unfortunately the current single chain variable antibody fragments (scFv) phage display technology is limited by the inability to produce sufficient scFv fragments for extensive testing in vivo. A larger production system with higher yield is therefore required to allow further characterisation and in vivo studies. Therefore, in a follow-up study, we manipulated expression of the low-yielding scFv in an attempt to enhance production (S-14). We improved the expression of the scFv in the cytoplasm of E-coli BL21 (DE3) by modifying the expression systems and optimizing the codons of the gene. We also evaluated two commercially available methods of protein recovery: in vitro refolding and the use of cold shock expression systems together with E.coli SHuffle®. This approach increased cytoplasic expression of the scFv expression in E. coli up to 5 times and delivered high functionality. Further improvements and/or upscaling the expression volume needs to be investigated in order to produce sufficient scFv to test for efficacy in animal models and, eventually, in humans. We used an antibody fragment library to select antibodies that bind to the VWF-propeptide to develop cost-effective assays to test for the differential diagnoses of VWF disease (S-15). In South Africa, commercially available assays are very expensive because the antibodies that are used in these assays are produced in animals and are also influenced by the unfavourable Rand/US$ exchange rate. We were the first to use human antibody proteins that were produced by phage display technology in our assays. Initially we displayed the VWF propeptide protein on yeast. We then selected antibody fragments (scFv) against the displayed VWF propeptide by using phage display technology. We selected two antibody fragments that bound with high specificity to the VWF propeptide. After purification of the antibody fragments, S-14 Vermeulen, J., Burt, F., Van Heerden, E., Cason, E. & Meiring, M. (2018). Evaluation of in vitro refolding vs cold shock expression: production of a low yielding single chain variable fragment. Protein Expression and Purification 151:62-71. (Cited by 1) S-15 Meiring, S.M., Setlai, P., Theron, C. & Bragg, R. (2018). The use of phage display and yeast based expression system for the development of a von Willebrand factor propeptide assay: Development of a von Willebrand factor propeptide assay. BioMed Research International 2018:632091:1-7 (https://doi.org/10.1155/2018/6232091). 20 we developed a cost-effective VWF propeptide assay that functions to detect VWF propeptide in normal plasma. We compared our assay’s performance to that of commercial VWF propeptide kits with antibodies (CLB-Pro 35 and CLB-pro 14.3, Cell Sciences, USA). Our selected antibodies showed a higher binding affinity for VWF propeptide than the commercially available antibodies. The limit of detection of our assay was 1.56%, where the commercial assay only detected VWF propeptide from 6.35%. Finally, we successfully secured a SA patent (2011/05426). In 2007, I established my laboratory as the reference laboratory for von Willebrand disease in South Africa and also developed it as a centre of international excellence to confirm diagnosis of von Willebrand disease. In order to achieve this, I developed and evaluated an assay that monitors VWF activity, the Collagen-Binding Assay (VWF:CB), using type III collagen. This assay is ten times more cost-effective than the commercially available methods and is also very sensitive to detect the collagen-binding activity in patients with VWD (S-16). The validation of our assay showed that the intra- and inter-assay coefficients of variation were <8% and <9% for normal values. The normal reference range varies between 51% and 143%. We could also demonstrate that this assay is sensitive to the presence of large VWF multimers. Type 2A VWD patients have a collagen-binding activity/antigen (CB/Ag) ratio of 0.18 and the CB/Ag ratio of type 2B patients is 0.4. Both these values indicate functional discordance. Furthermore, the mean CB/Ag ratios of type 2M and type 1 VWD, were 1.1 and 1.0, respectively. Although we did not compare our assay to commercial assays, our results were similar to literature findings on VWF:CB assays using type III collagen from Sigma, USA (Favaloro, 2000). I also developed a rapid and cost-effective test to visualise VWF multimer patterns to further differentially diagnose VWD (S-17). This test is used to identify VWD subclasses. Krizek & Rick (2000) described a highly sensitive S-16 Meiring, M., Badenhorst, P.N. & Kelderman, M. (2007). Performance and utility of a cost-effective collagen- binding assay for the laboratory diagnosis of von Willebrand disease. Clinical Chemistry and Laboratory Medicine. 45(8):1068-1072. (Cited by 19) S-17 Meiring, M., Badenhorst, P. & Kelderman, M. (2005). A rapid and cost-effective method to visualize von Willebrand factor multimers in plasma. Medical Technology SA 19(2):16-18. 21 and rapid method to visualise the multimeric structure of VWF using the Western blot technique. This test includes agarose gel electrophoresis followed by the transfer of proteins onto a polyvinylinine fluoride membrane. The multimeric pattern of VWF is visualised by immunolocalisation on the membrane and luminographic detection on an X-ray film. An advantage of this method is that no radioactivity is used. I modified this method comprehensively to increase its sensitivity and to reduce the cost whilst producing results quickly. I used one instead of two localisation antibodies and so reduced the immunolocalisation time by more than two hours. In the case of the Krizek test, results are available only after 3 additional hours. I further reduced the cost by using two carbon plates instead of the blotter instrument. This approach reduced the cost of the assay by at least 40%. In addition, degassing the agarose before casting and by using a 0.65% agarose gel instead of a 0.7% agarose, improved sensitivity to visualise type 1 VWD plasmas with low VWF levels sufficiently. My method also distinguishes between the multimer patterns of type 2M VWD and normal plasma, since plasma of type 2M patients shows a higher density of small multimers and a lower density of larger multimers than normal plasma. Since 2010, my reference laboratory for the diagnosis of VWD gained international recognition because we formed part of an international quality assurance programme (RCPA) where we blindly analyse test samples sent by other laboratories to assess the accuracy of their methods. In this regard I published, in collaboration with Prof. Emmanuel Favaloro from the Royal College of Pathologists of Australasia (RCPA), a definitive communiqué on the value and accuracy of the diagnostic tests for VWD. We conducted a cross-laboratory study where we compared how sensitive the three VWF activity assays are to recognise the loss of the high molecular weight multimers, represented by type 2A and type 2B VWD (S-18). We sent out a set of eight test samples, six of which had a stepwise reduction in the high S-18 Favaloro, E.J., Bonar, R., Chapman, K., Meiring, M. & Funk, D. (2012). Differential sensitivity of von Willebrand Factor (VWF) ‘activity’ assays to large and small molecular weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen-binding and mAb-based assays. Journal of Thrombosis and Haemostasis 10(6):1043-1054. (Cited by 47) 22 molecular weight multimers, to be tested by 51 different laboratories enrolled in the RCPA quality assurance programme, using a variety of assays. The collagen-binding assay was the most sensitive measure of the loss of high molecular weight VWF-multimers. The survey was recently repeated but included all the latest VWF activity assays; a total of five (S-19). In this study we sent out a set of four samples with a stepwise reduction in high molecular weight (HMW) VWF to be tested by more than 400 laboratories around the world, using all available assays. We also sent out a set of two samples representing type 1 vs 2A VWD-like plasma to be tested by a second set of 251 laboratories. Both the collagen- binding assay and the VWF-activity assay, based on spontaneous binding of VWF to a gain-of-function mutant GPIb fragment (VWF:GPIbM), showed the highest sensitivity to assess reduced HMW-multimers. The sensitivity of the ristocetin-cofactor assay and the VWF-activity assay that are based on ristocetin-induced binding of VWF to a recombinant wild type GPIb fragment (VWF:GPIbR) had intermediate sensitivity to assess reduced HMW- multimers. The VWF activity assays based on the binding of a monoclonal antibody to a VWF A1 domain epitope had the lowest sensitivity to assess reduced HMW-multimers. It is important to take these different assays into account, since they have significant clinical implications for both the diagnosis and monitoring of therapy for VWD. It follows that the least sensitive assays, and to a certain extent the intermediate sensitivities, can fail to diagnose patients with VWD, and that success of treatment may be misinterpreted. Von Willebrand factor multimers compete for clearance and for proteolysis through its cleavage protein in plasma, ADAMTS13. Exaggerated proteolytic cleavage by ADAMTS13 can cause von Willebrand disease. Conversely, S-19 Favaloro, E.J., Bonar, R., Hollestelle, M.J., Jennings, I., Mohammed, S., Meijer, P., Wood, T. & Meiring, M. (2018). Differential sensitivity of von Willebrand factor activity assays to reduced VWF molecular weight forms: A large international cross-laboratory study. Thrombosis Research 166:96-105. (Cited by 7) 23 defective proteolysis can cause thrombotic thrombocytopenic purpura (TTP). TTP used to be a rare disorder but its incidence has increased dramatically concurrent with the HIV/AIDS epidemic (Brecher et al., 2008). The underlying cause of the disease may be the result of low ADAMTS13 levels, functional abnormalities of ADAMTS13, endothelial dysfunction or the development of autoantibodies to ADAMTS13. Although ADAMTS13 antigen kits are available, the tests remain expensive, especially for small laboratories and laboratories in developing and lower-income countries. I developed a cost- effective ELISA test, using commercial antibodies to ADAMTS13 (S-20). We compared the result, given as a percentage of ADAMTS13 antigen levels of our in-house assay using two commercial antibodies. A murine anti-human ADAMTS13 antibody from R&D systems, USA was used to coat the ELISA plate and a rabbit anti-human detection antibody obtained from Santa Cruz Biotechnology, USA was used as a detection antibody. We compared our assay with a commercial ADAMTS13 antigen kit by using plasma of 40 HIV- associated TTP patients and 40 healthy individuals. The intra- and inter-assay coefficients of variation for our ADAMTS13 antigen assay were 8% and 7% respectively. The assay’s linearity ranged from 0.78% to 12.5% ADAMTS13. The limit of detection was 0.2% ADAMTS13 and the limit of quantification 0.8%. Our assay compared favorably with the commercial test kit from Sekisui Diagnostics (USA) with an R2-value of 0.9. The cost of our in-house assay was at least 90% less than the commercial kit. As a result, we now use this assay routinely to diagnose HIV-associated TTP. In a ground-breaking study published in 2012 as an educational report in European Oncology and Haematology we did not only show low levels of ADAMTS13 in HIV-associated TTP patients, but also extremely high levels of VWF and tissue factor (TF) (S-33). We had no ready explanation for the high levels of VWF. We then examined the effects that inflammatory cytokines and clotting factors such as tissue factor and thrombin and especially combinations of clotting factors and cytokines have on the synthesis and release of VWF, as well as the cleavage of ultra large von Willebrand factor S-20 Meiring, M. & Myneni, S. (2017). Evaluation of a cost-effective ADAMTS13 antigen assay. Medical Technology SA 31(1):1-4. 24 (ULVWF) by cultured human umbilical cord endothelial cells (HUVECs). We also measured endothelial micro-particle formation and the VWF content in these microparticles (S-21). We found that tissue factor alone and in combination with inflammatory cytokines substantially increased VWF release from cultured HUVECs. These studies resulted in two M.Med.Sc degrees titled: “The effect of inflammatory cytokines and coagulation factors on the von Willebrand factor synthesis and cleavage“ and “Microparticles derived from stimulation of human umbilical endothelium“. The results also mimic the findings where extremely high VWF levels are present in patients with thrombosis and inflammation (Lopez, 2006). Furthermore, less ADAMTS13 was secreted with all treatments. These results proved that the inflammatory cytokines IL-8 and TNF-alpha, clotting factors thrombin and tissue factor and combinations of cytokines and clotting factors stimulate human endothelial cells to release ULVWF and downregulate the release of ADAMTS13. We hypothesized that this resulted in an overload of ULVWF in plasma ready to bind platelets that lead to platelet aggregation and might lead to the almost always fatal HIV-associated TTP. The endothelial layer is fortunately responsible for many of the anti-thrombotic functions of blood vessels. However, endothelialisation of vascular grafts has been limited due to the cost and availability of reagents, and because it is difficult to get seeded endothelial cells to attach to the de-endothelialised blood vessels. This pertains to de-endothelialisation by balloon catheterisation or through air injury to the endothelium (Kakisis et al., 2005). In collaboration with Prof. Francis Smit from the Robert W.M. Frater Research Institute of the Department of Cardiothoracic Surgery at the University of the Free State, we prepared tissue-engineered decellularised small-vessel conduits using baboon arteries (S-22). We seeded the decellularised baboon arteries with cultured human umbilical endothelial cells (HUVECs) and compared its S-21 Meiring, M., Allers, W. & Le Roux, E. (2016). Tissue Factor: a potent stimulator of von Willebrand factor synthesis by human umbilical vein endothelial cells. International Journal of Medical Science 13(10):759-764. (Cited by 4) S-22 Meiring, M., Khemisi, M., Laker, L., Dohmen, P.M. & Smit, F.E. (2017). Tissue engineered small vessel conduits – the anti-thrombotic effect of re-endothelialization of decellularized baboon arteries: a preliminary experimental study. Medical Science Monitor Basic Research 23:344-351. (Cited by 6) 25 thrombogenicity to that of decellularised arteries. In both instances, the blood vessels were circulated for one hour with native blood in the flow chamber that I designed to study endothelial function. The decellularised arteries had no endothelial cell lining with an intact basement membrane as confirmed by scanning and transmission electron microscopy. The seeding process resulted in a complete endothelial layer on the surfaces of the arteries. Perfusion with native blood did not dislodge the seeded cells on the decellularised surface. After perfusion, no thrombi formed in intact control arteries and in the re-endothelialised vessels. In contrast, there was widespread platelet adhesion and activation in the decellularised vessels despite a relatively intact basal membrane. To our knowledge, this was the first study that achieved successful re-endothelialisation in small-diameter arteries (carotis, radial and femoral). Ultimately, the success of this approach needs to be assessed in vivo in baboons where decellularised recellularised vessels are implanted. 2.3 Part 3: Diagnosis of thrombotic and haemostatic disorders All the prior research work led to a substantial number of publications on the diagnosis of von Willebrand disease. The first was in 2005 where we used an algorithm of Prof. Federici from Italy (Federici et al., 2002) as a guideline for the diagnosis and treatment of von Willebrand disease type 1, 2A, 2B and 2M VWD (S-23). Importantly, we also studied four patients with VWD where the types and subtypes were not clear. In this study we also outlined some pitfalls to look out for when diagnosing VWD. These pitfalls are mainly due to the limited sensitivity, reproducibility and the huge interlaboratory variability of the ristocetin-cofactor assay and the ristocetin-induced platelet agglutination assay (Favaloro, 2000). We unexpectedly found that the low collagen-binding activity in type 2A and type 2B patients was more consistent with the reduction in high molecular weight VWF multimers than the ristocetin cofactor activity. As a result we strongly recommend that the collagen-binding assay S-23 Meiring M., Badenhorst, P.N. & Kelderman, M. (2005). The use of an algorithm for the laboratory diagnosis of von WIllebrand disease. Medical Technology SA 19(1):15-18. 26 should form part of the diagnosis of VWD. In 2009, as a result of my extensive experience in the field, I was invited by the editors of European Oncology and Haematology to write a review paper on the laboratory diagnosis of von Willebrand disease. In this paper I proposed a modified algorithm that must include all the diagnostic tests needed to diagnose VWD (S-24). These tests include the VWF antigen assay (VWF:Ag), the ristocetin cofactor assay (VWF:RCo), the collagen-binding assay (VWF:CB), the VWF propeptide assay (VWF:pp), the factor VIII-binding assay (VWF:FVIIIB), the ristocetin-induced platelet agglutination assay (RIPA) and RIPA mixing studies. We included an algorithm to diagnose type 1 VWD with increased clearance and also suggested how to distinguish between platelet-type VWD and type 2B VWD. We also included an approach to diagnose type 2N VWD. We ultimately proposed that it is vitally important to follow a systematic approach to diagnose VWD. As I mentioned in S-24, it is very important to keep in mind that each laboratory test only forms one part of the diagnostic puzzle. This makes it necessary to put all the puzzle pieces together to build the bigger diagnostic picture. In 2011, Seminars in Thrombosis and Haemostasis published an edition to include all the laboratories world-wide that diagnose VWD. I was asked to write a review paper on the laboratory diagnosis and management of VWD in South Africa. I listed the 17 Haemophilia Treatment Centres in South Africa where patients with VWD are cared for (S-25). I indicated that we do not know the prevalence of VWD in South Arica. I also included the genotypic data that we determined in five type 2 VWD patients in our laboratory. Two of these presented with type 2B VWD, two with type 2M VWD and one with type 2A VWD. We sequenced exon 28 and identified single nucleotide polymorphisms (SNPs). The 4641T/C SNP was found in all five patients. A 4141A/G SNP was found in three patients while a silent SNP 2923G/A was found in one patient and a new silent SNP 4923G/A in another patient. We also found that forty- S-24 Meiring, M., Kelderman, M. & Badenhorst, P.N. (2009). Laboratory diagnosis of von Willebrand Disease. European Oncology and Haematology 3(1):33-36. (Cited by 1) S-25 Meiring, M., Coetzee, M., Kelderman, M., & Badenhorst, P. (2011). Laboratory diagnosis and management of von Willebrand disease in South Africa. Seminars in Thrombosis and Haemostasis 37(5):576-580. (Cited by 9) 27 five percent of our type 1 VWD patients present with VWF:pp/VWF:Ag ratio of 1.9 ± 0.3. This indicates an increased VWF clearance phenotype. Our reference range in normal subjects for the VWF:pp/VWF:Ag ratio is 1.3 ± 0.2. We rejected approximately 15% of the samples we received mainly due to poor storage conditions. In order to scientifically show, and to stress the importance of storage conditions on the quality and accuracy of the diagnostic tests, we exposed normal plasma samples to diferent strorage conditions and measured the effect on the VWF:Ag, VWF:RCo, VWF:CB ratios and multimeric analysis. Storage at -20°C broke down the large VWF multimers. The functional assays confirmed this. Storage at -70°C had no measurable effect on the multimer pattern, even when samples were thawed and frozen up to five times. In this study we also point out that the VWF:Ag concentration in the FVIII/VWF concentrate that is used to treat patients in South Africa is approximately double that of the FVIII concentrate. This fact must be consided when this product is used to treat patients for VWD and is thus important for clinicians to be aware of. In 2017, I published a paper detailing the challenges we face to diagnose VWD in the laboratory and to manage these patients in our country (S-26). We reported on the distribution of VWD subtypes and determined the percentage misdiagnoses that occurred if the only two tests, the VWF:Ag and VWF:RCo assays were used. It must be noted that these two assays are mostly used in the other laboratories in South Africa. Retrospective analysis of data from 250 VWD patients indicated that 6% of type 1 VWD patients would have been misdiagnosed as type 2 VWD, 13% of type 2A would have been misdiagnosed as type 1 and 77% of them as type 2B, 8% of type 2B would have been misdiagnosed as type 1 and 55% of them as type 2A; 28% of type 2M patients would have been misdiagnosed as type 1 and 28% of them as type 2A or type 2B and 1% of type 3 VWD patients would have been misdiagnosed as type 1VWD. When the multimeric analysis was included together with the VWF:Ag and VWF:RCo tests, 20% of patients would still S-26 Meiring, M., Haupt, L., Conradie, C. & Joubert, J. (2017). Challenges in the laboratory diagnosis and management of von Willebrand disease in South Africa. Annals of Blood 2:19-25. (Cited by 2) 28 have been misdiagnosed. My research and publications on VWD led to continued collaboration with Prof Emmanuel Favaloro, head of the RCPA quality assurance programme in Australia. Our first collaboration on the diagnosis of VWD was a study in 2006 evaluating the diagnosis of type 2B VWD (S-27). Six different type 2B plasma samples were sent out for testing by 52 laboratories. Those laboratories that used only the VWF:Ag and VWF:RCo assays misdiagnosed 26% of the samples either as “normal” or as “type 1 VWD”. When the VWF:CB assay was added to the test panel, 11% misdiagnosed the samples. In addition, VWF sub-assays influenced the diagnosis of type 2 VWD. Of concern was that the use of automated platelet agglutination to assay VWF:RCo resulted in a more consistently functional discordance to identify VWF function as compared to the classic platelet aggregometry. Our in-house VWF:CB assays performed better than the commercial kits. The automated LIA-based VWF-activity assays also performed better than the ELISA-based assays. The majority of laboratories were proficient to test for VWD, but unfortunately incorrectly interpreted the results. We concluded that the correct diagnosis was more likely when more diagnostic tests were used and when the VWF:CB assay was included in the diagnosis. We finally provided a series of recommendations in the form of an algorithm to be used to properly identify type 2B VWD in the laboratory. I am happy to report that the majority of laboratories now follow our recommendations by referring VWD samples to expert VWD testing centres for investigation. Despite these helpful approaches, the laboratory diagnosis of VWD remains problematic for many laboratories world-wide. Together with Prof Emmanuel Favaloro’s quality assurance programme, in 2014 we evaluated the laboratory errors in the diagnosis of VWD (S-28). In this evaluation we used 29 plasma samples of both quantitative VWD deficiencies (type 1 and 3 VWD) and S-27 Favaloro, E.J., Bonar, R., Meiring, M., Street, A. & Marsden, K. (2007). 2B or not 2B? Disparate discrimination of functional VWF discordance using different assay panels or methodologies may lead to success or failure in the early identification of type 2B VWD. Thrombosis and Haemostasis 98:346-358. (Cited by 41) S-28 Favaloro, E.J., Bonar, R., Meiring, S.M., Duncan, E., Mohammed, S., Sioufi, J. & Marsden, K. (2014). Evaluating errors in the laboratory identification of von Willebrand disease in the real world. Thrombosis Research 134:393-403. (Cited by 49) 29 qualitative defects (type 2 VWD) that were tested by 55 participating laboratories. We found considerable variation between laboratories and between the different methods that led to errors in the identification of VWD. Samples with a moderate quantitative VWF deficiency were misdiagnosed as qualitative defects in 30 of 334 occasions (9% error rate). Qualitative VWF defects were also misdiagnosed as quantitative deficiencies at a 9% error rate. Of concern was that most laboratories misdiagnosed their own data. In addition, in most instances the misdiagnoses were due to limited or insufficient test panels. As a result we again stressed that laboratories should use at least the VWF:Ag and two other activity (functional) tests, e.g. the VWF:RCo and the VWF:CB to diagnose VWD. Adding multimeric analysis will possibly help to distinguish between qualitative and quantitative defects. It is important to note that only a small percentage (3.5%) of laboratories used this. In another collaborative study with Prof. Emmanuel Favaloro we assessed how many times qualitative type 2M VWD that is not associated with the loss of high molecular weight multimers, is misdiagnosed (S-29). We sent four type 2M VWD samples; two type 2B VWD and 2 type 2A-like samples to 60 laboratories world-wide. The results were alarming. Only 29% of laboratories identified the type 2M samples correctly; the error rate was 71%. These misdiagnoses were the result of insufficient test panels (42%), wrong interpretation of results (10%) and other analytical errors (13%). The findings clearly indicated that type 2M VWD was diagnosed more often incorrectly than correct. This clearly indicates that type incidence of 2M VWD is under- reported in literature. In an attempt to improve on the limited value or insufficient test panels that are used, we assessed if using a supplementary PFA-100 test together with VWD testing will minimize the misdiagnosis of VWD. FPA-100 measures S-29 Favaloro, E.J., Bonar, R.A., Mohammed, S., Arbelaez, A., Niemann, J., Freney, R., Meiring, M., Sioufi, J. & Marsden, K. (2016). Type 2M von Willebrand disease – more often misidentified than correctly identified. Haemophilia 22:e145-e55. 30 platelet adhesion to a collagen surface (S-30). We analysed retrospective data obtained from patients with type1, type 2A or type 2M VWD and also from patients with haemophilia and carriers of haemophilia. Laboratory analyses included factor VIII, VWF:Ag and VWF:RCo assays and the PFA- 100 closure times pre- and post-desmopressin treatment or pre- and post- FVIII-concentrate treatment. The normalisation of the PFA-100 closure times after desmopressin (DDAVP) treatment correlated well with the normalisation of the VWF:CB post-treatment. Based on this we strongly suggest that the supplementary use of the PFA-100 closure time and VWF:CB together with the VWF:Ag and VWF:RCo will reduce misdiagnosis of VWD and also improve the treatment management of patients with VWD. I also collaborated in studies with researchers at other universities e.g. the North-West University, University of Cape Town and the University of the Witwatersrand in South Africa by measuring VWF levels in several studies. In collaboration with Prof Leoné Malan from the subject group Physiology, Faculty of Health Sciences, Potchefstroom Campus, North-West University, we investigated possible associations between salivary testosterone levels with fibrinogen, di-dimer and VWF levels in plasma in 102 African and 165 Caucasian male teachers in the Northwest province of South Africa (S-31). Cardiovascular competency was assessed by blood pressure monitoring. The Stroop Colour-Word Conflict test was used to evoke acute mental stress responses (Stroop, 1935). The African sub-group showed elevated blood pressure with a slight inflammation response, hyperglycaemia and low testosterone levels. This might lead to a higher cardiovascular risk. The Caucasian group displayed only higher fibrinogen levels. Multiple regression analysis found an association of higher fibrinogen and VWF levels with a low testosterone environment. S-30 Favaloro, E.J., Thom, J., Patterson, D., Just, S., Baccala, M., Dixon, T., Meiring, S.M., Koutts, J., Rowell, J., Baker, R. (2009). Potential supplementary utility of combined PFA-100 a functional von Willebrand factor testing for the laboratory assessment of desmopressin and factor concentrate therapy in von Willebrand disease. Blood Coagulation and Fibrinolysis 20(6):475-483. (Cited by 22) S-31 Malan, N.T., Von Kanel, R., Schutte, A.E., Huisman, H.W., Smith, W., Schutte, R., Smith, W., Mels, C.M., Kruger, R., Meiring, M., Van Rooyen, J.M. & Malan, L. (2013). Testosterone and acute stress are associated with fibrinogen and von Willebrand factor in African men: The SABPA study. International Journal of Cardiology 168 (5):4638-4642. (Cited by 11) 31 In another study we compared in the same group as in S-31 the coagulation response to acute mental stress. We also studied the relationship of these coagualtion responses to physical well-being (S-32). We measured circulating levels of VWF:Ag, fibrinogen and D-dimer before and after the Stroop Colour- Word Conflict test. All participants completed health questionnaires to establish symptoms of depression. Africans showed lower VWF and fibrinogen reactivity in response to laboratory mental stress than Caucasians. The symptoms of depression in Africans were associated with a higher increase in VWF levels and decrease in fibrinogen plasma levels than in Caucasians. Thus we could postulate that depressive mood might be associated with an increased risk of hypercoagulability, as reported by Frasure-Smith & Lespérance (2010). The composition of the von Willebrand factor multimers in the circulation is a result of the competition between the liver clearance of VWF and proteolysis through its cleavage protein in plasma, ADAMTS13. Defective proteolysis of VWF by ADAMTS13 can result in thrombotic thrombocytopenic purpura (TTP), a fatal thrombotic disorder (Moake et al., 1982). TTP used to be a rare disorder but its incidence has increased dramatically concurrent with the HIV/AIDS epidemic (Gunther et al., 2006). The underlying cause is likely to be the result of low ADAMTS13 levels, fundamental abnormalities of ADAMTS- 13, endothelial dysfunction, or autoantibody production against ADAMTS13. In 2012, we established low levels of ADAMTS13 in HIV-associated TTP patients in conjunction with extremely high plasma VWF and tissue factor (TF) levels (S-33). Interestingly, we found that only 50% of our patients with HIV- associated TTP presented with autoantibodies to ADAMTS-13. Based on our results, we propose that the trigger for HIV-associated TTP is HIV infection that increases cytokine release from endothelium, resulting in increased VWF S-32 Von Kanel, R., Hamer, M., Malan, N.T., Scheepers, K., Meiring, M. & Malan, L. (2013). Procoagulant reactivity to laboratory acute mental stress in Africans and Caucasionas, and its relation to depressive symptoms: The SABPA Study. Thrombosis and Haemostasis 110(5):977-986. (Cited by 7) S-33 Meiring, M., Webb, M., Goedhals, D. & Louw, V. (2012). HIV associated Thrombotic Thrombocytopenic Purpura – What we know so far. European Oncology and Haematology, 8(2):89-91. (Cited by 10) 32 synthesis. VWF that is released by the endothelium consequently leads to microvascular thrombus formation. In addition, micronutrient deficiencies in HIV-infected patients also cause decreased ADAMTS13 synthesis that leads to microvascular thrombus formation (Kaiser et al., 2006). We also found that HIV infection increases the production of autoantibodies to ADAMTS13. It is plausible that some of these autoantibodies might inhibit the function of ADAMTS13 which can also lead to microvascular thrombus formation. Ultimately, haemolytic anemia and thrombocytopenia result in HIV-associated TTP, independent of the cause of microvascular thrombus formation. To my knowledge, this is the first proposed mechanism that explains the onset of HIV-associated TTP which has changed the paradigm about TTP. The findings in this study provided me with a highly focused, ongoing research project that will lead to further publications and provide research projects for postgraduate students. The treatment products for TTP in South Africa include fresh frozen plasma (FFP), cryosupernatant (CSP) and and Bioplasma FDP, a solvent/detergent- treated plasma (SDP). In another study on TTP (S-34), we compared the ADAMTS13 and VWF levels and activities, as well as the plasminogen levels in plasma products that are currently used to treat TTP. With this study, we explored possible differences in VWF and ADAMTS13 levels and activities between the products that may offer treatment advantages. All products had normal to high ADAMTS13 levels and activities, making them suitable to treat TTP. Bioplasma FDP had normal VWF levels and CSP had typically reduced VWF levels. FFP had a large inter-batch variation where some batches contained decreased VWF levels and some normal levels. Statistically higher ADAMTS13 activities were found on FFP samples from the O-blood group than the non-O blood groups. This finding might indicate that FFP of blood group O may be preferential to treat TTP patients. We concluded that one should consider the viral safety, the cost, the availability of the product and the impact of the within-product variations when choosing a product for S-34 Van Marle, A.C., Joubert, J. & Meiring, S.M. (2019). Comparison of ADAMTS13 and von Willebrand Factor levels and activities, and plasminogen levels, in plasma products currently available for the treatment of thrombotic thrombocytopenic purpura in South Africa. Transfusion and Apheresis Science 58:72-78. 33 treatment. The inclusion of ADAMTS13 testing in our routine diagnostic setup opened the way for collaboration on projects that require these results. Together with the group of Prof. Marc Combrinck from the Department of Neurology, Faculty of Health Sciences at the University of Cape Town, we investigated the role of VWF and ADAMTS13 in HIV-related strokes in young patients of 35 years of age (mean age) in a case-control study with three participant groups: HIV- positive antiretroviral therapy-naïve young stroke patients, HIV-negative young stroke patients and HIV-positive antiretroviral therapy-naïve non-stroke control patients (S-35). In the case of the stroke patients, plasma VWF and ADAMTS13 levels were measured five and seven days post-stroke. The HIV- positive non-antiretroviral stroke patients had significantly higher VWF levels and lower ADAMTS13 levels than the HIV-positive antiretroviral treated non- stroke controls. In addition, VWF levels in all HIV-positive participants correlated negatively with CD4 counts, i.e. the higher the VWF levels, the lower the CD4 counts. Thus, stroke in HIV infection was associated with a pro-thrombotic state with increased VWF and decreased ADAMTS13 levels. The pro-thrombotic state of HIV is not yet fully elucidated and require ongoing research. We have several studies planned, as well as studies in collaboration with the University of the Witwatersrand. A case study describes that a patient, treated with valproic acid (a commonly used anticonvulsant), developed serious bleeding during surgery (S-36). Platelet aggregometry in response to arachidonic acid, adenosine diphosphate, epinephrine, collagen and ristocetin was tested. The platelet aggregation in response to arachidonic acid was abolished and reduced in response to adenosine diphosphate, adrenaline, collagen and ristocetin. Valproic acid treatment was halted for two weeks, with close monitoring for epilepsy. All platelet functions returned to normal and no bleeding developed S-35 Allie, S., Stanley, A., Bryer, A., Meiring, M. & Combrinck, M.I. (2015). High levels of von Willebrand factor and low levels of its cleaving protease, ADAMTS13, are associated with stroke in young HIV-infected patients. International Journal of Stroke 10(8):1294-1296. (Cited by 13) S-36 Jafta, A.D., Meiring, M. & Conradie, C. (2010). Valproic acid associated platelet dysfunction: Case Report. African Journal of Haematology and Oncology 1(2):54-56. 34 during follow-up surgery. It is very likely that free fractions of valproic acid in plasma may reduce platelet aggregation, in spite of the fact that most patients on valproic acid do not experience platelet dysfunction. We are currently testing this hypothesis in collaboration with local neurologists from the Department of Neurology and pharmacologists from the Department of Pharmacology at the UFS. 35 3. GENERAL SUMMARY My main contribution to the knowledge of haemostatic disorders was to develop and test new and existing antithrombotic drugs, to develop cost- effective tests to diagnose these disorders, and to unravel the mechanisms underlying the diagnoses of these disorders. We investigated several monoclonal antibodies for their effect in platelet- dependent thrombogenesis. It is important to note that some caused almost immediate thrombocytopenia when injected into baboons. Others had no effect on the circulating platelet count. Antibody 6B4 and the F(ab)2 fraction thereof caused an immediate drop in platelet count (profound thrombocytopenia) (S3) when injected into baboons. Thehe F(ab)-fragment did not decrease the platelet count. The fact that monoclonal antibodies, as large proteins and foreign to the body, trigger the immune system should also be kept in mind. As a result second treatments will not be successful because they will be recognised by the immune system as foreign and immediately removed from the circulation. It is therefore advisable to use the smaller, non- immunogenic F(ab) fragments. In my group we reported on the first echistatin-like RGD-sequence containing monoclonal antibody MA-16N7C2 that significantly inhibited platelet glycoprotein IIb/IIIa function. In addition, it appeared that the antibody was internalised to prolong the inhibitory effect (S-2). We also described the first anti-human GpIb/IX monoclonal antibody MA-6B4 (S-3) that successfully prevented platelet adhesion to the subendothelium and subsequent thrombus formation. This finding confirmed the important role of GpIb/IX in platelet adhesion in vivo. An important finding was that the bleeding time was not markedly affected, while no blood loss, spontaneous bleeding or thrombocytopenia was observed. We also conclusively showed that inhibition of both GPIb/IX with antibody 6B4 and GPIIb/IIIa with antibody 16N7C2 was more effective and safer than using them separately (S-5). It is also important to note that the superior inhibition of platelet-dependent thrombosis was achieved at much lower individual doses when the antibodies were used in 36 combination. We also described, for the first time, the antithrombotic effect of the monoclonal antibody 82D6A3 (S-6) that effectively inhibits the VWF binding to fibrillar collagens type I and III in baboons without significantly lengthening the bleeding time, even at high doses. This antibody is therefore a safe target to prevent arterial thrombosis. We also showed that the factor VIII inhibitory monoclonal antibody LE2E9 (S-8) prevented platelet-dependent thrombosis without prolonging the bleeding time and so decreases the tendency of in vivo bleeding. This may present a novel approach to a long- acting antithrombotic agent with an optimal safety/efficacy profile. We also found that inhibiting VWF with antibody 82D6A3 (S-7) prevented stenosis in a clinically relevant model. We also developed a model to use to prevent restenosis following stent deployment. In my endeavours to develop cost-effective anti-thrombotic drugs, we identified a collagen-binding protein from the hookworm Necator americanus by using a cDNA-expressing phage display library (S-9). This was the first study using phage display technology to select proteins based on their binding properties in which I was involved. Through phage display technology we developed a competitive thrombin inhibition peptide (TIP), that inhibited thrombin-induced platelet activation in a dose-dependent manner (S-11). It also reduced platelet adhesion onto a human microvascular endothelial matrix in a parallel plate flow chamber under both arterial and venous shear conditions. Using the same cyclic heptapeptide library, we also selected a peptide that binds to factor VII and competes with tissue factor for binding to factor VII (S-12). This peptide prolonged the prothrombin time and reduced platelet adhesion to both human endothelial cells and tissue factor matrices in a flow chamber under arterial flow conditions. We also selected single chain antibody fragments that inhibit tissue factor (S-13), the initiator protein of blood coagulation. We expressed this single chain variable fragment (scFv) by increasing the levels in the cytoplasm of E. coli than what we had previously achieved in the periplasm (S-14). Both in vitro refolding and cold shock strategies were able to produce functional tissue factor inhibitory scFv, although with varying degrees of success. This approach successfully 37 improved the production of low-yielding single chain variable fragments. This will ultimately lead to the production of large amounts of product. We used an antibody fragment library to select antibodies that bind to the VWF-propeptide in order to develop cost-effective assays for diagnosis of haemostatic disorders (S-15). Our assay compared favourably with that of the available commercial kit. Of particular importance is that our antibodies had a higher binding affinity for VWF propeptide than the commercial antibodies. We also filed an South Africa patent 2011/05426. In order to widen our scope of diagnoses of VWD, I developed a cost-effective assay that monitors VWF activity by measuring its binding to type III collagen, the collagen-binding assay (S-16). This assay is ten times cheaper than other commercially available assays. It is also very sensitive to measure the collagen-binding activity in VWD patients. I also developed the VWF multimer analysis as an additional instrument to diagnose VWD (S-17); this is 10 times cheaper than the commercially available assay. We developed a cost-effective ELISA test, using commercial antibodies to ADAMTS13 (S-20). Our in-house assay was 90% cheaper than the commercial kit and we therefore recommend that this assay be used to diagnose HIV-associated TTP. Therefore, by developing these essays, we were able to significantly decrease the high cost of diagnoses as a result of purchasing these tests from commercial sources. In summary, I developed cost-effective assays to increase the scope of tests in order to improve the ability of my reference laboratory to diagnose VWD. In order to provide an extensive and accurate diagnostic service to the South African public, our battery of tests for the diagnosis of von Willebrand disease includes the VWF antigen test, the collagen-binding assay, the ristocetin cofactor assay, the multimeric analysis, the factor VIII-binding assay and the VWF propeptide assay. By developing cost-effective diagnostic tests, we can now diagnose VWD at a fraction of the cost that it would have been if we had used imported commercial assays. As I mentioned before, it is crucial to use a systematic approach to diagnose VWD. Each test only forms one piece of the diagnostic puzzle. It is therefore necessary to put all the puzzle pieces together for the whole diagnostic picture to emerge. 38 Building this puzzle, we found that 45% of our type 1 VWD patients had an increased VWF clearance phenotype (S-25). We also determined the inaccuracy of diagnosis if only the VWF:Ag and VWF:RCo assays are used (S-26). We did this because these tests are often the only ones used by the majority of laboratories in South Africa. Retrospective analyses of data of 250 VWD patients in our database indicated that 6% of type 1 VWD patients would have been misdiagnosed as type 2 VWD, 13% of type 2A would have been misdiagnosed as type 1 and 77% as type 2B; 8% of type 2B would have been misdiagnosed as type 1 and 55% as type 2A;, 28% of type 2M patients would been misdiagnosed as type 1 and 28% as type 2A or type 2B, and 1% of type 3 VWD patients would have been misdiagnosed as type 1 VWD. Even if multimeric analyses were added to the VWF:Ag and VWF:RCo tests, 20% of patients would still have been misdiagnosed. Importantly, storage at -20°C broke down the large VWF multimers; a finding confirmed by the functional assay. Storage at -70°C caused no breakdown of multimers (S-25). This will further exacerbate misdiagnosis. We also found a new silent SNP 4923G/A in one of our type 2 VWD patients (S-25). I emphasize the errors in diagnosis because it directly affects the success of treatment of patients with VWD. Errors in diagnosis lead to wrong treatment approaches that is mostly ineffective and expensive and may increase the risk to patients. In collaboration with Prof. Emmanuel Favaloro from the Royal College of Pathologists of Australasia (RCPA), we conclusively showed that the collagen-binding assay and the VWF activity assay based on spontaneous binding of VWF to a gain-of-function mutant GPIb fragment (VWF:GPIbM) had the highest sensitivity to assess the loss of the high molecular weight multimers. This confirms my finding on the sensitivity of the collagen-binding assay to assess VWF-multimer decrease (S-18). The ristocetin-cofactor assay and the VWF activity assay based on ristocetin-induced binding of VWF to a recombinant wild type GPIb fragment (VWF:GPIbR) displayed intermediate sensitivity to assess decreased HMW multimers (S-19). These results ultimately hold significant clinical implications to diagnose and to monitor the 39 outcome of therapy in patients with VWD. In a further study using a larger sample size (S-28), we discovered that most international laboratories were proficient in testing to diagnose VWD, but that their interpretation of the assays was incorrect when they used only the VWF:Ag and VWF:RCo assays. It became clear that inclusion of the VWF:CB assay improved the accuracy of diagnosis. Based on these results, we made a series of recommendations in the form of an algorithm on the tests to be used to enable a laboratory to correctly identify VWD. In most cases the misdiagnoses by laboratories were due to limited or insufficient test panels (S-27). We again stressed that laboratories should use an extensive test panel to diagnose VWD. If it is not possible, the VWF:Ag and two other functional tests, e.g. the VWF:RCo and the VWF:CB, may be used as the minimum tests to ensure accurate diagnosis. Similarly, the use of the multimeric VWF analysis will be important to distinguish between qualitative and quantitative defects, since the treatment depends on the correct classification of VWD. Unfortunately only a small percentage of laboratories perform this test. In the survey we also found that type 2M VWD is misdiagnosed more regularly than it is diagnosed correctly (S-29). This may explain the relative under-reported incidence of type 2M VWD in the literature. Because of this, we strongly recommended that supplementary combination of the PFA-100 assay be done together with measuring VWF:CB. This, together with the VWF:Ag and VWF:RCo tests, may possibly improve diagnosis and treatment efficacy in patients with VWD. Based on our findings on patients with TTP, we were the first to propose a mechanism to explain the initial onset of the disorder (S-33). We proposed that the trigger for HIV-associated TTP is HIV infection that increases cytokine release from endothelium, resulting in increased VWF synthesis and release by the endothelium leading to microvascular thrombus formation. This changed the paradigm of thought pertaining to TTP. We furthermore proved that human umbilical vein endothelial cells decrease VWF in reaction to tissue factor and cytokines interleukin 6 and tumor necrosis factor. This explained the extremely high VWF level in HIV and HIV-associated TTP patients. The proposed mechanism of onset of TTP was based on our findings that 40 inflammatory cytokines such as IL-8 and TNF-alpha, coagulation factors such as thrombin and tissue factor, as well as combinations thereof, stimulate the release of ULVWF and inhibit the release of ADAMTS13 in human umbilical endothelial cells (HUVECs, S-21). We also indicated that all products that are currently used to treat TTP have normal to high ADAMTS13 levels and activity. Of particular importance was the finding that the VWF varied considerably in the commercially available products that are used to treat these patients. Bioplasma contains normal VWF levels. Cryosupernatant has typically reduced VWF levels. Fresh frozen plasma had a large inter-batch variation where some batches contain decreased VWF levels and some normal levels. This may adversely impact on the success of treatment of HIV- positive patients presenting with TTP. Furthermore, statistical differences were found across products and ABO blood groups of donors. Fresh frozen plasma from blood group O donors contains more ADAMTS13 than those from non-O donors and would be more beneficial for the treatment of patients with HIV-associated TTP. In addition to the choice of product based on VWF and ADAMTS13 availability, we also emphasised that factors such as viral safety, costs and product availability must be taken into account. In collaborative studies, we were the first to show that stroke in HIV-infected patients was associated with a pro-thrombotic state, characterised by elevated VWF and low ADAMTS13 levels (S-35). In addition, depression was associated with higher plasma concentration of VWF increase in Africans compared to Caucasians. It is possible that ethnic differences in the vasculatory energic stress response may partially explain this difference. Multiple regression analysis suggested an association of higher circulating levels of fibrinogen and VWF with low testosterone levels (S-32). We also described a case where valproic acid, a commonly used anticonvulsant, was associated with excessive bleeding during surgery (S-36). In this patient in vitro platelet aggregometry was decreased in response to aggregation induced by with arachidonic acid, adenosine diphosphate, epinephrine, collagen and ristocetin. 41 In a new avenue of research that I recently started in collaboration with Prof. Francis Smit from the Robert W.M Frater Institute for Cardiovascular Research, we developed re-endothelialised tissue-engineered small-vessel conduits with no thrombogenicity (S-22). We studied re-endothelialisation of decellularised baboon arteries in the in vitro flow chamber described in S-22. We conclusively found that HUVECs irreversibly bind to the decellularised scaffold in an in vitro flow chamber. It is important to stress that, to our knowledge, this was the first study to show success with re-endothelialisation. Up to this point, attempted re-endothelialisation of arteries that were damaged by balloon catheterisation, or endothelial damage by air or magnetism, had been unsuccesfull. Although this research is in an infantile stage, we have high hopes of positive results following implantation of re-endothelialised decellularised arteries into baboons. In conclusion, my research contributed significantly to the knowledge base of thrombosis inhibition and accurate differential diagnosis of VWD. The success of particularly the research on VWD resulted in my laboratory functioning as the reference laboratory for VWD in RSA. In addition, my laboratory is also part of the world-wide quality assurance programme for the diagnosis of VWD. In this regard, I work closely with Prof. Emmanuel Favaloro from the Royal College of Pathologists’ quality assurance programme. 42 4. REFERENCES Bradshaw, D., Groenewald, P., Laubscher, R., Nannan, N., Nojilana, B., Normal, R., Pieterse, D., Schneider M., Bourne D.E., Timaeus I.M., Dorrington, R. & Johnson, L. (2003). Initial burden of disease estimates for South Africa 2000. South African Medical Journal 93(9):682-688. Brecher, M.E., Hay, S.N. & Park, Y.A. (2008). Is it HIV TTP or HIV-associated microaniopathy? Journal of Clinical Apheresis 23(6):186-190. (https://doi.org/10.1002/jca.20176) Cadroy Y., Horbett T.A. & Hanson S.R. (1989). Discrimination between platelet-mediated and coagulation-mediated mechanisms in a model of complex thrombus formation in vivo. Journal of Laboratory and Clinical Medicine 113:672-675 Eyal, A. & Veller, M. (2009). HIV and venous thrombotic event: vasculare surgery: review. South African Journal of Surgery 47(2):54. (https://doi.org/10.7196/sajs.427) Favaloro, E.J. (2000). Collagen binding assay of von Willebrand Factor detection of von Willebrand disease and discrimination of VWD subtypes dependent on collagen source. Thrombosis and Haemostasis 2000:83:127-135. Federici, A.B., Castaman, G. & Mannucci, P.M. (2002). Guidelines for the diagnosis and management of von Willebrand disease in Italy. Haemophilia 8:607-621. Frasure-Smith, N. & Lespérance, F. (2010). Depression and cardiac risk: present status and future directions. Heart 96:173-176. (http://dx.doi.org/10.1136/hrt.2009.186957) Gunther, K., Garizio, D. & Dhlamini, B. (2006). The pathogenesis of HIV- related thrombotic thrombocytopenic prupura – is it different? ISBT Science series 1:246-250. Hanson, S.R., Kotzé, H.F., Savage, B. & Harker, L.A. (1985). Platelet interactions with Dacron vascular grafts: A model of acute vascular thrombosis in baboons. Arteriosclerosis 5, 595-603. Harsfalvi, J., Stassen, J.M., Hoylaerts, M.F., Van Houtte, E., Sawyer, R.T., Vermylen, J. & Deckmyn, H. (1995). Calin from Hirudo medicinalis, an inhibitor of von Willebrand factor binding to collagen under static and flow conditions. Blood 85(3): 705-711. (https://doi.org/10.1182/blood.V85.3.705.bloodjournal853705) 43 Heidenreich, P.A., Trogdon, J.G., Khavjou, O.A. et al. (2011). Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation 123:933-44. (https://doi.org/10.1161/CIR.0b013e31820a55f5) Hirsh, J. & Weitz, J.I. (1999). New antithrombotic agents. Lancet 353(9162):1431-1436. Jacobson, B. (2013). Thrombosis – prevention is better than cure. South African Medical Journal 103(4): 231. (https://doi.org/10.7196/SAMJ.6808) Kaiser, J.D., Campa, A.M., Ondercin, J.P., Lioung, G.S., Pless, R.F. & Baum, M.K. (2006). Micronutrient supplementation increased CD4 count in HIV- infected individuals on highly active antiretroviral therapy: a prospecteve, double-blinded placebo-controlled trial. Journal on Acquired Immune Deficiency Syndrome. 42(5): 523-528. Kakisis, J.D., Liapis, C.D., Breuer, C. & Sumpio, B.E. (2005). Artificial blood vessel: The holy grail of peripheral vascular surgery. Journal of Vascular Surgery 41:349-359. Krizek, D.R. & Rick, M.E. (2000). A rapid method to visualize von Willebrand factor multimers by using agarose gel electrophoresis, immunolocalization and luminographic detection. Thrombosis Research 97:457-462. Lopez, J.A. (2006). Sticky business: von Willebrand factor in inflammation. Blood 108:3627. (https://doi.org/10.1182/blood-2006-09-046474) Mahlangu J.N. (2009). Haemophilia care in South Africa: 2004-2007 look back. Haemophilia 15(1):135-141. Moake, J.L., Rudy, C.K., Troll, J.H., Weinstein, M.J., Colannino, N.M., Azocar, J., Seder, R.H., Hong, S.L., Deykin. M.D. (1982). Unusually large plasma VIII: von Willebrand factor multimers in chronic relapsing thrombotic thrombocytopenic purpura. New England Journal of Medicine 307(23):1432-1435. (https://doi.org/10.1056/NEJM198212023072306) Nicolaides, N.C., Sass, P.M. & Grasso, L. (2006). Monoclonal antibodies: a morphing landscape for therapeutics. Drug Development Research 67:781-789. Sadler, J.E. (2005). Von Willebrand factor: two sides of a coin. Journal of Thrombosis and Haemostasis 3:1702-1709. Spyropoulos, A.C. (2010). Risk assessment of venous thromboembolism in hospitalized medical patients. Current Opinions in Pulmonary Medicine 16 (5):419-425. (https://doi.org/10.1097/MCP.0b013e32833b4669) 44 Stroop, J.R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology 18(6): 643-662. (https://doi.org/10.1037/h0054651) Tsai, J., Grant, A.M., Beckman, M.G., Grosse, S.D., Yusuf, H.R. & Richardson, L.C. (2015). Determinants of venous thromboembolism among hospitalizations of US adults: a multilevel analysis. PLoS ONE. 10(4):e0123842. (https://doi.org/10.1371/journal.pone.0123842) Tsongo Vululi, S., Bugeza, S., Zeridah, M., Ddungu, H., Openy, A.B., Frank, M. & Parkes-Ratanshi, R. (2018). Prevalence of lower limb deep venous thrombosis among adult HIV positive patients attending an outpatient clinic at Mulago Hospital. AIDS Research and Therapy 15:3. (https://doi.org/10.1186/s12981-018-0191-1) Wolberg, A.S., Aleman, M.M., Leiderman, K. & Machlus, K.R. (2012). Procoagulant Activity in Hemostasis and Thrombosis: Virchow's Triad Revisited. Anesthesia-Analgesia 114(2):275–85. (https://doi.org/10.1213/ANE.0b103e31823a088c) 45 5. LIST OF PUBLICATIONS RELATED TO THIS COMPILATION (Supplements) S-1 Meiring, S.M., Lötter, M.G., Badenhorst, P.N., Bucha, E., Nowak, G.& Kotzé, H.F. (1999). Sites of elimination and pharmacokinetics of recombinant 131I-lepirudin in baboons. Journal of Pharmaceutical Sciences 88(5):523-529. (Cited by 2) S-2 Kotzé, H.F., Badenhorst, P.N., Lamprecht, S., Meiring, M., Van Wyk, V., Nuyts, K., Stassen, J.M., Vermylen, J. & Deckmyn, H. (1995). Prolonged inhibition of acute arterial thrombosis by high dosing of a monoclonal anti-platelet glycoprotein IIb/IIIa antibody in a baboon model. Thrombosis and Haemostasis 74:751-757. S-3 Cauwenberghs, N., Meiring, S.M., Vautarin, S., Van Wyk, V., Lamprecht, S., Roodt, J.P., Novak, L., Harsfalvi, J., Deckmyn, H. & Kotzé, H.F. (2000). Antithrombotic effect of platelet glycoprotein Ib-blocking monoclonal antibody Fab fragments in nonhuman primates. Arteriosclerosis, Thrombosis and Vascular Biology 20:1347-1353. (Cited by 141) S-4 Fontayne, A., Meiring, M., Lamprecht, S., Roodt, J., Demarsin, E., Barbeaux, P. & Deckmyn, H. (2008). The humanized anti-glycoprotein 1b monoclonal antibody h6B4-Fab is a potent and safe antithrombotic in a high shear arterial thrombosis model in baboons. Thrombosis and Haemostasis 100:670-677. (Cited by 58) S-5 Wu, D., Meiring, M., Kotzé, H.F., Deckmyn, H. & Cauwenberghs, N. (2002). Inhibition of platelet glycoprotein Ib, glycoprotein IIb/IIIa, or both by monoclonal antibodies prevents arterial thrombosis in baboons. Arteriosclerosis, Thrombosis and Vascular Biology 22:323-328. (Cited by 84) S-6 Wu, D., Vanhoorelbeke, K., Cauwenberghs, N., Meiring, M., Depraetere, H., Kotzé, H.F. & Deckmyn, H. (2002). Inhibition of the von Willebrand- collagen interaction by an antihuman VWF monoclonal antibody results in abolition of in vivo arterial platelet thrombus formation in baboons. Blood 99(10):3623-3628. (Cited by 139) S-7 De Meyer, S.F., Staelens, S., Badenhorst, P.N., Pieters, H., Lamprecht, S., Roodt, J., Janssens, S., Meiring, M., Vanhoorelbeke, K., Bruwer, A., Brown, S, & Deckmyn, H. (2007). Coronary artery in-stent stenosis persists despite inhibition of the von Willebrand factor – collagen interaction in baboons. Thrombosis and Haemostasis 98: 1343-1349. (Cited by 18) S-8 Jaquemin, M., Stassen, J-M., Saint-Remy, J-M., Verhamme, P., Lavend’Homme, R., VanderElst, L., Meiring, M., Pieters, H., Lamprecht, S., Roodt, J. & Badenhorst, P. (2009). A human monoclonal antibody 46 inhibiting partially factor VIII activity reduces thrombus growth in baboons. Journal of Thrombosis and Haemostasis 7:429-437. S-9 Viaene, A., Crab, A., Meiring, M., Pritchard, D. & Deckmyn, H. (2001). Identification of a collagen-binding protein from Necator americanus by using a cDNA-expression phage library. Journal of Parasitology 87(3):619-625. S-10 Meiring, S.M., Kotzé, H.F., Pretorius, G.H.J. & Badenhorst, P.N. (1999). Die toepassing van peptiedblootlegging op fage in trombose en hemostase (The application of phage display technology in thrombosis and haemostasis). Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie (South African Journal of Science and Technology) 18 (3):76-81. S-11 Meiring, S.M., Littauer, D., Harsfalvi, J., Van Wyk, V., Badenhorst,P.N., Kotzé, H.F. (2002). In vitro effect of a thrombin inhibition peptide selected by phage display technology. Thrombosis Research 107:365- 371. (Cited by 16) S-12 Meiring, S.M., Roets, C.E. & Badenhorst, P.N. (2006). Funksionele beskrywing van ’n faktor VIIa inhiberende peptied, IP-7, geselekteer deur faagblootleggingstegnologie. (Functional characterisation of a factor VIIa inhibiting peptide, IP-7, selected by phage display technology) Die Suid Afrikaanse Tydskrif vir Wetenskap en Tegnologie (South African Journal of Science and Technology) 25(4):209-220. S-13 Meiring, S.M., Vermeulen, J. & Badenhorst, P.N. (2009). Development of an inhibitory antibody fragment to human tissue factor using phage display technology. Drug Development Research 70:199-205. (Cited by 2) S-14 Vermeulen, J., Burt, F., Van Heerden, E., Cason, E. & Meiring, M. (2018). Evaluation of in vitro refolding vs cold shock expression: production of a low yielding single chain variable fragment. Protein Expression and Purification 151:62-71. (Cited by 1) S-15 Meiring, S.M., Setlai, P., Theron, C. & Bragg, R. (2018). The use of phage display and yeast based expression system for the development of a von Willebrand factor propeptide assay: Development of a von Willebrand factor propeptide assay. BioMed Research International 2018:632091:1-7. S-16 Meiring, M., Badenhorst, P.N. & Kelderman, M. (2007). Performance and utility of a cost-effective collagen-binding assay for the laboratory diagnosis of von Willebrand disease. Clinical Chemistry and Laboratory Medicine. 45(8):1068-1072. (Cited by 19) 47 S-17 Meiring, M., Badenhorst, P. & Kelderman, M. (2005). A rapid and cost- effective method to visualize von Willebrand factor multimers in plasma. Medical Technology SA 19(2):16-18. S-18 Favaloro, E.J., Bonar, R., Chapman, K., Meiring, M. & Funk, D. (2012). Differential sensitivity of von Willebrand Factor (VWF) ‘activity’ assays to large and small molecular weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen-binding and mAb-based assays. Journal of Thrombosis and Haemostasis 10(6):1043-1054. (Cited by 47) S-19 Favaloro, E.J., Bonar, R., Hollestelle, M.J., Jennings, I., Mohammed, S., Meijer, P., Wood, T. & Meiring, M. (2018). Differential sensitivity of von Willebrand factor activity assays to reduced VWF molecular weight forms: A large international cross-laboratory study. Thrombosis Research 166:96-105. (Cited by 7) S-20 Meiring, M. & Myneni, S. (2017). Evaluation of a cost-effective ADAMTS13 antigen assay. Medical Technology SA 31(1):1-4. S-21 Meiring, M., Allers, W. & Le Roux, E. (2016). Tissue Factor: a potent stimulator of von Willebrand factor synthesis by human umbilical vein endothelial cells. International Journal of Medical Science 13(10):759- 764. (Cited by 4) S-22 Meiring, M., Khemisi, M., Laker, L., Dohmen, P.M. & Smit, F.E. (2017). Tissue engineered small vessel conduits – the anti-thrombotic effect of re-endothelialization of decellularized baboon arteries: a preliminary experimental study. Medical Science Monitor Basic Research 23:344- 351. (Cited by 6) S-23 Meiring M., Badenhorst, P.N. & Kelderman, M. (2005). The use of an algorithm for the laboratory diagnosis of von WIllebrand disease. Medical Technology SA 19(1):15-18. S-24 Meiring, M., Kelderman, M. & Badenhorst, P.N. (2009). Laboratory diagnosis of von Willebrand Disease. European Oncology and Haematology 3(1):33-36. (Cited by 1) S-25 Meiring, M., Coetzee, M., Kelderman, M., & Badenhorst, P. (2011). Laboratory diagnosis and management of von Willebrand disease in South Africa. Seminars in Thrombosis and Haemostasis 37(5):576-580. (Cited by 9) S-26 Meiring, M., Haupt, L., Conradie, C. & Joubert, J. (2017). Challenges in the laboratory diagnosis and management of von Willebrand disease in South Africa. Annals of Blood 2:19-25. (Cited by 2) 48 S-27 Favaloro, E.J., Bonar, R., Meiring, M., Street, A. & Marsden, K. (2007). 2B or not 2B? Disparate discrimination of functional VWF discordance using different assay panels or methodologies may lead to success or failure in the early identification of type 2B VWD. Thrombosis and Haemostasis 98:346-358. (Cited by 41) S-28 Favaloro, E.J., Bonar, R., Meiring, S.M., Duncan, E., Mohammed, S., Sioufi, J. & Marsden, K. (2014). Evaluating errors in the laboratory identification of von Willebrand disease in the real world. Thrombosis Research 134:393-403. (Cited by 49) S-29 Favaloro, E.J., Bonar, R.A., Mohammed, S., Arbelaez, A., Niemann, J., Freney, R., Meiring, M., Sioufi, J. & Marsden, K. (2016). Type 2M von Willebrand disease – more often misidentified than correctly identified. Haemophilia 22:e145-e55. S-30 Favaloro, E.J., Thom, J., Patterson, D., Just, S., Baccala, M., Dixon, T., Meiring, S.M., Koutts, J., Rowell, J., Baker, R. (2009). Potential supplementary utility of combined PFA-100 a functional von Willebrand factor testing for the laboratory assessment of desmopressin and factor concentrate therapy in von Willebrand disease. Blood Coagulation and Fibrinolysis 20(6):475-483. (Cited by 22) S-31 Malan, N.T., Von Kanel, R., Schutte, A.E., Huisman, H.W., Smith, W., Schutte, R., Smith, W., Mels, C.M., Kruger, R., Meiring, M., Van Rooyen, J.M. & Malan, L. (2013). Testosterone and acute stress are associated with fibrinogen and von Willebrand factor in African men: The SABPA study. International Journal of Cardiology 168 (5):4638-4642. (Cited by 11) S-32 Von Kanel, R., Hamer, M., Malan, N.T., Scheepers, K., Meiring, M. & Malan, L. (2013). Procoagulant reactivity to laboratory acute mental stress in Africans and Caucasionas, and its relation to depressive symptoms: The SABPA Study. Thrombosis and Haemostasis 110(5):977-986. (Cited by 7) S-33 Meiring, M., Webb, M., Goedhals, D. & Louw, V. (2012). HIV associated Thrombotic Thrombocytopenic Purpura – What we know so far. European Oncology and Haematology, 8(2):89-91. (Cited by 10) S-34 Van Marle, A.C., Joubert, J. & Meiring, S.M. (2019). Comparison of ADAMTS13 and von Willebrand Factor levels and activities, and plasminogen levels, in plasma products currently available for the treatment of thrombotic thrombocytopenic purpura in South Africa. Transfusion and Apheresis Science 58:72-78. 49 S-35 Allie, S., Stanley, A., Bryer, A., Meiring, M. & Combrinck, M.I. (2015). High levels of von Willebrand factor and low levels of its cleaving protease, ADAMTS13, are associated with stroke in young HIV-infected patients. International Journal of Stroke 10(8):1294-1296. (Cited by 13) S-36 Jafta, A.D., Meiring, M. & Conradie, C. (2010). Valproic acid associated platelet dysfunction: Case Report. African Journal of Haematology and Oncology 1(2):54-56. 50 Index of list of publications related to this compilation Supplement Page S-1 52 S-2 59 S-3 66 S-4 73 S-5 81 S-6 87 S-7 94 S-8 101 S-9 111 S-10 118 S-11 123 S-12 131 S-13 143 S-14 150 S-15 160 S-16 168 S-17 173 S-18 176 S-19 188 S-20 198 S-21 202 S-22 208 S-23 216 S-24 220 S-25 224 S-26 229 S-27 235 S-28 248 S-29 259 S-30 271 S-31 280 S-32 285 S-33 295 S-34 298 S-35 305 S-36 308 51 S-1 Site52s of Elimination and Pharmacokinetics of Recombinant [131I]Lepirudin in Baboons S. M. MEIRING,*,† M. G. LÖTTER,‡ P. N. BADENHORST,† E. BUCHA,§ G. NOWAK,§ AND H. F. KOTZEÄ † Contribution from Department of Haematology and Cell Biology, University of the Orange Free State, P.O. Box 339 (G2), Bloemfontein, Republic of South Africa, Department of Medical Physics, University of the Orange Free State, Bloemfontein, Republic of South Africa, and Max-Planck Gessellschaft, Pharmacological Haemostaseology, Friedich Schiller University, Jena, Germany. Received October 12, 1998. Final revised manuscript received January 28, 1999. Accepted for publication February 4, 1999. Abstract 0 Lepirudin has a short half-life, and only 50−60% of the Subjects and Methods intravenously administered dose is excreted by the kidneys. The fate Experimental AnimalssEight male baboons (Papio ursinus) of the remainder is unknown. We designed a study to determine the supporting permanent femoral arteriovenous shunts (A-V shunts) fate of this lepirudin. In each of six baboons, [131I]lepirudin was given of silicone rubber tubing7,8 were used. The animals were sedated intravenously as a bolus or infused over 30 min, 24 h apart. The in with intramuscular ketamine hydrochloride (10 mg/kg body mass, vivo redistribution of [131I]lepirudin was determined and quantified by Centaur Laboratories, South Africa) to enable handling. Anesthe-sia was maintained with subsequent administrations when neces- scintillation camera imaging. In all studies, the half-life of [131I]lepirudin, sary. One hundred milliliters of saline were given intravenously as determined from the disappearance of radioactivity, was 21 ± 3 (iv) to each baboon 30 min before the start of a study to ensure min. The half-life determined from the disappearance of lepirudin, that they were normovolaemic. measured by the Ecarin Clotting Time (ECT) method, was similar at Experimental ProtocolsSix baboons received [131I]lepirudin 23 ± 8 min. Results obtained with the labeled lepirudin are therefore as an iv bolus of 0.3 mg/kg as well as an infusion at a rate of 0.01 comparable with those obtained using the plasma concentration of mg/kg/min for 30 min on separate days. The treatment require- lepirudin. When lepirudin was administered as a bolus, the half-life ments were such that an equal number of baboons received thebolus or infusion on each day. The lepirudin (HBW 023, Hoechst was 18 ± 4 min, and lepirudin was cleared from the plasma at a rate AG, Frankfurt and Behringwerke AG, Marburg, Germany) was of 42 ± 12 mL/min and by the kidneys at 23 ± 2 mL/min. Following labeled with 131-iodine (131I; Radiochemical Centre, Amersham, infusion over 30 min, the half-life and total and renal clearances were UK) using the IODO-GEN method.9 Labeling efficiency was 98 ( not significantly different. In both studies, between 50 and 60% of the 1%. It is important to note that only labeled lepirudin was infused administered lepirudin was excreted by the kidney. Studies on and that we determined the plasma concentration of the labeled sacrificed baboons showed that appreciable amounts of lepirudin were lepirudin (see later). Therefore, when reference is made to lepi-rudin, it refers to the results obtained from the plasma concentra- present in the bile, indicating the liver as a contributor to the elimination tion as determined using the ECT method.10 Similarly, reference of lepirudin. to labeled lepirudin refers to results obtained from radioactive count rates of [131I]lepirudin. Scintillation Camera ImagingsImaging of the in vivo distribu- tion of labeled lepirudin was done with a Large Field of View Scintillation Camera fitted with a high-energy collimator. Image Introduction analysis was done with an A3-MDS data processing system that was interfaced with the camera. Imaging was done in two phases; Hirudin is regarded as the most potent direct inhibitor they are, a dynamic image acquisition phase and a static acquisi- of thrombin and its recombinant form, lepirudin ([leu1, tion phase. Thr2]-63-disulfatohirudin), has the same physiochemical Dynamic Image AcquisitionsThe baboons were positioned in characteristics and biochemical properties as the native front of the detector of the camera so as to include the kidneys, protein.1,2 Lepirudin is a 65 amino acid polypeptide (7 kDa) bladder, liver, and spleen in the field of view. Dynamic image 1 acquisition, consisting of 3-min images (64 × 64 word mode) forproduced by transfected yeast cells. The interaction be- 120 min (infusion study) and 90 min (bolus study), started tween hirudin and thrombin results in the formation of a simultaneously with the infusion of [131I]lepirudin. For analysis, stable, noncovalent stoichiometric 1:1 complex that inhibits a region of interest was selected for the kidneys, the bladder, the all functions of thrombin.1,2 Apart from its inhibition of spleen, and the liver, and the radioactivity in each region was coagulation, it also inhibits platelet-dependent arterial-type expressed as a percentage of injected [131I]labeled lepirudin, which thrombosis when given in high enough dosages.3-5 Al- was estimated by the geometrical mean method of quantification though the pharmacokinetics of lepirudin is well-known, (see Static Image Acquisition). The radioactive count rates of the there are still some unresolved issues. It is generally left and right kidneys were summed to obtain total kidney accepted that lepirudin is mainly excreted in an unchanged radioactivity. form by the kidneys.1,6 However, approximately 40 to 50% Static Image Acquisition and Quantification of In Vivo Distribu-tion of [131I]LepirudinsAfter the dynamic image acquisition, static of injected protein cannot be accounted for in the urine. anterior and posterior images of 3 min (64 × 64 word mode) were The sites of metabolism and/or uptake of this lepirudin are acquired of the head, thorax and abdomen, and legs. A background unknown. We designed this study to investigate the image was also acquired to correct whole body and region of metabolism and fate of lepirudin in baboons. interest (organ) radioactivity for background radioactivity. Before treatment on the second day, anterior and posterior images of the thorax and abdomen were acquired to determine and correct for * Corresponding author. Telephone: 0027-51-4053039. Fax: 0027- 131 131 51-4441036. E-mail: GNHMSMM@MED.UOVS.AC.ZA. residual I radioactivity as a result of infusion of [ I]lepirudin † Department of Haematology. on the previous day. ‡ Department of Medical Physics. To obtain anterior and posterior whole body radioactivity, the § Max-Planck Gessellschaft. radioactivity in the head, thorax and abdomen, and legs was © 1999, American Chemical Society and 10.1021/js980407q CCC: $18.00 Journal of Pharmaceutical Sciences / 523 American Pharmaceutical Association Published on Web 04/15/1999 Vol. 88, No. 5, May 1999 52 corre5ct3ed for background radioactivity and/or residual radioactivity For the bolus injection, CLrenal was calculated by dividing the and summed. Regions of interest for the kidneys, bladder, spleen, maximum radioactive count rate in the bladder by AUC0-∞ of the and liver were selected to determine organ radioactivity. On the plasma radioactivity (radioactive count rate/mL)-time profiles. first day of the experiment, these radioactivities were corrected Extravascular ClearancesThe plasma clearance of lepirudin by for background radioactivity. On the second day, organ radioactiv- the extravascular compartment was estimated by calculating the ity was corrected for background radioactivity and residual difference between the total clearance and the renal clearance. radioactivity, measured before the infusion of [131I]lepirudin. The Determination of Sites of Lepirudin Accumulation in in vivo organ distribution of [131I]lepirudin was quantified by the Sacrificed BaboonssTwo baboons were used. The one baboon geometric mean method.11 was sacrificed after 30 min of infusion of [131I]lepirudin. The other Blood Collection and AnalysissBlood, 4.5 mL in 0.5 mL of one received no lepirudin and was sacrificed to obtain organs and a 3.2% sodiumcitrate solution, was collected before a study was body fluids that served as controls for immunohistochemistry and started and then every 15 min thereafter. The radioactivity in 1 lepirudin determinations, respectively. The baboons were deeply mL of plasma was determined in a gamma counter. Plasma levels anesthetized. The test baboon was exsanguinated through the of lepirudin were determined by the ECT method, which was arterial side of the A-V shunt. About 2 L of saline at 37 °C were specifically developed to measure hirudin levels in plasma and simultaneously infused under pressure (140 mmHg) into the body fluids.10 The activated partial thromboplastin time (aPTT) venous side of the shunt to replace the blood that was lost. 131 was measured with a fibrinometer (Clotex II, Hyland Division, Ex Vivo Quantification of the Distribution of [ I]Lepirudins Travenol Laboratories, Costa Mesa, CA), and reagents were The liver, spleen, kidneys, and bladder of the sacrificed baboons supplied by the same company. The aPTT was measured to give were immediately excised by careful dissection. Residual radio- an indication of the level of anticoagulation achieved and to show activity in the body and that present in the liver, spleen, kidneys, that the labeled lepirudin was functional. and bladder were determined by scintillation camera imaging. 11 Images of 3 min (64 × 64 word mode) were acquired. Organ To determine if the labeling procedure affected the function of 131 radioactivity was then expressed as a percentage of total body[ I]lepirudin, unlabeled and labeled lepirudin were added to radioactivity, obtained by summation of residual and organ plasma obtained from 4 baboons. Plasma concentrations were 2.5 radioactivities. and 5.0 µg/mL of lepirudin. In addition, a 50/50 mixture of labeled Organ BiopsiessBiopsies of the kidneys (cortex and medulla), and unlabeled protein was also used. The aPTT was determined spleen, and liver were collected to establish the location of the on all samples. lepirudin by using immunohistochemistry techniques. A primary Urine Collection and AnalysissUrine was collected into a antibody, rabbit anti-hirudin antibody (diluted 1:100 in PBS; urine-collecting bag with a Teflon-coated latex Foley catheter (GRS Celsus Laboratories, Cincinnati, OH) was added to thin processed Medicals, Kelvin, SA) for up to 24 h and analyzed for lepirudin by sections of the tissues and incubated for 60 min. A secondary goat the ECT method.10 The total amount of lepirudin was calculated anti-rabbit antibody (Signet kit, Johnson & Johnson, Johannes- from the total urine volume and the concentration in urine. burg, Republic of South Africa) was added to bind to the primary Urine was also analyzed by sodium dodecylsulfate polyacryla- antibody and a third antibody complex, a peroxidase-antiperoxi- mide gel electrophoresis (SDS-PAGE). Samples analyzed were dase immune complex, was used as the marker complex. These [131I]lepirudin, urine without [131I]lepirudin (native urine), [131I]- antibody complexes become visible when the tissues are exposed lepirudin added to native urine, and urine collected from the to the substrate diaminobenzidine. By adding hydrogen peroxide treated baboons. The results strongly suggest that [131I]lepirudin to the diaminobenzidine solution, a brown precipitate forms where was excreted in an unchanged form. The results are not given. lepirudin is present. Light microscopy was used to interpret the CalculationssPlasma Half-Lives of Lepirudin and [131I]- stains. Lightmicrographs of all tissues were enlarged 200 times LepirudinsThe elimination half-lives of lepirudin and [131I]- and examined for the presence of lepirudin in cells. lepirudin were calculated by adjusting a one-phase exponential Biopsies of the kidneys, liver, and spleen were also homogenized function to the appropriate phases of the log-linear plasma in saline and centrifuged at 8000 g for 24 h. The ECT method 10 concentration-time and radioactive count rate-time profiles by was used to test for the presence of lepirudin in the supernatant the method of least-squares analysis. The value k, in the function of the homogenates. C x e-ktmax , was thus found, where Cmax is the maximum lepirudin Collection and Analysis of Duodenal and Gastric Juices and concentration or maximum radioactive count rate in plasma. The BilesDuodenal and gastric juices and bile were collected from the half-lives were calculated as t ) 0.693/k. sacrificed baboons to test for the presence of lepirudin by the ECT1/2 131 method. 10 Total Clearance of Lepirudin and [ I]LepirudinsFor the StatisticssStatistical comparison was done using Student’s t constant infusion, total clearance (CLtotal) was calculated by test for paired data. Values of p < 0.05 were considered significant. dividing the rate of infusion (mg/min or radioactive count rate/ The results are expressed as mean ( 1 standard deviation (SD). min) by the maximum concentration of lepirudin or radioactive count rate in the plasma. To estimate the maximum plasma concentration or plasma radioactive count rate, an exponential association function was fitted to the data points obtained during Results the infusion period and extrapolated to maximum (i.e., an estimate Changes in aPTT and Pharmacokinetics of of steady-state levels).12 LepirudinsIn the in vitro studies, concentrations of 2.5 For the bolus injection, CLtotal was calculated by dividing the and 5.0 µg/mL lepirudin lengthened the aPTT to 87 ( 4 total dose, or the total counts injected, by the area under the curve 131 (AUC0-∞) of the plasma lepirudin concentration (µg/mL)-time or and 174 ( 5 s. [ I]Lepirudin, at the same concentrations, plasma radioactivity (radioactive count rate/ml)-time profiles. The lengthened the aPTT to 92 ( 4 and 180 ( 6 s. When labeled area under the plasma concentration-time and plasma radioactive and unlabeled lepirudin were mixed, the aPTT lengthened counts rate-time curves up to 90 min (AUC0-90) were calculated to 90 ( 3 and 178 ( 4 s. The difference in measurements using the trapezoid rule. The curves were extrapolated to ∞ using was not significant. The labeling process therefore does not the formula AUC0-∞ ) AUC0-90 + Cp*/ke, where Cp* is the plasma affect the function of lepirudin. lepirudin concentration or plasma radioactive count rate at the The in vivo changes in aPTT are summarized in Table last sampling time (90 min). The terminal rate constant (ke) was 1. The labeled hirudin was biologically active because it determined from the slope of the terminal portion of the log-linear caused a 2-3-fold lengthening in the aPTT. When lepirudin Cp versus time curve. infusion was stopped and following the bolus injection, the Renal Clearance of [131I]LepirudinsFor the constant infusion, aPTT rapidly shortened to reach near pre-infusion values renal clearance (CLrenal) was calculated by dividing the rate of 131 after 90 min.appearance of [ I]lepirudin in the bladder (radioactive count rate/ min) by the maximum counts in the plasma. The latter was The changes in circulating lepirudin and labeled lepiru- estimated by fitting an exponential association function to the din are summarized in Figure 1. Following infusion or the available radioactive count rate/mL-time profiles and extrapo- bolus injection, the plasma levels of lepirudin and [131I]- lated to estimate the maximum value where a steady state was lepirudin decreased rapidly. In all studies, the plasma half- reached. life of lepirudin was 24 ( 9 min. When calculated from the 524 / Journal of Pharmaceutical Sciences Vol. 88, No. 5, May 1999 53 Table514sChanges in APTTa time (min) infusion bolus −30 40 ± 4 − −10 − − 0 131± 20 42 ± 6 30 79 ± 10 98 ± 16 60 64 ± 9 74 ± 13 90 56 ± 8 64 ± 12 a Results are given in seconds and are expressed as mean ± 1 SD. Figure 2sDynamic changes of [131I]lepirudin in the (A) kidneys and (B) bladder following infusion (open circles) and bolus administration (black circles). The radioactivity is expressed as a mean percentage ± 1 SD of whole body radioactivity at the end of each study. Table 3sOrgan Distribution of [131I]Lepirudin at the End of Treatment organ infusion bolus kidneys 4.3 ± 1.5 5.7 ± 2.3 bladder 51.2 ± 5.2 51.2 ± 10.3 liver 1.7 ± 0.8 1.8 ± 0.9 spleen 3.6 ± 2.4 3.0 ± 2.7 Figure 1sDynamic changes of (A) plasma lepirudin and (B) [131I]lepirudin a following infusion (open circles) or bolus administration (black circles). Values Percent of whole body radioactivity at the end of each study. Values are are given as a mean ± 1 SD. given as mean ± 1 SD. Table 2sTotal, Renal, and Extravascular Clearance of Lepirudin and Table 4sECT of Bile and Duodenal and Gastric Juices, and of the [131I]Lepirudin Supernatant of Homogenized Kidney, Liver, and Spleen Tissue a route clearance r-Lepirudin [131I]Lepirudin sample ECT, s infusion total 49 ± 16 52 ± 10 bile 229 renal − 32 ± 10 gastric juice 49 extravascular − 20 ± 8 duodenal juice 39 bolus total 42 ± 12 45 ± 12 kidney 77 renal − 23 ± 2b liver 44 extravascular − 22 ± 12 spleen 49 a aValues are given in mL/min and are expressed as mean ± 1 SD. b p < Control values for the ECT method are between 38 and 40 s. 0.05, infusion versus bolus (Student’s t test for paired data). In Vivo Distribution of [131I]LepirudinsThe changes disappearance of radioactivity from the blood, it was 21 ( in radioactivity in the kidneys and bladder are summarized 6 min. The difference was not statistically significant (p > in Figure 2, and the in vivo organ distribution at the end 0.05). After infusion of [131I]lepirudin was stopped, lepiru- of the study in Table 3. Only ≈5% of total injected [131I]- din was eliminated from the plasma with a half-life of 23 lepirudin was quantified in the liver and the spleen. The ( 8 min ([131I]lepirudin ) 21 ( 3 min). When the same results obtained in the sacrificed baboon, 0.1% of the amount of [131I]lepirudin was given as a bolus, the half- injected radioactivity in the spleen and 1.5% in the liver, life was 18 ( 4 min ([131I]lepirudin ) 19 ( 8 min). The also showed that the liver and spleen contained little [131I]- mean difference in the half-lives between infused and bolus lepirudin. Between 50 and 60% of the injected labeled injected lepirudin was 5 ( 10 min. The 95% confidence lepirudin could be accounted for in the bladder. At the end interval of between -5 and 16 min indicates that the of infusion or shortly after the bolus was injected, the difference was not statistically significant. kidneys contained ≈15% of the injected radioactivity Total, renal, and extravascular clearance of lepirudin is (Figure 2A). summarized in Table 2. Total clearance was ≈45 mL/min, The presence of lepirudin in the bile and gastric and whether it was calculated from the data obtained from the duodenal juices of the sacrificed baboon was measured by concentration of lepirudin or from labeled lepirudin. Be- the ECT method, and the results are shown in Table 4. Of tween 51 and 61% of this lepirudin was cleared by the these, only the bile contained appreciable amounts as kidneys. indicated by the markedly longer than normal ECT results. Journal of Pharmaceutical Sciences / 525 Vol. 88, No. 5, May 1999 54 55 Figure 3sLight micrographs (enlarged 200X) of the (A) renal cortex and (B) medulla. A1 and B1 are from baboons that received lepirudin. A2 and B2 are negative controls where no anti-hirudin antibody was used. A3 and B3 show the renal cortex and medulla of the control baboon that did not receive lepirudin but was treated with anti-hirudin antibody. The amount of lepirudin excreted by the kidneys was from baboons treated with lepirudin also stained negative calculated from the total urine volume and the concentra- when the tissues were not incubated with the antibody tion of r-hirudin in urine. In two cases (1 × bolus study (Figure 3, A2 and B2, and Figure 4, A2 and B2), indicating and 1 × infusion study), very little urine could be collected. that the positive staining obtained in the test samples was The results are nevertheless given. After infusion of labeled not a result of the staining procedure. lepirudin, 2.2 ( 1.9 mg of lepirudin was excreted, after the In the renal cortex, the cytoplasm of the epithelial cells bolus 2.2 ( 1.5 mg was excreted. This amount was ≈54% of the convoluted tubules, the cytoplasm of the endothelial of the total dose of lepirudin given. cells of the blood vessels, and the connective tissue stained ImmunohistochemistysThe results are presented in for lepirudin. The glomeruli stained negative (Figure 3, A1). Figures 3 and 4. The kidney, liver, and spleen of the baboon The renal medulla stained intensely. It is likely that that was not treated with lepirudin stained negative when lepirudin was present in the epithelium of the thick and the tissues were incubated with the anti-hirudin antibody thin segments of the loop of Henly and the collecting tubes. (Figure 3, A3 and B3, and Figure 4, A3 and B3). This result The cytoplasm of the endothelial cells of the vas recta and indicates that these tissues did not contain proteins that the connective tissues between the cells stained strongly interact with the antibody. Kidney, spleen, and liver tissue (Figure 3, B1). 526 / Journal of Pharmaceutical Sciences Vol. 88, No. 5, May 1999 55 56 Figure 4sLight micrographs (enlarged 200X) of the (A) liver and (B) spleen. A1 and B1 are from baboons that received lepirudin. A2 and B2 are negative controls where no anti-hirudin antibody was used. A3 and B3 show the liver and spleen of the control baboon that did not receive lepirudin but was treated with anti-hirudin antibody. The spleen did not stain for lepirudin (Figure 4, A1). On lepirudin, when labeled with 131I, was not markedly af- the other hand, staining of liver is ambiguous (Figure 4, fected by the labeling procedure. First, treatment with the B1). The cytoplasm of the hepatocytes did not stain labeled lepirudin lengthened the aPTT 3-4-fold (Table 1). strongly, but the cytoplasm of the endothelial cells of the In addition, when unlabeled and labeled hirudin, either hepatic artery and portal vein and that of the epithelial alone or as a 50/50 mixture, were added to plasma in vitro, cells of the bile duct stained strongly. the lengthening in aPTT was equivalent. This result indicated that the labeled lepirudin was functional. Second, Discussion when the half-life was measured from the radioactive countrate in blood, it was 21 ( 6 min, which was not significantly In a study of this nature, where the pharmacokinetics different from that estimated for lepirudin (24 ( 9 min). and mechanism of clearance of a compound labeled with a These estimates were also not significantly different from radioisotope is studied, it is vital that the labeled compound that obtained in another study.5 Third, after 90 min of reacts in the same way as the native compound. There are treatment with labeled lepirudin, no accumulation of 131I several lines of evidence to show that the behavior of was imaged in the thyroid or spleen, both of which rapidly Journal of Pharmaceutical Sciences / 527 Vol. 88, No. 5, May 1999 56 remo5v7e free iodine in plasma.13 It is therefore reasonable tered as a bolus (Table 2). However, when renal clearance to conclude that labeling of lepirudin with 131I did not was expressed as a percentage of plasma clearance, 62 ( adversely affect the function and plasma clearance of 16% of infused lepirudin was cleared by the kidneys. The lepirudin and that the 131I remained bound to the lepirudin. corresponding value for bolus administrated lepirudin, 55 The results obtained with the labeled lepirudin can there- ( 15%, was not significantly different (p > 0.05, Student’s fore be compared with those of lepirudin, and conclusions t test for paired data). Thus, although the rate of clearance relating to the pharmacokinetics of lepirudin can be made was different, the relative contribution of the kidneys to from the results obtained with labeled lepirudin. plasma clearance was the same. It is important to note that It is evident that lepirudin was mainly excreted by the in rhesus monkeys that received a bolus of lepirudin, renal kidneys (Figures 2A and B). Between 50 and 60% of the clearance also contributed ≈50% to plasma clearance.18 injected lepirudin was detected in the bladder 90 min after These results are similar to those obtained in this study. it was injected (Table 3), which agrees with results obtained The method of administration of the labeled lepirudin did in humans and rhesus monkeys.6,14,15 Because the clearance not affect the extravascular clearance (Table 2). of lepirudin approximates that of creatinine in humans, it was proposed that excretion be by glomerular filtration.14,15 In summary, whether 0.3 mg/kg of labeled lepirudin was We used immunohistochemical techniques in an attempt infused over 30 min or given as a bolus, it did not influence to verify this proposal (Figure 3). The results were rather its plasma half-life and clearance. This result suggests that confusing and do not confirm glomerular filtration. Glom- the study was performed with plasma concentrations below eruli in the renal cortex did not stain for lepirudin. Tissue the maximum threshold levels of excretion of lepirudin by surrounding the glomerulus and the blood vessel did stain. the kidneys. The mode of administration did not affect the Even more surprising was the fact that the tubuli in the clearance of lepirudin from plasma, nor that by the kidneys medulla, and especially the epithelial cells lining the or the extravascular compartment. The kidneys were the tubules, stained intensely, which suggests tubular reab- main sites of excretion of lepirudin, where between 50 and sorption and secretion. A well-planned study to investigate 60% was excreted, probably by tubular secretion. It has glomerular filtration and possible tubular reabsorption and been speculated that the remainder may be catabolized by secretion is called for. Such a study will provide definitive the kidneys and that the methods used to determine answers about the mechanisms by which the kidneys lepirudin are unable to recognize the degradation prod- excrete lepirudin. ucts.6,14,18,19 The quantification of the percentage labeled It is generally accepted that the fraction of lepirudin that lepirudin found in the bladder argues against renal catab- is not excreted by the kidneys is distributed in the olisation. If the labeled lepirudin was catabolized, much extravascular compartment.14 The in vitro results in this more than 50 to 60% of administered labeled lepirudin study however showed that the liver plays a part in the would have been detected in the bladder at the end of the catabolism of lepirudin and that lepirudin was excreted in study. In addition, SDS-PAGE suggested excretion in an the bile. Although we could not quantify appreciable unchanged form. A surprising finding was that the bile quantities of [131I]labeled lepirudin in the liver (Table 3), contained appreciable amounts of lepirudin. immunohistochemical studies showed that lepirudin was present in the hepatocytes and the tissues surrounding the blood vessels and bile ducts. The finding that lepirudin was present in the bile means that it has to be excreted into References and Notes the digestive system. We explain the absence of lepirudin 1. Adkins, J. C.; Wilde, M. I. Lepirudin: A review of its potential in the duodenal juices (Table 4) and the finding that place in the management of thrombotic disorders. Biodrugs lepirudin was not present in the feces of normal humans 1998, 10, 227-255. 24 h after it was administered15 by digestion and absorption 2. Harvey, R. P.; Degryse, E.; Stefani, L.; Schamber, F.; of lepirudin by the enzymes present in the duodenum. It Cazenave, J. P.; Courtney, M.; Tolsthev, P.; Lecocq, J. P. is reasonable to assume that the 40% of administered Cloning and expression of c DNA coding for the anticoagulant lepirudin that was not excreted by the kidneys, could be hirudin from the blood sucking leech, Hirudo medicinalis.Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 1084-1088. excreted by the liver into the bile. However, a well-planned 3. Kelly, A. B.; Marzec, U. M.; Krupski, W.; Bass, A.; Cadroy, study is needed to quantify the contribution of the liver to Y.; Hanson, S. R.; Harker, L. A. Hirudin interruption of the clearance of lepirudin from plasma. It is further evident heparin-resistant arterial thrombus formation in baboons. that the spleen plays no part in the catabolism of lepirudin Blood 1991, 77, 1006-1012. (Table 3, Figure 4). 4. Kotzé, H. F.; Lamprecht, S.; Badenhorst, P. N. A 4-h infusion The plasma elimination half-life of lepirudin was 18 ( of recombinant hirudin results in long-term inhibition ofarterial-type thrombosis in baboons. Blood 1995, 85, 3158- 4 min when it was injected as a bolus. When the same 3163. quantity was infused over 30 min, the estimated half-life 5. Kotzé, H. F.; Lamprecht, S.; Van Wyk, V.; Roodt, J. P.; was not significantly different (23 ( 8 min). This result is Badenhorst, P. N. Differences in platelet-dependent throm- similar to the estimated half-life that we determined in bosis and thrombin production does not affect the pharma- baboons where different dosages of lepirudin were infused cokinetics of r-hirudin (HBW 023) in baboons. Thromb. over 30 min.5 The half-lives estimated in the baboons are Haemost. 1995, 73, 1308 (abstract). shorter than the 55-70 min measured in humans, rhesus 6. Nowak, G. Pharmacokinetics of hirudin. Semin. Thromb. monkeys, rats, rabbits, and dogs.6,15-17 We have no easy Hemost. 1991, 17, 145-149. explanation for this discrepancy. 7. Harker, L. A.; Hanson, S. R. Experimental arterial throm-boembolism in baboons: Mechanism, quantification, and The total clearance of lepirudin from plasma was ≈45 pharmacologic prevention. J. Clin. Invest. 1979, 64, 559- mL/min, whether the same amount was infused or given 569. as a bolus. Similar results were obtained when clearance 8. Hanson, S. R.; Kotzé, H. F.; Savage, B.; Harker, L. A. Platelet was calculated from the radioactive radioactive count rates interactions with Dacron vascular grafts. Arteriosclerosis (Table 2). This value is not very different from that 1985, 5, 595-603. determined in rhesus monkeys,18 but differs greatly from 9. Salacinski, P. R. P.; Mclean, C.; Sykes, J. E. C.; Clement- the results reported for humans, rats, rabbits, and dogs.6,15 Jones, V. V.; Lowry, P. J. Iodination of proteins, glycoproteinsand peptides using a solid-phase oxidising agent, 1,3,4,5- Labeled lepirudin, infused over 30 min, was cleared by the tetrachloro-3R, 6R-diphenyl glycoluril (IODO-GEN). Anal. kidneys at a significantly higher rate than that adminis- Biochem 1981, 117, 136-146. 528 / Journal of Pharmaceutical Sciences Vol. 88, No. 5, May 1999 57 10. 5N8owak, G.; Bucha, E. Quantitative determination of hirudin 15. Meyer, B. H.; Luus, H. G.; Müller, F. O.; Badenhorst, P. N.; in blood and body fluids. Semin. Thromb. Hemost. 1996, 22, Röthig, H.-J. The pharmacology of recombinant hirudin, a 197-202. new anticoagulant. S. Afr. Med. J. 1990, 78, 268-270. 11. Kotzé, H. F.; Lötter, M. G.; Badenhorst, P. N.; Heyns, A. du 111 16. Markwardt, F.; Fink, G.; Kaiser, B.; Klöking, H, P.; Nowak,P. Kinetics of In-platelets in the baboons. II. In vivo G.; Richter, M.; Stürzebecher, J. Pharmacological survey of distribution and sites of sequestration. Thromb. Haemost. recombinant hirudin. Pharmazie 1988, 43, 202-207. 1985, 53, 408-410. 12. Lyer, L.; Koza, M.; Iqbal, O.; Calabria, R.; Fareed, J. Studies 17. Nowak, G.; Markwardt, F.; Fink, E. Pharmacokinetic studies on the pharmacokinetics and pharmacodynamics of recom- with recombinant hirudin in dogs. Folia Haematol. 1988, binant hirudin (rHV-Lys47) after intravenous and subcuta- 115, 70-74. neous administration in dogs. Thromb. Res. 1993, 69, 259- 18. Grötsch, H.; Hropot, M.; Berscheid, G.; Crause, P.; Malerczyk, 269. G.; Apidopoulos, G.; Haun, G.; Husak, B. Pharmacokinetic 13. Saha, G. B. Uses of radiochemicals in nuclear medicine. In investigations of the R-human thrombin-hirudin complex in Fundamentals of Nuclear Pharmacy; Saha, G. B., Ed.; rhesus monkeys. Thromb. Res. 1992, 66, 271-275. Springer-Verlag: New York, 1992; pp 227-241. 19. Grötsch, H.; Hropot, M. Degradation of rDNA hirudin and 14. Markwardt, F.; Nowak, G.; Stürzebecher, J.; Griessbach, U.; R-human thrombin-lepirudin complex in liver and kidney Walsmann, P.; Vogel, G. Parmacokinetics and anticoagulant homogenates from rat. Thromb. Res. 1991, 64, 763-767. effect of hirudin in man. Thromb. Haemost. 1984, 52, 160- 163. JS980407Q Journal of Pharmaceutical Sciences / 529 Vol. 88, No. 5, May 1999 58 59 S-2 59 60 60 61 61 62 62 63 63 64 64 65 65 S-3 66 Thrombosis Antithrombotic Effect of Platelet Glycoprotein Ib–Blocking Monoclonal Antibody Fab Fragments in Nonhuman Primates Nancy Cauwenberghs, Muriel Meiring, Stephan Vauterin, Veronika van Wyk, Seb Lamprecht, Jan P. Roodt, Levente Novák, Jolan Harsfalvi, Hans Deckmyn, Harry F. Kotzé Abstract—Platelet adhesion in arterial blood flow is mainly supported by the platelet receptor glycoprotein (GP) Ib, which interacts with von Willebrand factor (vWF) that is bound to collagen at the site of vessel wall injury. Antibody 6B4 is a monoclonal antibody (MoAb) raised against purified human GPIb. MoAb 6B4 inhibits both ristocetin- and botrocetin-induced, vWF-dependent human platelet agglutination. MoAb 6B4 furthermore blocks shear-induced adhesion of human platelets to collagen I. We studied the antithrombotic effect of this inhibitory murine MoAb 6B4 in a baboon model of arterial thrombosis. When injected into baboons, intact IgG and i9t)s2 Ffr(aagbments caused almost immediate thrombocytopenia, whereas injection of the Fab fragments alone did not. Fab fragments were subsequently used to investigate their in vivo effect on platelet deposition onto a thrombogenic device, consisting of collagen-rich, glutaraldehyde-fixed bovine pericardium (0.6 2c),mat a wall shear rate ranging from 700 to 100201. sBaboons were either pretreated with Fabs to study the effect of inhibition on platelet adhesion or treated 6 minutes after placement of the thrombogenic device to investigate the effect on interplatelet cohesion. Pretreatment of the animals with bolus doses ranging from 80 to 640mg/kg Fab fragments significantly reduce11d1In-labeled platelet deposition onto the collagen surface by'43% to 65%. Only the highest dose caused a significant prolongation (doubling) of the bleeding time. Ex vivo ristocetin-induced platelet agglutination was equally reduced. Treatment with a bolus mogf /1k1g0Fab fragments after a thrombus was allowed to form for 6 minutes had no effect on further platelet deposition. We therefore conclude that Fab fragments or derivatives of inhibitory anti-GPIb antibodies may be useful compounds to prevent thrombosis. (Arterioscler Thromb Vasc B. i2o0l00;20:1347-1353.) Key Words:platelet adhesion platelet aggregation thrombosisn glycoprotein Ibn monoclonal antibodies Blood platelets, through the processes of adhesion, acti- initial adhesion, platelets will aggregate, an interaction that isvation, shape change, the release reaction, and aggregaa- ssured through the binding of GPIIb-IIIa complexes with tion, form the first line of defense when blood vessels are multivalent ligands, in particular, vWF and fibrinoge5 nT.he damaged. Platelets form a hemostatic plug at the site of injury platelet GPIb receptor is thus important for initiating a to prevent excessive blood loss. Extensive platelet activation, thrombus at the site of the exposed lesion. Abnormalities in however, may overcome the normal thromboregulatory GPIb, as seen in Bernard-Soulier syndrome, result in defi- mechanisms that limit the size of the hemostatic plug. The cient platelet attachment to a site of vascular injury and a platelets then become major prothrombotic offenders predis- predisposition to clinical bleedin6g. posing to vaso-occlusive disea1s,2e. Lately, much effort has been directed to develop antibodies Platelet adhesion is regarded as the trigger for hemostasisand peptides that can block the binding of the adhesive and thrombosis. When subendothelial collagen is exposed atproteins to GPIIb-IIIa, and many of these are being tested in the site of vessel injury, circulating von Willebrand factor clinical trials.7–9 On the other hand, the development of (vWF) binds to it and, under the influence of arterial blood compounds that interfere with the vWF-GPIb axis has lagged flow, undergoes a conformational change enabling it to bind behind. Only a few in vivo studies that investigated the to its receptor, glycoprotein (GP) Ib, on the platelet mem- effects of inhibition of platelet adhesion on thrombogenesis brane3. As a result, motion of the platelets on the collagen are described. They include the use of anti-vWF monoclonal surface is decreased, which may then be followed by firm antibodies (MoAbs)1,0–12 GPIb-binding snake venom pro- adhesion mediated by other collagen recep4toArsft.er the teins like echiceti1n3 and crotalin1,4 aurintricarboxylic acid Received September 16, 1999; revision accepted January 4, 2000. From the Laboratory for Thrombosis Research (N.C., S.V., H.D.), Interdisciplinary Research Center, KU Leuven Campus Kortrijk, Kortrijk, Belgium; the Department of Haematology and Cell Biology (N.C., M.M., V.v.W., S.L., J.P.R., H.F.K.), University of the Orange Free State, Bloemfontein, South Africa; and the Department of Clinical Biochemistry and Molecular Pathology (L.N., J.H.), Medical School, University of Debrecen, Debrecen Hungary. Correspondence to Nancy Cauwenberghs, Laboratory for Thrombosis Research-IRC. K U Leuven Campus Kortrijk, E Sabbelaan 53, B-8500 Kortrijk, Belgium. E-mail Nancy.Cauwenberghs@kulak.ac.be © 2000 American Heart Association, Inc. Arterioscler Thromb Vasc Biiosl.available at http://www.atvbaha.org 66 1347 Downloaded from http://ahajournals.org by on October 22, 2018 1348 67 Arterioscler Thromb Vasc Biol.May 2000 that binds to vWF1,5 and recombinant vWF fragments like To prepare F(ab29)fragments, MoAb 6B4 was dialyzed overnight VCL,16–18 all of which inhibit vWF-GPIb interaction. All of against a 0.1 mol/L citrate buffer (pH 3.5). The antibody was these studies showed the potential feasibility of antithrom- digested by incubation with pepsin (Sigma Chemical Co; 1 part pepsin to 200 parts MoAb) for 1 hour at 37°C. Digestion was stopped botic therapy based on inhibition of the GPIb-vWF axis. by adding 1 volume of a 1 mol/L Tris-HCl buffer (pH 9) to 10 A number of potent, inhibitory anti-GPIb antibodies have volumes of antibody solution. been produced and were extensively tested with respect to their Monovalent Fab fragments were prepared by papain digestion. A in vitro effect under both static (platelet agglutination, vWF 1:10 (vol/vol) solution of 1 mol/L phosphate buffer (pH 7.3) was 19 added to the antibody. Papain (Sigma) was added at a ratio of 1binding) and flow conditions. We are aware of only 2 success- volume papain to 25 volumes of the phosphate buffer containing the ful in vivo studies on guinea pigs, in which F(9a)2bfragments of MoAb, 10 mmol/LL-cysteine-HCl (Sigma), and 15 mmol/L EDTA. PG-1, a monoclonal anti–guinea pig GPIb antibody, were After incubation for 3 hours at 37°C, digestion was stopped by used2.0,21 These fragments were shown to significantly prolong adding freshly prepared iodoacetamide solution (Sigma) to a final the time to arterial microvascular graft thrombosis without concentration of 30 mmol/L, which was then kept in the dark at room temperature for 30 minutes. prolonging the bleeding tim2e1.In another guinea pig model, the Both F(ab92) and Fab fragments were further purified from fragments could effectively reduce thrombus formation on a contaminating intact IgG and Fc fragments by using protein laser-induced injury20. However this PG-1 antibody is specific A–Sepharose. The purified fragments were finally dialyzed against for guinea pig platelets and does not cross-react with humanPBS. The purity of the fragments was determined by SDS–poly- platelets. Part of this rather surprising lack of in vivo studies is acrylamide gel electrophoresis, and the protein concentration wasmeasured by using bicinchoninic acid protein assay reagent A due to the low cross-reactivity of the anti-human GPIb MoAbs (Pierce Chemical Co). with platelets from commonly used laboratory animals. This situation necessitates the use of nonhuman primates as experiF- urther Antibody Characterization mental animals. However, even then, attempts to perform in vivo MoAb 6B4 binds to a (His1–Val289) recombinant GaPI(brGPIba) 26 studies are hampered because injection of the anti-GPIb MoAbs,fragment expressed by Chinese hamster ovary ceilnlsd,icating that its epitope is localized within the amino-terminal region of GaP.Ib as well as the snake venom protein echicetin that reacts withMoAb 6B4 Fabs were further tested for inhibition of ristocetin- and GPIb, invariably causes severe thrombocytope10n,1ia3,.22,23 botrocetin-induced binding of vWF to the rGPaIbfragment by We have studied the antithrombotic efficacy of a novel ELISA.27 Microtiter plates were coated with m5g/mL MoAb 2D4 murine MoAb, 6B4 (IgG1), raised against human platelet for 48 hours at 4°C. MoAb 2D4, an anti-GPIb MoAb that we GPIb, in in vitro and in vivo studies. In vitro, 6B4 and its produced, binds to the rGPaIbfragment but does not block vWF binding. Nonabsorbed sites were blocked with 3% skimmed milk, F(ab9)2 and Fab fragments all potently inhibited the binding whereafter the plates were washed with Tris-buffered saline contain- of vWF to human GPIb under both static and flow conditions, ing 0.1% Tween 20. Purified rGPaIbfragments were immobilized and they also bound to baboon platelets. In the in vivo studies on MoAb 2D4 by incubating 2mg/mL rGPIba for 2 hours at 37°C. in which 6B4 was injected into baboons, the intact MoAb After being washed with Tris-buffered saline–Tween 20, increasing concentrations of 6B4 Fab fragments diluted in Tris-buffered saline– caused immediate and severe thrombocytopenia; injection ofTween 20 were added, followed by 1.25 or 0m.6g/mL purified its F(ab9)2 fragments resulted in a moderate decrease in human vWF, respectively, when ristocetin (3m0g0/mL) or botroce- platelet count, whereas the Fab fragments did not have atin (0.5 mg/mL) was used as a modulator. Binding of vWF was major effect on platelet count. Furthermore, the Fab frag- determined by incubating the resulting mixture for 1 hour with ments were studied in a baboon model of platelet-dependenthorseradish peroxidase–conjugated polyclonal anti-vWF antibody (Dako), diluted 1/3000 in Tris-buffered saline–Tween 20. The color arterial thrombosis. Platelet deposition onto collagen-rich reaction, stopped by addition of 4 mol/L2SHO4, was generated with bovine pericardium was inhibited when the fragments were o-phenylenediamine (Sigma). Ristocetin was supplied by abp, and injected into the baboons before a thrombus was generated.purified vWF was purchased from the Red Cross, Belgium. The On the other hand, when the Fab fragments were injected purification of botrocetin from crudeBothrops jararacavenom (Sigma) was performed as previously describ28after a thrombus was allowed to form, no inhibition of further ed.The effect of 6B4 Fab on shear-induced platelet adhesion to thrombosis was observed. collagen was tested in a Sakariassen-type parallel-plate flow cham- ber at shear rates of 650, 1300, and 260–01 ass previously Methods described2.9 Human collagen type I (Sigma) was dissolved in 50 mmol/L acetic acid (1 mg/mL), dialyzed for 48 hours against Preparation and Purification of Intact MoAb 6B4, PBS, subsequently sprayed onto plastic Thermanox coverslips, and F(ab*)2, and Fab Fragments stored at room temperature overnight before use. Twelve milliliters The antibody used in these studies, 6B4 (subtype IgG1), is a murine of blood anticoagulated with low-molecular-weight heparin (25 MoAb raised against purified human GPIb. GPIb was purified from U/mL; Clexane, Rhône-Poulenc Rorer) was preincubated with 6B4 outdated platelet concentrates collected by the Belgian Red Cross inFab fragments at 37°C for 5 minutes. This was then used to perfuse Leuven, Belgium, according to the method described by Wicki and the collagen-coated coverslips. After 5 minutes of perfusion, the Clemetson2.4 BALB/c mice were immunized by intraperitoneal platelets were fixed with methanol and the coverslips stained with injection with the purified GPIb. Murine antibodies were prepared by May-Grünwald/Giemsa. Platelet adhesion (percent of total surface conventional hybridoma technology as previously descr2ib5eadn,d covered with platelets) was evaluated with a light microscope generated hybridoma cells were screened for antibody production in connected to an image analyzer. An average of 30 fields per an ELISA by using purified GPIb. Hybridoma cells producing coverslip were analyzed. Platelet adhesion was expressed as percent anti-GPIb MoAbs were grown and subsequently injected into maximal platelet adhesion obtained in the absence of inhibitor. pristane-primed BALB/c mice. After 10 days ascites fluid was collected. The IgG was extracted from the ascites by using protein Animal Studies A–Sepharose CL-4B (Pharmacia). These antibodies were screenedNormal male baboons (Papio ursinus) were used. The animals for inhibition of ristocetin-induced human platelet aggregation as weighed between 10 and 15 kg and were disease-free for at least 6 described below. MoAb 6B4 IgG totally abolishes both the weeks before the experiments. All procedures were approved by the ristocetin- and the botrocetin-induced aggregation of human plate- Ethics Committee for Animal Experimentation of the 6U7niversity of let-rich plasma (PRP). the Free State in accordance with the National Code for Animal Use Downloaded from http://ahajournals.org by on October 22, 2018 68 Cauwenberghs et al GPIb Inhibition Inhibits In Vivo Thrombosis1349 in Research, Education, Diagnosis and Testing of Drugs and Relatedplatelets were done as described in de3t0a,3il1. In brief, image Substances in South Africa. acquisition of the grafts, including proximal and distal Silastic The baboons had permanent polytetrafluoroethylene (Teflon)– segments, was done with a large-field-of-view scintillation camera silicone rubber (Silastic) arteriovenous shunts implanted in the fitted with a high-resolution collimator. The images were stored on femoral vessels30. Blood flow through the shunts varied between 100 and analyzed with a Medical Data System3scAomputer (Medtronic) and 120 mL/min, resulting in wall shear rates between 700 and 1000 interfaced with the scintillation camera. Dynamic 2-minute image s21, which compares well with the shear rates found in medium-size acquisition was started simultaneously with the start of blood flow arteries3.0 Handling of the baboons was achieved through anesthesia through the devices. A 2-minute image (1321828-byte mode) of a with ketamine hydrochloride ('10 mg/kg IM; Anaket-V, Centaur 3-mL autologous blood sample collected in EDTA was also acquired Laboratory). each time that the grafts were imaged to determine circulating blood radioactivity (blood standard). A region of interest of the graft Study Protocol segment was selected to determine the deposited and circulating To test the effect of the MoAb on platelet count, 6B4, its F(2ab9) radioactivity in each of the dynamic images. Radioactivity in a fragments, or its Fab fragments were administered to 3 different region of similar size in the proximal segment of the extension tubing baboons. The plasma volume was calculated by assuming a bloodwas determined and subtracted from the radioactivity in the graft volume of 65 mL/kg body mass and correcting for the hemato31crit. region to calculate deposited radioactivity. Platelet deposition was Platelet-dependent arterial thrombus formation was induced by using expressed as the total number of platelets deposited. The method to bovine pericardium (0.6 c2mfixed in buffered glutaraldehyde3)2.The calculate this has been described in de3t0ail. pericardium was built into the wall of Silastic tubing (3-mm inside diameter). The method of preparation of the thrombogenic device Laboratory Measurements has been described in det3a1ile, xcept that fixed bovine pericardium was used instead of Dacron vascular graft material. In each experi- Receptor Binding Studies ment, a thrombogenic device prefilled with saline to avoid a 6B4, its F(ab92) fragments, or Fab fragments were labeled with 12N5I a blood-air interface was incorporated as an extension segment into the(Amersham) by using the Iodo-Gen meth3o4dI.odo-Gen was pur- permanent arteriovenous shunt by means of Teflon conne3c0tors. chased from Pierce. Baboon PRP, adjusted with autologous plasma Two approaches were followed to determine the effect of 6B4 Fab to a count of 100 000 platelemtsL/ , was incubated with different fragments on platelet adhesion and interplatelet cohesion onto theconcentrations of iodinated 6B4, F(a2bf9r)agments, or Fab fragments collagen of the bovine pericardium. for 15 minutes at room temperature. The mixture was layered onto 20% sucrose buffer (wt/vol) containing 0.1% (wt/vol) BSA and Dose-Response Effect of 6B4 Fab Fragments on Plateletcentrifuged for 4 minutes at 10 000ing Eppendorf tubes. The top Adhesion and Deposition fluid, including the plasma, was removed and the pellets were Seven baboons were used in the first studies; in total 13 perfusion counted in a gamma counter. Binding studies were performed in experiments were performed. In the first 5 experiments (3 baboons), duplicate on the PRP of 2 baboons. a thrombogenic device was placed to determine the deposition of platelets (see the section below on graft imaging and quantification In Vitro and Ex Vivo Platelet Aggregation of platelet deposition). After 30 minutes, the device was removed The aggregation of platelets in response to ristocetin (ABP) was done and blood flow through the permanent arteriovenous shunt reestab-on 10 mL of blood collected in 1 mL of 3.2% trisodium citrate. PRP lished. Fifteen minutes after removal of the device, each baboon waswas prepared by differential centrifugat3io5nand the platelet count treated with a bolus of 8m0g/kg 6B4 Fab fragments (in 2 mL saline), adjusted to 200 000 platelemtsL/ with autologous plasma. The and 15 minutes after treatment, a second thrombogenic device wasaggregation response was measured in a Monitor IV Plus aggre- placed for 30 minutes to determine the effect of the Fab fragments on gometer (Helena Laboratories) and recorded for 5 minutes. The thrombogenesis. The device was again removed and blood flow percent aggregation at 5 minutes was calculated as the difference in through the permanent shunt established. This step was followed by a second bolus injection of Fab fragments (m8g0/kg) to attain a light transmission between PRP and platelet-poor plasma. cumulative dose of 160 g/kg. After 15 minutes, a third thrombo- In in vitro studies, the PRP was preincubated for 5 minutes withm genic device was placed for 30 minutes and platelet deposition was serial dilutions of intact IgG 6B4, F(a9)b2 fragments, or Fab frag- measured. In 4 other experiments (2 baboons), the same studyments before aggregation was initiated. Inhibition of aggregation protocol was used, except 2 doses of 3m2g0/kg were administered. was calculated from the difference in the aggregation response of Sham studies were done in 4 other experiments (2 baboons). In theseplatelets without and with antibody or fragments. In the ex vivo studies, the same protocol of placement of thrombogenic devices wasdeterminations, inhibition was calculated from the difference in the followed, but the baboons were not treated with Fab fragments. aggregation response of platelets before and after treatment of the baboons. Effect of Anti-GPIb 6B4 Fragments on Plasma levels of 6B4 Fab fragments were measured with a Interplatelet Cohesion sandwich ELISA. In brief, microtiter plates were coated overnight at In a second series of 12 experiments, 6 baboons were used. In all4°C with 5 mg/mL polyclonal goat anti-mouse IgG (Sigma). After baboons, a thrombogenic device was placed for 24 minutes. In 6 unoccupied binding sites were blocked with BSA, serial dilutions of experiments (3 baboons), the baboons received a bolus injection ofbaboon plasma were added to the wells and incubated for 2 hours. Fab fragments of 11m0g/kg. The fragments were injected 6 minutes Bound 6B4 Fab fragments were detected by using goat anti-mouse after placement of the thrombogenic device to allow enough platelets IgG (Fab-specific) conjugated to peroxidase (Sigma) and developed to be deposited to cover the collagen surface. In the other 6 by usingo-phenylenediamine as described above. Standard curves experiments (3 baboons), the baboons did not receive any Fabwere constructed by adding known amounts of 6B4 Fab fragments to fragments. baboon plasma. In both approaches, blood was collected at different time points to determine platelet count and hematocrit (EDTA), circulating and Bleeding Time platelet-associated radioactivity, the ex vivo aggregation of platelets The bleeding time was determined by using the Simplate II device in response to ristocetin, and the plasma concentrations of Fab (Organon Teknika) according to the instructions of the manufacturer. fragments (see the section on laboratory measurements). The timeThe volar surface of the forearm of the baboons was shaved, and a points at which the blood was collected are given in the figures. pressure cuff was applied and inflated to 40 mm Hg. Graft Imaging and Quantification of Statistical Analysis Platelet Deposition Student’s t test for paired data was used to test for statistically Autologous blood platelets were labeled w1it1h1In-tropolone3,3 and significant differences. Data given in the text are me6a8n6SE. Prob- imaging and quantification of the deposition of th11e1In-labeled ability values,0.05 were considered statistically significant. Downloaded from http://ahajournals.org by on October 22, 2018 1350 69 Arterioscler Thromb Vasc Biol.May 2000 Figure 3. Semilog binding curves of anti-GPIb 125I-6B4 IgG (f), 125I-F(ab9)2 (F), and 125I-Fab fragments (Œ) to baboon platelets in plasma. Data are mean of duplicate measurements in PRP from 2 baboons. with an IC50 of 1.8 mg/mL (36 nmol/L) and 2.5mg/mL (50 nmol/L), respectively. The 6B4 Fab fragments inhibited platelet adhesion to Figure 1. Semilog plots showing inhibitory effect of anti-GPIb human collagen type I in a concentration-dependent manner 6B4 Fab fragments on ristocetin- (0.3 mg/mL, A) or botrocetin- at shear rates of 650, 1300, and 260–10(Fsigure 2), with a (0.5 mg/mL, B) induced vWF binding to an rGPIba (His1– Val289)-fragment. The rGPIba fragment (2 mg/mL) was immobi- 50% reduction of surface coverage obtained at a concentra- lized on MoAb 2D4 (5 mg/mL) coated onto microtiter plates. tion of 3.5mg/mL (70 nmol/L), 1.1mg/mL (22 nmol/L), and vWF (1.25 or 0.6 mg/mL, respectively, when binding was 0.5 mg/mL (10 nmol/L), respectively. induced by ristocetin or botrocetin) was added together with increasing concentrations of 6B4 Fab fragments. vWF binding was detected. Data are mean6SD (n53). OD indicates optical In Vitro Effect of MoAb 6B4 and Its F(a*b)2 and density. Fab Fragments on Baboon Platelets Binding of the antibody and its fragments to baboon platelets was dose dependent and saturable. Half saturaKtioD5n)(wasResults 0obtained with 4.7 nmol/L for 6B4 IgG, 6.4 nmol/L for In Vitro Effect of 6B4 Fab Fragments on F(ab9)2, and 49.2 nmol/L for Fab (Figure 3). At saturating Human Platelets concentrations, ristocetin-induced aggregation was com- MoAb 6B4 Fab fragments blocked the ristocetin- (1 mg/mL) pletely abolished (Figure 4). The 5I0Cof platelet aggregation and botrocetin- (0.5mg/mL) induced human platelet aggluti- was obtained at 4.5 nmol/L, 7.7 nmol/L, and 40 nmol/L for nation, with an IC50 of 1.260.3mg/mL (2466 nmol/L) and 6B4 IgG, F(ab92), and Fab fragments, respectively. 2.060.5mg/mL (40610 nmol/L), respectively. For the intact 6B4 IgG, these values were 0m.3g/mL (2.0 nmol/L) and 0.8 Effect of Injection of MoAb 6B4, F(a)b2,*and Fab mg/mL (5.3 nmol/L), respectively. 6B4 binds to an epitope Fragments on Peripheral Platelet Count in Baboons localized on the amino-terminal part (His1–Val289) of In 1 baboon, 100mg/kg intact antibody caused a profound GPIba. The inability of 6B4 to bind to denatured GPIb during decrease in the blood platelet count (,3039/L1)0within 10 Western blotting (not shown) strongly suggests that this minutes after injection. After 48 hours, the platelet count was antibody binds to a conformation-sensitive epitope on GaP. Ib still ,1003109/L. When 100mg/kg 6B4 F(ab92) fragments The 6B4 Fab fragments dose-dependently inhibited both the was injected into 2 baboons, the platelet count decreased ristocetin- (300mg/mL) and botrocetin- (0.5mg/mL) induced rapidly to between 120 and 150391/L0, ie, by '60%, but binding of vWF to the (1–289) rGPaIbfragment (Figure 1), then reached preinfusion values within 24 hours. Finally, when 80 to 320mg/kg of the monovalent 6B4 Fab fragments was injected, the platelet count (45 minutes after injection) decreased by onl'y 10% to 20% and by only 26% when 640 Figure 2. Effect of anti-GPIb 6B4 Fab fragments on shear- induced adhesion of human platelets to human collagen I–coated coverslips. Blood was perfused at a constant shear rate of 650 s21 (F, n52), 1300 s21 (f, n53), and 2600 s21 (Œ, n53) for 5 minutes. Data are mean6SD and are expressed as Figure 4. Semilog plot showing effect of anti-GPIb 6B4 IgG (f), percent of maximal platelet adhesion relative to control value in F(ab9)2 (F), and Fab fragments (Œ) on ristocetin-induced baboon the absence of 6B4 Fab (35%, 33%, and 29%, respectively, at platelet aggregation. Data, expressed as percent inhibit6io9n, are 650, 1300 and 2600 s21). mean of duplicate measurements in PRP from 2 baboons. Downloaded from http://ahajournals.org by on October 22, 2018 70 Cauwenberghs et al GPIb Inhibition Inhibits In Vivo Thrombosis1351 Platelet Counts, Plasma Levels of 6B4 Fab Fragments, Ex Vivo Ristocetin-Induced Platelet Agglutination, and Bleeding Times After 6B4 Fab Administration Platelet Counts % Inhibition of Ex Vivo Plasma Levels, Ristocetin-Induced (1.5 mg/mL) Bleeding Dose, mg/kg Time, min 3103/mL % Decrease mg/mL Platelet Agglutination Times, s n55 0 Pre 307632 0 0.0760.03 0 190620 80 90 272622 11 1.7260.14 2669 160633 160 150 248619 19 4.8460.56 47612* 250645 270 315631 0.4560.09 863 ND n54 0 Pre 283623 0 0.0260.01 0 232642 320 90 219610 23 9.1360.48 25621 340663 640 150 210613 26 15.3561.38 8069* 405645* 270 238620 16 1.1960.09 1569 ND 24 h 236613 17 0.0460.01 763 ND ND indicates not determined. Administration of 80–640 mg/kg 6B4 Fab fragments to baboons was performed. Values are given as mean6SE. Statistical comparisons were made by Student’s t test for paired sample groups: *P,0.05. mg/kg was injected (the Table ). On the basis of this result, the respectively. Doses of 320 and 6m40g/kg (23320mg/kg 6B4 6B4 Fab fragments were used for further studies. Fab) significantly reduced platelet deposition by 56% and 65%, respectively. Effect of Different Doses of MoAb 6B4 Fab Plasma levels of 6B4 Fab fragments and inhibition of ex Fragments on Platelet Deposition vivo agglutination determined on samples obtained 45 min- Platelet adhesion and deposition onto thrombogenic devicesutes or 2 hours after administration changed in both a dose- sequentially placed 30 minutes apart are summarized in and time-dependent manner (the Table). Ex vivo ristocetin- Figure 5. In the sham studies (Figure 5A), placement of the induced platelet aggregation was significantly inhibited at previous graft had no significant effect on platelet deposition doses of 160 and 64m0g/kg. Both ristocetin-induced platelet that formed on subsequent grafts. aggregation and plasma values returned to baseline within 3 In the treatment studies (Figure 5B), dosages of 80 and 160hours after antibody injection. mg/kg (2380mg/kg 6B4 Fab) significantly inhibited platelet Bleeding times determined in the treatment studies before deposition in comparison with control b'y43% and 53%, and 45 minutes after injection of 80 to 32m0g/kg 6B4 Fab fragments were not significantly prolonged. Only a dose of 640 mg/kg significantly prolonged the bleeding time. Effect of 6B4 Fab Fragments on Interplatelet Cohesion Treatment of the baboons with 11m0g/kg 6B4 Fab did not affect platelet deposition when the animals were injected after a thrombus was allowed to form for an initial 6 minutes (Figure 6). Discussion The initial step in platelet adhesion consists of vWF binding to GPIb. We investigated the effect of inhibiting this inter- action on platelet function both in vitro and in vivo. A murine MoAb, 6B4, that binds to a conformational epitope in the amino-terminal part of GPIabwas used. The antibody and its fragments potently inhibited the binding of vWF to an rGPIba fragment (His1–Val289) and dose-dependently in- hibited vWF-dependent human platelet agglutination. The intact antibody and its fragments also dose-dependently Figure 5. Platelet deposition onto 3 thrombogenic devices con- inhibited human platelet adhesion to type I collagen in a flow 21 taining bovine pericardium placed consecutively at times 0 (F), chamber at wall shear rates of 650, 1300, and 260.0Tshis 60 (f,l), and 120 (Œ,) minutes for 30 minutes (top shaded inhibition was shear dependent, ie, more pronounced at bars). A, Sham experiments (n54). B, After injection of 0 (F) higher shear. (n59), 80 (f), and 160 (Œ) mg/kg 6B4 Fab fragments (n55) or 320 (l) and 640 () mg/kg (n54). Values are given as Of interest was the finding that 6B4 did no7t 0react with mean6SE. platelets from dogs, hamsters, pigs, or guinea pigs but did Downloaded from http://ahajournals.org by on October 22, 2018 1352 71 Arterioscler Thromb Vasc Biol.May 2000 compared with GPIIb-IIIa antagonists, like ReoP37roO.n the other hand, the GPIb-vWF interaction in contrast to the GPIIb-IIIa–fibrinogen interaction is the ultimate first step in platelet adhesion under fast blood flow. Because binding of vWF to GPIb also activates platele3t8s,3,9 it is reasonable to assume that by inhibiting vWF-GPIb binding, fewer platelets will be activated. The smaller thrombus that finally forms may therefore be a consequence of both fewer platelets that adhere to collagen and less platelet aggregation. Thus, where GPIIb-IIIa blockers mainly prevent platelet aggregation, in- terruption at an earlier stage by a GPIb blocker is expected not only to limit the platelet plug that is formed but also to Figure 6. Influence of late treatment of baboons with 6B4 Fab reduce additional platelet-dependent effects, such as granule fragments on platelet deposition. The thrombogenic device was placed at time 0 and platelet deposition determined for 24 min- release, thought to play a role in the development of arterio- utes (top shaded bar). After 6 minutes (arrow), animals were sclerosis and restenos4i0s,4.1 either left untreated (f) or treated with a bolus of 110 (F) mg/kg There are also indications that the GPIb-IX-V complex is 6B4 Fab fragments (n56). Values are given as mean6SE. involved in platelet-platelet interactions. Ruggeri et42al recently reported that blocking the GPIb-vWF interaction, bind to baboon platelets with much the same characteristicsafter platelets from PPACK-anticoagulated blood had ad- as to human platelets. As a result, baboons were used for inhered to bovine collagen in vitro for 100 seconds at 15201,0 s vivo and ex vivo studies. Both the intact antibody as well as prevented further thrombus growth measured after another its F(ab9)2 fragments caused immediate thrombocytopenia, 740 seconds, even at low shear rates that do not normally similar to what was seen when other anti-GPIb MoAbs were initiate vWF-dependent platelet adhesion. To test this concept injected into different experimental anima1l0s,2.2,23 On the in vivo, we performed a second series of studies to investigate other hand, the Fab fractions had only a moderate effect on the role of GPIb in platelet-platelet interactions at intermedi- the blood platelet count, and we therefore decided to use the ate shear rates. A thrombogenic device was placed as an Fab fractions to assess the antithrombotic effect of 6B4 in a extension segment in the permanent arteriovenous shunt and baboon model of arterial thrombos3i0s.The glutaraldehyde- exposed to native flowing blood. After 6 minutes the baboons fixed bovine pericardium was highly thrombogenic: after 30 were treated with the Fab fractions of 6B4. We postulated that minutes of exposure to native flowing bloo'd,33109 plate- a 6-minute exposure (number of platelets deposited was lets deposited on the area of 0.62c. mIn similar studies, only 9 '0.6310 9) was sufficient to allow ample coverage of the '0.7310 platelets accumulated in 30 minutes on Dacron 2 pericardium with adhering platelets. Inhibition of plateletvascular graft material (0.9 cm).31 It is therefore also not deposition due to treatment, when compared with sham surprising that a number of control thrombogenic devices be- studies, would therefore reflect inhibition of platelet-platelet came occluded before 30 minutes of exposure to flowing blood. interactions. Because no such effect was seen, it strongly Treatment of baboons with 6B4 Fab fragments inhibited suggested that GPIb does not play a major role in in vivo platelet deposition on the thrombogenic devices by 43% to platelet-platelet interactions under the conditions used in this 65%. The observed effect must be ascribed to the effect of the study. antibody, because sequential placement of thrombogenic In conclusion, we have reported on the first anti-human devices in untreated baboons caused no decreased deposition. GPIb antibody that can be used successfully to prevent No complete inhibition of platelet deposition was observed, platelet adhesion and thrombus formation in vivo, thereby even at high doses. It is possible that this incomplete inhibition was due to the confirming the predominant role of GPIb in platelet adhesion medium shear rates (700 to 100–01) sused in this study. In in vivo. Our studies, however, do not support the hypothesis general, in flow chambers a more effective inhibition with that GPIb also plays a part in platelet-platelet interactions in vWF-GPIb inhibitors is obtained at higher shear ra1t1e,1s2,,36 vivo. On the basis of our results on bleeding times and the which we also demonstrated here by finding a more pro- inhibition of thrombogenesis, we propose that the Fab frag- nounced effect of 6B4 on platelet adhesion to collagen when ments or derivatives of the anti-GPIb MoAb 6B4 may be a higher shear was applied in a flow chamber; also, in vivo an useful compounds in preventing arterial thrombosis in those arterial thrombus was more readily prevented than a venouspatients in whom thrombosis is expected, ie, after vascular one.11 This implies that inhibition of GPIb would result engraftment, endarterectomy, or balloon catheterization. especially in arterial effects, which in addition could result in less bleeding risk. Acknowledgments This investigation was supported by grants FWO V3/5, the Flamand- The doses used indeed caused a significant reduction inHungarian Bilateral, the SA Medical Research Council, the Central thrombus size without apparently increasing the risk of Research Fund of the University of the Orange Free State, and bleeding since bleeding time was only mildly prolonged at Biomed PL96.3517. N. Cauwenberghs was temporarily employed as the highest dose. This finding supports data obtained with a Visiting Scientist at the University of the Orange Free State and other GPIb-vWF–blocking agents (VCL, anti-vWF MoAb was supported by an International Relations Treaty between the University of Leuven and the University of the Orange Free State. AJvW-2) that lengthened the bleeding time only moderate- We are grateful to Drs S. Meyer and B. Steiner for providing the ly.11,12,17,18 This finding is important and might provide a Chinese hamster ovary cells expressing the rGaPfIrbag7m1 ent and major advantage in the development of antithrombotic agents Wim Noppe for his help with the purification of botrocetin. Downloaded from http://ahajournals.org by on October 22, 2018 72 Cauwenberghs et al GPIb Inhibition Inhibits In Vivo Thrombosis1353 References of platelet membrane glycoprotein Ib inhibition and thrombin inhibition. 1. Marcus AJ, Safier LB. Thromboregulation: multicellular modulation of Blood. 1993;82:126–134. platelet reactivity in hemostasis and thrombosFisA.SEB J. 1993;7: 22. Becker BH, Miller JL. Effects of an antiplatelet glycoprotein Ib antibody 516–522. on hemostatic function in the guinea piBgl.ood. 1989;74:690–694. 2. Davies MJ, Thomas AC. Plaque fissuring: the cause of acute myocardial 23. Ravanat C, Freund M, Moog S, Mangin P, Azorsa D, Schwartz C, Lanza infarction, sudden ischaemic death, and crescendo anBgirnHa.eart J. F, Cazenave J. Differential effects of monoclonal antibodies to rat 1985;53:363–373. GPIIb-IIIa, GPIba or GPV on platelet function and in vivo survival. 3. Siedlecki CA, Lestini BJ, Marchant KK, Eppell SJ, Wilson DL, Marchant Thromb Haemost. 1999;82(suppl):528a. Abstract. RE. Shear-dependent changes in the three-dimensional structure of 24. Wicki AN, Clemetson KJ. The glycoprotein Ib complex of human blood platelets.Eur J Biochem. 1987;163:43–50. human von Willebrand factorB. lood. 1996;88:2939–2950. 25. Kohler G, Milstein C. Continuous cultures of fused cells secreting 4. Clemetson KJ. Platelet collagen receptors: a new target for inhibition? antibody of predefined specificityN.ature. 1975;256:495–497. Haemostasis. 1999;29:16–26. 26. Meyer S, Kresbach G, Haring P, Schumpp Vonach B, Clemetson KJ, 5. Phillips DR, Charo IF, Scarborough RM. GPIIb-IIIa: the responsive Hadvary P, Steiner B. Expression and characterization of functionally integrin.Cell. 1991;65:359–362. active fragments of the platelet glycoprotein (GP) Ib-IX complex in 6. Caen JP, Nurden AT, Jeanneau C, Michel H, Tobelem G, Levy-Toledano mammalian cells: incorporation of GP aIbinto the cell surface S, Sultan Y, Valensi F, Bernard J. Bernard-Soulier syndrome: a new membrane.J Biol Chem. 1993;268:20555–20562. platelet glycoprotein abnormality: its relationship with platelet adhesion 27. Vanhoorelbeke K, Cauwenberghs N, Vauterin S, Schlammedinger A, to subendothelium and with the factor VIII von Willebrand proteJinL.ab Mazurier C, Deckmyn H. A new and reproducible vWF:RiCoF assay. Clin Med. 1976;87:586–596. Thromb Haemost2.000;83:107–113. 7. Tcheng JE. Platelet glycoprotein IIb/IIIa integrin blockade: recent clinical 28. Fujimura Y, Titani K, Usami Y, Suzuki M, Oyama R, Matsui T, Fukui H, trials in interventional cardiologyT.hromb Haemost. 1997;78:205–209. Sugimoto M, Ruggeri ZM. Isolation and chemical characterization of two 8. Hanson SR, Sakariassen KS. Blood flow and antithrombotic drug effects. structurally and functionally distinct forms of botrocetin, the platelet Am Heart J. 1998;135:S132–S145. coagglutinin isolated from the venom oBfothrops jararaca.Bio- 9. Coller BS. GPIIb/IIIa antagonists: pathophysiologic and therapeutic chemistry. 1991;30:1957–1964. insights from studies of c7E3 FaTb.hromb Haemost. 1997;78:730–735. 29. Harsfalvi J, Stassen JM, Hoylaerts MF, Van Houtte E, Sawyer RT, 10. Cadroy Y, Hanson SR, Kelly AB, Marzec UM, Evatt BL, Kunicki TJ, Vermylen J, Deckmyn H. Calin fromHirudo medicinalis, an inhibitor of Montgomery RR, Harker LA. Relative antithrombotic effects of mono- von Willebrand factor binding to collagen under static and flow con- clonal antibodies targeting different platelet glycoprotein-adhesive ditions.Blood. 1995;85:705–711. molecule interactions in nonhuman primatBelso.od. 1994;83:3218–3224. 30. Hanson SR, Kotze HF, Savage B, Harker LA. Platelet interactions with 11. Yamamoto H, Vreys I, Stassen JM, Yoshimoto R, Vermylen J, Hoylaerts Dacron vascular grafts: a model of acute thrombosis in baboAorntes-. MF. Antagonism of vWF inhibits both injury induced arterial and venous riosclerosis. 1985;5:595–603. thrombosis in the hamsteTr.hromb Haemost. 1998;79:202–210. 31. Kotze HF, Lamprecht S, Badenhorst PN, van Wyk V, Roodt JP, 12. Kageyama S, Yamamoto H, Nagano M, Arisaka H, Kayahara T, Alexander K. In vivo inhibition of acute platelet-dependent thrombosis in Yoshimoto R. Anti-thrombotic effects and bleeding risk of AJvW-2, a a baboon model by Bay U3405, a thromboxan2e-reAceptor antagonist. monoclonal antibody against human von Willebrand factBor. J Thromb Haemost. 1993;70:672–675. Pharmacol. 1997;122:165–171. 32. Quintero LJ, Lohre JM, Hernandez N, Meyer SC, McCarthy TJ, Lin DS, 13. Peng M, Lu W, Beviglia L, Niewiarowski S, Kirby EP. Echicetin: a snake Shen SH. Evaluation of in vivo models for studying calcification behavior venom protein that inhibits binding of von Willebrand factor and albo- of commercially available bovine pericardiumJ .Heart Valve Dis. 1998; aggregins to platelet glycoprotein IBb.lood. 1993;81:2321–2328. 7:262–267. 14. Chang MC, Lin HK, Peng HC, Huang TF. Antithrombotic effect of 33. Kotze HF, Heyns AD, Lotter MG, Pieters H, Roodt JP, Sweetlove MA, crotalin, a platelet membrane glycoprotein Ib antagonist from venom of Badenhorst PN. Comparison of oxine and tropolone methods for labeling Crotalus atrox.Blood. 1998;91:1582–1589. human platelets with indium-111J.Nucl Med. 1991;32:62–66. 15. Golino P, Ragni M, Cirillo P, Pascucci I, Ezekowitz MD, Pawashe A, 34. Fraker PJ, Speck JC. Proteins and cell membrane iodinations with a Scognamiglio A, Pace L, Guarino A, Chiariello M. Aurintricarboxylic sparingly soluble chloroamide, 1,3,4,6-tetrachloro-3a,6a- acid reduces platelet deposition in stenosed and endothelially injured diphenylglycoluril.Biochem Biophys Res Commun. 1978;80:849–857. rabbit carotid arteries more effectively than other antiplatelet inter- 35. van Wyk V, Heyns AD. Low molecular weight heparin as an antico- ventions.Thromb Haemost. 1995;74:974–979. agulant for in vitro platelet function studiesT.hromb Res. 1990;57: 16. Zahger D, Fishbein MC, Garfinkel LI, Shah PK, Forrester JS, Regnstrom 601–609. J, Yano J, Cercek B. VCL, an antagonist of the platelet GP1b receptor, 36. Ruggeri ZM. Glycoprotein Ib and von Willebrand factor in the process of markedly inhibits platelet adhesion and intimal thickening after balloon thrombus formationA. nn N Y Acad Sci. 1994;714:200–210. injury in the rat.Circulation. 1995;92:1269–1273. 37. Aguirre FV, Topol EJ, Ferguson JJ, Anderson K, Blankenship JC, Heuser 17. McGhie AI, McNatt J, Ezov N, Cui K, Mower LK, Hagay Y, Buja LM, RR, Sigmon K, Taylor M, Gottlieb R, Hanovich G. Bleeding compli- Garfinkel LI, Gorecki M, Willerson JT. Abolition of cyclic flow vari- cations with the chimeric antibody to platelet glycoprotein IIb/IIIa ations in stenosed, endothelium-injured coronary arteries in nonhuman integrin in patients undergoing percutaneous coronary intervention: EPIC primates with a peptide fragment (VCL) derived from human plasma von Investigators. Circulation. 1995;91:2882–2890. Willebrand factor-glycoprotein Ib binding domaiCn.irculation. 1994;90: 38. Kroll MH, Harris TS, Moake JL, Handin RI, Schafer AI. von Willebrand 2976–2981. factor binding to platelet GPbI initiates signals for platelet activation. 18. Yao SK, Ober JC, Garfinkel LI, Hagay Y, Ezov N, Ferguson JJ, Anderson J Clin Invest. 1991;88:1568–1573. HV, Panet A, Gorecki M, Buja LM. Blockade of platelet membrane 39. Asazuma N, Ozaki Y, Satoh K, Yatomi Y, Handa M, Fujimura Y, Miura glycoprotein Ib receptors delays intracoronary thrombogenesis, enhances S, Kume S. Glycoprotein Ib-von Willebrand factor interactions activate thrombolysis, and delays coronary artery reocclusion in doCgirsc.u- tyrosine kinases in human plateleBtsl.ood. 1997;90:4789–4798. lation. 1994;89:2822–2828. 40. Assoian RK, Grotendorst GR, Miller DM, Sporn MB. Cellular transfor- 19. Goto S, Salomon DR, Ikeda Y, Ruggeri ZM. Characterization of the mation by coordinate action of three peptide growth factors from human unique mechanism mediating the shear-dependent binding of soluble von platelets.Nature. 1984;309:804–806. Willebrand factor to plateletsJ. Biol Chem. 1995;270:23352–23361. 41. Ip JH, Fuster V, Israel D, Badimon L, Badimon J, Chesebro JH. The role 20. Miller JL, Thiam-Cisse M, Drouet LO. Reduction in thrombus formation of platelets, thrombin and hyperplasia in restenosis after coronary angio- by PG-1 F(ab92), an anti-guinea pig platelet glycoprotein Ib monoclonal plasty.J Am Coll Cardiol. 1991;17:77B–88B. antibody.Arterioscler Thromb. 1991;11:1231–1236. 42. Ruggeri ZM, Dent JA, Saldivar E. Contribution of distinct adhesive 21. Dascombe WH, Hong C, Garrett KO, White JG, Lyle VA, Miller JL, interactions to platelet aggregation in flowing blooBdl.ood. 1999;94: Johnson PC. Artificial microvascular graft thrombosis: the consequences 172–178. 72 Downloaded from http://ahajournals.org by on October 22, 2018 S-4 73 © 2008 Schattauer GmbH, Stuttgart Animal Models The humanized anti-glycoprotein Ib monoclonal antibody h6B4-Fab is a potent and safe antithrombotic in a high shear arterial thrombosis model in baboons Alexandre Fontayne1, Muriel Meiring2, Seb Lamprecht2, Jan Roodt2, Eddy Demarsin3, Philippe Barbeaux3, Hans Deckmyn1 1Laboratory for Thrombosis Research, IRC, KU Leuven Campus Kortrijk, Belgium; 2Department of Hematology and Cell Biology, Faculty of Health Sciences, University of the Free State, South Africa; 3Thrombogenics, Leuven, Belgium Summary The Fab-fragment of 6B4,a murine monoclonal antibody target- allowing for cyclic flow reductions (CFRs) which are measured ing the human platelet glycoprotein (GP) Ibα and blocking the on an extracorporeal femoral arteriovenous shunt. Injection of binding of von Willebrand factor (VWF), is a powerful anti- 0.5 mg/kg h6B4-Fab significantly reduced the CFRs by 80%, thrombotic. In baboons, this was without side effects such as whereas two extra injections, resulting in cumulative doses of bleeding or thrombocytopenia. Recently, we developed a fully 1.5 and 2.5 mg/kg,completely inhibited the CFRs.Platelet recep- recombinant and humanized version of 6B4-Fab-fragment, tor occupancy, plasma concentrations and effects ex vivo were h6B4-Fab,which maintains its inhibitory capacities in vitro and ex consistent with what was previously observed. Finally, minimal vivo after injection in baboons. We here investigated the anti- effects on bleeding time and blood loss, no spontaneous bleed- thrombotic properties, the effect on bleeding time and blood ing and no thrombocytopenia were observed. We therefore loss and initial pharmacokinetics of h6B4-Fab in baboons.The conclude that h6B4-Fab maintains the antithrombotic capacities antithrombotic effect of h6B4-Fab on acute platelet-mediated of the murine 6B4-Fab, without causing side effects and there- thrombosis was studied in baboons where thrombus formation fore can be used for further development. is induced at an injured and stenosed site of the femoral artery, Keywords Antithrombotic, glycoprotein Ib, humanized antibody, platelet adhesion, thrombosis Thromb Haemost 2008; 100: 670–677 Introduction antithrombotic effect in vivo in baboons (3, 4) but also strongly inhibits ristocetin-induced platelet aggregation ex vivo (3). In Platelets are a key factor in the maintenance of normal haemos- contrast to most antithrombotic drugs, but in line with observa- tasis (1). However, in some pathological situations, such as tions with other experimental compounds that interfere with the stroke or myocardial infarction, vessel damage and enhanced collagen-VWF-GPIb axis (5), 6B4-Fab administration does not shear rates cause platelet activation and thrombus formation induce a significant prolongation of the bleeding time. The epi- leading to vessel occlusion. The initial stage of this process de- tope recognized by 6B4 was first mapped using chimera human/ pends on the binding of platelet glycoprotein Ibα (GPIbα) to von dog rGPIbα to reside within the C-terminal flanking region, be- Willebrand factor (VWF), bound to the collagen matrix exposed tween residues 201 and 268 of GPIbα (6). More recently, by to the blood flow as a consequence of the vessel damage. using computer docking and mutagenesis we further could ident- Previously, we prepared and characterized a murine mono- ify Asp235 and Lys237 as dominant epitope residues, whereas clonal antibody (mAb) against human GPIbα, designated as the paratope of 6B4 is mainly formed by residues Tyr31, Lys32, 6B4. This mAb inhibits VWF-dependent platelet adhesion to Asp33 and Glu98 of the light chain CDR1 and 3 respectively and collagen under high shear-stress conditions, as was shown in Tyr106 of the heavy chain CDR3 (7). From these data, it is clear flow chambers (2). Injection of 6B4-Fab-fragments has a potent that 6B4 directly competes with the VWF A1-domain binding Correspondence to: Received February 6, 2008 Hans Deckmyn Accepted after major revision July 23, 2008 Laboratory for Thrombosis Research, IRC KU Leuven Campus Kortrijk Prepublished online September 5, 2008 E. Sabbelaan 53, B-8500 Kortrijk, Belgium doi:10.1160/TH08-02-0073 Tel.: +32 56 246422, Fax: +32 56 246997 E-mail: Hans.Deckmyn@kuleuven-kortrijk.be 670 73 Downloaded by: University of Free State. Copyrighted material. 74 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons Figure 1:Time schedule for the study. The first blood sample is drawn before the start of surgery. Blood flow is recorded 15 min before damaging the artery (injury). The first series of CFRs is recorded for 30 min (–120 to –90). During this period the second blood sample is drawn (-115 min). Saline is next injected and CFRs are recorded (-90 to –60). After the sa- line control, consecutive doses of h6B4 Fab or saline are administered to treated or control animals respectively and CFRs are recorded for 30min, (-60 to -30, -30 to 0 and 0 to 30 min). Each time a blood sample is drawn 5 min later, whereas consecutive samples are obtained at 30, 60, 150, 300, 600, 1200, 1440 min, 1 week and 1 month to determine full blood count, ristocetin-induced platelet aggregation, recep- tor occupancy and plasma concentration. site within GPIbα (8, 9). As a further step in the development of Animal study 6B4, we prepared a fully recombinant and humanized version of Housing, treatment, surgery and care for the animals used in the the 6B4-Fab-fragment, named h6B4-Fab (10).This molecule has study was approved by the Control Committee on Animal Ex- inhibitory capacities in vitro and ex vivo equivalent to the parent- perimentation of the University of the Free State, South Africa. al murine 6B4-Fab-fragment, and with a comparable Kd of 0.8 ± Seven healthy baboons (Papio hamadryas) of either sex, weigh- 0.1 x 10–8 (10). ing between 8 to 13 kg, were used in this study. Animals were an- In the present study, we validated the antithrombotic proper- aesthetized with ketamine (0.1 ml/kg IM/30 min) prior to sur- ties of h6B4-Fab in a baboon thrombosis model by studying the gery. Dissection was made on the leg to expose 4–5 cm of the fe- antiplatelet effects in the injured and stenosed femoral artery moral vessels.All nearby branches in the femoral artery and vein under high-shear stress conditions (Roodt et al., submitted) and were ligated. Next a small incision in both artery and vein was monitored the bleeding tendency and platelet counts. made in which a catheter was inserted. To the catheter tips, sili- cone tubing was connected, thus shunting the femoral artery to Materials and methods the femoral vein. This setup increases blood flow 3– to 5-fold, yielding flow rates of between 150 and 300 ml/minute. Blood Production and purification of murine and humanized flow was measured by a probe (Transonic systemsTS410, probe: recombinant 6B4-Fab fragments ME3PXL1OO8) attached on the silicone tubing. After a latency The murine and humanized 6B4-Fab-fragments were produced phase of 15 min to stabilize the flow, the endothelium of the fe- by transient expression with the FreeStyle™ 293 Expression moral artery was injured proximally to the vessel tip with a Mar- System (Invitrogen, Carlsbad, CA, USA) with the 293 Free- tin needle holder (Hegar-Baumgartner TC Gold 14cm Product Style™ cells as previously described. Briefly, cells were cultured code 20.634.14) by pressing on the vessel for 10 seconds at to obtain 1.1x106 cells per ml on the day of the transfection. The maximum depression. Two overlapping injuries are made. cells were pelleted before resuspension in fresh, pre-warmed Arterial stenosis was next applied by adjusting a plastic con- FreeStyle™ 293 Expression medium. Plasmid DNA coding for strictor device over the injury site to reduce the flow to 30 ml/min the recombinant 6B4-Fab fragment was diluted into OptiMEM® (80 to 90% stenosis). Injury and stenosis resulted in thrombus in the presence of 293fectin™. After incubation for 30 minutes formation in the femoral artery, detected as a reduction in the (min) at room temperature to allow the DNA-293fectin™ com- flow. When blood flow reached almost baseline, the thrombus plexes to form, the mixture was added to the cell suspension. The was dislodged by opening the constrictor. After this, the external transfected suspension was further incubated for 120 hours (h) at stenosis was restored inducing a new thrombus formation. This 37°C. The cells were discarded and the secreted 6B4-Fab-frag- repeated process results in cyclic flow reductions (CFR’s). Base- ments were dialyzed, concentrated and purified by ion-exchange line CFR’s were recorded for 30 min. Saline was injected and chromatography, first on a Phenyl Sepharose fast flow column CFR’s were monitored for a further 30 min which served as an in- (Pharmacia Biotech, Uppsala, Sweden) followed by a second pu- ternal control. In five baboons, the effect of a first injection of rification on a Q-Sepharose Fast Flow column, (Pharmacia Bio- 0.5 mg/kg of h6B4-Fab, followed by two injections of 1 mg/kg of tech) to remove contaminants. 6B4-Fab-fragments present in the h6B4-Fab separated by a 30-min interval, on the CFR’s was flowthrough were next dialyzed against PBS and their concen- monitored each time for 30 min. Control baboons (n=2) were in- tration determined. The molecules were kept at –20°C before jected with saline instead. Blood samples were drawn at set times use. (Fig. 1) to determine full blood count, GPIbα receptor occupan- cy, bleeding time, blood loss, platelet aggregation ex vivo in- duced by ristocetin and h6B4-Fab plasma levels. As a control 671 74 Downloaded by: University of Free State. Copyrighted material. 75 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons compound, Clopidogrel (Plavix™), purchased from Bristol- (FACSCalibur, BD Biosciences). The percentage of 6B4-b Myers Squibb/Sanofi Pharmaceuticals (New York, NY, USA), bound to baboon platelets was calculated, with 100% set in the was pulverised and suspended in methanol (150 mg/ml). This absence of h6B4-Fab. Receptor occupancy was determined by solution was further diluted in saline to a concentration of 10 mg/ subtracting the percentage of labeled molecules bound to the pla- ml and passed through a 0.22 µm filter. Oral administration does telets after treatment with h6B4-Fab from that obtained before not allow the performance of baseline measurements or the use treatment. of stepping doses and as the active metabolite was not available, we decided to follow the procedure described by Yao et al . (11), Plasma concentration which was done with the co-operation of the Sanofi Research de- h6B4-Fab plasma concentrations were determined in an ELISA partment. assay as described previously (10). Briefly, rGPIbα was captured by mAb 2D4 precoated on an ELISA plate. Baboon plasma was Platelet count and blood loss then added in a dilution series to the wells. After 1.5 h incubation The full blood count including platelet numbers was evaluated at RT, unbound proteins were removed and bound h6B4-Fab, was before the surgery, 5 min after each bolus injection of h6B4-Fab, detected with Fab-specific goat anti-human IgG HRP-labeled and 30 min, 1, 2.5, 5, 10, 20, 24 h and one week after the last in- (Sigma, St.Louis, MO, USA) and color development by addition jection. As described elsewhere (10), blood samples were col- of H2O2 and orthophenylenediamine (OPD, Sigma).The reaction lected in 3.13% trisodium citrate and measured on a Technicon was stopped with H2SO4 before optical density determination H3 blood cell analyzer (Bayer AG, Leverkusen, Germany). Pa- (490–630 nm) on a microplate reader EL340 (Biotek instrument, rameters related to bleeding tendency were monitored in two Winooski, VT, USA). The plasma concentration of h6B4-Fab in ways: the skin template bleeding time was determined at the sur- each sample was calculated using exponential increase first face of the shaved forearm as the time needed from the skin inci- order equations from standard curves. These standard curves sion to cessation of bleeding, as previously described (10). In ad- were constructed by in-vitro addition of known concentrations of dition, we evaluated the blood loss during 30 min from a stan- h6B4-Fab to baboon plasma. dardized incision of 2 cm long and 0.8 cm deep, made by using a Swann-Morton #15 (Roodt et al, submitted). Pre-weighed and Statistical analysis standardized gauzes were inserted in the wound and replaced The different parameters tested were compared by the Student after 30-min periods. Blood loss was calculated by subtracting t-test and the differences were considered to be statistically sig- gauze weight before and after use, which was more straightfor- nificant when p < 0.05. ward and more reproducible than measuring haemoglobin. Re- sults were expressed as ratio of the saline control to account for variation among individuals. Results Ristocetin-induced platelet aggregation Haematological parameters and thrombocytopenia Platelet aggregation ex vivo was investigated as previously de- The full blood count and platelet number were evaluated before scribed (10). Briefly, a baboon platelet suspension was prepared the surgery, 5 min after each bolus injection of h6B4-Fab, but at 300x103 platelets/µl by mixing platelet-rich plasma (PRP) and also 30 min, 1, 2.5, 5, 10, 20, 24 h, one week and one month after autologous platelet-poor plasma (PPP) prepared by differential the last injection. The white blood cells (WBC) increased from centrifugation from citrated blood. Platelet aggregation was in- 6.2 ± 0.3 103/µl up to 14.7 ± 0.6 103/µl with a predominance of duced by ristocetin A (ABP corp., Marlton, NJ, USA) and was neutrophils (Table 1); however, they were almost back to normal performed at 37°C, 1,000 rpm on a C560CA aggregometer after one day. This increase was also observed in the controls (Chrono-log corp, Halfway House, South Africa). A higher con- (and in baboons treated with other drugs, J Roodt unpublished centration of ristocetin (3 mg/ml) than used with human platelets data) albeit to a somewhat lesser extent, implying that the in- (1 mg/ml) was needed to make sure that platelets of all baboons crease is likely due to the surgical procedure. would agglutinate. The extent of aggregation was estimated No marked thrombocytopenia was observed throughout the quantitatively by measuring the curve amplitude, and expressed experiment (Table 1). A small decrease (max. 30%) in the pla- in percentage of the value from the pretreatment sample. telet count was observed in both treated and control animals near the end of the thrombosis study which would indicate that a non- Receptor occupancy negligible platelet consumption occurs in the model. Later on, GPIbα receptor occupancy by h6B4-Fab-fragment was esti- platelet counts increased and normalized within a day. mated using flow cytometry by competition with a fixed concen- tration of biotinylated 6B4 (6B4-b) (home made, [10]) in com- Antithrombotic effect: CFR inhibition bination with streptavidin/phycoerythrin (PE) (BD Biosciences, After the surgical procedure the CFRs were recorded for 30 min Erembodegem, Belgium) as described previously. Baboon pla- to ascertain stable periods. Injection of saline had no effect on the telets (7x106), prepared from citrated blood, were first incubated CFRs compared to the control before injection. On average 10,0 with 10 ng of 6B4-b followed by 5 µl of streptavidin/PE. Both in- CFRs were recorded during the 30-min period with a standard cubations were done for 15 min at room temperature. Samples deviation of 2.9 (n= 8), indicating that both the damage induced were finally fixed with 0.2% formaldehyde, 0.9% NaCl before to the vessel and the artificial stenosis applied, resulted in a re- determining platelet-associated fluorescence by flow cytometry producible model. In contrast, administration of 0.5 mg/kg 672 75 Downloaded by: University of Free State. Copyrighted material. 76 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons A B Figure 2: Cyclic flow reductions. Representative recording of the blood flow in the arteriovenous shunt (A). Every injection is marked by an arrow. Histogram of the CFRs (B).Values of the CFRs during 30 min recording are expressed as percentage of the number of CFRs following saline injection. Mean values ± SD (n=5). * statistically different p<0.05, ** p<0.005 t-test for paired sample groups. h6B4-Fab significantly reduced the number of CFRs to only 20% of the control (Fig. 2A, B). Two extra injections of 1 mg/kg h6B4-Fab completely inhibited the CFRs, resulting in a continu- ous patent artery. Bleeding time and blood loss The bleeding time was not significantly affected by injections of h6B4-Fab (Table 2), except for a mere two-fold, but statistically significant, increase 60 min after the highest dose, which re- turned to baseline within 24 h. Blood loss was determined by measuring the blood adsorbed by a standardized gauze in a standardized incision, over 30 min. As this parameter was relatively variable between but much less within baboons, the blood loss was expressed as the fold increase over the value seen at baseline in the same animal (Table 2). The maximum increase was about 2.6 at the cumulative dose of 1.5 mg/kg h6B4-Fab, which was statistically significantly different compared to saline at the same time point, and did not further in- crease when a full antithrombotic effect was obtained with 673 76 Table 1: White blood cell and platelet counts in baboons treated with saline or h6B4-Fab in function of time.Values are range (saline n=2) and mean ± SEM (h6B4-Fab n= 5). Per- centage of platelet count is between brackets. * statistically different p<0.05 t-test for paired sample groups. Time –125 min –115 min –55 min –25 min 5 min 30 min 60 min 150 min 300 min 600 min 1200 min 1440 min 1 week control baseline 0.5mg/kg 1.5mg/kg 2.5mg/kg - - - - - - - - White Blood Cells (103/µl) Saline 5.8–7.0 5.0–7.8 3.7–10.4 4.3–10.6 5.8–8.5 6.8–8.1 nd nd nd nd nd nd nd h6B4-Fab 6.2 ± 0. 3 5.7 ± 0.4 7.8 ± 0.4 9.3 ± 0.6 11.5 ± 0.5* 13.4 ± 0.8* 13.9 ± 0.7* 14.0 ± 0.6* 14.0 ± 0.8* 14.0 ± 0.8* 14.7 ± 0.6* 12 ± 0.4* 8.9 ± 0. 4 Platelets (103/µl) Saline 280–311 281–361 265–276 221–249 231–239 231–250 nd nd nd nd nd nd nd - (100 ± 0) (85.4 ± 6.3) (73.8 ± 3.4) (74.2 ± 5.6) (76.5 ± 8.8) h6B4-Fab 403.8 ± 21.2 363.2 ± 3. 2 328.8 ± 14. 5 309.8 ± 11. 9 292.0 ± 10.3 267.2 ± 13.4* 298.6 ± 9.9 312.4 ± 11.2 327.0 ± 13.7 330.4 ± 7.8 353.4 ± 11.3 336.2 ± 9.1 537.2 ± 18.9 - (100 ± 0) (91.3 ± 4.9) (85.8 ± 4) (80.9 ± 3.5) (74.2 ± 4.3) (82.7 ± 3.5) (86.6 ± 3.9) (90.9 ± 4.8) (91.5 ± 3.1) (98 ± 4.1) (93.2 ± 3.5) (148.6 ± 5.6) Downloaded by: University of Free State. Copyrighted material. 77 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons Table 2: Bleeding parameters in baboons treated with saline or h6B4-Fab in function of time. Template bleeding time is expressed in seconds. nd indicates not determined. Blood loss is expressed as the ratio of the volume of blood that was adsorbed during 30 min by a gauze from a standardized incision upon saline administration.Values are range (saline n=2) and mean ± SEM (h6B4-Fab n= 5). * statistically different p<0.05 t-test for paired sample groups. Time –125 min –115 min –55 min –25 min 5 min 30 min 60 min 600 min 1200 min 1440 min control baseline 0.5mg/kg 1.5mg/kg 2.5mg/kg - - - - - Bleeding time (s) Saline 225–345 nd 150–240 135–165 180–210 105–270 nd nd nd nd h6B4-Fab 84 ± 2.7 nd 156 ± 14.5 165 ± 21 177 ± 13.8 126 ± 7.5 165 ± 25.7* 132 ± 7.5 120 ± 12.7 99 ± 8.1 Blood loss (ratio over baseline) Saline 0.5–2.6 1.0–1.0 0.5–1.0 0.6–0.4 0.3–0.4 nd nd nd nd nd h6B4-Fab 3.5 ± 0.7 1.0 ± 0.0 2.1 ± 0.4 2.6 ± 0.3* 2.2 ± 0.4 nd nd nd nd nd 2.5 mg/kg (Fig. 3A). A similar full antithrombotic effect was level to 28.7 ± 3.2 µg/ml. The presence of the antibody in the seen after administration of 2.5 mg/kg clopidogrel where; how- blood stream then decreased rapidly and reached baseline in 24 ever, a 4.3 ± 2.6 fold increase in blood loss was observed, that h. Based on these measurements in function of time after the last even further increased to 8.0 ± 5.0 -fold at 5 mg/kg (Fig. 3B). dose, an initial pharmacokinetic analysis was performed using Also 1.6 mg/kg abciximab gave a full inhibition of the CFRs; WinNonlin® Compartmental Modeling Analysis (Pharsight, however, it increased the blood loss by some 20-fold (Roodt et Mountain View, CA, USA): h6B4-Fab is cleared from circu- al., unpublished data). lation with an initial half-life of approximately 15.5 min, with the predominant clearance mechanism occurring for 96% via a 6B4-Fab plasma concentration single compartment, indicating that the secondary half-life is not Intravenous administration of 0.5 mg/kg h6B4-Fab resulted in a relevant for 6B4 pharmacokinetics. plasma concentration of 6.3 ± 1.1 µg/ml 5 min after its injection (Fig. 4A).An additive dose of 1 mg/kg, 30 min after the first one, Receptor occupancy increased the plasma level of the drug to 25.9 ± 3.5 µg/ml. Fin- The first dose of h6B4-Fab only resulted in a partial GPIbα oc- ally, the last injection of h6B4-Fab further increased the plasma cupancy with a maximum of approximately 52% after 5 min (Fig. 4B). Injection of the two doses of 1 mg/kg resulted in 66% and 64% (Fig. 3B) of the GPIbα receptors occupied by h6B4-Fab. After 1 h, the remaining receptor occupancy was 27.8 A ± 4.9% and continued to decrease to 19.2 ± 3.4% 300 min post dose. Interestingly, the receptor occupancy increased again 10 h (600 min) after the last injection and to a lower extent than for the main peak. At this stage the receptor occupancy could be esti- mated to be approximately two times less important than for the initial peak with 35% and 65%, respectively. Finally, after 24 h, the GPIbα occupancy was still 16.5 ± 4%. B Platelet activity: Ristocetin-induced platelet aggregation Already the injection of a cumulative dose of 1.5 mg/kg h6B4-Fab completely inhibited ristocetin-induced platelet ag- gregation (RIPA) after 5 min (Fig. 4C).After the cumulative dose of 2.5 mg/kg, the RIPA was fully inhibited for 30 min more and then recovered slowly within 24 h. Expressing h6B4-Fab plasma concentration versus percen- tage of ristocetin-induced platelet aggregation, shows that 10 µg/ml is enough to fully inhibit ristocetin-induced platelet ag- gregation (Fig. 5A). At this same concentration also the GPIbα Figure 3: Relationship between antithrombotic effect and blood loss. occupancy was maximal. The relationship between the receptorThe influence of increasing doses of h6B4 (a, n= 5) and of clopidogrel (b, n= 3) on the number of CFRs per 30 min (black bars) occupancy and ristocetin-induced platelet aggregation revealed and the relative increase of blood loss over saline (white bars) (mean ± that an estimated receptor occupancy of ~ 50% is enough to fully SD). inhibit the ristocetin-induced platelet aggregation (Fig. 5B). 674 77 Downloaded by: University of Free State. Copyrighted material. 78 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons A A B B C Figure 5: Relationship between h6B4-Fab plasma level, the per- centage of GPIbα occupancy (●) and ristocetin induced pla- telet aggregation (RIPA) (❍) (A) and between percentage of Figure 4: Ex-vivo effects of h6B4-Fab administration to ba- GPIbα occupancy and RIPA (B). To construct the curves, mean valu- boons. Successive doses of h6B4-Fab were administered to baboons es for each parameter were used. and antibody plasma levels (A), GPIbα occupancy (B) and ex-vivo ris- tocetin-induced platelet aggregation (C) were determined as described in materials and methods. Data represent the mean ± SEM calculated We have tackled this problem via an alternative new ap- from five independent experiments. proach. Aspirin, clopidogrel or GPIIb/IIIa blockers prevent the development of thrombi, however, not exclusively under high shear stress conditions as seen in an atherosclerotic, stenotic en- Discussion vironment. On the other hand, it is known that bridge formation between platelets and collagen by VWF is a prerequisite to allow Cardiovascular diseases remain the main cause of death in West- platelets to withstand the elevated shear forces, whereas the role ern countries (12). Treatment normally consists of anticoagu- of VWF at lower shear is redundant at best. So interfering with lants and/or antiplatelet drugs, often in conjunction with percut- this adhesion pathway might result in a specific targeting of the aneous interventions. The antiplatelet agents commonly used are antithrombotic action to stenotic arteries, leaving the rest of the aspirin and clopidogrel. Clinical trials have demonstrated sig- circulation intact, with presumably less risk for spontaneous nificant efficacy of these drugs and their combination results in bleeding. improved outcomes, albeit with a small but significant inherent We have shown that the monoclonal anti-human GPIbα anti- risk of increased bleeding (13). The development of GPIIb/IIIa body, 6B4, inhibits platelet adhesion to human collagen type I in blockers has been a most remarkable improvement to antiplatelet a perfusion flow chamber and this more efficiently with increas- approaches; however, after many clinical trials, it seems that this ing shear rate (15). The antithrombotic property of the mAb was class of blockers only can be used effectively in the prevention of demonstrated in baboon models of arterial thrombosis (3, 4) in ischemic complications and mortality after percutaneous coron- which intravenous injection of 6B4-Fab-fragment at 0.6 mg/kg ary intervention (14). These studies also demonstrated that e.g. reduced significantly and at 2 mg/kg completely blocked platelet abciximab injection can induce thrombocytopenia and bleeding. dependent thrombus formation. This antithrombotic effect was Limited efficacy is likely due in part to the inability to reach not accompanied by a prolongation of the bleeding time, in maximal therapeutic efficacy limited by the increasing risk of contrast to what is normally found with regular antiplatelet drugs bleeding (13). (13). In view of these promising results, we recently prepared a 675 78 Downloaded by: University of Free State. Copyrighted material. 79 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons In the present study, we now furthermore studied the anti- What is known about this topic? thrombotic effect of h6B4-Fab-fragment in injured and stenosed – The murine anti-glycoprotein Ib antibody 6B4 is anti- baboon femoral arteries, a further modification of the original thrombotic in baboons canine model of Folts (21). Wu et al. (4) previously published – and does not prolong the template bleeding time. that a full inhibition of the CFRs was obtained by injecting 2 mg/ kg of murine 6B4-Fab in baboons. In our study, a cumulative What does this paper add? dose of 1.5 mg/kg h6B4-Fab already resulted in a complete in- – The humanised anti-glycoprotein Ib antibody 6B4 is anti- hibition of the CFRs. Therefore, the humanized version of thrombotic in baboons 6B4-Fab fragment retains the antithrombotic effect of the parent- – and does neither prolong the template bleeding time nor al murine 6B4, and this without a marked prolongation of the increase the blood loss from a standardised incision. bleeding time or increased blood loss, which may be indicators – A new cyclic flow reduction model in the baboon is used, of enhanced bleeding tendency. The blood loss is increased over where the damage and stenosis is applied to the femoral baseline; however, about 4 and 10 times less than observed with artery, and the changes in blood flow are determined on clopidogrel and abciximab, respectively (Roodt et al., sub- an extracorporeal femoral arteriovenous shunt. mitted). The lack of effect on the bleeding time is in agreement with results obtained with different blockers of the GPIb/VWF axis such as VCL, a recombinant VWF-A1 domain (22, 23), AjvW-2-Fab, an inhibitory antibody against the VWF-A1 do- fully recombinant and humanized version of 6B4 called main, or its humanized version,AJW200 (17, 24) or 82D6A3, an h6B4-Fab (10) with similar affinity and in-vitro and ex-vivo ef- anti-VWF-A3 antibody that prevents VWF binding to collagen ficacy in different models as the murine Fab. In line with that (25). Prolonged bleeding times, however, could be induced by study, we here demonstrated again that h6B4-Fab causes a dose- using doses higher than needed for the antithrombotic effect (17, and time-dependent inhibition of ristocetin-induced platelet ag- 26). Although it is indeed impossible to extrapolate bleeding risk gregation ex vivo. As we previously observed with the murine from bleeding times, very encouraging results recently were ob- 6B4-Fab (4), 60% of GPIbα occupied by h6B4-Fab is enough to tained with an inhibitory Fab-fragment against GPIbα in a mu- totally inhibit the thrombus formation but also to inhibit the ex- rine model of acute experimental stroke, where reduced brain in- vivo platelet aggregation induced by 3 mg/ml ristocetin. This is farct volumes and improved neurological status were not accom- in perfect agreement with the findings of van Zanten et al. (16), panied by an increase in bleeding complications (27). who reported that the minimal GPIbα number needed to affect In conclusion, the present study demonstrates that the hu- platelet adhesion under flow conditions is 50%. The value of manised 6B4-Fab is a powerful inhibitorof the GPIbα/VWF axis 50% seems to be really critical, since with a receptor occupancy and again proves that inhibiting platelet GPIbα is an efficient of 40%, the inhibition of the RIPA is almost negligible. In the way to prevent thrombus formation in the injured and stenosed same way Kageyama et al. (17) showed that 50% VWF occupied baboon femoral artery without causing prolongation of the by AJW200 was enough to totally inhibit the CFRs but also the bleeding time or increased blood loss. All together this results in botrocetin-induced platelet aggregation ex vivo. Finally, GPIbα a much larger therapeutic window than available with the current remains partially occupied (16.5%) 24 h post injection, which antithrombotics. Accordingly, inhibition of GPIbα by com- again is in line with what we previously observed with h6B4-Fab pounds such as h6B4-Fab, is a promising approach that needs to and murine 6B4-Fab (4, 10). It nevertheless is clear that the ac- be validated in clinical trials aimed to prevent acute arterial tual receptor occupancy likely might be somewhat underesti- thrombotic syndromes. mated, since the biotinylated 6B4, used to determine the remain- ing free binding sites, will to some extent compete with the Acknowledgements bound h6B4-Fab. Another explanation why even at the highest This work was supported by a grant from the Instituut voor de Aanmoedig- doses never an occupancy of 100% is reached could be that as a ing van Innovatie door Wetenschap en Technologie in Vlaanderen (IWT consequence of antibody binding GPIbα may become internal- 020473) and by a Bilateral Collaboration grant between Flanders and South ized or shed (18, 19), as is e.g. seen with antibodies against GPVI Africa (BIL/04/56). AF was a EU-RTN (HPRN-CT-2002–00253) postdoc- (20); this, however, awaits further verification. toral fellow. References 1. Ruggeri ZM. Mechanisms initiating platelet throm- human primates. Arteriosclerosis Thromb Vasc Biol 6. Cauwenberghs N, Schlammadinger A, Vauterin S, bus formation. Thromb Haemost 1997; 78: 611–616. 2000; 20: 1347–1353. et al. Fc-receptor dependent platelet aggregation in- 2. Cauwenberghs N, Vanhoorelbeke K, Vauterin S, et 4. Wu D, Meiring M, Kotze HF, et al. Inhibition of pla- duced by monoclonal antibodies against platelet glyco- al. Epitope mapping of inhibitory antibodies against telet glycoprotein Ib, glycoprotein IIb/IIIa, or both by protein Ib or von Willebrand factor. Thromb Haemost platelet glycoprotein Ibalpha reveals interaction be- monoclonal antibodies prevents arterial thrombosis in 2001; 85: 679–685. tween the leucine-rich repeat N-terminal and C-ter- baboons. Arteriosclerosis Thromb Vasc Biol 2002; 22: 7. Fontayne A, De Maeyer B, De Maeyer M, et al. minal flanking domains of glycoprotein Ibalpha. Blood 323–328. Paratope and epitope mapping of the antithrombotic 2001; 98: 652–660. 5. Vanhoorelbeke K, Ulrichts H,Van De WG, et al. In- antibody 6B4 in complex with platelet glycoprotein Ib- 3. Cauwenberghs N, Meiring M, Vauterin S, et al. hibition of platelet glycoprotein Ib and its antithrom- alpha. J Biol Chem 2007; 282: 23517–23524. Antithrombotic effect of platelet glycoprotein Ib- botic potential. Curr Pharm Design 2007; 13: 8. Huizinga EG, Tsuji S, Romijn RA, et al. Structures blocking monoclonal antibody Fab fragments in non- 2684–2697. of glycoprotein Ibalpha and its complex with von 676 79 Downloaded by: University of Free State. Copyrighted material. 80 Fontayne et al. Antithrombotic properties of h6B4-Fab in baboons Willebrand factor A1 domain. Science 2002; 297: clonal antibodies that differentiate between shear-in- fragment (VCL) derived from human plasma von 1176–1179. duced and ristocetin/botrocetin- induced glycoprotein Willebrand factor-glycoprotein Ib binding domain. 9. Uff S, Clemetson JM, Harrison T, et al. Crystal Ib-von Willebrand factor interaction. Haemostasis Circulation 1994; 90: 2976–2981. Structure of the Platelet Glycoprotein Ibalpha N-ter- 2000; 30: 139–148. 23. Azzam K, Garfinkel LI, Bal dit SC, et al. Anti- minal Domain Reveals an Unmasking Mechanism for 16. van Zanten GH, Heijnen HFG, WuYP, et al. A fifty thrombotic effect of a recombinant von Willebrand fac- Receptor Activation. J Biol Chem 2002; 277: percent reduction of platelet surface glycoprotein Ib tor, VCL, on nitrogen laser-induced thrombus 35657–35663. does not affect platelet adhesion under flow conditions. formation in guinea pig mesenteric arteries. Thromb 10. Fontayne A, Vanhoorelbeke K, Pareyn I, et al. Blood 1998; 91: 2353–2359. Haemost 1995; 73: 318–323. Rational humanization of the powerful antithrombotic 17. Kageyama S, Matsushita J, Yamamoto H. Effect of 24. Kageyama S, Yamamoto H, Nakazawa H, et al. anti-GPIba antibody: 6B4. Thromb Haemost 2006; 96: a humanized monoclonal antibody to von Willebrand Pharmacokinetics and pharmacodynamics ofAJW200, 671–684. factor in a canine model of coronary arterial thrombo- a humanized monoclonal antibody to von Willebrand 11. Yao SK, McNatt J, Cui K, et al. CombinedADP and sis. Eur J Pharmacol 2002; 443: 143–149. factor, in monkeys. Arterioscler Thromb Vasc Biol thromboxane A2 antagonism prevents cyclic flow vari- 18. Cramer EM, Lu H, Caen JP, et al. Differential redis- 2002; 22: 187–192. ations in stenosed and endothelium-injured arteries in tribution of platelet glycoproteins Ib and IIb-IIIa after 25. Wu D, Vanhoorelbeke K, Cauwenberghs N, et al. nonhuman primates. Circulation 1993; 88: 2888–2893. plasmin stimulation. Blood 1991; 77: 694–699. Inhibition of the von Willebrand (VWF)-collagen inter- 12. Murray CJ,.Lopez AD. Mortality by cause for eight 19. Gardiner EE, Karunakaran D, Shen Y, et al. Con- action by an antihuman VWF monoclonal antibody re- regions of the world: Global Burden of Disease Study. trolled shedding of platelet glycoprotein (GP)VI and sults in abolition of in vivo arterial platelet thrombus Lancet 1997; 349: 1269–1276. GPIb-IX-V by ADAM family metalloproteinases. J formation in baboons. Blood 2002; 99: 3623–3628. 13. Cooke GE, Goldschmidt-Clermont PJ. The safety Thromb Haemost 2007; 5: 1530–1537. 26. Kageyama S, Yamamoto H, Nakazawa H, et al. and efficacy of aspirin and clopidogrel as a com- 20. Nieswandt B, Schulte V, Bergmeier W, et al. Long- Anti-human vWF monoclonal antibody, AJvW-2 Fab, bination treatment in patients with coronary heart dis- term antithrombotic protection by in vivo depletion of inhibits repetitive coronary artery thrombosis without ease. Expert Opin Drug Saf 2006; 5: 815–826. platelet glycoprotein VI in mice. J Exp Med 2001; 193: bleeding time prolongation in dogs. Thromb Res 2001; 14. De Meyer SF, Vanhoorelbeke K, Ulrichts H, et al. 459–469. 101: 395–404. Development of monoclonal antibodies that inhibit 21. Folts JD, Crowell EB, Jr., Rowe GG. Platelet aggre- 27. Kleinschnitz C, Pozgajova M, Pham M, et al. Tar- platelet adhesion or aggregation as potential anti- gation in partially obstructed vessels and its elimin- geting platelets in acute experimental stroke: impact of thrombotic drugs. Cardiovasc Hematol Dis Drug Tar- ation with aspirin. Circulation 1976; 54: 365–370. glycoprotein Ib, VI, and IIb/IIIa blockade on infarct gets 2006; 6: 191–207. 22. McGhie AI, McNatt J, Ezov N, et al. Abolition of size, functional outcome, and intracranial bleeding. 15. Cauwenberghs N, Ajzenberg N, Vauterin S, et al. cyclic flow variations in stenosed, endothelium-injured Circulation 2007; 115: 2323–2330. Characterization of murine anti-glycoprotein Ib mono- coronary arteries in nonhuman primates with a peptide 677 80 Downloaded by: University of Free State. Copyrighted material. 81 S-5 Thrombosis Inhibition of Platelet Glycoprotein Ib, Glycoprotein IIb/IIIa, or Both by Monoclonal Antibodies Prevents Arterial Thrombosis in Baboons Dongmei Wu, Muriel Meiring, Harry F. Kotze, Hans Deckmyn, Nancy Cauwenberghs Abstract—The antithrombotic efficacy of the monoclonal antibodies 6B4-Fab and MA-16N7C2 against platelet glycoprotein (GP) Ib and GP IIb/IIIa, respectively, on acute platelet-mediated thrombosis was evaluated in a baboon model of femoral artery stenosis, which is a modification of the original Folts model: platelet thrombi form on the injured stenosed artery, producing cyclic flow reductions (CFRs). A dose of 0.6 mg/kg 6B4-Fab significantly reduced the CFRs by 5915%, whereas 2 mg/kg 6B4-Fab completely abolished the CFRs without prolongation of the bleeding time. MA-16N7C2 inhibited CFRs by 438% at a dose of 0.1 mg/kg and abolished the CFRs at a dose of 0.3 mg/kg but with a significant prolongation of the bleeding time. Finally, the combination of 0.6 mg/kg 6B4-Fab and 0.1 mg/kg MA-16N7C2 fully prevented the CFRs without prolongation of the bleeding time. The present study demonstrates that the inhibition of platelet GP Ib function by 6B4-Fab is a powerful intervention to prevent platelet thrombus formation in injured arteries without prolongation of the bleeding time; the latter is in contrast to the result after the inhibition of GP IIb/IIIa. Moreover, we demonstrate that combining a GP Ib blocker with a GP IIb/IIIa blocker can achieve a strong antithrombotic effect without increasing the bleeding time. This provides new information that will be beneficial in designing clinical therapeutic approaches. (Arterioscler Thromb Vasc Biol. 2002;22:323-328.) Key Words: platelet glycoprotein Ib  platelet glycoprotein IIb/IIIa  cyclic flow reductions  antithrombotic agents  bleeding time Platelets adhere to the subendothelium of damaged blood Blocking GP Ib9–11 or vWF12 results in an inability of thevessels through the collagen–von Willebrand factor platelets to attach to the exposed subendothelium. Therefore, (vWF)–platelet glycoprotein (GP) Ib axis. vWF forms the the GP Ib–vWF axis is an attractive target on which to focus bridge between platelets and collagen in the vessel wall for the prevention of thrombus formation in stenosed arteries. (especially under high-shear conditions) and in stenosed In a recent study in baboons,13 we showed that inhibition of arteries and microvessels, where it initiates the formation of platelet adhesion, through prevention of the binding of vWF platelet aggregates. When vWF binds to GP Ib, platelets are to GP Ib by monoclonal antibody 6B4-Fab fragments, mark- slowed down, allowing direct platelet–collagen receptor in- edly inhibited thrombosis under relatively low shear condi- teractions via, for example, integrin 21 and glycoprotein tions, without causing thrombocytopenia or major lengthen- VI, which activate platelets, finally resulting in a conforma- ing of the bleeding time. 6B4 is a potent GP Ib blocker, tional change in the GP IIb/IIIa receptor to enable binding to which, by interacting with GP Ib (amino acids 201 to 268),14 fibrinogen and vWF, leading to the formation of platelet prevents vWF binding induced by either ristocetin, botroce- aggregates.1,2 After extensive vessel injury or rupture of tin, or shear. In the present study, we investigated the effect atherosclerotic plaques with subsequent exposure of throm- of inhibition of platelet adhesion in a modified Folts model,15 bogenic surfaces, platelet aggregation can progress to result in which the cyclic flow reductions are caused by platelet- in complete thrombotic occlusion of the injured vessel.3,4 dependent thrombi that form under high-shear conditions at Various clinical studies and studies in experimental ani- injured stenosed sites of an artery. The model has been mals have clearly shown that inhibition of platelet aggrega- described to represent some of the events that occur in tion, through prevention of the binding of the adhesive patients with unstable angina and is widely accepted to be proteins to GP IIb/IIIa, is an effective approach to prevent effective and clinically relevant for testing potential anti- thrombosis.5–8 Unfortunately, this approach increases the risk thrombotic agents.16,17 of bleeding, especially at the doses that are effective in The aims of the study were 2-fold. First, we evaluated the preventing thrombotic episodes.8 antithrombotic efficacy of 6B4-Fab in high-shear conditions Received June 25, 2001; revision accepted October 3, 2001. From the Laboratory for Thrombosis Research (D.W., H.D., N.C.), Interdisciplinary Research Center, KU Leuven Campus Kortrijk, Kortrijk, Belgium, and the Department of Haematology and Cell Biology (M.M., H.F.K.), University of the Orange Free State, Bloemfontein, South Africa. Correspondence to Prof Dr Hans Deckmyn, Laboratory for Thrombosis Research-IRC, K U Leuven Campus Kortrijk, E. Sabbelaan 53, B-8500 Kortrijk, Belgium. E-mail Hans.Deckmyn@kulak.ac.be © 2002 American Heart Association, Inc. Arterioscler Thromb Vasc Biol. is available at http://www.atvbaha.org 323 81 Downloaded from http://ahajournals.org by on October 24, 2018 82 324 Arterioscler Thromb Vasc Biol. February 2002 in baboons and compared it with the efficacy of the GP measurements (platelet count, coagulation, platelet aggregation, IIb/IIIa blocking monoclonal antibody MA-16N7C27,18 under receptor occupation, and plasma levels) were drawn at 60, 0, 30, the same conditions. The second aim was to determine 60, 150, and 300 minutes and 24 hours after treatment. whether the combination of an anti–GP Ib and anti–GP Baboons were divided into 6 groups, and drug treatment was as follows: (1) vehicle (saline), n2; (2) 6B4-Fab at 0.6 mg/kg, n5; IIb/IIIa treatment could result in a synergistic antithrombotic (3) 6B4-Fab at 2.0 mg/kg, n3; (4) MA-16N7C2 at 0.1 mg/kg, n3; effect and a lowering of the bleeding time prolongation (5) MA-16N7C2 at 0.3 mg/kg, n2; and (6) 6B4 at 0.6 mg/kg plus associated with the anti–GP IIb/IIIa agent. MA-16N7C2 at 0.1 mg/kg, n2. All agents were diluted in saline for injection. In addition, 4 mg/kg 6B4-Fab was injected in 1 baboon to Methods determine the effect of such a high dose on platelet count, receptor occupation, bleeding time, and platelet aggregation. CFRs were not Preparation of Monoclonal Antibodies 6B4-Fab measured in this baboon. and MA-16N7C2 Antibodies 6B4, raised against human GP Ib,13,19 and MA-16N7C2, Platelet Count, Coagulation, and Bleeding Time directed against GP IIb/IIIa,7,18 were purified from ascitic fluid from All blood samples were collected in 0.32% (f.c.) trisodium citrate. BALB/c mice on protein A Sepharose. 6B4-Fab was prepared from The platelet count was determined by using a Technicon H2 blood the IgG by papain digestion, as previously described.13 cell analyzer (Bayer Diagnostics). Prothrombin time and activated partial thromboplastin time were measured at 37°C by use of a In Vitro Flow Studies coagulometer (Clotex II, Hyland). The effects of 6B4-Fab and MA-16N7C2 on platelet adhesion to The template bleeding time was measured at the surface of the collagen was studied in a Sakariassen-type parallel-plate flow forearm by using a Simplate II device (Organon Teknika). The volar chamber at a shear rate of 1500 s1, as previously described.19 Blood surface of the forearm was shaved, and a pressure cuff was applied was collected from healthy volunteers by using LMW-heparin (25 and inflated to 40 mm Hg. The time elapsed until the visual cessation U/mL, Clexane) as an anticoagulant. Glass coverslips were coated of blood loss was recorded as the bleeding time. The wound was with human collagen type I (Sigma Chemical Co), dissolved in carefully dabbed every 15 seconds with filter paper. 50 mmol/L acetic acid (1 mg/mL), and dialyzed against PBS for 48 hours. Fifteen milliliters of blood, preincubated with vehicle or compound at 37°C for 5 minutes, was perfused in the flow chamber Platelet Aggregation for 5 minutes at 37°C. The platelets were fixed with methanol and Ex vivo platelet aggregation studies were performed on citrated stained with May-Grünwald–Giemsa. Platelet surface coverage was blood with an aggregometer (Elvi 840, Pabisch). Platelet-rich plasma analyzed by using an image analyzer. An average of 10 fields per and platelet-poor plasma were prepared by differential centrifuga- coverslip were analyzed. The results were expressed as percentage of tion. The platelet count in platelet-rich plasma was adjusted to total surface covered with platelets. 200 000 platelets per microliter with autologous platelet-poor plasma. Platelet aggregation was induced by 1.5 mg/mL ristocetin Folts Model in Baboons (ABP), 20 mol/L ADP (Sigma), or 50 g/mL collagen (Nycomed Nineteen baboons (Papio ursinus) of either sex, weighing 12 to 18 Arzneimittel GmbH) and measured by the change in light kg, were used. All procedures were approved by the Ethics Com- transmission. mittee for Animal Experimentation of the University of the Orange Free State in accordance with the National Code for Animal Use in Plasma Concentration Research, Education, Diagnosis and Testing of Drug and Related Plasma levels of antibodies were measured by ELISA with the use of Substances in South Africa. plates coated with 5 g/mL goat anti-mouse IgG (Sigma) and The experimental procedure was modified from the one described developed with peroxidase-conjugated goat anti-mouse Fab-specific by Folts.15 Baboons were anesthetized with ketamine hydrochloride IgG (Sigma). The plasma concentration of the antibodies in each (10 mg/kg IM), intubated with a cuffed endotracheal tube, and sample was calculated from standard curves constructed by adding ventilated by a respirator with oxygen supplemented with 0.5% known amounts of these antibodies to baboon plasma. Fluothane to maintain anesthesia. Body temperature was maintained at 37°C with a heating table. A catheter was placed in a femoral vein for drug administration and blood sampling. A segment of another Receptor Occupancy 125 femoral artery was gently dissected free from surrounding tissue, and Ex vivo binding of I-labeled 6B4-Fab or MA-16N7C2 IgG was a perivascular ultrasonic flow probe (Transonic Systems Inc) was used to determine the number of receptors occupied by 6B4-Fab and placed around the distal dissection site. The mean and phasic blood MA-16N7C2 IgG, respectively. Labeling of the antibodies with flow were recorded continuously throughout the experiment. Ba- sodium 125I (Amersham) was performed by using Iodogen (Pierce boons were allowed to stabilize for 30 minutes. The proximal Chemical Co). Platelet-rich plasma (200 000 platelets per microliter) dissection site of the femoral artery was injured by applying 3 was incubated with a near saturating dose (2 g/mL) of 125I-MA- occlusions of the artery for 10 seconds at 2-mm intervals with the use 16N7C2 or 125I-6B4-Fab for 15 minutes at room temperature. A of a spring-loaded forceps. A spring-loaded clamp was placed in the fraction of this mixture was then layered onto 20% sucrose buffer middle of the injured site to produce an external stenosis of 65% to (wt/vol) containing 0.1% BSA in Eppendorf tubes and centrifuged 80%. A gradual decline in blood flow due to platelet adhesion and for 4 minutes at 10 000g. Platelet pellet–associated radioactivity was aggregation was observed. When flow reached zero, blood flow was determined, and the number of radiolabeled molecules bound per restored by pushing the spring of the clamp to mechanically dislodge platelet was calculated. The results were calculated by subtracting the platelet-rich thrombus. This repetitive pattern of decreasing the number of labeled molecules bound to the platelets after blood flow after mechanical restoration is referred to as cyclic flow treatment from that obtained before treatment and are expressed as reductions (CFRs). Additional endothelial injury and appropriate external stenosis selections were repeated if needed. Finally, stable the percentage of receptors occupied. CFRs were obtained in 17 of 18 baboons. After a 60-minute control period, test agents were given via an Statistical Analysis intravenous bolus injection, and monitoring was continued for up to Data are expressed as meanSEM. The Student t test (2-tailed) or 60 minutes after drug administration. The antithrombotic effect was 1-factor ANOVA followed by the Fisher test was used for statistical quantified by comparing the number of CFRs per hour before and evaluation. A value of P0.05 was considered to be statistically after drug administration. Blood samples for the different laboratory significant. 82 Downloaded from http://ahajournals.org by on October 24, 2018 83 Wu et al GP Ib and IIb/IIIa Blockade Averts CFRs in Baboons 325 Figure 1. Inhibition by 6B4-Fab and MA-16N7C2 of platelet- dependent surface coverage of collagen-coated coverslips as determined after 5-minute perfusion of whole blood at 1500 s1. A and E, PBS control. B, 6B4-Fab (1.5 g/mL). C, MA-16N7C2 Fig 2. Effect of 6B4-Fab and MA-16N7C2 administration on (0.5 g/mL). D, 6B4-Fab (1.5 g/mL) plus MA-16N7C2 (0.5 template bleeding time in baboons. Open square indicates 0.6 g/mL). F, 6B4-Fab (2.25 g/mL). G, MA-16N7C2 (0.75 g/mL). mg/kg 6B4; solid square, 2.0 mg/kg 6B4; solid circle, 0.1 mg/kg H, 6B4 Fab (2.25 g/mL) plus MA-16N7C2 (0.75 g/mL). Data MA-16N7C2; open circle, 0.3 mg/kg MA-16N7C2; and open tri- are given as meanSEM (n4). *P0.05 vs PBS control. angle, 0.6 mg/kg 6B40.1 mg/kg MA-16N7C2. Data represent meanSEM. *P0.05 vs pretreatment value. Results combination of 0.6 mg/kg 6B4-Fab and 0.1 mg/kg MA- In Vitro Flow Studies in the Flow Chamber 16N7C2 completely blocked the CFRs. 6B4-Fab completely inhibits platelet adhesion to collagen in flow at 5 g/mL at 1300 s1 and 2.5 g/mL at 2600 s1.13 In Platelet Aggregation the present study, we analyzed the combined effects of The effects of 6B4-Fab and MA-16N7C2 on ex vivo platelet partially inhibitory concentrations of 6B4-Fab and MA- aggregation are shown in Figure 4. 6B4-Fab inhibited the ex 16N7C2 on human platelet adhesion to human collagen type vivo ristocetin-induced platelet aggregation in a dose- and I in vitro after 5 minutes of perfusion at 1500 s1 in a flow time-dependent manner. Aggregation was totally absent 30 chamber. Concentrations of 1.5 and 2.25 g/mL 6B4-Fab minutes after injection of 0.6 mg/kg 6B4 Fab or injection of inhibited platelet adhesion by 38% and 53%, respectively, the combination of 6B4-Fab and MA-16N7C2. Significant whereas 0.5 and 0.75 g/mL MA-16N7C2 inhibited the inhibitory effects at these doses persisted for 150 minutes surface coverage by platelets by 7% and 44%, respectively, and returned to normal values within 24 hours. mainly by reducing platelet aggregate formation. A combi- ADP- and collagen-induced platelet aggregation was sig- nation of 1.5 g/mL 6B4-Fab and 0.5 g/mL MA-16N7C2 nificantly inhibited after the administration of 0.1 or 0.3 inhibited surface coverage by 76%, whereas 88% inhibition mg/kg MA-16N7C2 and also when 0.6 mg/kg 6B4-Fab and was achieved when 2.25 g/mL 6B4-Fab and 0.75 g/mL 0.1 mg/kg MA-16N7C2 were combined. It is of interest that  the combination of 0.6 mg/kg 6B4-Fab fragments and 0.1 MA-16N7C2 were given in combination (Figure 1). mg/kg MA-16N7C2 had a more pronounced effect on the Platelet Count, Coagulation, and Bleeding Time inhibition of ADP-induced platelet aggregation than did a The platelet count was not significantly affected by injection single dose of 0.1 mg/kg MA-16N7C2, the reason for which of up to 4 mg/kg 6B4-Fab or MA-16N7C2 (in agreement with is unclear at present. previous observations)7,13 or the combination of 6B4 Fab and MA-16N7C2. No significant changes of prothrombin time and activated partial thromboplastin time were observed in any of the groups. The bleeding time (baseline 2.040.59 minutes, interani- mal coefficient of variation 29%, n13) was prolonged only after the injection of 0.3 mg/kg MA-16N7C2, but it returned to normal levels 24 hours later (Figure 2). Antithrombotic Effect The frequency of the CFRs (Figure 3) was not significantly different between the different groups before the treatments were started, nor was it changed by the injection of saline (1077%). A dose of 0.6 mg/kg 6B4-Fab resulted in a significant reduction of the CFRs by 5915%. A dose of 2 mg/kg abolished the CFRs, which could not be restored by Figure 3. Effects of different doses of 6B4-Fab, MA-16N7C2, or increasing the intimal damage or by increasing the stenosis. their combination on CFRs in the baboon femoral artery. Saline injection appears as 0.0 mg/kg. The results are expressed as MA-16N7C2 significantly inhibited CFRs by 438% at a percentage of the pretreatment value. Data represent the dose of 0.1 mg/kg and abolished the CFRs at 0.3 mg/kg. The meanSEM. *P0.05, **P0.01, and ***P0.001. 83 Downloaded from http://ahajournals.org by on October 24, 2018 84 326 Arterioscler Thromb Vasc Biol. February 2002 decreased almost linearly, reaching baseline levels 24 hours after treatment. Thirty minutes after a bolus injection of 0.3 mg/kg MA-16N7C2, 71% of the GP IIb/IIIa receptors were occupied by the antibody. The decrease that followed also fit a linear function best, reaching a value of 40% at 24 hours after treatment. The GP IIb/IIIa receptor occupancy rates were 29% and 30% at 30 minutes after bolus injection of 0.1 mg/kg MA-16N7C2 and after the combination of 0.6 mg/kg 6B4-Fab and 0.1 mg/kg MA-16N7C2, respectively. The occupation remained almost at the same level 24 hours after treatment. The relationship between GP Ib receptor occupancy, ristocetin-induced agglutination, CFRs, and bleeding time sug- gests that occupancy of 20% and of 35% of receptors with 6B4-Fab results in 50% reduction of platelet agglutination and CFRs, respectively, whereas up to 80% occupancy does not result in a significantly prolonged bleeding time (Figure 6). Discussion The present study evaluated the antithrombotic effects of the Fab fragments of monoclonal anti–GP Ib antibody (6B4) and of the monoclonal anti–GP IIb/IIIa antibody (MA-16N7C2) in injured and stenosed baboon femoral arteries. Previous studies have shown that 6B4-Fab inhibited ristocetin- and botrocetin-induced vWF-dependent human platelet agglutination. 6B4-Fab inhibited human platelet ad- hesion to human collagen type I in a perfusion flow chamber with IC50 values of 3.5, 1.1, and 0.5 g/mL at shear rates of 650, 1300, and 2600 s1, respectively, indicating that this compound is more effective under high-shear conditions.13 In a baboon model with permanent arteriovenous shunt (700 to 1000 s1), 6B4-Fab significantly reduced platelet deposition onto collagen.13 The present study evaluated the antithrom- botic efficacy of 6B4-Fab under high-shear conditions by using a modified Folts model in baboons: 6B4-Fab signifi- cantly reduced CFRs at a dose of 0.6 mg/kg and completely abolished CFRs at 2.0 mg/kg without causing a marked prolongation of the bleeding time, even when a dose of 4 mg/kg 6B4-Fab was given, suggesting that 6B4-Fab is a powerful antithrombotic agent with, possibly, a low bleeding risk. Figure 4. Effects of 6B4-Fab, MA-16N7C2, or their combination Although administration of inhibitory antibodies against vWF on ex vivo ristocetin-induced (A), ADP-induced (B), or collagen- induced (C) platelet aggregation in baboons. Open square indi- resulted in bleeding times that were not prolonged or were only cates 0.6 mg/kg 6B4; solid square, 2.0 mg/kg 6B4; solid circle, minimally prolonged20,21 or in bleeding times that resulted in a 0.1 mg/kg MA-16N7C2; open circle, 0.3 mg/kg MA-16N7C2; nearly 7-fold prolongation,10 the minimal effect of blocking GP and open triangle, 0.6 mg/kg 6B40.1 mg/kg MA-16N7C2. Data represent meanSEM. Ib on the bleeding time as seen in the present study is correlated with previous findings obtained with antibodies13,22 or with a 11 Plasma Concentration and Receptor Occupation recombinant vWF-A1 domain, which resulted in a prolonga- The plasma concentrations of 6B4-Fab and MA-16N7C2 are tion of the bleeding time of, at most, 3-fold. Most likely, the shown in Figure 5. The plasma concentration of both anti- rather exceptional combination of an antiplatelet effect without bodies peaked at 30 minutes after the bolus injection and prolongation of the bleeding time, as observed in the present declined rapidly thereafter. The occupancy of GP Ib and GP study and in other settings, is due to the shear dependence of the IIb/IIIa receptors by 6B4-Fab and MA-16N7C2 also was vWF–GP Ib interaction, which would therefore play a lesser role maximal at 30 minutes; however, it remained elevated for an in the lower shear systems and might be more relevant for the extended period. GP Ib receptor occupancy was 26%, 69%, bleeding time measurements. The bleeding problems observed and 84% after bolus injections of 0.6, 2.0, or 4.0 mg/kg in patients with Bernard-Soulier syndrome may be due to either 6B4-Fab, respectively, and 26% after the combination of 0.6 the complete absence of GP Ib, in contrast to the maximum 80% mg/kg 6B4-Fab and 0.1 mg/kg MA-16N7C2. Occupation receptor occupancy that was obtained in the present study, or the 84 Downloaded from http://ahajournals.org by on October 24, 2018 85 Wu et al GP Ib and IIb/IIIa Blockade Averts CFRs in Baboons 327 Figure 5. Plasma levels of 6B4-Fab (A) and MA-16N7C2 (C) and of GP Ib (B) and GP IIb/IIIa (D) receptor occupancy by 6B4-Fab and MA-16N7C2, respectively, as function of the time after their administration to baboons. Open square indicates 0.6 mg/kg 6B4; solid square, 2.0 mg/kg 6B4; solid cir- cle, 0.1 mg/kg MA-16N7C2; open circle, 0.3 mg/kg MA-16N7C2; and, open triangle, 0.6 mg/kg 6B40.1 mg/kg MA-16N7C2. Data represent the meanSEM. combination of GP Ib deficiency, thrombocytopenia, and giant which is in accordance with the findings of van Zanten et al,25 platelets, which may be less readily arrested at the site of injury. who reported that lower available GP Ib numbers were On the other hand, we could not really demonstrate the needed to prevent platelet adhesion to collagen under flow- anticipated higher efficiency of 6B4-Fab in the present induced compared with ristocetin-induced aggregation. How- high-shear setting compared with the lower shear setting in a ever, the receptor occupancy needed to obtain an effect as previous experiment.13 This, to a large extent, may be due to observed in the present study may be somewhat underesti- the different thrombogenic stimulus that was presented: a mated in view of the relative low affinity of 6B4-Fab. collagen-rich glutaraldehyde-fixed bovine pericardium in an For purposes of comparison, we also used the Folts model extracorporeal arteriovenous shunt in the medium-shear ex- to investigate the effect of inhibition of GP IIb/IIIa with periments13 versus a physically damaged and stenosed fem- MA-16N7C2. A dose of 0.1 mg/kg MA-16N7C2 inhibited oral artery. Part of the explanation may be that in the CFR CFRs by 43%, and CFRs were abolished at a dose of 0.3 experiments, ultrahigh-shear levels might occur, which can mg/kg MA-16N7C2. Thus, inhibition of platelet adhesion and result in direct GP Ib-vWF–dependent platelet aggrega- inhibition of platelet aggregation are both viable avenues for tion23,24 and may be more resistant to inhibition. preventing arterial thrombosis. However, with the high dose 6B4-Fab caused a dose- and time-dependent inhibition of of MA-16N7C2, the bleeding time was significantly pro- ex vivo ristocetin-induced platelet aggregation. A lower longed. Therefore, inhibition of aggregation may increase the receptor occupancy was needed to inhibit ristocetin-induced risk of bleeding, in agreement with studies using other platelet agglutination than was needed to prevent CFRs, antiplatelet agents.26,27 The roles of GP Ib and of GP IIb/IIIa in the formation of a platelet-dependent thrombus are complementary. Indeed, we found that combined inhibition has a more pronounced effect on in vitro platelet accumulation onto collagen in a flow chamber and on the prevention of CFRs in the baboon. With this combination, in addition, a full in vivo antithrom- botic effect was obtained without prolongation of the bleed- ing time. In conclusion, the present study demonstrates that the inhibition of platelet GP Ib is a powerful intervention in the prevention of platelet thrombus formation in the injured and stenosed baboon femoral artery without prolongation of the bleeding time. The effect on bleeding time is definitely less during inhibition of GP Ib than during inhibition of GP Figure 6. Relationship between percent GP Ib receptor occu- IIb/IIIa. Accordingly, a GP Ib blocker, such as 6B4-Fab, pancy by 6B4-Fab and percent ristocetin-induced platelet seems to be promising for further development as a com- aggregation, percent CFRs, and bleeding time (BT). To con- struct the CFR curve, receptor occupancy 30 minutes after pound for the prevention of acute arterial thrombotic syn- administration was used. dromes. Moreover, we demonstrated that combining a low 85 Downloaded from http://ahajournals.org by on October 24, 2018 86 328 Arterioscler Thromb Vasc Biol. February 2002 dose of a GP Ib inhibitor with a low dose of a GP IIb/IIIa 13. Cauwenberghs N, Meiring M, Vauterin S, van Wyk V, Lamprecht S, inhibitor has a potent antithrombotic action, again with Roodt JP, Novak L, Harsfalvi J, Deckmyn H, Kotze HF. Antithrombotic minimal effects on the bleeding time. This provides new effect of platelet glycoprotein Ib-blocking monoclonal antibody Fab fragments in nonhuman primates. Arterioscler Thromb Vasc Biol. 2000; information that will be beneficial in designing clinical 20:1347–1353. therapeutic approaches. 14. Cauwenberghs N, Vauterin S, Vanhoorelbeke K, Westra DF, Romo G, Huizinga EG, Lopez JA, Berndt MC, Harsfalvi J, Deckmyn H. Epitope Acknowledgments mapping of inhibitory antibodies against platelet glycoprotein Ib reveals interaction between the leucine-rich repeat N-terminal and C-terminal This work was supported by a EU-Biomed grant (PC 96.3517) and flanking domains of glycoprotein Ib. Blood. 2001;98:652–660. a Bilateral Collaboration grant between the Flemish and South 15. Folts J. An in vivo model of experimental arterial stenosis, intimal African Government (BIL 98/64). D.W. was a Junior Fellow of the damage, and periodic thrombosis. Circulation. 1991;83(suppl IV):IV- KU Leuven. We thank Seb Lamprecht, Jan P. Roodt, Stephan 3–IV-14. Vauterin, and Griet Vandecasteele for excellent technical assistance 16. Ikeda H, Koga Y, Kuwano K, Nakayama H, Ueno T, Yoshida N, Adachi and Dr Jef Arnout for help with the statistical analysis. K, Park IS, Toshima H. Cyclic flow variations in a conscious dog model of coronary artery stenoses and endothelial injury correlate with acute References ischemic heart disease syndromes in humans. J Am Coll Cardiol. 1993; 1. Andrews RK, Shen Y, Gardiner EE, Dong JF, Lopez JA, Berndt MC. The 21:1008–1017. glycoprotein Ib-IX-V complex in platelet adhesion and signaling. Thromb 17. Willerson JT, Yao SK, McNatt J, Benedict CR, Anderson HV, Golino P, Haemost. 1999;82:357–364. Murphree SS, Buja LM. Frequency and severity of cyclic flow alterations 2. Cauwenberghs N, Vanhoorelbeke K, Vauterin S, Deckmyn H. Structural and platelet aggregation predict the severity of neointimal proliferation determinants within glycoprotein Ibalpha involved in its binding to von following experimental coronary stenosis and endothelial injury. Proc Willebrand factor. Platelets. 2000;11:373–378. Natl Acad Sci U S A. 1991;88:10624–10628. 3. Maseri A, Chierchia S, Davies G. Pathophysiology of coronary occlusion 18. Deckmyn H, Stanssens P, Hoet B, Declerck PJ, Lauwereys M, Gansemans Y, in acute infarction. Circulation. 1986;73:233–239. Tornai I, Vermylen J. An echistatin-like Arg-Gly-Asp (RGD)-containing 4. Haerem JW. Platelet aggregates in intramyocardial vessels of patients sequence in the heavy chain CDR3 of a murine monoclonal antibody that dying suddenly and unexpectedly from coronary artery disease. Athero- sclerosis. 1972;15:199–213. inhibits human platelet glycoprotein IIb/IIIa function. Br J Haematol. 1994; 5. Coller BS, Folts JD, Scudder LE, Smith SR. Antithrombotic effect of a 87:562–571. monoclonal antibody to the platelet glycoprotein IIb/IIIa receptor in an 19. Cauwenberghs N, Ajzenberg N, Vauterin S, Hoylaerts MF, Declerck PJ, experimental animal model. Blood. 1986;68:783–786. Baruch D, Deckmyn H. Characterization of murine anti-GPIb monoclonal 6. Gold HK, Coller BS, Yasuda T, Saito T, Fallon JT, Guerrero JL, antibodies that differentiate between shear-induced and risto/botrocetin- Leinbach RC, Ziskind AA, Collen D. Rapid and sustained coronary artery induced GPIb-vWF interaction. Haemostasis. 2000;30:139–148. recanalization with combined bolus injection of recombinant tissue-type 20. Kageyama S, Yamamoto H, Nagano M, Arisaka H, Kayahara T, plasminogen activator and monoclonal antiplatelet GPIIb/IIIa antibody in Yoshimoto R. Anti-thrombotic effects and bleeding risk of AJvW-2, a a canine preparation. Circulation. 1988;77:670–677. monoclonal antibody against human von Willebrand factor. Br J 7. Kotze HF, Badenhorst PN, Lamprecht S, Meiring M, Van Wyk V, Nuyts Pharmacol. 1997;122:165–171. K, Stassen JM, Vermylen J, Deckmyn H. Prolonged inhibition of acute 21. Kageyama S, Yamamoto H, Nakazawa H, Yoshimoto R. Anti-human arterial thrombosis by high dosing of a monoclonal anti-platelet glyco- vWF monoclonal antibody, AJvW-2 Fab, inhibits repetitive coronary protein IIb/IIIa antibody in a baboon model. Thromb Haemost. 1995;74: artery thrombosis without bleeding time prolongation in dogs. Thromb 751–757. Res. 2001;101:395–404. 8. Adgey AA. An overview of the results of clinical trials with glycoprotein 22. Becker BH, Miller JL. Effects of an antiplatelet glycoprotein Ib antibody IIb/IIIa inhibitors. Eur Heart J. 1998;19(suppl D):D10–D21. on hemostatic function in the guinea pig. Blood. 1989;74:690–694. 9. Sakariassen KS, Bolhuis PA, Sixma JJ. Human blood platelet adhesion to 23. Depraetere H, Ajzenberg N, Girma JP, Lacombe C, Meyer D, Deckmyn artery subendothelium is mediated by factor VIII-von Willebrand factor H, Baruch D. Platelet aggregation induced by a monoclonal antibody to bound to the subendothelium. Nature. 1979;279:635–638. the A1 domain of von Willebrand factor. Blood. 1998;91:3792–3799. 10. Cadroy Y, Hanson SR, Kelly AB, Marzec UM, Evatt BL, Kunicki TJ, 24. Kroll MH, Hellums JD, McIntire LV, Schaefer AI, Moake JL. Platelets Montgomery RR, Harker LA. Relative antithrombotic effects of mono- and shear stress. Blood. 1996;88:1525–1541. clonal antibodies targeting different platelet glycoprotein-adhesive molecule interactions in nonhuman primates. Blood. 1994;83:3218–3224. 25. van Zanten GH, Heijnen HF, Wu Y, Schut-Hese KM, Slootweg PJ, de 11. McGhie AI, McNatt J, Ezov N, Cui K, Mower LK, Hagay Y, Buja LM, Groot PG, Sixma JJ, Nieuwland R. A fifty percent reduction of platelet Garfinkel LI, Gorecki M, Willerson JT. Abolition of cyclic flow vari- surface glycoprotein Ib does not affect platelet adhesion under flow ations in stenosed, endothelium-injured coronary arteries in nonhuman conditions. Blood. 1998;91:2353–2359. primates with a peptide fragment (VCL) derived from human plasma von 26. Verstraete M. Synthetic inhibitors of platelet glycoprotein IIb/IIIa in Willebrand factor–glycoprotein Ib binding domain. Circulation. 1994;90: clinical development. Circulation. 2000;101:E76–E80. 2976–2981. 27. Tamai Y, Takami H, Nakahata R, Ono F, Munakata A. Comparison of the 12. Yamamoto H, Vreys I, Stassen JM, Yoshimoto R, Vermylen J, Hoylaerts effects of acetylsalicylic acid, ticlopidine and cilostazol on primary hemo- MF. Antagonism of vWF inhibits both injury induced arterial and venous stasis using a quantitative bleeding time test apparatus. Haemostasis. thrombosis in the hamster. Thromb Haemost. 1998;79:202–210. 1999;29:269–276. 86 Downloaded from http://ahajournals.org by on October 24, 2018 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 87 S-6 HEMOSTASIS, THROMBOSIS, AND VASCULAR BIOLOGY Inhibition of the von Willebrand (VWF)–collagen interaction by an antihuman VWF monoclonal antibody results in abolition of in vivo arterial platelet thrombus formation in baboons Dongmei Wu, Karen Vanhoorelbeke, Nancy Cauwenberghs, Muriel Meiring, Hilde Depraetere, Harry F. Kotze, and Hans Deckmyn The interaction between collagen, von tions (CFRs) were measured in the femo- VWF-collagen binding was observed. Willebrand factor (VWF), and glycopro- ral artery. Administering a dose of 100, 82D6A3 has a high affinity for VWF; after tein Ib is the first step in hemostasis and 300, and 600 g/kg resulted in a 58.3%, 48 hours still 68% VWF (300g/kg) was thrombosis especially under high shear 100%, and 100% reduction in the CFRs, occupied with a pharmacologic effect up conditions. We studied the inhibition of respectively. When 100 g/kg 82D6A3 was to 5 hours after administration (80%- the VWF-collagen interaction by using an infused into the baboons, 80% of VWF-A3 100% occupancy). In conclusion, these antihuman VWF monoclonal antibody domain was occupied, corresponding to results clearly indicate that the VWF- 82D6A3 to prevent arterial thrombosis in 30% to 36% ex vivo inhibition of VWF collagen interaction is important in vivo baboons to develop a new kind of anti- binding to collagen, with no prolongation in thrombosis under high shear condi- thrombotic strategy and determine for the of the bleeding time. The bleeding time tions and thus might be a new target for first time experimental in vivo data con- was also not significantly prolonged when preventing arterial thrombosis. (Blood. cerning the importance of the collagen- the CFRs were abolished at doses of 300 2002;99:3623-3628) VWF interaction. We used a modified Folts g/kg and 600 g/kg. At these doses model to study the antithrombotic effi- 100% of VWF was occupied by the anti- cacy of 82D6A3, where cyclic flow reduc- body and 100% ex vivo inhibition of the © 2002 by The American Society of Hematology Introduction In normal hemostasis and thrombosis, platelets adhere to tshneake venom proteins echicetin and crota9,l1i0na, n antiguinea pig subendothelium of damaged blood vessels through an interactGioPnIb antibody1,1 a recombinant A1 domain fragment (VC1L2)a, nd with von Willebrand factor (VWF), which forms a bridge betweenrecently an antihuman GPIb antibo1d3yor compounds that bind to collagen within the damaged vessel wall and the platelet recepVtoWr F, such as anti-A1-VWF-monoclonal antibodies (mA1b4,s15)or glycoprotein Ib (GPIb/V/IX), an interaction especially importantaurin tricarboxylic acid (ATA)1,6 indeed, inhibit thrombus forma- under high shear condition1sT. his reversible adhesion allows tion in vivo. platelets to roll over the damaged area, which is then followed by a Specific blockade of the VWF-collagen interaction in vivo has firm adhesion through the collagen receptors (GPIa-IIa, GPVnI,ot yet been demonstrated but based on in vitro data could be a GPIV, p65, TIIICBP)2-4 resulting in platelet activation. This leadsnovel strategy for the prevention of thrombus formation in stenosed to the conformational activation of the platelet GPIIb/IIIa receptora,rteries. We here describe for the first time the antithrombotic effect fibrinogen binding, and finally to platelet aggregat5ion. of a murine antihuman VWF mAb 82D6A3, known to bind to the The VWF subunit is composed of several homologous domainAs3 domain and to inhibit VWF binding to fibrillar collagens type I and III but not to collagen V1I.7 each covering different functions; VWF interacts through its A1 The antithrombotic efficacy of 82D6A3 was demonstrated in domain mainly with the GPIb/V/IX comple6x,whereas its A3 baboons by using a modified Folts model, where cyclic flow reductions domain predominantly interacts with fibrillar collagen fibe7 rs.(CFRs) due to thrombus formation and its dislodgmenmt aereasured in Under normal conditions platelets and VWF do not interacmt. echanically damaged, severely stenosed femoral art1e8ries. However, when VWF is bound to collagen at high shear rate, it is believed to undergo a conformational change allowing its binding with the platelet receptor GPIb/IX/8V. Materials and methods One line of search for antiplatelet drugs in the prevention of thrombosis is focusing on the inhibition of the VWF-GPIb axisP. urification of 82D6A3 Compounds that interact with GPIb, such as the GPIb-bindin8g2D6A3 was raised and purified from ascites by proteinAchromatogr1a7phy. From the Laboratory for Thrombosis Research, Interdisciplinary Research Reprints: Hans Deckmyn, Laboratory for Thrombosis Research-IRC, K U Center, K U Leuven Campus Kortrijk, Kortrijk, Belgium; and the Department of Leuven Campus Kortrijk, E. Sabbelaan 53, B-8500 Kortrijk, Belgium; e-mail: Haematology and Cell Biology, University of the Orange Free State, hans.deckmyn@kulak.ac.be. Bloemfontein, South Africa. The publication costs of this article were defrayed in part by page charge Submitted October 11, 2001; accepted January 8, 2002. payment. Therefore, and solely to indicate this fact, this article is hereby Supported by a EU-Biomed grant (PC 96.3517) and a Bilateral Collaboration grant marked ‘‘advertisement’’ in accordance with 18 U.S.C. section 1734. between the Flemish and SouthAfrican Government (BIL 98/64). D.W. was a Junior Fellow of the KU Leuven. D.W. and K.V. contributed equally to this work. © 2002 by The American Society of Hematology BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 3623 87 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 88 3624 WU et al BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 Surgical preparation Plates were blocked with 3% milk powder for 2 hours at room temperature (RT). Next dilution series (1:2 in PBS) of the plasma samples (prewarmed Baboons (Papio ursinus) used in this study were of either sex and weighed for 5 minutes at 37°C) were added for 2 hours at RT and goat antimouse IgG 12 to 18 kg. All procedures were approved by the Ethics Committee for labeled with horseradish peroxidase (HRP) was added for 1 hour at RT. Animal Experimentation of the University of the Free State and Free State Visualization was obtained with H2O2 and ortho-phenylenediamine (OPD, Provincial Administration in accordance with the National Code for Animal Sigma) and the coloring reaction was stopped with 4 moL/L H2SO4. The Use in Research, Education, Diagnosis and Testing of Drug and Related absorbance was determined at 490 nm. After each incubation step, plates Substances in South Africa. The experimental procedure followed19 was were washed with PBS, 0.1% Tween-20, 3 times after coating and blocking that of the original Folts model18 except that the femoral artery was injured steps and 12 times elsewhere. The plasma concentration of 82D6A3 in each with a spring-loaded forceps. Baboons were anesthetized with ketamine sample was calculated from a standard curve where known amounts of hydrochloride (10 mg/kg, intramuscularly), intubated with a cuffed endotra- 82D6A3 were added to baboon plasma. cheal tube, and ventilated by a respirator with oxygen supplemented with 0.5% Fluothane (halothane) to maintain anesthesia. Body temperature was maintained at 37°C with a heating table. A catheter was placed in a femoral VWF-antigen levels vein for drug administration and blood sampling. A segment of a femoral Determination of the VWF-antigen (Ag) levels was performed essentially artery was gently dissected free from surrounding tissue and a perivascular as described.20 Briefly, microtiter plates were coated with a polyclonal ultrasonic flow probe (Transonic Systems, Ithaca, NY) was placed around anti-VWF-Ig solution (Dako, Glostrup, Denmark), blocked with 3% milk the distal dissection site. The mean and phasic blood flow were recorded powder and plasma samples, preincubated for 5 minutes at 37°C, and added continuously throughout the experiment. Baboons were allowed to stabilize to the wells at 1:40 to 1:2560 dilutions. Bound VWF was detected with for 30 minutes. The proximal dissection site of the femoral artery was rabbit antihuman VWF HRP antibodies (Dako). VWF-Ag levels were injured by applying 3 occlusions of the artery for 10 seconds with 2-mm calculated from a standard curve obtained by adding 1:40 to 1:2560 interval using a spring-loaded forceps, and a spring-loaded clamp was dilutions to the coated wells of a human plasma pool, known to contain 10 placed in the middle of the injured site to produce an external stenosis of g/mL human VWF. 65% to 80%. A gradual decline in blood flow due to platelet adhesion and aggregation was observed. When flow was reduced by at least 50%, blood VWF occupancy flow was restored by pushing the spring of the clamp to mechanically dislodge the platelet-rich thrombus. This repetitive pattern of decreasing Microtiter plates (96-well) were coated overnight at 4°C with 125 L/well blood flow following mechanical restoration is referred to as cyclic flow of a polyclonal anti-VWF-Ig solution (Dako; 1:1000 in PBS). Plates were reductions (CFRs). Additional endothelial injury and appropriate external blocked with 3% milk powder solution for 2 hours at RT. A dilution series stenosis selection was repeated if needed to finally obtain stable CFRs in (1:2 in PBS) of the plasma samples (prewarmed for 5 minutes at 37°C) was these baboons. The number of times the thrombus needed to be dislodged added for 2 hours at RT. Samples containing 100% occupied VWF were determines the number of CFRs. obtained by adding a saturating amount of 82D6A3 (6 g/mL) to each After a 60-minute control period of reproducible CFRs (t  60 corresponding baboon plasma. Bound 82D6A3 was detected by addition of minutes to 0 minute), test agents (saline or 82D6A3) were given via an goat antimouse IgG-HRP (1 hour at RT). Visualization and wash steps were intravenous bolus injection (t  0) and monitoring was continued up to 60 performed as described above. The VWF occupancy of each sample was minutes after drug administration (t  60 minutes). The antithrombotic calculated as follows: (A490-nm sample/A490-nm sample saturated with effect was quantified by comparing the number of CFRs per hour before and 82D6A3)  100. after drug administration. If the last flow reduction in the 60-minute recording period was not cyclic, it was considered in the calculations. Blood Determination of the VWF-collagen binding activity samples for the different laboratory measurements were drawn at different time points. The doses of 82D6A3 were selected on the basis of preliminary The enzyme-linked immunosorbent assay (ELISA) was performed essen- dose-finding studies. Group 1 was a control group where the baboons tially as described.20 Briefly, microtiter plates were coated with human received saline (n  2). Group 2 received a high dose of 600 g/kg collagen type I (Sigma), blocked with 3% milk powder solution and 1⁄2 82D6A3 (n  2). In group 3, the baboons received an initial dose of 100 dilution series of baboon plasma (prewarmed for 5 minutes at 37°C) were g/kg 82D6A3 and after 60 minutes an additional 200 g/kg 82D6A3 was added. Bound VWF was detected with rabbit antihuman VWF-HRP given (n  3). In a preliminary study we found that 82D6A3 has a long antibodies. Binding of baboon VWF to collagen in the different plasma half-life. This therefore resulted in a final dose of 300 g/kg. All agents samples was compared to the binding of VWF in the respective blood were diluted with saline. sample taken at time zero (presample), which was set as 100%. Platelet count, coagulation, and bleeding time Inhibition of the VWF binding to collagen by 82D6A3 All blood samples were collected on 0.32% (final concentration) trisodium The ELISA was performed as described for the determination of the VWF- citrate. The platelet count was determined using a Technicon H blood cell collagen binding activity except that serial dilutions of 82D6A3 were either2 analyzer (Bayer Diagnostics, Tarrytown, NY). Prothrombin time (PT) and preincubated with constant amounts of VWF in a preblocked plate for 30 activated partial thromboplastin time (aPTT) were measured at 37°C using minutes or were directly added to the plate containing collagen-bound VWF. a coagulometer (Clotex II, Hyland, South Africa). The template bleeding time was measured using the Simplate II device (Organon Teknika, Statistics Durham, NC). The volar surface of the forearm was shaved and a pressure Data are expressed as mean  SD. Student t test (paired) or one-factor cuff was applied and inflated at 40 mm Hg. The wound was carefully ANOVA followed by Fisher test was used for statistical evaluation. P  .05 dabbed every 15 seconds with filter paper. Baseline bleeding time measured was considered as statistically significant. in 13 baboons was 2.04  0.59 minutes, with an interanimal coefficient of variation of 29%. All measurements were performed once at each time point. Results Plasma concentration of 82D6A3 In vitro interaction of 82D6A3 with baboon VWF Microtiter plates (96-well, Greiner, Frickenhausen, Germany) were coated overnight at 4 °C with 5 g/mL (in phosphate-buffered saline [PBS], 100 82D6A3, known to inhibit human VWF binding to collagen,17 L/well) goat antimouse IgG whole molecule (Sigma, St Louis, MO). cross-reacts with baboon VWF. 82D6A3 has a comparable affinity 88 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 89 BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 IN VIVO VON WILLEBRAND FACTOR-COLLAGEN INTERACTION 3625 Figure 1. Inhibition of baboon VWF binding to collagen by 82D6A3. Different concentrations of 82D6A3 (20 L) were incubated with undiluted baboon plasma (220 L) for 30 minutes before addition to a collagen-coated plate. Bound VWF was detected with rabbit antihuman VWF antibodies. The data are the mean of 4 measurements obtained with the plasma of 2 different baboons Figure 3. A representative tracing of CFR data. After a 60-minute control period, 100 g/kg 82D6A3 was administered to the baboon and changes in CFRs were for baboon VWF as for human VWF (median effective concentra- monitored for 60 minutes. tion [EC50] 200 ng/mL) and inhibited the binding of baboon VWF to collagen with an inhibitory concentration of 50% (IC50) of 3.5 g/mL (Figure 1). Coagulation, platelet count, and bleeding time 82D6A3 not only prevented binding of VWF to collagen, but also was able to remove the bound VWF from a collagen surface No significant changes in PT and aPTT were observed in any of the with somewhat lower efficacy (Figure 2). animals (data not shown). No major changes in platelet count were observed for the 100 g/kg (not significant [NS]), 300 g/kg (NS), Antithrombotic effect and 600 g/kg 82D6A3 doses (Tables 1 and 2). The bleeding time was not significantly prolonged after injection of 100 g/kg. After Using a Folts model, the antithrombotic efficacy of 82D6A3 was injection of 300 g/kg the bleeding time was increased 2.7 times at tested by administering different doses of 82D6A3 to baboons. In 60 minutes and 2.4 times at 150 minutes; however, these changes Figure 3, a representative tracing of CFR data is given. were not statistically significant. After injection of 600 g/kg the Injection of saline did not affect CFRs (107%  7%). In a first bleeding time was increased 1.9, 3.0, 2.8, and 1.7 times after 30, 60, study, a dose of 600 g/kg 82D6A3 resulted in a complete 150, and 300 minutes, respectively. disappearance of the CFRs (Figure 4). This dose was expected to be an overdose because according to theoretical calculations, a dose of 600 g/kg would result in a plasma concentration of 13 g/mL (45 82D6A3 plasma concentration, VWF-Ag levels, VWF mL/kg plasma volume), which in in vitro experiments results in a occupancy, and ex vivo VWF-collagen binding 100% inhibition of the VWF binding to collagen (Figure 1). As a The 82D6A3 plasma levels remained relatively constant during the result, in all subsequent in vivo studies lower doses of 82D6A3 first 3 hours of the experiment. Then, they decreased to 69%, 23%, were used. Administration of 100 g/kg 82D6A3 resulted in a and 7.6% after 300 minutes, 24 hours, and 48 hours, respectively, significant reduction of the CFRs by 58.3%  4.8% (Figure 4). when 300 g/kg 82D6A3 was administered (Table 1). Administration of a total doses of 300 g/kg completely abolished the CFRs (Figure 4), which could not be restored by increasing intimal damage or increasing stenosis. Figure 2. Inhibition of collagen-bound VWF by 82D6A3. Different concentrations Figure 4. Inhibition of femoral artery CFRs after administration of different of 82D6A3 were either preincubated with diluted human plasma (final dilution of 1:46, doses of 82D6A3 to baboons. Different doses of 82D6A3 were administered to ie, 217 ng/mL VWF) before addition to a collagen-coated plate (F) or were added to a baboons and the CFRs were measured during 60 minutes. Data represent the mean collagen-coated plate where VWF (final dilution of 1:46, ie, 217 ng/mL VWF) was with n  2 for 0 and 600 g/kg and mean  SD with n  3 for 100 and 300 g/kg mAb already bound for 30 minutes (E). Bound VWF was detected with rabbit antihuman 82D6A3. *  P  .05; †  P  .01. One-factor ANOVA followed by Fisher test was VWF antibodies. used for statistical evaluation. 89 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 90 3626 WU et al BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 Table 1. Ex vivo analysis of plasma samples after administration of 100 and 300 g/kg 82D6A3 to baboons Platelet count Bleeding time VWF-Ag levels 82D6A3 levels VWF occupancy Collagen binding Time (103/mm3) (min) (g/mL) (g/mL) (%) (%) 100 g/kg (n  3) 0 286  54 2.7  0.4 10.2  1.7 0 2.3  1.3 101  7 30 min 292  65 2.7  0.4 10.2  2.5 0.4  0.07† 80  10.8† 64  7* 60 min 289  49 3.5  2.1 8.9  1.4 0.4  0.1* 80  2.4‡ 69  9 300 g/kg (n  3) 0 286  54 2.7  0.4 10.2  1.7 0 2.3  1.3 101  7 30 min 265  41 4.6  0.6 8.8  1.4 2.9  0.3† 102  10.4† 4  1† 60 min 287  53 7.3  2.5 9.1  2.4 2.8  0.3† 99  10.6† 4  1† 150 min 309  83 6.4  3.1 9.7  2.7 2.6  0.1‡ 101  7.6† 4  1† 300 min 282  7 3.15  1.2 8.8  0.1 2.0  0.5 94  0.9† 4  1† 24 h 312  46 3.25  0.3 12.8  1.3 0.7  0.2* 74  31 91  18 48 h 306  79 3 13.2  0.8 0.2  0.01* 63  7.8* 93  0 Data are mean data  SD. At each time point, the plasma samples were measured 3 times in 3 different ELISAs for the 3 animal experiments. For statistical analysis, the mean data were used. Student t test (paired) was used for statistical evaluation. *P  .05. †P  .01. ‡P  .001. Thirty minutes after injection of the different doses of 82D6A3, plasma VWF-Ag levels decreased slightly and increased above Discussion baseline at 24 hours. None of these changes were significant (Tables 1 and 2). Platelet adhesion to a damaged vessel wall is the first step in arterial At 60 minutes after administration VWF occupancy was 80% thrombus formation. The tethering of platelets by VWF to the for the 100-g/kg dose and nearly 100% for the 300-g/kg and collagen exposed in the damaged vessel wall is especially impor- 600-g/kg doses. VWF remained occupied for an extended period; tant under high shear conditions. Antithrombotics that interfere even 48 hours after the injection of 300 g/kg 82D6A3, still 63% with the GPIb-VWF axis have been studied in animal models and of the VWF-binding sites were occupied by 82D6A3 (Table 1). were shown to be effective.13,14 The present study evaluated for the Injection of 100 g/kg 82D6A3 resulted in an ex vivo inhibition first time the antithrombotic effects of the inhibition of the of the VWF-collagen binding of 31% (blood sample taken after 1 VWF-collagen interaction in vivo. For this purpose, we used a hour; Table 1). At doses of 300 g/kg and 600 g/kg no interaction monoclonal antihuman VWF antibody 82D6A3, which by binding between baboon VWF and collagen was observed in samples taken to the VWF A3-domain, inhibits VWF binding to fibrillar collagens up to 5 hours after the administration of the antibody. Twenty-four types I and III.17 We here showed that 82D6A3 cross-reacts with hours after the injection of the drug the VWF-collagen interaction baboon VWF and inhibits baboon VWF binding to collagen type I recovers (Table 1 and 2). under static conditions. Moreover, in vitro 82D6A3 is able to Relation between the ex vivo 82D6A3 plasma levels, VWF remove collagen-bound VWF from its surface, which in view of the occupancy, and collagen binding presence of collagen-bound VWF in the damaged vessel wall, might be a prerequisite for a proper function in vivo. Because the The relationship between 82D6A3 plasma levels and VWF occu- subendothelium, however, not only includes different types of pancy is illustrated in Figure 5. At plasma concentrations between 1 collagen that may interact differently with VWF, and other matrix and 2 g/mL 82D6A3 nearly 100% of circulating VWF was proteins that may affect the binding of VWF, and because, in occupied (Figure 5). There was a good relation between the ex vivo addition, it may contain different VWF forms, such as ultralarge VWF occupancy and the VWF binding to collagen. To obtain forms unprocessed by the VWF-cleaving plasma protease, no inhibition of VWF binding to collagen (20%-40%), a VWF direct extrapolation is possible. occupancy of at least 80% was required, with complete inhibition A modified Folts model was used to evaluate the antithrombotic at 90% to 100% occupancy (Figure 6). These data were confirmed efficacy of 82D6A3 under high shear conditions18 in baboons. This by in vitro experiments, where different concentrations of 82D6A3 model allows study of CFRs due to platelet-dependent thrombi were added to baboon plasma (Figure 6). Occupancy levels of up to forming at the injured, stenotic site of the artery.21 We demonstrated 60% resulted in no inhibition of the VWF binding to collagen, that inhibition of the VWF-collagen interaction is accompanied by whereas up to complete inhibition was observed with 80% to 100% an antithrombotic effect in vivo. Indeed, in all 5 animals receiving VWF-occupancy levels. doses of 300 g/kg or higher, a 100% inhibition of the CFRs was Table 2. Ex vivo analysis of plasma samples after administration of 600 g/kg 82D6A3 to baboons Platelet count Bleeding time VWF-Ag levels 82D6A3 levels VWF occupancy Collagen binding (103/mm3) (min) (g/mL) (g/mL) (%) (%) 0 min 335 1.8 14  1.7 0 6.9  0.1 100  0 30 min 320 3.5 11.5  0.9 4.5  0.5 96  1 4  0.2 60 min 313 5.5 10.8  0.1 4.8  0.7 96  0.2 3.5  0.2 150 min 356 5 11.9  1.8 3.8  0.5 97  4 3.5  0.2 300 min 334 3 10.5 3.8  0.6 97 4 24 h 347 ND 22.9 1.4  0.01 88 45 Data are mean data  SD. At each time point, the plasma samples were measured 3 times in 3 different ELISAs for the 2 animal experiments. ND indicates not determined. 90 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 91 BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 IN VIVO VON WILLEBRAND FACTOR-COLLAGEN INTERACTION 3627 Figure 5. Relationship between the ex vivo VWF occupancy and 82D6A3 Figure 6. Relationship between the ex vivo and in vitro VWF binding to collagen plasma levels. All mean data measured at the different time points in the 3 different and the VWF occupancy. The ex vivo data are the mean data measured at the dosage studies were used (Tables 1 and 2). different time points in the 3 different dosage studies shown in Tables 1 and 2 (F) and the in vitro data were obtained by adding known concentrations of mAb 82D6A3 to undiluted baboon plasma and measuring VWF occupancy and collagen binding (E). obtained, whereas administration of 100 g/kg resulted in 58% Data are the mean of 2 determinations in duplicate. inhibition. Administration of the lowest 82D6A3 dose (100 g/kg) to the for 74%, the ratio VWF monomer/mAb is 8. So too many VWF baboons revealed that when 80% of the VWF-A3 domain sites subunits are devoid of mAb to have some effect in VWF-collagen were occupied (30%-36% inhibition of VWF binding to collagen), binding. It has to be noted, however, that at 100% VWF occupancy, 58% reduction in CFRs was observed with no prolongation of the the VWF monomer/mAb ratio (both in vitro and ex vivo) is 2.8, bleeding time. When a 100% decrease of CFRs was observed after which demonstrates that both in vitro and ex vivo not all VWF-A3 administration of 300 g/kg 82D6A3 (and also of 600 g/kg) the domains are accessible for 82D6A3 binding. bleeding time increased 2.7 times (60 minutes, 300 g/kg), which Although many in vitro studies demonstrated that the VWF- was not statistically significant. At these doses, 100% VWF was collagen interaction is vital for sustaining platelet adhesion under occupied with 82D6A3 and 100% ex vivo inhibition of the VWF high shear conditions, this study is the first experimental proof of binding to collagen was observed. It is interesting to note that when the in vivo importance of this interaction. It is of interest to note we studied the effect of 300 g/kg of an anti-GPIIb/IIIa mAb that for a very long time no bleeding disorders due to an isolated 16N7C2 in exactly the same animal model, a more than 15-fold defect in VWF binding to collagen were known. Only recently 2 increase in bleeding time was observed, which was statistically patients, mother and daughter, were identified with a moderate significant.19 Thus the therapeutic window of 82D6A3 seems to be bleeding disorder due to Ser968Thr mutation in the A3 domain of broader when compared to the one of an anti-GPIIb/IIIa blocker in VWF resulting in a defective binding of VWF to collagen.22 These our animal model.19 This suggests that bleeding problems might be recent results together with our data finally unequivocally demon- less expected when 82D6A3 is used as an antithrombotic agent. strate the in vivo relevance of the VWF-collagen interaction. The observation that no significant decreases in platelet count and In conclusion, blockade of VWF-collagen interaction by 82D6A3 VWF-Ag levels were observed furthermore pointed out that the reduced platelet thrombus formation in the injured and stenosed observed antithrombotic effect is indeed due to the specific baboon femoral arteries in a dose-dependent manner. In view of the inhibition of the VWF-collagen interaction. lack of effect on the bleeding time, it is worthwhile to develop this 82D6A3 has a high affinity for VWF (Kd 0.4 nmol17), which is approach further to determine its potential to treat acute arterial reflected here by the observation that VWF is still occupied by 63% thrombotic syndromes. with the antibody 48 hours after its administration. On the other hand, to be functionally active, at least up to 80% VWF occupancy is needed to have an ex vivo and in vitro inhibition of VWF binding Acknowledgments to collagen. This level was maintained for up to 5 hours after the administration of 300 and 600 g/kg 82D6A3. The observation We thank Miss Griet Vandecasteele, Mr Stephan Vauterin, Mr Seb that up to 80% VWF occupancy is needed to be functionally Lamprecht, and Mr Jan P. Roodt for their excellent technical relevant might be explained by the fact that when VWF is occupied assistance and Dr Jef Arnout for help with the statistical analysis. References 1. Sixma JJ, Wester J. The hemostatic plug. Semin 6. Berndt MC, Ward CM, Booth WJ, Castaldi PA, ture of human von Willebrand factor. Blood. 1996; Hematol. 1977;14:265-299. Mazurov AV, Andrews RK. Identification of aspar- 88:2939-2950. 2. Savage B, Saldivar E, Ruggeri ZM. Initiation of tic acid 514 through glutamic acid 542 as a glyco- 9. Peng M, Lu W, Beviglia L, Niewiarowski S, Kirby platelet adhesion by arrest onto fibrinogen or protein Ib-IX complex receptor recognition se- EP. Echicetin: a snake venom protein that inhibits translocation on von Willebrand factor. Cell. 1996; quence in von Willebrand factor. Mechanism of binding of von Willebrand factor and alboaggre- 84:289-297. modulation of von Willebrand factor by ristocetin gins to platelet glycoprotein Ib. Blood. 1993;81: 3. Kehrel B. Platelet-collagen interactions. Semin and botrocetin. Biochemistry. 1992;31:11144- 2321-2328. Thromb Hemost. 1995;21:123-129. 11151. 10. Chang MC, Lin HK, Peng HC, Huang TF. Anti- 7. Lankhof H, van Hoeij M, Schiphorst ME, et al. A3 thrombotic effect of crotalin, a platelet membrane4. Monnet E, Fauvel-Lafeve F. A new platelet recep- glycoprotein Ib antagonist from venom of Crota- tor specific to type III collagen. Type III collagen- domain is essential for interaction of von Wille- lus atrox. Blood. 1998;91:1582-1589. binding protein. J Biol Chem. 2000;275:10912- brand factor with collagen type III. Thromb Hae- most. 1996;75:950-958. 11. Miller JL, Thiam-Cisse M, Drouet LO. Reduction10917. in thrombus formation by PG-1 F(ab)2, an anti- 5. Phillips DR, Charo IF, Scarborough RM. GPIIb- 8. Siedlecki CA, Lestini BJ, Kottke-Marchant KK, guinea pig platelet glycoprotein Ib monoclonal IIIa: the responsive integrin. Cell. 1991;65:359- Eppell SJ, Wilson DL, Marchant RE. Shear-de- antibody. Arterioscler Thromb. 1991;11:1231- 362. pendent changes in the three-dimensional struc- 1236. 91 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 92 3628 WU et al BLOOD, 15 MAY 2002  VOLUME 99, NUMBER 10 12. McGhie AI, McNatt J, Ezov N, et al. Abolition of inhibits both injury induced arterial and venous antibodies prevents arterial thrombosis in ba- cyclic flow variations in stenosed, endothelium- thrombosis in the hamster. Thromb Haemost. boons. Arterioscler Thromb Vasc Biol. 2002;22: injured coronary arteries in nonhuman primates 1998;79:202-210. 323-328. with a peptide fragment (VCL) derived from hu- 16. Golino P, Ragni M, Cirillo P, et al. Aurintricarboxy- 20. Vanhoorelbeke K, Cauwenberghs N, Vauterin S, man plasma von Willebrand factor-glycoprotein Ib lic acid reduces platelet deposition in stenosed Schlammadinger A, Mazurier C, Deckmyn H. A binding domain. Circulation. 1994;90:2976-2981. and endothelially injured rabbit carotid arteries reliable and reproducible ELISA method to mea- 13. Cauwenberghs N, Meiring M, Vauterin S, et al. more effectively than other antiplatelet interven- sure ristocetin cofactor activity of von Willebrand Antithrombotic effect of platelet glycoprotein Ib- tions. Thromb Haemost. 1995;74:974-979. factor. Thromb Haemost. 2000;83:107-113. blocking monoclonal antibody Fab fragments in 17. Hoylaerts MF, Yamamoto H, Nuyts K, Vreys I, 21. Ikeda H, Koga Y, Kuwano K, et al. Cyclic flow nonhuman primates. Arterioscler Thromb Vasc Deckmyn H, Vermylen J. von Willebrand factor variations in a conscious dog model of coronary Biol. 2000;20:1347-1353. binds to native collagen VI primarily via its A1 do- artery stenosis and endothelial injury correlate 14. Cadroy Y, Hanson SR, Kelly AB, et al. Relative main. Biochem J. 1997;324:185-191. with acute ischemic heart disease syndromes in antithrombotic effects of monoclonal antibodies 18. Folts J. An in vivo model of experimental arterial humans. J Am Coll Cardiol. 1993;21:1008-1017. targeting different platelet glycoprotein-adhesive stenosis, intimal damage, and periodic thrombo- 22. Ribba AS, Loisel I, Lavergne JM, et al. Ser968Thr molecule interactions in nonhuman primates. sis. Circulation. 1991;83:IV3–14. mutation within the A3 domain of von Willebrand Blood. 1994;83:3218-3224. 19. Wu D, Meiring M, Kotze HF, Deckmyn H, Cau- factor (VWF) in two related patients leads to a 15. Yamamoto H, Vreys I, Stassen JM, Yoshimoto R, wenberghs N. Inhibition of platelet glycoprotein defective binding of VWF to collagen. Thromb Vermylen J, Hoylaerts MF. Antagonism of vWF Ib, glycoprotein IIb/IIIa, or both by monoclonal Haemost. 2001;86:848-854. 92 From www.bloodjournal.org by guest on October 22, 2018. For personal use only. 93 2002 99: 3623-3628 doi:10.1182/blood.V99.10.3623 Inhibition of the von Willebrand (VWF)−collagen interaction by an antihuman VWF monoclonal antibody results in abolition of in vivo arterial platelet thrombus formation in baboons Dongmei Wu, Karen Vanhoorelbeke, Nancy Cauwenberghs, Muriel Meiring, Hilde Depraetere, Harry F. Kotze and Hans Deckmyn Updated information and services can be found at: http://www.bloodjournal.org/content/99/10/3623.full.html Articles on similar topics can be found in the following Blood collections Hemostasis, Thrombosis, and Vascular Biology (2485 articles) Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved. 93 S-7 94 © 2007 Schattauer GmbH, Stuttgart Cardiovascular Biology and Cell Signalling Coronary artery in-stent stenosis persists despite inhibition of the vonWillebrand factor - collagen interaction in baboons Simon F. De Meyer1*, Stephanie Staelens1*, Philip N. Badenhorst2, Henry Pieters2, Seb Lamprecht2, Jan Roodt2, Stefan Janssens3, Muriel Meiring2, KarenVanhoorelbeke1, André Bruwer4, Stephen Brown4, Hans Deckmyn1 From the 1Laboratory for Thrombosis Research, KU Leuven Campus Kortrijk, Kortrijk, Belgium; 2Department of Haematology and Cell Biology, and 4Department of Paediatrics and Child Health, University of the Free State, Bloemfontein, South Africa; 3Department of Cardiology, KU Leuven, Leuven, Belgium Summary Revascularization techniques, such as angioplasty and stent im- bleeding time was observed. Our results show that the plantation, frequently lead to restenosis due to the formation of formation of neointima was not affected in the group of baboons neointima after platelet activation and the concomittant release where primary platelet adhesion was abolished with 82D6A3 of various smooth muscle cell mitogenic and attractant factors. when compared to the control group. Vascular injury scores We here investigate whether inhibition of initial platelet ad- were the same in both groups. Inhibition of platelet adhesion hesion after stent implantation can decrease neointima during the first three days after stenting, on top of standard formation in a clinically relevant baboon model of in-stent ste- treatment with aspirin,clopidogrel and heparin,had no effect on nosis using standard treatment with aspirin, clopidogrel and he- neo-intima formation in a baboon model of in-stent stenosis. parin. Inhibition of platelet adhesion was established by adminis- During the last decade,attempts to translate seemingly effective tration of the anti-von Willebrand factor (VWF) monoclonal therapies based on smaller animal experimentation to the clinic antibody 82D6A3, which inhibits VWF binding to collagen.Ad- have consistently failed.This study, using a non-human primate ministration of 82D6A3 resulted in a complete inhibition of model that more closely resembles the clinical situation, pres- VWF binding to collagen during the first three days after stent ents a model that may be of further clinical interest for studying implantation. No thrombocytopenia or prolongation of the the prevention of restenosis. Keywords vonWillebrand factor,stenosis,baboon model,stent,neointima Thromb Haemost 2007; 98: 1343–1349 Introduction Compared to bare metal stenting, the use of drug-eluting stents reduced the rate of in-stent restenosis significantly (2–5), but Coronary artery disease is a worldwide problem, being one of the concerns are rising regarding a possible increase in the rate of leading causes of morbidity and mortality. To treat arterial ste- stent thrombosis caused by antiproliferative agents eluting from noses, percutaneous transluminal coronary angioplasty (PTCA) the stents, possibly due to impaired reendothelialization (6) or al- and stent deployment has become a widely applied strategy. Im- tered endothelial gene expression (7). plantation of a stent efficiently prevents acute vessel recoil and Restenosis is a multi-factorial complication mediated by sev- limits late arterial shrinkage and remodeling, leading to chronic eral key processes, including platelet activation, which results in lumen loss after angioplasty. Nevertheless, angioplasty and stent the release of mitogenic and chemo-attractant factors that me- implantation can still lead to a renarrowing of the treated vessel, diate vascular smooth muscle cell (VSMC) migration and prolif- resulting in a clinically significant restenosis of the targeted area eration and the formation of neointima. Despite several ap- after 6–12 months in a substantial number of treated patients (1). proaches, targeting different agonists, the ‘magic bullet’ to pre- Correspondence to: Financial Support: Prof. Dr. Hans Deckmyn This work was supported by a grant from the IWT (IWT020473) and a bilateral collab- Laboratory for Thrombosis Research oration grant between Flanders and South Africa (BIL/04/56) and a grant from I.R.C., KULeuven Campus Kortrijk Thromb-X. E. Sabbelaan 53 *S.F.D.M and S.S contributed equally to this manuscript. 8500 Kortrijk, Belgium Tel.: +32 56 24 64 22, Fax: +32 56 24 69 97 Received May 9, 2007 E-mail: hans.deckmyn@kuleuven-kortrijk.be Accepted after resubmission September 6, 2007 Prepublished online November 9, 2007 doi:10.1160/TH07-05-0335 94 1343 Downloaded by: University of Free State. Copyrighted material. 95 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis vent this multi-factorial process has not yet been found. Anti- Angioplasty and stent implantation thrombotic agents, such as aspirin and clopidogrel, are routinely Before surgery, baboons were anaesthetized by intramuscular in- used during angioplasty procedures to prevent acute thrombus jection of 10 mg/kg/30 min ketamine (Ketalar®, Pfizer, Brussels, formation (8–11). These inhibit only one out of many amplifi- Belgium) and 2 mg/kg/30 min xylazine (Chanazine®, Bayer, Le- cation loops involved in platelet activation, resulting in only a verkusen, Germany). Electrocardiogram and arterial blood press- partial inhibition of acute platelet thrombus formation and no ure were continuously monitored. A 6F introducer sheath was clinical effect on restenosis. Inhibitors of platelet aggregation placed into the femoral artery for blood sampling, whereas drugs have been a promising target, but clinical trials using glycopro- were administered intravenous (iv.) via the forearm. Approxi- tein (GP) IIb/IIIa inhibitors such as ReoPro report conflicting re- mately 10 min before angioplasty, 10000 IU heparin (Heparine sults (12–16). Indeed, GPIIb/IIIa antagonists do not interfere Rorer,Aventis/sanofi-synthelabo, Brussels, Belgium) and 900 mg with the primary adhesion of blood platelets, and consequently, aspirin (Aspegic®, Aventis/sanofi-synthelabo) were injected iv. can not prevent initial platelet activation and the concomitant re- and treatment animals received a bolus of 1 mg/kg 82D6A3, lease of VSMC mitogenic and attractant factors that trigger whereas control animals received a comparable volume of PBS. neointimal hyperplasia. In contrast to those interfering with later The coronary artery was visualized using a 6F catheter and Ultra- steps in platelet dependent haemostasis, antithrombotic agents vist® (Berlex, Montréal, Canada) as a contrast agent. A 18 mm that effectively block primary platelet adhesion reduce initial MULTI-LINK VISION coronary stent (Guidant, Diegem, Bel- platelet activation events, making them attractive candidates for gium) was mounted on a conventional coronary angioplasty bal- the prevention of neointima formation. The effect of several in- loon catheter and deployed in a selected arterial segment of either hibitors of platelet adhesion on neointima formation after angio- the left anterior descending artery or the right coronary artery, plasty has been studied in small laboratory animals, including using an inflation pressure of 8 atm for 30 s. Vessel patency after guinea pig, hamster, rabbit and rat, all with promising results (11, stent implantation was confirmed by coronary angiography after 17–22). Unfortunately however, successful outcome in small intra-arterial injection of 0.25 mg nitroglycerin. Finally, the arte- animal models for the prevention of restenosis, repeatedly did riotomy was repaired and the dermal layers were closed using stan- not allow straightforward extrapolation to human clinical in- dard techniques.After surgical procedure, animals were intramus- stent restenosis (23, 24). cularly injected with enrofloxacine (Baytril®, Bayer; 5 mg/kg). In order to better predict the outcome in patients, we devel- Animals received additional boli of 1 mg/kg 82D6A3 (treatment oped a clinically relevant human-like baboon coronary artery animals) or a comparable volume of PBS (control animals) at 12 h, model of in-stent stenosis in which we assessed the effect of in- 24 h and 48 h after stent implantation via the femoral vein. On the hibition of primary platelet adhesion on neointima formation. day of surgery, blood samples were drawn before the first antibody Standard clinical procedures, closely mimicking the human (or PBS) injection and after the surgical procedure. The next two clinical situation were used, including the standard antithrom- days, blood samples were drawn via the femoral vein immediately botic treatment regimen with aspirin, heparin and clopidogrel. before and after antibody (or PBS) administration. All animals re- Inhibition of platelet adhesion was established using the mono- ceived clopidogrel (Plavix®, Aventis/sanofi-synthelabo, 75 mg/ clonal antibody (moab) 82D6A3, that binds to the VWF A3-do- day, peroral) during 28 days, starting from the day before surgery. main and inhibits theVWF-collagen interaction resulting in anti- After 28 days, the animals underwent a control angiogram thrombotic effects (25, 26). and blood samples were drawn before they were sacrificed by iv. injection of an overdose of pentabarbitone (Euthapent, Kryon Methods Laboratories, Johannesburg, South Africa). All experiments and analyses were performed by researchers who were blinded for Animals the treatment regimen of the animals. Baboons (Papio ursinus) of male sex, weighing 18–30 kg, were fed with a standard diet and drinking water was available ad libi- Platelet count, coagulation and bleeding time tum. Housing, treatment, surgery and caring of the animals was All blood samples were collected in Vacutainers containing approved by the Control Committee on Animal Experimentation 0.105 M citric acid (BD Biosciences, Plymouth, UK). Platelets of the University of the Free State, South Africa. were counted using a Technicon H3 (Bayer Healthcare, Isando, South Africa). Plasma was prepared by centrifugation for 15 min Production and purification of 82D6A3 at 1000 x g. The prothrombin time (PT) and activated partial Monoclonal antibody 82D6A3 was raised in mice against human thromboplastin time (aPTT) were measured at 37°C in a STA® VWF (27) and the antibody was produced using the INTEGRA Compact coagulometer (Roche, Mannheim, Germany) using CELLine CL 350 system (Elscolab, Kruibeke, Belgium) using STA®-Neoplastine® and STA®-PTT Automate 5 reagents (Diag- serum-free nutrient medium (TX-HYB, Thromb-X, Leuven, nostica Stago, Taverny, France) respectively. Bleeding time in Belgium) as described by the manufacturer. The moab 82D6A3 baboons was measured as described (25) on the chest using a was purified by Protein A chromatography (GE Healthcare, Surgicutt® device (ITC, Rodano, Italy). Uppsala, Sweden) and purity was checked in SDS-PAGE. Endot- oxin levels in the antibody preparations were measured using the Blood parameters QCL-1000® Chromogenic LAL Endpoint Assay (BioWhittaker, Blood samples were collected in Vacutainers containing spray Walkersville, ME). coated Li-Heparin and a polymer separator gel (BD Bios- ciences). Creatinine, alkaline phosphatase, aspartate amino- 95 1344 Downloaded by: University of Free State. Copyrighted material. 96 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis transferase (AST), alanine aminotransferase (ALT), total creati- Histological preparation of coronary segments nine kinase, creatinine kinase MB iso-enzyme, troponine-I, After sacrifice of the animal, the heart was removed and press- C-reactive protein (CRP) and lactate dehydrogenase (LDH) were ure-fixed at 80 mmHg with PBS containing 4% formaldehyde to analysed using standard laboratory techniques. preserve arterial wall dimensions. The coronary segments con- taining a stent were carefully dissected and fixed with 4% for- VWF:Ag determination maldehyde in PBS for 2–4 h at 4°C. After washing overnight at Antigen levels of VWF in plasma samples were determined 4°C in PBS containing 6.8% saccharose, the segments were using a sandwich ELISA as described (28, 29). The baseline ba- dehydrated for 8 h at 4°C with 100% ethanol before embedding boon plasma sample (taken before surgery) was used as an inter- in hydroxy-ethylmethacrylate using Technovit 8100 (Kulzer nal standard for the VWF:Ag determination in each baboon. Histo-Technik 8100, Heraeus Kulzer, Wehrheim, Germany) fol- lowing the manufacturer’s instructions. Starting from the proxi- Determination of theVWF-collagen binding activity mal part of the coronary segment, pairs of 5 µm cross-sections (VWF:CBA) were cut using a carbide tungsten knife with 100 µm intervals Functional binding of VWF to collagen was performed essen- and stained with hematoxylin and eosin for morphometric analy- tially as described (29). The absorbance of the baseline sample sis. Where possible, up to ten sections, equally distributed over was arbitrarily set as 100% VWF:CBA. the stent segments, were blindly selected for analysis. Plasma 82D6A3 levels andVWF occupancy Morphometric analysis of coronary segments The plasma 82D6A3 levels in plasma and VWF occupancy by Blinded morphometric analysis was performed under an Eclipse 82D6A3 were determined in ELISA as described (25). To calcu- TE200 microscope (Nikon, Tokyo, Japan) equipped with a late VWF occupancy, respective baseline plasma samples Basler 113C RGB color digital camera (Basler, Ahrensburg, supplemented with a saturating amount of 82D6A3 (15 µg/ml) Germany). Images were captured and processed using Lucia G were used as 100% occupancy standards. software version 4.81 (Laboratory Imaging, Prague, Czech Re- public). To determine the degree of neointima formation, the in- ternal elastic lamina (IEL) and the border of the lumen were Figure 1: Pharmacokinetics of 82D6A3 after multiple injection in baboons. Baboons were injected with 1 mg/kg 82D6A3 10 min before PTCA and stent implantation and at 12 h, 24 h and 48 h after the first injec- tion. Blood was drawn before the first injection (0 min), after surgery (30 min) and 5 min be- fore and after each of the subsequent 82D6A3 injections. Plasma levels of 82D6A3 (A), VWF:CBA (B, ◆),VWF occupancy (B, ·) and plasmaVWF concentrations (C) were deter- mined. Data are represented as mean ± SEM of double measurements of samples taken from five baboons. 96 1345 Downloaded by: University of Free State. Copyrighted material. 97 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis Table 1: Ex-vivo plasma analyses, blood platelet count and bleeding time in baboons injected with multiple boli of 82D6A3 or PBS (controls). 82D6A3 Control timepoint platelets bleeding time PT (s) aPTT (s) platelets bleeding time PT (s) aPTT (s) (x103/µl) (min) (x103/µl) (min) 0 min 362±30 <1.3 8.8±0.2 26.2±4.1 295±24 <1.3 9.4±0.3 31.4±2.0 30 min 342±30 n.d. 13.7±4.2 >120 277±20 n.d. 9.9±0.3 >120 11 h 55 min 345±38 n.d. 8.7±0.2 32.3±0.9 304±33 n.d. 10.0±0.0 32.0±1.1 23 h 55 min 341±30 n.d. 9.2±0.5 38.2±2.5 276±48 n.d. 9.4±0.2 32.1±0.5 24 h 5 min 339±26 4.9±2.5 9.1±0.5 35.3±1.4 341±50 <1.3 n.d. n.d. 47 h 55 min 342±27 n.d. 8.5±0.2 32.6±1.2 287±22 n.d. 9.5±0.5 31.5±1.2 48 h 5 min 348±30 3.2±0.8 8.6±0.4 32.0±1.0 n.d. n.d. n.d. n.d. 28 d 282±38 2.1±0.4 9.0±0.8 32.6±0.8 247±24 1.8±0.3 9.6±0.2 32.2±0.8 Data are represented as the mean ± SEM, n = 5; n.d. = not determined. traced and the areas enclosed by each were calculated. The Statistical analysis neointimal area was defined as the area enclosed by the IEL sub- The degree of neointima formation and injury score from control tracted by the area of the lumen (IEL area – lumen area), where- baboons and baboons treated with 82D6A3 were compared by as the percentage of stenosis was defined as the ratio between the the Student t-test. The differences were considered to be statis- neointimal area and the IEL area ((neointimal area/IEL tically significant when p<0.05. area)*100). The intima over media ratio (I/M) was calculated by dividing the area of the neointima by the area of the media. The Results mean injury score for each cross-section was calculated by deter- mining the extent of the injury caused by every individual coil Pharmacokinetics of 82D6A3 in baboons during and wire in the cross-section according to Schwartz et al. (30). Injury after PTCA and stent implantation scoring was performed on hematoxylin and eosin stained cross- The effects of multiple bolus injections of 82D6A3 during and sections using green light (510–560 nm). after PTCA were analysed. To achieve an optimal inhibition of platelet adhesion, a complete occupancy of VWF by 82D6A3 is essential (25).The first bolus injection of 1 mg/kg of 82D6A3 re- sulted in a full occupancy ofVWF by 82D6A3 that was sustained for at least 12 h (Fig. 1B).Additional bolus injections of 82D6A3 were given after 12 h, 24 h and 48 h, resulting in increasing plas- ma concentrations of 82D6A3 and a complete VWF occupancy during the first 48 h after the first injection (Fig. 1A, B). More- over, in view of the sustained complete occupation for 12 h after the first bolus injection, we assumed that the applied regimen of 82D6A3 administration resulted in a complete VWF occupancy for at least 60 h after PTCA and stent implantation. In accord- ance with the VWF occupancy data, the ex-vivo binding of VWF to collagen was completely blocked for 48 h after the first anti- body administration (Fig. 1B), suggesting a complete inhibition of VWF-mediated platelet adhesion. VWF:Ag levels were not affected by the 82D6A3 treatment and remained constant at physiologic values (10 µg/ml) throughout the experiment (Fig. Figure 2: Coronary cross-sec- 1C). VWF:Ag levels and VWF:CBA were normal in control ba- tions of the stented segments boons (data not shown). of control and treated ba- Inhibition of the VWF-collagen interaction by 82D6A3 did boons. Sections of 5 µm were not induce bleeding problems in any of the treated baboons. In stained with hematoxylin and eosin agreement with previous studies (25), administration of 82D6A3 for morphometric analysis and show the presence of the stent did not result in thrombocytopenia and was not accompanied struts in the media. In-stent with a significant prolongation of the bleeding time and PT neointima formation is clearly vis- (table 1). Although a prolongation of the aPTT was observed im- ible in both groups. mediately after surgery, it was completely restored 12 h after sur- 97 1346 Downloaded by: University of Free State. Copyrighted material. 98 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis gery. The same increase was present in the control baboons, at- viously described that the antibody completely abolishes throm- tributing this rise in aPTT to the administration of heparin during bus formation in a modified Folt’s model in baboons, indicating PTCA and stent implantation. its ability to efficiently block platelet adhesion in this animal. In- Finally, the procedure did not cause liver, cardiac or kidney terestingly however, this study shows no reduction of neointima dysfunction nor inflammation, since alkaline phosphatase, AST, formation 28 days after stenting when 82D6A3 was used to ALT, total creatinine kinase, creatinine kinase MB iso-enzyme, block platelet adhesion. This finding, most probably reflecting troponine-I, creatinine, CRP and LDH were within normal range the outcome in man, is in sharp contrast to a variety of studies throughout the experiments (data not shown). that reported beneficial effects of inhibitors of platelet adhesion on angioplasty-induced neointima formation (11, 17–22). Morphometric analysis of coronary sections Though, the present study differs from these in some substantial In order to investigate the effect of complete abolishment of the features which all may account for the observed discrepancy. VWF-collagen binding during the first three days after PTCA First, previous studies were performed in small laboratory and stenting, stented coronary segments were isolated after 28 rodents, i.e rabbit (17), hamster (18, 20), guinea pig (19) and rat days for morphometric analysis. Figure 2 shows a representative (11, 21, 22) in which the cardiovascular and physiological sys- section isolated from a control baboon and a baboon treated with 82D6A3, 28 days after PTCA and stenting. The injury score on all selected sections of both control and treated animals was assessed and did not differ significantly be- tween control animals and those receiving 82D6A3 (1.24 ± 0.23 vs. 1.46 ± 0.23; p>0.1, Fig. 3A) and resemble injury scores ob- tained in porcine in-stent restenosis studies (31, 32). The injury scores confirmed that a uniform degree of injury was induced in the coronary arteries by the PTCA and stenting procedure in both groups. Of all selected sections, the degree of stenosis was morpho- metrically determined by dividing the area of neointima by the area enclosed by the IEL. This analysis revealed a mean percent stenosis of 31.2 ± 5.9 (n=4) in control animals. However, no sig- nificant reduction in neointima formation could be observed in animals treated with 82D6A3 where a mean percent stenosis of 36.2 ± 9.0 (n=4) was observed (Fig. 3B). This observation was confirmed by the I/M ratios, which did not differ significantly between both groups of animals (Fig. 3C). Discussion Restenosis is a complex process, involving several physiological processes such as thrombus formation, VSMC proliferation and arterial remodeling. In addition to stent placement, a variety of approaches has been investigated to prevent thrombus and neointima formation, including brachytherapy, VSMC antipro- liferative and antithrombotic therapy. However, none of these can exclude that pathological restenosis still occurs in a substan- tial number of treated patients (1, 33, 34). Despite initial promis- ing results, obtained in a variety of small animal models, extra- polation to human clinical restenosis is far from straightforward (23, 24), indicating the need for more human related animal models. Therefore, we here developed a clinically relevant baboon model of in-stent stenosis to study whether the inhibition of VWF-binding to collagen can contribute to the reduction of neointima formation after PTCA and stenting using procedures closely mimicking the human clinical situation. Administration of the previously described anti-VWF moab 82D6A3 (25, 26, Figure 3: Morphometric analysis of treated arterial sections. A) 35) on top of the standard treatment regimen with aspirin, clopi- Both groups of animals showed the same degree of arterial injury in-duced by the PTCA and stenting procedure. B) No significant difference dogrel and heparin, resulted in a complete inhibition of the ex- could be observed in percent in-stent stenosis between the two groups. vivo collagen-VWF interaction during the first 60 hours after C) The I/M ratios from both groups of animals were not statistically dif- surgery, without prolongation of the bleeding time. We pre- ferent. Data are represented as mean ± SEM (n=4). 98 1347 Downloaded by: University of Free State. Copyrighted material. 99 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis tems are far from identical to those in man. Second, although an- dicated that short-term inhibition of initial platelet adhesion re- gioplasty is the standard technique for revascularization of cor- sults in decreased neointima formation already within two weeks onary arteries, which are muscular arteries, arterial injury in pre- after vascular damage (11, 17–20, 22, 43, 44) vious in-vivo studies was mainly performed in peripheral ar- Taken together, our study, using a coronary in-stent stenosis teries, namely femoral (11) and carotid arteries (18–22). Clearly, model in baboons that is closely related to the human situation, is both injury and response to injury may be very different in cor- the first report investigating the possible antistenotic effect of in- onary versus peripheral arteries (36). Third, no stents were im- hibitors of the VWF-collagen interaction after stent implan- planted in the previous studies. It is well known in man that pro- tation. Our results suggest that, in the current routine clinical set- cesses responsible for lumen loss after angioplasty are different ting, additional inhibition of platelet adhesion is not sufficient to from those after angioplasty and stent implantation (1). Finally, reduce neointimal hyperplasia, which could indicate an overesti- in previous studies, the anti-adhesive agent was studied alone, mation of the role of VSMC mitogenic and attractant factors that whereas the animals in our study were also treated with aspirin, are released upon platelet activation after PTCA and stenting. clopidogrel and heparin, i.e. the routine clinical regimen that is Whether inhibition of platelet adhesion affects other biological used to minimize the risk of acute thrombosis or other procedur- processes besides neointmial hyperplasia (e.g. alterations in al complications (8). Although it would be very interesting to see neointimal compostion) remains unclear and may warrant future the possible effect of 82D6A3 alone in our model, such a strategy research. It is very likely that the contribution of the individual would not be supported for further clinical studies in humans. elements that mediate restenosis, including platelet activation Therefore, in an effort to mimick clinical reality as close as pos- after platelet adhesion, may not be consistently proportional in sible, moab 82D6A3 was tested on top of the standard clinical every patient or might even differ between lesions in the same treatment. In fact, whereas aspirin inhibits subacute stent throm- patient. In the absence of a single ‘magic bullet’ therapy that ef- bosis, it is not effective in the inhibition of restenosis after angio- fectively addresses all of these pathogenic elements, the ability plasty (37) and long term administration of clopidogrel on top of to routinely prevent restenosis will likely require accurate clini- aspirin does not seem to result in increased clinical benefits fol- cal identification and understanding of the principal factors gov- lowing angioplasty (38, 39). erning the biology of a given restenotic lesion. Still, it is important to note that also this baboon model does Finally, in view of the repeated failure of extrapolating not perfectly mimic human clinical restenosis. Indeed, we used a beneficial effects of various agents on the prevention of stenosis healthy baboon coronary artery, which is significantly different obtained in small animal models to human clinical restenosis from the treated pathological human arteries, where intimal (23, 24), this study points out the importance of evaluating new (atherosclerotic) lesion already exists and several physiological antistenotic therapies in a clinically relevant model and in the responses, such as healing, are bound to be abnormal. meantime presents a model that may be of further clinical inter- Restenosis is a ~6 month phenomenon and in the current set- est for studying the prevention of restenosis. up with a three day inhibition of platelet adhesion and an end- point analysis after 28 days, we can not exclude possible long- Acknowledgements term effects of 82D6A3 on neointima formation. However, both The antibody used in this study was produced by Thromb-X, Leuven, Bel- experimental animal models, including baboons, and human gium. Simon De Meyer and Karen Vanhoorelbeke are postdoctoral fellows autopsy and atherectomy show that the thrombogenicity of the of the FWO (Fonds voor Wetenschappelijk Onderzoek), Vlaanderen, Bel- exposed vessel wall drastically decreases after a few days gium. Stephanie Staelens is supported by a postdoctoral fellowship of the (40–42), suggesting that platelet adhesion (or the inhibition IWT (Instituut ter bevordering van Innovatie door Wetenschap enTechnolo- thereof) would not be important in the late (>1 month) develop- gie in Vlaanderen) Vlaanderen Belgium (OZM0503110) We would like to thank Stephan Vauterin and Hilde Gillijns for technical assistance. ment of restenosis. In line with this, studies in small animals in- References 6. Joner M, Finn AV, Farb A, et al. Pathology of drug- 11. Zahger D, Fishbein MC, Garfinkel LI, et al. VCL, 1. Bennett MR, O'Sullivan M. Mechanisms of angio- eluting stents in humans: delayed healing and late an antagonist of the platelet GP1b receptor, markedly plasty and stent restenosis: implications for design of thrombotic risk. JAm Coll Cardiol 2006; 48: 193–202. inhibits platelet adhesion and intimal thickening after rational therapy. Pharmacol Ther 2001; 91: 149–166. 7. Muldowney JA, III, Stringham JR, Levy SE, et al. balloon injury in the rat. Circulation 1995; 92: 2. Morice MC, Serruys PW, Sousa JE, et al. A ran- Antiproliferative agents alter vascular plasminogen ac- 1269–1273. domized comparison of a sirolimus-eluting stent with a tivator inhibitor-1 expression: a potential prothrom- 12. Topol EJ, Califf RM, Weisman HF, et al. Randomis- standard stent for coronary revascularization. N Engl J botic mechanism of drug-eluting stents. Arterioscler ed trial of coronary intervention with antibody against Med 2002; 346: 1773–1780. Thromb Vasc Biol 2007; 27: 400–406. platelet IIb/IIIa integrin for reduction of clinical reste- 3. Moses JW, Leon MB, Popma JJ, et al. Sirolimus- 8. Steinhubl SR,.Topol EJ. Clopidogrel with aspirin is nosis: results at six months. The EPIC Investigators. eluting stents versus standard stents in patients with the optimal antiplatelet regimen for intracoronary Lancet 1994; 343: 881–886. stenosis in a native coronary artery. N Engl J Med 2003; stenting. J Thromb Thrombolysis 1999; 7: 227–231. 13. Neumann FJ, Blasini R, Schmitt C, et al. Effect of 349: 1315–1323. 9. Bhatt DL, Bertrand ME, Berger PB, et al. Meta- glycoprotein IIb/IIIa receptor blockade on recovery of 4. Stone GW, Ellis SG, Cox DA, et al. A polymer- analysis of randomized and registry comparisons of ti- coronary flow and left ventricular function after the based, paclitaxel-eluting stent in patients with coronary clopidine with clopidogrel after stenting. J Am Coll placement of coronary-artery stents in acute myo- artery disease. N Engl J Med 2004; 350: 221–231. Cardiol 2002; 39: 9–14. cardial infarction. Circulation 1998; 98: 2695–2701. 5. Gallo R, Padurean A, Jayaraman T, et al. Inhibition 10. Steinhubl SR, Berger PB, Mann JT, III, et al. Early 14. Topol EJ, Mark DB, LincoffAM, et al. Outcomes at of intimal thickening after balloon angioplasty in por- and sustained dual oral antiplatelet therapy following 1 year and economic implications of platelet glycopro- cine coronary arteries by targeting regulators of the cell percutaneous coronary intervention: a randomized tein IIb/IIIa blockade in patients undergoing coronary cycle. Circulation 1999; 99: 2164–2170. controlled trial. JAMA 2002; 288: 2411–2420. stenting: results from a multicentre randomised trial. 99 1348 Downloaded by: University of Free State. Copyrighted material. 100 De Meyer et al. Role of platelet adhesion in baboon in-stent stenosis EPISTENT Investigators. Evaluation of Platelet IIb/ outcome of clinical trials? Cardiovasc Res 1998; 39: 34. Teirstein P, Reilly JP. Late stent thrombosis in bra- IIIa Inhibitor for Stenting. Lancet 1999; 354: 50–59. chytherapy: the role of long-term antiplatelet therapy. J 2019–2024. 24. Libby P, Tanaka H. The molecular bases of resteno- Invasive Cardiol 2002; 14: 109–114. 15. Acute platelet inhibition with abciximab does not sis. Prog Cardiovasc Dis 1997; 40: 97–106. 35. Vanhoorelbeke K, Depraetere H, Romijn RAP, et reduce in-stent restenosis (ERASER study). The 25. Wu D, Vanhoorelbeke K, Cauwenberghs N, et al. al. A consensus tetrapeptide selected by phage display ERASER Investigators. Circulation 1999; 100: Inhibition of the von Willebrand (VWF)-collagen inter- adopts the conformation of a dominant discontinuous 799–806. action by an antihuman VWF monoclonal antibody re- epitope of a monoclonal Anti-VWF antibody that in- 16. Kastrati A, Mehilli J, Schuhlen H, et al. A clinical sults in abolition of in vivo arterial platelet thrombus hibits the von Willebrand factor-collagen interaction. J trial of abciximab in elective percutaneous coronary in- formation in baboons. Blood 2002; 99: 3623–3628. Biol Chem 2003; 278: 37815–37821. tervention after pretreatment with clopidogrel. N Engl 26. Staelens S, Hadders MA, Vauterin S, et al. Paratope 36. Muller DW, Ellis SG, Topol EJ. Experimental mod- J Med 2004; 350: 232–238. determination of the antithrombotic antibody 82D6A3 els of coronary artery restenosis. J Am Coll Cardiol 17. Waissbluth A, Ghigliotti G, Abendschein DR, et al. based on the crystal structure of its complex with the 1992; 19: 418–432. Aurintricarboxylic acid attenuates intimal thickening von Willebrand factor A3-domain. J Biol Chem 2006; 37. Goodnight SH. Aspirin therapy for cardiovascular after balloon injury of the rabbit aorta. Thromb Hae- 281: 2225–2231. disease. Curr Opin Hematol 1996; 3: 355–360. most 2002; 88: 668–672. 27. Hoylaerts MF, Yamamoto H, Nuyts K, et al. von 38. Eriksson P. Reply: Long-term clopidogrel follow- 18. Matsuno H, Kozawa O, Niwa M, et al. Multiple in- Willebrand factor binds to native collagen VI primarily ing PCI: marginal antithrombotic effects are offset by hibition of platelet activation by aurintricarboxylic acid via itsA1 domain. Biochem J 1997; 324 (Pt 1): 185–191. increased bleeding risks. Eur Heart J 2004; 25: prevents vascular stenosis after endothelial injury in 28. Ingerslev J. A sensitive ELISA for von Willebrand 2170–2171. hamster carotid artery. Thromb Haemost 1998; 79: factor (vWf:Ag). Scand J Clin Lab Invest 1987; 47: 39. Kandzari DE, Berger PB, Kastrati A, et al. In- 865–871. 143–149. fluence of treatment duration with a 600-mg dose of 19. Kageyama S, Yamamoto H, Yoshimoto R. Anti- 29. Vanhoorelbeke K, Cauwenberghs N, Vauterin S, et clopidogrel before percutaneous coronary revascular- human von willebrand factor monoclonal antibody al. A reliable and reproducible ELISA method to ization. J Am Coll Cardiol 2004; 44: 2133–2136. AJvW-2 prevents thrombus deposition and neointima measure ristocetin cofactor activity of von Willebrand 40. Kinlough-Rathbone RL, Packham MA, Mustard formation after balloon injury in guinea pigs. Arte- factor. Thromb Haemost 2000; 83: 107–113. JF. Vessel injury, platelet adherence, and platelet sur- rioscler Thromb Vasc Biol 2000; 20: 2303–2308. 30. Schwartz RS, Huber KC, Murphy JG, et al. Reste- vival. Arteriosclerosis 1983; 3: 529–546. 20. Matsuno H, Kozawa O, Niwa M, et al. Inhibition of nosis and the proportional neointimal response to cor- 41. Fuster V, Falk E, Fallon JT et al. The three processes von Willebrand factor binding to platelet GP Ib by a onary artery injury: results in a porcine model. J Am leading to post PTCA restenosis: dependence on the fractionated aurintricarboxylic acid prevents restenosis Coll Cardiol 1992; 19: 267–274. lesion substrate.Thromb Haemost 1995; 74: 552–559. after vascular injury in hamster carotid artery. Circu- 31. Banai S, Gertz SD, Gavish L, et al. Tyrphostin 42. Touchard AG, Schwartz RS. Preclinical restenosis lation 1997; 96: 1299–1304. AGL-2043 eluting stent reduces neointima formation models: challenges and successes. Toxicol Pathol 21. Cruz CP, Eidt J, Drouilhet J, et al. Saratin, an in- in porcine coronary arteries. Cardiovasc Res 2004; 64: 2006; 34: 11–18. hibitor of von Willebrand factor-dependent platelet ad- 165–171. 43. Phillips MD, Moake JL, Nolasco L, et al. Aurin tri- hesion, decreases platelet aggregation and intimal hy- 32. Sinnaeve P, Chiche JD, Gillijns H, et al. Over- carboxylic acid: a novel inhibitor of the association of perplasia in a rat carotid endarterectomy model. J Vasc expression of a constitutively active protein kinase G von Willebrand factor and platelets. Blood 1988; 72: Surg 2001; 34: 724–729. mutant reduces neointima formation and in-stent reste- 1898–1903. 22. Davis JA, Brown AT, Alshafie T, et al. Saratin (an nosis. Circulation 2002; 105: 2911–2916. 44. Cruz CP, Eidt J, Drouilhet J, et al. Saratin, an in- inhibitor of platelet-collagen interaction) decreases 33. Stone GW, Aronow HD. Long-term care after per- hibitor of von Willebrand factor-dependent platelet ad- platelet aggregation and homocysteine-mediated post- cutaneous coronary intervention: focus on the role of hesion, decreases platelet aggregation and intimal hy- carotid endarterectomy intimal hyperplasia in a dose- antiplatelet therapy. Mayo Clin Proc 2006; 81: perplasia in a rat carotid endarterectomy model. J Vasc dependent manner. Am J Surg 2004; 188: 778–785. 641–652. Surg 2001; 34: 724–729. 23. Lafont A, Faxon D. Why do animal models of post- angioplasty restenosis sometimes poorly predict the 100 1349 Downloaded by: University of Free State. Copyrighted material. S-8 Journ1a0l 1of Thrombosis and Haemostasis, 7: 429437 DOI: 10.1111/j.1538-7836.2008.03271.x ORIGINAL ARTICLE A human monoclonal antibody inhibiting partially factor VIII activity reduces thrombus growth in baboons M. JACQUEMIN ,* J . M . STASSEN , J . -M. SA INT-REMY,* P . V ERHAMME,* R . LAVEND HOMME,* L . VANDER ELST ,* M. ME IR ING , H. P I ETERS , S . LAMPR ECHT , J . ROODT and P . BADENHORST *Center for Molecular and Vascular Biology, University of Leuven, Leuven; ThromboGenics, Leuven, Belgium; and Department of Hematology, University of Free State, Bloemfontein, South Africa To cite this article: Jacquemin M, Stassen JM, Saint-Remy J-M, Verhamme P, Lavendhomme R, VanderElst L, Meiring M, Pieters H, Lamprecht S, Roodt J, Badenhorst P. A human monoclonal antibody inhibiting partially factor VIII activity reduces thrombus growth in baboons. J Thromb Haemost 2009; 7: 42937. Keywords: antibody, baboon, factor VIII, thrombosis. Summary. Background: The inhibitory activity of an anti- factor VIII (FVIII) antibody can be modulated through Introduction glycosylation of the antigen binding site, as has recently been described. This offers the opportunity to develop an optimized Most anticoagulant agents developed so far target enzymes anticoagulant agent targeting partial FVIII inhibition. involved in the coagulation cascade. This strategy is associated Objectives: We investigated in non-human primates the withwell-knownrisks.VitaminKantagonists exert theiractivity antithrombotic activity, pharmacokinetics,and pharmacody- not only on procoagulant enzymes but also on inhibitors of the namics of a human monoclonal antibody, Mab-LE2E9Q, coagulation cascade such as Protein C [1] and require monitor- inhibiting FVIII activity partially. Methods: The ability of ing. The development of synthetic inhibitors of coagulation Mab-LE2E9Q to prevent thrombosis was evaluated in enzymes has increased the specificity of the inhibition but has baboons after administration of 1.25 and 5 mg kg)1 antibody generated the possibility of unexpected toxicity, notably for the or saline as a single intravenous (i.v.) bolus. Thrombus liver [2]. Targeting the non-enzymatic cofactors of the coagu- development was recorded in expansion (venous) and in lation cascade therefore appears as a potentially attractive Dacron (arterial) thrombosis chambers incorporated in an alternative provided that limited inhibition of the activity of the extracorporeal arteriovenous shunt implanted between the target cofactor can be guaranteed so as to prevent bleeding. femoral vessels 1 h, 24 h and 7 days after the administration of Based on that concept, we have produced a human Mab-LE2E9Q. Results: Mab-LE2E9Q reduced thrombus monoclonal antibody, Mab-LE2E9Q, partially inhibiting fac- growth to a similar extend 1 h, 1 day and 1 week after tor VIII (FVIII) activity irrespective of the excess of antibody administration of the antibody. Ex vivo pharmacodynamic over FVIII. That antibody was generated by introducing a analysis indicated that the evaluation of the residual FVIII point mutation in the gene of an anti-FVIII antibody, Mab- activity was strongly dependent on the type of FVIII assay and LE2E9, derived by immortalization of B lymphocytes from a on the phospholipid concentration in the assay. No significant patient with mild hemophilia A, who had developed a strong difference in bleedings was observed between animals treated immune response to FVIII.Mab-LE2E9 inhibits 90%of FVIII withMab-LE2E9Q or with saline.Conclusions:Understanding activity and prevents FVIII from binding to von Willebrand the role of glycosylation in FVIII inhibition by a human factor (VWF) [3]. The point mutation introduced in the Mab- monoclonal antibody allowed selection of an antibody inhib- LE2E9 gene to generate Mab-LE2E9Q removes a glycosyla- iting only moderately FVIII activity while significantly reduc- tion site in the variable region of the heavy chain. As a result, ing thrombus development in a baboon extracorporeal model. Mab-LE2E9Q inhibits only about 40% of FVIII activity and As that antibody did not increase the bleeding tendency, it may does not prevent FVIII from binding to VWF [4]. represent a novel type of a long-acting antithrombotic agent The antithrombotic efficacy of Mab-LE2E9Q has been with an optimal safety/efficacy profile. evaluated in mice that carry a mutation in the heparin-binding site of antithrombin III (Atm/m) and display spontaneous chronic thrombosis at several sites [5]. This model does not Correspondence: Marc Jacquemin, Center for Molecular and Vascular require surgical intervention and is representative of thrombo- Biology, Herestraat 49, B-3000 leuven, Belgium. Tel.: +32 16 346018; fax: +32 16 345990. sis in patients with severe prothrombotic risk factors. Although E-mail: marc.jacquemin@med.kuleuven.ac.be it inhibits only about 40% of FVIII activity, administration of Mab-LE2E9Q completely prevented acute thrombosis in )/) Received 1 September 2008, accepted 12 December 2008 ATIII mice [4].  2008 International Society on Thrombosis and Haemostasis 101 430 M1.0J2acquemin et al Given the low concentration of FVIII in plasma and the 15 mm 15 mm long half-life of an IgG antibody, treatment with the Mab- LE2E9Q antibody could be very convenient, allowing one administration every month. In addition, because Mab- LE2E9Q inhibits FVIII activity only partially, FVIII activity can be normalized very rapidly by administration of FVIII Expansion Dacron independently of the antibody concentration in plasma [4]. Chamber Chamber Accordingly, Mab-LE2E9Q is so far the only anticoagulant “venous” “arterial” agent that can be neutralized specifically and without any thrombosis thrombosis delay. Mab-LE2E9Q therefore appears as a promising novel type of antithrombotic drug. Here, we evaluated the ability of Mab-LE2E9Q to reduce thrombus development in a thrombosis model in non-human Detector primates. Fig. 1. Experimental protocol for extracorporeal thrombosis. Arteriove- Materials and methods nous shunts were implanted in male baboon femoral vessels. Thrombo- genic devices prefilled with saline were incorporated as extension segments Thrombosis model in baboons into the permanent arteriovenous shunt. Platelet-dependent arterial thrombus was induced by inserting Dacron into the wall of Silastic tubing. Study design The study was randomized and investigators Coagulation-dependent venous thrombosis was generated in an expansion 111 were blinded towards treatment regimen.Male baboons (Papio chamber. The deposition of autologous In-labeled platelets was fol- ursinus) were used. The animals weighed between 7.5 and 19 kg lowed with a gamma scintillation camera. and were disease free for at least 2 months prior to the experiments. All procedures were approved by the Ethics The procedure was carried out simultaneously with one Committee for Animal Experimentation of the University of animal treated with 1.25 mg kg)1 Mab-LE2E9Q, one with the Free State in accordance with the National Code for 5 mg kg)1 Mab-LE2E9Q and one with control buffer. Animal Use in Research, Education, Diagnosis and Testing of Randomization of treatment was therefore restricted to sets Drugs and Related Substances in South Africa. of three animals. After the thrombosis experiments, animals Baboons were immobilized and handled under anesthesia were followed up for 53 days for clinical and biological with ketamine hydrochloride (Anaket-V; Centaur Laboratory, parameters. Johannesburg, South Africa). Permanent polytetrafluoroethyl- In the treated group, animals received either 1.25 or 5 mg ene (Teflon) and silicone rubber (Silastic) arteriovenous shunts kg)1 Mab-LE2E9Q. These doses of antibody were selected to were implanted in the baboon femoral vessels [6]. Blood flow allow for kinetic analyses and to ensure that Mab-LE2E9Q through the shunts varied between 100 and 120 mL min)1. In concentrations in plasma would remain above the concentra- each experiment, a thrombogenic device prefilled with saline to tion required to reach the maximal FVIII inhibition at least up avoid a blood-air interface was incorporated as an extension to day 7 when the last thrombosis experiment had to be segment into the permanent arteriovenous shunt by means of performed. Because the antibody inhibits FVIII only partially Teflon connectors (Fig. 1). even in large excess over FVIII, the residual FVIII activity was Platelet-dependent arterial thrombus was induced using expected to be in the same range for both treated groups up to Dacron (1.26 cm2; US Catheter Inc., Billerica, MA, USA) the end of the thrombosis experiments. insertedintothewallofSilastictubing(3-mminsidediameter) [7]. An expansion chamber (3.77 cm2) was used to generate coag- Thrombus imaging Autologous platelets were labeled with ulation-dependent venous thrombosis. Blood flowed through 111In-tropolone and reinjected into the animal 1 h before the the thrombogenic devices at a rate of approximately 120 mL start of the control experiment [7]. This allowed quantification min)1.The initial shear stresswas318 s)1 for theDacron section of thrombus formation within the chambers on day 0 and 1. To and 10 s)1 for the expansion chamber. provide image acquisition on day 7 the labeling procedure was In the control studies, the devices were kept in place for repeated. Image acquisition of the grafts was done with a 60 min or until they occluded. After removal of the device, gamma scintillation camera fitted with a high-resolution blood flow through the arteriovenous shunt was re-established. collimator. To determine baseline blood radioactivity, The baboons were then treated with a single intravenous (i.v.) imaging was also performed on an autologous blood sample bolus of 1.25 or 5 mg kg)1 Mab-LE2E9Q or saline. Thromb- each time the grafts were imaged. Regions of interest of the ogenic devices were placed for 60 min 1 h after antibody graft and expansion segments were selected to determine the injection. Additional 60-min studies were carried out at 24 h deposited and circulation radio-activity in the dynamic image. after the antibody bolus injection. The extracorporeal shunts The number of platelets deposited on the vascular graft were then removed. At day 7, the baboons were shunted on the material and in the expansion chamber was measured every opposite site and60-min thrombosis experimentswere repeated. 3 min.  2008 International Society on Thrombosis and Haemostasis 102 103 Anti-FVIII antibody 431 The highest platelet deposition (maximal platelet binding) (n = 5) or saline (n = 4). In addition, three animals received occurring during the observation period was determined for a single subcutaneous (s.c.) administration of Mab-LE2E9Q. each experiment. An alternative model to evaluate platelet Blood samples were taken at intervals after the deposition was also used to take into account that platelets can administration of a single i.v. or s.c. administration of detach from the thrombosis chamber, which is followed by Mab-LE2E9Q, either directly from the shunt or by further platelet binding. Therefore, the number of platelets venopuncture. bound during each time period of 3 min was calculated as the Mab-LE2E9Q concentration was measured in a competitive difference between the number of platelets at the end and at the assay in which the binding of biotinylated FVIII to insolubi- beginning of the time period. The cumulative platelet deposi- lized Mab-LE2E9Q was inhibited by Mab-LE2E9Q present in tion was then calculated as the sum of all platelet deposition the test sample. One volume of test sample, diluted in Tris >0 during each 3-min time period of an experiment (cumu- 50 mM, pH 7.3, containing 150 mM NaCl (TBS) and 1% lative platelet binding). bovine serum albumine (BSA) supplemented with 5% baboon plasma, was mixed with 1 volume of biotinylated FVIII. The mixture was added to 96-wells microtitration plates prealably Study endpoints coated with 50 lL Mab-LE2E9Q. Following a 2-h incubation The primary endpoint of the study was the evaluation of the period, plates were washed 3 times with TBS and bound platelet deposition during the observation period. To allow for biotinylated FVIII was detected by addition of streptavidine analysis of the data with a non-parametric test (MannWhit- peroxidase followed by O-phenylenediamine. The concentra- ney U-test), the smallest group had to include at least four tion of Mab-LE2E9Q in the test samples was determined by animals. Groups of five animals were therefore planned to comparing the inhibition of biotinylated FVIII binding to the allow for one potential casualty during the experimental inhibition of FVIII binding in mixtures containing known procedure. The secondary endpoints included TB-402 phar- concentrations of Mab-LE2E9Q. macodynamics and pharmacokinetics. Pharmacokinetic data were fitted to a two-compartment model by weighted non-linear regression analysis or by visual Mab-LE2E9Q ex vivo pharmacodynamics All animals examination. First, data were plotted and fitted to a sum of two used in this study were included in the extracorporeal exponential terms, C(t) = Ae)at + Be)bt, by weighted non- thrombosis study and were treated with a single intravenous linear regression analysis using GraFitR. Initial estimates of the bolus of 1.25 (n = 5) or 5 mg kg)1 Mab-LE2E9Q (n = 5) or model parameters were provided by visual examination. When saline (n = 5). Blood samples were taken either directly from the best fit obtained with GraFit did not provide a good the shunt or by venopuncture. estimation of the terminal elimination phase, data were fitted FVIII:C levels were measured in a batch analysis at the end by visual examination using Excel. Data were plotted and an of the study using a chromogenic assay (CoatestR FXa optimal curve corresponding to C(t) = Ae)at + Be)bt was generation assay; Chromogenix-Instrumentation Laboratory obtained by selecting different B, a and b parameters. A was SpA, Milano, Italy) according to the manufacturers instruc- calculated as the antibody concentration measured 5 min after tions. In addition, FVIII concentrations were evaluated using i.v. administration minus B. a modification of the CoatestR FXa, in which phospholipids The drug clearance parameters were calculated from the were diluted sixtyfold in tris(hydroxymethyl)-aminoethane extrapolated intercepts (A, B) and exponents (a, b) describing (Tris) 50 mM, pH 7.3, 150 mM NaCl, 0.2% bovine albumin the disposition of each administered product, using standard [8]. FVIII activity in test sample was determined by comparing formulas. The following clearance parameters were calculated: FXa generation with that obtained using dilutions of a pool of initial drug concentration in the blood: C0 = A + B; volume human plasma. The sensititivity of low- and high-phospho- of the central compartment:Vc = dose/(A + B); total volume lipids assays to Mab-LE2E9Q inhibitory activity was evalu- of distribution: Vd = dose/B; extrapolated area under the ated in an in vitro spiking experiment. Various concentrations curve: AUC = A/a + B/b; plasma clearance: Clp = dose/ of Mab-LE2E9Q were added to baboon plasma, supple- AUC; initial half-life: t1/2a = ln2/a; secondary half-life: t1/ mented with 25 mM HEPES and incubated at 37 C for 2 h 2c = ln2/b, mean residence time: MRT = Vd · AUC/dose. before measuring the residual FVIII:C activity, using a The pharmacokinetics parameters were determined for each CoatestR SP FXa generation assay in which phospholipids animal individually. The mean and standard deviation (SD) of were either undiluted or diluted sixtyfold, as above. In the pharmacokinetics parameters were then calculated per dose addition, FVIII:C levels were also measured in a 1-stage group. clotting assay (Actin FS; Dade Behring, Marburg, Germany), immediately after plasma collection. Laboratory values Mab-LE2E9Q ex vivo pharmacokinetics Animals used in Laboratory values were obtained at day 0, 1, 2, 5, 7, 9, 12, 16, this study participated in the extracorporeal thrombosis study 19, 23, 26, 30, 44 and 60. They included hemoglobin level, and the pharmacodynamic study and were treated with a hematocrit level, platelet count, white blood cell count, FVIII single i.v. bolus of 1.25 (n = 5) or 5 mg kg)1 Mab-LE2E9Q and TB-402 concentration.  2008 International Society on Thrombosis and Haemostasis 103 432 M1.0J4acquemin et al control animals treated with saline in both the venous and Immunogenicity arterial thrombosis chambers (Fig. 2A,B). The difference was IgG and IgM antibodies to Mab-LE2E9Q were detected using statistically significant in the arterial chamber at all time points ELISA. Mab-LE2E9Q (2 lg mL)1 in 20 mM glycine, 34 mM after Mab-LE2E9Q injection and 24 h after administration in NaCl, pH 9.2) was incubated inmicrotitration plates overnight. the venous chamber. After three washes with 140 mM NaCl, 67 mM KCl, 20 mM A fraction of the platelets bound in the venous or in the Na2HPO4, 4.4 mM KH2PO4, pH 7.4, Tween-80 0.005% (PBS- arterial thrombosis chamber detached from the thrombus in Tween), baboon PL diluted 1/20 and 1/100 in TBS with 1% some experiments. To take this phenomenon into account, a BSA 0.005% Tween-80 (Tris-BSA-Tween) was incubated in cumulative platelet deposition was calculated by adding up the wells at room temperature for 2 h. After three washes with platelet depositions during each time period of 3 min. The PBS-Tween, baboon antibodies bound to Mab-LE2E9Q were cumulative platelet deposition was 1.1 to 2.3 times higher than detected by addition of goat antibaboon antibodies conjugated the maximal platelet deposition in 21 and in 17 out of 60 to horseradish peroxidase (Biorad; ref. 170-6516) diluted experiments in the venous and arterial thrombosis chambers, thousandfold in Tris-BSA-Tween followed after three washes respectively (Fig. 3A,B). The cumulative platelet deposition with PBS-Tween by addition of O-phenylenediamine. Optical was significantly lower in treated animals than in controls in densities were read at 490 nm. both the venous and arterial thrombosis chambers at all time Statistical analysis A Venous thrombosis chamber 100 000 NS NS P = 0.008 NS For pharmacokinetics analysis, data obtained from animals treated with 1.25 and 5 mg kg)1 were analyzed separately. For comparison of hemoglobin levels and platelet deposition in thrombosis chambers, data from the animals treated with both 10 000 doses were combined and compared with data from untreated animals. Data were analyzed using theMannWhitneyU-test. Significance was defined as P £ 0.05. 1000 Saline Results Mab-LE2E9Q 1.25 mg kg–1 Mab-LE2E9Q 5 mg kg–1 Prevention of thrombosis 100 0 Before 1 h 24 h 7 days The ability of Mab-LE2E9Q to prevent thrombosis was After TB-402 administration evaluated in an extracorporeal thrombosis model in baboons. At the beginning of the experiment, an extracorporeal shunt Arterial thrombosis chamber B was inserted between the femoral artery and the femoral vein 100 000 NS P = 0.001 P = 0.019 P = 0.005 on one side of the animal. Thrombogenic devices were placed for 60 min 1 h before, 1 h and 24 h after Mab-LE2E9Q or saline injection. The extracorporeal shunts were then removed. At day 7, the baboons were shunted on the opposite site and a 10 000 thrombosis experiment was repeated. Platelet deposition was recorded as a function of time in the expansion (venous) thrombosis chamber and in the DacronR (arterial) thrombosis 1000 chamber incorporated in the extracorporeal arteriovenous Saline shunt. Mab-LE2E9Q 1.25 mg kg–1 In the control group, five animals were treated with saline. In Mab-LE2E9Q 5 mg kg–1 the treated groups, five animals received 5 mg kg)1 Mab- 100 0 Before 1 h 24 h 7 days LE2E9Q and five others 1.25 mg kg)1 antibody. Because the antibody inhibits FVIII only partially even in large excess over After TB-402 administration FVIII [5], the residual FVIII activity was expected to be in the Fig. 2. Reduction of thrombus growth in baboons treated with Mab- same range for both treatment groups at least up to day 7 when LE2E9Q. Thrombogenic devices were placed for 60 min 1 h before and the last thrombosis experiment was performed. Those assump- 1 h, 24 h and 7 days after Mab-LE2E9Q or saline injection. Platelet tions were validated in the pharmacodynamic and pharmaco- deposition was recorded over 1 h in the venous (A) and arterial (B) kinetic study (see below). thrombosis chambers. The maximal platelet depositions recorded at any time point over the entire observation period are indicated for each animal. In the 10 animals treated with Mab-LE2E9Q, the maximal The mean platelet deposition per dose group is also indicated. The sta- platelet deposition 1 h, 1 day and 7 days after administration tistical differences between treated and control groups were calculated of the antibody platelet deposition was lower than in the using the MannWhitney U-test (NS, non-statistically significant).  2008 International Society on Thrombosis and Haemostasis 104 Maximal platelet binding Maximal platelet binding 105 Anti-FVIII antibody 433 A Venous thrombosis chamber 100 FVIII assay: low PL 100 000 NS P = 0.04 P = 0.005 P = 0.055 FVIII assay: high PL 80 10 000 60 40 1000 saline 20 Mab-LE2E9Q 1.25 mg kg–1 Mab-LE2E9Q 5 mg kg–1 100 0 0 Before 1h 24h 7 days 0.001 0.01 0.1 1 10 TB-402 (µg mL–1) After TB-402 administration Fig. 4. Sensitivity of high and low phospholipids chromogenic assays to Mab-LE2E9Q inhibitory activity. Baboon plasma spiked with various B Arterial thrombosis chamber concentrations of Mab-LE2E9Q and incubated for 2 h at 37 C. The 100 000 NS P = 0.001 P = 0.019 P = 0.005 residual FVIII:C was measured with high- and low-phospholipid chro- mogenic assays. Results are expressed as mean ± SD of FVIII inhibition by comparison to control samples without Mab-LE2E9Q in two independent experiments. 10 000 tion of the former one, was rendered sensitive to the inhibitory activity of the antibody by reducing sixtyfold phospholipids 1000 saline concentrations. The second assay had previously been used to Mab-LE2E9Q 1.25 mg kg–1 detect the effect of a mutation responsible for mild hemophilia Mab-LE2E9Q 5 mg kg–1 A that impaired FVIII interaction with phospholipids [8]. 100 FVIII:C was measured in baboon plasma spiked with 0 Before 1h 24h 7 days various concentrations ofMab-LE2E9Q and incubated for 2 h After TB-402 administration at 37 C. The maximal FVIII inhibitions measured with high and low phospholipids assays were 29% and 50%, respectively Fig. 3. Reduction of cumulative platelet deposition in baboons treated (Fig. 4). Noteworthy, no FVIII inhibition was observed when with Mab-LE2E9Q. Platelet deposition was recorded as in Fig. 2 in the FVIII:C was measured with a one-stage coagulation assay venous (A) and arterial (B) thrombosis chambers. A cumulative platelet deposition was also calculated by summing platelet deposition during each (Actin FS R, data not shown). time period of 3 min. The mean cumulative platelet deposition per dose Comparison of FVIII:C measured with the low- and high- group is also indicated. The statistical differences between treated and phospholipids assays showed a clear decrease in FVIII activity control groups were calculated using theMannWhitneyU-test (NS, non- after Mab-LE2E9Q administration (Fig. 5A,B). FVIII levels statistically significant). were normalized by the end of the follow-up period in the group treated with 1.25 mg kg)1 whereas FVIII:C was still points after Mab-LE2E9Q administration, except at day 7 in reduced in the group treated with 5 mg kg)1. In contrast, the venous thrombosis chamber (P = 0.055). FVIII:C measured with both assays was similar in animals treated with saline (Fig. 5C). No difference was observed between the different groups when FVIII:C was measured with Pharmacodynamics the one-stage coagulation assay (Fig. 6). Blood samples were taken from baboons included in the Interestingly, the lack of a clear-cut increase or decrease of thrombosis prevention study at intervals after administration FVIII:C measured with the high phospholipids assay and with of a single i.v. bolus (1.25 mg kg)1 or 5.0 mg kg)1) of Mab- the one-stage FVIII assay, insensitive to Mab-LE2E9Q, up to LE2E9Q or saline. 60 days after administration of the antibody indicated that As preliminary experiments had shown that the measure- Mab-L2E9Q does not significantly modify the rates of FVIII ment of residual FVIII activity after incubation of human production and clearance. plasma with Mab-LE2E9Q varies according to the type of FVIII assay used, different types of FVIII assays were used to Pharmacokinetics evaluate FVIII:C in PL of baboons treated with Mab- LE2E9Q. One test, a commercial chromogenic assay, barely Blood samples were taken for the pharmacodynamic study. detected Mab-LE2E9Q whereas the second assay, a modifica- The concentrations of Mab-LE2E9Q were evaluated by  2008 International Society on Thrombosis and Haemostasis 105 Cumulative platelet binding Cumulative platelet binding FVIII inhibition (%) 434 M1.0J6acquemin et al A 300 300 200 200 100 5 mg kg–1 Mab-LE2E9Q FVIII assay: high PL saline h FVIII assay: low PL 100 TB-402 1.25 mg kg–1 –1 0 1 2 0 TB-402 5 mg kg–1 1 10 20 30 40 50 60 h mAb-LE2E9Q-L –1 0 2 injection days 1 0 1 10 20 30 40 50 60 B 300 mAb-LE2E9Q-L days injection Fig. 6. FVIII levels measured in a 1-stage coagulation assay in baboons treated with Mab-LE2E9Q. Blood samples from baboons treated with 200 Mab-LE2E9Q were harvested as in Fig. 5. The concentrations of FVIII were measured using a 1-stage coagulation FVIII assay, which is insensi- tive to Mab-LE2E9Q inhibitory activity (Actin FSR). Data are shown as mean ± SEM and as a function of time. 100 1.25 mg kg–1 Mab-LE2E9Q FVIII assay: high PL h FVIII assay: low PL 1000 –1 0 1 2 TB-402 5 mg kg –1 i.v. 0 TB-402 1.25 mg kg–1 i.v. 1 10 20 30 40 50 60 100 mAb-LE2E9Q-L injection days 10 C 300 1 200 0.1 0 10 20 30 40 50 60 70 Time after administration (days) 100 Fig. 7. Plasma clearance curve after a single intravenous injection ofMab-saline LE2E9Q in baboons. Antibody concentration were measured by ELISA FVIII assay: high PL h in plasma of baboons treated by one single subcutaneous administrationFVIII assay: low PL 0 of 1.25 or 5 mg kg )1 Mab-LE2E9Q and are expressed as mean ± SD. –1 1 2 0 Data were fitted to a biexponential disposition curve with the parameters 10 20 30 40 50 60 of Table 1. mAb-LE2E9Q-L injection days Fig. 5. FVIII levels measured in chromogenic assays in baboons treated distribution was 134 ± 46 mL kg)1 (Table 1). Similar bio- with Mab-LE2E9Q. Baboons were treated by one intravenous adminis- availability and terminal disposition phases were observed after tration of 5 (A) or 1.25 (B) mg kg)1 Mab-LE2E9Q or saline (C) as indi- s.c. administration of 5 mg kg)1 Mab-LE2E9Q. cated. Blood samples were harvested starting 1 h before administration. The concentrations of FVIII were measured using chromogenic assays with a high or low phospholipid concentration, which is less sensitive to Clinical and biological observations Mab-LE2E9Q inhibitory activity. Data are shown as mean ± SEM and as a function of time. The experimental procedure and the administration of Mab- LE2E9Q were well tolerated by all animals. In both groups, measuring the ability of Mab-LE2E9Q in the plasma samples most animals presented with small hematomas at the sites of to inhibit FVIII binding to insolubilized Mab-LE2E9Q. After venopuncture and oozing at the sites where the catheters had single i.v. bolus injections of 1.25 mg kg)1 Mab-LE2E9Q, a been inserted. One animal (treatment 5 mg kg)1 Mab- biphasic disappearance curve with rapid initial clearance and LE2E9Q) died 25 days after administration of the antibody. slow secondary clearances was observed (Fig. 7). The initial Neither external nor subcutaneous bleedings were observed in half-life of Mab-LE2E9Q was about 7.2 ± 4 h (mean ± SD) this animal. As no necropsy could be performed an internal and the secondary half-life was 305 ± 105 h. The volume of bleeding problem could not be excluded. However, the  2008 International Society on Thrombosis and Haemostasis 106 FVIII (ng mL–1) FVIII (ng mL–1) FVIII (ng mL–1) TB-402 (µg mL–1) FVIII (ng mL–1) 107 Anti-FVIII antibody 435 Table 1 Calculated pharmacokinetic parameters of Mab-LE2E9Q after a 5 mg kg)1 Mab-LE2E9Q was measured in sandwich ELISA. single intravenous bolus injection in baboons None of the animals developed a detectable humoral response Mab-LE2E9Q Mab-LE2E9Q to Mab-LE2E9Q (data not shown). 1.25 mg kg)1 5 mg kg)1 C0 39 ± 16 128 ± 46 lg mL)1 Discussion A 29 ± 14 88 ± 38 lg mL)1 B 10 ± 3 40 ± 11 lg mL)1 In this study, we sought evidence that Mab-LE2E9Q, a a 0.15 ± 0.12 0.16 ± 0.10 h)1 human monoclonal antibody inhibiting FVIII only partially, b 0.0026 ± 0.0012 0.0028 ± 0.0005 h)1 might constitute a novel approach to anticoagulant therapy, V 38 ± 18 46 ± 25 mL kg)1c )1 without the risk of overdosing or causing spontaneousVd 134 ± 46 132 ± 37 mL kg t 7.2 ± 4.0 5.5 ± 2.6 h bleeding. The data demonstrated that Mab-LE2E9Q signifi-1/2a t 305 ± 105 257 ± 49 h cantly reduced thrombus growth in an extracorporeal circu-1/2b 12.7 ± 4.4 10.7 ± 2 days lation model in baboons, without inducing excess bleeding. AUC 4916 ± 2648 15 611 ± 4990 lg h mL)1 The inhibition of thrombus growth was similar 1 h, 1 day and Clp 0.3 ± 0.2 0.3 ± 0.1 mL h)1 1 week after Mab-LE2E9Q, indicating a stable and long- MRT 472 ± 168 388 ± 73 h acting inhibition of FVIII when the antibody is in excess over FVIII. 10 Control The ability of Mab-LE2E9Q to reduce thrombus develop- Mab-LE2E9Q 1.25 mg kg–1 ment was evaluated in venous and arterial thrombosis Mab-LE2E9Q 5 mg kg–1 chambers inserted in an extracorporeal circulation shunt 8 between the femoral vessels. The venous thrombosis chamber was an extension chamber whereas the arterial chamber was covered by Dacron, a surface that allows platelet adhesion. In 6 both the venous and the arterial thrombosis chambers, thrombus growth was monitored by measuring platelet depo- sition. The rationale for the selection of that method was that 4 platelets are present in both venous and arterial thrombi, although their contribution to the physiopathological process are lower in the former than in the latter [9,10]. 2 Although treatment withMab-LE2E9Q did not increase the bleeding tendency, it resulted in a reduction of platelet 0 deposition in both chambers. The reduction of thrombus development in the arterial chamber may appear surprising but Fig. 8. Hemoglobin drop following Mab-LE2E9Q injection. Baboons is in agreement with the observation that the rate of myocardial were treated with one i.v. administration of 1.25 or 5 mg kg)1 antibody. infarction is lower whenFVIII is reduced [11] andwith thewell- Hemoglobin levels were followed up to 8 weeks after antibody adminis- tration. Results are expressed as the most severe hemoglobin drop, as established role of FVIII in arterial thrombus development in compared with hemoglobin concentrations before treatment, recorded experimental models [12]. However, it cannot be excluded that during the observation period. The statistical significance of differences platelet deposition in one chamber was influenced by the rate of between treated and control groups were tested using the MannWhitney platelet deposition in the other chamber and by the resulting U-test. flow alterations. The major risk of targeting FVIII with an anticoagulant hemoglobin levels in that animal were normal (13.9 g L)1) agent is that of causing complete inhibition of FVIII activity. 2 days before death indicating that that animal did not present The observation that Mab-LE2E9 VH N-glycosylation deter- with any chronic bleeding problem. mines the maximal inhibitory activity of the antibody [4] A key parameter for the selection of an anticoagulant agent offered a unique opportunity to develop an optimal anticoag- is the safety with regards to the induction of bleedings. Only a ulant agent targeting FVIII. Indeed, both modulation of the limited drop in hemoglobin level was observed in animals glycosylation and mixing deglycosylated, mutated antibodies treated with Mab-LE2E9Q. This drop was not significantly with native ones allowed us achieve inhibition ranging from the different from that observed in control animals treated with inhibitory activities of 40% to 90% [4]. saline (Fig. 8). Tail clipping experiments in mice treated withMab-LE2E9Q demonstrated that in vivo the antibody only partially neutral- izes FVIII activity. Similarly, all in vitro assays indicated that Immunogenicity the antibody inhibits FVIII activity only partially, even when The presence of anti- Mab-LE2E9Q IgG and IgM antibodies the antibody is in large excess over FVIII. However, different in plasma collected before administration of Mab-LE2E9Q types of FVIII assays displayed various sensitivities for the and 7, 14, 21, 30 and 60 days after administration of 0, 1.25 or detection of the inhibitory activity of Mab-LE2E9Q in baboon  2008 International Society on Thrombosis and Haemostasis 107 Maximal hemoglobin drop (g L–1) 436 M1.0J8acquemin et al plasma as well as in human plasma [13]. This was unexpected Interestingly, the use of the assays insensitive to Mab- because the type of FVIII assay is not considered as a major LE2E9Q (chromogenic assay with high phospholipids concen- cause of variability in the Bethesda assay, possibly because type tration and 1-stage coagulation assay) provided information I inhibitors are usually selected for multicenter studies aiming regarding the clearance of FVIII in the presence of the at standardization of the method. antibody. Thus, when measured with the FVIII assays A potential explanation of this phenomenon is that the insensitive to Mab-LE2E9Q, FVIII:C activity remained stable binding of a type II inhibitor antibody, such as Mab-LE2E9Q, in animals treated with 1.25 mg kg)1 or 5 mg kg)1 Mab- to FVIII induces a qualitative alteration of the FVIII molecule. LE2E9Q, indicating that FVIII clearance was not altered in the It is interesting to compare such alterations with those observed presence of the antibody. in patients with mild/moderate hemophilia A. Some FVIII Pharmacokinetic analyses indicated that the half-life of variants with reduced stability or impaired thrombin activation Mab-LE2E9Q is about 10 days in baboons, in agreement show different activities in one-stage vs. two-stages assays with the long half-life of IgG. No increase of the clearance of [1418]. This discrepancy has clinical implications as it may the antibody was observed over the entire follow-up period, lead tomisdiagnosed hemophilia A. Similarly, a deletion of one in agreement with the lack of immune response to the amino acid in the FVIII domain mediating FVIII binding to antibody. phospholipids significantly impairs FVIII interaction with Although the reduction of thrombus development in phospholipids. However, this alteration is not detected in the baboons treated with Mab-LE2E9Q cannot be directly conventional FVIII assays but requires assays with reduced extrapolated to a clinical situation in man, they are in phospholipids concentrations or platelets as a source of agreement with epidemiological observations that a limited phospholipids [8]. The last observations prompted the evalu- reduction of FVIII activity, such as that observed in carriers of ation of the inhibitory activity of Mab-LE2E9Q in different hemophilia A, has a positive impact on vascular disease [11]. types of FVIII assays. These observations may have important implications for the In this study, two chromogenic assays differing by the development of efficient, easy and safe strategies for the concentration of phospholipids and a one-stage coagulation prevention and treatment of thrombosis. This antibody repre- assay were used to measure residual FVIII activity in the sents an unique and new strategy to prevent thrombosis and plasma of animals treated with Mab-LE2E9Q. The one-stage warrants further investigation in other preclinical and clinical coagulation assay did not allow us detect a difference in FVIII studies. level between animals treated with Mab-LE2E9Q and control animals. This can be attributed to the poor sensitivity of that Addendum assay to the inhibitory activity of Mab-LE2E9Q. in contrast, the comparison of FVIII:C levels measured with the high- and M. Jacquemin, R. Lavendhomme, L. VanderElst were low-phospholipid concentrations allowed us to identify the responsible for in vitro coagulation studies. J.M. Stassen, inhibitory activity of Mab-LE2E9Q. In agreement with the J.-M. Saint-Remy, P. Badenhorst, H. Pieters, P. Verhamme pharmacokinetics data, administration of 1.25 mg resulted in contributed to the design of the study. M. Meiring, the reduction of FVIII activity up to day 44 whereas FVIII S. Lamprecht, J. Roodt were responsible for thrombosis activity was still clearly reduced 60 days after the administra- experiments in baboons. All authors contributed to the revision tion of 5 mg kg)1 Mab-LE2E9Q. of the manuscript. The variation of FVIII inhibition by Mab-LE2E9Q as a function of the FVIII assay used to measure residual FVIII:C Acknowledgements raises questions about the in vivo inhibitory activity of the antibody. The observation that mice treated with Mab- We thank S. Loosbergh and P. Barbeaux for human LE2E9Q survive a tail clipping experiment in contrast to monoclonal antibody production. This study was supported FVIII)/) mice indicates that the antibody only partially by grants G.0231.05 andG.0275.05 from the Flemish Research inactivates FVIII. Similarly, the lack of severe bleedings in Foundation. The CMVB is supported by an Excellentie baboons treated with Mab-LE2E9Q suggests that FVIII Financiering grant. inhibition is only partial. However, the exact level of FVIII inhibition remains unknown. In vivo local parameters, such as Disclosure of Conflict of Interests tissue damage, flow or phospholipids, may not only determine thrombus development but also influence the inhibitory activity The authors state that they have no conflict of interest. of the antibody. Such a mechanism may explain that the antithrombotic activity of Mab-LE2E9Q appears to vary References according to the thrombosis model. Thus, in mice with a strong prothrombotic phenotype, Mab-LE2E9Q completely pre- 1 Viganò S, Mannucci PM, Solinas S, Bottasso B, Mariani G. vented thrombus development [5], whereas in the extracorpo- Decrease in protein C antigen and formation of an abnormal protein soon after starting oral anticoagulant therapy.Br JHaematol 1984; 57: real circulation model in baboons, the antibody only partially 21320. reduced thrombus growth.  2008 International Society on Thrombosis and Haemostasis 108 109 Anti-FVIII antibody 437 2 Agnelli G, Eriksson BI, Cohen AT, Bergqvist D, Dahl OE, Lassen 10 Quarmby J, Smith A, Collins M, Cederholm-Williams S, Burnand K. MR, Mouret P, Rosencher N, Andersson M, Bylock A, Jensen E, Amodel of in vivo human venous thrombosis that confirms changes in Boberg B, on behalf of the EXTEND Study Group. Safety assessment the release of specific soluble cell adhesion molecules in experimental of new antithrombotic agents: lessons from the EXTEND study on venous thrombogenesis. J Vasc Surg 1999; 30: 13947. ximelagatran. Thromb Res 2008; 123: 48897. 11 Sramek A, Kriek M, Rosendaal FR. Decreased mortality of 3 Jacquemin M, Benhida A, Peerlinck K, Desqueper B, Vander Elst L, ischaemic heart disease among carriers of haemophilia. Lancet 2003; Lavendhomme R, dOiron R, Schwaab R, Bakkus M, Thielemans K, 362: 3514. Gilles JG, Vermylen J, Saint-Remy JM. A human antibody directed to 12 Kawasaki T, Kaida T, Arnout J, Vermylen J, Hoylaerts MF. A new the factor VIII C1 domain inhibits factor VIII cofactor activity and animal model of thrombophilia confirms that high plasma factor VIII binding to von Willebrand factor. Blood 2000; 95: 15663. levels are thrombogenic. Thromb Haemost 1999; 8: 30611. 4 Jacquemin M, Radcliffe CM, Lavendhomme R, Wormald MR, 13 JacqueminM, Lamprecht S, Lavendhomme R, Roodt J, Chatelain B, VanderElst L, Wallays G, Dewaele J, Collen D, Vermylen J, Dwek Deneys V, Dewaele J, Stassen JM, Saint-Remy JM,MeiringM, Pieters RA, Saint-Remy JM, Rudd PM, DewerchinM. Variable region heavy H, Badenhorst P. The determination of the inhibitory activity of some chain glycosylation determines the anticoagulant activity of a factor Factor VIII type II inhibitor antibodies is biased by the type of Factor VIII antibody. J Thromb Haemost 2006; 4: 104755. VIII assay. Blood 2005; 106: 894a (Abstract). 5 Dewerchin M, van der Elst L, Singh I, Grailly S, Saint-Remy JM, 14 Rudzki Z, Duncan EM, Casey GJ, Neumann M, Favaloro EJ, Lloyd Collen D, Jacquemin M. Inhibition of factor VIII with a partially JV. Mutations in a subgroup of patients with mild haemophilia A and inhibitory human recombinant monoclonal antibody prevents a familial discrepancy between the one-stage and two-stage factor thrombotic events in a transgenic model of type II HBS antithrombin VIII:C methods. Br J Haematol 1996; 94: 4006. deficiency in mice. J Thromb Haemost 2004; 2: 7784. 15 Mazurier C, Gaucher C, Jorieux S, Parquet-Gernez A. Mutations in 6 Kotze HF, Lamprecht S, Badenhorst PN, van Wyk V, Roodt JP, the FVIII gene in seven families with mild haemophilia A. Br J Hae- Alexander K. In vivo inhibition of acute platelet-dependent throm- matol 1997; 96: 4267. bosis in a baboon model by Bay U3405, a thromboxane A2-receptor 16 Keeling DM, Sukhu K, Kemball-Cook G, Waseem N, Bagnall R, antagonist. Thromb Haemost 1993; 70: 6725. Lloyd JV. Diagnostic importance of the two-stage factor VIII:C assay 7 Hanson SR, Kotze HF, Savage B, Harker LA. Platelet interactions demonstrated by a case of mild haemophilia associated with with Dacron vascular grafts. A model of acute thrombosis in baboons. His1954 fi Leu substitution in the factor VIII A3 domain. Br J Arteriosclerosis 1985; 5: 595603. Haematol 1999; 105: 11236. 8 DOiron R, Lavergne JM, Lavendhomme R, Benhida A, Bordet JC, 17 Schwaab R, Oldenburg J, Kemball-Cook G, Albert T, Juhler C, Negrier C, Peerlinck K, Vermylen J, Saint-Remy JM, Jacquemin M. Hanfland P, Ingerslev J. Assay discrepancy in mild haemophilia A due Deletion of alanine 2201 in the FVIII C2 domain results in mild to a factor VIII missense mutation (Asn694Ile) in a large Danish hemophilia A by impairing FVIII binding to VWF and phospholipids family. Br J Haematol 2000; 109: 5238. and destroys amajor FVIII antigenic determinant involved in inhibitor 18 Pipe SW, Saenko EL, Eickhorst AN, Kemball-Cook G, Kaufman development. Blood 2004; 103: 1557. RJ. Hemophilia A mutations associated with 1-stage/2-stage 9 Meijden PE, Heemskerk JW, Hamulyák K, Cate H. Classification activity discrepancy disrupt protein-protein interactions within the of venous thromboembolism (VTE). J Thromb Haemost 2005; 3: triplicated A domains of thrombin-activated factor VIIIa. Blood 2001; 25757. 97: 68591.  2008 International Society on Thrombosis and Haemostasis 109 110 110 S-9 111 111 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 112 112 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 113 113 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 114 114 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 115 115 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 116 116 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms 117 117 This content downloaded from 146.182.18.9 on Wed, 24 Oct 2018 14:39:10 UTC All use subject to https://about.jstor.org/terms S-10 76 118 ISSN 0254-3486 = SA T ydskrif vir N atuurwetenskap en Tegnologie 18, no. 3 1999 Navorsings- en oorsigartikels Die toepassing van peptiedblootlegging op fage in trombose en hemostase S.M. Meiring*, H.F. Kotzé, G.H.J. Pretorius en P.N. Badenhorst D epartem ent H em atologic en Selbiologie, U niversiteit van die Oranje-V rystaat, Posbus 339(G2), B loem fontein, 9300 *Outeur aan wie korrespondensie gerig kan word. Ontvang 6 Julie 1999, aanvaar 31 Augustus 1999 UITTREKSEL Die omvang en gebruike van die tegnologie van peptiedblootlegging op fage, en die toepassing daarvan op die gebied van trombose en hemostase en van geneesmiddelontwerp word bespreek. Die algemene beginsels van die tegniek word verduidelik. ABSTRACT The application o f phage display technology in thrombosis and haemostasis The scope and utility o fphage display are reviewed with emphasis on application in the fie ld o f thrombosis and haemostasis and drug design. General principles o f this technique are described. INLEIDING Die tegnologie van peptied- o f proteienblootlegging op fage is ’n kragtige en vinnige tegniek om protei'ene met spesifieke Hemostase is ’n beskermende meganisme wat oormatige bloed- antitrombotiese funksies te isoleer en/of te ontwerp en word verlies beperk wanneer bloedvate beskadig word en help om ’n reeds aangewend in die ontwikkeling van nuwe antitrombotiese fisiologiese homeostase te handhaaf. Normale hemostase word middels. Dit is egter nodig om die tegniek van peptied­ bewerkstellig deur ’n reeks komplekse interaksies tussen blootlegging op fage te verduidelik voordat die toepassing daar­ bloedvate, bloedselle, stollingsfaktore en -inhibeerders, asook van in trombose en hemostase bespreek word. die fibrinolitiese stelsel. Trombose, aan die anderkant, word allerweë as die teenbeeld van hemostase beskou. Na beskadiging PEPTIED BLO O TLEG G IN G OP FAGE van die bloedvat-endoteel, sodra die bloed in aanraking met die subendoteel kom, k leef die bloedplaatjies m.b.v. die gliko- Peptiedblootlegging op fage is ’n tegniek waar ’n peptied of proteren (GP) Ib/IX/V-reseptore op hulle membrane aan die prote'ien op die oppervlak van ’n filamenteuse faag blootgelê subendoteel en word geaktiveer. Geaktiveerde plaatjies stel die word. Dit word gedoen deur ’n gekose geen o f geenfragment wat reseptor GP Ilb/Illa op hulle oppervlak bloot, waaraan die vir ’n oppervlakprote'ien kodeer in die faaggeen in te bou. Hier­ klewingsproteiene, veral fibrinogeen en von Willebrandfaktor die inbouing bemvloed nie die noodsaaklike funksies van die (vWF) bind. Die klewingsproteiene bevat ’n Arg-Gly-Asp- opperviakproteïene nie en lei tot die blootlegging van ’n fusie- volgorde wat aan GP llb/IIIa bind. Terselfdertyd word stollings­ peptied o f -proteien op die oppervlak van die faag. Die peptied of faktore geaktiveer om trombien te vorm en word die plaatjieprop prote'ien kan dan as ’n ligand, ensiem, immunogeen o f enige ander aktiewe deelnemer in ’n biologiese reaksie optree. Hierdie versterk om oormatige bloedverlies te verhoed.' Endoteelselle in die omgewing van die beskadigde wand skei tromboreguleerders tegniek is eerste deur Smith in 1985“' gebruik. Hy het ’n ver- sameling o f biblioteek van lukraakgekose peptiedvolgordes aan (prostasiklien, stikstofoksied en fibrinolitiese middels) af, wat die amino-terminus van die mantelprote'ien, glikoprote'i'en (GP) die grootte van die prop beperk.^ Die plasminogeenaktiveerders, III, op die oppervlak van ’n faag M13 blootgelê. Wanneer nl. weefseltipe plasminogeenaktiveerder (tPA) en urokinase-tipe lukraakgekose DNA-volgordes in faaggenome ingebou word, plasminogeenaktiveerder (uPA) is belangrike fibrinolitiese word lukraak peptiede op die oppervlak van die fage blootgelê en middels wat die pro-ensiem plasminogeen aktiveer om plasmien dit lei tot ’n groot mengsel van faagklone wat elk ’n verskillende te vorm. Plasmien verteer die onoplosbare fibrien in die peptiedvolgorde op die oppervlak blootlê. Uit hierdie mengsel plaatjieprop na opiosbare afbraakprodukte en sodoende beperk van faagklone (faagbiblioteek) word fage wat aan ’n teiken- dit die grootte van die prop. Plasminogeenaktiveerder- prote'i'en bind, geselekteer. Die geselekteerde fage kan dan in inhibeerders (PAI) is protei'ene wat tPA en uPA inhibeer. bakterieë soos E. coli vermeerder word. Die aminosuurvolgorde Oormatige aktivering van plaatjies deur biomateriale en van die peptiede op die geselekteerde fage word afgelei deur die aterosklerotiese letsels, oorkom egter hierdie tromboregulerende ingevoegde DNA se volgorde m.b.v. bestaande tegnieke te meganismes. W anneer dit gebeur, verander die plaatjies en bepaal. Die krag van hierdie tegniek lê daarin dat funksionele stollingsfaktore in trombotiese oortreders wat trombo-embolieë aktiwiteit gekoppel kan word aan genetiese inligting. Peptied­ veroorsaak en bloedvate kan afsluit. Dit lei tot potensieel fatale blootlegging bring vinnige seleksie van ’n gekose prote'ien mee, siektes, soos akute miokardiale infarksie en beroerte.^ Die belang terwyl die genetiese inligting vinnige en betroubare verineerde- van trombo-emboliese komplikasies van arteriële ring van fage meebring. Dit lei tot die produksie van groot hoe- bloedvatsiektes word al hoe meer besef en dit het gelei tot die veelhede moontlike Interessante teikens. Die tegnologie word ook soektog na meer effektiewe antitrombotiese middels. gebruik om antiliggaam-epitope suksesvoi te bepaal. om 118 ISSN 0121594-3486 = SA T ydskrifv ir Natuurwetenskup en Tegnologie 18, no. 3 1999 77 struktuurfunksie-verwantskappe van sellulêre reseptore te bestu- ’n funksionele GP 111 te produseer. So byvoorbeeld sal daar vir deer en om ensiemsubstraat-spesifisiteit na te vors en te ver- ’n heksapeptiedbiblioteek, oligonukleotiede ingebou word wat ’n ander. Om die tegniek beter te verduidelik, word elke faset lukraak volgorde (NNK)f, bevat. In hierdie geval beteken dit dat daarvan in meer besonderhede bespreek. N gelyke hoeveelhede van al vier nukleotiede adenien, guanien, sitosien en timien bevat, terwyl K gelyke hoeveelhede guanien Filam enteuse faag en timien bevat, Deur dit so te doen, word die aantal verskillende Fiiamenteuse fage is ideaal geskik vir in v;7/-o-seleksie omdat trinukleotiede wat gevorm word, verminder van 4 x 4 x 4 = 64 na iiuile klein genome bevat waarin groot biblioteke van ’n verskei- 4 X 4 X 2 = 32, terwyl geen stopkodons gevorm word nie, Hierdie deniaeid verskillende gene maklik ingebou kan word. Die enkei- genetics gemanipuleerde faagvektore word dan gebruik om E. string-DNA-fage (fd o f M13) word meestal gebruik. Hierdie coli te infekteer. Die selle produseer fage wat geisoleer en fage het ’n deursnee van 6 nm en is 1 000 - 2 000 nm lank. Aan gesuiwer kan word, en die biblioteek is gereed vir seleksie. Die die een punt is daar 5 kopieë van elk van die mantelproteïene GP totale aantal verskillende heksapeptiede wat uit die 2 0 aminosure ill en VI (flguur 1). Die hoof-mantelproteien, GP VIII, kom in gevorm kan word, is 20'’ = 64 x 10“ fage, ’n Praktiese biblioteek 2 800 kopiee op die sye van die faag voor. Beide GP III en VIII bevat sowat I O'* tot 10''' fage wat ongeveer 70% van alle moont- stel hulle aminoterminale op die oppervlak van die faag bloot, like aminosuur-volgordekombinasies verteenwoordig, terwyl GP VI se karboksielkant ekstrasellulêr geleë is. Die faaggenoom bevat sowat 6 500 nukleotiede. Seleksie en verm eerdering Die lewensiklus van fd en M I3 filamenteuse fage begin wan- Seleksie van fage word gedoen deur eerstens ’n teikenproteïen neer hulle E. coli infekteer. Die faag bind aan die punt van die F- (reseptor, antiliggaam, ligand ens,) aan ’n plastiese immuunbuis tipe geslagspilus van die E. coli en word in die sel ingeneem. Die o f aan magnetiese korrels wat vooraf met kitsmelkpoeier of F-pilus van die E. coli breek dan a f sodat ’n E. coli slegs deur een beesserum-albumien (BSA) bedek is, te immobiliseer (figuur 2). faag met ’n spesifieke DNA-samestelling geïnfekteer kan word. Die kitsmelkpoeier o f BSA dien as ’n blokkeermiddel om die Die gevolg is die ontstaan van ’n bakteriële kolonie, wat fage bindingsetels wat nie deur die teikenproteïen beset word nie, te met dieselfde peptiedvolgorde produseer. Na inname word die beset. Die teikenproteïen word vervolgens met die faagbiblioteek enkelstring faag-DNA na ’n dubbelstring-vorm omgeskakel. geinkubeer. Na interaksie word die nie-bindende fage afgewas Daarna volg herhaaidelike replisering, transkripsie van faaggene, sodat net fage met peptiede wat aan die teikenproteien bind, proteiensintese en uiteindelike vrystelling as fage. Die mega- behoue bly. Die bindende fage kan dan nie-spesifiek met ’n sterk nismes wat hierdie prosesse reguleer, is goed bekend maar buite suur geëlueer word. Na suur-eluering word die faagoplossing die bestek van hierdie oorsig.* Dit is belangrik om daarop te let met ’n sterk basis geneutraliseer. Geselekteerde fage kan ook dat die faag nie die gasheer doodmaak nie, en dat ’n E. coli wat spesifiek geëlueer word. So byvoorbeeld kan antiliggaam- deur een faag geïnfekteer is, tussen 100 en I 000 nuwe fage gebonde fage met die natuurlike antigeen geëlueer word en tydens een generasietydperk produseer. reseptorgebonde fage met die ligand, ’n Voorbeeld hiervan is die isolasie van ’n inhibeerder van die trombienreseptor op plaatjies. K onstruksie van ’n faagbiblioteek In hierdie geval word ’n faagbiblioteek met intakte bloedplaatjies ’n Faagbiblioteek word gebou deur ’n versameling vreemde geinkubeer en die bindende fage met die reseptorligand geëlueer. oligonukleotiede in die M l3- o f fd-faaggenome in te bou. ’n Dit dui daarop dat spesifieke kennis oor die reseptor nie nodig is Tetrasiklien- (o f enige ander antibiotikum) weerstandsgeen en om binders, o f met ’n bietjie geluk, inhibeerders te selekteer nie.'’ restriksie-ensiemsnypunte word eerstens in die faaggenoom Die bindende fage word dan oornag met F-pilus-bevattende E. ingebou. Die tetrasiklien-weerstandsgeen bring mee dat geinfek- coli in hulle eksponensiele groeifase geinkubeer in die teenwoor- teerde E. coli weerstandbiedend is teen tetrasiklien. Vervolgens digheid van ’n antibiotikum (bv. tetrasiklien). Die geïnfekteerde word ’n stukkie vreemde DNA, wat as ’n vulsel dien, op so ’n E. coli produseer fage wat versamel word deur die E. co//-kultuur wyse m.b.v. restriksie-ensieme in een van die mantelproteien- te sentrifugeer en te suiwer deur ’n aantal polietileenglikol- gene (bv. GP III) ingebou dat die leesraam verskuif Fage wat natriumchloried (PEG-NaCI) presipiterings- en hersuspendering- hierdie DNA-vulsel bevat, sal ’n mutant van GPIII produseer wat siklusse. Hierdie geselekteerde fage word dan gebruik vir ’n verhoed dat die fage E. coli kan infekteer. Hierdie vuIsel-DNA tweede, derde of selfs vierde rondte van seleksie en vermeer­ word dan verwyder en die gaping word met genoegsame dering. Die aantal bindende fage kan op dié wyse met tussen oligonukleotiede gevul sodat die leesraam weer herstel word om I 000 en 100 000 keer per rondte vermeerder word. Die sukses 119 78 120 ISSN 0254-3486 = SA T ydskrif vir N atuurwetenskap en Tegnologie 18, no. 3 1999 Geïmmobiliseerde ligand X aantal rondtes Inkubasie met faagblootleggings- Vermeerder deur biblioteek E. coli-selle te infekteer Verskeie wasstappe A A A Elueer fage met hoë-affinteitklone ■0---- Verwydering fage met AAA lae-affiniteitklone -13 0 — ^ Figuur 2: 'n Skematiese voorstelling van peptiedblootlegging op fage. van die seleksieproses word bepaal deur ’n klein hoeveelheid van algemeen gebruik. ’n Variasie hiervan is om ’n sisteïen aan elk die geinfekteerde bakterieë wat na die laaste seleksierondte van die amino- en karboksieterminale van die peptiede in te bou versamel is, op agarplate uit te plaat sodat enkelkolonies kan sodat ’n sikliese peptied blootgelê word.’ Daar word algemeen opgroei. Omdat ’n faag siegs een bakterie kan infekteer, sal elke aanvaar dat die bindingseienskappe van ’n sikliese peptied beter kolonie ’n versameling van identiese fage wees. Enkelkolonies is as dié van ’n liniêre peptied.’ wat opgetel en in mikrotiterplaatputte gekweek word, produseer Die Arg-Gly-Asp-volgorde wat in die klewingsproteiene, fibri- dus monoklonale fage wat, na ’n sentrifugeringstap, versamel nogeen en fibrien voorkom, bind aan die GP Ilb/lIIa-reseptor op word. Hierdie fage word m.b.v. ’n ensiemgekoppelde immuun- die plaatjiemembraan en vorm sodoende ’n brug wat plaatjie- bepaling (ELISA)-tegniek getoets vir binding aan die teiken- aggregasie tot gevolg het.* ’n Peptiedbiblioteek waar ’n Arg-Gly- proteien. Kortliks kom dit daarop neer dat die fage oorgedra Asp-volgorde tussen twee stelle lukraak tripeptiede ingebou is, is word na mikrotiterputte waaraan die teikenproteïen gekoppel is. gebruik om ’n effektiewe inhibeerder van plaatjie-aggregasie te Na inkubasie word die nie-bindende fage afgewas en ’n antifaag- isoleer.^ Dit is gedoen deur die GP Ilb/llla-reseptor aan die wand antiliggaam, gekonjugeer aan peperwortelperoksidase (HRPO), van ’n immuunbuis te immobiliseer en met die faagbiblioteek te in die putte gevoeg. N a ’n verdere wasstap word ’n kleursub- inkubeer. Peptiede wat aan GP Ilb/lIIa bind, is geselekteer en straat bygevoeg om die bindende faagkolonies te identifiseer. getoets vir inhibisie van plaatjie-aggregasie. Volgordebepaling Daar is verskeie groepe wat peptiedbiblioteke gebruik het om die epitoop waarmee monoklonale antiliggame aan hulle teiken- Die E. co//-kolonies wat bindende fage bevat, word gekweek om protei'ene bind, te bepaal. So byvoorbeeld is die epitoop van ’n genoegsame hoeveelhede fage te produseer sodat die faag-DNA menslike antiliggaam teen plasminogeenaktiveerder-inhibeerder geisoleer kan word. Die DNA-volgorde van die oligonukleotied- tipe 1 en aj-m akroglobulien bepaal.'” In die laasgenoemde geval invoeging word dan m.b.v. klassieke volgordebepalingstegnieke is daar ’n stel van vyf oorvleuelende fragmente wat oor 54 ami­ bepaal. nosure strek (van Gly429, tot Cys4j44) van die P-ketting van die reseptor, geisoleer. DIE GEBRUIK VAN PEPTIEDBLOOTLEGGING IN Die trombienreseptor op die bloedplaatjiemembraan is ’n inte- TROMBOSE EN HEMOSTASE grale membraanproteien. Trombien bind daaraan en sny sewe Navorsers op die gebied van trombose en hemostase maak reeds aminosure van die C-terminaal a f Die sewe-aminosuurligand van peptied-, antiliggaam-, cDNA- en proteienbiblioteke ge- wat vorm, is dan in staat om as ’n trombienreseptor-agonis op te bruik, enersyds om kandidate te identifiseer wat as antitrom- tree en die plaatjies via die reseptor te aktiveer. Hierdie eienskap botiese middels ontwikkel kan word, en andersyds om van die reseptor is gebruik om ’n inhibeerder van die reseptor te struktuurfunksie-verwantskappe in trombose te bepaal. Wat hier- isoleer." ’n Faagbiblioteek is gemaak met peptiede wat gegrond na volg, is slegs ’n kort bespreking van die gebruik van hierdie is op die peptiedvolgorde van die ligand. Hierdie biblioteek is biblioteke en die inligting wat daaruit verkry is. met plaatjies geinkubeer en die bindende fage m.b.v. die trom­ bienreseptor-agonis geëlueer (figuur 3). Fage wat dus aan die Peptiedbiblioteke trombienreseptor bind, is geëlueer en getoets vir inhibisie van die Peptiede wat 7, 10, 15 o f 38 lukraakgekose aminosure as liniêre trombienreseptor. Die aminosuurvolgorde van die blootgelegde peptiede op die oppervlak van filamenteuse fage blootlê, word inhiberende peptiede is daarna bepaal en peptiede met dié 120 ISSN 0251421-3486 = SA T ydskrif vir Natuurwetenskap en Tegnologie 18, no. 3 1999 79 daarna die totale RNA uit die milt en beenmurg van die konyne ge-ekstraheer. Die RNA wat vir die K-ligte en y-swaarkettings kodeer, is ge'isoleer, na cDNA getranskribeer, en in ’n faagmied ingebou om ’n F„|,-biblioteek te maak. Ses-en-veertig faagklone wat aan PAl-1 bind, is m.b.v. hierdie biblioteek ge'isoleer. Peptiedbiblioteek Hierdie antiliggaamfragmente op die fage kan weer van die T rombienreseptor mantelprotei'en vrygestel w ord .'' Eerstens kan ’n stopkodon tus- .a sen die geen wat kodeer vir die antiliggaamfragment en dié vir die faag-mantelproteien ingebou word. Wanneer die stopkodon onderdruk word, bly die antiliggaamfragmente aan die mantelprotei'en gebind, so nie, word dit vrygestel.'”̂ Tweedens Inkubering kan die geen wat kodeer vir die antiliggaamfragment m.b.v. ’n restriksie-ensiem van die manteiprote'iengeen gesny word en in ’n ander vektor wat geskik is vir die sekresie van die antilig­ IQ Q Q Elueer met 'n peptied- agonis van die gaamfragmente, herkloneer word,“’ Geïmmobilíseerde plaatjies trombienreseptor cDNA-biblioteke F iguur 3: Seleksie van 'n peptiedantagonis van die trombien­ reseptor. Die cDNA-biblioteke vind al hoe meer toepassing in trombose- en hemostase-navorsing. Die totale mRNA van ’n sel, orgaan of organisme word voorberei, getrutranskribeer na cDNA en in die aminosuurvolgordes is gemaak. faaggeen van GP III o f IV ingebou. In die laasgenoemde geval Urokinase-tipe plasminogeen-aktiveerder (uPA) aktiveer fibri- word die vreemde prote'ien aan die karboksie-terminale kant van noliese deur plasminogeen na plasmien om te skakel. Sy reseptor GP VI geheg. Dit is ook die mees praktiese manier om die op bloedselle, die urokinasereseptor, reguleer die fibrinolitiese uitdrukking van ’n cDNA-biblioteek te verseker, aangesien die sisteem en bemiddel onder meer die metastasesering en in- vreemde cDNA met hulle 5 ’-kante aan die faag-DNA heg. Die dringing van tumorselle. ’n Nuwe teiken vir die terapeutiese voordeel is dat die 5’-kant nie ribosoombindingsetels het nie en behandeling van tumore is dus om die binding van uPA aan sy dat daar alreeds translasionele stopseine aan die 3 ’-kant is. Dit reseptor te blokkeer.'^ Deur gebruik te maak van ’n 15-amino- beteken dat die stoppe nie die blootlegging van die vreemde suur lukraak peptiedbiblioteek, is twee peptiede wat sterk aan die prote'ien op die mantelprotei'en kan be'mvloed nie.'^ ’n Voorbeeld urokinasereseptor bind, geïsoleer. Dit was interessant dat nie een hiervan is ’n cDNA-biblioteek wat gemaak is van die haakwurm van die twee peptiede se aminosuurvolgordes ooreengestem het Ancylostoma caninum. Hierdie biblioteek is in pD0NG6-faag- met die bindingsetels van uPA aan die reseptor nie. In hierdie miedvektore gekloneer en fage wat aan tripsien o f faktor Xa geval is daar dus twee peptiedligande van sel-oppervlakreseptore bind, is geselekteer. ’n Nuwe tripsieninhibeerder, struktureel ver- geïdentifiseer wat kan lei tot die ontwikkeling van inhibeerders want aan die Kunitz-tipe serumprotease-inhibeerders, en ’n fak­ van die reseptor.'^ tor Xa-inhibeerder is uit hierdie biblioteek ge'isoleer.'* Plaaslik het ons ’n lukraak sikliese heptapeptiedbiblioteek ge­ bruik om heptapeptiede wat aan a-trom bien bind, te isoleer. Proteienbiblioteke Hierdie peptiede herken spesifiek a-trom bien, maar inhibeer nie die funksie daarvan nie wat daarop dui dat hierdie peptiede aan In die geval van prote'ienbiblioteke word 'n hele prote'ien op die ’n epitoop bind wat nie die proteolitiese aktiwiteit van trombien oppervlak van die filamenteuse faag blootgelê. Hierdie prote'iene beïnvloed nie. Hierdie tipe peptiede kan aan radioaktiewe kan varieer van klein prote'iene soos groeihormoon'"^ en jodium gebind word en kan moontlik gebruik word om in vivo- pankreas-tripsieninhibeerder-® tot groter prote'iene soos alkaliese trombusvorming m.b.v. sintillasiekamera-tegnieke op te spoor.'^ fosfatase^' en a-laktam ase.-- Prote'ien-struktuurfunksiestudies Trombien word in ’n trombus vasgevang'"* en hierdie peptiede vind baie hierby baat, veral nadat verskeie groepe al aangetoon kan dan spesifiek aan trombien in die trombus bind waar dit het dat ’n groot verskeidenheid prote'iene op filamenteuse fage m.b.v. ’n sintillasiekamera opgespoor kan word. blootgele kan word sonder 'n verlies aan funksionele akti- witeit.'^'^^ Die tegniek het tot gevolg dat 'n groot aantal mutante prote'iene gegenereer kan word en verskaf ’n goeie seleksie- Antiliggaambiblioteke strategie om mutante met die gewenste fenotipe te identifiseer. In In hierdie geval word antiliggame o f gedeeltes daarvan op die die geval van groeihormoon byvoorbeeld, het die seleksie op oppervlak van fage blootgelê. Die doel is om veral mens- ge'immobiliseerde reseptore gelei tot die identifikasie van mu­ antiliggame te kloneer wat dan as geneesmiddels ontwikkel kan tante met ’n hoër affmiteit vir sy reseptor as in die geval van die word o f in navorsing gebruik kan word. Soortgelyk aan die B-sel oorspronklike prote'ien.''’ van die immuunstelsel wat antiliggaamgene bevat, en hulle op sy In trombose en hemostase is ensieme soos PAl-1 op fage bloot- oppervlak blootstel, word antiliggaambiblioteke so gemaak dat gelê. PAI-1 wat op fage blootgele word, behou sy vermoë om mens-antiliggame o f gedeeltes daarvan op die oppervlak van tPA te inhibeer.--’ ’n Biblioteek van PAI-1-mutante is gebruik om fage blootgestel word. Daardie fage wat aan die antigeen bind, te bepaal watter aminosuurvolgorde van PAI-1 aan trombien word dan geselekteer. bind,^^ Met behulp van antiliggaambiblioteke kan Fab-fragmente teen Jespers et a l} ‘' het ’n biblioteek van lukraak variante van komplekse, biologiese antigene gei'soleer word. In ’n poging om staphilokinase (SAK) gemaak en op dié wyse die epitope van 2 antiliggame teen plasmingeen-aktiveerder-inhibeerder-tipe 1 te monoklonale antiliggame teen staphilokinase (SAK) bepaal. selekteer, het Lang et al? ' soos volg te werk gegaan: Hulle het ’n Hierdie variante bevat enkel, dubbel o f trippel aminosuur- konyn met die a-granule van mens-plaatjies gefmmuniseer en mutante. Die werkswyse was om die biblioteek negatief te 121 80 122 ISSN 0254-3486 = SA T ydskrif vir N atuurwetenskap en Tegnologie 18, no. 3 1999 selekteer vir SAK-variante wat nie aan die monoiilonale affinities or activities. The separation o f phage particles antiliggame bind nie. Uit hierdie geseleicteerde fage is dan expressing different peptide inserts on the phage surface protein variante geseiekteer wat aan plasmien bind. Na verskeie is accomplished by affinity selection. seieksierondtes, is die aminosuurvolgorde van die geselekteerde fage bepaal. Die aminosuurvolgorde van daardie mutante wat die THE USE OF PHAGE DISPLAY IN THROMBOSIS AND hoogste voorkoms het, weerspieël dan die funksionele epitoop HAEMOSTASIS van die antiiiggaam. Scientists already use peptide, antibody, cDNA and protein phage libraries to identify and develop effective antithrombotics. SAM EV A TTIN G Peptide libraries were used to select peptides that bind to and Hierdie oorsig bespreek die tegnieke van peptied- en proteïen- inhibit the glycoprotein llb/llla-receptór on platelets and an blootlegging op fage, asook variasies daarvan. Toepassings op inhibitor o f the thrombin receptor on platelets was also isolated. die gebied van trombose en hemostase word ook gegee. Daar Both these peptides inhibit platelet aggregation. A peptide that moet egter besef word dat hierdie tegniek ook baie ander inhibits the urokinase receptor that may have anti-tumor toepassings buite die veld van trombose en hemostase het. So activities was also selected. We selected thrombin-binding byvoorbeeld word die tegniek gebruik om geneesmiddels en peptides that can be used to detect in vivo thrombus formation gene in geselekteerde weefsels te plaas deur peptiede te selekteer with a scintillation camera. wat vaskulêre selmerkers identifiseer.^* Dit word ook gebruik in Antibody libraries are formed by displaying antibodies or parts die ontwikkeling van entstowwe.^* Alhoewel hierdie tegniek nog o f antibodies on filamentous phages. In this way, antibodies to relatief nuut is, het dit al baie bygedra tot die uitbouing van ons the plasminogen activator inhibitor-type 1 (P A l-l) were selected kennis van veral struktuurfunksie-verwantskappe van ensiem- from a mixture of antibodies. reaksies, inhibisie van in v/vo-prosesse en die ontwikkeling van In the case o f cDNA libraries, the total mRNA o f a cell, organ geneesmiddels.^' '*■ Dit is baie bemoedigend om waar te neem or organism is collected and reverse transcribed to cDNA. The dat hierdie tegnologie alreeds ’n hoofrol in die gebied van cDNA is then inserted into the phage genome to display foreign trombose en hemostase begin speel. Daar word verwag dat, soos proteins on the phage surface. An example is a cDNA library hierdie tegnieke meer toegepas word, ’n verdere kennisont- made from the hookworm Ancylostoma caninum. Inhibitors of ploffing in die gebied sal plaasvind. trypsin and factor X„ were selected form this library. Protein libraries were made where enzymes, such as PAI-I, was S U M M A R Y displayed on the surface o f phages. A library o f PAI-I mutants INTR OD U CTIO N was then used to determine which sequence in PAl-I binds to thrombin. The functional epitope o f two monoclonal antibodies Haemostasis is a protective mechanism that limits excessive to staphylokinase were also determined by using a library of blood loss. Thrombosis, on the other hand, is the antitype o f random variants o f SAK. haemostasis. Following blood vessel injury, when the blood Although the technique o f phage display is relatively new, its comes into contact with the sub-endothelium, the blood platelets adhere to sub-endothelial structures and become activated. use has already made remarkable contributions in the field of Simultaneous activation o f the coagulation factors form thrombosis and haemostasis. It is not only suitable to select thrombin. In concert, platelets and thrombin form a platelet plug specific binders to proteins but also expand our knowledge of to prevent excessive blood loss. Endothelial cells in the vicinity structure function relationships and to develop effective agents for treatment. We expect that the increasing use o f phage display o f the platelet plug secrete prostacyclin, nitrogen-oxide and fibrinolytic agents that limit the size o f the plug. Excessive will result in an explosion o f our knowledge o f the mechanisms activation o f platelets by atherosclerotic lesions overcome these involved in thrombosis. thromboregulatory mechanisms and the platelets and coagulation factors then become thrombotic offenders that occlude the blood LITER A TU U RV ERW Y SIN G S vessel at the site o f injury or form thrombo-emboli that can block 1. Marcus, A.J., Satier, L.B. (1993). Thromboregiilation: imilticelliilar blood vessels distal to the site o f injury. In view o f this, more modulation o f platelet reactivity in lieniostasis and thrombosis. FASEB effective anti-thombotic agents are being investigated. For this Journal, 7, 512-522. purpose, phage display technology is ideally suited because it is 2. Gimbrone, M.A. (1986). Vascular endothelium; N ature’s blood container. In: Gimbrone, M.A. (ed). I'asculur eudolhclium in haemostasis and a powerful and rapid method to isolate or design proteins with thrombosis (Churchill Livingstone, New York), pp. 1-9. specific anti-thrombotic functions. 3. Gimbrone, M.A. (1986). Vascular endothelium: N ature’s blood container. In: Gimbrone, M.A. (ed). Vascular endothelium in haemostasis and PHAGE DISPLA Y thrombosis (Churchill Livingstone. New York), pp. 1-9. 4. Smith, G.P. (1985). Filamentous fusion phage: novel expression vectors Phage display is a technique that expresses or “displays” a that display cloned antigens on the surface o f the viron. Science. 228. peptide or protein on the surface o f a filamentous phage. This is 1315-1317. accomplished by the insertion o f a gene or gene fragment in a 5. Baas, P.D. (1985). DNA replication o f single-stranded Escherichia coli phage surface protein gene. The insert will result in a fusion DNA phages, Biochim. Biophys. Acta.. 825. 111-139. peptide/protein on the phage surface, provided that the reading 6. Doorbar, J.. Winter. J.A. (1994). Isolation o f a peptide antagonist to the frame is correct and that the insert does not interfere with the thrombin receptor using phage d isp lay ../ Mol. Biol., 244, 361-369. essential functions o f the surface protein. If the peptide is well 7. Scott, J.K., Smith, G.P. (1990). Searching for peptide ligands with an exposed on the surface it will be available to act as a ligand, epitope library. Science 249, 386-390. enzyme, and immunogen or otherwise actively participate in a 8. Philips, D.R., Chiiro, I F., Pari.se, L.V.. Fitzgerald, L A. (1988), The platelet membrane glycoprotein llb /llla complex. Blood. 71, 831-834 biochemical process. The insertion o f random oligonucleotide 9. Ke, S-H., Coombs, G.S., Tachias. K.. Navre. M.. Corey, D R., Madison, sequences provides a means o f constructing extensive peptide E.L, (1997). Distinguishing the specificities o f closely related proteases, libraries that may be screened to select peptides with specific J. Biol. C hem .,212. 16603-16609 122 ISSN 0215243-3486 = SA T ydskrifv ir Natuurwetenskap en Tegnologie 18, no. 3 1999 10. Van Zonncveld, A-J., vaii den Berg, B.M.M., van Meijer, M.. Pannekoek, 19. Bass, S., Greene. R.. Wells. J,A. (1990). I lormone phage: an enreichment H. (1995). Identification o f functional interactions sites on proteins using method for variant proteins with altered binding properties. Proteins, 8, bacteriophage-displayed random epitope libraries. Gene, 78, 1097-1103. 309-314. 11. Vu T-K.H., Hung, D.T., W heaton, V.I., Coughlin, S.K., (1991). 20. Markland, W., Roberts, B.L., Saxona, M..1.. Gutterman, S.K... Ladner, R.C. Molecular cloning o f a functional thrombin receptor reveals a novel (1991). Design, construction and function o f a multicopy display vector proteolytic mechanism o f receptor activation. Cell, 64, 361-369. using fusions to the major coat protein o f bacteriophage. Gene, 109, 13-19. 12. Goodson. R.J., Doyle, M.V., Kaufman, S.E., Rosenberg, S. (1994). High- 21. McCafferty, J., Hackson. R.H.. C'hriswell. I)..l. (1991) Phagc-cnzymes affinity urokinase receptor antagonists identified with bacteriophage expression and atTuiity chromatography o f functional alkaline peptide display, Proc. Natl. Acad. Sci. USA, 91, 7129-7133. phosphatase on the surface o f bacteriophage. Frot. Eng.. 48. 955-961. 13. Kotze, H.F., Badenhorst, P.N., Lamprecht, S., Meiring, S.M., Van Wyk, 22. Soumillion, P.. Jespers. L., Bouchet, M.. Marchand-Brynaert, J., Winter, V,, Nuyts, K., Stassen. J.M ., Vermylen, J., Deckmyn. II. (1995). G ., Fastrez. J. (1994). Selection o f ix-lactamase on filamentous Prolonged inhibition o f acute arterial thrombosis by high dosing o f a bacteiophage by catalytic ac tiv ity ../ Mol. ISiol.. 237, 415-422. monochlonal antiplatelet glycoprotein lib/IIIa antibody in a baboon 23. Pannekoek, H., van Meijer, M.. Scleef. R.. Loskutoff. D.J., Barbas III. model. Thromh. Haemosl., 74(2), 751-757. C.F. (1993). Functional display ofhunian plasniinogen-activator inhibitor 14. Stubbs, M.T., Bode, W. (1993). A player in many parts: the spotlight on (P A l-l) on phages: novel perspectives for sturcture-function analysis by thrombin’s structure, Thromb. Res , 69, 1-58. error-prone DNA synthesis. Gene. 128. 135-140. 15. Hoogenboom, H R., Griffiths, A.D., Johnson, K.S.. Chriswell, D.J., 24. Jespers, L.. Jenne, S., Lasters. I.. Collen. 1). (1997). Epitope mapping by Hudson, P., W inter, G. (1991). Multi-subunit proteins on the surface o f filantous phage m ethodologies for displaying antibody (Fab) heavy and negative selection o f randomized antigen libraries displayed on light chains. Nucleic Acids Res, 19, 4133-4137. filamentous phage, J.Mo/. Hiol.. 269, 704-718. 16. Barbas III, C.F., Lerner, R.A. (1991). Combinatorial immunoglobulin 25. Barry, M.A., Dower. W.J.. Johnson. S.A. (1996). Toward cell-targeting libraries on the surface o f phage: rapid selection o f antigen-specific Fabs, gene therapy vectors: .selection o f cell-binding peptides from random Methods. 2, 119-124. peptide-presenting phage libraries. Nature Med.. 2, 299-305. 17. Stubbs, M .Ï .. Bode, W. (1993). A player in many parts: the spotlight on 26. Barbas III, C.F.. Burton, D R. (1996) Selection and evolution o f high- throm bin's structure, Thromb. Res., 69, 1-58. affinity human anti-viral antibodies. Tihtech. 14, 230-234. 18. Gramatikoff, K., Georgiev, O., Schaffner, W. (1994). Direct interaction 27. Katz, B.A. (1997). Structural and mechanistic determinations o f affinity rescue, a novel filamentous phage technique to study protein-protein and specificity o f ligands discovered or engineerd by phage display, Annu. interactions. Nucleic. Acids., 22, 5761-5762. Rev. Biophys. Biomol. Struct.. 26, 27-45. Muriel Meiring is Eersle Mediese Nalitiirwelenskaplike in die Departemeni Hentalologie en Selhiologie aan die Universiteil van die Oranje-Vry.'itaal. Sy he! haar B.Sc.-graad en B.Sc.- Honneursgraad in Fisiotogie aan die Potchefstroomse Universileit vir Christelike Hoer Onderwys hehaal. Sy het n M.Med.Sc.-graad met lo f en n Ph.D.-graad, albei in Hema- tologie, aan die Universiteil van die Oranje-Vrystaat behaal. In 1997 het 'n hettrs van die Vlaantse Gemeenskap ontvang om haar postdoktorale slttdie in ISelgie te doen. Sy is ottlettr en mede-outeur van v y f joernaatarlikels. 123 S-11 124 Thrombosis Research 107 (2002) 365–371 Regular Article In vitro effect of a thrombin inhibition peptide selected by phage display technology Muriel S. Meiringa,*, Derek Litthauerb,1, Jolan Hársfalvic,2, Veronica van Wyka,3, Philip N. Badenhorsta,3, Harry F. Kotzéd,4 aDepartment of Haematology and Cell Biology, University of the Free State, P.O. Box 339(G2), Bloemfontein 9300, South Africa bDepartment of Microbiology, Biochemistry and Food Sciences, University of the Free State, Bloemfontein, South Africa cDepartment of Clinical Biochemistry and Molecular Pathology, University of Debrecen, P.O. Box 40, Debrecen H-4012, Hungary dWallace H. Coulter Department of Biomedical Engineering, Emory University, Atlanta, GA, USA Received 1 August 2002; received in revised form 24 October 2002; accepted 5 November 2002 Abstract A repeated selection of phages from a cyclic heptapeptide phage display library resulted in the enrichment of phages that bind to human a-thrombin. One clone of the binding phages that competed with PPACK for binding to thrombin and that had the best binding characteristics was chosen. The amino acid sequence of the peptide displayed on this phage was determined and a peptide with the sequence, Cys-Asn-Arg- Pro-Phe-Ile-Pro-Thr-Cys was synthesised. This peptide, thrombin-inhibiting peptide (TIP), is a full competitive inhibitor of thrombin with an inhibition constant (Ki) of 0.4974 mM. It lengthened the thrombin time and inhibited thrombin-induced platelet activation and the platelet release reaction, both in a dose-dependent manner. It also reduced platelet adhesion onto a human microvascular endothelial matrix in the parallel plate flow chamber under both arterial and venous shear conditions. Thus, we have selected and synthesised a cyclic heptapeptide that competes with PPACK to bind to thrombin and that can be developed as a direct antithrombin. D 2002 Elsevier Science Ltd. All rights reserved. Keywords: Thrombin antagonists and inhibitors; Thrombin; Peptide library; Recombinant fusion proteins; Bacteriophage M13 genetics; Phage display system 1. Introduction toward a target protein. Its power lies in the combination of functional activity with genetic information. Although not The technique of phage display is a powerful and quick yet extensively, this technique has been used in the field of way to select peptides with specific affinities and activities thrombosis and haemostasis. Peptide libraries were used to select peptides that bind to and inhibit the thrombin Abbreviations: PPACK, D-Phe-Pro-Arg-CH2Cl2; TIP, thrombin inhib- receptor on the platelet membrane and the urokinase ition peptide; Ki, inhibition constant; TAP, tick anticoagulant peptide; TBS, receptor [1,2]. Potent anticoagulants have also been Tris-buffered saline; E. coli, Escherichia coli; IPTG, isopropyl h-D- derived by targeting the tissue factor/factor VIIa complex thiogalactoside; X-Gal, 5-bromo-4-chloro-3-indonyl-h-D-galactoside; SM, with naı̈ve peptide libraries displayed on the M13 phage skimmed milk; HRP, horseradish peroxidase; OPD, ortho-phenylenedi- amine; ELISA, enzyme-linked immunosorbent assay; BSA, bovine serum [3]. albumin; ADP, adenosine diphosphate; FITC, fluorescein isothiocynate; PE, In this study, we used a cyclic heptapeptide phage phycoerythrin; HMEC, human dermal microvascular endothelial cells; library to select phages that bind to and inhibit a-thrombin. DNA, dioxyribonucleic acid. Thrombin was chosen for two reasons. First, it is a multi- * Corresponding author. Tel.: +27-51-405-3593; fax: +27-51-444- functional serine protease that plays a key role in thrombus 1036. E-mail addresses: GNHMSMM@med.uovs.ac.za (M.S. Meiring), formation and is an important participant in the sequelae Litthad.sci@mail.uovs.ac.za (D. Litthauer), Harsfalv@jaguar.dote.hu thereof [4]. It is also mainly responsible for maintaining an (J. Hársfalvi), GNHMPNB@med.uovs.ac.za (P.N. Badenhorst), intricate balance between the interactive elements of the hkotze@emory.edu (H.F. Kotzé). 1 vessel wall, coagulation factors and platelets [4]. Thus,Tel.: +27-51-401-2396; fax: +27-51-444-3219. 2 thrombin cleaves fibrinogen and activates factor XIII toTel.: +36-52-431-956; fax: +36-52-417-631. 3 Tel.: +27-51-405-3043; fax: +27-51-444-1036. form insoluble fibrin. It activates platelets and amplifies its 4 Tel.: +1-404-727-8009; fax: +1-404-727-0777. own generation (positive feedback) by activation of factor 0049-3848/02/$ - see front matter D 2002 Elsevier Science Ltd. All rights reserved. PII: S0049 -3848 (02 )00349 -3 124 125 366 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 V and factor VIII [5]. Normally, thrombin generation is 2.1.2. Isolation of thrombin-binding phages under stringent physiological control, because on the sur- A cyclic 7-mer phage display library was purchased from face of intact endothelial cells, thrombin in complex with New England Biolabs (Beverley, MA, USA). Streptavidin thrombomodulin activates protein C, which in turn inacti- magnetic particles (1.5 mg; Boehringer Mannheim, Ger- vates factors Va and VIIIa, resulting in a decline in many) were first blocked for 1 h with 2% skimmed milk thrombin generation (negative feedback). The thrombin (DIFCO, Detroit, MI, USA) in Tris-buffered Saline (TBS) at concentration in plasma is low due to inactivation by room temperature. Phages (2 1011) were added to half of circulating antithrombin III or heparin cofactor II bound the blocked streptavidin magnetic particles and incubated to dermatan sulphate at the endothelial cell surface [6]. for 1 h at room temperature. Ten micrograms of the Second, studies have shown that inhibition of thrombin biotinylated thrombin was added to the other half of the with recombinant hirudin, or its formation by inhibition of magnetic particles, and incubated for 30 min at room factor Xa with recombinant tick anticoagulant peptide temperature. The phage supernatant was then transferred (TAP) effectively inhibits coagulation and arterial throm- to the thrombin-bound magnetic particles and incubated bosis [7,8]. overnight at room temperature. After incubation, the mag- The central role of thrombin in thrombogenesis makes it netic particles were washed 10 times with TBS–Tween an ideal target for antithrombotic intervention [4]. Under (0.1% Tween-20 in TBS) to remove the nonbinding phages. pathological conditions, excessive thrombin generation Phages bound to thrombin were eluted with 50 Al of 0.1 M occurs. Since thrombin is formed continuously from pro- glycine, pH 2, for 15 min at room temperature and imme- thrombin at the surface of a thrombus, the rate of thrombin diately neutralised with 125 Al of 1 M Tris–HCl, pH 8. The inhibition, by a thrombin inhibitor, should at least be as titre of the eluted phages was estimated and half of the efficient as the rate of thrombin formation. Indirect, as well eluted fraction was used for amplification. Three additional as direct inhibitors of thrombin represent potentially useful rounds of panning were performed. During this last three drugs for the treatment of both venous and arterial throm- rounds, the thrombin-bound phages were eluted with 8 mg bosis. The clinical effectiveness of the indirect thrombin r-hirudin (Hoechst-Marion-Rousell, AGS Frankfurt, Ger- inhibitors such as heparin, and low-molecular-weight hep- many). arin, is hampered by their inability to inhibit clot-bound Phages were amplified after each round of panning by thrombin, and by the fact that they may cause thrombo- infection of Escherichia coli (E. coli) ER2537 cells and cytopenia [9]. Direct thrombin inhibitors of thrombin purified from the supernatant by two polyethylene glycol include hirudin, hirulog, etc. They are small peptides that precipitations. A fraction of the eluted phages of panning inhibit only thrombin. These compounds are very potent round 4 was subjected to serial 10-fold dilution with Luria- inhibitors, and there are now indications that hirudin may Broth (LB) culture media (DIFCO). To 10 Al of each of the be too potent and may promote bleeding tendencies when phage dilutions, 200 Al of E. coli cells in their log phase of the dosage is too high [10]. However, an advantage of growth was added and the mixture was incubated for 5 min these smaller peptides is that they are not immunogenic at room temperature to allow the phages to infect the E. coli [11]. cells. Each dilution was spread onto agar plates containing There is thus still a need for small, direct thrombin isopropyl h-D-thiogalactoside (IPTG) and 5-bromo-4- inhibitors, since small peptides are nonimmunogenic and chloro-3-indonyl-h-D-galactoside (X-Gal) and incubated nontoxic and direct thrombin inhibitors inhibit clot-bound overnight at 37 jC. Since the library phages were derived thrombin, which is largely involved in thrombosis [8]. In from the common cloning vector M13mp19, which carries this study, we selected a cyclic heptapeptide named throm- the LacZa gene, phage plaques with inserts appear blue bin inhibition peptide (TIP) from a cyclic heptapetide phage when plated on media containing X-Gal and IPTG. library by using phage display technology and tested its effect on thrombin function. 2.1.3. Phage ELISA A 96-well ELISA plate (Maxisorp surface, Nalge Nunc International, Roskilde, Denmark) was coated overnight 2. Materials and methods with 100 Ag/ml a-thrombin in 0.1 M NaHCO3 binding buffer at 4 jC in a humidified container. The wells were 2.1. Selection of thrombin-binding phages blocked for 2 h at room temperature with 200 Al 4% skimmed milk (SM) solution. Simultaneously, 96 blue 2.1.1. Biotinylation of human a-thrombin phage colonies were picked from the agar plate and grown Human a-thrombin (a kind gift from Dr. Pötzsch, Ker- overnight at 37 jC in 1 ml LB medium containing 1:100 ckhoff Klinik, Bud Nauhem, Germany) was biotinylated diluted E. coli cells from a preculture. Hundred microliters with EZ-Linkk sulfo-NHS-LC-Biotin (Pierce Chemical, of this supernatant was added to each well together with 100 Rockford, IL, USA) according to the instructions of the Al of 4% SM and incubated at room temperature for 2 h. manufacturer. Two molecules of biotin were bound to one Bound phages were detected after 1-h incubation with a molecule of thrombin. polyclonal anti-M13 phage horseradish peroxidase (HRP)- 125 126 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 367 conjugated antibody (Pharmacia, USA) and visualised by start the reaction. The total volume of the reaction was 200 adding ortho-phenylenediamine (OPD, Sigma, St. Louis, Al. All reagents were diluted in the reaction buffer, HEPES- MO, USA). The reaction was stopped with 4 M H2SO4 and BSA (50 mM HEPES, 0.05% BSA, pH 7.4). Kinetic absorbance determined at 490 nm. Between each incubation parameters, Km and Vm, were determined by least squares step, the plates were washed three to nine times with TBS– nonlinear regression using a single binding site model. Tween (0.1% Tween-20). The six phage colonies with the best thrombin-binding 2.3.3. Thrombin times characteristics were amplified by infecting E. coli cells and The thrombin time of a dilution series of the peptide in purified by polyethylene glycol precipitation. A dilution plasma of healthy volunteers was measured with a Cobas series of these individual phage clones were tested for Fibro fibrinometer (Diagnostica Stago, Asnières, France). thrombin binding in an ELISA. The same technique was Reagents were supplied by Dade Behring (Newark, DE, used as described for the previous ELISA. Again the best USA). binding phages were used for further testing. To determine if In vitro platelet aggregation: Blood of healthy volunteers these colonies bind to the active site of thrombin, an were collected in 3.8% sodium citrate (9:1 v/v). The in vitro inhibition ELISA was done where different concentrations aggregation response of platelets at different concentrations of D-Phe-Pro-Arg-CH2Cl2 (PPACK, a kind gift from Prof. of the peptide was measured turbid metrically in a Monitor S.R. Hanson, Emory University, School of Medicine, At- IV Plus aggregometer (Helena Laboratories, USA). The lanta, GA, USA) was added to the thrombin-coated wells. method is fully described [10]. The aggregation response, After incubation for 15 min, 5 1010 phages in 2% SM measured after 4 min following addition of ADP (20 AM) or were added into each well and incubated for 2 h. Bound collagen (0.05 g/l), was related to the difference in light phages were again detected after 1-h incubation with the transmission between platelet-rich and platelet-poor plasma anti-M13 antibody and visualisation done with OPD and and expressed as a percentage. absorbance measured at 490 nm. 2.3.4. Flow cytometry 2.1.4. Thrombin times Fluorescence flow cytometry was used to measure the Thrombin Times were done where a dilution series of inhibition of platelet activation and degranulation at differ- phages of colony A were added to 100 Al of normal human ent concentrations of the peptide. Samples were analyzed by plasma. For the control, phosphate-buffered saline (PBS) two-color fluorescence using fluorescein isothiocynate was added to the plasma. (FITC)- and phycoerythrin (PE)-conjugated antibodies. Whole blood was added to polystyrene tubes containing 5 2.2. Nucleotide sequencing Al thrombin (1 U/ml, Sigma). Gly-Pro-Arg-Pro (1.25 mM final concentration, Sigma), 2.5 Al, was added immediately Phage DNA of the thrombin-inhibiting phage colony was to delay fibrin polymerization. Ten microliters of TIP prepared using the ph.D-C7C phage display peptide library peptide (1000, 250 and 62.5 Ag/ml) and HEPES buffer (1 kit (New England Biolabs) and sequencing reaction were performed according to the DY Enamic ET terminator cycle sequencing premix kit (Pharmacia, Buckinghamshire, UK). 2.3. Peptide analysis 2.3.1. Peptide synthesis A peptide sequence that correlated with the peptide sequence of the best thrombin-binding phage clone was called thrombin inhibition peptide (TIP) and synthesized by Ansynth Service (Roosendaal, The Netherlands). 2.3.2. In vitro enzyme inhibition Chromogenic substrate hydrolysis was detected using the Bio-Tech EL312e microplate ELISA reader (Biotek instru- ments, VT, USA) at 405 nm equipped with a microplate Fig. 1. Binding curves of different concentrations of phages to a-thrombin. mixer in addition to a kinetic module software (Kinetic An ELISA plate was coated overnight with 100 Ag/ml human a-thrombin. Calc, Biotek). In several kinetic assays, different fixed TIP A dilution series of phages of the six best binding phage colonies (colonies concentrations (0–0.956 mM) were incubated with a- A to F) were added to the wells of the ELISA plate and incubated. An anti- M13 phage horseradish peroxidase (HRP)-conjugated antibody was added thrombin (0.049 pM) for 10 min followed by addition of to the wells and the thrombin-binding phages were visualised by adding different chromogenic substrate, chromozym TH (Roche ortho-phenylenediamine (OPD). The reaction was stopped with 4 M H2SO4 Biochemicals, Germany) concentrations (1.95–125 AM) to and absorbance determined at 490 nm. 126 127 368 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 Fig. 4. Kinetics of the inhibition of the a-thrombin-induced chromogenic substrate (chromozym TH) hydrolysis by TIP. The reactions were initiated by different substrate concentrations (0, 1.95, 3.91, 7.81, 15.63, 31.25, 62.5 Fig. 2. Prevention of phage colony A from binding to thrombin by PPACK and 125 AM) to a mixture of 0.049 pM a-thrombin and different TIP (D-Phe-Pro-Arg-CH2Cl2). Different concentrations of PPACK were added concentrations (0, 0.239, 0.478 and 0.956 mM). The Km value was to the thrombin-coated wells. After incubation for 15 min, 5 1010 phages calculated by one-site binding nonlinear regression (r2>0.990). The Ki in 2% skimmed milk (SM) were added into each well and incubated for 2 h. value was calculated from the x-intercept, which is Ki. Linear regression Bound phages were detected after 1-h incubation with the anti-M13 of the date yields a Ki of 0.4974 mM. The points of each figure are the antibody and visualisation done with OPD and absorbance measured at 490 meanF standard deviation of six independent experiments. nm. ilarly FITC-labelled isotopic controls were prepared using mM, pH 7.4) was added to a final volume of 50 Al. The whole blood. These controls were run with each experiment samples were gently mixed and incubated for 10 min at to determine nonspecific antibody binding. Controls of rest- room temperature. This was followed by the addition of 5 Al ing platelets were also used for each antibody to define a PE-conjugated CD41 antibodies against the platelet marker GPIIb/IIIa (Beckman Coulter). For measurement of platelet activation and degranulation, 5 Al FITC-conjugated anti- CD62P (P-selectin; Beckman Coulter) antibody or anti- CD63 (Lisosomal glycoprotein, Beckman Coulter) was added. The samples were then incubated at room temper- ature for 30 min. Finally, each sample was diluted and fixed with 450 Al of 1% p-formaldehyde in saline before analysis. The labelled platelets were analysed using a FAC-Calibur flow cytometer (Becton Dickinson, USA). Forward and side scatter as well as green (FITC) and red (PE) fluorescence signals were acquired with logarithmic amplification. The data from 10000 cells were acquired and processed. Sim- Fig. 3. Effect of a dilution series of phages from phage colony A on the thrombin time. Different concentrations of phages from phage colony A were added to 100 Al of normal human plasma. The phages of this colony Fig. 5. Thrombin times of a dilution series of the peptide (A) and hirudin also lengthened the thrombin time in a dose-dependent manner. (B) in plasma of healthy volunteers. 127 128 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 369 Table 1 Arzeimittel, Norway) or a matrix of human dermal micro- The effect of TIP on the platelet release reaction and platelet activation after vascular endothelial cells (HMEC-1). The matrix was pre- stimulation of the platelets with thrombin in the presence of Gly-Pro-Arg- pared by 0.1 N ammonia detachment of cells cultured as Pro (n = 4) described previously [12]. HMEC-1 cell line was a kind gift Platelet release reaction Platelet activation (CD62 expression) (CD63 expression) of Drs. E.W. Ades and T.J. Lawley (Centres for Disease Control and Prevention and Emory University School of Negative control 8.0F 3.4 12.8F 8.8 Positive control 94.5F .3 88.3F 5.7 Medicine, Atlanta, GA). Cells were cultured and character- TIP (1000 Ag/ml) 26.3F 22.7 14.8F 6.7 ized as described previously [1,2]. Perfusions were carried TIP (250 Ag/ml) 39.5F 28.8 26.5F 11.3 out at 37 jC and by recirculating heparinized blood over the TIP (62.5 Ag/ml) 74.0F 14.5 67.3F 18.9 coated coverslips for 5 min. After perfusion, the coverslips were immersed in methanol and stained with May-Grün- wald-Giemsa. Platelet adhesion was evaluated with a light platelet population negative for activation markers. Positive microscope connected to an image analyser. An average of controls of activated, nontreated platelets were included in 30 fields per coverslip was analysed. Platelet adhesion was each assay. Platelets were gated on CD41-positive binding. expressed as percentage surface coverage by platelets. Platelet activation was calculated from the percentage of CD41- and CD62- or CD41- and CD63-positive platelets, respectively, with and without addition of the thrombin 3. Results inhibition peptide. Platelet adhesion with parallel plate flow chambers: The 3.1. Selection of phages that bind to thrombin inhibitory effect of TIP was studied at shear rates of 100 and 2600 s 1 using parallel plate flow chambers with appro- Repeated selection of phages from the cyclic heptapep- priate slit height (0.4–1 mm) and following a previously tide phage display library resulted in an enrichment of described method [11]. Coverslips were coated with either a phages that bind to human a-thrombin. After the fourth mixture of types I and III collagen (1 mg/ml, Nycomed round of enrichment (panning), individual colonies were Fig. 6. The inhibitory effect of TIP at shear rates of 100 and 2600 s 1 using parallel-plate flow chambers. Coverslips were coated with a matrix of human dermal microvascular endothelial cells (HMEC-1). Heparinized blood was recirculated over the coated coverslips for 5 min at 37 jC. After perfusion, the coverslips were stained and platelet adhesion was expressed as percentage surface coverage by platelets. 128 129 370 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 tested for its binding affinity to thrombin. The six strongest It thus had the potential to form a disulphide linked nano- binding colonies were identified and amplified in E. coli for peptide loop with restricted conformational freedom. We further characterisation. Two of the six colonies bound very used the cyclic peptide phage library for three reasons. First, strongly and in a dose-dependent manner to thrombin. They peptides that are based on the cyclic sequence are generally did not bind to uncoated wells, indicating that the binding more stable than those with a linear sequence. Second, a was very specific (Fig. 1). In a competition experiment with ligand sequence, where the imposed constraint allows for a PPACK, we found that PPACK prevented, in a dose- productive binding conformation, will bind more tightly dependent manner, phages of one of the colonies to bind than the same linear sequence due to improved binding to thrombin. The estimated IC 1150 was 0.8 Ag/ml at 5 10 entropy, and third, cyclic peptides are more resistant to phages/ml (Fig. 2). The phages of this colony also length- proteolysis than linear peptides [13,16–19]. After four ened the thrombin time in a dose-dependent manner (Fig. 3). rounds of enrichment (panning), two phage clones that bind By using specific primers, the sequence of the Cys-flanked the strongest to thrombin (Fig. 1) were tested for binding to heptapeptide insert of the thrombin-inhibitory phage clone thrombin in competition with PPACK. One of these two was deducted. The amino acid sequence was Cys-Asn-Arg- clones competed with PPACK to bind to thrombin (Fig. 2). Pro-Phe-Ile-Pro-Thr-Cys. We had the peptide synthesised This strongly suggested that these clones display the Phe- and named it thrombin inhibition peptide (TIP). Pro-Arg sequence on their surfaces. This sequence binds to the active site of thrombin [13]. This phage clone also 3.2. Determination of the inhibitory activity of TIP lengthened the thrombin time in a dose-dependent manner (Fig. 3). Because of their size, only limited number phages Since TIP contains the reverse sequence of the well- can be used in any one assay. In view of this, no further known thrombin active site binding peptide Phe-Pro-Arg results on inhibition could be obtained by using the phages. [13,14], it is possible that it acts as a second substrate of A cyclic peptide, TIP, with the sequence, Cys-Asn-Arg- which the hydrolysis cannot be detected in the assay. A Pro-Phe-Ile-Pro-Thr-Cys, to correspond with that displayed substrate analogue, which binds reversibly to the enzyme, on the phage surfaces, was synthesised. It is of interest to excluding the substrate, acts as a purely competitive inhib- note that the active site binding sequence, Phe-Pro-Arg, was itor. Competitive inhibitors have the effect of increasing the reversed in TIP. Thrombin-binding peptides with this apparent Km of the reaction [15]. This is indeed the case in reverse Phe-Pro-Arg sequence were also selected from Fig. 4, where the apparent Km is plotted against different peptide-encoded peptide libraries by Vagner et al. [20]. TIP concentrations. The Km value increases with increasing In substrate hydrolysis reactions, TIP acts as a second inhibitor concentrations. substrate of which the hydrolysis cannot be detected in the TIP lengthened the thrombin time in a dose-dependent assay. The system behave as if it a full competitive inhibitor manner (Fig. 5A). For purposes of comparison, the length- (Fig. 4). A plot of Km vs. TIP concentration is a straight line ening of the thrombin time at different concentrations of r- (Fig. 4). The Ki value is calculated at 0.4974 mM. This is hirudin is also summarised in Fig. 5B. TIP had no effect on much higher than the Ki value for the inhibition of thrombin in vitro platelet aggregation in response to ADP or collagen. of other published thrombin inhibitors, such as hirudin, The data are not shown. Fluorescence flow cytometry, on hirunorm IV, argatroban, efegatran and anophelin, which the other hand, showed that TIP inhibited the release are all in the nanomolar range [21–23]. The Ki value of TIP reaction and activation of platelets in a dose-dependent is also higher than that of the Arg-Gly-Pro-D-Phe peptide in manner when the platelets were stimulated with thrombin the literature, which also contains the reverse sequence of (Table 1). the active site binding peptide of thrombin, namely Phe-Pro- TIP (25 Ag/ml) inhibited platelet adhesion onto human Arg [20]. microvascular endothelial matrix in the parallel plate flow TIP lengthens the thrombin time in a dose-dependent chamber under both arterial and venous shear conditions, manner. The effect was, however, much less pronounced i.e. 2600 and 100 s 1 (Fig. 6). Like hirudin and TAP, it did than that of hirudin (Fig. 5). As was expected, TIP had no not inhibit platelet adhesion onto the collagen-coated cover- effect on ADP or collagen-stimulated platelet aggregation, slips in the parallel plate flow chamber under similar similar to results obtained with hirudin. On the other hand, it conditions. The data are not shown. inhibited the release reaction and activation of platelets when the platelets were stimulated with the thrombin mimetic peptide, Gly-Pro-Arg-Pro (Table 1). 4. Discussion TIP has antithrombin activity in spite of the fact that the active site binding sequence Phe-Pro-Arg was reversed. In an attempt to identify peptides that may inhibit the This is also shown in the literature that this reverse sequence function of thrombin, phages from a phage display library also inhibits thrombin. However, the peptide Arg-Gly-Pro- were selected on their binding characteristics to human a- D-Phe in the literature inhibits thrombin much stronger [20]. thrombin. A heptapeptide phage display library, in which It is known that thrombin cleaves the peptide bond between the heptapeptides were flanked by two cysteines, was used. Arg-Pro in the bifunctional inhibitor D-Phe-Pro-Arg-Pro- 129 130 M.S. Meiring et al. / Thrombosis Research 107 (2002) 365–371 371 (Gly) 4-hirudin-thrombin complex [19]. Thrombin also [7] Kotzé HF, Lamprecht S, Badenhorst PN, Roodt JP, Van Wyk V. hydrolyses fibrinogen by the selective cleavage of 4 Arg- Transient interruption of arterial thrombosis by inhibition of factor Xa results in long-term antithrombotic effects in baboons. Thromb Gly bonds and it cleaved the chromogenic substrate chro- Haemost 1997;77:1137–42. mozym TH at the carboxyl end of Arg. So it seems that [8] Kotzé HF, Lamprecht S, Badenhorst PN. A four-hour infusion of thrombin cleaves peptides at the carboxyl end of an Arg. In recombinant hirudin results in long-term inhibition of arterial-type this peptide TIP, thrombin may also cleave it at the Arg-Pro thrombosis in baboons. Blood 1995;85:3158–63. bond. If this is the case, one may argue that the advantages [9] Olsen ST, Bjork I. Regulation of thrombin activity by antithrombin and heparin. Semin Thromb Hemost 1994;20:373–408. of cyclic peptides may be get lost, because the peptide is [10] Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound cleaved and act as a linear peptide. This may also be the thrombin is protected from inhibition by heparin–antithrombin III but reason for the low binding and weak inhibition (high Ki is susceptible to inactivation by anti-thrombin III-independent inhib- value) of TIP. Another reason may be the fact that TIP is a itors. J Clin Invest 1990;86:385–91. bigger peptide than PPACK and therefor do not bind as [11] Markwardt F. Hirudin and derivates as anticoagulant agents. Thromb Haemost 1991;66:141–52. strong because of its allosteric conformation. [12] Harsfalvi J, Stassen JM, Hoylaerts MF, Van Houtte E, Sawyer RT, TIP also inhibits platelet adhesion onto human micro- Vermylen J, et al. Calin from Hirudo medicinalis, an inhibitor of von vascular endothelial matrix in the parallel plate flow cham- Willebrand factor binding to collagen under static and flow condi- ber under both arterial and venous shear conditions. It, tions. Blood 1995;5:595–603. however, does not inhibit platelet adhesion to the colla- [13] Maraganore JM, Bourbon P, Jablonski J, Ramachandran KL, Fenton JW. Design and characterization of hirulogs: a novel class of bivalent gen-coated surface collagen in the flow chamber. Neither peptide inhibitors of thrombin. Biochemistry 1990;29(30):7095–101. PPACK, nor hirudin, has an effect on platelet adhesion to [14] Chirgadze NY, Sall DJ, Briggs SL, Clawson DK, Shang M, Smith GF, collagen. The reason for this can be that thrombin does not et al. The crystal structures of human alpha-thrombin complexed with play a role in the binding of platelets to collagen. active site-directed diamino benzo[b]thiophene derivatives: a binding TIP has the potential to be developed as an antithrom- mode for a structurally novel class of inhibitors. Protein Sci 2000, Jan;9(1):29–36. botic compound. It inhibits thrombin by binding to the [15] Dixon M, Webb EC. Enzymes. Longman Group: Academic Press; active site. It also inhibits platelet activation and the release 1979. p. 334 reaction when platelets were stimulated with a thrombin- [16] Fabioloa GF, Bobde V, Damodharon L, Pattabhi V, Durani S. Con- mimetic peptide, Gly-Pro-Arg-Pro, and it inhibits platelet formational preferences of heterochiral peptides. Crystal structures adhesion onto a human microvascular endothelial matrix in of heterochiral peptides Boc-(D)-Val-(D)-Ala-Leu-Ala-Ome and Boc- Val-Ala-Leu-(D)-Ala-Ome enhanced stability of beta sheet through C– the parallel plate flow chamber under both arterial and H–O hydrogen bonds. J Biomol Struct Dyn 2001, Feb;18(4):579–94. venous shear conditions. The stability and specificity of [17] Das B, Merovitch H. Optimization of solvation models for predict- this peptide in comparison with other serine proteases, such ing the structure of surface loops in proteins. Proteins 2001, May 15; as factor Xa and factor VIIa, will be determined and in vivo 43(3):303–14. studies with this peptide are necessary to compare its effect [18] Pearce J. Going round in circles to avoid proteolysis. Trends Biochem Sci 2001, May;26(5):282. with that of the other well-known thrombin inhibitors. [19] Priestle JP, Rahuel J, Rink H, Tones M, Grutter MG. Changes in interactions in complexes of hirudin derivatives and human alpha- thrombin due to different crystal forms. Protein Sci 1993, Oct;2(10): References 1630–42. [20] Vagner J, Barany G, Lam KS, Krchnák V, Sepetov NF, Ostrem JA, et al. Enzyme-mediated spatial segregation on individual polymeric [1] Doorbar J, Winter G. Isolation of a peptide antagonist to the thrombin support beads: application to generation and screening of encoded receptor using phage display. J Mol Biol 1994;244:361–9. combinatorial libraries. Proc Natl Acad Sci U S A 1996, August; [2] Goodson RJ, Doyle MV, Kaufman SE, Rosenberg S. High-affinity 93:8194–9. urokinase receptor antagonists identified with bacteriophage peptide [21] Cappiello M, Vilardo PG, Lippi A, Criscuoli M, Del-Corso A, Mura display. Proc Natl Acad Sci U S A 1994;91:7129–33. U. Kinetics of human thrombin inhibition by two novel peptide in- [3] Dennis MS, Eigenbrot C, Skelton NJ, Ultsch MH, Santell L, Dwyer hibitors (hirunorm IV and hirunorm V). Biochem Pharmacol 1996, MA, et al. Peptide exosite inhibitors of factor VIIa as anticoagulants. Oct. 25;52(8):1141–6. Nature 2000;404(6777):465–70. [22] Nilsson T, Sjoling-Ericksson A, Deinum J. The mechanism of binding [4] Badimon L, Meyer BJ, Badiman JJ. Thrombin in arterial thrombosis. of low-molecular-weight active site inhibitors to human alpha-throm- Haemostasis 1994;24:69–80. bin. J Enzyme Inhib 1998, Feb;13(1):11–29. [5] Bloom AL. Physiology of blood coagulation. Haemostasis 1990; [23] Francischetti IM, Valenzuela JG, Ribeiro JM. Anophelin: kinetics 20(Suppl 1):14–29. and mechanism of thrombin inhibition. Biochemistry 1999, Dec. 14; [6] Davie EW, Fujikawa K, Kissel W. The coagulation cascade: initiation, 38(50):16678–85. maintenance and regulation. Biochemistry 1991;30:10363–70. 130 S-12 131 209 Funksionele beskrywing van ’n faktor VIIa inhiberende peptied, IP-7, geselekteer deur faagblootleggingstegnologie SM MEIRING*, CE ROETS** & PN BADENHORST Departement Hematologie en Selbiologie, Fakulteit Gesondheidswetenskappe, Universiteit van die Vrystaat, Posbus 339 (G2), Bloemfontein, 9300 * Persoon aan wie korrespondensie gerig moet word. ** Huidige adres: Farmovs: Parexel, Privaat sak X09, Brandhof, 9324 UITTREKSEL Die tegniek van faagblootlegging is gebruik om ’n sikliese heptapeptied te selekteer wat met weefselfaktor (WF) kompeteer vir binding aan stollingsfaktor VIIa. Die aminosuurvolgorde van die peptied is Cys-Ala- Trp-Pro-His-Thr-Pro-Asp-Cys (C-AWPHTPD-C) en dit verleng die protrombientyd (PT) op ’n konsentrasie-afhanklike wyse. Die peptied beperk plaatjieklewing aan beide menslike endoteelsel- en weefselfaktormatrikse in ’n vloeikamermodel onder arteriële vloeitoestande. Die peptied funksioneer as ’n volledig mededingende inhibeerder van faktor VIIa met ’n inhibisiekonstante (Ki) van 123,2 M. In sy huidige vorm is die peptied waarskynlik nie sterk genoeg om verder as antitrombotiese middel ontwikkel te word nie, maar verskillende strategieë kan gevolg word om die werking daarvan te versterk. ABSTRACT Functional characterisation of a factor VIIa inhibiting peptide, IP-7 selected by phage display technology By using the technique of phage display, we selected a cyclic heptapeptide sequence Cys-Ala-Trp-Pro-His- Thr-Pro-Asp-Cys (C-AWPHTPD-C) that competes with tissue factor for binding to coagulation factor VII. This peptide prolongs the prothrombin time (PT) in a concentration dependent way. It also reduces platelet adhesion to both human endothelial cell and tissue factor matrixes in a flow chamber under arterial flow conditions. Furthermore, it acts as a full competitive inhibitor of factor VIIa with an inhibition constant (Ki) of 123,2 M. In its current form the peptide is probably not sufficiently potent for development as an antithrombotic agent, but different strategies could be followed to reinforce its performance. INLEIDING Bloedstolling begin met die beskadiging van ’n bloedvat en die blootstelling van sirkulerende bloed aan weefselfaktor (WF) in die subendoteel. Stollingsfaktor VII (FVII) wat in die bloed sirkuleer, bind aan WF om ’n FVII/WF-kompleks te vorm. Hierdie kompleks aktiveer faktor X wat op sy beurt protrombien aktiveer om trombien te vorm. Trombien skakel fibrinogeen na fibrien om en aktiveer bloedplaatjies om ’n trombus of stolsel te vorm.1 Ten spyte van jare se navorsing is daar slegs twee antistolmiddels wat ruimskoots deur dokters gebruik word, naamlik kumarien en heparien. Kumarien versteur die funksionering van vitamien K-afhanklike proteïene wat beide pro-stollingsfaktore (trombien en faktore X, IX en VII) en anti-stollingsfaktore (proteïene C en S) insluit, terwyl heparien die inhibisie van trombien en faktor Xa deur antitrombien III bevorder. Die nieselektiewe inhibisie van stolling deur beide hierdie antistolmiddels is waarskynlik verantwoordelik vir die onvermoë daarvan om ’n terapeutiese balans tussen trombose en hemostase te handhaaf.2,3 Tans word baie aandag bestee aan die ontwikkeling van middels wat faktor X en trombien direk inhibeer. Die feit dat faktor VII hoër op in die stollingsbaan funksioneer, Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 131 132 210 beteken dat daar teoreties minder inhibeerder nodig is om stolling op hierdie vlak te stuit. Die vorming van die FVIIa/WF-kompleks lei die stollingsproses in en bied ’n goeie teiken vir die ontwikkeling van antistolmiddels. Weefselfaktorbaaninhibeerder (WFBI) is ’n natuurlike inhibeerder van FVIIa/WF en sy werking is van faktor Xa afhanklik.4 Alhoewel daar nie FVIIa- inhibeerders kommersieel beskikbaar is nie, is dit al voorheen verken.5,6 Verbindings wat die binding van WF aan FVIIa blokkeer, sal die aktivering van faktor X verhoed en aldus die vorming van ’n stolsel of trombus bekamp. Ons het ’n studie ontwerp om peptiedinhibeerders van FVIIa te ontwikkel deur van faagblootleggingstegnologie gebruik te maak. Die tegniek van faagblootlegging maak die sifting van groot getalle faagklone moontlik. Meer as ’n miljard verskillende peptiedvolgordes kan op die manier gesif word en dit gee aan faagblootlegging ’n voordeel bo ander metodes.7 Ons het besluit om ’n kort oligonukleotiede te selekteer omdat klein peptiede minder immunogenies is.8 Verskillende FVIIa-bindende faagkolonies is getoets vir hulle vermoë om FVIIa te inhibeer. Die DNS-volgorde van dié wat die sterkste inhibisie getoon het, is bepaal en ooreenstemmende peptiede gesintetiseer. Op die manier is ’n sikliese heptapeptied met die volgorde C-AWPHTPD-C gesintetiseer en deur middel van ’n protrombientyd (PT) tipeer. Ons het ook die peptied se kinetiese eienskappe ten opsigte van FVIIa-inhibisie bepaal en die effek daarvan op plaatjieklewing aan endoteelselle en WF gemeet. METODES 1. Seleksie van faktor VIIa-bindende fage Die faagblootleggingstegniek wat gebruik is, is reeds volledig beskryf.9 Twee peptiedblootleggingsbiblioteke, ’n sikliese heptameer- en ’n lineêre dodekameerbiblioteek (New England Biolabs, Beverly, MA, VSA) is gebruik. Twintig g FVIIa (ADI, Greenwich, CT, VSA) opgelos in 1 ml fosfaatgebufferde fisiologiese soutoplossing (PBS) is oornag by 4°C aan ’n immuunbuis (Nunc, IL, VSA) gebind. Fage (2 x 1011) van elke biblioteek in PBS is tot die buisies toegevoeg en daarna vir 1½ uur by kamertemperatuur geïnkubeer om sodoende aan die faktor VIIa te bind. Die ongebonde fage is afgewas en die FVIIa-bindende fage niespesifiek met ’n sterk suuroplossing (0,2 M glisien by pH 2) geëlueer. Drie seleksierondtes is uitgevoer. Tydens die tweede en derde rondte is die VIIa-bindende fage spesifiek met 10 g van ’n muis- anti-mens-FVIIa-monoklonale antiliggaam (ADI, Greenwich, CT, VSA) geëlueer. Die FVIIa- bindende fage is na elke rondte vermeerder deur log-fase Escherichia coli (E.coli) selle oornag te infekteer en te suiwer deur twee poliëtileenglikol-presipitasies. Om enkel VIIa-bindende kolonies voor te berei, is seriële tienvoudige verdunnings van ’n fraksie van die geëlueerde fage gemaak en uitgeplaat. Eenhonderd vier-en-veertig (3 X 48) enkel kolonies van elke biblioteek is na die derde seleksierondte uitgekies en oornag by 37 C in E coli-selle gekweek. Die supernatante, wat die geamplifiseerde fage bevat, is met behulp van ’n bindings-ELISA vir binding aan faktor VIIa getoets. a. Bindings-ELISA ’n 96-putjie ELISA-plaat is oornag met 20 g/ml FVIIa by 4 C bevestig. Na blokkering met 4% afgeroomde melk (DIFCO, Detroit. MI, VSA) in PBS, is die supernatante (100 µl) van die gekweekte kolonies by die putjies toegevoeg en vir 2 uur by kamertemperatuur geïnkubeer. ’n Peperwortelperoksidase-gekonjugeerde anti-faagantiliggaam (Amersham Pharmacia Biotech, NJ,VSA), is bygevoeg om die FVIIa-bindende fage op te spoor. Ses kolonies van elke Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 132 133 211 biblioteek het sterk binding getoon en is verder gekweek. Verdunningsreekse van bogenoemde 12 kolonies (ses van elke biblioteek) is vir konsentrasie-afhanklike binding aan FVIIa in ’n verdunnings-ELISA getoets. b. Inhibisie-ELISA ’n Inhibisie-ELISA is gedoen om te bepaal of WF in staat is om te verhoed dat die twaalf kolonies aan FVIIa bind. Verskillende konsentrasies WF (ADI, Greenwich, CT, VSA) is by die FVIIa-bedekte putjies gevoeg voordat die fage bygevoeg is. Die res van die ELISA is uitgevoer soos bo beskryf. c. Protrombientyd (PT) Die PT is gedoen deur 100 µl normale mensplasma by 50 µl van verskillende konsentrasies van die twaalf sterkste FVIIa-bindende fage te voeg en vir 10 min te inkubeer. Die faagkonsentrasies het van 1 x 109 tot 2,2 x 1012 fage gewissel. Daarna is 200 µl weefselfaktor (Innovin van Dade Behring, Marburg, Duitsland) bygevoeg en die stoltyd met ’n STart 4 stollingsmeter (Diagnostica Stago, Asnieres, Frankryk) gemeet. 2. Nukleotiedvolgordebepaling DNS is van beide die kolonies wat die PT verleng het, voorberei en die DNS-volgorde is met behulp van die DYEnamic ET Terminator Cycle Sequencing Premix Kitsstel (Amersham Pharmacia Biotec Inc., NJ, VSA) bepaal. ’n Sikliese heptapeptied met dieselfde volgorde as die blootgelegde peptied op die sikliese kolonie is deur ADI, Greenwich, GT, VSA, gesintetiseer. Ons het die peptied IP-7 genoem. 3. Peptiedanalise a. Protrombientyd (PT) Die effek wat drie verskillende peptiedkonsentrasies (1,17mM; 0,58 mM; en 0,295 mM) op die PT het, is bepaal. PBS en ’n peptied wat nie aan FVII bind nie is as negatiewe kontrole gebruik. Die PT is gedoen soos bo beskryf. b. Perfusiestudies met endoteelselle Die antitrombotiese effek van die peptied IP-7 is met ’n vloeikamer, soos deur Sakaríassen beskryf, getoets.10 Thermanox plastiese dekglasies (Nunc, IL, VSA) is met menslike endoteelselle bedek wat goedgunstiglik geskenk is deur E.W. Ades en T.J. Lawley van die Centres for Disease Control and Prevention and Emory University School of Medicine, Atlanta, GA, VSA. Drie verskillende peptiedkonsentrasies (2,34 M; 4,7 M; en 9,35 M) is met 10 ml bloed vir 10 minute by 37 C geïnkubeer. Konstante bloedvloei is met ’n peristaltiese pomp gehandhaaf en die bloed is vir 5 minute oor die dekglasies hersirkuleer. Verskillende skuiftempo’s van 1000 s-1 en 200 s-1 is gebruik. Die dekglasies is daarna verwyder, met HEPES buffer gespoel, in metanol fikseer en volgens die May-Grünwald-Giemsa-metode gekleur. Dertig velde per dekglasie is ontleed en die persentasie plaatjiebedekking is met behulp van ’n Zeiss ligmikroskoop gemeet.11,12 Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 133 134 212 c. Perfusiestudies met weefselfaktor Die dekglasies is met 100 l weefselfaktor (Innovin van Dade Behring, Marburg, Duitsland) bedek. Die perfusiestudies is uitgevoer soos hierbo beskryf, behalwe dat die skuiftempo’s in hierdie geval 200 s-1, 650 s-1 en 1300 s-1 was. Finale peptiedkonsentrasies van 29,2 M; 58,5 M en 117 M is gebruik. d. In vitro-ensieminhibisie Die ensiemkinetika van die inhibisie van Faktor VIIa met peptied IP-7 is met behulp van chromogeniese substraathidrolise bepaal. Verskeie kinetiese reaksies is uitgevoer met peptiedkonsentrasies wat van 0 tot 0.456 mM gewissel het en waarby verskillende konsentrasies van die chromogeniese substraat van faktor VIIa (Spectrosyme FVIIa van American Diagnostica, VSA) gevoeg is. Die substraatkonsentrasies het gewissel van 0.56 tot 0 mM. ’n Optimale FVIIa konsentrasie van 400 nM is gebruik en Innovin (Dade Behring, Marburg, Duitsland) is as bron van WF gebruik. Die totale reaksievolume was 200 l en al die reagense is in die reaksiebuffer (0.05 M Tris, 0.1 M NaCl pH 8.4) verdun. Chromogeniese substraathidrolise is vir 40 minute by ’n golflengte van 405 nm gevolg en deur middel van ’n EL312e mikroplaat biokinetiese leser (Bio-tek instruments, Vermont, VSA) gelees. Die kinetiese parameters Km en Vm by elke peptiedkonsentrasie is bereken deur gebruik te maak van enkelplekbinding nielineêre regressie. Om die tipe inhibisie van die peptied te bepaal, is ’n Lineweaver-Burk grafiek (1/Vo vs 1/substraat) gestip (figuur 6). Die inhibisiekonstante (Ki -waarde) is bepaal deur die oënskynlike Km-waardes teenoor die peptiedkonsentrasies te stip (kyk figuur 7). Die Ki–waarde is as die x-as afsnit beskou. RESULTATE 1. Faagseleksie en DNS volgordebepaling Die bindings-ELISA het twaalf kolonies geïdentifiseer wat binding aan FVIIa getoon het. Ses van hierdie kolonies was van die dodekapeptiedbiblioteek en ses van die sikliese pentapeptiedbiblioteek afkomstig. ’n Verdunnings-ELISA is uitgevoer op die twaalf kolonies wat FVIIa-bindingsaffiniteit getoon het. ’n Irrelevante faagkolonie is ook getoets vir binding aan FVIIa en is as ’n negatiewe kontrole gebruik. Om die grafiek te vereenvoudig is die bindingsaffiniteit van een kolonie uit elke biblioteek in figuur 1 gestip. Die optiese digtheid het met stygende faagkonsentrasies verhoog, wat daarop dui dat die fage spesifiek aan FVIIa bind. Die faagkolonies van die sikliese heptapeptiedbiblioteek het groter bindingsaffiniteit getoon as dié van die dodekapeptiedbiblioteek (n=3). Die faagkolonie wat as ’n negatiewe kontrole gebruik is, het geen binding aan FVIIa getoon nie. ’n Inhibisie-ELISA is gedoen om te bepaal of WF daartoe in staat is om te voorkom dat die kolonies aan FVIIa bind. Die lineêre kolonies het geen inhibisie getoon nie en slegs een vanaf die sikliese biblioteek het matige inhibisie getoon (figuur 2). Laasgenoemde kolonie sowel as een kolonie vanaf die lineêre biblioteek het die PT verleng (figure 3 en 4). Die negatiewe kontrole het geen effek op die PT gehad nie. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 134 135 213 0.4 n=3 0.3 Sikliese heptapeptied OD (490nm) 0.2 Lineêre dodekapeptied 0.1 0.0 0.0 1.0×1010 2.0×1010 3.0×1010 fage/ml Figuur 1: Verdunnings-ELISA van beide sikliese heptapeptied- en lineêre dodekapeptied- kolonies aan FVIIa. Die OD490 vermeerder met stygende faagkonsentrasies. Die sikliese heptapeptied-kolonie bind sterker as die dodekapeptied-kolonie (n=3). 0.9 0.8 OD 0.7 (490nm) 0.6 0.5 Sikliese heptapeptied 0.4 0.3 0.2 0.1 0.0 0.00 0.01 0.02 0.03 0.04 0.05 WF (µM) Figuur 2: Inhibisie-ELISA van die sikliese heptapeptied-kolonie by verskillende WF- konsentrasies. WF-konsentrasies van 0 tot 0.05 M is by FVIIa-bedekte putjies gevoeg en vir 15 min geïnkubeer. Daar is 5.1010 fage van die sikliese heptapeptied-kolonie by die putjies gevoeg en vir 2 uur geïnkubeer waarna die FVIIa-gebonde fage met ’n anti-faagantiliggaam aangetoon is (n=3). Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 135 136 214 Beide kolonies wat die PT verleng het se DNS-volgorde is bepaal. Dit is interessant om daarop te let dat die volgorde van die kolonie vanuit die sikliese heptameerbiblioteek presies in die middelste deel van die volgorde van die kolonie uit die dodekapeptiedbiblioteek gepas het. Omdat die sikliese kolonie sterker aan FVIIa gebind het, het ons besluit om ’n sikliese peptied met dieselfde volgorde as dié wat op die sikliese kolonie blootgelê is, te sintetiseer. Die volgorde van hierdie peptied is C-AWPHTPD-C. Dit vergelyk nie met die volgorde van ’n reeds gepubliseerde faktor VII inhibeerder PN7051 (C-QY-C) nie. 25 n=3 PT (s) Sikliese heptapeptied20 15 10 5 0 0.0 2.0×10 11 4.0×10 11 6.0×10 11 8.0×10 11 fage/ml Figuur 3: PT’s met stygende konsentrasies van die sikliese heptapeptied-kolonie. Stygende konsentrasies van hierdie faagkolonie is vir 10 minute met mensplasma geïnkubeer voor toevoeging van die PT-reagens. Die PT is met 8.3 sekondes by die hoogste faagkoloniekonsentrasie (5.8.1011 fage) verleng 22.5 n=3 20.0 Lineêre dodekapeptied PT 17.5 (s) 15.0 12.5 10.0 7.5 5.0 2.5 0.0 0.0 1.0×10 11 2.0×10 11 3.0×10 11 fage/ml Figuur 4: PT’s met stygende konsentrasies van die dodekapeptied-kolonie. Stygende konsen- trasies van hierdie faagkolonie is vir 10 minute met menslike plasma geïnkubeer voordat die PT- reagens toegevoeg is. Hierdie faagkolonie verleng die PT op ’n konsentrasie-afhanklike manier. By die hoogste faagkonsentrasie (2.3.1011 fage) word die PT met 5,4 sekondes verleng (n=3). Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 136 137 215 50 n=3 PT (s) 40 30 20 10 0 0.00 0.25 0.50 0.75 1.00 1.25 IP-7 (finale konsentrasie, mM) Figuur 5: Verlenging van die PT in mensplasma. Verskillende peptiedkonsentrasies is by normale gepoelde mensplasma gevoeg en vir 10 minute geïnkubeer voor toevoeging van die PT- reagens. By die hoogste peptiedkonsentrasie, 1.17 mM, is die PT met 30 sekondes verleng. (n=3). 2. Toetse uitgevoer op die peptied, IP-7. IP-7 het die PT op ’n dosisafhanklike manier verleng, en peptiedkonsentrasies van hoër as 0,58 mM was hiervoor nodig (figuur 5). ’n Irrelevante peptied is as negatiewe kontrole gebruik en het geen effek op die PT gehad nie. Die persentasie inhibisie van plaatjieklewing aan endoteelselbedekte dekglasies word in tabel 1 opgesom. Die peptied het plaatjieklewing op ’n dosisafhanklike manier by beide skuiftempo’s geïnhibeer. Die inhiberende effek was meer uitgesproke by die arteriële vloeitoestande (1000 s-1) omdat die hoogste peptiedkonsentrasie (9.35 M) plaatjieklewing aan die dekglasies met 74% onderdruk het. By die lae skuiftempo (200 s-1) was die onderdrukking 54% vir dieselfde peptiedkonsentrasie. TABEL 1: Persentasie inhibisie van plaatjieklewing aan endoteelselbedekte dekglasies by skuiftempo’s van 200 s-1 en 1000 s-1 (n=3) Persentasie inhibisie by ’n Persentasie inhibisie by ’n skuiftempo van 200 s-1 skuiftempo van 1000 s-1 Kontrole (0 M IP-7) 0% 0% 2.34 M IP-7 19% 22% 4.70 M IP-7 55% 47% 9.35 M IP-7 54% 74% Die peptied het ook die klewing van plaatjies aan WF geïnhibeer. Die persentasie inhibisie van plaatjieklewing aan die dekglasies word in tabel 2 opgesom. Die peptied het ’n meer uitgesproke inhibisie van plaatjieklewing by die hoër skuiftempo veroorsaak. By ’n skuiftempo van 1300 s-1 is plaatjieklewing aan die WF-bedekte dekglasies op ’n dosisafhanklike manier onderdruk. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 137 138 216 TABEL 2: Persentasie inhibisie van plaatjieklewing aan weefselfaktorbedekte dekglasies by skuiftempo’s van 200 s-1, 650 s-1 and 1300 s-1 (n=2) Persentasie inhibisie Persentasie inhibisie Persentasie inhibisie by by ’n skuiftempo van by ’n skuiftempo van ’n skuiftempo van 1300 200 s- 650 s-1 s-1 Kontrole (0 µm IP-7) 0% 0% 0% 29.2 M IP-7 0% 0% 53% 58.5 M IP-7 33% 59% 87% 117 M IP-7 Nie bepaal nie Nie bepaal nie 85% Met die kinetiese studies het die maksimum omloopsnelheid (Vm-waardes) by die verskillende peptiedkonsentrasies onveranderd gebly terwyl die Michaelis-Menten konstante (Km-waarde) verhoog het met stygende konsentrasies. Dit kon op mededingende inhibisie dui en om dit te bevestig het ons ‘n Lineweaver-Burk grafiek gestip (figuur 6). Hiermee is mededingende inhibisie bevestig omdat die helling van die grafiek toegeneem het met stygende peptiedkonsentrasies. Die inhibisiekonstante (Ki-waarde) is bepaal as die x-as afsnit by 123, 2 µM soos aangetoon in figuur 7. 3000 IP-7-konsentrasies 0 mM 2500 0.057 mM 1/V0 2000 0.114 mM 0.171 mM 1500 0.228 mM 1000 500 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 1/substraat (mM) Figuur 6: Lineweaver-Burk-stip wat die effek van mededingende inhibisie aantoon. Die helling van die grafiek neem toe met stygende peptiedkonsentrasies, wat dui op mededingende inhibisie van FVIIa. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 138 139 217 20 18 Km' 16 14 12 10 8 6 4 2 -0.3 -0.2 -0.1 -0.0 0.1 0.2 0.3 0.4 0.5 -Ki = -123.2 M IP-7 (mM) Figuur 7: Die Km-waarde van die reaksies met verskeie IP-7 konsentrasies is gestip teenoor die betrokke peptiedkonsentrasies. Met behulp van hierdie grafiek kon die Ki bepaal word deur die X-afsnit van ’n lineêre regressie te meet. (n=4) BESPREKING Met hierdie ondersoek het ons die tegniek van faagblootlegging gebruik om peptiede wat FVIIa inhibeer, te selekteer. Daar is op peptiede besluit omdat kleiner molekules minder immunogenies as groter molekules is.8 Ons het gepoog om peptiede wat aan die katalitiese domein van FVIIa bind, te selekteer deur die FVIIa-bindende fage in die laaste 2 seleksierondtes met ’n muis monoklonale antiliggaam wat aan die aktiewe setel van FVIIa bind, te elueer.13 Die feit dat die faagkolonies van die sikliese heptapeptiedbiblioteek sterker aan FVIIa as dié van die lineêre een bind, is nie onverwags nie omdat dit bekend is dat sikliese peptiede hegter as lineêre peptiede aan hulle onderskeie ligande bind vanweë verbeterde bindingsentropie.14 Verder het WF die binding van een van die kolonies van die sikliese peptiedbiblioteek verhoed, wat daarop dui dat hierdie spesifieke sikliese peptied aan dieselfde bindingsetel as WF bind. Die effek is egter beperk omdat WF nie die binding van die fage volledig voorkom het nie (kyk figuur 2). ’n Sikliese peptied met dieselfde aminosuurvolgorde as dié van die sikliese heptapeptiedkolonie, is gesintetiseer. Geen ooreenstemming tussen die volgorde van die gesintetiseerde peptied, IP-7, en dié van WF, FX of WFBI is gevind nie. Daar was twee redes waarom ons besluit het om die sikliese peptied te sintetiseer, naamlik omdat dit sterker aan FVIIa gebind het, en dit bekend is dat sikliese peptiede meer stabiel is as dié met ’n lineêre volgorde en meer bestand is teen proteolise.15,16 IP-7 het plaatjieklewing op ’n dosisafhanklike manier by beide skuiftempo’s inhibeer, maar die inhibisie was meer uitgesproke by arteriële vloeitoestande (skuiftempo van 1000 s-1). Die rede hiervoor is dat plaatjies meer by arteriële trombose betrokke is en fibrien by veneuse trombose.17 Dieselfde resultate is ook tydens die perfusiestudies met WF-bedekte dekglasies gevind. Uit die literatuur blyk dit dat soortgelyke resultate verkry is met ’n ander TF/FVIIa- inhibeerder, PN7051.18 Lg. peptied het 50% inhibisie met ’n konsentrasie van 500 M met die protrombien tyd getoon in vergelyking met IP-7 wat 58% inhibisie met 58,5 M met die Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 139 140 218 protrombientyd getoon het. IP-7 blyk dus ’n sterker onderdrukkende effek te hê op trombusvorming as PN7051.18 Dit is belangrik om daarop te let dat die perfusiestudies met gehepariniseerde bloed uitgevoer is. Dit is bekend dat lae molekulêre gewig heparien (LMWH) nie alle stollingsaktiwiteit blokkeer nie en ’n mate van trombienvorming toelaat sodat fibrienneerlegging op die oppervlakte plaasvind.17 Fibriennetwerke kan duidelik waargeneem word op die WF- matriks by lae skuiftempo’s. Die inhibisiekonstante (Ki-waarde) is as 123,2 µM bepaal, wat beteken dat IP-7 nie ’n sterk FVIIa-inhibeerder is in vergelyking met van die ander FVIIa-inhibeerders nie. Byvoorbeeld, WFBI inhibeer die aktivering van FXa met ’n Ki van 30 nM, en Kunitz domeinvariante van WFBI inhibeer die FVIIa/WF-kompleks met ’n K van 1,9 nM.7i ’n Ander peptiedinhibeerder, A-183, wat aan die eksosetel van die proteasedomein van FVIIa bind, inhibeer FX aktivering met ’n Ki van 200 pM.6,19 IP-7 verleng die PT op ’n dosisafhanklike manier. In vergelyking met ander inhibeerders van WF-afhanklike stolling, het ’n peptiedkonsentrasie van 1,17nM die PT drievoudig verleng. PN7051 verleng ook die stoltyd op ’n dosisafhanklike manier met ’n IC50 van 1.3 mM.18 Die rede vir die verlenging van die PT kan toegeskryf word aan die binding van die peptied aan FVIIa om die vorming van die FVIIa/WF-kompleks te verhinder en sodoende die aktivering van FX te verhoed. In ’n studie waar die effek van twee verskillende peptiedinhibeerders van die eksosetel van FVIIa met PT gemeet is, het die verlenging van die PT by konsentrasies van groter as 100 nM plaasgevind. IP-7 is dus nie so potent as eksosetel inhibeerders van FVIIa nie6, wat waarskynlik aan die kort lengte van die peptied toegeskryf kan word. Dit is in die literatuur aangetoon dat langer peptiede meer aktief is as kort peptiede. ’n 27- aminosuur peptied het byvoorbeeld ’n sewevoudige hoër aktiwiteit as ’n 15-aminosuur peptied getoon.6 Die kort sikliese peptied PN7051 (Cys-Gln-Tyr-Cys) het ook ’n baie laer aktiwiteit as langer peptiede getoon.6,18 Verdere studies is nodig om die aktiwiteit van IP-7 te verhoog en sluit onder meer in die koppeling van twee of selfs drie van die peptiede, of die kombinering van die peptied met ander faktor VIIa of WF inhiberende peptiede.6 Aangesien die faagblootleggingstegnologie so ’n kragtige instrument in die ontwikkeling van geneesmiddels is, kan dit ook aangewend word in die ontwikkeling van antistolmiddels wat die FVIIa/WF-kompleks inhibeer. 7 BIBLIOGRAFIE 1. Hoffman, M. (2003). Remodelling the blood coagulation cascade. Journal of Thrombosis and Thrombolysis, 16, 17-20. 2. Hirsh, J., Dalen, J.E., Deykin, D., Poller, L. (1992). Heparin: mechanism of action, pharmacokinetics, dosing considerations, monitoring, efficacy, and safety, Chest, 102, 337S-351S. 3. Philips, D.R., Conley, P.B., Sinha, U., Andre, P. (2005). Terapeutic approaches in aterial thrombosis, Journal of Thrombosis and Haemostasis, 3, 1577-1589. 4. Broze, G.J., Warren, L.A., Novotny, W.F., Higuchi, D., Girard, J.J., Miletich, J.P. (1988). The lipoprotein-associated coagulation inhibitor that inhibits the factor VII-tissue factor complex also inhibits factor Xa: insight into its possible mechanism of action, Blood, 71, 335-343. 5. Eigenbrot,C. (2002). Structure, function and activation of coagulation factor VII, Current Protein and Peptide Sciences, 3, 287-299. 6. Lazarus, R.A., Olivero, A.G., Eigenbrot, C., Kirchofer, D. (2004). Inhibitors of tissue factor VIIa for anticoagulant therapy, Current Medicinal Chemistry, 11, 2275-2290. 7. Meiring, S.M., Kotzè, H.F., Pretorius, G.H.J., Badenhorst, P.N. (1999). Die toepassing van peptiedblootlegging op fage in trombose en hemostase, SA Tydskrif vir Wetenskap en Tegnologie, 18(3), 76-81. 8. Markwardt, F. (1990). Hirudin and derivates as anticoagulant agents, Thrombosis and Haemostasis, 66,141-152. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 140 141 219 9. Meiring, S.M., Littauer, D., Hàrsfalvi, J., van Wyk, V., Badenhorst, P.N., Kotze, H.F. (2002). In vitro effect of a thrombin inhibition peptide selected by phage display technology, Thrombosis Research, 107, 365-371. 10. Sakaríassen, K.S., Aarts, P.A.M.M., DeGroot, P., Houdijk, W.P.M., Sixma, J.J. (1983). A perfusion chamber developed to investigate platelet interaction in flowing blood with human vessel wall cells, their extracellular matrix, and purified components, Journal of Laboratory and Clinical Medicine, 102, 522-535. 11. Bonnefoy, A., Hàrsfalvi, J., Pflieger, G., Fauvel-Lafève, F., Legrand, C. (2001). The subendothelium of the HMEC-1 cell line supports thrombus formation in the absence of Von Willebrand factor and collagen types I, III and VI, Thrombosis and Haemostasis, 85, 552-559. 12. Hàrsfalvi, J., Stassen, J.M., Hoylaerts, M.F., Van Houtte, E., Sawyer, R.T., Vermylen, J., Deckmyn, H. (1995). Calin from Hirudo medicinalis, an inhibitor of Von Willebrand factor binding to collagen under static and flow conditions, Blood, 85, 705-711. 13. Clarke, B.J., Ofosu, F.A., Sridhara, S., Bona, R.D., Rickles, .FR., Blajchman, M.A. (1992). The first epidermal growth factor domain of human coagulation factor VII is essential for binding with tissue factor, FEBS letters, 298, 206-210. 14. Das, B., Meirovitch, H. (2001). Optimisation of solvation models for predicting the structure of surface loops in proteins, Proteins, 43, 303-314. 15. Fabiola, G.F., Bobde, V., Damodharan, L., Pattabhi, V., Durani, S. (2001). Conformational preferences of heterochiral peptides. Crystal structures of heterochiral peptides Boc-(D)Val-(D)Ala- Leu-Ala-Ome and Boc-Val-Ala-Leu-(D)Ala-Ome-enhanced stability of beta-sheet through C-H-O hydrogen bonds, Journal of Bio molecular Structure Dynamics, 18, 579-594. 16. Pearce, J. (2001). Going round in circles to avoid proteolysis, Trends in Biochemical Sciences, 26, 282. 17. Sakaríassen, K.S., Hanson, S.R., Cadroy, Y. (2001). Methods and models to evaluate shear-dependent and surface reactivity-dependent antithrombotic efficacy, Thrombosis Research, 104, 149-174. 18. Örning, L., Fischer, P.M., Hu, C., Agner, E., Engebretsen, M., Husbyn, M., Petersen, L.B., Ørvim, U., Llinas, M., Sakaríassen, K.S. (2002). A cyclic pentapeptide derived from the second EGF-like domain of factor VII is an inhibitor of tissue factor dependent coagulation and thrombus formation, Thrombosis and Haemostasis, 87, 13-21. 19. Dennis, M.S., Roberge, M., Quan, C., Lazarus, R.A. (2001). Selection and characterization of a new class of peptide exosite inhibitors of coagulation factor VIIa, Biochemistry, 40, 9513-9521. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 141 142 220 VERKORTE CURRICULA VITAE Muriel Meiring is medeprofessor in die Departement Hematologie aan die Universiteit van die Vrystaat. Sy behaal haar PhD graad in Hematologie in 1996 waarna sy ’n beurs van die Vlaamse regering ontvang het om haar post-doktorale studie in 1997 in België te doen. In 1998 sluit sy samewerkingsooreenkomste met navorsingsgroepe in België en Hongarye met wie sy nog steeds saamwerk. Sy is outeur van verskeie navorsingsartikels en publiseer in Afrikaans en Engels. Sy het die THUTHUKA-toekenning vir Vroue in Navorsing ontvang en doen ook kontraknavorsing vir verskeie farmaseutiese maatskappye in die ontwikkeling van nuwe antistolmiddels. Lizel Roets is tans ’n kliniese ouditeur by Farmovs-Parexel. Sy het haar PhD in Hematologie in 2002 behaal vir navorsing in samewerking met American Diagnostica Inc. Met die navorsing het sy gepoog om ’n antistolmiddel te ontwikkel. Huidig voer sy verskeie interne en eksterne ouditte uit om te verseker dat daar aan alle regulasies en riglyne voldoen word. Sy is ook betrokke by sisteemouditte in die verskillende afdelings en help met die aanbieding van interne en eksterne kursusse en interne opleidingsessies. Philip Badenhorst is hoof van die Departement Hematologie en Selbiologie aan die Universiteit van die Vrystaat en hoofspesialis by die NHLS (National Health Laboratory Service). Hy behaal die MB, ChB grade aan die Universiteit van Stellenbosch en MMed en MD aan die Universiteit van die Vrystaat. Hy is ook genoot van die Suid-Afrikaanse Kollege vir Geneeskunde, lid van verskeie nasionale en internasionale professionele en wetenskaplike verenigings, sommige waarvan hy op die bestuur dien. Hy het reeds verskeie eerbewyse onvang, waaronder die Hendrik Verwoerd Navorsingstoekenning en die Universiteit van die Vrystaat Eeufeesmedalje. Hy is outeur of mede-outeur van vier boeke, 20 hoofstukke in boeke en 90 navorsingsartikels. Suid-Afrikaanse Tydskrif vir Natuurwetenskap en Tegnologie, Jaargang 25 No 4: Desember 2006 142 S-13 143 DRUG DEVELOPMENT RESEARCH 70 : 199–205 (2009) Research Article Development of an Inhibitory Antibody Fragment to Human Tissue Factor Using Phage Display Technology S.M. Meiring, J. Vermeulen, and P.N. Badenhorst Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, South Africa Strategy, Management and Health Policy Enabling Preclinical Development Clinical Development Technology, Preclinical Toxicology, Formulation Phases I-III Postmarketing Genomics, Research Drug Delivery, Regulatory, Quality, Phase IV Proteomics Pharmacokinetics Manufacturing ABSTRACT Tissue factor is involved in the etiology of thrombotic diseases initiating the thrombosis associated with the inflammation that occurs during infection. The prevention of blood coagulation and inflammation is of primary importance in a number of pathological situations. A single-chain variable antibody fragment of molecular weight of 26 kD that inhibits the action of human tissue factor was selected by phage display technology, purified and tested for its tissue factor inhibitory effect, purified on a protein A column, and its purity evaluated on SDS-PAGE. The effects of the antibody fragment on prothrombin times, Factor Xa production, and thrombin generation were assessed with increasing fragment concentrations, using chromogenic and fluorometric substrates. The antibody fragment dose-dependently prolonged the prothrombin time (IC505 0.5 mM) and delayed the lag phase before the thrombin generation burst and the peak thrombin concentration in the thrombin generation assay. The effect on thrombin generation was more pronounced in thrombophilic plasma than in normal plasma. Antibody-based tissue factor inhibitors therefore may provide an effective treatment for thrombotic disease without serious bleeding complications. Drug Dev Res 70 : 199–205, 2009. r 2009 Wiley-Liss, Inc. Key words: phage display; tissue factor; anticoagulants INTRODUCTION to fibrin, ultimately resulting in the formation of a fibrin Recent developments in the field of thrombosis clot. Thrombin further amplifies the coagulation process and hemostasis have highlighted the crucial role of the by activating cofactors, e.g., FV and FVIII and zymogens tissue factor/Factor VIIa complex in the initiation of such as FXI. Moreover, thrombin activates platelets coagulation processes. Coagulation is triggered by leading to platelet aggregation, which is necessary for exposure of zymogen Factor VII (FVII) to its mem- the formation of a hemostatic plug [Mann et al., 2003]. brane-bound cofactor, tissue factor (TF), to form the A major challenge in the field of cardiovascular TF/FVII complex [Mann et al., 2003]. The TF/FVII therapeutics is to develop antithrombotic agents to complex is converted to the enzymatically active TF/ FVIIa complex by Factor Xa (FXa) or autocatalytically Grant sponsor: National Research Foundation of South by TF/FVIIa. The TF/FVIIa complex can then activate Africa. Factor IX (FIX) to form FIXa, FX to FXa, and FVII to Correspondence to: Muriel Meiring, Department of FVIIa. FX can also be activated to FXa by the Haematology and Cell Biology, University of the Free State, complexing of FIXa with its cofactor FVIIIa, and FIX P.O. Box 339 (G2), Mandela Dive, Bloemfontein, 9300, South can be activated to FIXa by FXIa. FXa in complex with Africa. E-mail: GNHMSMM.MD@mail.uovs.ac.za its cofactor Factor V (FV) activates the conversion of Published online in Wiley InterScience (www.interscience.wiley. prothrombin to thrombin. Thrombin cleaves fibrinogen com). DOI: 10.1002/ddr.20295 c 2009 Wiley-Liss, Inc. 143 DDR 144 200 MEIRING ET AL. prevent uncontrolled thrombosis, but with less bleeding scFv_tomlinsonIJ.jsp) were kindly provided by Professor complications than with the existing widespread clinically Greg Winter (MRC Centre for Protein Engineering, used drugs (coumarins and heparins). Heparins enhance Cambridge, UK). Four Immune tubes were coated antithromin III-mediated inhibition of thrombin and overnight with 2 ml of each of two TF preparations. Two factor Xa, whereas coumarins impair the function of the tubes were coated with the prothrombin time reagent vitamin K-dependent proteins including procoagulants Thromborel (Dade Behring, GmbH, Marburg Germany) (thrombin, FXa, FIXa, and FVIIa) and anticoagulants and two tubes with Innovin (Dade-Behring, South (activated protein C and protein S). Although unfractio- Africa), respectively. After 3 washes with phosphate- nated heparins and coumarins are of great clinical value, buffered saline (PBS), the immune tubes were blocked both require careful dosing and frequent monitoring. with 4 ml of PBS containing 2% skimmed milk powder Significant progress has been made by introducing low (SM) (Difco Laboratories, Becton Dickinson, MD) for molecular weight heparin and pentasaccharides, but 2 h at room temperature. The tubes were washed 3 times there is still a need for improved anticoagulants with a with PBS after the blocking solution was removed. Then broad therapeutic window [Hirsh and Weitz, 1999]. 1 ml of the primary I and J libraries (1 1012 phages) The TF/FVIIa complex is an interesting target in was added to each of the Thromborel and Innovin- thrombosis-related disease because TF/FVIIa-directed coated tubes together with 3 ml of 4% SM solution and inhibitors might achieve anticoagulant efficacy without rotated for 1 h at room temperature. After another 1-h significantly interfering with normal hemostasis [Frédér- incubation at room temperature, the unspecific binding ick et al., 2005]. Intravascular ‘‘blood-borne’’ TF exhibits phages were washed away with 0.1% Tween PBS 10 procoagulant activity and can be incorporated into times (20 times for the second and third round). Bound platelet thrombi and contribute to thrombosis. There- phages were eluted by 500ml trypsin-PBS (50ml of fore, specific anti-TF antibodies may cause less bleeding 10 mg/ml trypsin stock solution added to 450ml PBS) and as they inhibit intravascular TF at concentrations that rotated for 10 min at room temperature. are far below those necessary to block the high amounts Escherichia coli TG1 cells (optical density [OD] of the hemostatic extravascular TF [Giesen et al., 1999]. 6005 0.4) cultured in 2TY medium (16 g tryptone, Antibodies and antibody fragments represent the 10 g yeast extract, and 5 g NaCl in 1 L distilled water) fastest growing segment of the biopharmaceutical was infected by the eluted phages at 371C for 30 min. A market [Nicolaides et al., 2006]. This is mainly due to portion of the infected E. coli TG1 was serially diluted their drug safety profiles and because it could elicit into 100, 101, 102, 103, and 104 dilutions. Ten ml of each clinical benefit by antagonizing a specific antigen dilution was spotted separately on TYE plates (15 g without the common side effects that are prevalent Bacto-Agar, 8 g NaCl, 10 g tryptone, 5 g yeast extract with small chemical entities due to their nonspecific in1 L distilled water including 100mg/ml ampicilin, and effect on homeostatic biochemical pathways. 1% glucose). These plates were incubated at 371C Phage display is a powerful and rapid means to overnight to titer the eluted phages by counting the select antibody fragments with specific affinities and clones on the TYE plates. At the same time, the rest of activities toward a target protein. Its power lies in the infected E. coli TG1 cells were all spread out on large combination of functional activity and genetic informa- TYE plates (including 100mg/ml ampicilin and 1% tion, although this technique has been used to a limited glucose) and also incubated overnight at 371C. The degree in the field of thrombosis and hemostasis. next day, bacteria on the large plates were scrapped Antibody libraries were used to select antibodies that and added into 100ml 2TY (including 100 mg/ml bind to the platelet receptor GPIIb/IIIa [Schwarz et al., ampicilin and 1% glucose) for amplification. Phages 2006] and human vascular endothelial growth factor amplified in E. coli TG1 were rescued as follows: [Lin et al., 2008]. We used two single-chain variable Helper phages (5 1011 KM13) were added to 10 ml of fraction antibody libraries to select an inhibitory the amplified E. coli TG1 cells and incubated for antibody fragment to human TF using phage display; 30 min at 371C. After centrifugation for 10 min at we selected a single-chain variable fraction (scFv) 3000 g, the supernatant was discarded and cells were antibody fragment that inhibits the in vitro function of resuspended in 50 ml 2TY medium containing TF, which therefore inhibits coagulation. 100mg/ml ampicilin, 50mg/ml kanamycin, and 0.1% glucose. These cells were amplified overnight at 301C, METHODS AND MATERIALS centrifuged at 3000 g for 15 min and the supernatant added to 10 ml of polyethylene glycol (PEG) for 1 h on Selection of Tissue Factor Binding Phages ice. After another centrifugation, the rescued phages Human single fold scFv libraries I and J (Tomlinson were diluted in 2 ml of PBS and stored at 41C for the I1J; http://www.geneservice.co.uk/products/proteomic/ next selection round in which 1012 rescued phages Drug Dev. Res. 144 145 TISSUE FACTOR INHIBITORY ANTIBODY FRAGMENT 201 were added into 2% SM solution to start the next plates to 50 ml 2TY (with 1% glucose) and grown at selection round. Four selection rounds were per- 371C for 2 h. Ten ml of these cultures was then added to formed. During the last selection round, the TF a centrifuge tube and 5 1010 helper phages were binding phages were specifically eluted with an excess added and incubated for 30 min at 371C. After amount of Factor VII to select inhibitory clones. centrifugation at 300 g for 10 min, the pellets were resuspended in 50 ml 2TY containing 100mg/ml Preparation of scFv Phage for Monoclonal Enzyme- ampicilin, 50mg/ml kanamycin and 0.1% glucose and Linked Immunosorbent Assay (ELISA) shaken at 301C overnight. The phages from the single From the last selection round, 400 single clones colonies were rescued the next day by centrifuging the from each library were randomly picked out from the overnight culture at 3000 g for 15 min. The supernatants TYE plates and cultured in different wells containing were added to 10 ml PEG/NaCl solution (20% poly- 100ml 2TY medium (including 100mg/ml ampicillin ethylene glycol [PEG] 6000, 2.5 M NaCl) for 1 h at 41C and 1% glucose) of four 96-well microculture plates. to precipitate the phages. After another centrifugation of After culturing overnight at 371C, a small inoculum (2ml) 3000 g for 30 min, the phage pellets were dissolved in was transferred to other plates with 200ml 2TY 2 ml PBS and the phage concentration determined as (including ampicilin and glucose) in each well. The follows. Phage concentration (phages/ml)5OD260 original plates were stored at 41C temporarily. The dilution 2.214 1011. The phages from the 14 stron- transferred plates were shaken at 371C for 2 h, and gest binding colonies were stored in PBS with 15% the phages were rescued in E. coli TG1 by adding 109 glycerol at 701C for further testing. helper phages to each well and shaken for 1 h at 371C before spinning the plate at 1800 g for 10 min. Pellets TF Activity Assay were suspended in 200ml 2TY (including 100mg/ml The Actichrome TF assay kit from American ampicilin and 50mg/ml kanamycin) and shaken overnight Diagnostica (USA) was used to determine the TF at 301C. The next day, culture plates were spun down at 1800 g for 10 min, and 100m inhibitory activity of the14 strongest TF bindingl of the supernatant contain- phages. An aliquot (450 ml) of the amplified phages of ing scFv-phage was used in a monoclonal phage ELISA. each of the 14 colonies was precipitated on PEG/NaCl Monoclonal phage ELISA and the pellet dissolved in 200ml assay buffer of the kit. The phage concentration of each colony was deter- Eight ELISA plates (Nunc. Maxisorp) were coated mined as previously described. An amount of 5 1010 overnight at 41C with 50ml of the TF preparation (4 phages of each colony was put into each of 2 wells on plates with Innovin and 4 plates with Thromborel). The an ELISA plate. A phage that did not bind or inhibit plates were washed 3 times with PBS and blocked with m TF was use as a control. TF (15 pM) was added to the2% SM (200 l per well) for 2 h at room temperature. m phages and incubated for 10 min at room temperatureAfter 3 washing steps with PBS, 50 l of the scFv followed by 25ml of Factor VII and 25 ml of Factor X. supernatant of each well of the cell culture plates were Following incubation for 15 min at 371C, 25ml of the added to each of the wells of the coated ELISA plates m substrate, Spectrozym Xa was added to each well andtogether with 50 l of 4% SM. After incubation for 1 h at the absorbance measured at 405 nm every 20 min for a room temperature, plates were washed 3 times with m period of 80 min. The 6 colonies with the lowest TFPBS-0.1% Tween-20. Subsequently, 100 l of a horse- activity, e.g., the six strongest inhibitors of the reaction radish peroxidase-anti-M13 antibody were added to m were diluted 1:2 for 4 times and the kinetic reactioneach well in a 1:5000 dilution on PBS-2%SM (100 l per repeated to determine concentration-dependent inhi- well) and incubated for another hour at room tempera- bition of TF by these colonies. ture. After 3 more washes with PBS-0.1% Tween-20, 100ml substrate solution (100mg/ml TMB in 100 mM sodium acetate, pH 6.0, with 30% hydrogen peroxide Prothrombin Times directly before use) was added to each well. The Prothrombin times (PT) were conducted on the reaction was stopped by adding 50ml 2 M H2SO4. three strongest inhibitory phage colonies selected via Absorbance at OD450 and OD650 was determined and the TF activity assay; 5 1010 phages were incubated the 14 colonies with the highest OD450–650 values were with the PT reagent, Thromborel for 10 min, after amplified (see next section). normal plasma was added and the time measured until the clot formed using a STAR-4 coagulation instrument Amplification of Single Colonies (Diagnostica Stago, Asnières sur Seine, France). The The 14 strongest tissue factor binding phages were strongest TF inhibitory clone was selected for the amplified by adding 50ml of the cell stock of the culture production of soluble scFv antibody fragments. Drug Dev. Res. 145 146 202 MEIRING ET AL. Production of Soluble scFv Antibody Fragments substrate, Spectrozym Xa were added to each well and The selected phage clone, JT C5 was used to the absorbance at 405 nm measured every 15 min for infect HB2151, a nonsuppressor E. coli strain that is 90 min. then induced to provide soluble expression of antibody fragments. An aliquot (3 ml) of overnight activated Prothrombin Times E. coli HB2151 (infected by the selected scFv phage) Prothrombin times were evaluated using antibody culture was transferred to 300 ml 2TY (0.1% glucose fragment concentrations of 1.5mM, 0.75mM and and 100 m/ml ampicilin) and shaken at 250 rpm at 371C 0.375mM. The antibody fragments were incubated for 73 h until the OD6005 0.9. Isopropyl b-D-thioga- with the PT reagent, Thromborel, and buffer (TBS) for lactoside (IPTG, final concentration 1 mM) was then 10 min. Normal plasma was then added and the time added and shaking continued overnight at 301C. measured until clot formation using a STAR-4 coagula- After centrifugation at 1800 g for 10 min, the super- tion instrument (Diagnostica Stago). natant containing the soluble antibody fragments was stored at 41C. 1.5 Purification of JT C5 scFv Antibody Fragments JT C5 scFv antibody fragments were purified from the cell culture medium by using a Protein A 1.0 column. Five ml of Protein A Sepharose Fast Flow Control JT C5 (Pierce, IL) was packed into a column. After rinsing 0.5 with 10 ml of binding buffer (0.1 M phosphate, 0.15 M NaCl, pH5 7.2), the cell culture medium was diluted 1:1 with binding buffer and applied to the column. 0.0 After the medium was run through, the column was 0 20 40 60 80 washed with a further 15 ml of binding buffer. The TIme (min) antibody fragments were eluted in 1-ml fractions with Fig. 1. Tissue factor activity of the 14 strongest tissue factor binding an acidic elution buffer (0.1 M glycine, pH 2–3). The phage colonies. 51010 phages of each colony were added to each of protein content of each fraction was measured at 2 wells on an ELISA plate. TF (15 pM) was added to the phages and OD280 and the fractions with the highest protein incubated for 10 min at room temperature. Factor VII (25 ml) was content were pooled. Purified scFv was evaluated by added to each well followed by Factor X (25ml). After incubation for 12% SDS-PAGE and Western blotting using a mouse 15 min at 371C, 25ml of the substrate, Spectrozym Xa was added to each well and the absorbance at 405 nm measured every 20 min for a anti-cMyc antibody (Pierce). The purified antibody period of 80 min. fragment was dialyzed with Tris-buffered saline (TBS) and the concentration determined via absorbance at 280 nm. JT C5 Functional Characterization of the Tissue Factor 1.5 Inhibiting Antibody Fragment The selected antibody fragment JT C5 was tested 1.0 0 for TF inhibitory activity using the Actichrome TF 6.125e9 1.25e10 activity assay. Prothrombin times were conducted with 2.5e10 different concentrations of the antibody fragment and 0.5 5.10e10 the effect of the antibody fragment was also tested on thrombin generation in normal plasma as well as plasma from a thrombophilic patient. 0.0 0 25 50 75 100 TF Activity Assay Time (min) Antibody fragment concentrations of 625ng, Fig. 2. Concentration-dependent tissue factor activity of the strongest 313ng, 156ng and 78ng were added in duplicate to an inhibiting phage clone, JT C5. Phage concentrations ranging from 9 ELISA plate. TF(15 pM) and assay buffer to a volume 6.12510 –510 10 were added to each of 2 wells on an ELISA plate. of 75ml were added to the antibody fragment and TF (15 pM) was added to the phages and incubated for 10 min at RT. Factor VII (25ml) was added to each well followed by Factor X (25 ml). incubated for 10 min at RT. Factor VII (25ml) was After incubation of 15 min at 371C, 25ml of the substrate, Spectrozym added to each well followed by Factor X (25ml factor). Xa were added to each well and the absorbance at 405 nm measured After incubation for 15 min at 371C, 25ml of the every 15 min for a period of 90 min (n5 3). Drug Dev. Res. 146 OD (405 nm) OD 405 nm 147 TISSUE FACTOR INHIBITORY ANTIBODY FRAGMENT 203 Thrombin Generation Assay AMC and 15 mM CaCl2 was added to start each The TECHNOTHROMBIN TGA kit (Techno- reaction that were conducted in duplicate. Fluores- clone, Surrey, UK) was used to determine thrombin cence was measured at 360 nm/460 nm (excitation/ generation with time in platelet-poor plasma upon emission) using the SYNERGY 2 Kinetic reader activation of the clotting cascade by TF that was (BIOTEK) at 1-min intervals for 2 h. incubated with different concentrations of the anti-TF RESULTS antibody fragment. The TGA reagent C that contains approximately 5 pM of TF was incubated with different FT Activity of the Strongest Tissue Factor-Binding JT C5 scFv concentrations (13, 6.25, and 3.125 mM) Phage Colonies and TBS for 10 min at 371C. Normal pooled plasma Figure 1 shows the effect of the 14 strongest TF and plasma from a thrombophilic patient with low free binding phages on the tissue factor activity assay. protein S level of 16% were then added in parallel to Approximately 6 phage clones inhibited the TF activity the reactions (40 ml per reaction). Fluorogenic TGA strongly. Colony JT C5 had the strongest inhibitory substrate (50ml/reaction) containing 1 mM Z-G-C-R- 60 PT's of different phage colonies 120 50 100 80 60 JT D1 40 40 JI A1 25 JT C5 20 30 15 10 5 20 0 0.0 0.5 1.0 1.5 2.0 0.0 1.5×1010 3.0×1010 4.5×1010 6.0×1010 µM anti-TF Antibody Phage concentration Fig. 5. Prothrombin times of the purified C5 scFv antibody fragment. Fig. 3. Prothrombin times of the 3 strongest tissue factor inhibiting Prothrombin times were conducted at antibody fragment concentra- phage clones. 51010 phages were incubated with the PT reagent, tions of 1.5, 0.75, and 0.375mM. The antibody fragments were Thromborel for 10 min, after which normal plasma was added and the incubated with the PT reagent, Thromborel, and buffer (TBS) for time measures until the clot was formed (n5 3). 10 min, after which normal plasma was added and the time measured until clot formation (n53). 1.00 kD 0 nM 66 0.75 53 nM 45 36 0.50 29 106 nM 212 nM 0.25 26 kD 24 26 kD 0.00 14 0 15 30 45 60 75 90 Time (minutes) JIA1 JTD1 JTC5 Fig. 6. Concentration-dependent inhibition of tissue factor activity by the purified JT C5 scFv antibody fragment. Antibody fragment JTD1 JIA1 Protein concentrations of 212, 106, and 53 nM were added in duplicate to JTC5 Marker the wells of an ELISA plate. TF (15 pM) and assay buffer to a volume of 75ml were added to the antibody fragment and incubated for 10 min at Fig. 4. SDS-PAGE and Western blotting of the antibody of the three room temperature. Factor VII (25ml) was added to each well followed strongest inhibiting phage clones. The purified scFv was evaluated by a by Factor X (25ml). After incubation for 15 min at 371C, 25 ml of the 12% SDS-PAGE and Western blotting, using a mouse anti-cMyc substrate, Spectrozym Xa were added to each well and the absorbance antibody. at 405 nm measured every 15 min for a period of 90 min (n53). Drug Dev. Res. 147 Time (seconds) OD (405 nm) Time (sec) 148 204 MEIRING ET AL. effect. There were three colonies that even increased Evaluation of the Strongest Inhibiting scFv Antibody the TF activity. The inhibitory effect of JT C5 is shown Fractions by SDS-PAGE and Western Blotting in Figure 2, where different concentrations of this SDS-PAGE and Western blotting (Fig. 4) showed phage were tested. At a concentration of 5 1010 that the scFv antibody fragments of the 3 strongest almost no Factor Xa activity was measured. The other inhibiting phage clones were expressed in the E. coli five inhibitory phages (JI B1, JI E3, JI A3, JI A1, and culture medium. The expression of scFv JTC5 and JT D1) were also tested for concentration-dependent JTD1 was better than that of JTA1. TF inhibition, but the results are not shown here as the inhibitory effects were not as strong as that Functional Characterization of the TF Inhibiting of JT C5. Antibody Fragment Prothrombin Times of the Three Strongest Tissue Purified JT C5 scFv increased the prothrombin Factor Inhibiting Phage Clones time dose-dependently (Fig. 5) with 1.5mM JT C5 scFv The effect of the three strongest inhibitory phage increasing PT twofold. Factor Xa activity was also clones (JT C5, JI A1, and JT D1) on the thrombin times inhibited by the JT C5 scFv (IC505 100 nM; Fig. 6). of normal plasma is shown in Figure 3. Again the Thrombin generation of normal plasma and of JT C5 phage colony showed the strongest inhibitory thrombophilic plasma was decreased with increasing effect. concentrations of the JT C5 scFv. The effect of JT C5 A Normal plasma 600.00 500.00 400.00 300.00 200.00 100.00 0.00 -100.00 Time (min) 13 M 6.5 M 3.125 M Control B Thrombophilia 500.00 400.00 300.00 200.00 100.00 0.00 -100.00 Time (min) 13 M 6.5 M 3.125 M Control Fig. 7. Thrombin generation assays of different C5 scFv antibody fragment concentrations with normal plasma (A) and thrombophilic plasma (B). The TGA reagent C from the Technothrombin TGA kit (Technoclone, Surrey, UK) that contains about 5 pM TF, was incubated with different C5 scFv concentrations (13, 6.25, and 3.125mM) and TBS for 10 min at 371C. Normal pooled plasma and plasma from a thrombophilic patient were then added in parallel to the reactions (40 ml per reaction). The fluorogenic TGA substrate (50 ml per reaction) was added to start each reaction. Fluorescence was measured at 360 nm/460 nm (excitation/emission) at 1-min intervals for 2 h. Drug Dev. Res. 148 nM Thrombin nM Thrombin 00:00:00 00:00:00 00:10:00 00:10:00 00:20:00 00:20:00 00:30:00 00:30:00 00:40:00 00:40:00 00:50:00 00:50:00 01:00:00 01:00:00 01:10:00 01:10:00 01:20:00 01:20:00 01:30:00 01:30:00 01:40:00 01:40:00 01:50:00 01:50:00 02:00:00 02:00:00 149 TISSUE FACTOR INHIBITORY ANTIBODY FRAGMENT 205 was more pronounced in thrombophilic plasma, since In summary, we selected a single chain variable the lag phase before the thrombin generation burst was antibody fragment that inhibits tissue factor activity in a also increased, which was not the case with normal concentration-dependent manner by using phage dis- plasma (Fig. 7). play technology. The effect on thrombin generation is more pronounced in thrombophilic plasma than in DISCUSSION normal plasma. TF inhibitors may provide an effective To our knowledge, JT C5 is the first human treatment for thrombotic disease without serious single-chain variable antibody fragment selected that bleeding complications. Additionally, phage display is bound to and inhibited the action of human TF using a powerful tool to select novel antithrombotics. phage display technology. JT C5 was identified using REFERENCES four rounds of selection. Most phage display applica- tions use 3–4 selection rounds [Schwarz et al., 2006; Boström SL, Hansson GFH, Sarich TC, Woltz M. 2004. The inhibitory effect of melagtran, the active form of the oral direct Lin et al., 2008]. Some 800 TF binding phages were theomin inhibitor ximelagatran, compared with enoxaparin and tested in the monoclonal phage ELISA for TF binding r-hirudin on ex-vivo thrombin generation in human plasma. larger than other efforts [Zhang et al., 2007; Schwarz Thromb Res 113:85–91. et al., 2008; Lin et al., 2008]. Frédérick R, Pochet L, Charlier C, Masereel B. 2005. Modulators of Expression of JT C5 scFv was evaluated using the coagulation cascade: focus and recent advances in inhibitors of SDS-PAGE and Western blotting. The scFv has Tissue Factor, Factor VIIa and their complex. Curr Med Chem 12:397–417. a molecular weight of 26 kD and was purified on a protein A column, since it gives a better yield than a Giesen PLA, Rauch U, Bohrmann B, Killing D, Roque M, Fallon JT, Badiman JJ, Himber J, Riedere MA, Nemerson Y. 1999. Proc Natl protein L column. The yield, however, is insufficient Acad Sci USA 96:2311. for in vivo testing with only about 1 mg of scFv from 1 L Hirsh J, Weitz JI. 1999. Thrombosis: new antithrombotic agents. of E. coli culture. Larger E. coli cultures as well as Lancet 353:1431–1436. larger columns are required to ensure a better yield of Kirchhofer D, Moran P, Bullens S, Peale F, Bunting S. 2005. the purified scFv fragment. Purified JT C5 scFv A monoclonal antibody that inhibits mouse tissue factor function. increased prothrombin time in a concentration-depen- J Thromb Haemost 3:1–2. dent manner with a two-fold prolongation at 1.5mM Lin Z, Cao P, Lei H. 2008. Identification of neutralizing scFv (39 ng/ml). This is 1,000 times stronger inhibition than binding to human vascular endothelial growth factor 165 the rat anti-mouse monoclonal antibody, 1H1 [Kirch- (VEGF165) using a phage display antibody library. Appl Biochem Biotech 144:15–26. hofer et al., 2005]. It also slowed Factor Xa production concentration-dependently in the TF activity assay with Mann KG, Butenas S, Brummer K. 2003. The dynamics of thrombin formation. Arterioscler Thromb Vasc Biol 23:17. an IC50 value of 100 nM, 10 times stronger than that Miura M, Seki N, Koike T, Ishihara T, Hirayame F, Shigengaga T, observed for the monoclonal anti-TF antibody, TF4A12 Sakai-Moritani Y, Tagawa A, Kawasaki T, Sakamoto S, Okada M, [Peng et al., 2007]. It is also a stronger inhibitor than Ohta M, Tsukamato S. 2007. Design, synthesis and biological other orally available TF/FVIIa complex inhibitors activity of selective and orally available TF/FVIIa complex [Miura et al., 2007]. inhibitors containing non-amide P1 ligands. Bioorg Med Chem The thrombin generation assay shows a reduction 15:160–173. in the peak thrombin concentration as well as an Nicolaides NC, Sass PM, Grasso L. 2006. Monoclonal antibodies: a increase in the lag phase before thrombin is generated. morphing landscape for therapeutics. Drug Dev Res 67:781–789. The effect was more pronounced in thrombophilic Peng ZP, Cai X, Zhang Y, Kong D, Guo H, Liang W, Tang Q, Song H, Ma D. 2007. A novel anti-tissue factor monoclonal plasma than in normal plasma, indicating that TF antibody with anticoagulant potency derived from synthesized inhibitors may achieve anti-coagulant activity in multiple antigenic peptides though blocking FX combination with thromboplilic plasma without interfering with hemos- TF. Thromb Res 121:85–93. tasis in normal plasma. The reduction in peak thrombin Schwarz M, Meade G, Stoll P, Ylanne J, Bassler N, Chen Y, generation by JT C5 scFv was about 30 times less than Hagemeyer C, Ahrens I, Moran N, Kenny D, Fitzgerald D, the inhibitory effect of direct thrombin inhibitors such Bode C, Peter K. 2006. Conformation-specific blockade of the as Ximelagatran [Boström et al., 2004]. integrin GPIIb/IIIa: a novel antiplatelet strategy that selectively targets activated platelets. Circ Res 99:25–33. The limitation of the current scFv phage display Zhang B, Zhang Y, Wang J, Zhang Y, Chen J, Pan Y, Pen L, Hu Z, technology is the production of sufficient scFv Zhao J, Liao M, Wang S. 2007. Screening and identification of a fragments. A larger production system is needed to targeting peptide to hepatocarcinoma from a phage display enable further testing in in vivo studies. peptide library. Mol Med 13:246–254. Drug Dev. Res. 149 S-14 150 Protein Expression and Purification 151 (2018) 62–71 Contents lists available at ScienceDirect Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep Evaluation of in vitro refolding vs cold shock expression: Production of a low T yielding single chain variable fragment Jan-G. Vermeulena,∗, Felicity Burtb,c, Esta van Heerdend, Errol Casond, Muriel Meiringa,b a Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, South Africa bNational Health Laboratory Service, Universitas, Bloemfontein, South Africa c Division of Virology, Faculty of Health Sciences, University of the Free State, South Africa d Department of Microbial, Biochemical and Food Biotechnology, Faculty of Agricultural Sciences, University of the Free State, South Africa A R T I C L E I N F O A B S T R A C T Keywords: The development of therapeutic antibodies in their various forms has been a constant challenge since the de- Inclusion bodies velopment of the first monoclonal antibodies in 1975. This is especially true for the development of therapeutic Single chain variable fragment single chain variable (scFv) fragments in Escherichia coli. In a previous study the selection of a tissue factor In vitro protein refolding inhibiting single chain variable fragment (TFI-scFv) isolated from the Thomlinson I + J phage libraries was Cold shock described. Although the initial findings were promising, additional characterization of the antibody fragment Disulfide bridge Tissue factor and subsequent application was hampered due low protein yield. This study reports on: i) the improved ex- Fusion partner pression of a previously low yielding TFI-scFv in the cytoplasm of E. coli BL21 (DE3) through modifications to the expression systems in conjunction with codon optimization ii) evaluation of two commercial methods of protein recovery: in vitro refolding and the utilization of cold shock expression systems in conjunction with E. coli SHuffle. Results showed that TFI-scFv could be expressed at higher levels in the cytoplasm of E. coli than pre- viously achieved in the periplasm. Both the in vitro refolding and cold shock strategies were capable of producing functional TFI-scFv with varying degrees of success. These procedures could be applied to improve the pro- duction of other problematic low yielding scFv isolated from phage display repositories in order to facilitate their characterization. 1. Introduction disease and mortality [6,7]. As a result, the inhibition of TF holds great potential for the development of novel anti-thrombotic agents. The development of antibodies as therapeutic agents has always Despite phage display's exceptional ability to generate antibodies been an attractive prospect due to their high specificity, affinity to a against a variety of targets, the utilization of the scFv remains limited wide variety of molecules and stability [1]. Furthermore, the use of due to complications that arise during protein production. Expression in antibody sub-fragments is especially attractive due to the relatively Escherichia coli systems are often preferred for production of ther- small size (27 kDa), low immunogenicity and relative ease of selection apeutics, for laboratory as well as industrial scales, due to low cost and by means of phage display [2]. Previously we reported on the utiliza- simplicity of cultivation. However, the successful production of func- tion of phage display technology to select a scFv from the Tomlinson tional protein is often hampered due to codon bias and incorrect folding I + J Human Single Fold Phage Libraries that functionally inhibits of the target protein [8–12]. Thus, a constant challenge in the field of human tissue factor [3]. Tissue factor is constitutively expressed in protein engineering is to overcome the shortcomings of the native vascularized organs [4]. The distribution of tissue factor in vital organs molecular mechanisms through alternative expression strategies [13]. provide a protective hemostatic barrier in order to minimize blood loss Due to the high frequency of scFv aggregation, the regeneration of during injury [5]. Despite this critical role in hemostasis, excessive or functional scFv from insoluble aggregates through in vitro refolding aberrant expression of tissue factor has been linked to inflammatory techniques has become standard practice [14–16]. Various refolding and thrombotic disorders, which are major contributors to global techniques have been developed to refold scFv, which include direct Abbreviations: Codon Adaptation Index, CAI; Dimethyl sulfoxide, DMSO; E.coli Codon Usage Analyzer, ECUA; Escherichia coli, E.coli; Guanidinium-HCl, GdnHCl; Heavy chain variable domains, VH; Immobilized Metal Affinity Chromatography, IMAC; Isopropyl β-D-1-thiogalactopyranoside, IPTG; Light chain variable domains, VL; Medical Research Council, MRC; Oxidized glutathione, GSSG; Prothrombin Time, PT; Reduced glutathione, GSH; Single chain variable fragment, scFv; Tissue factor, TF; Tissue Factor Inhibitor scFv, TFI-scFv; Tris Buffered Saline, TBS ∗ Corresponding author. Department of Microbial, Biochemical and Food Biotechnology, Faculty of Agricultural Sciences, University of the Free State, South Africa. E-mail address: slimskim@gmail.com (J.-G. Vermeulen). https://doi.org/10.1016/j.pep.2018.06.005 Received 4 June 2018; Accepted 8 June 2018 Available online 09 June 2018 1046-5928/ © 2018 Elsevier Inc. All rights reserved. 150 151 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 dilution, dialysis, diafiltration, chromatographic methods, matrix-as- 2.3. Homology modelling of TFI-scFv sisted refolding, and chemical-assisted refolding with varying levels of success [10,17–20]. The major disadvantage to commonly used in vitro DNA sequence of TFI-scFv was translated to the final amino acid refolding techniques is that they are often complex and require in- sequence using online translations tool ExPASy: translate from Swiss tensive optimization of multiple operational steps in order to obtain Institute of Bioinformatics. The amino acid sequence was used to model functional protein [20]. Target protein recovery yields are also highly a three-dimensional structure of TFI-scFv. Homology modelling of TFI- variable due to the protein specific nature of in vitro refolding techni- scFv was performed using Yet Another Scientific Artificial Reality ques [21]. Application (YASARA) version 16.4.6.L.64 with the following set As an alternative approach, cold shock expression is a popular parameters; Modelling speed: Slow, Number of PSI-BLAST iterations in method used to overcome protein aggregation. It utilizes the native template search (PsiBLASTs): 3, Maximum allowed (PSI-) BLAST E- response of E. coli to sudden reductions in temperature to overcome value to consider template (EValue Max): 0.5, Maximum number of protein aggregation [22]. Sudden temperature fluctuations strongly templates to be used (Templates Total): 5, Maximum number of tem- influence gene regulation in mesophilic bacteria such as E. coli [23–25] plates with same sequence (Templates SameSeq): 1, Maximum oligo- Additionally, a number of chaperones are up-regulated with increased merization state (OligoState): 4 (tetrameric), Maximum number of activity at lower temperatures which in turn facilitates appropriate alignment variations per template: (Alignments): 5, Maximum number protein folding [26]. These expression mechanisms have also been of conformations tried per loop (LoopSamples): 50 and Maximum successfully utilized to facilitate the solubilization of proteins con- number of residues added to the termini (TermExtension): 10 [37]. taining multiple disulfide bonds [27]. In conjunction with cold shock, the utilization of fusion partner expression systems has also been shown 2.4. Construction of TFI-scFv expression vectors to facilitate the production of soluble recombinant proteins in E. coli [28]. Fusion expression systems make use of a highly soluble partner The codon-optimized gene was synthesized for expression in E. coli that is fused to the N-terminus of the target protein in order to improve by GeneART (Regensburg, Germany). The codon-optimized TFI-scFv the solubility of the target protein [29]. sequence was synthesized to incorporate 5′ NdeI and 3′ XhoI restriction An additional complication to consider is the presence of inter do- sites for downstream cloning. The optimized scFv gene was amplified main stabilizing disulfide bonds. All antibodies have highly conserved using sense primer 5′-GAAGGAAGGCCGTCAAGG-3′, anti-sense primer: disulfide bonds and scFv contains one intra-domain chain disulfide 5′-CTCGAGATTACGTTTAATTTCAACTTTGG-3′ and sub cloned into bond for each of the variable regions [30]. The reducing conditions of NdeI/XhoI linearized expression vector pET-22 b (+), pCOLD DNA II, the E. coli cytoplasm hampers the formation of disulfide bonds that are and pCOLD TF expression vectors. The new constructs were designated necessary for stable tertiary structure formation [31,32]. It has been as pET22 TFI-scFv (Supplementary Fig. 1), pCOLD DNA II TFI-scFv, and shown that, modification of the expression host itself can promote pCOLD TrF TFI-scFv (Supplementary Fig. 2) respectively. Constructs disulfide bond formation in the cytoplasm. The enhanced oxidizing were verified by Sanger DNA sequencing. cytoplasmic conditions in combination with improved isomerization capability have been shown to contribute to improve target protein 2.5. Sanger DNA sequencing of the scFv gene fragment solubilization through improved disulfide bond formation [33]. In this study we report on the (i) improved expression of a pre- All DNA sequencing reactions were performed using the BigDye® viously low yielding TFI-scFv against human tissue factor by means of Terminator v3.1 cycle sequencing kit (Applied Biosystems, USA) in codon optimization, (ii) we evaluate the effectiveness of two commonly conjunction with the 3130xl Genetic Analyzer HITACHI (Applied utilized strategies for the generation of functional scFv namely: the Biosystems, USA) according to manufacturer's specifications. The ori- directed regeneration of inclusion bodies through in vitro refolding and ginal pIT2 plasmid containing the TFI-scFv gene was sequenced using the expression of functional TFI-scFv using a commercial (TAKARA) LBR (sense) 5′-CAGGAAACAGCTATGAC-3′ and pHEN (anti-sense) cold shock pCOLD DNA II and pCOLD Trigger Factor expression vectors 5′-CTATGCGGCCCCATTCA-’3 primers. The pET22-TFI expression in conjunction with E. coli BL21 (DE3) and E. coli SHuffle strains. vector was sequenced using generic T7 Promoter Primer: 5′TAATACG ACTCACTATAGGG 3′ and T7 Terminator Primer: 5′GCTAGTTATTGCT CAGCGG 3’ (Integrated DNA Technologies, Belgium. The pCOLD DNA II 2. Materials and methods TFI-scFv expression vector construct was sequenced using sense primer 5′-ACGCCATATCGCCGAAAGG-3′ and universal anti-sense primer 2.1. Bacterial cell lines and plasmids 5′-GGCAGGGATCTTAGATTCTG-3’. The pCOLD TF expression vector constructs were sequenced using 5′-CCACTTTCAACGAGCTGATGA-3′ in E. coli strain TG1 (MRC Centre for Protein Engineering, Cambridge, conjunction with universal anti-sense primer 5′-GGCAGGGATCTTAGA UK) was used for phagemid production. Cloning host, E. coli Top10 TTCTG-3’ (Integrated DNA Technologies, Belgium). (Invitrogen, USA) was used for the in vitro amplification of pSMART (Lucigen, USA) cloning vectors. All scFvs constructs were expressed 2.6. In vitro refolding of TFI-scFv inclusion bodies using either pET22 (Novagen, Germany), pCOLD DNA II or pCOLD TF expression vectors (TAKARA Bio Inc, Japan) utilizing E. coli BL21 (DE3) 2.6.1. Expression of pET22 TFI-scFv in E. coli BL21 (DE3) (Invitrogen, USA) or E. coli SHuffle® (New England Biolabs, USA). E. coli BL21 (DE3) cells transformed with pET22 TFI-scFv in parallel with a control harboring empty pET22b (+) plasmid were incubated in 2.2. Rare codon analysis LB-media (5 g L−1 NaCl, 5 g L−1 yeast extract, 10 g L−1 Bacto-tryptone) containing 100 μgmL−1 ampicillin for 16 h at 37 °C, 160 rpm. The Rare codons were identified using E. coli Codon Usage Analyzer v2.1 overnight cultures were used at 1:100 dilution to inoculate a 5-L (ECUA) (http://www.faculty.ucr.edu/). The frequency and distribution Sartorius bioreactor filled with 4 L LB media containing 100 μgmL−1 of codons was analyzed using Graphical codon usage analyzer v2.0 ampicillin, controlled by BIOSTAT B Plus universal controller unit utilizing codon usage table for E. coli [34]. The Codon Adaptation Index (Sartorius, Germany). The cultures were incubated at 37 °C, 300 rpm (CAI) value was calculated using the CAIcal server [35]. Distribution of agitation and 2 L per minute (lpm) aeration up to mid-log phase rare codons within the native TFI-scFv (JTC5-scFv) was visualized using (OD600= 0.6) and induced by 1mM isopropyl β-D-1-thiogalactopyr- web-based application Protter: Interactive protein feature visualization anoside (IPTG). Cultures were harvested by centrifugation at 15,000 g and integration with experimental proteomic data v1.0 [36]. for 10min at 4 °C, the supernatant discarded, and cell pellet stored at 63 151 152 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 −20 °C. Harvested cells were disrupted using the one-shot homogenizer USA) as described by Stanton and co-workers [40]. The soluble fraction (Constant Systems, UK) at 30,000 KPSI. Soluble and insoluble fractions was isolated by centrifugation at 15,000 g for 10min at 4 °C and eval- were separated by centrifugation at 15,000 g for 10min at 4 °C. The uated by 10% SDS-PAGE. Densitometric analysis was performed using soluble fraction was further fractionated using ultra-centrifugation at ImageJ v 1.51 d (http://imagej.nih.gov/ij/) image analysis software. 100,000 g for 20min at 4 °C. All fractions were evaluated by 10% SDS- PAGE as described by Laemmli (1970) and stained as described by 2.7. Cold shock expression Fairbanks and co-workers (1971) [38,39]. 2.7.1. Expression of pCOLD DNA II TFI-scFv and pCOLD TrF TFI-scFv in E. 2.6.2. Isolation of inclusion bodies coli BL21 (DE3) and E. coli Shuffle Cell pellets were thawed on ice and washed twice using disruption Chemically competent E. coli BL21 (DE3) and E. coli SHuffle cultures buffer (50mM Tris buffer containing 500mM NaCl, 10mM EDTA, were transformed using pCOLD DNA II TFI-scFv, pCOLD TrF TFI-scFv, 5 mM DTT and 2% Triton X-100 at pH 7.5). Cells were disrupted using pCOLD DNA II (Empty Control), and pCOLD TrF (Empty Control). The the one-shot homogenizer (Constant Systems, UK) at 30,000 KPSI. transformants were incubated as described in section 2.4.1 with slight Soluble and insoluble fractions were separated by centrifugation at modification. Once mid-log phase (OD600= 0.6) was reached, cold 15,000 g for 10min at 4 °C. The insoluble fraction was again washed shock was initiated. The incubation chamber was submerged in ice twice using disruption buffer and a third time using a chelating buffer slurry to enable rapid temperature drop required for cold shock. Once (50mM Tris and 10mM EDTA at pH 7.5). The pellets were suspended the cultures reached 4 °C the bioreactor was removed from the ice- in a minimal volume of dimethyl sulfoxide (DMSO) and kept at room slurry and incubated for an additional 30min at 15 °C. Expression was temperature for 30min. Pellets were solubilized in denaturing buffer induced by adding 1mM IPTG (final concentration) and 24 h incuba- (50mM Tris at pH 7.5 containing 6M guanidinium-HCl (GdnHCl) and tion at 15 °C with 300 rpm agitation. The following day cultures were 5mM dithiothreitol) at room temperature for 3 h with slow magnetic harvested by centrifugation at 15,000 g for 10min at 4 °C stored at stirring. The soluble and insoluble fractions were further separated −20 °C. Harvested cells were disrupted using the One-Shot using ultra-centrifugation at 100,000 g for 20min at 4 °C in order to Homogenizer (Constant Systems, UK) at 30,000 KPSI. The soluble- and remove residual membrane fraction. Protein concentrations of the so- insoluble fractions were separated by means of centrifugation at lubilized fractions were determined using the Q-Bit Illuminometer and 15,000 g for 10min at 4 °C. The membrane fraction was then pelleted Q-Bit Protein assay (ThermoFisher Scientific, USA) according to man- by means of ultra-centrifugation at 100,000 g for 90min at 4 °C. All ufacturer's specifications. fractions were evaluated by 10% SDS-PAGE and densitometry as pre- viously described. 2.6.3. Purification of TFI-scFv using IMAC under denaturing conditions TFI-scFv inclusion bodies were purified under denaturing conditions 2.8. Purification of TFI-scFv by immobilized metal affinity chromatography using a 5-mL HisTrap FF column (GE Healthcare, USA) in conjunction with the ÄKTAprime plus (Amersham Biosciences, UK) chromatography Fractions containing the soluble TFI-scFv and pCOLD TF and fusion system according to manufacturer's specification. The system was construct Trigger factor -TFI-scFv (TrF:TFI-scFv) were purified using a equilibrated with 100mL binding buffer at 5mLmin−1 (50mM Tris 5-mL HisTrap FF column (GE Healthcare, USA) in conjunction with the containing 6M Gdn-HCl, 0.5 M NaCl and 20mM imidazole at pH 7.5). ÄKTAprime plus (Amersham Biosciences, UK) chromatography system The denatured sample was diluted (1:10) with binding buffer and according to manufacturer's specifications. The column and system loaded at 1mLmin−1. The column was then washed with 50mL were initially equilibrated with 100mL binding buffer (50mM Tris, binding buffer at 1mLmin−1. Elution was performed in binding buffer 150mM NaCl containing 20mM Imidazole at pH 7.5) at 5mLmin−1. using a 100mL linear Imidazole gradient (20–300mM) at 1mLmin−1. The TFI-scFv and TrF-TFI-scFv sample was diluted (1:10) with binding Protein concentration of the solubilized fractions was determined using buffer and loaded at 1mLmin−1. The column was then washed with 10 the Q-Bit illuminometer and Q-Bit Protein assay. Purified TFI-scFv was column volumes binding buffer at 1mLmin−1. Target protein was resolved by means of 10% SDS-PAGE as described previously. eluted from the column using a 100-mL linear imidazole gradient (20–500mM) at 1mLmin−1. The TrF:TFI-scFv construct contains a 2.6.4. Screening of refolding conditions thrombin cleavage (Leu-Pro-Val-Arg-Gly-Ser) site nested between the Refolding conditions were screened using the Pierce Protein trigger factor and the insert (Supplementary Fig. 2). Target TFI-scFv Refolding Kit (Thermo scientific, USA) according to manufacturer's was cleaved from the TrF-TFI-scFv fusion construct using the Thrombin specifications. Due to the presence of intra-domain disulfide bonds, Cleancleave™ Kit (Sigma-Aldrich, USA) per manufacturer's instructions. redox conditions were investigated using oxidized (GSSG) and reduced Briefly, 1mL batch reactions were performed by adding 1mg of fusion (GSH) glutathione (Supplementary Material: Table 1). Batch-wise re- construct to agarose immobilized thrombin resin. Reactions were in- folding reactions were performed by a 20-fold dilution of the denatured cubated for 8 h at room temperature and then stored overnight at 4 °C. TFI-scFv in each of the 1mL screening reactions. The final refolding The following day the cleavage reactions were centrifugation at 500g reactions containing 500 μgmL−1 TFI-scFv were performed overnight for 5min and the supernatant collected and pooled together. The at 4 °C. Buffer exchange was performed by overnight dialysis at 4 °C cleaved trigger factor and TFI-scFv were separated by means of a sec- against 3 L TBS (50mM Tris containing 150mM NaCl, pH 7.5) using ondary purification using a 5-mL HisTrap FF column in conjunction 10,000MWCO SnakeSkin pleated dialysis tubing (Thermo scientific, with the ÄKTAprime as described previously. The purified TFI-scFv Table 1 Isolation and purification of denatured TFI-scFv. Total biomass (wet weight in g) a Isolated Inclusion aPurity aCalculated TFI-scFv in bTotal Purified bPurity bRetention bodies inclusion body fraction Denatured TFI-scFv 5.5 g L−1 54mg L−1 > 80% 37mg L−1 4.25mg L−1 > 95% 11.% The in vitrorefolding of TFI-scFv. a Prior to IMAC. b Following IMAC. 64 152 153 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 Fig. 1. Homology model of TFI-scFv and the distribution of rare codons. fractions were pooled together and concentrated using 10,000MWCO VIVASPIN 20 columns (Sartorius, Germany). The protein concentra- tions of the solubilized fractions were determined using the Q-Bit Illu- minometer and Q-Bit Protein assay. All fractions were evaluated by 10% SDS-PAGE and densitometry as previously described. 2.9. Prothrombin times Diluted prothrombin time tests (PTs) were performed using Dade C Control platelet poor plasma (Siemens Healthcare, Germany) in con- junction with the STAR-4 coagulation instrument (Diagnostica Stago). Tissue factor (Dade Tromborel S, Siemens Healthcare, Germany) was individually incubated with escalating dosages (0.0 mgml−1 to 0.2 mgml−1) of TFI-scFv at 37 °C for 10min prior to testing. PTs were performed in triplicate and the average clotting time was calculated. The inhibition effect (extended clotting time) of the TFI-scFv fractions was calculated as the difference between the average clotting time and the blank (negative control). Fig. 2. Codon frequency of usage profile. 3. Results 3.1. Rare codon optimization Analysis of the TFI-scFv gene was performed using ECUA software with a 10% codon frequency threshold based on highly and con- tinuously expressed genes during exponential growth phase (Class II genes). The analysis of the gene identified a total of 45 (approximately 20%) rare codons distributed throughout the Heavy- and Light chains of the scFv (Fig. 1 and suppl. Fig. 3), which contribute to low expression levels. A total of 150 out of 242 codons (approximately 62%) were substituted in the optimized gene. The CAI index value for the native gene increased from 0.61 to 0.91 for expression in E. coli (Fig. 2). Two ribbon diagrams of the TFI-scFv homology model are shown at Fig. 3. Cytoplasmic expression of TFI-scFv. 180° rotations with the flexible (GGGGS)×3 linker indicated as pre- dicted by YASARA software. The distribution and positions rare codons in TFI-scFv are indicated in red, with the Cysteine disulfide bridges case of rare codons may lead to the truncation of proteins. Codons with located within the heavy- and light chains in yellow. C-terminal His- high relative frequency (highly utilized) generally improve expression and Myc-tag has been omitted from diagram. efficiency. The comparative summary of codon frequency between the native Multiple alignment of the native and optimized DNA sequences TFI-scFv and optimized TFI-scFv. A value of 0–100 was assigned to reveals the extent of modification to the original sequence synonymous codons for each amino acid based on their frequency of (Supplementary Material: Table 2) while the final amino acid sequences utilization according to the E. coli codon usage database. Codons with have remained unchanged (Supplementary Material: Table 3). low relative frequency values indicate less preferred or rare codons. Low frequency codons have lower expression efficiency and as with the 65 153 154 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 Table 2 Summary of solubilization efficacy. Refolding Buffer Denaturant Aggregated TFI-scFv (μg) Soluble TFI-scFv (μg) % Solubility 1 Range 1 439 61.3 12 2 490 71.0 14 3 374 126 25 4 Range 2 462 37.4 7.5 5 300 200 40 6 372 128 26 7 Range 3 198 302 60 8 261 239 48 9 219 280 56 Table 3 Analysis of prothrombin times. Refolding Buffer Average PT Time extended Final Concentration mg.mL−1 Relative Inhibition Specific activity Time extended/ Effectiveness (%) Concentration s/mg.mL−1 Blank 20.6 0 0 0 0 1 23.1 2.4 0.245 14 9.79 2 24.8 4.1 0.282 24 14.5 3 23.7 3.1 0.503 18 6.16 4 22.3 1.6 0.150 10 10.6 5 25.1 4.5 0.8 26 5.63 6 35.8 15.2 0.512 100 29.6 7 21.2 0.6 0.121 4 4.96 8 24.8 4.1 0.956 24 4.28 9 20.9 0.3 0.112 2 2.68 3.2. In vitrorefolding of TFI-scFv inclusion bodies purity (Fig. 4A: Lane 7), which correlates to production of approxi- mately 37mg L−1 of TFI-scFv. During IMAC purification of the solubi- 3.2.1. Cytoplasmic expression of codon-optimized TFI-scFv lized inclusion bodies under denaturing conditions, a protein absor- SDS-PAGE analysis revealed (Fig. 3) that a protein with a molecular bance peak, corresponding to the 70–100mL fractions, were pooled weight of approximately 27 kDa was highly expressed (lane 2) which together (Fig. 4B) and was analyzed by means of SDS-PAGE. The de- correlates with the theoretical molecular weight of TFI-scFv. TFI-scFv natured 27 kDa TFI-scFv was thus isolated with high purity (Fig. 4C: constitutes approximately 40% of the total cellular protein. Generally, Lane 1). Densitometry analysis (Supplementary Material: Fig. 4) in- expression ranged from 20 to 30% of total cellular protein. The majority dicated that the denatured 27 kDa TFI-scFv was successfully isolated of expressed TFI-scFv accumulated in the insoluble cytoplasmic fraction and purified with a purity greater than 95%. The denatured TFI-scFv (Lane 4) while the soluble fraction (Lane 8) only contained minimal yield was approximately 4.25mg L−1 TFI-scFv. Isolation and purifica- amount (< 10%) of the 27 kDa protein TFI-scFv. tion of denatured TFI-scFv is summarized in Table 1. SDS-PAGE analysis of codon-optimized TFI-scFv expression. The A) SDS-PAGE analysis of the insoluble cytoplasmic fraction. Lane M: expected molecular weight of TFI-scFv is 27 kDa. Lane M, protein Protein marker. Lane 1: Whole cell fraction. Lane 2: Disruption buffer marker. Lane 1: Whole cell fraction of E. coli BL21 transformed with wash I. Lane 3: Disruption buffer wash II. Lane 4: Chelating buffer empty pET22b (+) plasmid (negative control). Lane 2: Whole cell wash. Lane. 5: Solubilized inclusion bodies (12-fold dilution). Lane 6: fraction of E. coli BL21 transformed with pET22 TFI-scFv. Lane 3: Solubilized inclusion bodies (6-fold dilution). Lane 7: Solubilized in- Insoluble cytoplasmic fraction (negative control). Lane 4: Insoluble clusion bodies (30-fold dilution). B) AKTÄ Purification Profile of TFI- cytoplasmic fraction (pET22 TFI-scFv). Lane 5: Membrane fraction scFv under denaturing conditions. C) SDS-PAGE analysis of the purified (negative control). Lane 6: Membrane fraction (pET22 TFI-scFv). Lane denatured TFI-scFv (70–100mL pooled fraction). Lane M: Protein 7: Ultracentrifuge soluble fraction (negative control). Lane 8: Ultra marker. Lane 1: Purified TFI-scFv. centrifuge soluble fraction (pET22 TFI-scFv). Refolding reactions displayed varying levels of protein aggregations in all the samples (Supplementary Material: Fig. 5). Refolding buffer 4 3.3. Isolation and purifications of TFI-scFv produced the lowest level of solubilization (7.5%) while refolding buffer 7 produced the highest level of solubility (60%). Solubilization The SDS-PAGE analysis (Fig. 4A) of the insoluble cytoplasmic data is summarized in Table 2. fraction. TFI-scFv contributed to roughly 22% of the total cellular The redox ratio of 1 to 1 oxidized/reduced glutathione (1mM) protein of the whole cell fraction (Fig. 4A lane 1) with. The washing and (sample 3, 5 and 7) produced the highest level of solubility of each of denaturation steps (Fig. 4A: Lanes 2–4) removed most inclusion body the three redox ratios examined across all three denaturant ranges. The associated proteins. The low-quality migration observed in Fig. 4A lowest levels of solubilization was achieved with oxidized/reduced lanes 5–7 is due to the high concentration of guanidinium chloride glutathione ratio of 1–10 in range 1 (Buffer 1), a ratio 1 to 5 in range 2 (6M) required to denature the target protein. As the dilution ratio in- (Buffer 4) and a ratio of 1–10 in range 3 (Buffer 8). creased (6–30-fold dilution) the migration normalized. The final pur- ified inclusion body fraction contained TFI-scFv with greater that 80% 66 154 155 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 Fig. 4. Isolation and purification of scFv from insoluble cytoplasmic fraction. Fig. 5. Expression levels of TFI-scFv and TrF-TFI-scFv. 3.4. Prothrombin times of the pCOLD DNA II TFI-scFv construct (lane 2) resulted in the pro- duction of TFI-scFv (27 kDa) which consisted of approximately 15% of Diluted Prothombin Times (PTs) were performed in triplicate and the total protein. The over expression of modified fusion partner the average PT was calculated (Table 3). The inhibition effect of each (Trigger factor 52 kDa) by pCOLD TrF (Empty Control) is clearly visible refolded TFI-scFv was calculated as the difference between the average in lane 3. The Trigger factor constituted approximately 25% of the total PT and the blank (negative control). All refolded samples were capable protein. The 79 kDa TrF-TFI-scFv fusion construct (Trigger factor of extending the baseline PT value (Negative control of 20,6 s). The (52 kDa) + TFI-scFv (27 kDa)) is also clearly visible in lane 4. Simi- highest inhibition effect was observed with sample 6 (15.2 s), which is larly, the fusion construct constituted approximately 25% of total an approximate 84% extension of the baseline PT value. As a result, protein production. Both of the constructs were also successfully ex- inhibition effectiveness of the refolded TFI-scFv was expressed as a pressed by E. coli SHuffle (Fig. 5B). The 27 kDa antibody expressed from percentage value relative to sample 6. Refolding buffer 2, 5 and 8 were the pCOLD DNA II TFI-scFv vector was approximately 10% of total capable of relative inhibition of approximately 25% while refolding protein (lane 2). Interestingly, the production levels of the trigger factor buffers 4, 7 and 9 had a relative effectiveness of less than 10% relative fusion partner (lane 3) is slightly reduced to 15% of total protein in to refolding buffer 6. The specific activity was calculated as the time comparison to the 25% of total protein attained when expressed by E. extended in seconds per mg.mL−1 of TFI-scFv added to each reaction. coli BL21 in Fig. 5A lane 3. The TrF-TFI-scFv fusion construct (lane 4) constituted approximately 25% of total protein. 3.5. Summary of best in vitro refolding reaction Whole cell fraction SDS-PAGE analyses of pCOLD DNA II and pCOLD TrF expression profiles in A) E. coli BL21 (DE3) and B) E. coli When considering the level of solubilization in conjunction specific SHuffle. Lane M: Protein Marker, Lane 1: pCOLD DNA II (Empty activity of various refolding condition is clear that refolding buffer 6 Control), B) Lane 2: pCOLD DNA II TFI-scFv, Lane 3: pCOLD TrF (Empty (800mM guanidium, 800mM L-arginine, 2 mM glutathione (GSH), Control) and lane 4: pCOLD TrF TF-scFv. The 27 kDa TFI-scFv is in- 0.2 mM glutathione disulfide (GSSG)) was the most effective in pro- dicated as TFI (lane 2), the 52 kDa modified trigger factor is indicated ducing functional TFI-scFv. Although several of the refolding were more as TrF (lane 3) and the 79 kDa fusion construct is indicated as TrF:TFI successful in producing soluble TFI-scFv, they did not display good (Lane 4). inhibition. This suggest that despite being soluble, the TFI-scFv is misfolded and incapable of recognizing its target. 3.6.2. Solubilization levels of TFI-scFv and TrF:TFI-scFv 3.6. Cold shock expression of TFI-scFv The expression of the pCOLD DNA II TFI-scFv vector by E coli BL21 (Fig. 6 A Lane 1 and 2) resulted in production of approximately 19% 3.6.1. Expression TFI-scFv and TrF-TFI-scFv soluble TFI-scFv with the majority (> 80%) of the protein produced as The scFv antibody was successfully expressed by E. coli BL21 (DE3) insoluble inclusion bodies. The expression of the pCOLD TrF-TFI-scFv using both pCOLD DNA II and pCOLD TrF vectors (Fig. 5A). Expression vector by E. coli BL21 resulted in the solubilization of approximately 67 155 156 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 Fig. 6. Solubilization levels of TFI-scFv and TrF:TFI-scFv. Table 4 Summary of large scale expression and protein yields. m Plasmids Biomass wet weight (in g.L−1) Fraction of total cellular protein (%) Total purified TFI-scFv protein (mg.L−1) Purity E. coli BL 21 (DE3) pCOLD DNA II 7.3 15% 2.13 > 80 pCOLD TrF 8.2 25% 4.83 > 80 E. coli SHuffle pCOLD DNA II 3.6 10% 1.32 > 95 pCOLD TrF 4.2 25% 3.26 > 95 38% of the total TrF-TFI-scFv produced (Fig. 6 A Lane 3 and 4). Ex- purified from the fusion construct was higher than 80% for both strains pression of the two vectors by E. coli SHuffle (Fig. 6 B) resulted in the utilized. Although the low purity is not ideal, it was deemed sufficiently solubilization of approximately 20% of the TFI-scFv produced by pure for initial activity screening. However, as these fractions did not pCOLD DNA II (Lane 1 and 2) and 37% of the fusion constructs pro- provide any notable inhibition effect on the actions of tissue factor, no duced by pCOLD TrF TFI-scFv (lane 3 and 4). additional attempts were made to improve the purity of the product. SDS-PAGE analyses of pCOLD DNA II TFI-scFv and pCOLD TrF TFI- The final biomass production and protein purification details are scFv expression profiles A) E. coli BL21 (DE3) and B) E. coli SHuffle. summarized in Table 4. Lane M: Protein Marker, Lane 1: pCOLD DNA II TFI-scFV (Soluble Fraction), Lane 2: pCOLD DNA II TFI-scFv (Insoluble Fraction), Lane 3: pCOLD TrF TFI-scFV (Soluble Fraction) and Lane 4: pCOLD TF TF-scFv 3.6.4. Prothrombin times (Insoluble Fraction). The 27 kDa TFI-scFv is indicated as TFI and the 79 kDa fusion construct is indicated as TrF:TFI. Of the four different purified fractions that were tested only the non- fusion construct produced by E. coli SHuffle pCOLD DNA II was capable 3.6.3. Cytoplasmic expression and purification of notably extending the prothrombin times (Fig. 7). The prothrombin time at the highest concentration (0.2 mgml−1) of TFI-scFv was ex- The use of the bioreactors resulted in a final yield of roughly 7 to 9 tended by 16.7 s which is an approximate 77% extension of the baseline (wet weight) grams of E. coli BL21 per liter and 3–5 g of E. coli SHuffle prothrombin time with a specific activity of 83.8 s/mg.mL −1. The re- per liter. The final protein production ranged from 2.0 mg L−1 to maining isolates of TFI-scFv fractions had very little inhibition effect, 5 mg L−1 for E. coli BL21 and 1.0mg L−1 to 3.5. mg.L−1 for E. coli with all fractions only extending the prothrombin times less than 3 s. SHuffle. Overall the E. coli BL21 (DE3) strains are capable of producing The inhibition efficacy of the various isolated fraction of TFI-scFv much higher biomasses that the E. coli SHuffle strains when grown was tested by means of prothrombin times. Only the non-fusion con- under cold shock conditions. The 27 kDa TFI-scFv was successfully struct produced by E. coli SHuffle using pCOLD DNA II expression vector isolated and purified form the cytoplasm of both E. coli strains was capable of extending the prothrombin times in a dose dependent (Supplementary Material: Fig. 6). The non-fusion construct was isolated manner. The remaining isolates were incapable of notably extending with a purity greater than 95%, while the purity of the antibody the prothrombin times. Fig. 7. Elongation of prothrombin times. 68 156 157 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 4. Discussion producing soluble TFI-scFv, the antibodies were folded in incorrect intermediates that were not capable of recognizing the tissue factor It was suspected that due to the heterologous nature of the native epitope. scFv gene that the low levels of expression initially attained were due to Thus, as expected, the redox potential had a significant effect on the codon bias. Codon usage has been identified as the single most im- production of functional antibody due to the presence of intra-chain portant factor in prokaryotic gene expression [40]. The rare codon disulfide bonds, although this did not correlate with solubility analysis of TFI-scFv revealed that 45 codons (approximately 20% in [11,12,48]. Disulfide bond formation plays an important role in the total) was distributed throughout the gene. Of these rare codons, 3 stabilization of folding intermediates as the protein advances to ma- consecutive repeats and 2 single rare codons were identified near the N- turity [18,49,50]. At this refolding rate one could extrapolate that ap- terminal of the scFv gene a phenomenon which has been show to fur- proximately 6mg of active TFI-scFv can be produced during one round ther exacerbated protein expression levels [40]. Expression of the of expression. However, it must be mentioned that several attempts to codon optimized gene in E. coli BL21 (DE3) using pET 22 resulted in the scale up these results using an on-column refolding approach with improved expression of TFI-scFv with the antibody fragment re- identical experimental conditions, failed to produce functional TFI-scFv presenting approximately 22% of the total cellular protein at (Unpublished data). It is also clear form the purification table (Table 2) 40mg L−1, a 40-fold improvement on previously reported levels [3]. that large fraction of the inclusion bodies is lost during the isolation Although it is clear that the codon optimization improved TFI-scFv process. Most of the isolated TFI-scFv was lost during IMAC purification expression, the majority (> 70%) of TFI-scFv was accumulated as in- as the columns do perform optimally under denaturing and reducing clusion bodies in the insoluble cytoplasmic fraction. This result is in- conditions. This only demonstrates the complexity and sensitivity of in dicative of protein aggregation due to misfolding [41], however, the vitro refolding procedures and emphasizes need for target specific op- formation of insoluble aggregate folding intermediate during high-level timized procedures in order to balance target protein quality against protein in the cytoplasm production is a common occurrence especially quantity. with scFv [42,43]. It is believed that the resulting high expression rates Due to the drawbacks of in vitro refolding, the feasibility of produ- in combination with the presence of disulfide bridges, reducing cyto- cing a functional TFI-scFv in the cytoplasm of E. coli was explored as an plasmic environment, insufficient chaperone assistance and the lack of alternative approach. Solubilization of the scFv was achieved though post-translation modification contribute to protein aggregation ob- commercial cold shock expression vectors (pCOLD DNA II and pCOLD served during bacterial expression [8–10]. Initial attempts to alleviate Trigger factor) in conjunction with E. coli BL21 (DE3) and specialized the production of the inclusion bodies through expression at lower disulfide bond promoting strain (E. coli SHuffle). When comparing the temperatures ranges, and the co-expression of chaperones in E. coli expression of TFI-scFv using the pCOLD DNA II cold shock vector by E. BL21 and Origami strains, only produced minor improvement in bio- coli BL21 (DE3)- and SHuffle strains, it is interesting to note that the mass and solubilization levels of TFI-scFv, nonetheless these techniques expression level by E. coli BL21 was only marginally higher (5%), while were unsuccessful in producing functional antibody (Unpublished the level of solubilization between the two strains remained similar at data). 20% of the target protein. This was unexpected, as the mutations to E. The generation of biologically active antibodies from insoluble ag- coli SHuffle to reduce aggregation did not yield a higher level of solu- gregates has become common practice when working with scFvs [44]. bilization than that of E. coli BL21. The similarities in protein solubi- The challenge is to convert the inactive and insoluble scFv aggregates lization when comparing the two cell lines might be attributed to the into soluble biologically active products [45]. A dilution method was comprehensive effect of cold shock itself. Protein expression at low used for the screening of the various refolding conditions due to; low temperature results in improved solubilization of the target protein cost, availability of equipment, and relative simplicity, thus placing this likely due to the reduction in translation rate of mRNA to the poly- method with in the capabilities of most research laboratories. The re- peptides [41,51,52]. By comparison, the expression levels of the TrF- folding matrix explores three stages of escalating concentrations of TFI-scFv fusion constructs using pCOLD TF were roughly 10% higher. denaturant (guanidinium chloride) in combination with varying L-Ar- This increase in protein expression is largely due to the high expression ginine concentrations while maintaining the pH at 8.0. Although there levels of the fusion partner itself. Trigger factor constitutes 15–25% of is a clear increase in solubility with an increase in denaturant, it is total protein yield in the respective controls. The TrF-TFI-scFv fusion important to note that solubilization does not necessarily translate into constructs were expressed at similar levels (25% of the total protein). functionality. The addition of arginine in refolding buffers in order to The solubilization levels were also higher than that of the pCOLD DNA suppress protein aggregation is a well-established practice in protein II constructs. A total of 37–38% of the expressed fusion construct was refolding. Is has been proposed that the guanidinium group of arginine produced as soluble TFI-scFv. However, of the four isolated fractions, as well as the side chains of tryptophan may be responsible for sup- only the TFI-scFv produced via the expression of pCOLD DNA II plasmid pression of protein aggregation by arginine [46]. The addition of ar- in E. coli SHuffle,was capable of tissue factor inhibition as demonstrated ginine does not, however, suppress aggregation in some disulfide-con- by the PT assay. taining proteins [47]. Due to this unpredictable effect of arginine on Functional antibody fragments were produced through a combina- disulfide bond formation, the effect of arginine was examined across a tion of cold shock, pCOLD DNA II expression vector in conjunction with large concentration range (0–800mM). In general, the refolding buffers E. coli SHuffle. The SHuffle strain has deletions of the glutathione re- containing the lower levels the guanidinium chloride (range 1) had the ductase and thioredoxin reductase (Δ gor and Δ trxB) genes, which higher levels of aggregation. Interestingly, the highest level of solubi- allows disulfide bonds to form in the cytoplasm. In addition, SHuffle lization was observed in sample 7, which contains no L-arginine, while expresses a version of the periplasmic disulfide bond isomerase DsbC in contrast samples 1 and 4 (also L-arginine free) produced the lowest which lacks a signal sequence, retaining it within the cytoplasm in levels of solubilization for their respective ranges. The specific activity order to facilitate disulfide bond formation. These specific mutations to analysis provides a good indication of which buffer produced the most E. coli SHuffle seemed to be the key to producing functional TFI-scFv in functional TFI-scFv capable of recognizing the tissue factor epitope, the bacterial cytoplasm. However, a major drawback to cold shock oppose to producing soluble but non-functional TFI-scFv. Refolding expression systems is the reduction in target protein yield as well as Buffer 6 (29.6 s/mg.mL−1) and refolding buffer 2 (14.5 s/ final biomass yields, caused by to the lower growth- and protein mgmL−1) had the highest specific activity. Interestingly, refolding synthesis rates [53]. This issue becomes further compounded by the buffer 5, 7, 8 and 9 which all produced a higher level of solubility than slower growth rate of E. coli SHuffle. refolding buffer 6, had a considerable lower specific activity. This in- When comparing the expression strategies, the E. coli BL21 (DE3) dicates that although these refolding buffers were more effective at strains yielded roughly double the biomass compared to E. coli SHuffle 69 157 158 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 when grown in the 5-L bioreactors. However, the TFI-scFv produced by Appendix A. Supplementary data cold shock had a specific activity of 83.8 s/mg.ml−1 opposed to 29.6 s/ mg.mL-1 of the refolded TFI-scFv. This difference suggests that al- Supplementary data related to this article can be found at http://dx. though functional scFv was produced by in vitro refolding, much of the doi.org/10.1016/j.pep.2018.06.005. refolded fraction was soluble but inactive. We could speculate that is may be due to due incorrect refolding or may be due to partial de- References gradation (Supplementary Material: Fig. 5). One of the major chal- lenges of working with a novel antibody is the fact that there is no [1] P. Chames, M. Van Regenmortel, E. Weiss, D. Baty, Therapeutic antibodies: suc- commercially available counterpart to serve as reference point. It is cesses, limitations and hopes for the future, Br. J. Pharmacol. 157 (2009) 220–233, http://dx.doi.org/10.1111/j.1476-5381.2009.00190.x. complications such as these that makes working with novel hetero- [2] X. Yang, W. Hu, F. Li, H. Xia, Z. Zhang, Gene cloning, bacterial expression, in vitro logous proteins a complex and challenging task. Overall, the cold shock refolding, and characterization of a single-chain Fv antibody against PreS1(21-47) expression systems provide a simpler, less time consuming, and more fragment of HBsAg, Protein Expr. Purif. 41 (2005) 341–348, http://dx.doi.org/10. cost-e ective approach that is simpler to scale up in when compared 1016/j.pep.2005.02.005.ff [3] S.M. Meiring, J. Vermeulen, P.N. Badenhorst, Development of an inhibitory anti- with the in vitro refolding strategy. body fragment to human tissue factor using phage display technology, Drug Dev. Res. 70 (2009) 199–205, http://dx.doi.org/10.1002/ddr.20295. [4] A.J. Chu, Tissue factor, blood coagulation, and beyond: an overview, Int. J. 5. Conclusion Inflamm. (2011) 1–30, http://dx.doi.org/10.4061/2011/367284. [5] N. Mackman, Role of tissue factor in hemostasis and thrombosis, Blood Cells Mol. The challenge in recombinant protein expression is to find the Dis. 36 (2006) 104–107, http://dx.doi.org/10.1016/j.bcmd.2005.12.008. [6] A.P. Owens, N. Mackman, Tissue factor and thrombosis: the clot starts here, balance between quantity and quality of target protein while con- Thromb. Haemostasis 104 (2010) 432–439, http://dx.doi.org/10.1160/TH09-11- sidering practicality and the cost involved. The heterologous produc- 0771. tion of scFv in the cytoplasm of E. coli is notoriously difficult with [7] G.E. Raskob, P. Angchaisuksiri, A.N. Blanco, H. Buller, A. Gallus, B.J. Hunt, E.M. Hylek, A. Kakkar, S.V. Konstantinides, M. McCumber, Y. Ozaki, A. Wendelboe, various factors contributing to reduced expression and improper J.I. Weitz, Thrombosis: a major contributor to global disease burden, Arterioscler. folding. Although various strategies exist to overcome these difficulties, Thromb. Vasc. Biol. 34 (2014) 2363–2371, http://dx.doi.org/10.1161/ATVBAHA. they are generally protein specific and usually do not translate to other 114.304488. cases of scFv expression. In this study, codon optimization appears to be [8] A. Mitraki, J. King, Protein folding intermediates and inclusion body formation, BioTechnol. 7 (1989) 690–697, http://dx.doi.org/10.1038/nbt0789-690. the most effective strategy to increase expression levels but contributed [9] J. Song, Why do proteins aggregate? “Intrinsically insoluble proteins” and “dark to folding complication. Subsequently, several general optimization mediators” revealed by studies on “insoluble proteins” solubilized in pure water, strategies were explored to improve the production of soluble TFI-scFv F1000Research 2 (2013) 94, http://dx.doi.org/10.12688/f1000research.2-94.v1.[10] A. Singh, V. Upadhyay, A.K. Upadhyay, S.M. Singh, A.K. Panda, Protein recovery but were unsuccessful. This study indicates that, at least in small scale from inclusion bodies of Escherichia coli using mild solubilization process, Microb. experiments, in vitro refolding is feasible however not optimal. At larger Cell Factories 14 (2015) 41, http://dx.doi.org/10.1186/s12934-015-0222-8. scales, specialized expression hosts and vectors are required to obtain [11] E.D.B. Clark, Protein refolding for industrial processes, Curr. Opin. Biotechnol. 12 (2001) 202–207, http://dx.doi.org/10.1016/S0958-1669(00)00200-7. functional antibody. When comparing these two strategies the use of [12] T. Arakawa, D. Ejima, Refolding technologies for antibody fragments, Antibodies 3 cold shock in conjunction with of E. coli SHuffle provided a simpler and (2014) 232–241, http://dx.doi.org/10.3390/antib3020232. more cost-effective method for the production of function TFI-scFv [13] M. Kamionka, Engineering of therapeutic proteins production in Escherichia coli, Curr. Pharmaceut. Biotechnol. 12 (2011) 268–274, http://dx.doi.org/10.2174/ appose to the more complex in vitro refolding strategies. This strategy 138920111794295693. can be followed when expression other problematic low yielding scFv. [14] L. Sánchez, M. Ayala, F. Freyre, I. Pedroso, H. Bell, V. Falcón, J.V. Gavilondo, High cytoplasmic expression in E. coli, purification, and in vitro refolding of a single chain Fv antibody fragment against the hepatitis B surface antigen, J. Biotechnol. 72 Formatting of funding sources (1999) 13–20, http://dx.doi.org/10.1016/S0168-1656(99)00036-X. [15] S. Tapryal, L. Krishnan, J.K. Batra, K.J. Kaur, D.M. Salunke, Cloning, expression and This work was funded by the National Research Foundation (NRF) efficient refolding of carbohydrate-peptide mimicry recognizing single chain anti-body 2D10, Protein Expr. Purif. 72 (2010) 162–168, http://dx.doi.org/10.1016/j. of South Africa. pep.2010.03.024. [16] Z.A. Ahmad, S.K. Yeap, A.M. Ali, W.Y. Ho, N.B.M. Alitheen, M. Hamid, ScFv anti- body: principles and clinical application, Clin. Dev. Immunol. (2012), http://dx.doi. Conflicts of interest org/10.1155/2012/980250. [17] K. Sushma, C.J. Bilgimol, M.A. Vijayalakshmi, P.K. Satheeshkumar, Recovery of None. active anti TNF-α ScFv through matrix-assisted refolding of bacterial inclusion bodies using CIM monolithic support, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 891–892 (2012) 90–93, http://dx.doi.org/10.1016/j.jchromb.2012.02.011. Author contributions [18] P.K. Devaraneni, J.J. Devereaux, F.I. Valiyaveetil, In vitro folding of K vAP, a vol- tage-gated K + channel, Biochemistry 50 (2011) 10442–10450, http://dx.doi.org/ 10.1021/bi2012965. Dr. Jan-G Vermeulen: Principle researcher and main author. [19] R. Gutiérrez, E.M. Martín del Valle, M. a. Galán, Immobilized metal-ion affinity Dr. Errol Cason: Experimental design and manuscript editing. chromatography: status and trends, Separ. Purif. Rev. 36 (2007) 71–111, http://dx. Prof SM Meiring: Study leader, - Experimental design and manu- doi.org/10.1080/15422110601166007. [20] A. Singh, V. Upadhyay, A.K. Panda, Solubilization and refolding of inclusion body script editing. proteins, Insoluble Proteins Methods Protoc 99 (2014) 283–291, http://dx.doi.org/ Prof. FJ Burt: Co Study Leader - Manuscript editing and infra- 10.1007/978-1-4939-2205-5_15. structure support. [21] J.A. Vasina, F. Baneyx, Recombinant protein expression at low temperatures under the transcriptional control of the major Escherichia coli cold shock promotor cspA, Prof. E van Heerden: Co study leader – Infrastructure and monetary Appl. Environ. Microbiol. 62 (1996) 1444–1447. support. [22] H.P. Sørensen, K.K. Mortensen, Soluble expression of recombinant proteins in the cytoplasm of Escherichia coli, Microb. Cell Factories 4 (2005) 1, http://dx.doi.org/ 10.1186/1475-2859-4-1. Acknowledgements [23] C.A. White-Ziegler, S. Um, N.M. Pérez, A.L. Berns, A.J. Malhowski, S. Young, Low temperature (23 °C) increases expression of biofilm-, cold-shock- and RpoS-de- I would like to thank the Department of Haematology and Cell pendent genes in Escherichia coli K-12, Microbiology 154 (2008) 148–166, http://dx.doi.org/10.1099/mic.0.2007/012021-0. Biology, Division of Virology of the Faculty of Health Sciences as well as [24] C.O. Gualerzi, A.M. Giuliodori, C.L. Pon, Transcriptional and post-transcriptional the Department of Microbial, Biochemical and Food Biotechnology, control of cold-shock genes, J. Mol. Biol. 331 (2003) 527–539, http://dx.doi.org/ Faculty of Agricultural Sciences at the University of the Free State as 10.1016/S0022-2836(03)00732-0. [25] S. Phadtare, Recent developments in bacterial cold-shock response, Curr. Issues well as those involved in the research for their assistance throughout Mol. Biol. 6 (2004) 125–136. the project. [26] A. Mogk, M.P. Mayer, E. Deuerling, Mechanisms of protein folding: molecular 70 158 159 J.-G. Vermeulen et al. Protein Expression and Purification 151 (2018) 62–71 chaperones and their application in biotechnology, Chembiochem 3 (2002) http://dx.doi.org/10.1021/bi00789a030. 807–814, http://dx.doi.org/10.1002/1439-7633(20020902)3. [40] C. Gustafsson, S. Govindarajan, J. Minshull, Codon bias and heterologous protein [27] T. Rathnayaka, M. Tawa, S. Sohya, M. Yohda, Y. Kuroda, Biophysical character- expression, Trends Biotechnol. 22 (2004) 346–353, http://dx.doi.org/10.1016/j. ization of highly active recombinant Gaussia luciferase expressed in Escherichia coli, tibtech.2004.04.006. Biochim. Biophys. Acta 1804 (2010) 1902–1907, http://dx.doi.org/10.1016/j. [41] G.P. Subedi, T. Satoh, S. Hanashima, A. Ikeda, H. Nakada, R. Sato, M. Mizuno, bbapap.2010.04.014. N. Yuasa, Y. Fujita-Yamaguchi, Y. Yamaguchi, Overproduction of anti-Tn antibody [28] G. Hannig, S.C. Makrides, Strategies for optimizing heterologous protein expression MLS128 single-chain Fv fragment in Escherichia coli cytoplasm using a novel pCold- in Escherichia coli, Trends Biotechnol. 16 (1998) 54–60, http://dx.doi.org/10.1016/ PDI vector, Protein Expr. Purif. 82 (2012) 197–204, http://dx.doi.org/10.1016/j. S0167-7799(97)01155-4. pep.2011.12.010. [29] K. Terpe, Overview of tag protein fusions: from molecular and biochemical fun- [42] G. Hannig, S.C. Makrides, Strategies for optimizing heterologous protein expression damentals to commercial systems, Appl. Microbiol. Biotechnol. 60 (2003) 523–533, in Escherichia coli, Trends Biotechnol. 16 (1998) 54–60, http://dx.doi.org/10.1016/ http://dx.doi.org/10.1007/s00253-002-1158-6. S0167-7799(97)01155-4. [30] R. Glockshuber, T. Schmidt, A. Plueckthun, The disulfide bonds in antibody variable [43] M. Kudou, D. Ejima, H. Sato, R. Yumioka, T. Arakawa, K. Tsumoto, Refolding single- domains: effects on stability, folding in vitro, and functional expression in chain antibody (scFv) using lauroyl-L-glutamate as a solubilization detergent and Escherichia coli, Biochemistry 31 (1992) 1270–1279, http://dx.doi.org/10.1021/ arginine as a refolding additive, Protein Expr. Purif. 77 (2011) 68–74, http://dx. bi00120a002. doi.org/10.1016/j.pep.2010.12.007. [31] A. de Marco, Strategies for successful recombinant expression of disulfide bond- [44] K. Ikeda, Y. Kumada, T. Katsuda, H. Yamaji, S. Katoh, Refolding of single-chain Fv dependent proteins in Escherichia coli, Microb. Cell Factories 8 (2009) 26, http://dx. by use of an antigen-coupled column, Biochem. Eng. J. 44 (2009) 289–291, http:// doi.org/10.1186/1475-2859-8-26. dx.doi.org/10.1016/j.bej.2008.12.014. [32] I. Kumagai, R. Asano, T. Nakanishi, K. Hashikami, S. Tanaka, A. Badran, H. Sanada, [45] M. Li, Z.G. Su, J.C. Janson, In vitro protein refolding by chromatographic proce- M. Umetsu, Integration of PEGylation and refolding for renaturation of recombinant dures, Protein Expr. Purif. 33 (2004) 1–10, http://dx.doi.org/10.1016/j.pep.2003. proteins from insoluble aggregates produced in bacteria: application to a single- 08.023. chain Fv fragment, J. Biosci. Bioeng. 109 (2010) 447–452, http://dx.doi.org/10. [46] K. Tsumoto, M. Umetsu, I. Kumagai, D. Ejima, J.S. Philo, T. Arakawa, Role of ar- 1016/j.jbiosc.2009.10.016. ginine in protein refolding, solubilization, and purification, Biotechnol. Prog. 20 [33] M. Berkmen, Production of disulfide-bonded proteins in Escherichia coli, Protein (2004) 1301–1308, http://dx.doi.org/10.1021/bp0498793. Expr. Purif. 82 (2012) 240–251, http://dx.doi.org/10.1016/j.pep.2011.10.009. [47] J. Chen, Y. Liu, Y. Wang, H. Ding, Z. Su, Different effects of L-Arginine on protein [34] M. Fuhrmann, A. Hausherr, L. Ferbitz, T. Schödl, P. Hegemann, Monitoring dy- refolding: suppressing aggregates of hydrophobic interaction, not covalent binding, namic expression of nuclear genes in Chlamydomonas reinhardtii by using a synthetic Biotechnol. Prog. 24 (2008) 1365–1372, http://dx.doi.org/10.1021/bp.93. luciferase reporter gene, Plant Mol. Biol. 55 (2004) 869–881, http://dx.doi.org/10. [48] W. Swietnicki, Folding aggregated proteins into functionally active forms, Curr. 1007/s11103-005-2150-1. Opin. Biotechnol. 17 (2006) 367–372, http://dx.doi.org/10.1016/j.copbio.2006. [35] P. Puigbò, I.G. Bravo, S. Garcia-Vallve, CAIcal: a combined set of tools to assess 05.011. codon usage adaptation, Biol. Direct 3 (2008) 38, http://dx.doi.org/10.1186/1745- [49] S.C. Makrides, Strategies for achieving high-level expression of genes in Escherichia 6150-3-38. coli, Microbiol. Rev. 60 (1996) 512–538. [36] U. Omasits, C.H. Ahrens, S. Müller, B. Wollscheid, Protter: Interactive protein fea- [50] K. Proba, A. Wörn, A. Honegger, A. Plückthun, Antibody scFv fragments without ture visualization and integration with experimental proteomic data, Bioinformatics disulfide bonds, made by molecular evolution, J. Mol. Biol. 275 (1998) 245–253, 30 (2014) 884–886, http://dx.doi.org/10.1093/bioinformatics/btt607. http://dx.doi.org/10.1006/jmbi.1997.1457. [37] E. Krieger, G. Vriend, YASARA, View–molecular graphics for all devices–from [51] S.C. Makrides, Strategies for achieving high-level expression of genes in Escherichia smartphones to workstations, Bioinformatics 30 (2014) 2981–2982, http://dx.doi. coli, Microbiol. Rev. 60 (1996) 512–538. org/10.1093/bioinformatics/btu426. [52] S. Costa, A. Almeida, A. Castro, L. Domingues, Fusion tags for protein solubility, [38] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of purification, and immunogenicity in Escherichia coli: the novel Fh8 system, Front. bacteriophage T4, Nature 227 (1970) 680–685, http://dx.doi.org/10.1038/ Microbiol. 5 (2014) 1–20, http://dx.doi.org/10.3389/fmicb.2014.00063. 227680a0. [53] G.L. Rosano, E.A. Ceccarelli, Recombinant protein expression in Escherichia coli: [39] G. Fairbanks, T.L. Steck, D.F. Wallach, Electrophoretic analysis of the major poly- advances and challenges, Front. Microbiol. 5 (2014) 1–17, http://dx.doi.org/10. peptides of the human erythrocyte membrane, Biochemistry 10 (1971) 2606–2617, 3389/fmicb.2014.00172. 71 159 S-15 160 Hindawi BioMed Research International Volume 2018, Article ID 6232091, 7 pages https://doi.org/10.1155/2018/6232091 Research Article The Use of Phage Display and Yeast Based Expression System for the Development of a Von Willebrand Factor Propeptide Assay: Development of a Von Willebrand Factor Propeptide Assay S. M. Meiring ,1,2 B. D. P. Setlai,2 C. Theron,3 and R. Bragg3 1National Health Laboratory Services, Universitas Hospital, South Africa 2Department of Haematology and Cell Biology, University of the Free State, Bloemfontein 9301, South Africa 3Department of Microbial, Biochemical and Food Biotechnology, University of the Free State, Bloemfontein 9301, South Africa Correspondence should be addressed to S. M. Meiring; meiringsm@ufs.ac.za Received 24 October 2017; Revised 12 March 2018; Accepted 25 March 2018; Published 24 May 2018 Academic Editor: Wolfgang Miesbach Copyright © 2018 S. M. Meiring et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. The diagnosis of von Willebrand disease is complex due to the heterogeneity of the disease. About eighty percent of vonWillebrand disease patients are diagnosed with a quantitative defect of vonWillebrand factor (VWF) where fifty percent is due to an increased clearance of von Willebrand factor. These patients do not respond well to the treatment of choice, Desmopressin (DDAVP) due to decreased efficacy. The ratio between the VWF propeptide and the mature VWF antigen is used to diagnose these patients. Commercial VWF propeptide assays are too expensive for use in developing countries. In this study, we developed a cost-effective ELISA assay.Methods. We first displayed VWF propeptide on yeast. Antibody fragments were selected against the displayedVWFpropeptide by using phage display technology.The antibodies were used to develop a cost-effectiveVWFpropeptide assay and compared to a commercial VWF propeptide assay. Results. Two of these antibody fragments bound specific to the VWF propeptide and not to the yeast used for the expression of the propeptides. These purified antibody fragments were able to detect VWF propeptide in normal plasma. Conclusion. Our assay performed well when compared to a commercial kit. It also showed a higher binding affinity for VWF propeptide in plasma at especially lower plasma concentrations. 1. Background of VWF need to receive different treatment strategies and the correct diagnosis is crucial, since DDAVP, the treatment Von Willebrand disease (VWD) is the most common bleed- of choice for type 1 VWD patients, would not be effective ing disorder in the world with a prevalence of one percent at all in these patients. An increased ratio between von the in the general population [1]. It is classified into three types. Willebrand factor propeptide (VWFpp) and thematureVWF Types 1 and 3 are quantitative defects of von Willebrand antigen can be used to diagnose these patients. This can factor (VWF) and Type 2 is a qualitative defect VWF [1]. The be interpreted as a shortened half-life of plasma VWF and diagnosis of VWD is complex due to the heterogeneity of the therefore increased clearance of VWF [1]. disease. Variousmutations of theVWFgene result in a variety Furthermore, the level of VWFpp in the circulation can of phenotypes that makes the diagnoses of these patients very also be used as a marker of VWF synthesis. Studies showed challenging [1]. It is therefore not surprising that patients with that VWFpp levels are more reliable than mature VWF VWD are largely under-diagnosed or misdiagnosed. About levels in the assessment of endothelial cell activation and 80% of VWD patients are diagnosed with type 1 VWD and endothelial damage, since the plasma VWFpp concentration about fifty percent of these patients present with an increased is not influenced by blood groups or factors such as adhesive clearance rate of VWF [1]. Patients with increased clearance properties and catabolism [2]. Increased plasma VWFpp 160 161 2 BioMed Research International concentrations have been described in vasculopathies such 2.2. Selection of Antibody Fragments. Two phage display as hypertension and diabetes, systemic sclerosis, myocardial libraries (Tomlinson I & J) from theMedical Research Coun- infarction, and cardiovascular mortality [2, 3]. cil Centre for Protein Engineering, Cambridge in London, The current commercially available assays that are used to UK, were amplified. The I library consists of diversified side test the levels ofVWFpp in plasma are expensive and the price chains on 18 residues while the J library has NNK side that medical aids pay does not cover the costs of the tests. chains on these residues (N=GACT and K=GT). Antibody With this research, we developed a cost-effective VWFpp selection was done as previously described [8]. In short, diagnostic laboratory assay using antibodies produced by Nunc-Immuno tubes (Thermo Scientific, USA)were coated phage display technology. Phage display is an in vitromethod withVWFppdisplaying yeast cells (110 g/l in PBS) and rotated that allows rapid production of antibodies during cycles overnight at 4∘C. After washing (3 times) and blocking with of biopanning and propagation without the use of animal 2% skimmed milk powder in PBS at room temperature models [4–6]. for 2 hours, 1012 phages of each library (I and J) were added to the tubes and rotated at room temperature 2 2. Methods hours. After 10 times washing, the unbound phages were discarded and the bound phages were eluted with 500 𝜇l 2.1. Display of VWF Propeptide on Yeast. Since a commercial trypsin-PBS (10mg/ml trypsin, 50mM Tris-HCl, 1mM CaCl2 preparation of the VWFpp is not available, the VWFpp in PBS, and pH7.4) for 10min at room temperature. TG1 E- (P04275, amino acids 23-763 of VWF) were displayed on coli cells (1.75ml) were infected with the eluted phages for yeast according to the method described by Lin-Cereghino 30 minutes at 37∘C in a water bath until OD reach 0.4-600 et al. (2005) [7]. In short, the VWF propeptide-encoding 0.6. The infected TG1 E-coli cells were then centrifuged at sequence was transferred from the pMK-RQ plasmid to 11600g in a microcentrifuge for 5min; the pellet suspended pINA1317-CWP110 using the SfiI andHindIII restriction sites. in 50𝜇l 2TY medium (16g bacto-tryptone, 10g yeast extract, Recombinant expression vectors were used to transform the and 5g NaCl in 1 liter distilled water) and grown on trypsin- Y. lipolytica strain P01h. All plasmids and related reagents yeast plates (15g agar, 10g bacto-tryptone, 8g NaCl, and 5g were purchased from TaKaRa, Japan. A negative control yeast extract in 1 liter distilled water) containing 100 𝜇g/ml strain was constructed by transformation of Y. lipolytica ampicillin and 1% glucose overnight at 37∘C. Cells were strain P01h with the original pINA1317-CWP110 vector. loosened from the plates with 2ml 2TY medium containing Transformants were randomly selected from yeast nitrogen 15% glycerol and a small volume (50𝜇l) was grown in 50ml base selective plates [yeast nitrogen base without ammonium 2TY containing 100 𝜇g/ml ampicillin and 1% glucose at sulphate and amino acids: 0.17%, ammonium chloride: 0.4%, 37∘C until OD of 0.4-0.6 was reached. The culture was600 glucose: 1%, casamino acids: 0.2%, agar: 2%, and pH: 6.5]. then infected with 5 × 1010 helper phages, incubated at Chromosomal integration in transformants was confirmed 37∘C in a water bath for 30 min, centrifuged at 3000g for by PCR using genomic DNA as template. Y. lipolytica trans- 10min and resuspended in 50ml 2TY medium containing formants were cultivated in yeast extract peptone dextrose 100 𝜇g/ml ampicillin, 50𝜇g/ml kanamycin, and 0.1% glucose, broth [yeast extract: 1%, peptone: 2%, and glucose: 2%] on and amplified overnight shaking at 30∘C. After centrifugation a rotary shaker at 30∘C. After 48h cultivation, cells were at 3300g for 15min, poly-ethylene glycol was added to 40ml harvested by centrifugation at 5000 g for 5 min. Cell pellets supernatant and mixed well and the phages were precipitated were washed with phosphate buffered saline (PBS) and on ice for 2hrs. The phage mixture was centrifuged again at resuspended to a final concentration of 50 g/L wet cell weight 3300g for 30min at room temperature, resuspended in PBS, prior to assay. A control yeast with no VWFpp displayed on and centrifuged shortly to remove access bacterial cells. The it was also prepared. precipitated phages were used for the next round of selection. To detect the VWFpp on the yeast cells, VWFpp dis- Three selection rounds were performed. playing yeast cells and control yeast cells were coated onto a Ninety-six single colonies from the third selection round 96-well plate (Nunc, Thermo Scientific, USA) at decreasing of the I and J libraries, respectively, were amplified in 96- concentrations ranging from 27.5 g/L to 0 g/L (only PBS) well microculture plates. The individual colonies were grown wet cell weight and fixed onto the plate by adding ice cold overnight shaking (250 rpm) at 37∘C in 100𝜇l 2TY media methanol (100%) to each well and dried again overnight containing 100 𝜇g/ml ampicillin and 1% glucose. Two 𝜇l of at 37∘C. After blocking with PBS/0.1% Tween-20 with 1% the above cultures was inoculated into each well of the plates Bovine SerumAlbumin for 2hrs at 37∘C, the plate was washed containing 200 𝜇l 2TY with 100 𝜇g/ml ampicillin and 1% three times with PBS/0.1% Tween-20. The bound VWFpp glucose and grown shaking at 37∘C until an OD of 0.4-600 was detected with a commercial anti-VWFpp horseradish 0.6 was reached. After infection with 1 × 109 helper phages peroxidase-conjugated antibody (CLB-Pro 14.3, 1:1000 dilu- for 30 min, the cultures were pelleted and resuspended in tion in PBS/Tween/BSA, Biocom Africa, SA) at 37∘C for 2TY medium containing 100 𝜇g/ml ampicillin and 50 𝜇g/ml 2hrs. The reaction was developed with o-phenylenediamine kanamycin and grown overnight at 30∘C. The cultures were dihydrochloride (OPD) substrate (10ml 0.1M Citric acid, then centrifuged at 1,800g for 10 min and the supernatant 10 ml 0.1M Na HPO , 200 𝜇l OPD (1mg/ml), and 8𝜇l from each of the single colonies was tested for binding to the 2 4 H O ) and stopped with 30𝜇l of 4M sulphuric acid after VWFpp. 2 2 15min. Absorbance was measured at 450 – 650nm using the Two ELISA plates (one of each library) were coated with SynergyHT ELISA reader (Biotek, USA). the VWFpp displaying yeast and another two plates with 161 162 BioMed Research International 3 the control yeast. Fifty 𝜇l of the supernatant of each of 0.6 the different monoclonal phage colonies was added to both plates and incubated for 2 hours at 37∘C. The plates were then washed and 100𝜇l of a 1:5000 dilution of HRP-anti-M13 antibody (Amersham, South Africa) in PBS/2% skimmed 0.4 milk was added and incubated for 1 hour at 37∘C. After washing again, the reaction was developed as mentioned in previous section. The six colonies with the highest binding 0.2 affinity (OD > 0.5) were upscaled and tested for 490−630𝑛𝑚 concentration depended binding to VWFpp. The following phage concentrationswere used: 5×1010 phages/ml, 2.5×1010, 0.0 1.25 × 1010, 6.25 × 109, 3.125 × 109 and a blank, etc. The two colonies with the strongest and most specific binding to 0 10 20 30 the VWFpp on the displaying yeast were used to produced g/l yeast soluble antibody fragments. VWFpp yeast Control yeast 2.3. Production of Soluble Antibody Fragments. Phages from Figure 1: Existence of the displayed VWFpp.The VWFpp displayed each of the two strongest binders were used to infect yeast is indicated by dots and the control yeast indicated by squares. exponentially growing HB2151 bacteria (OD of 0.4-0.6). 600 The binding is expressed as average ± 1 standard deviation (SD) In order to obtain soluble antibody fragments, isopropyl- where 𝑛 = 2. 𝛽-D-thiogalactoside (IPTG) (Thermo Scientific, USA) was added to a final concentration of 1mM to the cultures and grown overnight at 30∘C. The overnight cultures with IPTG were centrifuged at 1,800 g for 10min and the scFv in (5) Coating antibody: A9, Detection antibody: G7. the supernatant was concentrated with the minimate TFF filtration system (Thermo Scientific, USA). Protein A-HRP was used as previously described for detec- tion in the second scenario. The assay conditions were again 2.4. Protein Purification. A Protein A IgG purification kit similar to that of the polyclonal ELISA. This assay was (Thermo Scientific, USA) was used according to the instruc- repeated twice. As validation of our assay, we compared the tions of the manufacturers. Elution fractions with the highest standard curves of using both commercial antibodies and of absorbance values were pooled together and dialysed in PBS using our two scFv. The sensitivity and interrun accuracy of overnight using a Slide-A-Lyzer Dialysis cassette (Thermo the assay were also determined from the standard curve. Scientific, USA) with 10 000kD cut-off and stored at 4∘C for further use. The two purified antibody fragments were then 3. Results characterised in aVWFpp assay and compared to commercial anti-VWFpp antibodies. The displayed propeptide onto the yeast is confirmed in Figure 1. The displayed propeptide binds to the commercial 2.5. Assay Development. A sandwich ELISA was performed antibody (CBL-Pro 35) that is directed against the VWFpp. to show that the antibody fragments are able to identify the The control yeast however did not bind to the anti-VWFpp VWFpp in normal human plasma (WHO 6th International antibody. standard for FVIII and VWF in plasma). One antibody From the third selection round, three colonies from the fragment (A9) was used to coat the plate and the other one I library and 6 from the J library were identified as strong (G7) was conjugated with HRP (EZ-Link activated peroxi- binders to the VWF displayed yeast and not the control yeast, dase antibody labelling kit, Thermo Fisher Scientific, USA) but only two of them bound specific to the VWFpp yeast and used as the detection antibody. We also used our scFv and not the control yeast. The results of the concentration as coating antibodies with commercially available antibody dependent ELISA are shown in Figure 2.The twomonoclonal as the detection antibody and vice versa. Furthermore, the phages, JA9 and JG7, showed the most specific affinity for commercially available antibodies were also used alone for VWFpp. comparison. The outline of these assays was as follows: Soluble antibody fragments grown from these two colonies bound concentration dependently to the VWFpp (1) Coating antibody: CLB-Pro 35, Detection antibody: displayed on the yeast. Neither of them bound to the yeast CLB-Pro 14.3 itself. (2) Coating antibody: CLB-Pro 35, Detection antibody: These two colonies were then purified on a protein A A9 columns. Both proteins appeared in the first 2 fractions during the elution process. The protein concentrations of (3) Coating antibody: A9, Detection antibody: CLB-Pro these antibody fragmentswere calculated as 185𝜇g/ml for JA9 14.3 and 191 𝜇g/ml for JG7 after dialysis with PBS. (4) Coating antibody: G7, Detection antibody: CLB-Pro Our VWFpp assay where standard plasma (6th WHO 14.3 VWF: FVIII standard, NIBSC, UK) was used and compared 162 OD (490nm) 163 4 BioMed Research International JA9 0.6 cost of our assay amounts to approximately 60% of the costof a commercial assay kit. This will allow our medical aids to support funding for patients with increased VWF clearance who needs the outcome assay for diagnostic purposes. 0.4 4. Discussion 0.2 TheVWFpp assay has recently been included in the diagnos- tic setup of von Willebrand disease (VWD) [10]. The ratio of the propeptide to antigen (VWFpp/VWF:Ag) is used to determine the clearance rate of VWF from the circulation 0.0 0 10 10 10 [11]. The VWFpp assay also measures endothelial cell activa-2×10 4×10 6×10 tion during endothelial damage [12, 13]. phages/ml Since the current commercially available assays that JA9 VWFpp yeast measure the levels of VWFpp in plasma are so expensive, JA9 Control yeast we developed a cost-effective VWFpp diagnostic laboratory 0.6 JG7 assay using antibodies produced by phage display technology. Phage display technology was used because it is a molecular diversity technology that allows the presentation of large amounts of various proteins on the surfaces of filamentous 0.4 phages. It also allows rapid production of antibodies during cycles of biopanning and propagation without the use of animal models. Phage display libraries thus permit the selec- 0.2 tion of peptides and proteins, including antibodies, with high affinity and specificity for almost any target [14]. The use of phage display also allows the user to manipulate the protocol 0.0 according to the requirements of the target antigen. As long 0 2×1010 4×1010 6×1010 as the target is immobilised to a support and the exposed phages/ml solutions containing phage are immobilised to the target, the changes of the experiment succeeding are almost a hundred JG7 VWFpp yeast JG7 Control yeast percent [6]. We used two single chain variable antibodyfragment libraries (Tomlinson I and J libraries) from which Figure 2: Concentration dependent binding of single colonies that we selected the VWFpp binding antibody fragments [15]. showed specific binding affinity for VWFpp displayed on yeast. The advantage of using small antibody fragments is that Phage binding to the VWFpp yeast is indicated by dots and to the they easily penetrate the cellular or tissuemembraneswithout control yeast by squares. The binding is expressed as average ± 1 compromising their affinity and specificity. Antibodies frag- standard deviation (SD) where 𝑛 = 2. ments are easier and faster to produce. They can be easily purified with commonly used purification systems such as protein A [16, 17]. to a commercial kit with antibodies (CLB-Pro 35 & CLB- The VWFpp is not commercially available and therefore Pro 14.3, Biocom Africa, SA). Figure 4 showed that our we needed to express the recombinant protein. E.coli protein assay using both our antibody fragments compared well to expression systems are widely used but lack proper protein that of the assay using commercial antibody fragments. Our folding and posttranslational modifications [17]. Mammalian assay detects VWFpp concentrations form 1.5625%, while the cells produce very low yields and are also very costly [18]. We commercial assay only detects VWFpp from 6.25%. Further- made use of a yeast display system to display theVWFpp anti- more, a total %CV of 16% between our duplicate runs were gen (amino acids 23-763 of VWF). Yeast display is more likely found. Thus this standard curve serves as proof that our two to produce soluble functional proteins with the appropriate antibody fragments are suitable to be used in a VWFpp assay. posttranslational modifications [19]. Furthermore, proteins The differences in the protein sequences of our JG7 and displayed on the surface of the yeast can be purified easily as JA9 antibody fragments are shown in Box 1. The light and an active yeast particle by centrifugation. A yeast cell is also heavy chains of the two antibody fragments do differ and thus capable of displaying up to 50–2000 copies of the antigen on are suitable to be used in a sandwich ELISA assay, since it is its surface [20]. unlikely that they might bind to the same binding site on the The combination of phage display and yeast expression VWFpp. systems in protein expression provides the opportunity to The cost of producing 1mg of a monoclonal antibody produce antibodies without the need of soluble target protein amounts to approximately 45US Dollar (4500USD for [21]. The reason why we did not use yeast display for the 100mg). However, it will be much more cost-effective to selection process is because the library sizes are smaller than produce single chain variable fragments in large scale using those of other display systems. The existence of the displayed metal-affinity columns (about 3000USD for 100mg) [9].The VWFpp on the yeast was confirmed (Figure 1). 163 OD (490nm) OD (490nm) 164 BioMed Research International 5 JG7 PIT2 vector: DSGNSYDHDYAKLACKFYFKETVIMKYLLPTAAAGLLLLAAQPAMAEV Lightchain:QLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSIVETGPETSYADSVKGRFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKSPQL Linker sequence:FDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQ Heavychain:SPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYRASSLQSGVPSRFSGSGSGTDFTLTIS SLQPEDFATYYCQQGHPTPGTFGQGTKVEIKR PIT2 vector:AAAHHHHHHGAAEQKLISEEDLNGAAEITS JA9 PIT2 vector:DSGNSYDHDYAKLACKFYFKETVIMKYLLPTAAAGLLLLAAQPAMAEV Lightchain:QLLESGGGLVQPGGSLRLSCAASGFTFSSYAMSWVRQAPGKGLEWVSSIYEEGVLTMYADSVKGRFTIS RDNSKNTLYLQMNSLRAEDTAVYYCAKRSVIR Linker: FDYWGQGTLVTVSSGGGGSGGGGSGGGGSTDIQMTQ Heavychain:SPSSLSASVGDRVTITCRASQSISSYLNWYQQKPGKAPKLLIYAASCKVGSPSRFSGSGSGTDFTLTISS LQPEDFATYYCQQENCEYTFGQGTKVEIKR PIT2 sequence: AAAHHHHHHGAAEQKLISEEDLNGAXEITS Box 1: Protein sequence of antibody fragments JA9 and JG7. The differences between the two antibody fragments are underlined. We used the Tomlinson I and J Phage display single chain Soluble ab fragments JA9 variable fragment libraries since they have a diversity of over 0.25 100 million different antibody fragments (Tomlinson I & J 0.20 protocol).Three selection rounds aremostly used in literature [22] so we also performed 3 rounds of selection. Ninety-six 0.15 colonies of each library’s selectionwere chosen from the third 0.10 selection round. Only two colonies (JA9 and JG7) showed 0.05 specific binding to the VWFpp and not to the yeast (see Figure 2). 0.00 0 50 100 150 Some studies showed that phage colonies might lose their l ab fragment specificity after conversion into soluble antibody fragments [16, 22]. This however did not happen in our study. The JA9 VWFpp yeast antibody fragments of both colonies (JA9 and JG7) still JA9 Control yeast bound concentration dependently to the displayed VWFpp Soluble ab fragments JG7 and not to the control yeast (see Figure 3). 0.25 The availability of purification tags that can be used to purify antibodies generated with phage display makes the 0.20 purification process easy and reliable [21]. The soluble scFv 0.15 antibodies JA9 and JG7 were purified on protein A column 0.10 and successfully eluted with 3-5ml elution buffer and aliquots pooled into single tube. 0.05 The amount of purified antibodies was just enough to 0.00 perform a VWFpp assay where standard human plasma was 0 50 100 150 used. The WHO 6th International standard for FVIII and l ab fragment VWF in plasma was used, since it is the only standard with JG7 VWFpp yeast a given true value for the VWFpp. Although there is not a JG7 Control yeast gold standard assay available to compare our assay with, we Figure 3: Concentration dependent binding curves of soluble single compared the standard curve of the VWFpp assay using our chain variable antibody fragments that bind specific to the VWFpp 2 antibody fragments to the standard curve of the VWFpp on the yeast. Binding to the VWFpp yeast is indicated by dots and assay using two commercial antibodies to the VWFpp. Our to the control yeast by squares. The binding is expressed as average assay shows higher binding affinity at especially lower plasma ± 1 standard deviation (SD) where 𝑛 = 2. levels and might thus be more specific than the commercial one. Due to funding constraints and a lack of upscaling facilities, the amount of purified and concentrated antibody fragments was not enough to test patient samples and to do a The sensitivity of our assay was determined at 1.5625% full validation. However, the plasma that we used to set up VWF:pp in plasma, which is better than those of the com- a standard curve is of human origin and was pooled from mercial assay of 6.25%. The robustness of the assay was not between 20 and 40 human plasmas with an assigned value determined in this study; however, it is known that the full- specific for the VWFpp. length VWF levels are influenced by cold storage at 4∘C, but 164 OD (490 nm) OD (490 nm) 165 6 BioMed Research International 2.0 [2] D. S. Frankel, J. B. Meigs, J. M. Massaro et al., “Von Willebrand factor, type 2 diabetes mellitus, and risk of cardiovascular 1.5 disease:The FraminghamOffspring Study,”Circulation, vol. 118, no. 24, pp. 2533–2539, 2008. 1.0 [3] A. Scheja, A. Åkesson, P. Geborek et al., “VonWillebrand factor propeptide as a marker of disease activity in systemic sclerosis 0.5 (scleroderma),” Arthritis Research & Therapy, vol. 3, no. 3, pp. 178–182, 2001. 0.0 [4] T. Schirrmann, T. Meyer, M. Schütte, A. Frenzel, and M. Hust, 0 20 40 60 80 100 “Phage display for the generation of antibodies for proteome % VWF:pp research, diagnostics and therapy,” Molecules, vol. 16, no. 1, pp. Coat A9 Detect CLB 412–426, 2011. Coat G7 Detect CLB [5] M. Lavin, S. Aguila, S. Schneppenheim et al., “Novel insights Our assay Coat A9 Detect G7 into the clinical phenotype andpathophysiology underlying low Commercial assay VWF levels,” Blood, vol. 130, no. 21, pp. 2344–2353, 2017. Coat CLB Detect A9 [6] W. G. T. Willats, “Phage display: Practicalities and prospects,” Figure 4: VWFpp ELISA assays with our 2 purified antibody Plant Molecular Biology, vol. 50, no. 6, pp. 837–854, 2002. fragments, commercially available VWFpp antibodies and combi- [7] J. Lin-Cereghino, W. W. Wong, S. Xiong et al., “Condensed nations of both. Our assay (diamonds), commercial assay (dots), protocol for competent cell preparation and transformation of commercial coating antibody (ab) and JA9 detection ab (squares), the methylotrophic yeast Pichia pastoris,” BioTechniques, vol. JG7 coating ab and commercial detection ab (triangle upside-down) 38, no. 1, pp. 44–48, 2005. and JA9 coating ab, and commercial detection ab (triangle upright). The data are expressed as average ± 1 standard deviation (SD) where [8] S. M. Meiring, J. Vermeulen, and P. N. Badenhorst, “Develop- 𝑛 = 2. ment of an inhibitory antibody fragment to human tissue factor using phage display technology,” Drug Development Research, vol. 70, no. 3, pp. 199–205, 2009. not by freezing at -80∘C [23]. In this study, all plasma samples [9] G. L. Rosano and E. A. Cecarelli, “Recombinant protein expres- were stored at -80∘C. sion in Esthericia coli: dvances and challenges,” Frontiers in Specificity was however not determined in the final assay. Microbiology, vol. 5, no. 172, 2014. We however showed that the soluble antibody fragments bind [10] M. Meiring, M. Coetzee, M. Kelderman, and P. Badenhorst, to the propeptide yeast and not to the control yeast; that “Laboratory diagnosis andmanagement of vonWillebrandDis- indicated the specificity of the antibody fragments (Figure 3). ease in South Africa,” Seminars in Thrombosis and Hemostasis, Commercial assays, however, show a high total %CV of vol. 37, no. 5, pp. 576–580, 2011. 25% for the between-run accuracy [24]. We found a total [11] A. Casonato, V. Daidone, and R. Padrini, “Assessment of von %CV of 16% between our duplicate runs. Willebrand factor propeptide improves the diagnosis of von Since the amino acid sequence of the two antibody Willebrand disease,” Seminars in Thrombosis and Hemostasis, fragments differs significantly (Box 1), it might be used in a vol. 37, no. 5, pp. 456–463, 2011. sandwich ELISA.The sequences were also compared to other [12] S. M. Kawut, E. M. Horn, K. K. Berekashvili, A. C. Widlitz, scFv sequences in a database and more than 90% homology E. B. Rosenzweig, and R. J. Barst, “Von Willebrand factor was found. It is known that all scFv from the I and J libraries independently predicts long-term survival in patients with are more than 90% homologous. No exact sequences to ours pulmonary arterial hypertension,” CHEST, vol. 128, no. 4, pp. were found. 2355–2362, 2005. In conclusion, our antibody fragments can be used in a [13] A. Y. Nossent, V. Van Marion, N. H. Van Tilburg et al., “von sandwich ELISA to determine the VWFpp levels in plasma. A Willebrand factor and its propeptide: the influence of secre- follow-up study is however needed to fully validate the assay. tion and clearance on protein levels and the risk of venous thrombosis,” Journal of Thrombosis and Haemostasis, vol. 4, no. Conflicts of Interest 12, pp. 2556–2562, 2006. [14] C.Wu, I. Liu, R. Lu, andH.Wu, “Advancement and applications The authors declare that there are no conflicts of interest of peptide phage display technology in biomedical science,” regarding the publication of this paper. Journal of Biomedical Science, vol. 23, no. 1, 2016. [15] B. Arza and J. Félez, “The Emerging Impact of Phage Display Acknowledgments Technology in Thrombosis and Haemostasis,” Thrombosis and Financial support for this study was provided by the Technol- Haemostasis, vol. 80, no. 09, pp. 354–362, 2017. ogy Innovation Agency of South Africa. [16] P. Pansri, N. Jaruseranee, K. Rangnoi, P. Kristensen, and M. Yamabhai, “A compact phage display human scFv library for References selection of antibodies to a wide variety of antigens,” BMC Bio-technology, vol. 9, article 6, 2009. [1] J. E. Sadler, “VonWillebrand factor: Two sides of a coin,” Journal [17] S. Hober, K. Nord, and M. Linhult, “Protein A chromatography of Thrombosis and Haemostasis, vol. 3, no. 8, pp. 1702–1709, for antibody purification,” Journal of Chromatography B, vol. 2005. 848, no. 1, pp. 40–47, 2007. 165 OD (490 nm) 166 BioMed Research International 7 [18] L. Baldi, D. L. Hacker, M. Adam, and F. M. Wurm, “Recombi- nant protein production by large-scale transient gene expres- sion in mammalian cells: State of the art and future perspec- tives,” Biotechnology Letters, vol. 29, no. 5, pp. 677–684, 2007. [19] N. Gera, M. Hussain, and B. M. Rao, “Protein selection using yeast surface display,”Methods, vol. 60, no. 1, pp. 15–26, 2013. [20] B. J. Tillotson, Y. K. Cho, and E. V. Shusta, “Cells and cell lysates: A direct approach for engineering antibodies againstmembrane proteins using yeast surface display,”Methods, vol. 60, no. 1, pp. 27–37, 2013. [21] J. Sheehan andW. A.Marasco, “Phage and yeast display,”Micro- biology Spectrum, vol. 3, no. 1, Article ID AID-0028-2014, 2015. [22] C. M. Y. Lee, N. Iorno, F. Sierro, and D. Christ, “Selection of human antibody fragments by phage display,” Nature Protocols, vol. 2, no. 11, pp. 3001–3008, 2007. [23] P. Goswami, D. Saini, and S. Sinha, “Phage Displayed scFv: PIII scaffold may fine tune binding specificity,” Hybridoma, vol. 28, no. 5, pp. 327–331, 2009. [24] M. Böhm, S. Täschner, E. Kretzschmar, R. Gerlach, E. J. Favaloro, and I. Scharrer, “Cold storage of citrated whole blood induces drastic time-dependent losses in factor VIII and von Willebrand factor: Potential for misdiagnosis of haemophilia and von Willebrand disease,” Blood Coagulation & Fibrinolysis, vol. 17, no. 1, pp. 39–45, 2006. 166 167 International Journal of Journal of Peptides Nucleic Acids The Scientific International Journal of International Journal of World Journal Cell Biology Microbiology Hindawi Publishing Corporation Hindawi Hindawi Hindawi Hindawi whtwtpw:/./hwinwdwaw.hii.ncodmawi.com Volume 20183 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Anatomy Biochemistry Research International Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Submit your manuscripts at www.hindawi.com Advances in Genetics Bioinformatics Research International Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Advances in Virolog y Z oInteornaltioonalg Joyurnal ofInternational Journal of Stem Cells BioMed Genomics International Research International Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 Neuroscience Journal Enzyme Journal of Journal of Research Parasitology Research Marine Biology Archaea Hindawi Hindawi Hindawi Hindawi Hindawi www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 www.hindawi.com Volume 2018 167 S-16 168 168 169 169 170 170 171 171 172 172 S-17 173 MEDICAL TECHNOLOGY SA A rapid and cost-effective method to visualize von Willebrand factor multimers in plasma 1. MURIEL MEIRING, Ph D, Department of Haematology and Cell Biology, University of the Free State and NHLS, Bloemfontein, 9300 (corresponding author) 2. PHILIP N BADENHORST, MD, Department of Haematology and Cell Biology, University of the Free State and NHLS, Bloemfontein, 9300 3. MARELI KELDERMAN, Diploma in Medical Technology, Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, 9300 Abstract The multimeric analysis of von Willebrand factor (VWF) is utilised for von Willebrand factor disease (VWD) subclass identification and is important for treating purposes. A highly sensitive and rapid method for the visualisation of the multimeric structure of VWF in plasma and platelets was described by Krizek et al in 2000 [1]. This method uses a western blot technique where horizontal agarose electrophoresis is followed by the transfer of the VWF onto a polyvinylidine fluoride (PVDF) membrane. The multimeric pattern of VWF is visualized by immunolocalisation and luminographic detection and no radioactivity is used. We modified this method com prehensively to increase its sensitivity and to reduce the cost and duration of the test. We used one in stead of two localisation anti bodies and thereby reduced the immunolocalisation time by more than two hours. This also reduced the cost of the procedure. We further reduced the cost by using two carbon plates for blotting in stead of a blotter instrument. A total cost reduction of 40% could be achieved. A higher sensitivity was obtained by degassing the agarose before the casting process. The higher sensitivity is reflect ed by the fact that differences between the multimer patterns of type 2M and normal patients could be detected. Introduction nol blue is added 1:10 to the diluted plasma and centrifuged at 14000 rpm for 1 minute in an Eppendorf centrifuge Von Willebrand factor (VWF) is a multimeric glycoprotein that plays a dual role in haemostasis. Firstly, it forms a complex with coagulation factor VIII Electrophoresis and it protects it from degradation. Secondly, VWF contributes to platelet ad hesion and aggregation by acting as a molecular bridge between sub endo The gel is set in place and the comb removed. Pre cooled electrophoresis thelial collagen and platelets [2]. The ability of VWF to support platelet ad buffer is poured onto the gel to overlay it with no more than 2 3 mm and 10 hesion and aggregation increases with multimer size. In the circulation mul µl of diluted sample is loaded into each well. The power is set on 30 mamps timer size is controlled by several mechanisms including proteolytic cleav (constant amp) for 30 minutes where after it is set at 50 mamps and run for age by the VWF cleaving protease, ADAMTS 13 [3, 4]. The steady state a further 4 6 hours until the dye front had migrated (±8cm). concentration of plasma VWF reflects the equilibrium between synthesis and Western Blot clearance. In addition, the multimer distribution of plasma VWF depends on competition between the mechanisms of clearance and of proteolysis by Two litres of transfer buffer (2.5 mM Tris, 19.2 mM glycine, 20% methanol, ADAMTS 13 [4, 5]. 0.01% SDS, pH 8.8) are prepared and stored at 4°C. After electrophoresis is Inherited VWD has been subdivided into three types that reflect its patho complete, the gel is placed in transfer buffer and equilibrated for 30 minutes. physiology. Types 1 and 3 VWD reflect respectively, the partial or virtually Polyvinylidene fluoride (PVDF) 0.45 µm membrane are cut only a bit small complete deficiency of VWF. Type 2 VWD is a qualitative defect that is sub er than the size of the gel, pre soaked in methanol for 1 2 minutes and stored divided into 4 subtypes (2A, 2B, 2M and 2N). Type 2A refers to variants with in transfer buffer until use. A western blot "sandwich" is assembled by plac decreased platelet dependent function and is associated with the absence of ing the cathode carbon block with black electrode at the bottom, 2 Scotch high molecular weight multimers. Type 2B refers to variants with increased Brite pads on top of it, 2 thick filter papers (Whatman CHR3, Whatman affinity for platelet glycoprotein 1b. Type 2M refers to variants with decreas International, England), 2 thin filter papers (Whatman no 1), the gel, the ed platelet dependent function not caused by the absence of high molecular PVDF membrane, 2 thin filter papers, 2 thick filter papers, 2 Scotch Brite weight multimers and Type 2N to variants with markedly decreased affinity pads and the anode block with red electrode on top. The pads and filter for factor VIII. papers are soaked in transfer buffer before the assembly. The whole sand No single test is available that provides appropriate information about the wich is placed into a plastic container at 4°C and covered with plastic wrap various functions of VWF and the laboratory diagnosis of VWD is based on to prevent evaporation. The transfer conditions are 70 mille ampere (con a panel of tests that includes the bleeding time, the measurement of factor stant ampere) overnight (15 17 hours). VIII coagulant activity (VIIIC), VWF antigen (VWF:Ag), VWF activity as measured by the ristocetin cofactor activity (VWF:RCo) and the collagen Blocking and immunolocalization of VWF multimers binding assay (VWF:CBA), VWF multimer analysis, ristocetin induced pla After blotting, the PVDF membrane is placed in the blocking agent that con telet agglutination (RIPA) and the factor VIII binding assay of plasma VWF. tains 5% skimmed milk powder in TBS for 1.5 hours at room temperature. The evaluation of the multimeric analysis of VWF is utilised for subclass Three washing steps are done with 50 ml TBS Tween for 1 minute and 3 identification of VWD. It is important to identify the specific subtype, be times 100 ml TBS Tween for 7.5 minutes each. The membrane is then placed cause treatment of VWD is based on the classification of the disease. A high into a 1:4000 dilution of an anti human VWF HRP conjugated antibody in ly sensitive and rapid method for the visualisation of the VWF multimers is TBS Tween for 1 hour 15 minutes. Washing steps are done as before where described [1], which we modified significantly to make it even more rapid, after the membrane was placed onto a mat surface with the protein side sensitive and cost effective. upwards. Equal volumes of ACL western blot detection reagent 1 and 2 (AEC Amersham, UK) are mixed and poured onto the membrane. The rea Materials and Methods gent must cover the entire surface of the membrane and must be held to the Preparation of SDS Agarose Gel surface by surface tension. After 1 minute incubation, the excess detection reagent is drained off, the membrane wrapped into plastic wrap and placed A 0.65% agarose gel is prepared in 100ml Tris acetate electrophoresis buffer in the dark on an x ray film and exposed for 1 2 minutes. The film is then (40 mM Tris, 0.1% SDS, 1 mM EDTA, pH 7.8) by melting the agarose until removed and developed in an automated film developer (Kodak, USA). The clear. After degassing by vacuum suction, the agarose is poured into a hori density of the high, intermediate and low molecular weight multimers are zontal gel apparatus with a 20 tooth comb in place and after solidification, determined using a Geldoc XR geldocumentation system (Bio Rad, CA, the gel is placed at 4°C for 30 min. USA). The total duration of this multimer analysis procedure is less than 28 hours. Sample preparation Results Whole blood of patients with a history of a bleeding tendency who was re ferred to our Haematology Clinic is collected into two vacutainer tubes con Typical patterns of VWF multimers of normal pooled plasma and vWD types taining 0.105 M sodium citrate in a ratio of 1:9 with blood. Platelet poor 1, 2A, 2B, and 2M are shown in figure 1. The density of the different bands plasma is prepared by centrifugation of whole blood at 2000 g for 20 min was plotted in figure 2 against the relative front (Rf value) of the different utes at room temperature. Samples are stored in 200 µl aliquots in poly lanes. The relative front is the distance of a band from the top of its lane, propylene tubes at 70°C until analysed. All tests are performed on aliquots divided by the total length of the lane. The larger multimers are represented that were not previously thawed. Each plasma sample is thawed at 37°C and by the smallest values on the x axis in figure 2. The first set of lanes in fig diluted 1:30 in sample buffer (0.01M Na2HPO4, 37 mM iodoacetamide and ure 1 contains plasma from a pool of 20 normal human volunteers. The high 1% SDS, pH 7.0). After incubation at 37°C for 60 minutes, 10 µl bromphe est density is in the region of the larger multimers (broken line in figure 2). 16 DECEMBER 2005 V1O7L3. 19 NO. 2 174 MEDICAL TECHNOLOGY SA Figure 1 The multimeric structure of VWF in normal plasma (NP) (lane sets 1 and 6), type 1VWD (lane set 2). Type 2A VWD (lane set 3), Type 2B VWD (lane set 4) and type 2M VWD (lane set 5). Figure 2 Densitometric tracing of lane sets 2 to 5. The density of the different bands is plotted against the relative front (Rf value) of the different lanes. VOL. 19 NO. 2 DECEMBER 2005 174 17 175 MEDICAL TECHNOLOGY SA Plasma from a patient with type 1 VWD was run in the second set of lanes patients, which is not the case with the method of Krizek et al in 2000 [1]. in figure 1. The highest density is also in the range of the larger multimers By using only one antibody to VWF instead of the two that is described in and due to the low VWF:Ag levels, the bands are much lighter (lower densi the literature, we shortened the duration of the test with more than 2 hours ty) than those of normal persons (figure 2A). Plasma from a patient with type and further reduced the cost of the method. 2A vWD was run in the lane set 3 (figure 1). No high molecular weight mul In conclusion, our modified luminographic method is a rapid, highly sen timers can be detected and an increase in the small multimers is visible. The sitive and cost effective visualisation method for VWF multimers in plasma low molecular weight multimers also show a higher density than those in and it will be wise to include the density profiles in the diagnostic work up normal plasma (figure 2C). Plasma from a patient with type 2B VWD was of von Willebrand disease. run in lane set 4 in figure 1. Only the intermediate and small multimers are visible. The density of these multimers is also higher than those in normal References plasma (figure 2B). Lane set 5 in figure 1 contains plasma from a type 2M patient. It is interesting to note the difference in distribution of the different 1. Krizek D.R., Rick M.E. (2000). A rapid method to visualize von size multimers compared to that of normal plasma. The density profile shows Willebrand Factor multimers by using Agarose Gel Electrophoresis, an almost even distribution of all multimers (figure 2D). The density of the Immunolocalization and Luminographic Detection. Thrombosis small multimers is higher than those of normal plasma and those of the large Research 97: 457 462. multimers less. 2. Tuddenham E.G. (1989). Von Willebrand factor and its disorders . An Discussion overview of recent molecular studies. Blood Reviews 3: 251 262. 3. Furlan M., Robles R., Lämmle B. (1996). Partial purification and char This method distinguishes type 1 from type 2A and 2B VWD. Type 2A acterization of a protease from human plasma cleaving von Willebrand shows a total absence of the high and intermediate molecular weight multi factor to fragments produced by in vivo proteolysis. Blood 87: 4223 mers and type 2B shows an absence of only the highest molecular weight 4234. multimers. Type 1 VWD shows a multimer pattern similar to that of normal plasma. Plasma from type 2M VWD patients contains a higher concentration 4. Tsai H M. (1996). Physiologic cleavage of von Willebrand factor by a of small multimers than normal plasma, although large, intermediate and plasma protease is dependent on its conformation and requires calcium small multimers are present. Perhaps the greatest advantages of our method ion. Blood 87: 4235 4244. are the rapid processing (less than 28 hours), high sensitivity to low concen trations of VWF, no radioactivity and the low cost of this procedure. A cost 5. Lenting P.J., Westein E., Terraube V., Ribba A.S., Huizinga E.G., Meyer comparison between this method and those described in the literature shows D., de Groot P.G., Denis C.V. (2004). An experimental model to study a reduction of 40% in favour of our method when all the apparatus and the in vivo survival of von Willebrand factor. Basic aspects and applica reagent costs are taken into account. We could also demonstrate the differ tion to the R1205H mutation. Journal of Biological Chemistry : 12102 ence in multimer pattern between normal plasma and that of type 2M 12109. 18 DECEMBER 2005 V1O7L5. 19 NO. 2 S-18 Journ1a7l 6of Thrombosis and Haemostasis, 10: 1043–1054 DOI: 10.1111/j.1538-7836.2012.04729.x ORIGINAL ARTICLE Differential sensitivity of von Willebrand factor (VWF) activity assays to large and small VWF molecular weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen-binding and mAb-based assays E . J . FAVALORO,* R . BONAR , K . CHAP MAN, M. ME IR ING§ and D . FUNK (ADCOCK)– *Haematology, Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, Westmead, NSW; Royal College of Pathologists of Australasia (RCPA) Haematology Quality Assurance Program (QAP), Northmead, NSW; Haematology Department, Hunter Area Pathology Service, Newcastle, NSW, Australia; §Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa; and –Esoterix Inc., Englewood, CO, USA To cite this article: Favaloro EJ, Bonar R, Chapman K, Meiring M, Funk (Adcock) D. Differential sensitivity of von Willebrand factor (VWF) activity assays to large and small VWF molecular weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen-binding and mAb-based assays. J Thromb Haemost 2012; 10: 1043–54. that these results have significant clinical implications for the Summary. Background: von Willebrand disease (VWD), the diagnosis of VWD and monitoring of its therapy, as well as for most common inherited bleeding disorder, is caused by the future diagnosis and therapy monitoring of thrombotic deficiencies and/or defects in von Willebrand factor (VWF). thrombocytopenic purpura. An effective diagnostic and VWD typing strategy requires plasma testing for factor VIII, and VWF antigen plus one or Keywords: diagnosis, high molecular weight sensitivity, labo- more VWF activity assays. VWF activity is classically assessed ratory testing, von Willebrand disease, von Willebrand factor. by using VWF ristocetin cofactor activity (VWF:RCo), although VWF collagen-binding (VWF:CB) and VWF mAb- Introduction based (VWF activity [VWF:Act]) assays are used by some laboratories. Objective: To perform a cross-laboratory study von Willebrand disease (VWD) is considered to be the most to specifically evaluate these three VWF activity assays for common inherited bleeding disorder. VWD is suspected comparative sensitivity to loss of high molecular weight following clinical and physical examination in individuals with (HMW) VWF, representing the form of VWF that is most personal and familial evidence of mucocutaneous bleeding, and functionally active and that is absent in some types of VWD, the diagnosis is confirmed by laboratory testing [1,2]. VWD namely 2A and 2B. Methods: A set of eight samples, including reflects quantitative or qualitative defects or deficiency in six selectively representing stepwise reduction in HMW VWF, plasma von Willebrand factor (VWF), and is classified into six were tested by 51 different laboratories using a variety of types. Types 1 and 3 define quantitative disorders with partial assays. Results: The combined data showed that the VWF:CB or complete loss of VWF, respectively, and types 2A, 2B, 2M and VWF:RCo assays had higher sensitivity to the loss of and 2N define qualitative disorders [1,2]. Notably, type 2A HMW VWF than did the VWF:Act assay. Moreover, within- VWD is characterized by decreased VWF-dependent platelet method analysis identified better HMW VWF sensitivity of adhesion, owing to selective deficiency of high molecular some VWF:CB assays than of others, with all VWF:CB assays weight (HMW) VWF multimers, type 2B VWD by increased still showing better sensitivity than the VWF:Act assay. affinity of VWF for platelet glycoprotein (GP)Ib, often also Differences were also identified between VWF:RCo methodol- associated with loss of HMW VWF, and type 2M VWD by ogies on the basis of either platelet aggregometry or as impaired VWF-dependent platelet adhesion without selective performed on automated analyzers. Conclusions: We believe deficiency of HMW VWF [2]. Focused laboratory testing generally includes assessment of plasma factor VIII coagulant activity (FVIII:C), and VWF Correspondence: Emmanuel J. Favaloro, Department of antigen (VWF:Ag) and VWF activity, supplemented with Haematology, Institute of Clinical Pathology and Medical Research additional tests on a case-by-case basis [1–3]. VWF activity is (ICPMR), Westmead Hospital, Westmead, NSW 2145, Australia. Tel.: +612 98456618; fax: +612 96892331. classically assessed with the VWF ristocetin cofactor E-mail: emmanuel.favaloro@swahs.health.nsw.gov.au (VWF:RCo) assay, originally described in the early 1970s [3– 5]. Problemswith this assay, including complexity, performance Received 20 December 2011, accepted 3 April 2012 time, poor reproducibility, and poor sensitivity to low levels of  2012 International Society on Thrombosis and Haemostasis 176 1044 E1.77J. Favaloro et al VWF (reviewed elsewhere [3]), has seen attention focused on A systematic, specific and comparative evaluation of VWF newer assays, primarily the VWF collagen-binding (VWF:CB) activity assays in terms of sensitivity to specific loss of HMW assay and various so-called VWF activity (VWF:Act) assays VWF has not, to our knowledge, ever been performed. based on mAb technology. Used in conjunction with VWF:Ag Accordingly, we describe a study that aimed to compare and FVIII activity, VWF activity assays have the capacity to results obtained with the three most commonly used activity identify and discriminate the various types of VWD. For assays (VWF:RCo, VWF:CB, and VWF:Act) by testing of a example, VWF:RCo/VWF:Ag (RCo/Ag) and VWF:CB/ normal sample as well as samples selectively and sequentially VWF:Ag (CB/Ag) ratios can be used to differentiate type 1 depleted of HMW and intermediate molecular weight (IMW) from types 2A, 2B and 2M VWD [1–3,6], given that these VWF forms, using a cross-laboratory exercise involving 51 assays are sensitive to loss of HMWVWF and may also reflect laboratories across a broad geographic region. VWF function (i.e. binding of VWF to platelet GPIb or collagen, respectively). Activity assays are also used as surro- Materials and methods gate laboratory markers of HMW VWF, either for VWD diagnosis [1–3], VWF concentrate production, therapy, and Preparation and initial testing of main study samples pharmacokinetic studies [7], or assessment of VWF protease (e.g. ADAMTS-13 activity or inhibition for investigation of The main study sample set is summarized in Table 1, and thrombotic thrombocytopenic purpura [TTP]) [8]. extended information is provided in Table S1 and Data S1. The VWF:CB assay is an ELISA-based assay first described The samples were derived primarily from a plasma pool of by Brown and Bosak in 1986 [9]. mAb-based VWF:Act individual normal plasmas. A series of eight plasma samples immunoassays, originally described in 1995, utilize a mAb was then initially produced from this pool by one of us (E.J.F.). directed against a functional epitope on VWF as the capture These comprised the otherwise unmodified pool (identified as and/or detection antibody [3,10–12]. Later independent vali- sample V1 in Table 1), and seven similar-volume aliquots of dation studies suggested inferiority of the commercial mAb this normal pool, each of which was differentially treated to ELISA-based assays to VWF:RCo and VWF:CB or even in- produce increasing stepwise loss of HMW and then IMW house mAb-based ELISA assays for discrimination of HMW VWF forms (Data S1, Table S1), This was achieved through a VWF-deficient VWD types such as 2A and 2B [13–15], and as propriety process of disulfide bond reduction with N-acetyl- most recently highlighted for a case study of type 2A VWD cysteine (NAC), similar to that recently described by Chen [16]. The latex agglutination assay developed by Instrumenta- et al. [23]. Although such treatment results in the loss of HMW tion Laboratory is currently the most popular mAb-based VWF, and also IMW VWF with continued application, the VWF:Act assay [17,18]. The true utility of this particular assay native VWF is otherwise essentially normally active [23]. The in VWD is only now emerging. Despite early reported high generated study samples were later anonymized to permit correlation with the VWF:RCo assay [17,18], these two assays blinded testing for the main study (Data S1, Table S1). The are increasingly recognized as providing different values with first few samples in the series (e.g. V2 and V3 in Table 1) many cases of VWD, notably qualitative variants [19,20]. represent a loss of HMW with a minor loss of IMW VWF, Assessments of the respective performance of VWF activity which might occur with a normal sample subsequent to a assays within cross-laboratory testing exercises are limited. The preanalytic event (e.g. filtration of a sample also intended for Royal College of Pathologists of Australasia (RCPA) Haema- lupus anticoagulant testing, or refrigeration of a whole blood tology Quality Assurance Program (QAP) last reported sample) [24–26]. Subsequent samples (e.g. V4 and V5 in findings in this area in 2007 [21]. At that time, the mAb Table 1) would then show increasing loss of HMW and IMW latex-based VWF:Act assay had recently emerged, and was VWF, potentially reflective of type 2A or 2B VWD-like performed by approximately 20% of participant laboratories, plasma, but with a relatively high level of VWF, which is a often as a replacement for the VWF:RCo assay. Nevertheless, pattern that might occur in type 2A or 2B VWD during discrepant behavior of the three different activity assays (i.e. pregnancy. The final sample in the series (V6 in Table 1) was VWF:RCo, VWF:CB, and VWF:Act) for discrimination of meant to be depleted of HMW and IMW VWF, and could several type 2B VWD cases was reported. More recently, the reflect a type 2A VWD pattern. North American Specialized Coagulation Laboratory Associ- The initial set comprised eight plasma samples (Table S1), ation (NASCOLA) reported their experience [22], highlighting which were tested for FVIII activity and various VWF test the emerging trend of the mAb VWF:Act assay replacing the parameters at the host laboratory (Institute of Clinical VWF:RCo assay in many North American laboratories, and Pathology and Medical Research [ICPMR], Westmead), prior despite any evidence of equivalence. In that study, overall to being frozen in both aliquot form and as large volume sets diagnostic interpretation error rates ranged from 3% for with stabilizers in preparation for lyophilization and later normal samples, to 28% for type 1 VWD, and to a staggering stability and homogeneity testing. Samples were subsequently 60% for type 2 VWD.Notably, the type 2 VWD samples were lyophilized in vials in 0.5-mL volumes by a commercial identified correctly by all laboratories using CB/Ag ratios, but lyophilization process, and thereafter stored refrigerated at by only one-third using RCo/Ag or VWF:Act/VWF:Ag (Act/ 4 C until required. All samples were retested by the host Ag) ratios. laboratory in a validation study, including homogeneity and  2012 International Society on Thrombosis and Haemostasis 177 178 VWF activity vs. VWF size 1045 Table 1 Summary of test sample set and characteristics Sample ID (current report) Sample comprised: Sample intended to, or that could feasibly, represent: V1 Pool of normal plasma samples Normal plasma V2 Sample V1 treated to yield minor loss of HMW VWF Normal plasma with minor loss of HMW VWF, as might be caused by a preanalytic event (see main text) V3 Sample V1 treated to yield loss of HMW and minor Normal plasma with mild loss of HMW VWF, as loss of IMW VWF might be caused by a preanalytic event (see main text) V4 Sample V1 treated to yield loss of HMW VWF and Type 2A or 2B VWD-like plasma as might be obtained moderate loss of IMW VWF in pregnancy V5 Sample V1 treated to yield loss of HMW VWF and Type 2A or 2B VWD-like plasma as might be obtained high loss of IMW VWF in pregnancy V6 Sample V1 treated to yield complete loss of HMW Type 2A VWD-like plasma. and IMW VWF V7 Mixture of V1 and VWF-deficient plasma Moderate type 1 VWD-like plasma with target VWF:Ag and FVIII:C = 20–30 U dL)1 V8 Mixture of HMW VWF-deficient sample and Type 2A or 2B VWD-like plasma with target VWF-deficient plasma VWF:Ag and FVIII:C = 20–30 U dL)1 FVIII:C, factor VIII coagulant; HMW, high molecular weight; IMW, intermediate molecular weight; VWD, von Willebrand disease; VWF, von Willebrand factor; VWF:Ag, von Willebrand factor antigen. Sample V1 represents a normal pool plasma and the main plasma used to generate subsequent test samples. Samples V2–V6 were derived from sample V1, and reflect sequential increasing depletion of HMW VWF, followed by sequential loss of IMW VWF. Samples V7 and V8 were, respectively, derived from mixtures of VWF-deficient plasma and sample V1 or an HMW VWF-deficient sample to produce samples reflective of moderate type 1 VWD and 2A/2B VWD, but with a similar level of VWF:Ag. stability, as extensively detailed in the Data S1. There were no mAbs against VWF [14,27,28] (Data S1). These mAbs bind obvious deleterious effects observed. Sample vials were also to various (different) sites on the VWF molecule, and include subsequently sent to four collaborative laboratories for local several that are reactive with the platelet GPIb-binding site blind testing of FVIII:C and VWF test parameters including on VWF, and which otherwise differentially block various multimer analysis to further validate study samples (details in VWF activities [27,28]. Notably, there was no evidence that Data S1). This testing also confirmed the selective and production of the samples adversely affected the core VWF sequential loss of HMW and, in some cases, IMW VWF structure, as represented by a structurally intact GPIb- (Fig. S1A,B), as well as the comparative reduction in func- binding site, despite samples reflecting a sequential loss of tional VWF test parameters (Fig. S1C,D). A subset of six HMW and IMW VWF (Data S1). However, it was also samples (Table 1, V1–V6) was finally selected as representing noted that: (i) mAb data were not identical to each other or the best differential stepwise pattern for distribution to to polyclonal antibody (RaVWF) data; (ii) the sample participants of the RCPA Haematology QAP for the main reflecting the greatest loss of HMW and IMW VWF (i.e. cross-laboratory study. V6) yielded a lower VWF protein level than the other On the basis of the preliminary testing, another two samples samples, with all VWF detection test systems (i.e. RaVWF were subsequently prepared tomimic amoderate type 1VWD- and all mAbs; Fig. S2) – thus, some alteration of core like and a more classical type 2A VWD-like plasma (Table 1; structure or internal disulfide reduction cannot be excluded V7 and V8). These were differentially and respectively prepared for this sample; and (iii) the levels of VWF protein appeared by dilution of either the normal pool sample or an HMW and to initially rise slightly in the first few samples, with a loss of IMW VWF-depleted sample into commercial VWF-deficient HMW and IMW VWF (Fig. S1C). This would be consistent plasma (see Data S1). Notably, the respective plasmas were with the test sample generation process providing improved specifically designed to produce two test plasmas with similar access for VWF antibody binding in these samples, rather levels of VWF:Ag (target of 25 U dL)1) and FVIII:C (target than elevation of VWF as such (see Data S1 and [29]). of 20–25 U dL)1), but with comparably differing VWF Another series of experiments were later performed with the activities, reflective, respectively, of a type 1 or 2A VWD (see same mAbs as VWF capture antibodies (i.e. rather than as Table 1 and Results). detection antibodies as above) (seeData S1).We have shown in the past that, used in this way, some mAbs show some HMW selectivity [14], as is also shown by the commercial mAb-based Further confirmation of test sample integrity activity assays. mAb-based assay ELISA testing The sample production process was intended to produce an incremental loss of Main cross-laboratory study HMW VWF, followed by a loss of IMW VWF, but was not intended or expected to substantially affect the core structure All laboratories (n = 55) enrolled in the standard VWF/VWD or activity of VWF [23]. In order to help confirm this, the test module (C) of the RCPAHaematology QAP (http://www. samples were tested with several various well-characterized rcpaqap.com.au/haematology/) were invited to participate  2012 International Society on Thrombosis and Haemostasis 178 1046 E1.79J. Favaloro et al in the main cross-laboratory study. This included three All data for all samples and all laboratories for the main laboratories that also participated in the validation study study test indices, namely VWF:Ag, VWF:RCo, VWF:CB, described above (see also Data S1). Two additional interna- VWF:Act, and FVIII:C, as well as assay ratios, are shown in tional collaborating laboratories that were also involved in the Fig. 1 as a scatter plot. This shows individual data and validation study were similarly invited to participate in the general trends, but also permits identification of outlier data, main study. All invited laboratories (total n = 57) were sent which may reflect aberrant testing from a few laboratories or three vials of each of eight anonymized samples (see Data S1; potential transcription errors. In brief, outlier data were only reidentified as V1 to V8 in Table 1), with sample reconstitution identified in the case of activity assays, and values > 3 and test instructions, requesting laboratories to perform their standard deviations from the mean were thereafter excluded normal test panels for testing of VWD. Laboratories were also from analysis. Outlier data were also checked for data source. asked for test methodology details and to interpret their test It was noted that five laboratories were responsible for nearly results. The choices offered included normal, equivocal, type 1 half of all the data outliers identified in Fig. 1. These outlier VWD (mild, moderate, or severe), type 2A or 2B VWD, data were also thereafter excluded from analysis. Summarized type 2M VWD, type 2N VWD, type 3 VWD, or other (with a data (excluding outliers noted above), including correspond- request to specify). Some of these choices (e.g. type 2N VWD ing assay ratios, are shown in Fig. 2, which attempts to focus and type 3 VWD) did not reflect any of the samples, but a on assay performance rather than participant performance. complete range of possible interpretations was provided to Note that VWF:Ag and FVIII:C testing identified similar avoid any potential biasing of returned data. Laboratory levels across all samples tested except for sample V6, similar numerical data were used as reported by participants, except to the trend identified prior to the main study by collaborative that values reported as < a given value were corrected laboratories (Fig. S1). In contrast, and as expected, downwards to permit numerical analysis (for example, an assay VWF:RCo, VWF:CB and VWF:Act levels fell sequentially value of < 10 U dL)1 was corrected to 9 U dL)1, and an according to increasing loss (or decreasing levels) of HMW assay ratio of < 0.2 was corrected to 0.1). This is consistent and IMW VWF. Notably, VWF:RCo and VWF:CB showed with our normal External Quality Assurance (EQA) practice. the greatest (similar) falls, particularly for samples showing These events occurred only for samples with high loss ofHMW the highest loss of HMW VWF, whereas VWF:Act showed VWF (i.e. V6 and V8), and only for VWF:RCo and VWF:CB intermediate results. Consistent with the trends observed for testing (respectively: five events for V6 VWF:RCo, four events VWF:Ag and FVIII, the FVIII/VWF:Ag (FVIII:C/Ag) ratio for V6 VWF:CB, four events for V8 VWF:RCo, and three showed only a moderate trend to reduction. Act/Ag ratios events for V8 VWF:CB). also showed a gradual trend to reduction, but this was overshadowed by the greater reductions observed in RCo/Ag and CB/Ag ratios. Comparative data for samples V7 and V8, Statistical analysis intended to represent a moderate type 1 VWD and a type 2A Numerical data were analyzed by means of comparative VWD sample, respectively, are shown in Fig. 2C,D. The medians, means, ranges and interassay (interlaboratory) coef- type 1 VWD mimic sample (V7) yielded similar numerical ficients of variation, with GRAPHPAD PRISM (GraphPad data for all test parameters (between 20 and 30 U mL)1), with Software, La Jolla, CA, USA; http://www.graphpad.com). normal assay ratios (all > 0.7). In contrast, although the type 2A VWD mimic (V8) yielded similar numerical data for VWF:Ag and FVIII:C (around 20–30 U mL)1) and similar Results data to sample V7, the VWF activity data (VWF:RCo, Validation data for the samples prepared and dispatched in VWF:CB, and VWF:Act) showed decreased values, as did the main study are detailed in Data S1 and Figs 1–3. The respective ratio data, and as expected for this sample. main study data are derived from 51 participant laboratories Consistent with the findings in Fig. 2B, VWF:RCo and (as not all invited laboratories [n = 57 in total] returned VWF:CB (and thus RCo/Ag and CB/Ag ratios) provided the data), and are presented in Figs 1–3, with multimer patterns lowest comparative values. from one of the collaborating laboratories used as a reference A subanalysis of data according to type of VWF:CB and site shown in Fig. 4. A breakdown of laboratory tests as used VWF:RCo assay is shown in Fig. 3. Although data should be by laboratories, as well as summary statistical data, is given in interpreted cautiously, given low subsample test numbers, some Table 2. VWF:CB assays appeared to be more sensitive to the loss of Fig. 1. Main study data as derived from all samples tested where reported by 51 participating laboratories and shown as a scatter plot. The data for the main study test indices, namely von Willebrand factor (VWF) antigen (VWF:Ag), VWF ristocetin cofactor (VWF:RCo), VWF collagen binding (VWF:CB), VWF activity (VWF:Act), and factor VIII coagulant activity (FVIII:C), are shown on the left side of each figure. The horizontal dashed line represents a nominal normal cut-off value of 50 U dL)1. Data for assay ratios are shown on the right side of each figure. The horizontal dashed line represents a nominal normal cut-off value of 0.6. (A)–(H), respectively, show data for samples V1–V8 (see Table 1). Some outlier data are evident; data in squares indicate values that are> 3 standard deviations from themean.Data in circles show other visually appearing outlier data, in general between 2 and 3 standard deviations from the mean. All identified outlier data were checked for transcription error and for source of data.  2012 International Society on Thrombosis and Haemostasis 179 180 VWF activity vs. VWF size 1047 A V1 E V5 120 1.6 120 1.6 110 1.5 110 1.5 1.4 1.4 100 1.3 100 1.3 90 1.2 90 1.2 80 1.1 80 1.1 1.0 1.0 70 0.9 70 0.9 60 0.8 60 0.8 50 0.7 50 0.7 0.6 0.6 40 0.5 40 0.5 30 0.4 30 0.4 20 0.3 20 0.3 0.2 0.2 10 0.1 10 0.1 0 0.0 0 0.0 Ag Co ct: B :C g g /A g Ag Ag Ag Co ct B I:C /A Ag Ag AgF :R :A C I A / /W F F F : I / / / : A C I / W W FV I Co ct B :C F :R F: F: VI Co t B C V W V R A C III W F c W W F R A C III : V V V VF V W V V FV B V2 F V6 120 1.6 110 1.5 120 1.6 1.4 1.5 100 1101.3 1.4 100 90 1.2 1.3 80 1.1 90 Stago and 1.2 1.0 80 the rest 1.1 70 0.9 1.070 60 0.8 0.9 60 Siemens 0.8 50 0.7 and Vidas 0.6 50 0.7 40 0.60.5 40 0.5 30 0.4 30 0.4 20 0.3 0.3 0.2 20 10 0.20.1 10 0.1 0 0.0 0 0.0 Ag C o ct CB C: A II: /A g Ag g g g F :R : : I o ct/ B/ A /A g Co ct B :C A g g gA F F F V C : A C I:C F :R F: A :C III o/ t/A /Ac B C/ A VW W W VW F R II W F F VV V V W W VW F RC A C II: V F V V I FV C V3120 G1.6 V7 120 1.6 110 1.5 1.4 110 1.5 100 1.3 1.4100 90 1.2 1.3 90 80 1.1 1.2 1.0 80 1.1 70 1.00.9 70 60 0.90.8 60 0.8 50 0.7 50 0.70.6 40 0.60.5 40 0.5 30 0.4 30 0.4 20 0.3 20 0.3 0.2 0.2 10 0.1 10 0.1 0 0.0 0 0.0 g Co ct B :C A g g g g g o t C g:A :R :A :C II I o/ t/A /A /A A C Ac CB: : II:R : I o/ A t/A g Ag g F / / A W F F F V C W W W F R A c CB I:CI F F:I W F F V C c B :C V V V V V V W VW W F R A C III F V V FV D V4 H V8 120 1.6 120 1.6 110 1.5 110 1.5 1.4 1.4 100 1.3 100 1.3 90 1.2 90 1.2 80 1.1 80 1.1 1.0 1.0 70 0.9 70 0.9 60 0.8 60 0.8 50 0.7 50 0.7 0.6 0.6 40 0.5 40 0.5 30 0.4 30 0.4 20 0.3 20 0.3 0.2 0.2 10 0.1 10 0.1 0 0.0 0 0.0 g o t B C g g g g g o :A C Ac c t B C g g g g F :R F: F: C III : o/ A t/A B/ A /A :A RC :A :C III : /A A A C F : o / A F FV C Ac C II: F F F FV C Ac t/ B/ C VW W W VW R I W W W R C III : V V WFV V V V V FV  2012 International Society on Thrombosis and Haemostasis 180 VWF or FVIII (U dL–1) VWF or FVIII (U dL–1) VWF or FVIII (U dL–1) VWF or FVIII (U dL–1) –1 VWF or FVIII (U dL–1VWF or FVIII (U dL ) ) VWF or FVIII (U dL–1) VWF or FVIII (U dL–1) Ratio Ratio Ratio Ratio Ratio Ratio Ratio Ratio 1048 E1.81J. Favaloro et al A C 35 1.2 120 120 1.1 110 110 30 1.0 100 VWF:Ag 100 0.9 90 90 25 0.8 80 80 20 0.7 70 70 0.6 60 FVIII:C 60 15 0.5 50 50 0.4 40 VWF:RCo 40 10 VWF:Act 0.3 30 30 0.2 20 20 5 0.1 10 10 VWF:CB 0 0.0 0 0 g o ctA C g g g: R :A :C B :C III /A t/A /A /A g V1 V2 V3 V4 V5 V6 FW F: F F o c B C Sample ID V W VW VW FV RC A C II : V FV I Decreasing HMW VMF B 1.2 1.2 D 35 1.2 1.1 1.1 1.1 1.0 1.0 30 1.0 0.9 0.9 0.9 0.8 25FVIII:C/Ag 0.8 0.8 0.7 0.7 20 0.7 0.6 0.6 0.6 0.5 0.5 RCo/Ag Act/Ag 15 0.5 0.4 0.4 0.4 0.3 0.3 10 0.3 0.2 0.2 5 0.2 0.1 CB/Ag 0.1 0.1 0.0 0.0 1 2 3 4 5 6 0 0.0V V V V V V g:A C o Ac t CB I:C g g g g Sample ID F R : : II / A t/A /A /A W F: F F V C o Ac CB II: C Decreasing HMW VMF V W VWV V W F R FV I Fig. 2. Summarized main study data as derived from all samples tested where reported by 51 participating laboratories and shown as mean ± standard error of the mean. Some outlier data identified in Fig. 1 have been removed (refer to text) to permit a focus on method-based differences (i.e. to exclude potential participant-based problems). (A) Main study test indices, namely von Willebrand factor (VWF) antigen (VWF:Ag), VWF ristocetin cofactor activity (VWF:RCo), VWF collagen binding (VWF:CB), VWF activity (VWF:Act), and factor VIII coagulant activity (FVIII:C), as reported by all laboratories. The horizontal dashed line represents a nominal normal cut-off value of 50 U dL)1. (B) Data for corresponding assay ratios. The horizontal dashed line represents a nominal normal cut-off value of 0.6. (C, D) Comparative data for samples V7 (C) and V8 (D), intended to, respectively, represent a moderate type 1 von Willebrand disease (VWD) and a type 2A VWD sample. HMWVWF than others. Similarly, there also appeared to be a In order to help explain some of the above observations, small difference in sensitivity between aggregometer-based and further studies (as outlined in Materials and methods) were automated test processes for VWF:RCo. Figure 3C shows performed by the host laboratory using several mAbs, includ- comparative data between the best and least well performing ing those known to react with VWF at the functional GPIb- commercial VWF:CB assays, with VWF:RCo (aggregometry) binding site, which are presumed to be analogous to the mAb and VWF:Act. used in the commercial VWF:Act assay. The data are shown in The reference laboratory multimer analysis shown in Fig. 4 Data S1 (Fig. S3). In brief, when used as the VWF capture is consistent with previous observations, and confirms the system, some mAbs could be shown to yield HMW sensitivity sequential loss of HMW and IMW VWF in the generated data that were very similar to those of the VWF:Act assay. This samples from V1 to V6, as well as the loss in sample V8 sensitivity was increased by manipulation of the assay, most (type 2AVWDmimic), but not in sample V7 (moderate type 1 notably by reducing the concentration of coating mAb used VWD mimic). The multimer analysis results submitted by two (see Discussion). study participant laboratories yielded similar findings (data not Interpretative data from participants are shown in Fig. 5. shown). Notably, all participants identified sample V1 as being normal,  2012 International Society on Thrombosis and Haemostasis 181 Ratio VWF or FVIII:C (U dL –1) VWF or FVIII:C (U dL–1) VWF or FVIII:C (U dL–1) Ratio Ratio VWF or FVIII:C (U dL–1) Ratio 182 VWF activity vs. VWF size 1049 A 1.2 1.2 HMW multimer level of normal 1.1 1.1 1.0 1.0 0.9 TC 0.9 0.8 0.8 S 0.7 0.7 0.6 0.6 0.5 0.5 0.4 LD 0.4 In-house 0.3 0.3 Nor 2B 1 2 3 4 5 6 7 8 2B Nor 0.2 0.2 PBI 0.1 0.1 Fig. 4. Multimer patterns obtained on themain study sample set from one collaborating laboratory, and used as the reference multimers, and con- 0.0 0.0 V1 V2 V3 V4 V5 V6 firming sequential loss in high molecular weight (HMW) and intermediate Sample ID molecular weight VWF in samples V1, V2, V3, V4, V5, and V6, as well as Decreasing HMW VWF the loss in sample V8 (type 2A vonWillebrand disease [VWD]mimic), but B not sample V7 (moderate type 1 VWD mimic). Nor, normal plasma 1.2 1.2 control; 2B, type 2B VWD plasma control. 1.1 1.1 1.0 1.0 which was reassuring. The vast majority of participants 0.9 0.9 identified sample V6 (devoid of HMW VWF) as being 0.8 0.8 type 2A, 2B or 2M VWD-like. There was also a trend for 0.7 0.7 0.6 0.6 decreasing identification of the samples as normal, to that of 0.5 Agg 0.5 them being increasingly identified as type 2A, 2B or 2MVWD- 0.4 0.4 like, as the samples progressively lost HMW and then IMW 0.3 Auto 0.3 VWF.Most participants identified sample V7 as type 1 VWD- 0.2 0.2 like, and most also identified sample V8 as type 2A, 2B or 2M 0.1 0.1 VWD-like. Somewhat concerning, perhaps, was that eight 0.0 0.0 participants identified HMW VWF-deficient sample V4 as V1 V2 V3 V4 V5 V6 normal, with three of these same participants also identifying Sample ID Decreasing HMW VWF sample V5 as normal. Similarly concerning was that 10 C participants identified the HMW VWF-deficient sample V8 0.7 0.7 as a type 1 VWD. 0.6 0.6 Discussion 0.5 0.5 Act/Ag To our knowledge, this study represents the first comparative 0.4 0.4 assessment of themost widely usedVWF activity-based assays (i.e., VWF:Co, VWF:CB, and VWF:Act) for their specific 0.3 0.3 sensitivity to loss of HMW VWF. VWF:RCo, the original 0.2 RCo (Agg)/Ag) 0.2 functional VWF assay described in the early 1970s [4,5], reflects CB(TC)/Ag the ability of VWF to bind to its major platelet receptor 0.1 0.1 CB (PBI)/Ag (GPIb). The VWF:CB assay, first described in 1986 [9], 0.0 0.0 represents another activity of VWF, namely its ability to bind V3 V4 V5 to collagen, a subendothelial matrix component. TheVWF:Act Sample ID assay, first reported as an immunoradiometric assay in 1985, Decreasing HMW VWF and later as an ELISA assay, is now most commonly Fig. 3. Main study data from Fig. 2, but now showing submethod anal- performed with an immunolatex procedure [10–13,16,17]. ysis for ratios of von Willebrand factor (VWF) collagen binding The mAb-based VWF:Act assay is marketed as an activity (VWF:CB)/VWF antigen (VWF:Ag) (CB/Ag) (A) and VWF ristocetin assay on the basis that the mAb used to capture VWF in the cofactor activity (VWF:RCo)/VWF antigen (VWF:Ag) (RCo/Ag) (B). (C) Comparative data for best (Precision BioLogic Inc. [PBC]) and least well test sample recognizes the GPIb-binding site of VWF.Whether (TechnoClone [TC]) performing method in the VWF:CB group vs. the this then bestows functionality to the assay is debated. RCo (Agg)/Ag and VWF activity (VWF:Act)/VWF antigen (VWF:Ag) Nevertheless, the assay has been embraced by a high propor- (Act/Ag) groups for main three samples (V4, V5, and V6) showing the tion of laboratories, and in many cases inappropriately used as greatest loss of high molecular weight (HMW) VWF (see Fig. 4). Dashed a surrogate for the VWF:RCo assay. Although the RCPA horizontal line at 0.6 represents a typical nominal normal cut-off value. Agg, agglutination; Auto, automated; LD, Life Diagnostics; S, Stago. Haematology QAP has noted such a trend in Australia [21], this was also recently noted in North America by the  2012 International Society on Thrombosis and Haemostasis 182 Ratio RCo/Ag ratio CB/Ag ratio Ratio CB/Ag ratio RCo/Ag ratio 1050 E1.83J. Favaloro et al  2012 International Society on Thrombosis and Haemostasis 183 Table 2 Summary test statistics and breakdown of main study methods used by study participants No. of Median (range); CV (of all data)§ participants Test/ reporting % of % of V1 V2 V3 V4 V5 V6 V7 V8 methodology results* total method FVIII:C 51 100 100 66.9 58.7 62.0 60.6 58.2 44.0 21.0 18.0 (47.0–87.0); (39.0–73.0); (44.0–75.0); (44.0–78.0); (40.0–73.0); (27.0–61.0); (14.0–30.0); (11.0–29.0); 9.5 11.2 10.7 10.4 11.9 18.7 14.5 20.1 VWF:Ag 52 100 100 82.1 78.3 89.0 93.0 98.0 74.0 26.0 26.1 (67.0–96.9); (65.0–107); (71.0–117); (70.2–118); (68.4–115); (37.0–99.0); (21.0–33.0); (19.0–33.0); 7.2 10.5 9.6 10.0 11.1 24.3 10.1 12.0 LIA 41 80.4 80.4 – – – – – – – – ELISA 6 11.8 11.8 – – – – – – – – ELFA 4 7.8 7.8 – – – – – – – – VWF:RCo 30 60.8 100 69.0 61.5 47.0 37.5 22.2 9.0 (0–37.0); 20.0 9.0 (61.0–97.0); (46.0–81.0); (39.0–85.0); (27.0–89.0); (11.0–67.0); 85.3 (13.0–37.0); (0–18.0); 12.4 14.1 20.3 35.6 27.3 24.1 58.1 Aggregometer 9 17.6 29.0 – – – – – – – – Automated 22 43.1 71.0 – – – – – – – – VWF:CB 33 60.8 100 83.0 70.0 57.0 39.0 18.5 4.0 (0–14.0); 23.0 5.0 (55.0–113); (53.0–107); (33.0–91.0); (19.0–71.0); (5.0–43.0); 84.7 (17.0–36.0); (1.0–19.0); 14.2 17.2 22.2 30.7 45.5 18.6 65.2 In-house 6 11.8 19.4 – – – – – – – – Stago 7 13.7 22.6 – – – – – – – – Technoclone 11 21.6 35.5 – – – – – – – – Life 5 9.8 16.1 – – – – – – – – Therapeutics Group– Precision 2 3.9 6.5 – – – – – – – – Biologic Inc. VWF:Act** 11 21.6 100 77.2 62.0 57.2 51.0 36.0 22.0 28.9 14.0 (60.0–98); (41.0–79.0); (48.0–73.0); (47.0–64.0); (30.0–48.0); (10.0–46.0); (21.0–45.0); (11.0–24.0); 14.9 19.0 12.2 12.6 17.4 37.9 25.4 30.6 Multimers 4 7.8 NA NA NA NA NA NA NA NA NA CV, coefficient of variation; ELFA, enzyme-linked fluorescence assay; FVIII:C, factor VIII coagulant activity; LIA, latex immunoassay; NA, not applicable; VWF:Act, von Willebrand factor activity; VWF:Ag, von Willebrand factor antigen; VWF:CB, von Willebrand factor collagen binding; VWF:RCo, von Willebrand factor ristocetin cofactor. *A total of 57 laboratories were invited to participate, but only 51 laboratories provided test data for main study samples. Some laboratories reported results for multiple test systems and methodologies, whereas other laboratories reported data only for selective tests and methodologies; thus, test numbers will differ in each case, and do not always add up to 51. Percentage of total participants reporting results for each test group. Percentage of participants reporting results for specific test or methodology groups (VWF:Ag, VWF:RCo, or VWF:CB, etc.). §Includes outlier data; data not reported for submethodology. –Comprises group of Life Therapeutics Ltd, Reaads Corgenix (Corgenix Medical Corporation) and Nusep Inc. **All participants reported using the Instrumentation Laboratory assay. 184 VWF activity vs. VWF size 1051 50 RCo/Ag and CB/Ag ratios) generally perform better than the Normal Equiv 2A/B/M 2N 1 3 Other 45 VWF:Act assay (and Act/Ag ratios) [21]. Interestingly, the NASCOLA study also identified better performance of the 40 VWF:CB assay than either the VWF:Act or VWF:RCo assay 35 in the context of such discrimination [22,30,31]. The current 30 study expands on these findings, and in part may also explain 25 possible reasons behind them, given that the VWF:Act assay 20 appears to be less sensitive to the loss ofHMWVWF than both the VWF:RCo and VWF:CB assays. 15 Although the VWF:CB and VWF:RCo assays (and thus the 10 CB/Ag and RCo/Ag ratios) showed similar trends, the data 5 were not identical. Methodology subanalysis was also per- 0 formed, and although small subgroup numbers prevent any V1 V2 V3 V4 V5 V6 V7 V8 definitive conclusions, there did appear to be some differences Sample ID ‘VWD1’ ‘VWD2A’ in relation to methodology (Fig. 3). For example, the least Decreasing HMW VWF sensitive VWF:CB assay for HMW VWF appeared to be that Fig. 5. Participant interpretations for tested samples. Note that: (A) not produced by Technoclone GmbH, an observation that is quite all participant laboratories provided an interpretation for test samples, so consistent with many previous evaluations of VWD cases by numbers are generally < 51; (B) all participants providing an interpreta- the host laboratory [15,32]. Even so, this least sensitive tion identified sample V1 as being normal; (C) the vast majority identified the high molecular weight (HMW) von Willebrand factor (VWF)-devoid VWF:CB assay for HMW VWF still appeared to be more sample (V6) as being type 2A, 2B or 2M von Willebrand disease (VWD)- sensitive than the Act/Ag ratio (Fig. 3C). For RCo/Ag ratios, like; (D) there was a trend for decreasing identification of the samples as the automated VWF:RComethod was as sensitive, if not more normal, to them increasingly being identified as type 2A, 2B or 2MVWD- so, to the increasing loss of HMW and IMWVWF, except for like, as the samples progressively lost HMW and then intermediate sample V6, where it is suspected that low limit of VWF molecular weight VWF (i.e. from sample V2 to sample V6); (E) most participants identified sample V7 as type 1 VWD-like; (F) most partici- sensitivity issues compromise the assays utility (that is, many )1 pants identified sample V8 as type 2A, 2B or 2M VWD-like. However, laboratories cannot report assay values below 10 U dL with some participants identified HMW VWF-deficient samples V4 and V5 as this assay [33]). This lower limit of sensitivity can be improved normal, and some participants identified HMW-deficient sample V8 as a by use of a low assay curve, as recently reported [34,35]. type 1 VWD. Participants in the main study provided interpretations of their data that were generally consistent with the sample type NASCOLA [22]. This trend is presumably related to the assays tested (Fig. 5). On occasion, however, interpretations appeared ease of use, and early (but incomplete) studies showing to be at odds with the sample type. Notably, several partic- behavior similar to that of the VWF:RCo assay by regression ipants identified HMW VWF-decreased samples V4 (eight analysis of selected samples. In some cases, preferential usage laboratories) and V5 (three laboratories) as normal, and may also be driven by regulatory requirements. For example, sample V8 (10 laboratories) as type 1 VWD. Interestingly, an although the VWF:Act assay has been cleared by the FDA for analysis of these findings appears to identify the major problem in vitro diagnostic use in North America, a VWF:CB assay has as being that of limited test panels, rather than problems with not. As clearly shown in the current study, the VWF:Act assay activity assays as such. Thus, these 21 occasions reflected does show some selective discrimination of HMW VWF, but testing by 15 laboratories, all of which performed the VWF:Ag the VWF:RCo andVWF:CB assays aremuchmore effective in assay. Three of the 15 laboratories performed no activity assay this regard. This sensitivity of the VWF:Act assay for HMW of any kind, eight laboratories performed the VWF:RCo assay, VWFmay be related to the manner in which the manufacturer five performed the VWF:CB assay, and three performed the has controlled the assay conditions, something that the host VWF:Act assay, with eight of 15 (53.3%) therefore performing laboratory for the current study has shown is possible for only a single activity assay, and only four performing two ELISA mAb-based capture systems, simply by reducing the activity assays. Thus, 11 of 15 (73.3%) laboratories performed mAb concentration used (Fig. S3, Data S1, and [14]). Thus, the only one or noVWFactivity assays, compromising their ability use of a lower concentration of mAb as a capture systemmakes to identify a loss of HMW VWF in this study (and a potential the ELISA-based assay more selective for HMW VWF, qualitative type 2 VWD otherwise). This finding is also presumably because low molecular weight VWF does not consistent with our previous experience [36]. contain sufficient binding sites to permit stable binding of the The current study utilized a propriety process that employed VWF to themicroplate. In a latex agglutination assay, it can be NAC to achieve a stepwise reduction in HMW and then IMW hypothesized that a similar process may occur; that is, cross- VWF. This in vitro process is thought to mimic a natural latex agglutination may also require VWF of a certain mass, in vivo process, described but incompletely characterized, that according to the amount of latex-bound mAb. We have permits reduction of VWF in the absence of ADAMTS-13 [37], previously shown that, in discrimination of type 2 vs. type 1 and also possibly that assists in the formation of VWF VWD cases, the VWF:RCo and VWF:CB assays (and thus the complexes at a thrombus by a process of self-association [38].  2012 International Society on Thrombosis and Haemostasis 184 Number of participants 1052 E1.85J. Favaloro et al The in vitro processing of VWF with NAC has most recently similar and showing the highest sensitivity, and the VWF:Act been described by Chen et al. [23] as a potential therapeutic aid assay showing lower sensitivity. We believe that these findings in TTP, and, indeed, we believe that clinical trials of this agent have significant implications for clinical practice in a variety of have recently begun or being planned [37]. Should this agent settings, namely diagnostic and therapy management practice become a treatment of choice for TTP, it can be envisaged that for both VWD and TTP. In particular, diagnostic and therapy laboratories may be called upon to monitor treatment in TTP management of both VWD and TTP rely, in part, on assays by using VWF assays, including the VWF:Ag assay and an that are defined as being sensitive to the loss of HMW VWF, HMWVWF activity surrogate. The current study would then but not all activity assays show similar sensitivities to the loss suggest caution in regard to the use of the VWF:Act assay for of HMW VWF. There now remains a stock of reserved this purpose. samples prepared for this study that can be used for ongoing Finally, VWD therapy primarily involves the use of either EQA, or for a more extensive, perhaps expert laboratory-based desmopressin (DDAVP) or VWF factor concentrate [1,2]. The study, similar to that previously reported for patient samples same assays that are used to diagnose VWD are also used to [40]. A clinical validation study, showing comparative findings monitor therapy for VWD, and to assess the potential clinical of plasma samples fromNAC-treated TTP patients in a clinical utility of factor concentrates. Of additional interest, DDAVP trial setting and cotested by different VWF activity assays, therapy can also be used to assist in VWDdiagnosis and typing would also seem to be warranted. [6], with various test patterns being observed in different VWD cases. Recently, the VWF:Act assay was proposed as a possible Addendum suitable alternative to theVWF:RCoassay in termsofassessing VWF factor concentrates [39]. The current study, however, E. J. Favaloro: conceived, designed and coordinated the study, would caution against the expectation that the VWF:Act and prepared the study samples, undertook preliminary testing of VWF:RCo assays will provide the same information in such study samples, participated in method validation, undertook assessments. Indeed, although a given factor concentrate may data analysis, and wrote the manuscript; R. Bonar: arranged provide the same value for VWF:RCo and for VWF:Act, this is andoversaw the logistics of themain study, arranged for sample not the same as identifying these assays as being equivalent for lyophilization, and contributed to data analysis; K. Chapman: this purpose. Thus, on the basis of the current study, it can be assisted in study design, participated in preliminary testing of predicted that the VWF:Act assay will provide higher VWF study samples and method validation, and participated in the values than theVWF:RCoassay forVWFconcentrates that are main study; M. Meiring: participated in preliminary testing of somewhat devoid of HMWVWF. Although potentially favor- study samples and method validation, and participated in the able to manufacturers of VWF concentrates, as this gives the main study. D. Funk (Adcock): participated in preliminary impression of more favorable characteristics, this is not recom- testing of study samples and method validation, and partic- mended practice. ipated in the main study. D. Funk (Adcock)s laboratory acted Although the sample production used in this study reflected as the studyreferencecenter formultimeranalysis.All coauthors an in vitro process to reduce the levels of HMW and IMW contributed tomanuscript revision, and have approved the final VWF, we do not believe that this caused any substantial manuscript for publication. reduction in core VWF function as such. This was shown by testing with various mAbs against VWF, including functional Acknowledgements sites on VWF, which showed comparable data to those obtained with polyclonal RaVWF material (Data S1). This The authors would like to thank the following individuals: N. would, in essence, suggest an intact VWF functional GPIb- Yan and M. Bolan, from Precision Biologic Inc. (Dartmouth, binding site. This is also consistent with current knowledge that Nova Scotia, Canada), for performing testing of study samples identifies very few cysteine molecules within the A domains of as part of both the initial validation andmain study evaluation; VWF [38]. Moreover, the FVIII/VWF:Ag ratio also remained S. Mohammed, J. McDonald and E. Grezchnik, from the fairly stable across the range of samples, again suggesting an ICPMRlaboratory,Westmead, forperformingsomeof thehost intact VWF functional FVIII-binding site. Although there was laboratory testing reported in this study; M. Kelderman, for a slight drop in the FVIII/VWF:Ag ratio across the samples, performing VWF:RCo assays at the Department of Haematol- this was more likely related to the slight increase in VWF:Ag ogy and Cell Biology, University of the Free State; C. Jordan observed than to any fall in FVIII:C (Fig. 2A), with this MT (ASCP), from Esoterix Inc., for performing multimer presumably reflecting greater accessibility of the antibodies to analysis; K. Marsden, Chair of the RCPAHaematology QAP, the VWF in the samples [29] rather than an increase in for critical review of the manuscript; and all RCPAHaematol- VWF:Ag as such. ogy QAP laboratories and staff who participated in the VWF In conclusion, we report on a cross-laboratory evaluation of special exercise reported in part within this paper. VWF testing with a range of assays and samples selectively depleted in HMWand, in some cases, IMWVWF. Differences Disclosure of Conflict of Interests in the ability of activity assays to detect this loss were observed, with the VWF:RCo and VWF:CB assays being The authors state that they have no conflict of interest.  2012 International Society on Thrombosis and Haemostasis 185 186 VWF activity vs. VWF size 1053 Willebrand factor – demonstration that VWF function resides in a Supporting Information conformational epitope. Thromb Haemost 1986; 55: 318–24. 12 Murdock PJ, Woodhams BJ, Mathews KB, Pasi KJ, Goodall AH. Additional Supporting Informationmay be found in the online von Willebrand factor activity detected in a monoclonal antibody- version of this article: based ELISA: an alternative to the ristocetin cofactor platelet agglutination assay for diagnostic use. Thromb Haemost 1997; 78: Data S1. Extended description of materials and methods 1272–7. related to the validation study. 13 Nitu-Whalley IC, Riddell A, Lee CA, Pasi KJ, Owens D, Enayat MS, Figure S1. Main initial validation dataset. Perkins SJ, Jenkins PV. Identification of type 2 vonWillebrand disease Figure S2. VWF:Ag testing of test samples by the host in previously diagnosed type 1 patients: a reappraisal using pheno- laboratory with either rabbit antibody (RaVWF) or various types, genotypes and molecular modelling. Thromb Haemost 2000; 84: 998–1004. mAbs (CR1, CR2, 5D2, 6G1) used as VWF detection 14 Favaloro EJ, Henniker A, Facey D, Hertzberg M. Discrimination of antibodies. von Willebrand disease (VWD) subtypes: direct comparison of von Figure S3.VWF testing by ELISAwith mAbs as VWF capture Willebrand factor:collagen binding activity/assay (VWF:CBA) with antibodies. monoclonal antibody (MAB) based ELISA VWF-detection systems. Table S1. Detailed summary of test sample set and character- Thromb Haemost 2000; 84: 541–7. 15 Favaloro EJ. Discrimination of von Willebrand Disease (VWD) sub- istics. types: direct comparison of commercial ELISA-based options used to Please note:Wiley-Blackwell are not responsible for the content detect qualitative von Willebrand factor (VWF) defects. Am J Clin Pathol 2000; 114: 608–18. or functionality of any supporting materials supplied by the 16 Ahmad S, Ptashkin B, Digiovanni C, Cines DB, Konkle BA, Cuker A. authors. Any queries (other than missing material) should be False normal von Willebrand factor activity by monoclonal antibody- directed to the corresponding author for the article. based ELISA in a patient with type 2A(IID) von Willebrand disease. Thromb Haemost 2011; 106: 1224–5. 17 DeVleeschauwerA,DevreeseK. Comparison of a new automated von References Willebrand factor activity assay with an aggregation von Willebrand ristocetin cofactor activity assay for the diagnosis of von Willebrand 1 Favaloro EJ. vonWillebrand disease: local diagnosis and management disease. Blood Coagul Fibrinolysis 2006; 17: 353–8. of a globally distributed bleeding disorder. Semin Thromb Hemost 18 Sucker C, Senft B, Scharf RE, Zotz RB. Determination of von 2011; 37: 440–55. Willebrand factor activity: evaluation of the HaemosIL assay in 2 Sadler JE, Budde U, Eikenboom JC, Favaloro EJ, Hill FGH, Holm- comparison with established procedures. Clin Appl Thromb Hemost berg L, Ingerslev J, Lee CA, Lillicrap D, Mannucci PM, Mazurier C, 2006; 12: 305–10. Meyer D, Nichols WL, Nishino M, Peake IR, Rodeghiero F, Sch- 19 Rodgers SE, Lloyd JV, Mangos HM, Duncan EM, McRae SJ. neppenheim R, Ruggeri ZM, Srivastava A, Montgomery RR, et al. Diagnosis and management of adult patients with von Willebrand Update on the pathophysiology and classification of von Willebrand disease in South Australia. Semin Thromb Hemost 2011; 37: 535–41. disease: a report of the Subcommittee on von Willebrand Factor. J 20 Chen D, Tange JI, Meyers BJ, Pruthi RK, Nichols WL, Heit JA. Thromb Haemost 2006; 4: 2103–14. Validation of an automated latex particle-enhanced immunoturbidi- 3 Favaloro EJ. Diagnosis and classification of vonWillebrand disease: a metric von Willebrand factor activity assay. J Thromb Haemost 2011; review of the differential utility of various functional von Willebrand 9: 1993–2002. factor assays. Blood Coagul Fibrinolysis 2011; 22: 553–64. 21 Favaloro EJ, Bonar R, MeiringM, Street A, Marsden K (on behalf of 4 HowardMA, Firkin BG. Ristocetin: a new tool in the investigation of the RCPA QAP in Haematology). 2B or not 2B? Disparate discrimi- platelet aggregation. Thromb Diath Haemorrh 1971; 26: 362–9. nation of functional VWF discordance using different assay panels or 5 Koutts J. A short history of diagnostic tests for von Willebrand dis- methodologies may lead to success or failure in the early identification ease: in memory of Barry Firkin (1930 to 2001) and Ted Zimmerman of Type 2B VWD. Thromb Haemost 2007; 98: 346–58. (1937 to 1988). Semin Thromb Hemost 2006; 32: 445–55. 22 Chandler WL, Peerschke EIB, Castellone DD, Meijer P, on behalf of 6 FavaloroEJ,ThomJ,PattersonD,JustS,DixonT,KouttsJ,BaccalaM, the NASCOLAProficiency Testing Committee. vonWillebrand factor Rowell J, Baker R. Desmopressin therapy to assist the functional assay proficiency testing. The North American specialized coagulation identification and characterisation of von Willebrand disease: differ- laboratory association experience. Am J Clin Pathol 2011; 135: 862–9. ential utility from combining two (VWF:CB and VWF:RCo) von 23 Chen J, Reheman A, Gushiken FC, Nolasco L, Fu X, Moake JL, Ni Willebrand factor activity assays? Thromb Res 2009; 123: 862–8. H, López JA. N-acetylcysteine reduces the size and activity of von 7 Favaloro EJ, Kershaw G, McLachlan AJ, Lloyd J. Time to Willebrand factor in human plasma and mice. J Clin Invest 2011; 121: think outside the box? Proposals for a new approach to future phar- 593–603. macokinetic studies of von Willebrand factor concentrates in 24 Favaloro EJ, Mohammed A, Coombs R, Mehrabani PA. Filtered people with von Willebrand disease. Semin Thromb Hemost 2007; 33: plasma as a potential cause of clinical misdiagnosis: inappropriate 745–58. testing in a haematology laboratory. Br J Biomed Sci 1995; 52: 243–48. 8 Just S. Methodologies and clinical utility of ADAMTS-13 activity 25 FavaloroEJ, Soltani S,McDonald J. Potential laboratorymisdiagnosis testing. Semin Thromb Hemost 2010; 36: 82–90. of haemophilia and von Willebrand disorder due to cold activation of 9 Brown JE, Bosak JO. An ELISA test for the binding of von Wille- blood samples for testing. Am J Clin Pathol 2004; 122: 686–92. brand antigen to collagen. Thromb Res 1986; 43: 303–11. 26 Favaloro EJ, Lippi G, Adcock DM. Preanalytical and postanalytical 10 Goodall AH, Jarvis J, Chand S, Rawlings E, OBrien DP,McGraw A, variables: the leading causes of diagnostic error in haemostasis? Semin Hutton R, Tuddenham EG. An immunoradiometric assay for human Thromb Hemost 2008; 34: 612–34. factor VIII von Willebrand factor (VIII:VWF) using a monoclonal 27 De Luca M, Facey DA, Favaloro EJ, Hertzberg MS, Whisstock JC, antibody that defines a functional epitope. Br J Haematol 1985; 59: McNally T, Andrews RK, Berndt MC. Structure and function of the 565–77. von Willebrand factor A1 domain. Analysis with monoclonal anti- 11 Chand S, McCraw A, Hutton R, Tuddenham EGD, Goodall AH. A bodies reveals distinct binding sites involved in recognition of the two-site, monoclonal antibody based immunoassay for von  2012 International Society on Thrombosis and Haemostasis 186 1054 E1.87J. Favaloro et al platelet membrane glycoprotein Ib–IX–V complex and ristocetin- factor ristocetin cofactor activity assay. J Thromb Haemost 2010; 8: dependent activation. Blood 2000; 95: 164–72. 2842–4. 28 Favaloro EJ, Facey D, Henniker A. Use of a novel Platelet Function 36 Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Baker R, Hertzberg M, Analyser (PFA-100TM) with high sensitivity to disturbances in von Street A, Marsden K (on behalf of the RCPA QAP in Haematology). Willebrand factor to screen for von Willebrands disease and other Reducing errors in identification of von Willebrand disease: the disorders. Am J Hematol 1999; 62: 165–74. experience of the Royal college of Pathologists of Australasia Quality 29 Tan FL, Ginsburg D. What a polyclonal antibody sees in von Wille- Assurance Program. Semin Thromb Hemost 2006; 32: 505–13. brand factor. Thromb Res 2008; 121: 519–26. 37 Turner T, Nolasco L, Moake J. Generation and breakdown of soluble 30 Marques MB, Fritsma GA. von Willebrand disease laboratory diag- ultra-large von Willebrand factor multimers. Semin Thromb Hemost nosis: the saga continues. Am J Clin Pathol 2011; 135: 818–20. 2012; 38: 38–46. 31 Favaloro EJ. von Willebrand factor assay proficiency testing contin- 38 Ganderton T, Wong JW, Schroeder C, Hogg PJ. Lateral self-associ- ued. Am J Clin Pathol 2011; 136: 657–9. ation of VWF involves the Cys2431–Cys2453 disulfide/dithiol in the 32 Favaloro EJ. Evaluation of commercial von Willebrand factor colla- C2 domain. Blood 2011; 118: 5312–18. gen binding assays to assist the discrimination of types 1 and 2 von 39 Mori F, Rossi P, Nardini I, Gambelli D, Farina C. Evaluation of von Willebrand disease. Thromb Haemost 2010; 104: 1009–21. Willebrand factor activity in factor VIII/von Willebrand factor con- 33 Favaloro EJ, Bonar R, Marsden K (on behalf of the RCPA QAP centrates with the automated von Willebrand factor: activity IL test. Haemostasis Committee). Lower limit of assay sensitivity: an under- Blood Coagul Fibrinolysis 2010; 21: 221–8. recognised and significant problem in von Willebrand disease identi- 40 Lee CA, Hubbard A, Sabin A, Budde U, Castaman G, Favaloro fication and classification. Clin Lab Sci 2008; 21: 178–85. EJ, Friedman KD, Mazurier C, Srivastava A, Weinstein M, 34 Hillarp A, StadlerM, Haderer C,Weinberger J, Kessler CM, Romisch Montgomery RR, Lillicrap D, Federici AB. For the working party J. Improved performance characteristics of the von Willebrand factor on VWF assays in VWD diagnosis ISTH-SSC-SC on VWF. Diag- ristocetin cofactor activity assay using a novel automated assay pro- nostic repertoire for laboratory diagnosis of von Willebrand disease: tocol. J Thromb Haemost 2010; 8: 2216–23. results of a blind study in 32 centres worldwide. A Scientific Stan- 35 Favaloro EJ, Mohammed S, McDonald J. Validation of improved dardization Committee Communication. J Thromb Haemost 2011; 9: performance characteristics for the automated von Willebrand 220–2.  2012 International Society on Thrombosis and Haemostasis 187 S-19 188 Thrombosis Research 166 (2018) 96–105 Contents lists available at ScienceDirect Thrombosis Research journal homepage: www.elsevier.com/locate/thromres Full Length Article Differential sensitivity of von Willebrand factor activity assays to reduced T VWF molecular weight forms: A large international cross-laboratory study Emmanuel J. Favaloroa,b,⁎, Roslyn Bonarc, Martine J. Hollestelled, Ian Jenningse, Soma Mohammeda, Piet Meijerd, Timothy Woodse, Muriel Meiringf aHaematology, Institute of Clinical Pathology and Medical Research (ICPMR), NSW Health Pathology, Westmead Hospital, Westmead, NSW, Australia b Sydney Centres for Thrombosis and Haemostasis, Westmead, NSW, Australia c Royal College of Pathology of Australasia Quality Assurance Program (RCPAQAP) Haematology, St Leonards, NSW, Australia d ECAT Foundation (External quality Control for Assays and Tests), Voorschoten, The Netherlands eUK NEQAS for Blood Coagulation, Sheffield, UK fDepartment of Haematology and Cell Biology, University of the Free State and National Health Laboratory Services, Bloemfontein, South Africa A R T I C L E I N F O A B S T R A C T Keywords: Introduction: von Willebrand disease (VWD), the most common inherited bleeding disorder, is due to defi- von Willebrand disease, VWD ciencies/defects in von Willebrand factor (VWF). Effective diagnosis requires testing for FVIII, VWF antigen and von Willebrand factor, VWF one or more VWF ‘activity’ assays. Classically, ‘activity’ is assessed using ristocetin cofactor (VWF:RCo), but High molecular weight sensitivity collagen binding (VWF:CB) and/or other assays are used by many laboratories. This extensive international Laboratory testing cross-laboratory study has specifically evaluated contemporary VWF activity assays for comparative sensitivity VWD diagnosis to reduction in high molecular weight (HMW) VWF, and their ability to differentiate type 1 vs 2A VWD-like samples. Materials and methods: A set of four samples representing step wise reduction in HMW VWF were tested by over 400 laboratories worldwide using various assays. A second set of two samples representing type 1 or type 2A VWD-like plasma was tested by a subset of 251 laboratories. Results: Combined data identified some differences between VWF activity assays, with sensitivity for reduction of HMW being highest for VWF:CB and VWF:GPIbM, intermediate for VWF:RCo and VWF:GPIbR, and lowest for VWF:Ab. ‘Within’ method analysis identified the Stago method as the most sensitive VWF:CB assay. A large variation in inter-laboratory CV (e.g., 7–24% for the normal sample) was also demonstrated for various methods. Although performance of various methods differed significantly, most laboratories correctly differentiated be- tween type 1 and 2 samples, irrespective of VWF activity assay employed. Conclusions: These results hold significant clinical implications for diagnosis and therapy monitoring of VWD, as well as potential future diagnosis and therapy monitoring of thrombotic thrombocytopenic purpura (TTP). 1. Introduction 2A, 2B, 2M and 2N define qualitative disorders [1,2]. Particularly, 2A VWD is characterized by decreased VWF-dependent platelet adhesion von Willebrand disease (VWD) is reportedly the most common in- due to selective deficiency of high-molecular-weight (HMW) VWF herited bleeding disorder, and suspected following clinical and physical multimers, 2B VWD by an increased affinity of VWF for platelet gly- examination in individuals with personal and familial evidence of coprotein Ib (GPIb), often also associated with loss of HMW VWF, 2M mucocutaneous bleeding, as later confirmed by laboratory testing [1,2]. VWD by impaired VWF-dependent platelet adhesion without a selective VWD is classified into six types, representing quantitative or qualitative deficiency of HMW VWF, and 2N VWD by impaired VWF-Factor VIII defects/deficiency in plasma von Willebrand factor (VWF). Types 1 and binding [2]. For type 1 VWD, most guidelines now advise that diagnosis 3 VWD respectively define partial or complete loss of VWF, and types of VWD should be restricted to those with VWF levels below 30 U/dL, Abbreviations: ELISA, enzyme linked immunosorbent assay; GPIb, glycoprotein Ib; HMW, high molecular weight (VWF); NAC, N‑acetycysteine; TTP, thrombotic thrombocytopenic purpura; VWF, von Willebrand factor; VWF:Ab, VWF activity assays based on the binding of a monoclonal antibody (MAB) to a VWF A1 domain epitope; VWF:Ag, VWF antigen (protein level); VWF:CB, VWF collagen binding; VWF:GPIbM, VWF activity assays based on spontaneous binding of VWF to a gain-of-function mutant GPIb fragment; VWF:GPIbR, VWF activity assays based on ristocetin-induced binding of VWF to a recombinant wild type GPIb fragment; VWF:RCo, VWF ristocetin cofactor; VWD, von Willebrand disease ⁎ Corresponding author at: Department of Haematology, Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, Westmead, NSW 2145, Australia. E-mail address: emmanuel.favaloro@health.nsw.gov.au (E.J. Favaloro). https://doi.org/10.1016/j.thromres.2018.04.015 Received 18 December 2017; Received in revised form 22 February 2018; Accepted 16 April 2018 Available online 19 April 2018 0049-3848/ Crown Copyright © 2018 Published by Elsevier Ltd. All rights reserved. 188 189 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 with those having levels above 30, but lower than the normal cut-off These samples comprised the normal plasma pool, and three similar value, be considered as representing ‘low VWF’ as a risk factor for volume aliquots of this pool after treatment to produce step-wise re- bleeding [1,3]. VWD can also be identified as an acquired disorder, duction of HMW VWF multimers, using a propriety process of dis- where it is usually termed acquired von Willebrand syndrome (AVWS). ulphide bond reduction with N‑acetycysteine (NAC), as per a previous Laboratory testing is focused on assessment of plasma factor VIII smaller study [9]. Although such treatment results in reduction of HMW coagulant (FVIII:C) activity, VWF antigen (VWF:Ag), and various VWF VWF multimers, the native VWF remains normally active [9,10]. Given ‘activities’, supplemented as required with other tests on a case by case our desire to purpose generate material with predefined characteristics, basis [1–4]. Classically, VWF ‘activity’ was assessed using ristocetin namely a starting sample with ~100 U/dL VWF:Ag, and because the cofactor (VWF:RCo), originally described in the early 1970s [4–6]. pool initially provided VWF levels closer to 115 U/dL, all samples were However, inherent problems, including complexity, long performance diluted slightly (~15%) with saline. Minor dilution of samples with time, poor reproducibility, and poor sensitivity to low levels of VWF saline or else dilution with plasma deficient in VWF is now a standard (recently reviewed in [4]), has focused attention on newer assays, in- tool for preparation of ‘artificial VWD’ test samples for use in the Royal cluding collagen binding (VWF:CB), and several others based either on College of Pathologists (RCPAQAP) Haematology external quality as- GPIb binding or monoclonal antibody (MAB) technology [4,6]. This has sessment (EQA) practice, and in such experience, does not adversely also seen recent emergence of new proposed assay classifications for the affect lyophilization or selectively affect functional VWF properties. ‘platelet dependent VWF activity’ group of assays [4,7], with the ad- The first sample in the series of four (i.e., ‘V1’; see Table 1) thus re- ditional recommended abbreviations VWF:GPIbR, VWF:GPIbM and presented an essentially normal pool sample for the purpose of this VWF:Ab. These assays may or may not utilize ristocetin, or even pla- exercise, with starting VWF:Ag level of ~100 U/dL, and similar levels of telets, to identify VWF ‘activity’ [4,6,7]. VWF ‘activity’ and FVIII. The second sample (‘V2’) would represent a Used in conjunction with VWF:Ag and FVIII:C, VWF activity assays minor loss of HMW VWF multimers, which might occur with a normal have the capacity to identify and discriminate various types of VWD. In sample subsequent to a pre-analytical event (e.g., refrigeration of whole particular, low levels of VWF with concordance of antigen and ‘activity’ blood sample; provision of serum instead of citrate plasma; provision of suggests a type 1 VWD (or ‘low VWF’), but discordance of VWF ‘activity’ clotted plasma sample) [11]. The last two samples (i.e., ‘V3’, ‘V4’) to antigen (usually expressed as a ratio below 0.6) suggests a type 2 would show increasing reduction of HMW VWF multimers and then VWD. Also relevant here is the increasing recognition of AVWS [1], intermediate molecular weight (IMW) VWF multimers, potentially re- especially those associated with cardiac defects or cardiac assist de- flective of type 2A or 2B VWD-like plasma, but with a relative high VWF vices, where loss of HMW VWF is associated with increased risk of level, which is a pattern that might occur in type 2A or 2B VWD during bleeding [8]. pregnancy [12], or in AVWS associated with cardiac defects/assist de- Nevertheless, the comparative behavior of different VWF ‘activity’ vices [8]. The main aim of the ‘Part 1’ study was to identify the com- assays is not well recognized, either in diagnosis of VWD or for relative parative sensitivity of different VWF activity assays to identify the sensitivity to HMW VWF multimers. In the current study, we have (sequential) reduction in HMW VWF multimers in these samples, with evaluated the comparative sensitivity of different VWF ‘activity’ assays associated inference on VWD diagnostics. to reduction in HMW VWF multimers, as well as for relative utility to Secondly, ‘Part 2’ comprised cross-laboratory testing of two addi- discriminate type 1 vs type 2 VWD-like samples. This study represents a tional plasma samples, also derived subsequent to the above described large international cross-laboratory investigation comprising over 400 process. Both samples were purpose constructed to generate samples laboratories distributed all over the world. with similar levels (~25–30 U/dL) of VWF:Ag, but one sample was engineered to behave like a type 1 VWD sample (VWD-1 ‘mimic’), with 2. Material & methods VWF activity remaining concordant to VWF:Ag, and with normal VWF activity/antigen ratios (i.e., > 0.7; identified as ‘V5’ in Table 1), whilst 2.1. Study aims and initial preparation and testing of study samples the second sample was engineered to behave like a type 2A VWD sample (VWD-2A ‘mimic’), with VWF activity discordant to VWF:Ag, The study sample set as prepared and later dispatched to labora- and low VWF activity/antigen ratios (i.e., < 0.5; identified as ‘V6’ in tories is summarized in Table 1. The samples were produced by the lead Table 1). These samples were also produced in a similar manner to author (EJF) in the laboratory of Haematology (Institute of Clinical those of the earlier smaller study [9]. There were several aims to the Pathology and Medical Research, ICPMR, Westmead), primarily fol- ‘Part 2’ study, including: (i) identifying the differential utility of diverse lowing treatment of a plasma pool of normal individual plasmas. There VWF activity assays to identify type 1 vs 2 VWD test patterns; (ii) were two parts to this study. In study ‘Part 1’, four samples (‘V1’ to ‘V4’; whether artificially generated samples could be utilized in cross-la- Table 1) were prepared to reflect sequential reduction of HMW VWF boratory studies (e.g., for EQA purposes) instead of native patient multimers, whilst aiming to retain levels of VWF relatively constant. samples, especially given ethical and practical difficulties in obtaining Table 1 Summary of test sample set and characteristics. Sample ID Sample comprised Sample intended to, or could feasibly, represent V1 Pool of normal plasmas Normal plasma. VWF:Ag level of ~100 U/dL, and similar levels of VWF ‘activity’ and FVIII. V2 Sample V1 treated to yield minor loss of HMW VWF Normal plasma with minor loss of HMW VWF as might be caused by a pre-analytical event (see main text) V3 Sample V1 treated to yield greater reduction of HMW VWF Type 2A (or 2B) VWD like plasma as might be obtained in pregnancy. V4 Sample V1 treated to yield high reduction of HMW VWF Type 2A (or 2B) VWD like plasma as might be obtained in pregnancy. V5 Mixture of V1 and VWF deficient plasma Moderate/mild type 1 VWD like plasma with target VWF:Ag 25–30 U/dL and normal VWF:Activity/Ag ratio (> 0.7) V6 Mixture of HMW VWF deficient sample and VWF deficient Type 2A/2B VWD like plasma with target VWF:Ag 25–30 U/dL and low VWF:Activity/Ag ratio (< 0.5) plasma Sample V1 represents a normal pool plasma and the main plasma used to then generate subsequent test samples. Samples V2, V3, V4 were derived from sample V1 and reflect sequentially increasing reduction of HMW/IMW VWF. Samples V5 and V6 were respectively derived from mixtures of VWF deficient plasma and V1 or a HMW VWF deficient sample to produce samples reflective of moderate type 1 VWD and 2A/2B VWD, but with a similar level of VWF:Ag. Abbreviations: HMW, high molecular weight; IMW, intermediate molecular weight; VWD, von Willebrand disease; VWF, von Willebrand factor. 97 189 190 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 Table 2 Summary test statistics and breakdown of main study methods used by study participants.a Test/methodology No. of study resultsb % of totalc % of methodd Mean values in U/dL (CV; %) Mean low reference value in U/dL (CV; %)e V1 V2 V3 V4 VWF:Ag 406 100.0 100.0 97.1 (8.0) 92.7 (9.8) 74.4 (13.8) 92.9 (19.6) 51.5 (13.0) LIA 328 80.8 80.8 97.3 (6.9) 92.0 (7.4) 74.2 (11.9) 92.7 (18.5) 51.8 (12.3) ELISA (E) 30 7.4 7.4 96.6 (14.4) 96.2 (17.5) 73.6 (16.0) 90.5 (25.4) 48.3 (14.4) VWF:RCo 222 54.7 100.0 74.3 (16.7) 63.4 (18.1) 27.3 (30.1) 15.8 (> 40) 51.2 (11.6) Aggregometer (Agg) 63 15.5 28.4 71.4 (20.7) 63.2 (23.8) 30.2 (30.9) 13.7 (> 40) 51.1 (13.0) Automated (Auto) 88 21.7 39.6 71.1 (14.4) 61.5 (14.5) 24.1 (32.3) 15.1 (> 40) 51.4 (10.9) VWF:GPIbR 53 13.1 23.9 83.4 (9.9) 67.3 (10.5) 28.6 (17.0) 19.3 (26.9) 50.6 (11.9) VWF:CB 128 31.5 100.0 89.2 (18.2) 76.5 (21.2) 29.0 (> 40) 11.0 (> 40) 51.2 (14.7) In-House 14 3.4 10.9 85.4 (19.2) 72.1 (21.7) 24.1 (> 40) 8.8 (> 40) 48.5 (19.9) Stago 21 5.2 16.4 90.4 (24.4) 66.6 (15.1) 17.2 (27.3) 6.9 (> 40) 53.7 (12.6) Technoclone 41 10.1 32.0 94.1 (13.8) 86.7 (15.4) 37.4 (22.7) 16.5 (37.0) 49.7 (13.6) VWF activity other VWF:GPIbM 87 21.4 100.0 74.1 (8.1) 62.3 (8.3) 24.9 (27.0) 11.5 (33.5) 49.3 (15.8) VWF:Ab 85 20.9 100.0 80.2 (13.6) 66.7 (13.9) 35.2 (28.0) 25.0 (> 40) 50.3 (15.6) Multimers 36 8.9 100.0 NA NA NA NA Abbreviations: CV, coefficient of variation; VWF:Ag, von Willebrand factor antigen; LIA, latex immuno assay; ELISA, enzyme linked immunosorbant assay; N/A, not applicable; VWF:CB, von Willebrand factor collagen binding; VWF:RCo, von Willebrand factor ristocetin cofactor. VWF:GPIbR, VWF:GPIbM and VWF:Ab refer to methodologies as recently proposed abbreviations from the VWF ISTH SSC [7] – refer to Material & methods text for additional details. a Summary data for the ‘part 1’ study, comprising data from all participants. b From all participants; all provided test data for VWF:Ag, mostly as performed using LIA, with smaller group using ELISA (E) and other/non specified methods (data not shown). Thereafter, varied tests for VWF ‘activity’ performed by various laboratories as shown. c Percent of total participants reporting results for respective test group. d Percent of participants reporting results within a specific methodology group (e.g., for VWF:Ag, or VWF:RCo, or VWF:CB). VWF:GPIbM, VWF:Ab and Multimers are considered as individual overall methods (i.e. = 100% of that method group). e Mean of study participant lower reference range value (=cut-off for normal/abnormal), and inter-laboratory CVs for these values. large quantities of diverse patient plasma for such purpose. Choices offered included normal/not VWD, VWD type 1 (mild, mod- The entire set of six samples (i.e., V1–V6; Table 1) were lyophilized erate, severe or unspecified), VWD type 2 (2A, 2B, 2M, 2N or un- in vials in 1.0 mL volumes, and thereafter stored refrigerated at 4 °C, defined), VWD type 3, or ‘other interpretation’ (with a request to ac- prior to sending to three collaborative EQA organizations (as anon- cordingly specify). Some of these choices (e.g., type 2N VWD, type 3 ymized, sequence relabeled sets), for further distribution to participants VWD) did not reflect any of the samples, but a complete range of of the cross-laboratory study. Lyophilized samples passed internal possible interpretations was provided to avoid any biasing of returned homogeneity and stability testing prior to use in the study (data not data. The various type 2 choices were largely aimed to identify parti- shown). However, it was noted that sample ‘V3’ yielded an un- cipant responses for samples V3, V4 and V6, with ‘type 2 undefined’ expectedly lower level of VWF:Ag than samples V1, V2 and V4. We permitted because respective identification of type 2 ‘subtypes’ would cannot explain this finding, although it is feasible that during produc- require additional investigations to those reported in this study (e.g., tion, sample ‘V3’ was inadvertently double diluted with saline to ~30% ristocetin induced platelet aggregation to differentiate 2A and 2B (i.e., 2×~15%). If this is the explanation, then this would not ex- VWD). The various type 1 choices were largely aimed to identify par- pectedly affect the main study aim, being the comparative sensitivity of ticipant responses for sample V5, and to identify laboratory perceptions VWF activity assays to reduction in HMW VWF or even activity/Ag around VWD ‘severity’. ratios, since all test analytes would be similarly ‘diluted’. Moreover: (a) all samples were tested blind by participants, using a relabeled se- 2.3. Data/statistical analysis quence of vials, and who did not know that samples V1 to V4 were ‘related’; (b) all samples were also interpreted as discrete samples, and Numerical data was largely analyzed by way of comparative med- so the interpretations for sample V3 would be for the sample as pre- ians, means, ranges and inter-assay (inter-laboratory) CVs, using sented, and thus still valid; (c) there would be no expected adverse GraphPad Prism (GraphPad Software Inc., La Jolla, CA 92037 USA; outcome in terms of coefficients of variation (CVs). [www.graphpad.com]). When reported, statistical comparisons reflect non-parametric assessment using the Mann-Whitney test and a two- 2.2. Cross-laboratory study tailed analysis, with a p-value<0.05 being considered statistically significant. Laboratory numerical data was used as reported by parti- All laboratories performing VWF testing and currently enrolled in cipants, except that values reported as ‘< ’ a given value were adjusted any of the noted EQA programs (i.e., RCPAQAP Haematology; UK downwards to permit numerical analysis (e.g., an assay value of< 10 NEQAS (National External Quality Assessment Schemes) coagulation; was adjusted to 9, and an assay ratio< 0.2 was adjusted to 0.1). This is and ECAT (External quality Control for Assays and Tests)) were invited consistent with EQA practice of some programs. These events generally to participate in this cross-laboratory study. Participants of RCPA occurred only for samples with high loss of HMW VWF multimers. Data (n=68) and ECAT (n=197) tested all six samples (V1–V6); partici- for FVIII has been excluded from this report, which has focused on VWF pants of NEQAS (n=163) tested only the four samples (V1–V4) com- activity assays. prising study Part 1. All invited laboratories received vials of the Although different laboratories apply different normal references anonymized samples, together with sample reconstitution and test in- ranges to VWF tests, values above 50 U/dL are generally considered as structions, essentially requesting laboratories to perform their normal ‘normal’, and so this ‘cut-off’ is used in this report as a general guide. test panels for investigation of VWD. Laboratories were also asked for Similarly, laboratories apply different ranges to VWF activity to antigen details of test methodology and to attempt to interpret their test results. ratios; however, values above 0.7 are generally considered as ‘normal’, 98 190 191 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 99 191 Fig. 1. Main study (‘Part 1’) data as derived from samples V1–V4, as tested by all participating laboratories and shown as a box plot (median, with 25th and 75th percentile identified by the box, and 10th and 90th percentile identified by the error bar), for assay groups as per Table 2. VWF test indices on left portion of each figure with values on left y-axis and the horizontal dashed line representing a nominal ‘normal/abnormal’ cut- off value of 50 U/dL. Assay ratios shown on the right portion of each figure, with horizontal dashed lines at 0.5 and 0.7 representing nominal ‘abnormal’ and ‘normal’ cut-off values. Panels A–D respectively show data for samples V1–V4 (see Table 1). 192 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 100 192 Fig. 2. Main study (‘Part 1’) data as derived from samples V1-V4, as tested by all participating laboratories and shown as plots of mean data (with 95% confidence intervals), for assay groups as per Table 2. VWF test indices on left portion of each figure with values on left y-axis and the horizontal dashed line representing a nominal ‘normal/abnormal’ cut-off value of 50 U/dL. Assay ratios shown on the right portion of each figure, with horizontal dashed lines at 0.5 and 0.7 representing nominal ‘abnormal’ and ‘normal’ cut-off values. Panels A–D respectively show data for all assay ratios to reflect spread of all data (A), main VWF activity groups (B), GPIb binding assays (C), VWF:CB (D). 193 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 Fig. 3. Inter-assay variation as co-efficient of variation (CV; %) for main (Part 1) study data for sample V1 for different assays and assay ratios. values below 0.5 are generally considered as ‘abnormal’, and values as having VWD earlier, and overall more consistently, should the uti- between 0.5 and 0.7 are variably assigned to either category depending lized assay measure lower VWF activity levels compared to an assay on the laboratory and the VWF activity assay [4]. that measures much higher levels in the same sample, and having a similar reference interval. Of interest, all assay reference range lower 2.4. Additional notes on ‘VWF activity assays’ and nomenclature cut-off values had similar mean values (all close to 50 U/dL; Table 2). Similarly, VWF activity to antigen ratios also fell sequentially in sam- For the purposes of this study, we have essentially followed the ples V1 to V4, being essentially within normal limits for V1, but ab- recommended nomenclature and abbreviations of the VWF ISTH SSC, normal for sample V4 (Table 2; Figs. 1 and 2). Better detail around VWF as previously published [7,13] and recently reviewed [4], in order to activity to antigen ratios is provided in Fig. 2, where data has been compare study outcomes. We have also assessed for differential sensi- separated according to VWF activity assay. Although all VWF activity/ tivity for major ‘sub-methodologies’, essentially as summarized in Ag ratios fell sequentially according to reduction in HMW VWF, there Table 2. We have included in the overall ‘VWF:RCo’ group, all meth- was notable variability in sensitivity according to VWF activity assay. odologies that employ ristocetin, whether or not employing platelets, More specifically, sensitivity for reduction of HMW was overall highest but also show data separately for methods employing platelets and for VWF:CB, as compared with VWF:Ab (p < 0.0001), VWF:GPIbR tested by aggregometer (‘Agg’) or automated instrument (‘Auto’), as (p < 0.0001), VWF:RCo (p < 0.0001), but was similar to VWF:GPIbM well as ristocetin based methods not employing platelets (p= 0.053). Moreover, ‘within’ method analysis identified better HMW (‘VWF:GPIbR’). For VWF:CB, we have assessed ‘all VWF:CB data’, as VWF sensitivity of the Stago VWF:CB assay, which yielded for sample well as the three main ‘sub-methodologies’ in use (namely, ‘in-house’, V4 the lowest values compared to other separate methods (e.g., Diagnostica Stago, and Technoclone), although also recognizing that p < 0.0001 for comparison with Technoclone). ‘in-house’ VWF:CB assays are quite heterogeneous in methodology. Fi- Assay variation is also identified (as inter-laboratory/inter-method nally, we have also separately analyzed the other major VWF activity CVs) in Table 2 (summarized for all samples V1–V4) and Fig. 3 (for assays (VWF:GPIbM and VWF:Ab [4,6,7]). sample V1). In brief, CVs ranged from 7 to 24% for the normal sample (V1), but not unexpectedly increased as the VWF activity level fell 3. Results (Table 2). The lowest CVs (below 12.5%) were observed for VWF:Ag (LIA), VWF:GPIbM and VWF:GPIbR methods. The main study data is summarized in Table 2 and main ndings for Summary test results for the type 1 and 2A VWD-mimic samplesfi both study Parts 1 and 2 shown in Figs. 1–4. Sample VWF multimer (respectively V5 and V6; Table 1) are shown in Fig. 4. As expected, both patterns are shown in Fig. 5, together with participant interpretations samples yielded similar VWF:Ag test results (~25–30 U/dL). However, for VWF multimer patterns for all test samples, and nal participant whilst V5 expressed similar VWF activity levels, and thus normal VWFfi interpretations for all test samples. Densitometry patterns for VWF activity/antigen ratios, consistent with a type 1 sample, sample V6 multimers were as expected (data not shown). As expected, normal test yielded discordant VWF activity levels, and thus abnormal VWF ac- results were generally reported for sample V1 for all VWF tests, and for tivity/Ag ratios, consistent with a type 2 sample. VWF:Ag for all samples in Study Part 1 (i.e., V1, V2, V3, V4; Table 2 and Interpretations for VWF multimer patterns is shown in Fig. 5. As Fig. 1). However, VWF activity assay values for the ‘Study Part 1’ expected, most participants identified sample V1 as expressing normal samples showed sequential reductions, in line with sequential reduction multimers, and samples V3, V4 and V6 as expressing ‘abnormal mul- in HMW VWF multimers (Table 2; Figs. 1 and 2). Relative sensitivity of timer patterns’ or ‘type 2A/2M/2B/2U’ VWD (NB: U for 'undefined'). the various VWF activity assays for the reduction in HMW VWF mul- Furthermore, most participants identified sample V5 as expressing timers can be compared by determining which assay measured the ‘normal multimer patterns’ or type 1 VWD. lowest VWF activity level in the samples with the highest reduction in Lastly, final interpretations for all test samples is also shown in HMW VWF (i.e., V4). Of relevance, a true patient sample will be de ned Fig. 5. As expected, most participants identified sample V1 as normal,fi 101 193 194 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 Fig. 4. Summary of ‘Part 2’ study data as derived from samples V5 (A) and V6 (B), as tested by participating laboratories and shown as a box plot (median, with 25th and 75th percentile identified by the box, and 10th and 90th percentile identified by the error bar), for main VWF assay groups. VWF test indices on left portion of each figure with values on left y-axis and assay ratios shown on the right portion of each figure, where horizontal dashed lines at 0.5 and 0.7 represent nominal ‘abnormal’ and ‘normal’ cut-off values. and samples V3, V4 and V6 as ‘type 2A/2M/2B/2U’ VWD. Most parti- activity, namely collagen (sub-endothelial matrix) binding. A plethora cipants identified sample V5 as type 1 VWD, with perceived ‘severity’ of new VWF ‘activity’ assays have emerged [4,6,7], many of which also sub-analysis also reported by some participants. reflect GPIb binding, but may or may not involve ristocetin, or even platelets. Irrespective, laboratories are now utilizing a wide variety of VWF assays, and appreciation of relative behaviors is lacking. 4. Discussion Some of us previously reported on a similar comparative assessment with fewer study participants, a single EQA provider, and the VWF To our knowledge, this represents the largest comparative multi- activity assays as then available [9]. The current study in part has si- center assessment of the most widely used VWF ‘activity’ assays for milar design, but a much larger global distribution (N=409 partici- their relative sensitivity to reduction of HMW VWF multimers. pants: 67.7% Europe, 10.5% North America and 17.4% Asia/Aus- VWF:RCo, the first functional VWF assay described in the early 1970s tralasia) and an extensive panel of evaluated VWF activity assays, as [5,6], reflects a VWF - GPIb binding assay, albeit facilitated by risto- contemporarily available. Similar to before [9], these assays show cetin. VWF:CB, first described in 1986 [14], reflects a different VWF 102 194 195 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 Fig. 5. A. Representative VWF multimer patterns obtained with study samples, and confirming sequential reduction in HMW and IMW VWF multimers in generated samples from V1 to V4, as well as type 1 VWD pattern for V5, and type 2 VWD pattern for V6. VWF multimers shown with light (V1–V6) and dark (V5, V6) gel exposure. Arising densitometry patterns were consistent with shown multimers (data not shown). B. Participant interpretations for VWF multimers, expressed as a percentage of returned responses. Importantly, most participants identified samples V1 and V5 as expressing normal multimer distribution, and samples V3, V4, and V6 as expressing abnormal multimer distribution (or being type 2 (2A, 2B, 2M or 2U) VWD). However, occasional participants identified the multimer pattern of the normal sample V1 and the type 1 VWD sample as being ‘abnormal’ or ‘equivocal’, and some participants identified the multimer pattern of the HMW VWF deficient samples V3, V4 and V6 as being ‘normal’. Sample V2 yielded a mixture of interpretations, with some laboratories identifying abnormal multimer patterns, but most reporting normal multimer patterns. C. Participant ‘final interpretations’ expressed as a percentage of returned responses. Importantly, most participants identi- fied sample V1 as being normal/not VWD; V3, V4 and V6 as being type 2 (2A, 2B, 2M or 2U) VWD; V5 as being type 1 VWD. However, occasional participants identified HMW VWF deficient samples V3 and V4 as normal, the type 1 VWD mimic sample (V5) as type 2 (2A, 2B, 2M or 2U) VWD, and the HMW deficient sample/type 2A VWD mimic (V6) as a type 1 VWD. More detailed information for V5 is shown in the far-right portion of the figure (S - severe, Mod – moderate, NS – not specified). 103 195 196 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 different sensitivity to reduction in HMW VWFmultimers, with VWF:CB VWD samples reported similar findings to ours [26]. However, our still ‘maintaining’ greatest sensitivity, and VWF:Ab showing least sen- participants reflect heterogeneous expertise in VWD diagnostics, in- sitivity. The current study further identifies the lower sensitivity of dicative of standard diagnostic practice, whereas the study of Lee et al. VWF:GPIbR and intermediate sensitivity with VWF:RCo and generally comprised ‘expert/reference’ laboratories [26]. Still, a follow VWF:GPIbM, as well as method ‘subgroup’ differences. For VWF:CB, the up study reflecting ‘expert/reference laboratories’ with contemporary most popular method (Technoclone), incongruously showed highest VWF activity assays may be useful to confirm our findings. Also, a assay variation (Table 2) and least HMW VWF sensitivity (Fig. 2D), commutability study co-assessing artificial samples with genuine VWD which replicates our previous conclusion [9], and is consistent with samples could be performed, albeit on a smaller scale to manage patient prior evaluation of commercial VWF:CB assays [15]. Nevertheless, this donations. Irrespective, participant interpretations in our study were relatively ‘insensitive’ VWF:CB assay still appeared more sensitive than largely consistent with sample type, and error-rates not unlike those VWF:Ab (Table 2; Figs. 1 and 2). Notably, overall patterns of HMW previously reported in similar cross laboratory evaluations using gen- VWF sensitivity data in this study are similar to that of genuine VWD uine VWD samples [16–19,26]. samples [16]. In conclusion, we report a large international cross-laboratory Our data also demonstrates large variations in CVs among activity evaluation of VWF testing using different VWF assays, and test samples methods, with VWF:GPIbM and VWF:GPIbR assays showing lowest CVs, showing sequential reduction in HMW VWF multimers, as well as and CVs being highest for VWF:RCo and VWF:CB methods. For samples reflecting type 1 or 2A VWD patterns. VWF ‘activity’ assays VWF:RCo, high CVs are well recognized, in part reflect the variability of showed differential sensitivity to HMW VWF multimer reduction, even ‘submethods’, and potential non-automation, also helping to drive the within the same assay category (e.g., VWF:CB). Furthermore, assay CVs development of other assays such as VWF:GPIbM and VWF:GPIbR. For ranged from 7 to 24% in the normal pooled plasma, demonstrating VWF:CB, relative high CVs may be due to use of ELISA technology, and imprecision among different methods. Lastly, all VWF ‘activity’ assays manual and semi-automated processes. In contrast, VWF:GPIbM and contributed to differentiation of type 1 and 2 VWD patterns and enabled VWF:GPIbR both reflect highly automated assays, with homogeneous effective identification of reduction in HMW VWF multimers in the type reagents. In total, such data is also comparable to previous reported 2 sample. We believe these findings hold significance for clinical values derived from various EQA reports, including genuine VWD practice, including diagnosis and therapy management for VWD (con- samples [16–19]. As heterogeneity of VWD requires performance of genital and acquired) and TTP. Findings might also be relevant to multiple diagnostic tests [1–4,6,7,20], since no single test completely identification of AVWS, particularly as associated with cardiac ab- defines VWD, the current study will help laboratories and clinicians normalities and/or implantable devices [8,27], especially given that better understand respective performance specifications of the various identification of HMW loss in these cases may be difficult with some assays used in daily practice. assays. Participants usually interpreted test results appropriately (Fig. 5), although, on occasion, interpretations were at odds with sample type. Disclosure Notably, HMW VWF multimer reduced samples V3 and V4 were occa- sionally identified as ‘normal/not VWD’, and samples V5 and V6 as The authors state that they have no interest that might be perceived ‘type 2’ and ‘type 1’ VWD respectively. Such ‘errors’ in interpretation, as as posing a conflict or bias. related to laboratory identification/exclusion of VWD have been pre- viously reported [16–19], inclusive of genuine VWD samples, and are Acknowledgements consistent with occasional ‘misdiagnoses’ of VWD. These events may be due to transcription error, limitations in VWF test panels employed, The authors would like to thank the following individuals: Jane and/or misinterpretations of test data (which is instead consistent with McDonald, Monica Ahuja and Ella Grezchnik, from the ICPMR labora- sample type). tory, Westmead, for performing some of the host laboratory testing The current study, like the earlier small study [9], has utilized a reported in this study; Aletta Veninga is acknowledged for her excellent propriety process that employed N‑acetycysteine (NAC) to achieve step- data processing for ECAT; finally, all participant laboratories and staff wise reduction in HMW VWF multimers. This in vitro process may that otherwise contributed data for this report. mimic a natural in vivo process, described but incompletely char- acterized, that permits reduction of VWF in the absence of ADAMTS-13 Author contributions [21], and also possibly assists formation of VWF complexes at a thrombus by a process of self-association [22]. NAC and ‘ADAMTS-13 EJF: conceived, designed and coordinated the study, prepared study alternative’ in vivo VWF processing activities may also have potential samples, undertook preliminary testing of samples, participated in therapeutic benefit in thrombotic thrombocytopenic purpura (TTP), method validation, undertook data analysis and wrote the original with clinical trials in process [23–25]. Should NAC or ‘analogous’ manuscript. RB: responsible for arranging and overseeing the logistics therapies become utilized in TTP, then laboratories may need to of the study from the RCPAQAP, arranged for sample lyophilizing, monitor such treatment, and the current study may provide additional contributed to data analysis. SM: Co-ordinated sample testing at the context for this. ICPMR. MJH, PM, IJ, TW: contributed to data analysis and responsible Naturally, our study has several limitations. First, all assessed for arranging and overseeing the logistics of the study from the per- samples were artificially generated and did not reflect genuine VWD spective of ECAT/NEQAS. All co-authors: contributed to manuscript patient samples. Nevertheless, study findings reflect positively and revision and have approved the final manuscript for publication. suggest such samples can be used as ‘VWD-mimics’, at least for EQA and cross-laboratory evaluations. Study findings are also consistent with References previous reports, including observations with genuine cases of VWD (e.g., [9,15–19]). Importantly, logistical and ethical barriers hinder [1] M.A. Laffan, W. Lester, J.S. O'Donnell, A. Will, R.C. Tait, A. Goodeve, C.M. Millar, sample utilization from individuals with specific VWD types for large D.M. Keeling, The diagnosis and management of von Willebrand disease: a United Kingdom Haemophilia Centre Doctors Organization guideline approved by the cross-laboratory studies. Moreover, specifically engineered material can British Committee for Standards in Haematology, Br. J. Haematol. 167 (2014) be purpose generated (e.g., samples V5 and V6 had similar low levels of 453–465. VWF:Ag, but differed according to VWF activity levels), and this would [2] J.E. Sadler, U. Budde, J.C. Eikenboom, E.J. Favaloro, F.G.H. Hill, L. Holmberg, assist to control for interpretation errors based on di ering VWF:Ag J. Ingerslev, C.A. Lee, D. Lillicrap, P.M. Mannucci, C. Mazurier, D. Meyer,ff W.L. Nichols, M. Nishino, I.R. Peake, F. Rodeghiero, R. Schneppenheim, levels. Furthermore, a recent multi-laboratory study utilizing genuine 104 196 197 E.J. Favaloro et al. Thrombosis Research 166 (2018) 96–105 Z.M. Ruggeri, A. Srivastava, R.R. Montgomery, A.B. Federici, the Working Party on assays to assist the discrimination of types 1 and 2 von Willebrand disease, Thromb. von Willebrand Disease Classification, Update on the pathophysiology and classi- Haemost. 104 (2010) 1009–1021. fication of von Willebrand disease: a report of the Subcommittee on von Willebrand [16] E.J. Favaloro, R.A. Bonar, M. Meiring, E. Duncan, S. Mohammed, J. Sioufi, Factor, J. Thromb. Haemost. 4 (2006) 2103–2114. K. Marsden, Evaluating errors in the laboratory identification of von Willebrand [3] W.L. Nichols, M.B. Hultin, A.H. James, et al., von Willebrand disease (VWD): evi- disease in the real world, Thromb. Res. 134 (2014) 393–403. dence-based diagnosis and management guidelines, the National Heart, Lung, and [17] P. Meijer, F. Haverkate, An external quality assessment program for von Willebrand Blood Institute (NHLBI) Expert Panel report (USA), Haemophilia 14 (2008) factor laboratory analysis: an overview from the European concerted action on 171–232. thrombosis and disabilities foundation, Semin. Thromb. Hemost. 32 (2006) [4] E.J. Favaloro, L. Pasalic, J. Curnow, Laboratory tests used to help diagnose von 485–491. Willebrand disease: an update, Pathology 48 (2016) 303–318. [18] W.L. Chandler, E.I.B. Peerschke, D.D. Castellone, P. Meijer, on behalf of the [5] M.A. Howard, B.G. Firkin, Ristocetin: a new tool in the investigation of platelet NASCOLA Proficiency Testing Committee. von Willebrand Factor Assay Proficiency aggregation, Thromb. Diath. Haemorrh. 26 (1971) 362–369. Testing, The North American specialized coagulation laboratory association ex- [6] S. Just, Laboratory testing for von Willebrand disease: the past, present, and future perience, Am. J. Clin. Pathol. 135 (2011) 862–869. state of play for von Willebrand factor assays that measure platelet binding activity, [19] S. Kitchen, I. Jennings, T.A. Woods, D.P. Kitchen, I.D. Walker, F.E. Preston, with or without ristocetin, Semin. Thromb. Hemost. 43 (2017) 75–91. Laboratory tests for measurement of von Willebrand factor show poor agreement [7] I. Bodo, J. Eikenboom, R. Montgomery, J. Patzke, R. Schneppenheim, J. Di Paola, among different centers: results from the UK NEQA Scheme for Blood Coagulation, on behalf of Subcommittee on von Willebrand Factor, Platelet dependent von Semin. Thromb. Hemost. 32 (2006) 492–498. Willebrand factor activity. Nomenclature and methodology: communication from [20] A. De Jong, J. Eikenboom, Developments in the diagnostic procedures for von the SSC of the ISTH, J. Thromb. Haemost. 13 (2015) 1345–1350. Willebrand disease, J. Thromb. Haemost. 14 (2016) 449–460. [8] A. Nascimbene, S. Neelamegham, O.H. Frazier, J.L. Moake, J.F. Dong, Acquired von [21] T. Turner, L. Nolasco, J. Moake, Generation and breakdown of soluble ultra-large Willebrand syndrome associated with left ventricular assist device, Blood 127 von Willebrand factor multimers, Semin. Thromb. Hemost. 38 (2012) 38–46. (2016) 3133–3141. [22] T. Ganderton, J.W. Wong, C. Schroeder, P.J. Hogg, Lateral self-association of VWF [9] E.J. Favaloro, R. Bonar, K. Chapman, M. Meiring, D.F. Adcock, Differential sensi- involves the Cys2431-Cys2453 disulfide/dithiol in the C2 domain, Blood 118 tivity of von Willebrand factor ‘activity’ assays to large and small VWF molecular (2011) 5312–5418. weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen [23] G. Cabanillas, A. Popescu-Martinez, N‑acetylcysteine for relapsing thrombotic binding and monoclonal antibody based assays, J. Thromb. Haemost. 10 (2012) thrombocytopenic purpura: more evidence of a promising drug, Am. J. Ther. 23 1043–1054. (2016) e1277–9. [10] J. Chen, A. Reheman, F.C. Gushiken, L. Nolasco, X. Fu, J.L. Moake, H. Ni, [24] A. Rottenstreich, S. Hochberg-Klein, D. Rund, Y. Kalish, The role of N‑acetylcysteine J.A. López, N‑acetylcysteine reduces the size and activity of von Willebrand factor in the treatment of thrombotic thrombocytopenic purpura, J. Thromb. in human plasma and mice, J. Clin. Invest. 121 (2011) 593–603. Thrombolysis 41 (2016) 678–683. [11] E.J. Favaloro, G. Lippi, Pre-analytical issues that may cause misdiagnosis in hae- [25] G.W. Li, S. Rambally, J. Kamboj, S. Reilly, J.L. Moake, M.M. Udden, M.P. Mims, mophilia and von Willebrand disease, Haemophilia 24 (2) (2018 Mar) 198–210. Treatment of refractory thrombotic thrombocytopenic purpura with [12] G. Castaman, Changes of von Willebrand factor during pregnancy in women with N‑acetylcysteine: a case report, Transfusion 54 (2014) 1221–1224. and without von Willebrand disease, Mediterr. J. Hematol. Infect. Dis. 5 (2013) [26] C.A. Lee, A. Hubbard, A. Sabin, G. Castaman, E.J. Favaloro, K.D. Friedman, e2013052. C. Mazurier, A. Srivastava, M. Weinstein, R.R. Montgomery, D. Lillicrap, [13] C. Mazurier, F. Rodeghiero, von Willebrand Factor Subcommittee of the Scientific A.B. Federici, For the Working Party on VWF assays in VWD diagnosis ISTH-SSC-SC and Standardization Committee of the International Society on Thrombosis and on VWF, Diagnostic repertoire for laboratory diagnosis of von Willebrand disease: Haemostasis, Recommended abbreviations for von Willebrand Factor and its ac- results of a blind study in 32 centres worldwide. A scientific standardization com- tivities, Thromb. Haemost. 86 (2001) 712. mittee communication, J. Thromb. Haemost. 9 (2011) 220–222. [14] J.E. Brown, J.O. Bosak, An ELISA test for the binding of von Willebrand antigen to [27] H.C. Kwaan, Complications of implanted nonbiologic devices-an overview, Semin. collagen, Thromb. Res. 43 (1986) 303–311. Thromb. Hemost. 44 (2018) 7–11. [15] E.J. Favaloro, Evaluation of commercial von Willebrand factor collagen binding 105 197 198 OPEN ACCESS Medical Technology SA | Volume 31 No. 1 | June 2017 S-20 Peer reviewed ORIGINAL ARTICLE Evaluation of a cost-EffEctivE aDaMts13 antigEn assay M Meiring PhD | S Myneni B.Sc Hons Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, South Africa Corresponding author: Muriel Meiring | tel: +27 51 405 3593 | email: meiringsm@ufs.ac.za ABSTRACT Thrombotic thrombocytopenic purpura (TTP) is life-threatening and is characterised by platelet deposition in the microvasculature with thrombus formation in particular organs. This results in thrombocytopenia, microangiopathic haemolytic anaemia, kidney failure and neurological symptoms. It is a rare disorder, but can occur in patients infected with the human immunodeficiency virus (HIV). ADAMTS13 is the 13th member of a disintegrin and metalloprotease with thrombospondin type motifs. Its deficiency causes TTP. Therefore, the measurement of the ADAMTS13 levels in plasma is vital in the diagnosis of TTP and also important in distin- guishing it from other thrombotic microangiopathies (TMA’s). Unfortunately, commercial ADAMTS13 antigen assays are expensive for healthcare service providers in developing countries. However, several antibodies and antibody pairs have been produced against ADAMTS13 and are commercially available. In this study, we evaluated an in-house ADAMTS13 antigen assay using two different commercial antibodies and compared the outcomes to that of a commercial ADAMTS13 antigen kit by using the plasma of 40 patients with possible HIV-associated TTP and 40 healthy subjects. The Intra- and inter-assay coefficients of variation were calculated as 8% and 7% respectively. The assay gave linear results between 0.78 to 12.5% ADAMTS13. The limit of detection was 0.2%, and the limit of quantification was 0.8%. The correlation of our assay compared to the commercial test kit was excellent, with a R2 value of 0.9. In addition, the cost of our ADAMTS13 antigen assay was lower than that of the commercial ADAMTS13 antigen test kit. Our cost-effective in-house ADAMTS13 antigen test also produced reliable results. We therefore recommend that this assay be used to diagnose HIV-associated TTP. KEYWORDS ADAMTS13 antigen; human immunodeficiency virus (HIV); Thrombotic thrombocytopenic purpura (TTP) INTRODUCTION ELISA test, using commercial antibodies to ADAMTS13. We have also evaluated the technical performance of this assay Thrombotic thrombocytopenic purpura (TTP) is a prothrom- with the use of plasma samples from 40 healthy subjects and 40 botic disorder that is characterised by microvascular platelet patients with suspectedHIV-associated TTP and compared it to clumping, resulting in microangiopathic haemolytic anaemia, a commercial test kit. fragmented erythrocytes (schistocytes), consumptive thrombo- cytopenia, renal dysfunction and neurological symptoms.[1] Hu- MATERIALS AND METHODS man immunodeficiency virus (HIV) infection precipitates TTP, probably by infecting vascular endothelial cells which results Subjects in dysfunction of these cells, thrombin generation and the con- Peripheral blood from 40 healthy subjects and 40 patients with sumption of the Von Willebrand factor protease, ADAMTS13.[2] suspected HIV-associated TTP were collected into sodium ci- The diagnosis of TTP and HIV-associated TTP remains based trate. The Health Sciences Research Ethics Committee of the on clinical history, examination of the patient and the blood University of the Free State approved this study in 2016 with film.[1] Assays for measuring ADAMTS13 levels and its function ethics approval number: 92/2016. in plasma help to confirm the diagnosis, differentiate TTP and Sample collection and preparation HIV-associated TTP from other thrombotic microangiopathies (TMA’s) and to monitor the efficacy of treatment. These assays Peripheral blood was collected into VacutainerTM tubes (BD are also helpful in the decision to begin additional or alterna- Vacutainer Systems, Plymouth, UK) containing 0.105M sodium tive therapy.[3] In a previous study, we suggested the use of AD- citrate with a ratio of 1:9 to blood. Platelet-poor plasma was AMTS13 levels as a confirmatory test for HIV-associated TTP. prepared by centrifugation at 200xg for 20 minutes at room The measurement of ADAMTS13 activity using the FRET assay temperature and samples were stored in polypropylene tubes did not confirm the disease and auto-antibodies to ADAMTS13 at -70°C until analysed. All tests were performed on the original aliquots that had not been previously thawed. were found in only 50% of patients with HIV-associated TTP.[4] Although ADAMTS13 antigen kits are available, it is still an Laboratory tests expensive test for small laboratories and for the patients from In-house assay: lower income countries. We have developed a cost-effective A 96-well plate (Nunc) was coated for two days at 4°C with ISSN 1011 5528 | www.smltsa.org.za 1 198 199 OPEN ACCESS Medical Technology SA | Volume 31 No. 1 | June 2017 50ng/ml of a mouse anti-human ADAMTS13 antibody (R&D of the Clinical and Laboratory Standards Institute.[5] In order to Systems, USA; 100µl per well). After washing with PBS with determine the inter-assay precision, we repeat the assay eight 0.1% Tween-20 (Sigma, USA), 100µl of the patient’s plasma, times on the same patient’s plasma in the same run using a pa- volunteer’s plasma or standard plasma was added and incu- tient with a low ADAMTS13 level (i.e. only one level was used). bated for 2 hours at 37°C. Plasma samples from patients and For the intra-assay precision, we used normal pooled plasma volunteers were diluted 1:10 in PBS/Tween-20 of which 100µl (combined from 20 persons) in 5 consecutive tests. was added to two wells and anysed in duplicate. The first In- Normal reference range: ternational standard for ADAMTS13 (NIBSC, UK) was used as We determined the plasma levels of ADAMTS13 in 20 healthy a calibrator and added in the following dilutions: 100%, 50%, males and 20 healthy females using our new assay and calcu- 25%, 12.5%, 6.25%, 3.125% and 1.56%. A blank sample was lated the reference range as the mean ± two standard devia- also included and served as a negative control. After washing tions (SD). the plate four times with a microplate washer (VACUTEC, South Africa), 2µg/ml of a rabbit anti- human ADAMTS13 detection Test comparison: antibody (Santa Cruz, USA; 100µl per well) was added and We compared our method to the Imubind ADAMTS13 ELISA kit from Sekisui Diagnostics (MA, USA) by using plasma from 40 incubated for 1 hour at room temperature. The plate washing patients with possible TTP. was repeated four times and 200ng/ml of a goat anti-rabbit horseradish peroxidase (HRP) conjugated antibody (100µl per Statistical analysis well) was added and incubated for 1 hour, at room temperature. We calculated the reference intervals as the mean ± 2 SD and The plate was further washed for another four times. The colour the Intra- and inter-assay precision by the mean, standard devia- was developed for 20 minutes by adding a substrate solution tion (SD) and the coefficient of variation (CV) from the results of ortho-phenylenediamine (0.05%) in 0.1M citrate-phosphate normal healthy volunteers. The precision goal for the Inter-assay buffer, pH 5.0 containing 0.03% H2O2 (90µl per well). The re- and intra-assay was not to exceed 10% of the CV in the ELISA action was then stopped by adding 100µl of 4M H2SO4. The assay.[6] We calculated the limits of detection and quantification optical density (OD490-630) of the solution on the plate was from the standard curve. The theoretical lower limits of detec- read in a Synergy HT spectrophotometer (Biotek, USA) and an tion (LLD) and the theoretical lower limit of quantification (LLQ) eight-point standard curve with the known ADAMTS13 concen- were then calculated as the minimum detectable concentration trations of the International Standard (Fig 1A) was drawn. The and reliable detection limits, respectively. We compared our ADAMTS13 level for each patient and normal volunteer was assay to the commercial assay by the Bland-Altman and Dem- read from the standard curve. ing regression plots using the 40 patient samples. We defined Commercial assay: the commercial kit as the reference method and calculated the The Immubind ADAMTS13 antigen kit from Sekisui Diagnostics mean bias and 95% limits of agreement between both assays. (USA) was used according to the manufacturer’s instruction. Briefly, plasma and standard samples were added to an anti- RESULTS ADAMTS13 antibody coated ELISA plate. After incubation, a The standard curve for our ADAMTS13 antigen assay (Figure 1a horseradish peroxidase conjugated anti-ADAMTS13 antibody and b) shows a good dose-response, with fast substrate colour was added and the plate was coloured with a substrate for per- development and steep upward linearity. The closeness of fit oxidase. The same ADAMTS13 standard was used to compare shows an r2 value of > 0.99. the two assays. Table 1 summarises the technical performance of our method. Precision studies: The intra- and inter-assay CV's were 8.2% and 7.2% respec- Precision studies were carried out according to the guidelines tively. The reference range for the 40 standard samples varied Figure 1a. Standard curve of our ADAMTS13 antigen assay. The first International Figure 1b. Dilutional linearity fit of the assay with 95% confidence intervals. Standard for ADAMTS13 from the NIBSC was used as a standard for ADAMTS13 Dilutions range from 1.56% to 6.25% ADAMTS13. concentrations ranging from 0% to 100% on the X-axis. Each point represents the mean of duplicate reading for a single representative experimental data set. ISSN 1011 5528 | www.smltsa.org.za 2 199 200 OPEN ACCESS Medical Technology SA | Volume 31 No. 1 | June 2017 Table 1. Acceptance criteria and performance characteristics of our ADAMTS13 antigen assay VALIDATION PARAMETER ACCEPTANCE CRITERIA OBSERVED RESULTS MEAN SD %CV Intra-assay precision CV ≤ 10 % 42.05 2.835 6.74 Inter-assay precision CV ≤10 % 81.2 5.848 7.2 Limit of detection (LOD) 1.56% ADAMTS13 Limit of quantification (LOQ) 1.56% ADAMTS13 Dilution linearity • R2 0.99 0.9981 • Slope 0,03523 ± 0,001089 • Y-Intercept 0,0623 ± 0,003899 • X-Intercept -1,769 • range 1.56 – 6.25 % ADAMTS13 Figure 2a. Regression analysis with 95% confidence intervals comparing our Figure 2b. Bland-Altman plot of the ADAMTS13 assay comparison. The mean method to that of the commercial assay using 40 patient samples. relative bias is represented by a solid line, while the broken lines illustrate the 95% limits of agreement. from 48% to 138% (93% ± 46%). The limit of detection and commercial kit method was found to be 5.2% and 7.2% respec- quantification was 1.56% ADAMTS13. tively, which correlated well with the corresponding values of Figure 2a shows the Deming regression plot comparing our as- 6.7% and 7.2% of our assay. For viability It has been previously [5] say to the that of the commercial test kit. The two tests compared recommended that the CV's must be less than 10%. Both the commercial assay and our assay performed well within these excellently with an r2 of > 0.9 and a slope of almost 1. Figure 2b limits. shows the Bland-Altman plot comparing the two methods. The bias was calculated as 4.575, while the 95% limits of agreement Regression analysis showed that both assays were linear when were between 24.12 and 14.97 difference. measuring ADAMTS13 in different dilutions of human plas- ma. The two assays were comparable in the measurement of DISCUSSION ADAMTS13 in the plasma of patients with possible HIV-asso- Acute TTP is diagnosed if the plasma ADSAMTS13 level is ciated TTP. found to be less than 10 percent.[7] However, it is difficult to The range of detection reported by the manufacturers of the distinguish between TTP and atypical haemolytic uraemic syn- commercial assay is from 2% up to 150% ADAMTS13. Our drome (aHUS) without an ADAMTS13 antigen assay. We have test's detection range was up to 100%. However, our detection developed an in-house ELISA method to measure ADAMTS13 range is appropriate for patients with possible HIV-associated levels in plasma by using two different commercial antibodies TTP patients who are suspected to have very low levels of [6] which recognise epitopes on ADAMTS13 in a sandwich ELISA. ADAMTS13. The purpose of the evaluation studies was to ensure that ana- The lower limit of detection of our assay was 1.56%. This cor- lytical methods can detect the corresponding analyte and and relates well with that of the commercial test of 2%. in addition provide repetitive and accurate results.[8] No other Compared to the cost of the commercial kit, our method was assays to date, have been validated against a commercial assay 90% less than the commercial kit. The commercial kit costs in current literature. In this assay, we measured the Intra and R500 per test minus the labour costs. If the labour costs were to inter-assay precision, as well as the limit of detection and limit be included, the assay costs would exceed that of the current of quantification of our assay. The intra and inter-assay CV of the medical aid rates. In comparison our assay costs R50 per test ISSN 1011 5528 | www.smltsa.org.za 3 200 201 OPEN ACCESS Medical Technology SA | Volume 31 No. 1 | June 2017 minus the labour costs. Labour costs were designated to be the 2. Brecher ME, Hay SN, Park Y. Is it HIV TTP or HIV-associated same for both assays. thrombotic microangiopathy? J Clin Apheresis 2008; 23:186-190. 3. Hassan S, Westwood JP, Ellis D, Laing C, McGuckin S, Benjamin Despite a good correlation, the Bland-Altman test showed S, Scully M. The utility of ADAMTS13 in differentiating TTP from a mean bias of 4.575. The Deming regression analysis of the other acute thrombotic microangiopathies: results from the UK two assays showed a slope of 1.009 ± 0.04166 in which the TTP Registry. Br J Hematol 2015; 171:830-835. 95% confidence intervals included the slope of 1.0, This is well 4. Meiring M, Webb M, Goedhals D, Louw V. HIV-associated TTP – within the acceptable limits. What we know so far. Eur Oncol Hematol 2012; 8:89-91. 5. Andreasson U, Perret-Liaudet A, Van Waalwijk van Doorn LJC et CONCLUSION al. A practicag huide to immuno assay method validation. Front Our assay showed excellent inter- and intra-assay precision that Neurol 2015; 6:179. can detect different ADAMTS13 levels in plasma up to 100%. 6. Rajasekariah GHR, Kay GE, Russel NV and Smithyman AM. As- sessment of Assay Sensitivity and Precision in a Malaria Antibody We are therefore confident that our assay can successfully be ELISA. J Immunoassay Immunochem 2003; 24:89-112. used to diagnose patients with HIV-associated TTP since the re- 7. Yenerel MN. Atypical Hemolytic Uremic Syndrome: Differential sults compare favourably to those of a commercial available kit. diagnosis from TTP/HUS and management. Turkish J Hematol 2014; 31:216-225. REFERENCES 8. Tecles F, Juentes P, Martinez-Subiela S, Parra MD, Mun-oz A, 1. Sarig G. ADAMTS13 in the Diagnosis and Management of Ceron JJ. Analytical validation of the commercially available Thrombotic Microangiopathies. Rambam Maimonides Med J method for accurate phase protein quantification in pigs. Res Vet 2014;5:1-15. Sci 2007; 83:133-139. © The Society of Medical Laboratory Technologists of South Africa ISSN 1011 5528 | www.smltsa.org.za 4 201 S-21 202 Int. J. Med. Sci. 2016, Vol. 13 759 Ivyspring International Publisher International Journal of Medical Sciences 2016; 13(10): 759-764. doi: 10.7150/ijms.15688 Research Paper Tissue factor: A potent stimulator of Von Willebrand factor synthesis by human umbilical vein endothelial cells Muriel Meiring, W. Allers, E. Le Roux Department of Haematology and Cell Biology, University of the Free State Bloemfontein, South Africa.  Corresponding author: Muriel Meiring; meiringsm@ufs.ac.za © Ivyspring International Publisher. Reproduction is permitted for personal, noncommercial use, provided that the article is in whole, unmodified, and properly cited. See http://ivyspring.com/terms for terms and conditions. Received: 2016.03.29; Accepted: 2016.08.15; Published: 2016.09.20 Abstract Inflammation and dysfunction of endothelial cells are thought to be triggers for the secretion of Von Willebrand factor. The aim of this study was to examine the effects of the inflammatory cytokines interleukin-6 (IL-6), interleukin-8 (IL-8) and tumour necrosis factor-alpha (TNF-α) and the coagulation factors, tissue factor and thrombin on the release and cleavage potential of ultra-large von Willebrand factor (ULVWF) and its cleavage protease by cultured human umbilical vein endothelial cells (HUVEC). HUVEC were treated with IL-6, IL-8, and TNF-α, tissue factor (TF) and thrombin, and combinations thereof for 24 hours under static conditions. The cells were then exposed to shear stress after which the VWF-propeptide levels and the VWF cleavage protease, ADAMTS13 content were measured. All treatments and their combinations, excluding IL-6, significantly stimulated the secretion of VWF from HUVEC. The VWF secretion from the HUVEC was stimulated most by the combination of TF with TNF-α. Slightly lower levels of ADAMTS13 secretion were found with all treatments. This may explain the thrombogenicity of patients with inflammation where extremely high VWF levels and slightly lower ADAMTS13 levels are present. Key words: Endothelium, Von Willebrand factor, ADAMTS13, Inflammation. Introduction To date, extensive progress has been made in the effect on ADAMTS13 synthesis has not been studied. knowledge of inflammation. It is now known that, The stimulatory effect of TF on endothelial cells to pro-inflammatory mediators are released or produced release ULVWF and ADAMTS13 has also not been from the surrounding tissue and cellular components studied. such as mast cells after injury [1]. Under inflammatory The increased amounts of VWF multimers due to conditions, the endothelium responds by regulating entothelium stimulation with these cytokines, then its own permeability and releases pro-inflammatory might bind platelets to form thrombi in small vessels. mediators such as cytokines. Among the A disintegrin-like metalloprotease with pro-inflammatory cytokines, tumour necrosis factor-α thrombospondin type I repeats-no 13 (ADAMTS13) (TNF-α), interleukin-6 (IL-6) and interleukin-8 (IL-8) released from endothelial cells cleaves the VWF have been implicated as the primary endogenous multimers into smaller and less active forms. These mediators of inflammation [2]. IL-6 up-regulates VWF multimers mediate the initial adhesion of fibrinogen, tissue factor, Von Willebrand factor activated platelets, the first step in thrombus (VWF), and factor VIII levels [3,4]. IL-6, IL-8, and formation. This process may be affected by the TNF-α were shown to have stimulatory effects on the amount of VWF and ADAMTS13 secreted [7]. endothelial cell release of ULVWF multimers, but not Ultimately, the increased VWF levels and the on the cleavage of ULVWF by ADAMTS13 [5]. decreased ADAMTS13 levels contribute to the Thrombin has also stimulatory effects on the development of thrombotic and inflammatory endothelial cell release of ULVWF multimers [6]. Its diseases, such as sepsis, antiphospholipid syndrome, http://www.med2s0c2i.org 203 Int. J. Med. Sci. 2016, Vol. 13 760 atherosclerosis, systemic schlerosis, diabetes and makes it a viable option for VWF research [17, 18]. thrombotic thrombocytopenic purpura (TTP) [8-11]. The extremely high VWF levels and slightly lower Cell culture treatments under shear stress ADAMTS13 levels in these patients are not The 4th-passaged HUVEC were grown until completely understood. confluent on all tissue culture dishes. Thereafter, the This study examines the effects of the most old medium was discarded and the dishes prepared commonly present inflammatory cytokines (IL6, IL8 for treatment. The HUVEC were stimulated with and TNF-α) and coagulation factors (tissue factor and inflammatory cytokines, IL-6, IL-8 and TNF-α thrombin) and especially combinations thereof on the (Invitrogen, USA) and also with human TF (Innovin, secretion of VWF and ADAMTS13 by cultured Siemens, South Africa) and bovine thrombin HUVEC. For optimal VWF-propeptide and (Beckman Coulter, South Africa). This was done by ADAMTS13 secretion, we used sub-confluent and incubating the cells for 24 hrs with the different fourth passage cells [12-14]. This might enable us to compounds and combinations of the cytokines determine the highest contributing factors to the respectively before applying shear stress. Six culture extremely high VWF levels and slightly lower flasks were used for each stimulant, three for the ADAMTS13 levels in patients with thrombotic control (untreated) and three for the treatment disorders. (treated). The final concentration for each treatment was prepared in 5 mL of supplemented Medium 200. Materials and Methods The following final concentrations were used for the inflammatory cytokines: 0 ng/mL as control, and 100 Procedure ng/mL as treatment. For thrombin, 0 IUnits/mL as Cultured human umbilical vein endothelial cells control, and 2 IUnits/mL as treatment and for tissue (HUVEC) were stimulated with cytokines IL-6 (100 factor 0 ng/mL as control, and 662 µl/mL as ng/mL), IL-8, (100 ng/mL) and TNF-α (100 ng/mL) treatment were used. The same concentrations of the and also with coagulation factors, thrombin (2 stimulants were used for the combinations as for the IUn/mL) and TF (662ng/mL) and combinations of different compounds alone. The cytokine these compounds. The combinations include that of concentrations were based on the experiments done IL-8 and thrombin, IL-8 and TF, TNF-α and thrombin by Bernardo et al. (2004) and Cao et al. (2008) who used and TNF-α and TF. The control for the model was the same concentrations [5, 7]. The coagulation factor untreated cells for each treated sample. concentrations were determined by a previous Endothelial cell culture non-published dose-response study. After the treatment period, the control and HUVEC cell line C-003-5C (Invitrogen, USA) stimulated dishes were carefully removed from the were maintained in round 22.1cm2 fibronectin incubator, and placed onto a ROTEM orbital shaker (10ng/mL, Invitrogen, USA) coated tissue culture for 1 hour to generate a wall shear stress of about 2.5 dishes at a concentration of 1.25 x 104 cells/mL with a dynes/cm2. It is known that under shear stress volume of 5 ml per culture dish. The cells were conditions, VWF becomes more susceptible to cultured in Medium 200 supplemented with Low proteolysis by ADAMTS13. This shear stress is Serum Growth Supplement (LSGS, Invitrogen) necessary for HUVEC to exposed VWF to proteolysis containing foetal bovine serum (FBS), hydrocortisone, [19-20]. The concentration of the treatment cytokines human epidermal growth factor, basic fibroblast and coagulation factors were assumed to stay growth factor and heparin in a humidified 37°C, 5% constant during the treatment and rotation process, CO2 / 95% air cell culture incubator. We cultured the since evaporation was negligible. Lastly, the flasks HUVEC in the flasks for approximately 3 days until were removed from the ROTEM and the perfusate the confluency was estimated to be 80%. collected, aliquoted and stored at -80ºC until the Sub-culturing was done with Trypsin/EDTA (0.025% measurements were performed. Trypsin / 0.01% EDTA solution) at room temperature for 3 minutes whereafter the action of Trypsin/EDTA ADAMTS13 levels was blocked by 3 mL of Trypsin Neutralizer solution Due to the expected low ADAMTS13 levels, the (Invitrogen, USA). Only 4th-passaged HUVEC were ADAMTS13 release was detected using a sodium used for all experiments. HUVEC were used, since it dodecyl-sulphide – poly-acrylamide gel-electro- is a major endothelial cell type that produces phorsis (SDS-PAGE), followed by Western-blot ADAMTS13 in humans and it is readily available and detection. The SDS-PAGE was prepared and the results are comparable [5, 12-16]. Human performed using a 12% separating gel and a 4% umbilical vein endothelial cells are also capable of stacking gel. Following SDS-PAGE, the gel was expressing VWF and also carry ADAMTS13, which http://www.med2s0c3i.org 204 Int. J. Med. Sci. 2016, Vol. 13 761 blotted onto a PVDF membrane (Thermo Scientific, added in duplicate (100 µl/well) and incubated for 2 USA). The PVDF membrane first blocked with 2% hours at 37ºC. An HRP-conjugated monoclonal skimmed milk powder in TBS-0.1% Tween-20 for 1 antibody against VWF-propeptide (CLB-Pro 14.3, hour at room temperature and washed 6 times with Euro-immune, Germany) was added in a 1:100 TBS-0.1% Tween-20. The membrane was then dilution after another wash step, and incubated for 1 incubated for 2 hours at room temperature in a 1:100 hour at room temperature in order to detect the dilution of a rabbit polyclonal IgG antibody against concentration of VWF-propeptide in the perfusate. human ADAMTS13 (Santa Cruz Biotechnology, CA, We used OPD (50 mg.L-1) as the substrate for HRP. USA). After washing again, it was incubated for 1 The WHO’s (World Health Organisation) 6th hour at room temperature with a polyclonal goat anti- FVIII/VWF standard was used as the standard rabbit antibody conjugated with horseradish against which the perfusates were measured. The peroxidise (HRP) (1:2,000 dilution, Santa Cruz results were expressed as percentage difference from Biotechnology, CA, USA) to detect the presence of the control samples. Thus (VWF-propeptide level of ADAMTS-13. Equal volumes of ECL Western-blot stimulated sample minus VWF-propeptide level of detection reagent 1 and 2 (AEC Amersham, UK) were control) / VWF-propeptide level of control * 100. mixed and poured onto the membrane for 1 minute whereafter the membrane was sealed with plastic film Statistics and exposed to an X-ray film for 1 to 10 minutes in the It is important to notice that the whole dark. Finally, the film was developed in an automated experiment was done three times. Thus for each film developer (Kodak, CA, USA) and the picture was stimulant we culture 6 flasks, three control and three scanned into a computer. The strength of the signal of treated flasks, and each perfusate was measured in ADAMTS13 was quantified by densitometric analysis duplicate for the ADAMTS13 and VWF-propeptide using the ImageJ software (Thermo Scientific, USA). levels. The percentage increase or decrease from the itometric analysis of ADAMTS13 content was done control samples was calculated and all the and the percentage difference from the control was experimental data were presented as mean ± SD calculated by dividing the difference in density for (standard deviation). The unpaired 2-tailed Student each treated sample from the density for the control t-test was used to test for differences and P- values by the density of the control sample. less than 0.05 (P<0.05) were considered as statistically significant. VWF-propeptide levels VWF-propeptide levels in the perfusates were Results determined using an in-house ELISA. In short, a Figure 1 shows the densitometic analysis of the 96-well ELISA plate was coated overnight at 4ºC with ADAMTS13 content in the stimulated samples. This is a monoclonal antibody against the VWF-propeptide expressed as percentage difference from the control (CLB-Pro 35, Euro-Immune, Germany, 1:100 dilution samples. The original Western-blot picture is shown in PBS, 100 µl per well). The plate was blocked with in Figure 2. An overall decrease in density of the 4% bovine serum albumin (BSA) in PBS (200 µl/well) stimulated samples compared to that of the controls is for 2 hours at room temperature and washed four clearly visible. times with PBS/0.1%Tween-20. The perfusates were Figure 1. Effect of cytokines (IL6. and TNF), coagulation factors (Thr and TF) and combinations thereoff (TNF+Thr, TNF+TF) on ADAMTS13 levels in the perfusates. The bar graphs shows the mean % difference from the control of the 3 experiments done in duplicate (n=3) and the error bars indicates the standard deviation. http://www.med2s0c4i.org 205 Int. J. Med. Sci. 2016, Vol. 13 762 Figure 2. Western Blot indicating the presence of the ADAMTS13 protein in all samples. The positions of the molecular markers are shown on the left side. For illustration purposes, only some of the samples are shown. Not stimulated sample (lane 1), IL6 (lane 2), IL8 (lane 3), TNFα (lane 4), thrombin (lane 5), tissue factor (lane 6), IL8 + thrombin (lane 7), TNFα + thrombin (lane 8), IL8 + tissue factor (9) and TNFα + tissue factor (lane 10). Figure 3. Effect of cytokines (IL6, IL8 and TNF), coagulation factors (Thr and TF) and combinations thereof (TNF+Thr, TNF+TF, IL8+Thr and IL8+TF) on VWF-propeptide levels. The bar graphs shows the mean % difference from the control of the 3 experiments done in duplicate and the error bars indicates the standard deviation. The * indicates statistic significance (p<0.05). The percentage difference of all stimulated thrombin and TNF-α and tissue factor shows a samples, except for IL-6 to that of the control samples significant increase in VWF propeptide levels ranged between a ratio of 30 to 60%. Although this compared to that of the control. was not a quantitative measurement, it nevertheless showed that the ADAMTS13 secretion was decreased when endothelial cells were stimulated with IL-8, Table 1. Effect of cytokines (IL6, IL8 and TNF), coagulation TNF-α, thrombin and TF and combinations of TF and factors (Thr and TF) and combinations thereoff (TNF+Thr, TNF+TF, IL8+The and IL8+TF) on VWF-propeptide levels (%). thrombin with the cytokines (IL-6 had no effect). In summary, endothelial cells (HUVEC) Stimulant % VWF-propeptide % VWF-propeptide Percentage of Control of Treated increase stimulated with cytokines and coagulation factors (mean ± SD) (n=6) (mean ± SD) (n=6) from control secreted more VWF and less ADAMTS13. (mean ± SD) The VWF-propeptide levels rather than VWF IL-6 56.5 ± 1 52 ± 1 (-)8 ± 1 TNF 54 ± 1.5 104.5 ± 2.5 48 ± 5 antigen levels were measured in the perfusate, since it Thr 73.5 ± 4 140 ± 5 90 ± 5 * provides a more accurate measurement of VWF TF 47.5 ± 2 109.5 ± 4.5 57 ± 4 synthesis [21]. Figure 3 shows the mean percentage TNF + Thr 64.5 ± 2 139 ± 5 116 ± 6 * change from the controls, while table 1 provides the TNF + TF 54 ± 2 130.5 ± 5 142 ± 4 * IL-8 21 ± 1 23.5 ± 1.5 5 ± 1 original data of the VWF-propeptide levels in the IL-8+ Thr 27 ± 1 44.5 ± 2 39 ± 2 control and treated samples. IL-8 + TF 20 ± 3 30.5 ± 2 53 ± 6 VWF-propeptide were secreted by all The results are given as the mean of 3 experiments in duplicate ± 1 standard deviation (n=6). The * indicates statistic significance at p<0.05. treatments, except for IL-6. The amount of VWF-propeptide was secreted from the lowest to the highest values by the following treatments: IL-8, The levels of secreted VWF-propeptide, except IL-8+thrombin, TNF-α, thrombin, TNF-α, for IL-6 which cause a decrease, increased in the TNF-α+thrombin, tissue factor, IL-8+tissue factor, and following order from low to high: IL-8, with the highest increase seen at TNF-α+tissue factor IL-8+thrombin, TNF-α, thrombin, TNF-α, treatment. Treatments with thrombin, TNF-α and TNF-α+thrombin, tissue factor, IL-8+tissue factor, and http://www.med2s0c5i.org 206 Int. J. Med. Sci. 2016, Vol. 13 763 with the highest increase the TNF-α+tissue factor We found no evident change in the HUVEC treatment. Only stimulation with Thrombin, ADAMTS13 release upon stimulation with TF. This TNF-α+thrombin and TNF-α+tissue factor increased correlates with studies where shiga toxin was used to the VWF-propeptide levels significantly. stimulate TF expression. Cells treated with shiga toxin also did not increase or decrease the ADAMTS13 Discussion levels [26, 27]. We hypothesized that the combination of certain This is also the first time where the effect of inflammatory cytokines and coagulation factors that tissue factor was tested on VWF secretion in HUVEC are released during inflammation may stimulate the cells. Tissue factor increased VWF secretion markedly. synthesis and release of VWF and simultaneously The highest increase was when HUVEC was slightly decrease the synthesis of ADAMTS13. This stimulated by the combination of tissue factor and might result in a thrombotic state where excessive TNF- α (Table 1). It is however not clear if TF can be amounts of VWF-platelet strings cannot be linked to increased VWF-propeptide levels in vivo degradated and might occlude small vessels. [28-30]. However, in HIV where TF levels are To test this hypothesis, we determined the effects increased, it is possible that the constant stimulation of 3 inflammatory cytokines, interleukin-6 (IL-6), of VWF release by tissue factor might contribute to the interleukin-8 (IL-8) and tumour necrosis factor-α extremely high levels of VWF in these patients [31]. (TNF-α), which are released during the early stages of Our results with IL-8 and TNF-α-stimulation systemic inflammation on the secretion of VWF and were consistent with a study done by Bernardo et al., its cleaving protease ADAMTS13 in human primary (2004) who indicated a similar effect of IL-8 and HUVEC. This study presents the first in vitro evidence TNF-α on VWF synthesis by endothelial cells [5]. on the effects of thrombin and tissue factor, combined We found total opposite results with with inflammatory cytokines, on VWF synthesis and ADAMTS13 compared to VWF-propeptide in the cleavage in HUVEC. perfusates. IL-8, TNF-α, tissue factor, IL-8+tissue Since VWF becomes more susceptible to factor and TNF-α+tissue factor decreased the levels of proteolysis by ADAMTS13 during shear stress, we ADAMTS13 (Fig. 1). Although the results with the used a fluid shear stress of 2.5 dyne/cm2 for a small different stimulation regimes were not significantly time period to ensure the expression of the different, the decrease with all stimulation regimes VWF-propeptide and ADAMTS13 [20, 21]. The were substantial. The affected decrease in release of VWF-propeptide levels were measured instead of the ADAMTS13 in human umbilical endothelial cells by VWF antigen levels, since it is a more accurate IL-8, TNF-α, thrombin and tissue factor and their measurement of VWF secretion [21]. combined effects, may offer a logical explanation of IL-6 was the only cytokine that did not affect how systemic inflammation and/or infection might VWF-propeptide and ADAMTS13 secretion (Table 1, trigger a thrombotic condition. Figure 1). However, this was expected because We suggest that certain inflammatory cytokines HUVEC do not have receptors for IL-6 [22, 23]. This and coagulation factors that are released during result confirmed the findings of Bernardo et al. (2004), pathological conditions may affect the balance who reported that IL-6 alone does not stimulate the between the quantity of VWF multimers and VWF synthesis from HUVEC [5]. However, IL-6 in ADAMTS13 released from endothelial cells. More complex with its receptor on other endothelial cells VWF and VWF-propeptide might be secreted and less induced a small but significant release of VWF [5]. ADAMTS13. As a result, the excessive amounts of The VWF-propeptide levels increased when VWF, with lesser amounts of the VWF cleavage HUVEC was stimulated with TNF-α (Table 1). This protease, ADAMTS13, might cause obstructions of supports the results of other similar studies [5, 16, 24]. platelet-VWF plugs in smaller vessels, leading to The precise mechanism through which TNF-α thrombosis [27, 29-31]. This should be investigated induces this increase is unknown [25]. In accordance further in more detail. with Cao et al. (2008), which found lower ADAMTS-13 mRNA levels after 24-hour treatment with 10 ng/mL Conclusion TNF-α, we also found that TNF-α induced a We investigated the effect of inflammatory and decreased ADAMTS13 release [7]. thrombotic stimuli and the combination thereof on the IL-8 did not stimulate the secretion of VWF by release of VWF and its cleavage protease ADAMTS13 HUVEC. It however induced a decrease in HUVEC by human umbilical vein endothelial cells (HUVEC). ADAMTS13 levels (Fig. 1). To date, no studies have Tissue factor and especially the combination of TF and been done on the effect of IL-8 treatment on TNF-α was the most potent stimulator of von ADAMTS13 release. Willebrand factor secretion by HUVEC. Thus, the http://www.med2s0c6i.org 207 Int. J. Med. Sci. 2016, Vol. 13 764 results in this study may provide a link between 22. Romano M, Sironi M, Toniatti C, et al. Role of IL-6 and its soluble receptor in induction of chemokines and leukocyte recruitment. Immun. 1997; 6: 315-325. inflammation and thrombosis, which may also be of 23. Peters M, Müller AM, Rose-John S. Interleukin-6 and soluble interleukin-6 therapeutic importance. It may also help to receptor: direct stimulation of gp130 and hematopoiesis. Blood 1998; 92: 3495-3504. understand the mechanisms that lead to thrombotic 24. Jy W, Jimenez JJ, Mauro LM, et al. Endothelial Microparticles Induce disorders with increased VWF levels and lower Formation of Platelet Aggregates Via a Von Willebrand Factor/RistocetinDependent Pathway, Rendering them Resistant to Dissociation. J Thromb ADAMTS13 levels. Haemost. 2005; 3: 1301-1308. 25. Curtis AM, Wilkinson PF, Gui M, et al. P38 Mitogen-Activated Protein Kinase Acknowledgement Targets the Production of Proinflammatory Endothelial Microparticles. JThromb Haemost. 2009; 7: 701-709. 26. Nestoridi E, Kushak RI, Duguerre D, et al. Up-Regulation of Tissue Factor This study was done with a grant from the Activity on Human Proximal Tubular Epithelial Cells in Response to Shiga National Research Foundation of South Africa. Toxin. Kidney Int. 2005; 67: 2254-2266. 27. Nolasco LH, Turner NA, Bernardo A, et al. Hemolytic Uremic Syndrome-Associated Shiga Toxins Promote Endothelial-Cell Secretion and Competing Interests Impair ADAMTS13 Cleavage of Unusually Large Von Willebrand Factor Multimers. Blood. 2005; 106: 4199-4209. The authors have declared that no competing 28. Lip GYH. The Prothrombotic State in Atrial Fibrillation: The Atrium, The interest exists. Endothelium and Tissue Factor? Thromb Res. 2003; 111: 133-135. 29. Sato M, Suzuki A, Nagata K, et al. Increased Von Willebrand Factor in Acute Stroke Patients with Atrial Fibrillation. J Stroke Cerebrovasc Dis. 2006; 15: 1-7. References 30. Yamashita A, Sumi T, Goto S, et al. Detection of Von Willebrand Factor andTissue Factor in Platelets-Fibrin Rich Coronary Thrombi in Acute Myocardial 1. Granger DN and Kubes P. The microcirculation and inflammation: Infarction. Am J Cardiol. 2006; 97: 26-28. modulation of leukocyte-endothelial cell adhesion. J Leukoc Biol. 1994; 55: 31. Meiring M, Webb M, Goedhals D, et al. HIV-associated thrombotic 662-675. thrombocytopenic Purpura – What we know so far. Eur Oncol Haematol. 2. Tracey KJ and Cerami A. Tumor necrosis factor: a pleiotropic cytokine and 2012; 8(2): 89-91. therapeutic target. Annu Rev Med. 1994; 45: 491-503. 3. Neumann FJ, Ott I, Marx N, et al. Effect of human recombinant interleukin-6 and interleukin-8 on monocyte procoagulant activity. Arterioscler Thromb Vasc Biol. 1997; 17: 3399-3405. 4. Kerr R, Stirling D, Ludlam C. Interleukin-6 and haemostasis. Br J Haematol. 2001; 115: 3-12. 5. Bernardo A, Ball C, Nolasco L, et al. Effects of Inflammatory Cytokines on the Release and Cleavage of the Endothelial Cell-Derived Ultra-large Von Willebrand Factor Multimers Under Flow. Blood. 2004; 104: 100-106. 6. Wagner DD. New links between inflammation and thrombosis. Arterioscler Thromb Vasc Biol. 2005; 25: 1321-1324. 7. Wang A, Duan Q, Liu X, et al. All-trans retinoic acid modulates the balance of ADAMTS13 and VWF in human micro-vascular endothelial cells. Microvasc Res. 2015; 102: 6-10. 8. Habe K, Wada H, Ito-Habe N, et al. Plasma ADAMTS13, Von Willebrand Factor (VWF) and VWF Propeptide Profiles in patients with DIC and related diseases. Thromb Res. 2012; 129: 598-602. 9. Scheja A, Akesson A, Geborek P, et al. Von Willebrand Factor propeptide as a marker of disease activity in systemic sclerosis (scleroderma). Arthritis Res. 2001; 3: 178-182. 10. Vischer UM, Emeis JJ, Bilo HJG, et al. Von Willebrand Factor as a plasma Marker of endothelial activation in diabetes: Improved reliability with parallel determination of the VWF propeptide (VWF:AgII). Thromb Haemost. 1998; 80: 1002-7. 11. Van Mourik J, Romani de Wit T, et al. Von Willebrand Factor Propeptide in vascular disorders. Thromb Haemost. 2001; 86: 164-71. 12. Hussein MNA, Bӧing AN, Biró É, et al. Phospholipid Composition of In Vitro Endothelial Microparticles and Their In Vivo Thrombogenic Properties. Thromb Res. 2008; 121: 865-871. 13. Luu NT, Rahman M, Stone PC, et al. Responses of Endothelial Cells From Different Vessels to Inflammatory Cytokines and Shear Stress: Evidence for the Pliability of Endothelial Phenotype. J Vasc Res. 2010; 47: 451-461. 14. Laflamme K, Roberge CJ, Pouliot S, et al. Tissue-engineerd human vascular media produced in vitor by the self-assembly approach present ifunctional properties similar to those of their native blood vessels. Tissue Eng. 2006; 12: 2275-2281. 15. Inoguchi H, Tanaka T, Maehara Y et al. The effect of gradually graded shear stress on the morphological integrity of huvec-seeded compliant small-diameter vascular graft. Biomaterials. 2007; 28: 486-495. 16. Combes V, Simon A, Grau G, et al. In Vitro Generation of Endothelial Microaprticles and Possible Prothrombotic Activity in Patients with Lupus Anticoagulant. J Clin Invest. 1999; 104: 93-102. 17. Turner N, Nolasco L, Tao Z, et al. Human Endothelial Cells Synthesise and Release ADAMTS-13. J Thromb Haemost. 2006; 4: 1396-1404. 18. Liu J, Yuan L, Molema G, et al. Vascular Bed-Specific Regulation of the Von Willebrand Factor Promotor in the Heart and Sceletal muscle. Blood. 2011; 117: 342-351. 19. Sargent CY, Berguig GY, Kinney MA, Hiatt LA, et al. Hydrodynamic modulation of embryonic stem cell differentiation by rotary orbital suspension culture. Biotechnol Bioeng. 2010; 105: 611-626. 20. Dong J, Moake JL, Nolasco L, et al. ADAMTS-13 Rapidly Cleaves Newly Secreted Ultralarge Von Willebrand Factor Multimers on the Endothelial Surface Under Flowing Conditions. Blood. 2002; 100: 4033-4039. 21. Ragni MV. On the cutting edge: von Willebrand factor propeptide and thrombosis. J Thromb Haemost. 2006; 4: 2553-2555. http://www.med2s0c7i.org IN VITRO STUDIES 208 eISSN 2325-4416 © Med Sci Monit Basic Res, 2017; 23: 344-351 DOI: 10.12659/MSMBR.905978 S-22 Received: 2017.06.30 Accepted: 2017.08.18 Tissue Engineered Small Vessel Conduits Published: 2017.10.30 – The Anti-Thrombotic Effect of Re-Endothelialization of Decellularized Baboon Arteries: A Preliminary Experimental Study Author s’ Contribution: AE 1,2 Muriel Meiring* 1 Department of Haematology and Cell Biology, University of the Free State, Study Design A BEF 1 Mmakgabu Khemisi* Bloemfontein, South Africa Data Collection B 2 National Health Laboratory Services, Universitas Hospital, Bloemfontein, Statis tical Analysis C C 3 Leana Laker South Africa Data In terpretation D AE 3,4 Pascal M. Dohmen 3 Department of Cardiothoracic Surgery, University of the Free State, Bloemfontein, Manuscrip t Preparation E AEG 3 Francis E. Smit South Africa Literature Search F 4 Department of Cardiac Surgery, University Hospital Oldenburg, Carl von Ossietzky Funds Collection G University, Oldenburg, Germany * These authors contributed equally in this project Corresponding Author: Muriel Meiring, e-mail: meiringsm@ufs.ac.za Source of support: This work was supported by a grant from the National Research Foundation of South Africa (Grant no: CPRR14080687541) Background: The use of decellularized biological scaffolds for the reconstruction of small-diameter vascular grafts remains a challenge in tissue engineering. Thrombogenicity is an important cause of obstruction in these vessels due to decellularization. Seeding of the decellularized vascular constructs with endothelial cells is therefore a pre- requisite for the prevention of thrombosis. The aim of this study was to seed decellularized baboon arteries with endothelial cells and to compare the thrombogenicity to that of decellularized arteries after circulation of blood. Material/Methods: Carotid, radial, and femoral arteries (12 arteries in total) were harvested from 2 Papio ursinus baboons. Ten ar- teries were decellularized. Normal morphology was confirmed in the control vessels. The effect of re-endothe- lialization was studied in the vessel scaffolds using scanning electron microscopy (SEM) and transmission elec- tron microscopy (TEM). Decellularization resulted in vessel scaffolds with well-preserved extracellular matrix and intact basal membranes. Six of the decellularized vessel scaffolds were seeded with viable human umbili- cal vein endothelial cells (HUVEC). Luminal endothelialization was established after 7 days in a bioreactor and SEM confirmed confluency. Two control, 4 decellularized, and 6 decellularized re-endothelialized vessel scaf- folds were studied in an in vitro flow chamber using baboon blood. Results: The decellularized arteries showed an absence of endothelial lining, and an intact basement membrane. The seeding process produced a complete endothelial layer on the surfaces of the arteries. After perfusion with whole blood, no thrombi were formed in the control arteries and re-endothelialized vessels. Widespread plate- let activation and adhesion occurred in the decellularized vessels despite a relatively intact basal membrane. Conclusions: This study supports the development of re-endothelialized tissue engineered small-vessel conduits. MeSH Keywords: Blood Coagulation • Blood Vessel Prosthesis • Endothelial Cells • Tissue Engineering • Tissue Scaffolds Full-text PDF: https://www.basic.medscimonit.com/abstract/index/idArt/905978 3090 1 5 18 208 This work is licensed under Creative Common Attribution- 344 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: The anti-thrombotic effect of re-endothelialized baboon arteries IN VITRO STUDIES © Med S2ci0 M9onit Basic Res, 2017; 23: 344-351 Background endothelialization of vascular grafts has been limited due to the cost and availability of reagents, and because it is diffi- Vascular diseases are responsible for more than 25% of all cult for endothelial cells to stay attached to the scaffold [11]. deaths worldwide [1]. Therapies for vascular diseases often require bypassing or replacement of the diseased vessels with The aim of the present study was to re-endothelialize small-di- vascular grafts. However, many patients do not have healthy ameter (<6 mm) decellularized baboon arteries using cultured vessels available for grafting due to pre-existing vascular con- HUVECs. Additionally, the re-endothelialized arterial scaffolds ditions, size mismatch, or available autograft conduits [2,3]. were perfused with baboon blood at high shear stress and Currently, arteries, such as the aorta or the iliac arteries, are compared to those of fresh baboon arteries. reconstructed using synthetic grafts that are made of expand- ed polytetrafluoroethylene (ePTFE) or Dacron. These synthetic Baboon models possess similar hemostatic characteristics to grafts are also used for reconstruction of small-diameter ar- humans. Their coagulation system and platelet behavior close- teries; however, the patency rates are not favorable because ly resembles that of humans, whereas other animal species of thrombogenicity and limited re-endothelialization capaci- such as dogs, sheep, and pigs do not. Baboon vascular endo- ty in vivo [4]. Some autologous endothelial cell seeding grafts thelial cell growth characteristics are also thought to be simi- have been implanted with promising results even in very small lar to that of humans. Furthermore, they share about 98% ho- (4 mm)-diameter grafts that were clinically used [4,5]. The dis- mology to human genes, possess similar protein structures to advantages, however, were still the use of foreign materials, humans, and reflect the anatomical, physiological, and behav- as well as complicated production and waiting periods of at ioral makeup of humans [3,12,13]. least 4–6 weeks, for use of these grafts [4,5]. Thus, there is still a worldwide shortage of small-diameter (<6 Material and Methods mm) conduits with sufficient patency rates that can be used to bypass or replace small peripheral diseased arteries [2,3,6]. All experiments were performed in accordance with the Autologous arteries are still the criterion standard for vascular “Principles of the Laboratory Animal Care” prepared by the replacement due to their inherent physiological properties [7]. National Society of Medical Research, and the “Guide for Care and Use of Laboratory Animals” developed by the US National Tissue engineered vessels can potentially be used to replace Institutes of Health, (NIH, revised 1996) and the South African diseased and damaged native blood vessels [8]. Decellularized laws for animal welfare. Ethics approval for this study was ob- biological scaffold material from both xenograft and allograft tained from the Interfaculty Animal Ethics Committee of the origin can be used in constructing tissues and organs to re- Faculty of Health Sciences of the University of the Free State store or establish normal function [4,6,7], aiming to develop liv- (NR 17/2014). Two male baboons (Papio ursinus) weighing ing autologous grafts with the capacity for growth, repair, and between 10 and 15 kg were anaesthetized with 0.01 ml/kg remodelling. Decellularized allograft tissue can also attenuate Rompun (Bayer, Johannesburg, South Africa) and 10 mg/kg immune response-related degeneration as a result of chronic Ketamine hydrochloride (SigmaAldich, Johannesburg, South rejection by recipients [4]. However, thrombogenicity remains Africa) and euthanized with an overdose of potassium chloride a major concern, as decellularized arteries have no endothe- (Bayer, Johannesburg, South Africa). Heparin (17 IU/kg, Pfizer, lial lining, thus exposing collagen fibers to circulating blood. Johannesburg, South Africa) was given before euthanasia. Twelve This direct exposure of collagen results in thrombosis due to medium-sized arteries with diameter smaller than 6 mm and a platelet adhesion and activation in circulating blood [9]. The minimum length of 6 cm were harvested from different regions absence of an endothelial lining is also associated with accel- of the 2 sacrificed baboon bodies. Six arteries were collected erated vessel calcification and degeneration [10]. from each baboon. The main source of blood vessels was from arteries branching from the aorta: the right (R) and left (L) ca- Surface coating of small-diameter grafts with angiogenic growth rotid, R and L radial, and R and L femoral arteries. Ten of the factors has some promise but does not solve the problem com- baboon arteries were decellularized. Two femoral arteries were pletely due to the inability to form a monolayer [11]. Recently, used as controls. Six decellularized arteries (2 from each group) a more promising approach for construction of small-diame- were seeded with HUVECs (Clonestics™ HUVEC systems, Lonza ter vascular grafts is the re-endothelialization of decellularized Walkersville, Inc., MD, USA). The remaining 2 decellularized ba- vascular constructs with autologous vascular endothelial cells boon arteries were not seeded. HUVECs were used because they before implantation [1,10]. Other recent investigations have are a readily available cell source for tissue engineering without shown that decellularized scaffolds have no negative effect sacrificing intact vascular tissues. Many other studies have suc- on cell seeding [2]. The endothelial layer incorporates many of cessfully used them for seeding vascular grafts [9]. Furthermore, the anti-thrombogenic properties of blood vessels. However, they are cost-effective and yield large quantities. 209 This work is licensed under Creative Common Attribution- 345 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: IN VITRO STUDIES The anti-thrombotic effect of re-endothelialized baboon arteries 210 © Med Sci Monit Basic Res, 2017; 23: 344-351 Decellularization A density of 2.5×103 endothelial cells/cm2 were used to seed the freshly coated graft surfaces. The EC culture was suspend- Ten baboon arteries (carotid (n=4), radial (n=4) and femo- ed into the sutured graft within the bioreactor and the air was ral (n=2)) were washed with sterile phosphate-buffered sa- removed. After substituting ECs, the bioreactor was placed into line (PBS, pH 7.4; Invitrogen, Carlsbad, CA) to remove residu- a biostrabilizor (Biegler Medizinelektronik GmbH, Mauerbach, al blood clots. Decellularization was accomplished using the Austria) to perform a standardized EC [16]. The bioreactor was acid- and detergent-based method [4]. Arteries were decellu- rotated to expose the entire luminal surface of the arteries to larized using a combination of 1% sodium deoxycholic acid, achieve optimal attachment conditions. After 3 h of seeding, 0.05% sodium dodecyl sulphate, and 0.05% triton-X100 at 37°C. the artery constructs were rinsed with PBS to remove non-ad- Extensive rinsing steps with saline followed this. herent cells. The arterial grafts were then maintained in fresh culture medium overnight at 37°C in a 5% CO2 incubator to A control sample of a circumferential 0.5 cm resection from all allow ECs to grow onto the arteries. This was performed for explanted arteries was obtained for histology, scanning elec- an additional 7 days, while changing the medium every 48 h. tron microscopy (SEM), and transmission electron microsco- We cut small circular pieces (0.5 cm) from each seeded artery py (TEM) analysis before and after decellularization. We then after days 1 and 7 of the seeding procedure to verify the suc- stored the arteries at 4ºC under sterile conditions in PBS con- cess of endothelialization using SEM analysis. taining penicillin (200 U/ml), streptomycin (200 µg/ml), ampho- tericin B (10 µg/ml), ciprofloxacin (50 µg/ml), and cefuroxime Cell viability on the scaffold was determined with the same (750 µg/ml) (ScienCell Research Laboratories, Carlsbad, CA). MTT assay kit (MTT Cell Viability and Proliferation Assay Kit, ScienCell, Carlsbad, CA). Two small circular pieces (0.5 cm) Cell culture and seeding from each seeded artery were cut; the cells were trypsinized and then cultured in a 24-well plate for 24 h. HUVECs from Laboratory procedures the original cell culture were used as a positive control. Only viable cells were counted. HUVECs were cultured according to the manufacturer’s in- structions (Clonestics™ HUVEC systems; Lonza Walkersville, Perfusion study Inc., MD, USA) in a laminar flow cabinet under sterile condi- tions. The HUVECs were passaged only 3 times for the pur- Blood samples (50 ml) were collected from 4 healthy baboons pose of this study, to retain the unique function of endotheli- in 3.2% sodium citrate and used within 4 h for in vitro perfu- al cells [14]. To optimize cell culturing, the morphology of the sion. An in vitro flow chamber connected to a peristaltic pump cells was assessed for homogenous cobblestone morphology was used to provide a closed system, which delivers laminar present throughout the culture, but no pictures were taken of flow to the arteries. The luminal surfaces of 2 normal, 4 decel- the cells. Cell viability and proliferation rate were determined lularized, and 4 seeded arteries were first washed (2×5 min) by the MTT assay (MTT Cell Viability and Proliferation Assay with 50 ml of PBS supplemented with penicillin and strepto- Kit, ScienCell, Carlsbad, CA). mycin. Then, 50 ml of blood was circulated for 2 h at 37°C with a pressure of 120/80 mmHg through the sutured baboon ar- First, the luminal surfaces of the 6 decellularized arteries (ca- tery. Calcium chloride (0.6 M) was added to the blood at the rotid (n=2), radial (n=2), and femoral (n=2)) were coated with time perfusion was started. After perfusion, the vessels were a solution of 10 µg/ml fibronectin (Human plasma fibronectin, washed once for 5 min with PBS. Small circular pieces (0.5 Gibco®, Grand Island, NY) in 1 ml PBS. The arteries were then cm) were cut for SEM and TEM analysis. SEM and TEM analy- incubated at 37°C for 45 min to allow the fibronectin to bind sis were done to assess the morphological differences of ar- onto the ECM. After that, we aspirated the unbound fibronectin teries before and after the decellularization, re-endothelializa- and washed the arteries twice with sterile PBS. The decellular- tion, and the perfusion experiment. TEM analysis was done ized coated arteries were immediately seeded using a bioreac- according to standard diagnostic methods that are widely used tor filled with endothelial growth medium (Clonestics™ EGM-2 for diagnostic purposes. TEM showed the presence and condi- Bullet kit medium, Lonza Walkersville, Inc., MD) and kept in a tion of the basement membrane on the luminal surface of the 5% CO2/air atmosphere incubator at 98% humidity at 37°C, decellularized arteries. Although histology studies were done as previously described [15]. The pH level was maintained at as well, this is not included in this report, since the SEM and 7.4 by modulating the CO2 supply. The static environment pro- TEM analysis shows more detail. vided a low shear stress environment during the seeding pro- cess. The artery constructs were then washed twice with PBS and supplemented with penicillin and streptomycin to limit any contamination on the vascular grafts. 210 This work is licensed under Creative Common Attribution- 346 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: The anti-thrombotic effect of re-endothelialized baboon arteries IN VITRO STUDIES © Med S2ci1 M1onit Basic Res, 2017; 23: 344-351 Electron microscopy surface, indicating that there are EC on the luminal surface. The decellularized artery had an intact basal membrane with Scanning electron microscopy limited areas with exposed collagen fibers, proving the ab- sence of an endothelial lining. Each vascular graft tissue sample for SEM was prepared by the Centre for Microscopy at the University of the Free HUVECs culture State. All samples were fixed in 2.5% glutaraldehyde (Merck, Johannesburg, South Africa). Tissue specimens were dried using Although not shown, HUVECs had the typical cobblestone mor- the critical point method (Tousimis critical point dryer, Rockville, phology of ECs in growing cultures. Their cell viability in cul- MD, USA, ethanol dehydration, and carbon dioxide drying gas) tures exceeded 90% (Table 1). and were metallized using gold (BIO-RAD, Microscience Division Coating System, London, UK; Au/Ar sputter coating @ 50-60 Seeding nm). Evaluations were performed with a Shimadzu SSX 550 scanning electron microscope (Kyoto, Japan, with integral im- Scanning electron microscopy (SEM) analysis of seeded aging). The surface area of each specimen was examined and arteries photographed in different positions. SEM micrographs were used to assess endothelial integrity and to evaluate the qual- The decellularized arteries supported re-endothelialization ity of the extracellular basal membrane. (Figure 3). Endothelial cells adhered to the decellularized ar- tery direct after seeding (Figure 3A). Figure 3B shows the pro- Transmission electron microscopy liferation and migration of the ECs on the decellularized artery after 1 day of seeding (indicated by the red arrows), forming Vascular graft samples were fixed in 3.0% glutaraldehyde over- an almost confluent monolayer. The ECs had formed an almost night, post-fixated in Palade’s osmium tetroxide, and dehydrat- confluent monolayer in the middle section of the arterial con- ed in a graded acetone series [15]. Dehydrated samples were struct. Seven days after seeding, a complete endothelial layer impregnated/embedded in epoxy to facilitate the making of formed on the surfaces of the arteries (Figure 3C). ultra-thin sections for the TEM evaluation. Ultra-thin sections were cut from the sample embedded in the epoxy using an Viability of seeded HUVECs ultra-microtome (Leica Ultracut UC7, Vienna, Austria). After sectioning the samples, they were stained with uranyl acetate Seeded HUVECs showed increased mitochondrial activity with and lead citrate. Sections of the leaflet samples were evaluat- an increased number of cells (Figure 4). ed by using a transmission electron microscope (CM100, FEI, The Netherlands) and photographed using an Olympus Soft Perfusion Imaging System Megaview III digital camera with Soft Imaging System digital image analysis and documentation software Scanning electron microscopy of perfused arteries (Olympus, Tokyo, Japan). Figure 5 represents SEM images of a decellularized artery (a), a normal artery (b), and a seeded decellularized artery (c) after Results perfusion with whole blood. The normal artery and the seed- ed decellularized arteries were devoid of thrombi on their lu- Decellularization efficacy minal surfaces. There were, however, areas on both with few isolated spots of platelets adhesion. This might be due to pos- Transmission electron microscopy analysis sible damage to the endothelial layer or mishandling of the ar- teries during the seeding process. However, the decellularized TEM examination of normal baboon arteries showed a normal arteries (a) had more platelet adhesion and activation on the endothelial lining (Figure 1A) with an intact basement mem- ECM after perfusion with whole blood, indicating that the de- brane. In a decellularized artery (Figure 1B), the endothelial cellularized scaffold promotes thrombosis. monolayer was absent. The decellularized artery contained cellular debris and cellular components. Discussion Scanning electron microscopy analysis The decellularized arterial scaffolds contained some cellu- SEM clearly showed differences between normal and decel- lar debris but no cells were present, as shown by TEM and lularized arteries (Figure 2). The normal artery had a smooth SEM (Figures 1, 2). SEM further confirmed the absence of an 211 This work is licensed under Creative Common Attribution- 347 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: IN VITRO STUDIES The anti-thrombotic effect of re-endothelialized baboon arteries 212 © Med Sci Monit Basic Res, 2017; 23: 344-351 A B Figure 1. TEM images of a normal artery (A) and a decellularized artery (B). Both a and b were taken at 1250× magnification. a shows an endothelial monolayer of a normal blood vessel indicated by the red arrows, with an intact basement membrane indicated by the blue arrows (n=2). (B) Shows the missing endothelial layer of a decellularized artery. There was cellular debris on the intimal surface (purple arrows) and ruptured remaining cellular material (orange arrows) throughout the matrix scaffold. It also has an intact basement membrane (dark blue arrows) (n=4). A B Figure 2. S EM images of a normal baboon artery (A) (n=2) and a de-endothelialized baboon artery (B) (n=4). (Images at 1000× magnification). Table 1. Total cell counts and percentage cell viability of the primary culture, after the first passage and after the second passage. Values are expressed as mean (n=2). The total cell count increased with each passage. The percentage viability of the HUVEC cells remained high in the primary culture and increased with each passage. Primary culture First passage Second passage Total cell count * 106 0.97 1.99 2.75 % Viability 94 96 97 endothelial lining. It is important to note that the basement [6]. Furthermore, a preserved extracellular matrix contributes membrane was intact after decellurization (Figure 1). This is an to maintaining a non-thrombotic environment in graft mate- important prerequisite to re-endothelialization because it mod- rial if used in bypass surgery [2]. The decellurization process ulates cell-matrix interactions by supporting cell adhesion, mi- needs to be refined since cellular debris could still be observed. gration, and proliferation during development and regeneration It is, however, important not to damage the basal membrane. 212 This work is licensed under Creative Common Attribution- 348 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: The anti-thrombotic effect of re-endothelialized baboon arteries IN VITRO STUDIES © Med S2ci1 M3onit Basic Res, 2017; 23: 344-351 A B C Figure 3. SEM images of a seeded decellularized artery. Attached ECs are visible on the scaffold after seeding (A: 1000x magnification, n=3). After day 1, ECs started to form a confluent monolayer and appeared to be migrating towards the direction of the red arrows in image B (B: 450× magnification, n=3). Notice the smooth surface on the seeded area of the decellularized scaffold compared to the rough unseeded area of the artery in image (B). After 7 days of seeding, image (C) indicates a complete endothelial layer on the surface of the arterial sections (magnification 100×, n=3). Notice the smooth surface of the endothelium. The internal elastic lamina separating the tunica intima from the tunica media could be observed from the cross-section. This confirmed the presence of the endothelium. In vitro seeding techniques using cultured autologous cells are required to endothelialize smaller vascular constructs before 0.8 implantation, since in vivo repopulation of smaller decellular- ized conduits causes thrombosis [15]. We used HUVECs to suc- 0.6 cessfully re-endothelialize the de-endothelialized baboon ar- teries (Figure 3) because HUVECs are mostly used to obtain 0.4 ECs for seeding vascular grafts surfaces [11]. Other sources of ECs include endothelial progenitor cells derived from peripheral 0.2 blood, bone marrow, and umbilical cord vein ECs [5]. We used HUVECs cell cultures because they are readily available com- 0.0 mercially. They can also be easily cultured under optimal con- 0 1000 2000 3000 ditions with a good growth potential. EC cultures were viable Number of cells per well and had excellent proliferation capabilities (Table 1). Figure 4. MTT assay of seeded HUVECs after 24-h cell culturing The decellularized baboon arterial grafts were successfully from trypsinized small circular pieces (0.5 cm) of the seeded with HUVECs. The seeded grafts were endothelialized seeded arteries (n=2). The seeded HUVECs showed along their entire length from day 1 (Figure 3B), and a conflu- increased mitochondrial activity with increased number of cells. ent monolayer of ECs was observed after 7 days (Figure 3C). Although a slight breakage of the endothelial layer occurred due to twisting during sectioning, a confluent monolayer could human factor VIII antibodies. Response of baboon platelets to still be observed. collagen ristocetin is similar to humans; however, response to ADP and EPI is slightly reduced and response to arachidonic Baboon vessels were used because the hemostatic mecha- acid is slightly increased. nisms closely resemble that of humans, whereas other ani- mal species such as dogs, sheep, and pigs do not. Their vas- The aim of this study was to determine if re-endothelialized, cular EC growth characteristics are also thought to be similar decellularized biological scaffolds could reduce thrombogenicity to that of humans. They also share about 98% homology to when compared to the decellularized scaffolds. Decellularized human genes and possess similar protein structures [17]. scaffolds promote thrombosis formation due to the absence Furthermore, the baboon immunological and coagulation sys- of an endothelial lining responsible for maintaining an an- tem is closest to that of humans. Baboons show similar fibrin- ti-thrombotic surface [17,18]. The perfusion studies support ogen level and thrombin time to that of humans. FVIII activity this finding, as unseeded decellularized arteries showed wide- is also similar to humans, and FVIII antigen cross-reacts with spread platelet activation during the perfusion experiment 213 This work is licensed under Creative Common Attribution- 349 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Mean OD (570–690) Meiring M. et al.: IN VITRO STUDIES The anti-thrombotic effect of re-endothelialized baboon arteries 214 © Med Sci Monit Basic Res, 2017; 23: 344-351 A B C Figure 5. SEM results of decellularized (A), normal (B), and seeded (C) arteries after perfusion with whole blood. Image (A) (2000× magnification) demonstrates platelet adhesion and activation on the surface of a decellularized artery (red arrows) (n=4). (B, C) (450× magnification) represent normal (n=2) and seeded (n=4) arteries, respectively. There were few locations (red arrows) of minimal platelet adhesion. with baboon blood (Figure 5). The seeded arterial grafts and Study limitations the healthy artery, on the other hand, showed no thrombus formation. However, there were areas where few isolated Limitations of the present study include the small number of platelets adhered to the surface. This might have been due to arteries that were seeded. Since this study was done to prove damaged ECs, thus exposing collagen fibers or the underly- that baboon arteries can be seeded with human endothelial ing basal membrane in those areas. A similar result of plate- cells, we did not use many arteries. A follow-up study would let adhesion on areas where the EC lining that was not intact be worthwhile with a large number of small arteries, in which is found in the literature [11,17]. In addition, exposed colla- autologous endothelial cells or progenitor cells with increased gen following vascular injury leads to immediate activation of growth potential will be isolated from blood and seeded onto platelets through the activation of the coagulation pathway these arteries. to seal the wound. It is therefore not surprising that the ex- posed fibrous structures of decellularized arteries would pro- mote platelet activation. Future Prospects and Conclusions It is important to note that perfusion did not dislodge the seed- HUVECs were successfully seeded on decellularized baboon ar- ed endothelial cells (Figure 5). Endothelialization of vascular teries. The decellularization did not alter the morphology of the grafts is limited by the inability of endothelial cells to remain extracellular matrix of the arteries and, importantly, the bas- attached to the scaffold after exposure to flow [1]. The seed- al membrane remained intact. Endothelialization clearly pre- ing process used in this study was clearly effective in creat- vented thrombus formation on the decellularized arterial scaf- ing an intact endothelial cell lining that withstood high shear fold surfaces after perfusion with whole blood at high shear stress during whole blood perfusion. It is also encouraging that rate. Importantly, the 2-h perfusion did not damage the seed- endothelial viability could be demonstrated after the perfu- ed endothelial cells, and these results were similar to those sion study (Figure 5). obtained in the perfused control. It is therefore possible to speculate that the findings in a primate model can be extrap- olated to humans, and this topic warrants further investigation. Conflict of Interest None. References: 1. Fukunishi T, Best CA, Sugiura T et al: Tissue-engineered small diameter ar- 2. Bäcker H, Polgár L, Soós P et al: Impedimetric analysis of the effect of de- terial vascular grafts from cell-free nanofiber PCL/chitosan scaffolds in a cellularized porcine heart scaffold on human fibrosarcoma, endothelial, sheep model. PLoS One, 2016; 11(7): e0158555 and cardiomyocyte cell lines. Med Sci Monit, 2017; 23: 2232–40 214 This work is licensed under Creative Common Attribution- 350 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] Meiring M. et al.: The anti-thrombotic effect of re-endothelialized baboon arteries IN VITRO STUDIES © Med S2ci1 M5onit Basic Res, 2017; 23: 344-351 3. Keough EM, Callow AD, Connolly RJ et al: Healing pattern of small calibre 11. Kakisis JD, Liapis CD, Breuer C, Sumpio BE: Artificial blood vessel: The holy dacron grafts in the baboon: An animal model for the study of vascular grail of peripheral vascular surgery. J Vasc Surg, 2005; 41: 349–59 prostheses. J Biomed Mater Res, 1984; 18(3): 281–92 12. Shi Q, Hodara V, Simerly CR et al: Ex vivo reconstitution of arterial endotheli- 4. Dohmen PM, Pruss A, Koch C et al: Six years of clinical follow-up with en- um by embryonic stem cell-derived endothelial progenitor cells in Baboons. dothelial cell-seeded small-diameter vascular grafts during coronary by- J Cell Mol Med, 2013: 22(4): 631–42 pass surgery. J Tissue Eng, 2013; 4: 2041731413504777 13. Sherpard AD, Connolly RJ, Callow AD et al: Endothelial cell seeding of small 5. Gabbieri D, Dohmen PM, Koch C et al: Aortocoronary endothelial cell-seed- calibre synthetic vascular prostheses in the primate: Sequential indium 111 ed polytetrafluoroethylene graft: 9-year patency. Ann Thorac Surg, 2007; platelet studies. Surgery Forum, 1984; 35: 432–34 83: 1166–68 14. Spurr AR: A low viscosity epoxy resin embedding medium for electron mi- 6. Deutsch M, Meinhart J, Fischlein T et al: Clinical autologous in vitro endo- croscopy. J Ultrastr Res, 1969; 26: 3–43 thelialization of infrainguinal ePTFE grafts in 100 patients: A 9-year expe- 15. Cox LA, Comuzzie AG, Havill LM et al: Baboons as a model to study genet- rience. Surgery, 1999; 126: 847–55 ics and epigenetics of human disease. ILAR J, 2013; 54: 106–21 7. Jordan JE, Williams JK, Lee SJ et al: Bioengineered self-seeding heart valves. 16. Dohmen PM, da Costa F, Yoschi S et al: Can autologous vascular endothelial J Thorac Cardiovascr Surg, 2012; 143: 201–8 cell seeding increase the patency rate of small-diameter No-React-treated 8. Smit FE, Dohmen PM: Cardiovascular tissue engineering: Where we come bovine internal mammary arteries? An in vivo study in juvenile sheep. Med from and where are we now? Med Sci Monit Basic Res, 2015; 21: 1–3 Sci Monit, 2007; 13: BR188–93 9. Barron V, Lyons E, Stenson-Cox C et al: Bioreactors for cardiovascular cell 17. Conklin BS, Richter ER, Kreutziger KL et al: Development and evaluation of and tissue growth: A review. Ann Biomed Eng, 2003; 31: 1017–30 a novel decellularized vascular xenograft. J Med Eng Phys, 2002; 24: 173–83 10. Heyligers JMM, Arts CHP, Verhagen HJM et al: Improving small-diameter 18. Schlegel F, Appler M, Halling M et al: Reprogramming bone marrow stem vascular grafts: From the application of an endothelial cell lining to the cells to functional endothelial cells in a mini pig animal model. Med Sci construction of a tissue-engineered blood vessel. Ann Vasc Surg, 2005; 19: Monit Basic Res, 2017: 23: 285–94 448–56 215 This work is licensed under Creative Common Attribution- 351 Indexed in: [Index Medicus/MEDLINE] [EMBASE/Excerpta Medica] NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) [Chemical Abstracts/CAS] S-23 216 MEDICAL TECHNOLOGY SA The use of an algorithm for the laboratory diagnosis of von Willebrand disease Corresponding Author: MURIEL MEIRING, Ph D Department of Haematology and Cell Biology, University of the Free State and NHLS, Bloemfontein, 9300 Other Authors: PHILIP N BADENHORST, MD Department of Haematology and Cell Biology, University of the Free State and NHLS, Bloemfontein, 9300 MARELI KELDERMAN, Diploma in Medical Technology Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, 9300 ABSTRACT Von Willebrand Disease (vWD) is a bleeding disorder caused by either quantitative (type 1 and 3) or qualitative (type 2) defects of von Willebrand factor (vWF). No single test is available that provides appropriate information about the various functions of vWF and the laboratory diagnosis of vWD is based on a panel of tests that includes the measurement of factor VIII coagulant activity (VIIIC), vWF antigen (vWF:Ag), vWF activity as measured by ristocetin cofactor activity (vWFR:Co), vWF multimer analysis, ris tocetin induced platelet agglutination (RIPA), the factor VIII binding assay of plasma vWF and the bleeding time. Due to the het erogeneity of vWF defects and the variables that interfere with vWF levels, a correct diagnosis of types and subtypes may sometimes be difficult but is very important for therapy. Furthermore, the Ristocetin Cofactor test and the RIPA test are based on platelet agglu tination in reaction with the non physiological antibiotic, ristocetin. These tests also have low sensitivity and are difficult to stan dardise. Therefore, several analyses (tests) are required to diagnose vWD and it is important to take the pitfalls that these tests are subject to in consideration in the diagnosis of vWD. In this article, the laboratory diagnosis of vWD is presented on patients with type 1, 2A, 2B and 2M vWD. The diagnosis is done by using an algorithm that is proposed by the guidelines for diagnosis and treatment of vWD in Italy. The pitfalls in this diagnosis of vWD are outlined by 4 other patients. Introduction each patient with the STart 4 coagulation monitor, (Diagnostica Stago, Von Willebrand Disease (vWD) is a bleeding disorder caused by a quantita Asnières, France). Furthermore, the blood group of each patient was deter tive or qualitative defect of vWF. vWF is a high molecular weight glycopro mined. Plasma factor VIII levels were also determined on the ACL coagu tein that plays an essential part in the early phases of haemostasis by pro lation analyser (Instrumentation Laboratories, Italy). moting platelet adhesion to the sub endothelium and platelet aggregation VWF antigen (vWF:AG) under high shear stress conditions [1]. vWF is also the carrier of factor VIII The vWF concentration in plasma was measured with an enzyme linked in plasma and a deficiency or abnormality of vWF also results in an impair immuno adsorbent assay (ELISA). ment of blood coagulation. By the non covalent interaction between vWF Ristocetin cofactor assay (vWF:RCO) and factor VIII, factor VIII is protected against binding to membrane surfaces This assay is done with the ristocetin cofactor assay kit from Helena and to proteolytic attack by a variety of serine proteases, including activated Laboratories (France). Formalin fixed washed platelets do not agglutinate in protein C [2]. In the majority of cases, vWD is a congenital disease that is the presence of the antibiotic ristocetin unless normal plasma is added as a inherited in an autosomal dominant fashion. Patients with vWD may have a source of vWF. The agglutination follows a dose response curve that is mild, moderate or severe bleeding tendency since childhood, usually propor dependent on the amount of plasma vWF added. The test was done on the tional to the degree of the vWF defect. Inherited vWD has been subdivided Monitor IV Plus platelet aggregometer (Helena Laboratories, France). A into three types that reflect its pathophysiology. Types 1 and 3 vWD reflect standard curve of calibrated human plasma is used as the standard against respectively, the partial or virtually complete deficiency of vWF. Type 2 vWD which the patient’s plasma is measured. is a defect that is subdivided into 4 subtypes (2A, 2B, 2M and 2N). Type 2A Ristocetin induced platelet agglutination (RIPA) refers to variants with decreased platelet dependent function and is associat RIPA is measured by mixing different concentrations of ristocetin ranging ed with the absence of high molecular weight multimers. Type 2B refers to from 0.2 to 2 mg/ml with increments of 0.1 mg/ml with the patient’s platelet variants with increased affinity for platelet glycoprotein 1b. Type 2M refers rich plasma (PRP) in an aggregometer. The results are expressed as the con to variants with decreased platelet dependent function not caused by the centrations of ristocetin (mg/ml) able to induce 30% agglutination. Types absence of high molecular weight multimers and Type 2N to variants with 2A and 2M vWD show a low response to ristocetin, i.e. ristocetin concen markedly decreased affinity for factor VIII. trations more than 1.2 mg/ml are needed to induce 30% agglutination. An The spectrum and severity of vWD is wide, ranging from few, doubtful important exception is type 2B vWD in which there is a hyper responsive haemorrhagic symptoms to severe life threatening bleeding episodes. This is ness to ristocetin, due to a higher than normal affinity of vWF to platelet due not only to the heterogeneous vWF gene which may impair its haemo GP1b [5]. In these patients low ristocetin concentrations (0.2 to 0.8 mg/ml) static function, but also to the influence exerted by other genes (e.g. those for are needed to induce 30% agglutination. ABO blood groups) [3]. In addition many acquired conditions, either physio vWF multimers logic (stress, pregnancy) or pathologic (inflammation), can induce fluctua The multimeric structure of vWF in plasma was determined by a highly sen tions in vWF levels [3]. This, highly variable clinical picture and the presence sitive and rapid method described by Krizek and Rick in 2000 [6]. This of many different defects in the vWF molecule, complicate the diagnosis of method utilises submerged horizontal agarose gel electrophoresis, followed vWD [3]. by transfer of the vWF onto a polyvinylidine fluoride membrane, and The guidelines for diagnosis and treatment of VWD in Italy propose the immuno localisation and luminographic visualisation of the vWF multimer use of an algorithm (Fig 2). We adopted these guidelines in our Haematology pattern. This method distinguishes type 1 from type 2A and 2B vWD. Type Clinic and tested the usefulness thereof. 2A shows a total absence of the high and intermediate molecular weight mul timers and type 2B shows an absence of only the high molecular weight mul Materials and Methods timers. All the other types or subtypes show a multimer pattern similar to Subjects that of normal plasma. Advantages to this method include rapid processing, Patients referred to the Haematology Clinic with a history of a bleeding ten simplicity of gel preparation, high sensitivity to low concentrations of vWF dency were screened for von Willebrand Disease. and elimination of radioactivity. Figure 1 shows multimer patterns of nor Sample collection mal plasma, type1, and 2A and 2B vWD. Blood samples were collected into two Vacutainer tubes containing 0.105M Factor VIII binding assay sodium citrate in a ratio of 1:9 with blood. Platelet poor plasma was prepared The capacity of plasma vWF to bind exogenous FVIII was measured with an by centrifugation of whole blood at 2000 g for 20 minutes at room tempera ELISA. A micro plate (Maxisorp, Nunc, Denmark) was coated by incuba ture. Samples were stored in polypropylene tubes at 70°C until analysed. All tion for 2 days at 4°C with 2 µg/ml of rabbit polyclonal antihuman vWF tests (except the vWF multimer test) were done on original aliquots that were (Dako, Denmark). After washing with Tris 50 mmol/l NaCl 100 mmol/l, not previously thawed. pH 8.0 (TBS) buffer containing 0.1% bovine serum albumin (BSA) and Screening tests 0.05% Tween, the wells were saturated with TBS containing 3% BSA. Then The platelet count was determined on the TECHNICON H1 blood cell 100 µl of serial dilutions of plasma from patients and normal pooled plasma analyser (Bayer Diagnostics, Germany). A bleeding time, prothrombin time were added and incubated overnight at 4°C. Each patient sample was tested (PT) and activated partial thromboplastin time (APTT) were determined on VOL. 19 NO. 1 JUNE 2005 15 216 217 MEDICAL TECHNOLOGY SA in six serial dilutions, the first being adjusted to 5% vWF antigen level. After multimers were absent in patient 7 and in patient 8, all other results indicate removal of endogenous FVIII using 350 mmol/l CaCl2 (10 min, twice), a type 2M vWD. Laffan et al. (2004) also found the RIPA normal in patients 70mU of recombinant FVIII were added to each well. After incubation for with an RCo value of less than 30%. And they found the RIPA only 2 hours at 37°C and washing, bound FVIII was quantified using 1 µg/ml of decreased in severe forms of vWD. peroxidase conjugated sheep polyclonal antihuman FVIII (Kordia, The Multimeric analysis in plasma is necessary to distinguish between type 2A Netherlands). After washing, immobilised VWF was measured using vWD (lack of the largest and intermediate multimers) and type 2M vWD (all 0.1µg.ml of peroxidase conjugated rabbit polyclonal anti humanVWF multimers present). Our method includes rapid processing, simplicity of gel (Dako). The colour was developed by addition of OPD and the OD was read preparation, high sensitivity to low concentrations of vWF and elimination at 490nm. Two reference curves were established in parallel, one for the of radioactivity. quantification of immobilised vWF and one for the quantification of bound Type 2N vWD can be suspected in case of discrepant values of factor VIII. rFVIII by diluting normal pooled plasma. For each plasma dilution, the val Thus, in case of factor VIII levels lower than the vWF:Ag, diagnosis of type ues of bound rFVIII were plotted against the amount of immobilised vWF. 2N vWD should be confirmed by the factor VIII binding assay. The slopes of the obtained regression lines reflect binding capacity of vWF Since vWD is such a complex diseases to diagnose, this systematic diag to FVIII. nosis process makes the diagnosis of vWD more accurate, which is very important for the treatment of the disease. There are however pitfalls in this Results diagnosis process that are due to the limitations in sensitivity, reproducibili Diagnosis of vWD ty and interlaboratory variability of the agglutination based RCo and RIPA The algorithm used for the laboratory diagnosis of vWD is outlined in Figure tests (Favaloro et al., 1999). 2. A proportional reduction of both vWF:Ag and vWF:RCo with a RCo/Ag Another functional assay that more laboratories are starting to use over the ratio >0.7 suggests type 1 vWD. If the RCo/Ag ratio <0.7, type 2 is diag last 7 years is the collagen binding assay (CBA) of vWF (Favaloro et al., nosed. Type 2B vWD can be identified in cases of an enhanced RIPA 2004). The CBA is based on the ability of the HMW multimers of vWF to (<0.8mg/ml) while type 2A and 2M in a low RIPA (>1.2mg/ml). Multimeric preferentially bind collagen. This is an ELISA based assay where dilutions analysis in plasma is necessary to distinguish between type 2A vWD (lack of of the patient’s plasma is added to a collagen coated ELISA plate and the largest and intermediate multimers) and type 2M vWD (all the multimers are amount of bound vWF evaluated using an anti HRP conjugated vWF anti present as in normal plasma). In type 1 vWD the ratio between factor VIII body. The values are expressed in U/dl, considering the optical density and vWF:Ag is always >1. When this ratio <1, type 2N vWD is suspected observed in the normal pooled plasma dilution as 100. This assay has been and this type of vWD can be confirmed by performing a factor VIII binding shown to be sensitive in the discrimination of Type 1 and types 2A and 2B assay. vWD (Favaloro & Koutts, 1997). Casonato et al., 2001 demonstrated that The results of patients who have been diagnosed with the different types the CBA was consistently more sensitive to large and intermediate vWF mul and subtypes of vWD are given in table 1. Patient 1 has low vWF:Ag and timer representation than the RCo assay, since none of type 1 vWD patients vWF: RCo values with a RCo/Ag ratio of 0.74, therefore >0.7, which sug studied showed CBA more decreased than the Ag levels. That was not the gests type 1 vWD. The RIPA is normal and high molecular weight multimers case with the RCo assay. The decreased values of the CBA in type 2A and are present. The ratio between factor VIII and vWF: Ag is also >1. Patient 2 2B patients were more consistent than that of the RCo assay. The CBA is has a very low RCo value with a RCo/Ag ratio of 0.44. This suggests type 2 however insensitive to type 2M vWD patients. vWD. The RIPA was decreased (30% agglutination occurs at 1.5 mg/ml ris In conclusion, it is crucial to use this systematic way to diagnose vWD. tocetin) and the high and intermediate vWF multimers were absent, which The CBA should be included as a useful diagnostic test in the profile of vWD indicates type 2A vWD. The multimeric pattern of this patient is showed in diagnosis to counteract the pitfalls in the diagnosis of this disease. figure 1. Patient 3 also has a RCo/Ag ratio <0.7 which suggest type 2 vWD. The RIPA is however increased (30% agglutination occurs at 0.6 mg/ml ris References tocetin) and the high molecular weight multimers are absent. This indicates 1. Tuddenham E.G. (1989). Von Willebrand factor and its disorders . An a type 2B vWD. Patient 4 is also a type 2vWD, since the RCo/Ag ratio is overview of recent molecular studies. Blood Reviews 3: 251 262. 0.55. The RIPA is decreased and the multimer pattern is normal. This patient 2. Vlot A.J., Koppelman S.J., Bouman B.N., Sixma J.J (1998). Factor VIII is therefore diagnosed as a type 2M vWD. and von Willebrand Factor. Thrombosis and Haemostasis 79: 456 465. The results of 4 patients where the diagnosis of the types and subtypes of 3. Dacie J.V., Lewis S.M. (1995). Practical Haematology. Eighth Edition. von Willebrand disease are not clear are outlined in table 2. Patient 5 has an 4. Ruggeri Z.M., Pareti F.I., Mannucci P.M., Ciavarella N, Limmerman RCo/Ag ratio <0.7 which suggest type 2 vWD. The multimers were normal T.S. (1980). Heightened interaction between platelets and factor VII and suggested a type 2B, but the RIPA, however, was normal. Other mem von Willebrand Factor in a new subtype of von Willebrand Disease. bers of this patient’s family were diagnosed with type 1 vWD. The RCo New England Journal of Medicine 302: 1047 1051. value (or vWF:Ag) could be wrong in this case. In patient 6 the RCo value 5. Krizek D.R., Rick M.E. (2000). A rapid method to visualize von was almost twice as much as that of the VWF:Ag. This indicates a “super” Willebrand Factor multimers by using Agarose Gel Electrophoresis, functional vWF that does not exist. All the other results of this patient were Immunolocalization and Luminographic Detection. Thrombosis normal. Patient 7 indicated a type 2 vWD, because of a RiCo/Ag ratio <0.7. Research 97: 457 462. The FVIII/Ag ratio was also discrepant and indicated a type 2N vWD. The 6. Federici A.B., Castman G., Mannucci P.M. (2002). Guidelines for the RIPA was also normal that fits in with a type 2N vWD. The HMW multimers diagnosis and management of von Willebrand Disease in Italy. , however, were absent, which indicates a type 2A vWD. Therefore the RIPA Haemophilia 8: 607 621. and/or the factor VIII level seem to be wrong in this case. Patient 8 also indi 7. Laffan M., Brown S.A., Collins P.W., Cuming A.M., Hill F.G.H., cates type 2 vWD since the RiCo/Ag ratio <0.7. The RIPA is however nor Keeling D., Peake I.R., Pasi K.J. (2004). The diagnosis of von mal and the HMW multimers were present. This indicates on the type 2M Willebrand disease: a guideline from the UK Haemophilia Centre vWD, but the RIPA cannot be explained. Doctors’ Organization. Haemophilia 10: 199 217. 8. Casonato A., Pontara E., Bertomoro A., Sartorello E., Girolai A (1999). Discussion Which assay is the most suitale to investigate von Willebrand Factor We have outlined a systematic way to diagnose vWD that is recommended functional assay? Thrombosis and Haemostasis 81: 994 995. by the International Society on Thrombosis and Haemostasis. It is important 9. Scott J.P., Montgomery R.R., Retzinger G.S. (1991). Dimeric ristocetin to note that the RCo/Ag ratio is necessary to distinguish between type 1 and flocculates proteins, binds to platelets, and mediates von Willebrand fac type 2 vWD. The RCo test however has a poor sensitivity (50%) (Scott et tor dependent agglutination of platelets. Journal of Biological al., 1991), is difficult to standardise (Casonato et al., 1999), and lack physi Chemistry 266: 8149 8151. ological analogue. It is however still the standard method for measuring 10. Favaloro E., Smith J., Petinos P., Hertzberg M., Koutts J. (1999). vWF activity that is approved by the Standardization Committee of the Laboratory testing for von Willebrand’s disease: and assessment of cur International Society on Thrombosis and Haemostasis (ISTH). The low sen rent diagnostic pratice and efficacy by means of a multi laboratory sur sitivity of the RCo assay makes diagnosis of vWD difficult as seen in vey. Thrombosis and Haemostasis 82: 1276 1282. patients 5 and 6. In patient 5 the family history and the results of the other 11. Favaloro E.J., Bonar R., Kershaw J., Sioufi M., Hertzberg M., Street A., tests indicate a type 1 vWD, but the RCo value was not proportionate to the Lloyd J., Marsden K. (2004). Laboratory diagnosis of von Willebrand’s Ag level. In patient 6 the RCo value was twice as much as the Ag level that disorder: quality and diagnostic improvement driven by peer review in indicates an over active vWF that does not exist in vWD patients. a multilaboratory test process. Haemophilia 10: 232 242. The RIPA test is necessary to distinguish between type 2B and type 2A 12. Favaloro E., Koutts J. (1997). Laboratory assays for von Willebrand vWD and the multimer pattern will indicate type 2A or 2M vWD. It is also factor: relative contribution to the diagnosis of von Willebrand’s disease. important to do the RIPA test with ristocetin concentrations ranging from 0.2 Patology 29: 385 390. mg/ml to 2 mg/ml with increments of 0.1 mg/ml. This will indicate at which 13. Casonato A., Pontara E., Bertomoro A., Sartorello F., Cattini M.G., concentrations maximum agglutination value of 30% occur. The RIPA test Girolami A. (2001). Von Willebrand factor collagen binding activity in that is currently used by most laboratories in South Africa is done with only the diagnosis of von Willebrand disease: an alternative to ristocetin 3 ristocetin concentrations (0.5, 1.0 and 1.25 mg/ml). It is sometimes diffi cofactor activity? British Journal of Haematology 112: 578 583. cult to determine if the RIPA is enhanced or reduced on only 3 values. This was seen in patients 7 and 8 where the RIPA seems normal, but the HMW 16 JUNE 2005 VOL. 19 NO. 1 217 218 MEDICAL TECHNOLOGY SA Figure 1 The multimeric structure of vWF in normal plasma (NP) (lines1 and 2) Type 2B vWD (lines 3 and 4), Type 2A vWD (lines 5 and 6) and type 1vWD (lines 7 and 8). Figure 2 Algorithm for the laboratory diagnosis of vWD VOL. 19 NO. 1 JUNE 2005 17 218 219 MEDICAL TECHNOLOGY SA TABLE 1 Laboratory results of 4 patients with different type and subtypes of von Willebrand disease. TABLE 2 Laboratory results of 4 patients where the diagnosis of the types and subtypes of von Willebrand disease are not clear. 18 JUNE 2005 VOL. 19 NO. 1 219 Meiring_subbed.qxp 11/8/09 10:26 am Page 33 220 Coagulation Disorders S-24 Laboratory Diagnosis of von Willebrand Disease Murie l Meir ing, 1 Phi l ip N Badenhorst 2 and Marel i Kelderman 3 1. Associate Professor and Specialist Scientist; 2. Professor and Chairman; 3. Medical Technologist, Department of Haematology and Cell Biology, University of the Free State DOI: 10.17925/EOH.2009.03.1.33 Abstract von Willebrand disease (VWD) is a bleeding disorder caused by either quantitative (type 1 and 3) or qualitative (type 2) defects of von Willebrand factor (VWF). No single available test provides appropriate information about the various functions of VWF, and the laboratory diagnosis of VWD is based on a panel of tests, including the measurement of factor VIII coagulant activity (FVIIIC), VWF antigen levels (VWF:Ag), VWF activity as measured by the ristocetin co-factor activity (VWF:RCo), the collagen-binding activity of VWF (VWF:CB), VWF multimer analysis, ristocetin- induced platelet agglutination (RIPA), the factor-VIII-binding assay of plasma VWF and VWF propeptide levels. Due to the heterogeneity of VWF defects and the variables that interfere with VWF levels, a correct diagnosis of types and subtypes may sometimes be difficult, but is very important for therapy. Furthermore, the RCo assay and the RIPA test are based on platelet agglutination in reaction with the non-physiological antibiotic ristocetin. These tests also have low sensitivity and are difficult to standardise. Therefore, several analyses (tests) are required to diagnose VWD and it is important to be aware of the pitfalls to which these tests are subjected in terms of the diagnosis. In this article, the laboratory diagnosis of patients with type 1, 2A, 2B, 2M, 2N and 3 VWD will be explained by using a modified algorithm that was first proposed by the guidelines for diagnosis and treatment of VWD in Italy. Keywords von Willebrand factor, von Willebrand disease, classification, subtypes Disclosure: The authors have no conflicts of interest to declare. Received: 14 May 2009 Accepted: 4 June 2009 Correspondence: Muriel Meiring, Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, 9300, South Africa. E: gnhmsmm.md@ufs.ac.za von Willebrand disease (VWD) is a bleeding disorder caused by Type 3 VWD is the least common subtype and reflects a virtually quantitative or qualitative defects of von Willebrand factor (VWF). VWF complete absence of VWF. Again, DDAVP treatment is not effective in is a high-molecular-weight (HMW) glycoprotein that plays an essential type 3 VWD. The current treatment of choice for type 3 VWD, as well part in the early phases of haemostasis by promoting platelet as most type 2 VWD, is transfusional therapy with plasma-derived adhesion to the subendothelium and platelet aggregation under high FVIII/VWF concentrates.3 shear stress conditions.1 VWF is also the carrier of factor VIII (FVIII) in plasma, and a deficiency or abnormality of VWF also results in an Type 2 VWD is a quantitative defect that is subdivided into four impairment of blood coagulation. By the non-covalent interaction subtypes (2A, 2B, 2M and 2N VWD). Type 2A refers to variants with between VWF and FVIII, FVIII is protected against binding to decreased platelet-dependent function and is associated with the membrane surfaces and to proteolytic attack by a variety of serine absence of HMW multimers; type 2B refers to variants with proteases, including activated protein C.2 In the majority of cases, increased affinity for platelet glycoprotein 1bα; type 2M refers to VWD is a congenital disease that is inherited in an autosomal- variants with decreased platelet-dependent function not caused by dominant fashion. Patients with VWD may have had a mild, moderate the absence of HMW multimers; and type 2N refers to variants with or severe bleeding tendency since childhood, usually proportional to markedly decreased affinity for FVIII. DDAVP treatment is not the degree of the VWF defect. Inherited VWD has been classified into recommended in patients with type 2 VWD, as it would only increase three types that reflect its pathophysiology. Type 1 VWD is the dysfunctional VWF. characterised by partial quantitative deficiency of VWF. It is the most common type, with a prevalence of about 80%. The mechanisms Platelet-type VWD (PT-VWD) is a rare autosomal-dominant bleeding involved include reduced synthesis and secretion of VWF or increased disorder. The genetic defect is in platelets rather than VWF, and clearance of VWF from plasma.3 The treatment is simple, as the disease is characterised by abnormally high binding affinity of the 1-deamino-8-D-arginine vasopressin (DDAVP) causes the release of platelets to the VWF, similar to type 2B VWD.4 Therefore, most people structurally normal VWF from endothelial stores. However, in patients with PT-VWD are misdiagnosed as type 2B VWD. However, these may with increased clearance of VWF, DDAVP treatment would not be require different therapeutic management, so discrimination is clinically effective as VWF in plasma is cleared very quickly from the circulation. important. The spectrum and severity of VWD is wide, ranging from a Therefore, it is important to diagnose patients with an increased few doubtful haemorrhagic symptoms to severe life-threatening clearance rate of VWF. bleeding episodes. This is due not only to the heterogeneous VWF gene, © T O U C H B R I E F I N G S 2 0 0 9 220 33 Meiring_subbed.qxp 12/8/09 5:12 pm Page 34 Coag2u21lation Disorders Figure 1: Algorithm for a Laboratory Diagnosis of plasma samples should be thawed to 37ºC before performing the von Willebrand Disease diagnostic tests. Special care should be taken to ensure that no cryoprecipitate is present in the samples; therefore, it must be Plasma VWF:Ag dissolved before the tests are performed as it will influence the results.8 Present Absent Screening Tests Plasma VWF:RCo versus VWF:Ag Screening tests for bleeding disorders include a platelet count, a VWF:CB versus VWF:Ag Type 3 bleeding time, prothrombin time (PT) and activated partial Proportionate Discrepant thromboplastin time (APTT), plasma FVIII levels and the blood group (0.7–1.2) (<0.7) of a patient. These tests are usually carried out by a routine coagulation laboratory. Plasma FVIII versus VWF:Ag Type 1 Type 2 VWF:pp versus VWF:Ag Laboratory Tests The first line of tests includes the VWF concentration in plasma Discrepant >2 Proportionate Discrepant (VWF:Ag), the ristocetin co-factor assay (VWF:RCo) and the collagen If low levels are detected FVIII-binding assay binding assay (VWF:CB). VWF:Ag is measured with an enzyme-linked immunoabsorbent assay (ELISA). An ELISA plate is coated with a Type 1 Type 2N specific rabbit antihuman VWF antibody that captures the VWF to be With increased measured. The plasma to be measured is added in 1:50 and 1:100 clearance of VWF Type 2 dilutions in blocking buffer. Afterwards, a rabbit antihuman VWF Multimer analysis antibody conjugated to peroxidase is added. This antibody binds to the remaining free antigenic determinants of VWF, subsequently HMW multimers forming a ‘sandwich’. The bound enzyme peroxidase is revealed by RIPA Present Absent its activity in a pre-determined time on the substrate ortho- phenylenediamine (OPD) in the presence of hydrogen peroxide. After stopping the reaction with a strong acid, the intensity of the colour Type 2M Type 2A Type 2B produced bears a direct relationship to the VWF concentration Increased Decreased initially present in the plasma sample. A standard curve of calibrated (0.2–0.8mg/ml) (>1.2mg/ml) VWF:CB normal human plasma is used as the standard against which the patient’s plasma is measured. RIPA-mixing studies The RCo assay is performed with formalin-fixed washed platelets in Plasma defect Platelet defect an aggregometer with software suited for the test. Formalin-fixed washed platelets can be obtained commercially or can be self- Type 2B PT-VWD prepared from normal platelet-rich plasma (PRP). Washed platelets do VWF = von Willebrand factor; RCo = ristocetin co-factor; Ag = antigen; CB = collagen-binding; not agglutinate in the presence of the antibiotic ristocetin unless pp = propeptide; FVIII = factor VIII; HMW = high-molecular-weight; RIPA = ristocetin-induced normal plasma is added as a source of VWF. The agglutination follows platelet agglutination; PT = platelet type. a dose–response curve that is dependent on the amount of plasma VWF added. Usually, plasma concentrations of 1:2 and 1:4 in Tris- which may impair its haemostatic function, but also to the influence exerted by other genes (e.g. those for ABO blood groups).6 In addition, many acquired conditions – either physiological (stress, pregnancy) or pathological (inflammation) – can induce fluctuations in VWF levels).6 A functional assay that more laboratories This highly variable clinical picture and the presence of many different have been starting to use over the last defects in the VWF molecule complicate the diagnosis of VWD.6 The guidelines for diagnosis and treatment of VWD in Italy7 propose the use seven years is the collagen-binding assay of an algorithm (see Figure 1). We adopted and modified these of von Willebrand factor. guidelines in our VWD testing facility. Laboratory Diagnosis of von Willebrand Disease Sample Collection buffered saline (TBS) are used. A standard curve of calibrated human Blood samples must be collected into tubes containing 0.105M sodium plasma is used as the standard against which the patient’s plasma citrate in a ratio of 1:9 with blood. Platelet-poor plasma (PPP) is is measured. Another functional assay that more laboratories have prepared by centrifugation of whole blood at 2,000g for 20 minutes at been starting to use over the last seven years is the collagen-binding room temperature. Samples must be stored immediately after assay (CBA) of VWF.9 The CBA is based on the ability of the HMW centrifugation in polypropylene tubes at -70ºC until analysed. It is multimers of VWF that preferentially bind to collagen. This is an ELISA- important to note that a cryoprecipitate may form if plasma samples based assay where dilutions of the patient’s plasma (comparable to are stored at temperatures over -70ºC. Cryoprecipitate contains large the VWF:Ag 1:50 and 1:100 dilutions) are added to a collagen-coated quantities of VWF, especially HMW multimers.8 All tests must be ELISA plate. The type of collagen seems to be important, but performed on original aliquots that were not previously thawed, and discordance still exists about which type of collagen to use (type 1, 34 E U R O P E A N H A2E2M1A T O L O G Y Meiring_subbed.qxp 12/8/09 5:13 pm Page 35 222 Laboratory Diagnosis of von Willebrand Disease type 3 or a combination thereof).10 The amount of bound VWF is Figure 2: The Multimeric Structure of von Willebrand determined by using an anti-horseradish peroxidase (HRP)- Factor in Normal Plasma and von Willebrand Disease Types 1, 2A, 2B and 2M conjugated VWF antibody. The values are expressed in U/dl or per cent. A standard curve of calibrated human plasma is used as the 1 2 3 4 5 standard against which the patient’s plasma is measured. This assay has been shown to be sensitive in the discrimination of type 1, 2A and Large multimers 2B VWD.11 Casonato et al.12 demonstrated that the CBA was consistently more sensitive to large- and intermediate-VWF multimer representation than the RCo assay. The decreased values of the CBA Intermediate multimers in type 2A and 2B patients are more consistent than that of the RCo assay. However, the CBA is insensitive to type 2M VWD patients.13 Small multimers The ristocetin-induced platelet agglutination (RIPA) and the VWF multimer patterns are performed to diagnose the type 2 subtypes. RIPA is measured by mixing different concentrations of ristocetin ranging from 0.2 to 2mg/ml in increments of 0.1mg/ml with the patient’s PRP in an aggregometer. The results are expressed as the Source: Meiring et al., 2005.15 concentration of ristocetin (mg/ml) able to induce 30% agglutination. RIPA-mixing studies are performed to distinguish between type 2B Figure 3: Densitometric Tracing of Lane Sets 2–5 VWD and PT-VWD. In short, the patient’s PRP and PRP from a control A C 4,000 4,000 person are centrifuged and gently resuspended to 200x106 platelets Type 1 Type 2A per millilitre in the following manner: patient platelets/patient plasma; 3,000 3,000 normal platelets/normal plasma; control platelets/patient plasma; and 2,000 2,000 patient platelets/normal plasma. Different ristocetin concentrations 1,000 1,000 are added to each of the platelet suspensions and the agglutination is 0 0 measured in an aggregometer, similar to the RIPA assay.5 PT-VWD is 0 0.25 0.50 0.75 1.00 0 0.25 0.50 0.75 1.00 Rf Rf diagnosed when the RIPA-mixing studies confirmed a platelet origin. The multimeric structure of VWF in plasma is determined by a highly B D 4,000 4,000 sensitive and rapid method originally described by Krizek and Rick in Type 2B Type 2M 2000.14 This method utilises submerged horizontal agarose gel 3,000 3,000 electrophoresis followed by transfer of the VWF onto a polyvinylidine 2,000 2,000 fluoride membrane and immunolocalisation and luminographic 1,000 1,000 visualisation of the VWF multimer pattern. This method distinguishes type 1 from type 2A and 2B VWD. The density of the high-, 0 0 0 0.25 0.50 0.75 1.00 0 0.25 0.50 0.75 1.00 intermediate- and low-molecular-weight multimers of each multimer Rf Rf pattern is determined using a gel-documentation system.15 VWF Normal plasma Patient plasma multimer patterns from normal plasma and type 1, 2A, 2B and 2M VWD The density of the different bands is plotted against the relative front (Rf value) of the is shown in Figure 2 and the density graphs in Figure 3. different lanes. The capacity of plasma VWF to bind exogenous FVIII is measured with The VWF propeptide (VWF:pp) level in the patient’s plasma is also an ELISA. A microplate (Maxisorp, Nunc, Denmark) is coated by measured with an ELISA. Microtitre wells are coated with CLB-Pro 35 incubation for two days at 4°C with a rabbit polyclonal antihuman antibody and incubated with the plasma sample. Subsequently, the VWF. After washing with a TBS buffer of 0.1% bovine serum albumin wells are washed and the bound propeptide-containing protein is (BSA) and 0.05% Tween, the wells are saturated with TBS containing detected with CLB-Pro 14.3 coupled to peroxidase. Calibrators for the 3% BSA. Next, 100µl of serial dilutions of plasma from patients and VWF:propeptide assay are used to set up a standard curve against normal pooled plasma are added and incubated overnight at 4°C. which the patient’s plasma is measured.17 Each patient sample is tested in six serial dilutions, the first adjusted to 5% VWF-antigen level. After removal of endogenous FVIII using Diagnosis of von Willebrand Disease 350mmol/l CaCl2 (twice for 10 minutes), 70mU of recombinant FVIII is The algorithm used for the laboratory diagnosis of VWD is outlined in added to each well. After incubation for two hours at 37°C and Figure 1. A proportional reduction of both VWF:Ag and VWF:RCo with a washing, bound FVIII is quantified using 1µg/ml of peroxidase- RCo/Ag ratio >0.7 as well as a proportional reduction of both VWF:Ag conjugated sheep polyclonal antihuman FVIII. After washing, and VWF:CB with a CB/Ag ratio >0.7 suggest type 1 VWD. If type 1 VWD immobilised VWF is measured using a peroxidase-conjugated rabbit is diagnosed, it is important to determine the clearance rate of VWF; polyclonal antihuman VWF. The colour is developed by addition of consequently, the VWF:pp is performed. If the ratio between the OPD and the optical density read at 490nm. Two reference curves are VWF:pp and the VWF:Ag is more than two, an increased clearance rate established in parallel. One is for the quantification of immobilised of VWF is suspected in this patient. VWF and the other one for the quantification of bound rFVIII. For each plasma dilution, the values of bound rFVIII are plotted against the If the RCo/Ag ratio and/or the CB/Ag ratio is <0.7, type 2 VWD is amount of immobilised VWF. The slopes of the obtained regression diagnosed. Type 2B VWD can be identified with an enhanced RIPA lines reflect the binding capacity of VWF to FVIII.16 (<0.8mg/ml). Type 2B VWD is distinguished from a PT-VWD (pseudo- E U R O P E A N H A E M A T O L O G Y 222 35 Intensity Intensity Intensity Intensity Meiring_subbed.qxp 12/8/09 5:19 pm Page 36 Coag2u23lation Disorders VWD) by performing the RIPA-mixing studies. Type 2A and 2M may only three values. Laffan et al.21 also found RIPA to be normal in have a low RIPA (>1.2mg/ml). Multimeric analysis in plasma is patients with a VWF:RCo value of less than 30%. They found RIPA necessary to distinguish between type 2A VWD (lack of largest and decreased only in severe forms of VWD. When RIPA is increased and intermediate multimers) and type 2M VWD (all the multimers are type 2B is suspected, it is important to perform the RIPA-mixing present). The VWF:CB is usually normal in type 2M VWD due to the studies to distinguish PT-VWD from type 2B VWF. presence of the HMW multimers, except where a collagen-binding defect is diagnosed in patients with type 2M VWD. In type 1 VWD, the VWF multimeric analysis in plasma is necessary to distinguish between ratio between FVIII and VWF:Ag is always concordant. When this ratio subtypes of type 2 VWD. Our method includes rapid processing, is discrepant with a FVIII level of <20%, type 2N VWD is suspected, and simplicity of gel preparation, high sensitivity to low concentrations of this type of VWD can be confirmed by performing a FVIII-binding assay. VWF and elimination of radioactivity.15 Type 2N VWD can be suspected in case of discrepant values of FVIII. In typical type 2N cases, the FVIII Conclusion level is usually less than 20%.22 The diagnosis of type 2N VWD should We have outlined a systematic way to diagnose VWD. This algorithm be confirmed by the FVIII-binding assay. As VWD is such a complex is also recommended by the International Society for Thrombosis and disease to diagnose, this systematic approach makes the diagnosis of Haemostasis (ISTH). It is important to note that the RCo/Ag ratio VWD more accurate, which is of vital importance for the treatment of and the CB/Ag ratio are necessary to distinguish between type 1 and the disease. However, there are pitfalls in this diagnostic process that type 2 VWD. are due to the limitations in sensitivity, reproducibility and inter- laboratory variability of the agglutination-based RCo and RIPA tests.23 The CBA is extremely sensitive in the discrimination of type 1 and types 2A and 2B VWD.11 In contrast, however, the RCo test has a poor In conclusion, it is crucial to use a systematic method to diagnose sensitivity of 50%,18 which is difficult to standardise,19 and lacks a VWD. Each laboratory test only forms one piece of the diagnostic physiological analogue; however, it remains the standard method for puzzle and therefore it is necessary to put all of the puzzle pieces measuring VWF activity approved by the Standardisation Committee together for the whole diagnostic picture to emerge. ■ of the ISTH. Casonato et al.12 demonstrated that the VWF:CB was consistently more sensitive to large and intermediate VWF-multimer Muriel Meiring is an Associate Professor and Specialist representation than the VWF:RCo. None of the type 1 VWD patients Scientist in the Department of Haematology and Cell studied showed a greater decrease in CBA than in Ag levels, which Biology at the University of the Free State at was not the case with the RCo assay. The decreased values of the Bloemfontein in South Africa. She established the Specialised Haemostasis Laboratory at the University of CBA in type 2A and 2B patients were more consistent than those of the Free State as a reference centre for von Willebrand the RCo assay. The VWF:CB is normal in type 2M VWD patients due to disease (VWD). Professor Meiring’s list of publications the presence of the HMW multimers. This assists in the diagnosis of and congress presentations has won various awards at local and national congresses. type 2M VWD. Philip N Badenhorst is Chairman of the Department and The mechanisms involved in type 1 VWD include reduced synthesis a Professor of Haematology and Cell Biology at the University of the Free State and the National Health and secretion of VWF or increased clearance of VWF from plasma.20 In Laboratory Service at Bloemfontein in South Africa. patients with increased clearance of VWF, the DDAVP treatment would Professor Badenhorst is a former President of the South not be effective, as VWF in plasma is cleared from the circulation very African Society of Haematology (SASH). He has received the University of the Free State Centenary Medal in quickly. The ratio between the propeptide and the antigen levels is recognition for his strong role and national profile in used as an indication of VWF clearance in these patients.3 establishing and developing haematology as a discipline at the university. The RIPA test is necessary to distinguish between type 2B and type 2A Mareli Kelderman is a Medical Technologist (specialising VWD. It is also important to carry out the RIPA test with ristocetin in haematology) in the Department of Haematology and concentrations ranging from 0.2 to 2mg/ml in increments of 0.1mg/ml. Cell Biology at the University of the Free State in Bloemfontein in South Africa, and is a member of the This will ensure a range of concentrations to determine the maximum Society of Medical Laboratory Technologists of South agglutination value of 30%. Currently, the RIPA test that is being used Africa (SMLTSA). She developed, streamlined and by most laboratories in developing countries is performed with only implemented several tests for diagnosing von Willebrand disease (VWD), especially the von Willebrand factor three ristocetin concentrations (0.5, 1 and 1.25mg/ml). It is sometimes (VWF) multimer test and the factor VIII-binding assay. difficult to determine whether the RIPA is enhanced or reduced on 1. Tuddenham EG, Blood Rev, 1989;3: 251–62. 2004;10:232–42. 1996;88:2951–8. 2. Vlot AJ, Koppelman SJ, Bouman BN, Sixma JJ, Thromb 10. Favaloro E, Thromb Haemost, 2000;83(1):127–35. 18. Scott JP, Montgomery RR, Retzinger GS, J Biol Chem, Haemost, 1998;79: 456–65. 11. Favaloro E, Koutts J, Patology, 1997;29:385–90. 1991;266:8149–1. 3. Sadler JE, Thromb Haemost, 2005;3:1702–9 12. Casonato A, Pontara E, Bertomoro A, et al., Br J Haematol, 19. Casonato A, Pontara E, Bertomoro A, et al., Thromb 4. Kessler C, Eur J Haematol, 2007;2(1):16–17. 2001;112:578–83. Haemost, 1999:81:994–5. 5. Favaloro EJ, Patterson D, Denholm A, et al., Br J Haematol, 13. Meiring SM, Badenhorst PN, Kelderman M, Clin Chem Lab 20. Haberichter S, Castaman G, Budde U, et al., Blood, 2007;139:621–8. Med, 2007;45(8):1068–72. 2008;111:4979–85. 6. Laffan M, Manning R, Dacie and Lewis Practical Haematology, 14. Krizek DR, Rick ME, Thromb Res, 2000;97:457–62. 21. Laffan M, Brown SA, Collins PW, et al., Haemophilia, Philadelphia: Elsevier, 1995;379–440. 15. Meiring SM, Badenhorst PN, Kelderman M, Medical 2004;10:199–217. 7. Federici AB, Castman G, Mannucci PM, Haemophilia, Technology SA, 2005;19(2):15–20. 22. Schneppenheim R, Budd U, Krey S, et al. Thromb Haemost, 2002;8:607–21. 16. Casonato A, Pontara E, Zerbinati P, et al., Am J Clin Pathol, 1996;76:598–602. 8. Favaloro EJ, Mehrabani PA, Haemophilia, 1996;2:218–23. 1998;109(3):347–52. 23. Favaloro E, Smith J, Petinos P, et al., Thromb Haemost, 9. Favaloro EJ, Bonar R, Kershaw J, et al., Haemophilia, 17. Borchiellini A, Fijnvandraat K, Ten Cate JW, et al., Blood, 1999;82:1276–82. 36 E U R O P E A N H A2E2M3A T O L O G Y 224 S-25 Laboratory Diagnosis and Management of Von Willebrand Disease in South Africa Muriel Meiring, Ph.D.,1 Marius Coetzee, M.Med.,1 Mareli Kelderman, D.M.T.,1 and Philip Badenhorst, M.D.1 ABSTRACT Patients with von Willebrand disease (VWD) in South Africa are cared for in 17 Hemophilia Treatment Centers. The exact prevalence of the disease is uncertain, but 539 patients are annotated in registries. VWD patients are mostly diagnosed in the five largest academic centers, and the classification of the subtypes is performed by one of these, the VWD testing facility. An algorithm is used for the diagnosis of VWD. The distribution of subtypes diagnosed by the VWD reference center is 38%, 58%, and 4% for type 1, 2, and 3, respectively, and 15% of plasma samples received are rejected due to poor storage and transport conditions. A novel single nucleotide polymorphism has been found in an African patient with type 2B VWD. From the type 1 VWD patients who were diagnosed by the VWD testing facility, 45% seem to have an increased VWF clearance phenotype with a propeptide-to-antigen ratio of 1.9 0.3. VWD patients are treated with desmopressin, factor (F)VIII/VWF concentrate (Haemosolvate FVIII; National Bio- products Institute, Durban, South Africa), and tranexamic acid. Haemosolvate FVIII contains a VWF antigen concentration of 167 27 IU/mL, a ristocetin cofactor activity of 100 29 IU/mL, a collagen binding activity of 99 29 IU/mL, normal VWF multimers, and a FVIII concentration of 50 IU/mL. Not all patients with VWD are currently classified, and many VWD patients in South Africa are probably undiagnosed. KEYWORDS: Von Willebrand disease, classification, diagnosis, hemophilia treatment centers, South Africa South Africa has a total land area of slightly Haemophilia Foundation (the national members’ organ- more than 1.2-million km2, making it roughly the same ization), the Medical and Scientific Council of South size as Nigeria, Angola, Mali, or Colombia. It measures Africa, and the National Haemophilia Nurses Commit- some 1600 km from north to south, roughly the same tee to ensure optimal management of patients with from east to west, and has a population of 49 million. bleeding disease, including von Willebrand disease Patients with bleeding diatheses in South Africa (VWD).1 are cared for in 17 Hemophilia Treatment Centers Hemophilia care data collected from 2004 to (HTCs) distributed all over the country. The HTCs 2007 shows that > 2200 patients with bleeding diatheses function in collaboration with the South African were cared for in this period by 79 professionals in 17 National Department of Health, the South African HTCs. Of these patients, 59% had hemophilia A, 21% 1Department of Haematology and Cell Biology, University of the Free of a Globally Distributed Bleeding Disorder; Guest Editor, State and NHLS, Bloemfontein, South Africa. Emmanuel J. Favaloro, Ph.D., F.F.Sc. (RCPA). Address for correspondence and reprint requests: Muriel Meiring, Semin Thromb Hemost 2011;37:576–580. Copyright # Ph.D., Professor, Department of Haematology and Cell Biology, 2011 by Thieme Medical Publishers, Inc., 333 Seventh Ave- University of the Free State and NHLS, Bloemfontein 9300, South nue, New York, NY 10001, USA. Tel: +1(212) 584-4662. Africa (e-mail: gnhmsmm@ufs.ac.za). DOI: http://dx.doi.org/10.1055/s-0031-1281045. Von Willebrand Disease: Local Diagnosis and Management ISSN 0094-6176. 576 224 Downloaded by: University of Free State. Copyrighted material. 225 VON WILLEBRAND DISEASE IN SOUTH AFRICA/MEIRING ET AL 577 VWD, and 12% hemophilia B; the remainder had rare (response only at > 1.2 mg/mL). Multimeric analysis bleeding diatheses and thrombocytopathies.1 in plasma is necessary to distinguish between type 2A The exact prevalence of VWD in South Africa is VWD (lack of largest and intermediate multimers) and uncertain. The central plateau of South Africa is rela- type 2M VWD (all the multimers are present). The tively dry. Epistaxis in the general population is therefore VWF:CB is usually normal in type 2M VWD due to frequent, and there is not a high index of suspicion of the presence of high molecular weight multimers, bleeding disorders. Because of lack of emphasis during except where a collagen binding defect is diagnosed training, menorrhagia is termed functional if there is no in patients with type 2M VWD. In type 1 VWD the obvious gynecologic cause. The woman is then offered ratio between factor VIII and VWF:Ag is always oral contraceptives if she is younger and a hysterectomy concordant. When this ratio is discordant with a if she is older. The 2008 Global Survey of the World FVIII level < 20%, type 2N VWD is suspected, and Federation of Hemophilia estimated there are 522 diag- this type of VWD can be confirmed by performing a nosed patients with VWD in South Africa. The access- factor VIII binding assay. based Hemophilia Registry mentions 539 patients in Genotypic data are only obtained for patients South Africa.2 African patients might be grossly under- with a functional abnormality of VWD. In an ongoing diagnosed because only 3.9% of patients mentioned in study we have searched for mutations in exon 28 of the the registry are of African origin. Women form 63.8% of VWF gene in five patients with functional defects of patients identified with VWD in South Africa.3 VWF to set up the method for genetic analysis of VWD patients are mostly diagnosed in the five VWD. We used two patients with type 2M, two with largest academic centers, and the classification of the type 2B, and one with type 2A VWD in this study. The subtypes is done by only one of these five centers, whole exon 28 was analyzed in four specific fragments, the VWD testing facility. This article concentrates on using polymerase chain reaction with primers that the VWD testing facility’s findings regarding the mismatch the pseudogene. The mutations were iden- statistics, diagnosis, and challenges in the laboratory tified by automatic sequencing of the different frag- diagnosis of VWD in South Africa. ments. The following polymorphisms were detected. A silent single nucleotide polymorphism (SNP) 4641T/C in all five patients, the SNP 4141A/G in three patients, LABORATORY DIAGNOSIS OF VON a silent SNP 3795G/A in one patient, and a new silent WILLEBRAND DISEASE SNP 4923G/A in a patient from the African popula- The following diagnostic tests are performed by the tion. It is important to note that no polymorphisms in VWD testing facility: von Willebrand factor antigen exon 28 were previously reported from African popu- (VWF:Ag), ristocetin cofactor (VWF:RCo), collagen lations. binding activity of VWF (VWF:CB), VWF propeptide The VWFpp levels are only performed on type 1 levels (VWF:pp), multimeric analysis of VWF, the VWD patients. We have found that 45% of our type 1 factor (F)VIII binding assay of VWF, and mixing studies VWD patients seem to have an increased VWF clear- to identify platelet-type VWD (PT-VWD). ance phenotype with a pp:Ag ratio of 1.9 0.3. Our The VWD testing facility adopted and modified normal range for the pp:Ag ratio of normal subjects is the guidelines for diagnosis and treatment of VWD in 1.3 0.24. Italy according to the algorithm outlined in Fig. 1.4,5 A proportional reduction of both VWF:Ag and VWF:RCo with a RCo:Ag ratio > 0.7 as well as a VON WILLEBRAND DISEASE DIAGNOSTIC proportional reduction of both VWF:Ag and VWF:CB STATISTICS with a CB:Ag ratio > 0.7 suggest type 1 VWD if the The VWD testing facility is situated in Bloemfontein, VWF:Ag level is low (< 45%). If type 1 VWD is the legislative capital of South Africa, in the central part diagnosed, it is important to determine the clearance of the country. The academic complex in Bloemfontein rate of VWF. The VWF:pp is then performed. If the serves patients from the Free State and the Northern ratio between the VWF:pp and the VWF:Ag is > 2, an Cape provinces with a total population of 4 million. increased clearance rate of VWF is suspected for that The VWD testing facility, however, receives patient patient. samples from all over the country for diagnosis but If the RCo:Ag ratio and/or the CB:Ag ratio is especially for the classification of VWD. < 0.7, type 2 VWD is diagnosed. Type 2B VWD can be Table 1 outlines the relative proportion of the identified with of an enhanced ristocetin-induced plate- various VWD types diagnosed by the VWD reference let agglutination (RIPA) (response with < 0.8 mg/mL). center. A total of 250 patients were included in this data Type 2B VWD is distinguished from a PT-VWD set. (pseudo-VWD) by performing the RIPA mixing stud- Because the VWD testing facility receives mostly ies. Type 2A and 2M typically have reduced RIPA VWD samples to be classified and not to be diagnosed, 225 Downloaded by: University of Free State. Copyrighted material. 226 578 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 37, NUMBER 5 2011 Figure 1 Algorithm for the laboratory diagnosis of von Willebrand disease (VWD) used in our center. FVIII, factor VIIII; PT- VWD, platelet-type VWD; RIPA, ristocetin-induced platelet agglutination; VWF:Ag, von Willebrand factor antigen; VWF:CB, VWF collagen binding; VWFpp, VWF propeptide; VWF:RCo, VWF ristocetin cofactor. the distribution of subtypes diagnosed is 38%, 58%, and CHALLENGES IN THE LABORATORY 4% for types 1, 2, and 3, respectively. From the type 2 DIAGNOSIS OF VON WILLEBRAND VWD patients, 23% were diagnosed with type 2A, 22% DISEASE with type 2B, 13% with type 2M, and none with type Samples must be stored immediately after centrifuga- 2N. No patients with PT-VWD have so far been tion in polypropylene tubes at 708C until analyzed.5 It diagnosed, probably due to the unavailability of platelets is thus important to note that a cryoprecipitate might from these patients, because most of the type 2 VWD form if plasma samples are stored at temperatures samples are referrals (i.e., transported plasma) from warmer than 708C. Cryoprecipitates contain large larger HTCs in the country. quantities of VWF and especially high molecular weight 226 Downloaded by: University of Free State. Copyrighted material. 227 VON WILLEBRAND DISEASE IN SOUTH AFRICA/MEIRING ET AL 579 Table 1 Relative Proportion of the Various Von Table 2 Test Results of a Normal Plasma Sample Willebrand Disease Types Stored at S 20-C in a Domestic Chest Freezer with No Frost-Free Facility versus When Stored at S 70-C for Percentage (Number) of Subtype of VWD Total Patients Diagnosed* 3 Weeks before Testing Tests Stored at S 20-C Stored at S 70-C Mild type 1 30.7% (75) Moderate to severe type 1 8.2% (20) VWF:Ag 71 74 Type 2A 22.5% (55) VWF:CB 0 87 Type 2B 23.4% (57) VWF:RCo 0 73 Type 2M 13.1% (32) Multimer LMW multimers LMW multimers Type 2N 0% (0) patten absent present Type 3 2.1% (5) VWF:Ag, von Willebrand factor antigen; VWF:CB, VWF collagen Platelet type 0% (0) binding; VWF:RCo, VWF ristocetin cofactor; LMW, low molecular weight. Total 100% (244) See Fig. 2 for multimer patterns. *Total number of patients¼ 250. VWD, Von Willebrand disease. multimers.6 All tests therefore must be done on original depends on the severity of the bleeding or the type of aliquots that were not previously thawed, and plasma surgical or dental intervention. DDAVP is usually the samples should be thawed at 378C before performing first product of choice for treatment of new patients with diagnostic tests. Special care should be taken to ensure VWD. that no cryoprecipitate is present in the samples. There- The FVIII/VWF concentrate used in South fore it must be dissolved before the tests are performed; Africa is Haemosolvate FVIII (National Bioproducts otherwise it will influence the results.6 Institute, Durban, South Africa). Haemosolvate Factor The VWD testing facility receives plasma samples VIII is an intermediate purity factor VIII concentrate, for subtyping of VWD from most HTCs countrywide. currently used for the treatment of hemophilia A and About 15% of samples received are rejected due to poor VWD. In a previous study we determined the concen- plasma storage and transport conditions. Samples are tration and activity of VWF in Haemosolvate Factor now rejected if unfrozen upon arrival or if sent refriger- VIII. We received 32 batches of the concentrate from ated only. This decision was made after performing a the National Bioproducts Institute in Pinetown, South study where the VWF levels, activity, and multimer Africa, and performed the VWF:Ag assay to determine distribution were measured on plasma samples following the VWF levels. The functional activity of VWF was different storage conditions. Normal plasma samples determined by performing the VWF:RCo and VWF:CB were exposed to different storage conditions and time assays. We also determined the FVIII levels and the intervals. We found that the VWF:Ag, VWF:CB, and multimeric analysis of VWF in these concentrates. For VWF:RCo results remain normal after storage all the tests, we needed to dilute the concentration at 708C. The multimer patterns also remain normal. However, after storage at 208C in a household chest freezer (not frost free), both the functional assays showed a decreased activity of VWF, and the multimer analysis showed an absence of the high molecular weight multi- mers in some samples. The multimer pattern thus stays stable at 708C but not at 208C. Table 2 shows the results of one such normal plasma sample, and Fig. 2 shows the associated multimer pattern. Many laborato- ries in South Africa use household chest freezers. We also found that even when a plasma sample is thawed and frozen up to five times, the multimer pattern stays normal when stored at 708C. Fig. 3 shows the results of a sample that was frozen five times at 708C. TREATMENT OF VON WILLEBRAND DISEASE IN SOUTH AFRICA VWD patients are treated with desmopressin Figure 2 Multimer patterns of a normal plasma sample (DDAVP), FVIII/VWF concentrate, tranexamic acid, stored at 708C versus when stored in a domestic chest and oral contraceptives. The choice of medication freezer at 208C for 3 weeks. 227 Downloaded by: University of Free State. Copyrighted material. 228 580 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 37, NUMBER 5 2011 Figure 4 Von Willebrand factor multimer patterns of eight batches of Haemosolvate FVIII. Figure 3 Multimer patterns of a normal plasma sample academic centers. This is mostly due to expensive ship- that was repeatedly thawed and frozen five times at 708C. ping costs. The inherent limitations in sensitivity, re- producibility, and interlaboratory variability of the extensively. The VWF:Ag concentration of all batches agglutination-based VWF:RCo and RIPA tests are had a mean value of 167 27 IU/mL, a VWF:RCo well known.7 Thus, given the limited tests applied in activity of 100 29 IU/mL, a VWF:CB activity of most centers, many VWD patients in South Africa 99 29 IU/mL, and a factor VIII concentration of 50 might be misdiagnosed or remain undiagnosed. IU/mL. The multimeric analysis showed a normal mul- timer pattern as seen in Fig. 4. We thus found that the VWF levels and activities in Haemosolvate Factor VIII REFERENCES are more than twice that of the FVIII level. This is now 1. Mahlangu JN; Medical and Scientific Council of the South taken into account when this product is administered to African Haemophilia Foundation. Haemophilia care in South patients for the treatment of VWD. Africa: 2004–2007 look back. Haemophilia 2009;15(1): The antifibrinolytic drug tranexamic acid is often 135–141 given as treatment for mucocutaneous bleedings, if 2. Karabus C. Reports on the South African Haemophilia necessary in combination with DDAVP or Haemosol- Registry. Paper presented at: Annual meeting of the Medical vate Factor VIII. It is also given before and after surgical and Scientific Advisory Council of the South African or dental procedures. Haemophilia Foundation; Johannesburg, South Africa; 2010 3. Coetzee MJ, Coetzee M, Jooste P, Goga Y. The Web-based Registry of the South African Haemophilia Foundation. Available at:http://www.hemophilia.org.za/registry. Accessed CONCLUSION January 25, 2011 Patients with VWD in South Africa are cared for in 17 4. Federici AB, Castaman G, Mannucci PM; Italian Association HTCs distributed throughout the country. These pa- of Hemophilia Centers (AICE). Guidelines for the diagnosis tients are treated according to international guidelines and management of von Willebrand disease in Italy. and with the VWF/FVIII concentrate produced and Haemophilia 2002;8(5):607–621 5. Meiring SM, Kelderman M, Badenhorst PN. Laboratory used for VWD in South Africa, Haemosolvate FVIII, a diagnosis of von Willebrand disease. European Haematology highly active VWF concentrate. The diagnosis of VWD 2009;3(1):33–36 is done mostly in only the five largest academic centers. 6. Favaloro EJ, Mehrabani PA. Laboratory assessment of von Except for the VWD testing facility, all these centers Willebrand factor: differential influence of prolonged ambient diagnose VWD mostly by the VWF:Ag and the temperature specimen storage on assay results. Haemophilia VWF:RCo or automated VWF activity assays. A dis- 1996;2:218–223 crepancy between these two tests would indicate type 2 7. Favaloro EJ, Smith J, Petinos P, Hertzberg M, Koutts J; RCPA Quality Assurance Program (QAP) in Haematology VWD. The further typing of VWD is done only by the Haemostasis Scientific Advisory Panel. Laboratory testing for VWD testing facility, situated in Bloemfontein. How- von Willebrand’s disease: an assessment of current diagnostic ever, only a very limited number of patient samples are practice and efficacy by means of a multi-laboratory survey. referred to the VWD testing facility from the large Thromb Haemost 1999;82(4):1276–1282 228 Downloaded by: University of Free State. Copyrighted material. S-26 229 Original Article Page 1 of 6 Challenges in the laboratory diagnosis and management of von Willebrand disease in South Africa Muriel Meiring1,2, Leriska Haupt1,2, Charmainé Conradie1,2, Jaco Joubert1,2 1National Health Laboratory Service, Universitas Hospital, Bloemfontein, South Africa; 2Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, South Africa Contributions: (I) Conception and design: M Meiring, L Haupt, J Joubert; (II) Administrative support: M Meiring, C Conradie; (III) Provision of study materials or patients: L Haupt, J Joubert; (IV) Collection and assembly data: M Meiring, C Conradie; (V) Data analysis and interpretation: M Meiring, C Conradie; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors. Correspondence to: Muriel Meiring. Department of Haematology and Cell Biology, University of the Free State, 205 Mandela Drive, Bloemfontein, 9301, South Africa. Email: MeiringSM@ufs.ac.za. Background: South African von Willebrand disease (VWD) care is challenged by the lack of laboratory testing standardisation, lack of national clinical guidelines, limited undergraduate exposure to bleeding disorders and limited pharmacological resources. As the only reference testing facility performing VWD confirmatory testing, our experience in the diagnosis of the majority of VWD cases in South Africa has highlighted many areas where improvements are desperately needed. Methods: We conductedd a retrospective audit of 250 VWD cases at our laboratory where we calculated the percentage misdiagnoses should only the von Willebrand factor (VWF) levels and the ristocetin cofactor activity assays be used. Results: We have shown that performing a limited panel of VWD testing (VWF antigen and ristocetin cofactor activity only) could lead to the misdiagnosis of VWD in up to 77% of cases analysed at our laboratory. If the multimeric analysis was to be included in the diagnostic setup, together with the limited panel, still 20% of patients would be misdiagnosed. The underdiagnosis and under-reporting of VWD greatly underestimates the burden of the disease and the impact on the healthcare system. Conclusions: Future advancements in VWD care in South Africa should be focused on correcting these shortcomings. Keywords: Von Willebrand disease (VWD); diagnosis; haemophilia treatment centres (HTCs);South Africa Received: 01 October 2017; Accepted: 17 November 2017; Published: 18 December 2017. doi: 10.21037/aob.2017.11.01 View this article at: http://dx.doi.org/10.21037/aob.2017.11.01 Introduction with bleeding disorders such as von Willebrand disease (VWD) (1). According to the South African Haemophilia Foundation, The 2016 Global Survey of the World Federation of patients with bleeding diatheses in South Africa are Haemophilia estimated that there are 632 patients diagnosed cared for in 20 haemophilia treatment centres (HTCs) with VWD in South Africa; 375 female and 257 male. Of nationwide. The function of HTCs, in collaboration with these patients, 431 have mild, 61 moderate and 42 severe the South African National Department of Health, the VWD. Most of these patients are diagnosed between the South African Haemophilia Foundation (the national ages of 14 and 44 years. VWD patients are predominantly members’ organization) with its Medical and Scientific diagnosed in five academic centres and a single central Advisory Council, and the National Haemophilia Nurses reference laboratory does the confirmatory testing that aids Committee, is to ensure optimal management of patients in the subclassification of VWD. This paper will concentrate © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 229 230 Page 2 of 6 Annals of Blood, 2017 on the reference laboratory’s findings regarding these This VWD reference laboratory is also the only statistics, as well as challenges in the laboratory diagnosis and laboratory in the country that performs the VWF:CB, management of VWD in South Africa. the mutimeric analysis of VWF, the VWF:pp and the VWF:FVIII binding assay (3). The diagnostic screening tests for VWD include Methods the VWF:Ag, WF:RCo, VWF:CB and the multimeric Laboratory diagnosis and management of VWD distribution of VWF. Confirmatory tests include RIPA and the VWF:FVIII binding assay. However, most other The VWD reference laboratory adopted and modified the laboratories in South Africa, except for the reference guidelines for diagnosis and treatment of VWD in Italy (2,3) laboratory, only perform the VWF:Ag and VWF:RCo according to the algorithm outlined in Figure 1. In assays (i.e., essentially screening for VWD). concordance with these guidelines the following diagnostic Retrospective data of 250 VWD cases were gathered and tests are offered at the reference laboratory: von Willebrand the percentage misdiagnoses were calculated, should only factor (VWF) antigen (VWF:Ag), ristocetin cofactor the VWF:Ag and VWF:RCo assays be used. The results are activity (VWF:RCo), collagen binding assay (VWF:CB), shown in Figure 2. multimeric analysis, VWF propeptide levels (VWF:pp), The relative distribution of the various VWD subtypes factor VIII (FVIII) binding assay, ristocetin induced platelet diagnosed by the VWD reference laboratory from 2011 to agglutination (RIPA), and RIPA mixing studies to identify 2016 is outlined in Table 1. Laboratory statistics, as well as platelet type VWD (PT-VWD) (3). challenges in the laboratory diagnosis and management of A proportional reductixon of both VWF:Ag and VWD in South Africa were determined by the reference VWF:RCo with a RCo/Ag ratio ≥0.7 as well as a laboratory. Information regarding the available treatment proportional reduction of both VWF:Ag and VWF:CB with modalities for VWD and challenges thereoff were gathered a CB/Ag ratio ≥0.7 suggests type 1 VWD. If type 1 VWD is from all heamophilia treatment centers nationwide. diagnosed, the clearance rate of VWF is determined and the VWF:pp then performed. If the ratio between the VWF:pp and the VWF:Ag is more than 2, an increased clearance rate Results of VWF is suspected for this patient (i.e., type 1 Vicenza). Challenges in the laboratory diagnosis of VWD If the RCo/Ag ratio and/or the CB/Ag ratio is <0.7, then type 2 VWD is diagnosed. Type 2B VWD can be identified By performing only the two most popular tests (VWF:Ag with an enhanced RIPA (response to <0.8 mg/mL ristocetin). and VWF:RCo), 6% of type 1 VWD patients would have Type 2B VWD is distinguished from a PT-VWD (pseudo been misdiagnosed as type 2 VWD; 13% of type 2A patients VWD) by performing RIPA mixing studies. Type 2A and 2M would be classified as type 1 and 77% as type 2B disease; 8% might show a reduced RIPA (response only to >1.2 mg/mL of type 2B VWD patients would be misdiagnosed as type 1 ristocetin). Multimeric analysis in plasma is necessary to and 55% as type 2B disease; 28% of type 2M VWD patients distinguish between type 2A VWD (lack of largest and would be misdiagnosed as type 1 and 48% as type 2A or 2B intermediate multimers) and type 2M VWD (all multimers VWD; 1% of type 3 VWD will be misdiagnosed as type 2 are present). The multimeric distribution pattern of type VWD. Moreover, even if the multimeric analysis were to 2M VWD often differs from that of normal plasma in the be included in this diagnostic setup (i.e., together with the lower density of high molecular weight multimers and two most popular tests), still 20% of patients would still be the higher density of lower molecular weight multimers. misdiagnosed. The VWF:CB is usually normal in type 2M VWD due As previously published, the vast distances between to the presence of the high molecular weight multimers, referral laboratories and the reference laboratory leads to except where a collagen binding defect is diagnosed in thermal sample compromise and testing challenges. When such patients. In type 1 VWD, the ratio between FVIII samples are stored at temperatures warmer than ‒70 ℃, level and VWF:Ag is always concordant. When this ratio a cryoprecipite might be formed which greatly affects is discrepant (i.e., FVIII/Ag <0.7) with a FVIII level of less multimeric analysis (3). In the time period between 2012 than 20%, type 2N VWD is suspected and this subtype can and 2017, 8% of referred samples were rejected due to poor be confirmed by performing a FVIII binding assay (3). sample storage and transport conditions. The challenges © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 230 231 Annals of Blood, 2017 Page 3 of 6 Plasma VWF:Ag Present Absent Plasma VWF:RCo vs. VWF:Ag VWF:CB vs. VWF:Ag TYPE 3 Proportionate (0.7–1.2) Discrepant (<0.7) TYPE 1 TYPE 2 VWF:pp vs. VWF:Ag Plasma FVIII vs. VWF:Ag Discrepant (<0.7) Discrepant >2 Proportionate FVIII < 20% TYPE 1 TYPE 2N With increased clearance of VWF FVIII binding assa TYPE 2 Multimer analysis HMW multimers Present Absent RIPA TYPE 2M TYPE 2A TYPE 2B VWF:CB normal Decreased Increased (0.2–0.8 mg/mL) (>1.2 mg/mL) RIPA Mixing studies Plasma defect Platelet defect TYPE 2B PT-VWD Figure 1 Algorithm for the laboratory diagnosis of VWD (3). VWD, von Willebrand disease; VWF, von Willebrand factor; HMW, high molecular weight; RIPA, ristocetin induced platelet agglutination. © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 231 232 Page 4 of 6 Annals of Blood, 2017 90 multimeric analysis and the propeptide assay (Meiring et al., 80 2017 unpublished data). The VWF:RCo has challenges of its 70 60 own, since the sensitivity is reduced at low levels (<15%) (4). 50 It is thus of utmost importance to perform all diagnostic 40 tests to prevent misdiagnosis of the disorder. 30 20 10 VWD diagnostic statistics 0 The VWD reference laboratory is situated in Bloemfontein, the legislative capital of South Africa, in the central part of the country. The academic complex in Bloemfontein Figure 2 Percentage misdiagnosis of VWD subtyte, should serves patients from the Free State and the Northern only the VWF:Ag and VWF:RCo assays be used. VWD, von Cape provinces (a geographically vast area) with a total Willebrand disease; VWF, von Willebrand factor. population of about 4 million (5). The VWD reference laboratory, however, receives patient samples country-wide for both initial screening and diagnosis, but especially for Table 1 Distribution of VWD subtypes diagnosed at the VWD reference laboratory the classification of VWD. The relative distribution of the various VWD subtypes diagnosed by the VWD reference Percentage [number] of total Subtype of VWD laboratory from 2011 to 2016 is outlined in Table 1. The patients diagnosed, N=255 reference laboratory still receives mostly type 2 VWD Mild type 1 23 [59] samples to be sub-classified, as also mentioned in 2011 (3). Moderate to severe type 1 13 [33] Still no type 2N VWD patients have been diagnosed. The Type 2A 28 [73] reason might be the improbability of an autosomal recessive disorder in our diverse population or the possibility of Type 2B 12 [30] misdiagnoses of type 2N with haemophilia carriers by the Type 2M 20 [50] HTCs. Type 2N 0 [0] Type 3 4 [10] Treatment of VWD in South Africa Platelet type 0 [0] Available treatment modalities include desmopressin VWD, von Willebrand disease. (DDAVP), FVIII/VWF concentrates, antifibrinolytic drugs, topical therapies and hormonal treatment modalities for posed by the geographically vast area of Southern Africa, women. DDAVP is the treatment of choice in patients with type 1, 2A and 2M. Treatment with DDAVP is subject to as covered by the reference laboratory, and the previously a proven plasma response of FVIII levels and VWF:RCo described issues with specimen preparation, freezing and activity. transport remain, and have furthermore been exacerbated Two intermediate purity FVIII/VWF concentrates by a general lack of dry ice for specimen transport (3). are available in South Africa for the treatment of VWD: Another challenge in the laboratory diagnosis of VWD, Haemosolvate® produced by the National Bioproducts is the cost of the assays and the investigation. In the South Institute (NBI) from pooled fresh human plasma, and African private medical sector, laboratory testing costs are Virally Inactivated Factor VIII (VIAHF) produced by the covered by various medical aid schemes. However, the price Western Province Blood Transfusion Service (WPBTS) that the medical aids are authorised to extend for these from small pools (5–6 bags) of cryoprecipitate. Both tests is often less than the costs of the latest test kits. Due to products are prepared from plasma obtained from voluntary these cost constraints, the reference laboratory developed non-remunerated donors after individual donation serologic and validated in-house assays for more cost-effective VWD and nucleic acid amplification and exclusion of human testing. These include the VWF:Ag and VWF:CB, the immunodeficiency virus (HIV), hepatitis B virus (HBV), © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 232 % misdiagnosis 1 as 2A/B/M 2A as 1 2A as 2B 2B as 1 2B as 2A 2M as 1 2M as 2A/B 3 as 2A/B/M 233 Annals of Blood, 2017 Page 5 of 6 and hepatitis C virus (HCV) infection. Both undergo viral the treatment of haemophilia in South Africa (10), but no inactivation steps which inactivate HIV, HBV and HCV: formal guidelines focusing on the management of VWD in Haemosolvate® using a solvent-detergent process, and South Africa have been published—a significant obstacle to VIAHF via 80 ℃ heating for 72 hours. Haemosolvate® safe, standardised care. Locally tailored guidelines similar or VIAHF are considered the treatment of choice when to the United Kingdom Haemophilia Centre Doctors DDAVP is not indicated or ineffective, as both concentrates Organization guideline approved by the British Committee contain high molecular weight VWF multimers (6). In a for Standards in Haematology (11) or the National Heart, previous study by our laboraotry, we have proven VWF:Ag Lung, and Blood Institute guidelines (6) should ideally be levels in Haemosolvate® to be as high as double that of drafted on a national level. Guidance on issues specifically FVIII, and our centre is also involved in continuous quality applicable to resource constrained environments like South control monitoring of this product (3). Africa, such as prophylaxis, genetic testing, and management Tranexamic acid (Cyclokapron®) is available in capsule of surgical interventions, may be particularly useful. Such and intravenous formulations. Recent studies confirm the guidelines may also be of value to other developing countries, effectiveness of prophylactic tranexamic acid in reducing where the use of American and British guidelines may not be the number of mild and moderate bleeding episodes (7,8); appropriate. however, this has not yet been explored at our centre. The use of tranexamic acid on an on-demand basis with bleeding Discussion episodes, is advocated locally. Hormonal treatments for women with VWD include The probable underdiagnosis of VWD in South Africa, combined oral contraceptives and hormonal intra-uterine may be a consequence of poor physician awareness of the devices, which are readily available in our country. disease, which may be a result of inadequate undergraduate Available topical therapy is limited to fibrin glue medical training on the bleeding disorders in general, (Thromboseel® and Tisseel®, South African National Blood and VWD in particular. VWD as a specific disease entity Service and Adcock Ingram Critical Care). receives only approximately 20 minutes of dedicated lecture time in the 5-year undergraduate medical curriculum at the University of the Free State. This may be difficult to address in light of the heavy Challenges in the treatment of VWD in South Africa demands placed on undergraduate medical curricula to In South Africa, the route of administration of DDAVP is accommodate vast amounts of information from different limited to a single DDAVP intravenous formulation which disciplines, but may be amenable to interventions at post- can also be administered subcutaneously. The intranasal graduate level. Training interventions targeted at interns and preparation of DDAVP, although available in South Africa other junior doctors may be effective in improving practice for the treatment of enuresis (0.1 mg/mL), is not sufficient standards, as was recently shown locally in transfusion in concentration for use in VWD, where a concentration of medicine (12,13). Such training interventions should ideally 1.5 mg/mL is quoted in international guidelines (6). This also target physicians that are most likely to be faced with is a challenging limitation in paediatric VWD care where VWD patients at presentation, such as gynaecologists, intranasal DDAVP would be a convenient choice. paediatricians, otorhinolarynologists and general practitioners. Although intermediate purity FVIII/VWF concentrates Misdiagnosis (now thought to possibly be a bigger are relatively freely available locally, no high purity VWF problem than previously suspected) may be reduced by concentrate is currently available in South Africa. When routinely using a more extensive testing profile, which is used to treat VWD, FVIII/VWF concentrates can result currently not standardised in South Africa. This reiterates in markedly elevated FVIII levels (disproportionate to the need for national guidelines on both the laboratory the VWF levels achieved with treatment) which has been testing and treatment of VWD, which are currently associated with thrombosis (9). A high purity VWF product lacking. Product issues (lack of suitable intranasal DDAVP could prevent such disproportionate increases in FVIII preparations and high-purity VWF concentrates) remain levels, further improving treatment safety (6). a challenge as well as a general lack of physician awareness The Medical and Scientific Advisory Council of the South and limited undergraduate training, which may contribute African Haemophilia Foundation has issued guidelines for to underdiagnosis of the disease. © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 233 234 Page 6 of 6 Annals of Blood, 2017 Conclusions Basis results – Municipalities. Statistical Release P301.1 6. Nichols WL, Hultin MB, James AH, et al. von Willebrand The advancement of VWD care in South Africa, and disease (VWD): evidence-based diagnosis and management perhaps many other developing countries , would guidelines, the National Heart, Lung, and Blood Institute likely benefit most from the establishment of national (NHLBI) Expert Panel report (USA). Haemophilia guidelines, targeted postgraduate training interventions 2008;14:171-232. and improved undergraduate medical training, an 7. South African National Blood Service (SANBS), Western expanded treatment product repertoire and standardised Province Blood Transfusion Service (WPBTS). Clinical laboratory testing. guidelines for the use of blood products in South Africa 2014; 5th ed. Available online: http://www.wpblood. org.za/village/wpbnew/sites/default/files/clinical_ Acknowledgements guidelines_5th%20Edition_2014.pdf None. 8. Eghbali A, Melikof L, Taherahmadi H, et al. Efficacy of tranexamic acid for the prevention of bleeding in patients with von Willebrand disease and Glanzmann Footnote thrombasthenia: a controlled, before and after trial. Conflicts of Interest: The authors have no conflicts of interest Haemophilia 2016;22:e423-6. to declare. 9. Mannucci PM. Treatment of von Willebrand’s disease. New Engl J Med 2004;351:683-94. Ethical Statement: As the entire study was based on pre- 10. Mahlangu JN, Gilham A, Medical and Scientific Advisory existing and anoymized samples, specific informed consent Council of the South African Haemophilia Foundation. were deemed unneccessary according to local regulations. Guideline for the Treatment of Haemophilia in South Africa. S Afr Med J 2008;98:126-40. 11. Laffan MA, Lester W, O’Donnell JS, et al. The diagnosis References and management of von Willebrand disease: a United 1. Mahlangu JN. Haemophilia care in South Africa: 2004- Kingdom Haemophilia Centre Doctors Organization 2007 look back. Haemophilia 2009;15:135-41. guideline approved by the British Committee for Standards 2. Federici AB, Castman G, Mannucci PM. Guidelines for in Haematology. Brit J Haematol 2014;167:453-65. the diagnosis and management of von Willebrand disease 12. Joubert S, Bosman M, Joubert G, et al. The utilization in Italy. Haemophilia 2002;8:607-21. of red cell concentrates at Kimberley Hospital Complex, 3. Meiring SM, Coetzee MJ, Kelderman M, et al. Laboratory Northern Cape Province, South Africa. Transfus Apher Sci diagnosis and management of von Willebrnad disease in 2013;49:522-7. South Africa. Semin Thromb Hemost 2011;37:576-80. 13. Joubert J, Joubert S, Raubenheimer J, et al. The long-term 4. Favaloro EJ, Bonar RA, Meiring M, et al. Evaluating effects of training interventions on transfusion practice: errors in the laboratory identification of von Willebrand A follow-up audit of red cell concentrate utilisation at disease in the real world. Thromb Res 2014;134:393-403. Kimberley Hospital, South Africa. Transfus Apher Sci 5. Statistics South Africa, 2008. Commuity Survey 2007; 2014;51:25-32. doi: 10.21037/aob.2017.11.01 Cite this article as: Meiring M, Haupt L, Conradie C, Joubert J. Challenges in the laboratory diagnosis and management of von Willebrand disease in South Africa. Ann Blood 2017;2:19. © Annals of Blood. All rights reserved. aob.amegroups.com Ann Blood 2017;2:19 234 S-27 235 © 2007 Schattauer GmbH, Stuttgart Blood Coagulation, Fibrinolysis and Cellular Haemostasis 2B or not 2B? Disparate discrimination of functionalVWF discor- dance using different assay panels or methodologies may lead to success or failure in the early identification of type 2BVWD Emmanuel J. Favaloro1, Roslyn Bonar1, Muriel Meiring2, Alison Street1, Katherine Marsden1, on behalf of the RCPA QAP in Haematology 1Departments of Haematology and RCPA Quality Assurance Program (QAP), Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, New South Wales, Australia; 2Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa Summary Laboratory proficiency in the identification of functional von fication of samples as potentially type 2 VWD, and VWF func- Willebrand factor (VWF) discordance in type 2B von Wille- tional discordance was more consistently identified when lab- brand disease (VWD) was assessed by external quality assur- oratories used (i) automated platelet agglutination for ance surveys conducted by the RCPA Haematology QAP, and VWF:RCo compared to classical platelet aggregometry, (ii) in- using six different type 2BVWD plasma samples (three histori- house VWF:CB assays compared to commercial kit methods, cal and three previously unpublished) tested by up to 52 labora- and (iii) automated LIA-based ‘VWF:Activity’ assays compared tories. For the three most recent samples, functional VWF dis- to ELISA based assays.We conclude that:(i) laboratories are gen- cordance was either not identified in testing or by interpretation erally proficient in tests for VWD but interpretative diagnostic with misidentification as ‘normal’ or ‘type 1 VWD’, on average errors do occur; (ii) correct diagnosis is more likely when test for 25.7% of test occasions when laboratories performed panels are more comprehensive and include the VWF:CB; (iii) VWF:Ag and VWF:RCo as their primary VWF test panel, but sub-methodology influences the appropriate identification of somewhat fewer occasions (10.9%) for laboratories that incor- VWF functional discordance. On the basis of these findings, we porated VWF:CB as an additional functional VWF assay.VWF provide a series of recommendations to enable the appropriate assay sub-methodologies also influenced the appropriate identi- laboratory identification ofVWD, in particular type 2BVWD. Keywords von Willebrand factor, VWF, von Willebrand's disorder, von Willebrand's disease,VWD, laboratory assessment, survey, hae- mostasis testing, diagnostic practice, quality control, quality as- surance Thromb Haemost 2007; 98: 346–358 Introduction tion of a diagnosis (6–9), genetic testing is in fact rarely under- taken in practice. Testing for VWD is, in most laboratories, von Willebrand's disorder/disease (VWD) is the most common undertaken in a staged or step-wise process, with initial ‘screen- inherited bleeding ailment, and is represented by a reduction ing’ using a von Willebrand factor (VWF) antigen (VWF:Ag) and/or abnormal function of von Willebrand factor (VWF), an assay, together with factor VIII coagulant (FVIII:C) and one or adhesive plasma protein essential for primary haemostasis (1–5). more functional VWF assays (e.g. ristocetin cofactor VWD is a heterogeneous disorder, subtyped using clinical and [VWF:RCo], collagen binding [VWF:CB]) to determine the laboratory criteria, and diagnosis typically requires a compre- presence or absence of functional-VWF discordance (absent in hensive panel of tests (1–5). Although genetic testing may be type 1, but present in type 2 VWD). Dependent on the results of undertaken for specific difficult cases or for eventual confirma- this initial testing, subsequent more extensive testing (e.g. ris- Correspondence to: Received December 6, 2006 Dr. E. J. Favaloro Accepted after resubmission April 12, 2007 Department of Haematology Institute of Clinical Pathology and Medical Research (ICPMR) Prepublished online July 6, 2007 SWAHS, Westmead, New South Wales, 2145, Australia doi:10.1160/TH06–12–0693 Tel.: +612 9845 6618, Fax: +612 9689 2331 E-mail: emmanuel@icpmr.wsahs.nsw.gov.au (or emmanuel.favaloro@swahs.health.nsw.gov.au from May, 2007) 235 346 Downloaded by: University of Free State. Copyrighted material. 236 Favaloro et al. Early identification of type 2BVWD tocetin-induced platelet aggregation [RIPA], multimers, 2A, 2N, 2M and normal samples (11–16; and unpublished data). VWF:FVIII binding assay) may be performed, in particular to The 2B VWD samples comprised three new (unpublished) further classify type 2 VWD. Selection of an appropriate initial samples dispatched in 2006, with findings compared to those of test panel, which includes appropriate functional VWF assays, is three other 2B VWD samples dispatched in earlier surveys. De- therefore critical for ensuring that further proper testing, or refer- tails of the previously reported (11, 12) cases are only summa- ral to expert VWD test laboratories or reference centres, will be rised here for comparison. All six 2B VWD patients had a sig- undertaken. nificant personal history of bleeding problems, and all samples Of primary importance, RIPA is recognised as the critical test showed enhanced responsiveness to ristocetin (≤0.5 mg/ml) in a for the correct identification of type 2B VWD but this test may RIPA assay as requisite for identification of type 2B VWD (1). only be performed, or referred onwards to a reference laboratory, RIPA mixing studies (17) also confirmed that the enhanced if the initial VWF test panel identifies an abnormal finding, or at RIPA responsiveness was plasma based for all (i.e. 2B VWD least identifies evidence of VWF functional discordance. Alter- rather than pseudo/platelet-type-VWD). natively, the necessary RIPA testing may not be performed if in- For the new case set, sample 2006.1 was obtained from a itial test findings appear essentially ‘normal’or are suggestive of 67-year-old female previously ‘diagnosed’ with ‘thrombocy- (mild) type 1 VWD (VWF:Ag and functional VWF assay results topenia’, and baseline blood confirmed thrombocytopenia are somewhat low but concordant). The findings of external (Table 1), but showed normal routine coagulation test times. quality assurance surveys conducted by the RCPA Haematology Sample 2006.2 from her 40-year-old daughter demonstrated QAP are reported with a particular emphasis on diagnostic errors mild thrombocytopenia (Table 1), with slightly prolonged APTT associated with testing for type 2B VWD, according to the meth- (38 seconds [sec]; normal reference range = 24–36 sec). Sample odology utilised. 2006.4 was from an unrelated 49-year-old female, and baseline blood demonstrated a borderline normal platelet count (Table 1), Materials and methods with normal coagulation test times. Baseline FVIII:C and VWF parameters for each patient are shown in Table 1, and VWF:Multimer results are shown in Figure 1. PFA-100® CTs General were maximally prolonged (i.e. >250 sec) with both collagen/ Survey samples were distributed to laboratories participating in ADP and collagen/epinephrine cartridges for each patient.Addi- the Special Haemostasis module of the RCPA Haematology tional details of the RCPA QAP’s sample collection and process- QAP (10). Survey participant numbers have varied over the ing procedures, and laboratory details for historical 2B VWD years, but are currently 55. These participants derive from a wide samples, are provided in previous reports (11, 12). All six pa- geographical area and primarily Australia (n=32) and New Zeal- tients showed loss of high-molecular-weight (HMW) VWF in and (n=6), but also Hong Kong (4), Malaysia (5), Singapore (1), multimer analysis, plus functional VWF discordance, although India (1), South Africa (3), United Arab Emirates (2) and Sultan- the level of HMW VWF loss and functional VWF discordance ate of Oman (1). Sample processing instructions were provided, differed between patients (see Discussion). Extensive stability and laboratories (n=52 in 2006) were asked to perform all plas- and homogeneity studies were performed on all samples pro- ma-based tests that they normally used for laboratory identifica- vided to participants, and the samples detailed in this report pass- tion (/diagnosis) of VWD (i.e. FVIII:C and specific VWF tests ed stringent test stability/homogeneity criteria. such as VWF:Ag, VWF:CB, VWF:RCo, VWF:Multimers, as locally available). Samples were blinded and identified only by a Evaluation of interpretative data code ID. To help assess test panel diagnostic efficacy (i.e. the relative power for identification of VWD), laboratories were asked to Sample sets and sample testing provide numerical test data plus comment on the likelihood of The results from six well-characterised type 2B samples pro- VWD, as well as the particular subtype suggested by their own vided over the last eight-year period were analysed. These 2B laboratory findings. The interpretative choices comprised: VWD samples were dispersed within a total of 47 samples dis- (i) normal/not VWD, (ii) equivocal (neither clearly normal nor patched over this period, the other samples including type 1, 3, VWD), (iii) type 1VWD (mild/severe), (iv) type 2VWD (further Table 1: Baseline platelet counts, and phenotypic FVIII:C andVWF test data for the three unpublished type 2BVWD samples as- sessed in this report. Sample ID Platelet count FVIII:C VWF:Ag VWF:RCo VWF:CB Ag/ RCo Ag/ CB RCo/ Ag CB/ Ag (x109/L) 2006.1 042 62 71 40 39 1.8 1.8 0.6 0.6 2006.2 071 25 45 <10 20 5.0 2.3 0.2 0.4 2006.4 166 55 47 33 06 1.4 7.8 0.7 0.1 Normal reference 150–400 45–180 45–200 35–350 50–250 0.5–1.5 0.5–1.5 >0.7 >0.7 range (NRR; 18) 236 347 Downloaded by: University of Free State. Copyrighted material. 237 Favaloro et al. Early identification of type 2BVWD Figure 1:VWF:Multimer patterns for the new set of three type 2BVWD plasma samples detailed in the current report (shown in duplicate at differing inten- sities), together with a normal control sample and some resultant densitometry patterns. Note the relative loss of high-mo- lecular-weight (HMW)VWF in each sample such that sample 2006.1 showed the least loss of HMWVWF, followed by sample 2006.2, and lastly sample 2006.4. LMW, low-molecular- weight. subtyping as 2A, 2B, 2M, 2N was optional), or (v) type 3 VWD. panel is VWF:RCo plus FVIII:C). Current methodology for In all surveys, participants could specify an alternative con- VWF:Ag is primarily by either immuno-turbimetric (or latex- clusion or else choose not to complete this section. This report immuno-assay; LIA) or by enzyme-linked-immunosorbent- focuses on 2B VWD samples and assesses the relative power of assay (ELISA; Fig. 2). In earlier surveys, >90% of laboratories different laboratory ‘VWD-diagnostic-screening’ test panels in performed VWF:RCo assays, whereas currently <60% of lab- identifying these samples as VWD (or not), in determining oratories still perform this assay. Thus, some laboratories have whether a type 2 VWD type pattern was evident (e.g. evidence of either abandoned this assay or have replaced VWF:RCo assays VWF functional discordance), and whether such results would with alternative ‘functional’ assays (see later). Approximately indicate, in normal laboratory and clinical practice, the need for equal numbers of current participants perform VWF:RCo by further investigation (i.e. in particular for the current report, the newer automated platelet agglutination assays or by ‘classical’ need to refer patients to expert VWD test centres, or to perform platelet aggregometry (Fig. 2). The number of laboratories per- RIPA to help identify type 2BVWD).An ‘interpretive diagnostic forming VWF:CB has been more consistent throughout the sur- error’ was therefore defined to occur when 2B VWD samples veys at close to 50%. Current methodology for VWF:CB is by were identified as ‘normal’, ‘type 1’ or ‘type 3’ VWD. ELISA, with most laboratories using a commercial kit (Life Therapeutics [previously Gradipore], Sydney, Australia) and Evaluation of error rates fewer using in-house assays (Fig. 2).All in-house assays use type ‘Diagnostic error rates’ were determined by calculating the pro- 1/3 collagen mixture products in the coating step, in line with portion of ‘interpretive (or diagnostic) errors’ that occurred di- local recommendations (19). ‘VWF:Act’ is a new category for vided by the total number of diagnostic interpretations submitted the RCPA QAP and represents assays marketed by some manu- for that sample and test group, expressed as a percentage, as pre- facturers as ‘VWF:activity’ assays. These are monoclonal anti- viously outlined (15, 16). body (MAB) based assays (non-ristocetin-based and non-col- lagen-based) performed by either ELISA or LIA technology Results (Fig. 2). The former (ELISA) are marketed commercially by either Axis-Shield (Dundee, Scotland) or by Reaads Corgenix VWF test types and methodologies performed by (distributed byAbacus Diagnostics, Brisbane,Australia), and are laboratories historically based on the work of Goodall et al. (20–22). The All participant laboratories currently perform FVIII:C testing, latter (LIA) represents a new assay available from Instrumen- and also did so for previous surveys. In earlier surveys, all lab- tation Laboratory (‘HemosIL von Willebrand Factor Activity’ oratories performed VWF:Ag assays, but currently one partici- assay; Beckman-Coulter, Sydney Australia). Only three labora- pant laboratory does not perform this assay (VWD screening tories currently perform multimer analysis, although not rou- 237 348 Downloaded by: University of Free State. Copyrighted material. 238 Favaloro et al. Early identification of type 2BVWD Figure 2: Summary of current relative sub-methodologies for strumentation Laboratory/Beckman-Coulter (n=7)]; b) ELISA is per- VWF assays used by participants of the RCPA QAP. Number of formed using either commercial kits (Diagnostica Stago, n=2; Life Thera- laboratories (‘labs’; left y-axis) or percentage of laboratories (‘labs’; right peutics, previously Gradipore, Sydney, Australia; n=4) or in-house meth- y-axis) performing particular assay methods forVWF:Ag (LIA, ELISA or ods (n=8) primarily using antibodies from Dako (Sydney, Australia); VIDAS/ELFA),VWF:RCo (Automated agglutination or classical platelet c)VIDAS instrumentation (ELFA [enzyme-linked fluorescent assay] tech- aggregometry),VWF:CB (in-house or commercial kit), or ‘VWF:Activity’ nology; BioMerieux, Sydney, Australia) is used by only three laboratories. (LIA or ELISA). ForVWF:Ag: a) LIA is primarily performed using ForVWF:RCo: a) ‘agglutination’ is primarily performed (n=18) using the reagents from Diagnostica Stago (Sydney, Australia; n=23), Dade-Behring von Willebrand reagent from Dade-Behring (Sydney, Australia) and vari- (Sydney, Australia; n=6) or Instrumentation Laboratory (Beckman- ous instrumentation [but primarily Dade-Behring (n=7), or Diagnostica Coulter, Sydney, Australia; n=6) and using instrumentation from the Stago instruments (n=9)], and b) ‘aggregometry’ is performed using a same companies [i.e. Diagnostica Stago (n=18), Dade-Behring (n=11), In- wide variety of reagents and instrumentation. tinely for all samples. The specific breakdown of sub-methodol- VWF:RCo, VWF:CB and/or VWF:Act). The actual VWF panel ogies for VWF:Ag and VWF:RCo is detailed in Figure 2. used by these laboratories varied widely as follows: VWF:Ag and VWF:RCo only = 11 (21.2%), VWF:Ag and VWF:CB only VWF test panels performed by laboratories = 5 (9.6%), VWF:Ag and VWF:Act only = 1 (1.9%), VWF:Ag, Some laboratories have a very limited test panel [e.g. only VWF:RCo and VWF:CB = 16 (30.7%), VWF:Ag, VWF:RCo FVIII:C andVWF:RCo (n=1), or FVIII:C andVWF:Ag (n=10)]. and VWF:Act = 2 (3.8%), VWF:Ag, VWF:CB and VWF:Act = 4 However, around 80% of participant laboratories perform (11.5%), and VWF:Ag, VWF:RCo, VWF:CB and VWF:Act = 2 VWF:Ag, FVIII:C and at least one VWF functional assay (i.e. (3.8%). Table 2: Summary of returned phenotypic data and diagnostic error rates for type 2BVWD samples assessed by the RCPA Hae- matology QAP. Sample No. Returned phenotypic data (median values):*** Correct diag- Diagnostic interpretative error ID* Labs nostic inter- rates (%):***** ** VWF: VWF: VWF: ‘VWF: FVIII: Ag/ Ag/ Ag/ pretations: Normal ‘equivocal’ Type 1 Ag RCo CB Act’ C RCo CB ‘Act’ n (%)**** VWD 1998.4 25 44.5 19.0 11.2 NA 55.5 4.1 4.2 NA 19 (76.0) 0 8.0 8.0 1998.8 25 36.2 15.5 7.2 NA 36.0 4.2 5.7 NA 16 (64.0) 0 0 24.0 2000.2 19 42.4 19.5 10.8 NA 55.1 3.4 6.7 NA 14 (73.7) 0 0 10.5 2006.1 52 70.0 31.0 32.0 24.0 64.0 2.4 2.1 2.8 31 (59.6) 11.5 9.6 9.6 2006.2 52 45.0 15.0 15.0 15.5 34.0 2.7 2.7 2.5 27 (51.9) 0 0 34.6 2006.4 52 28.0 8.0 9.0 13.5 25.0 3.4 2.9 2.2 38 (73.1) 0 0 11.5 * First three samples listed have been the subject of previous reports (11,12) and are listed here for comparative purposes only. Last three samples represent new 2BVWD sample data. ** Number of partici- pant laboratories per survey has varied, but equalled 52 in 2006. *** Data shown as median values returned for each sample in each respective survey. **** Number (percent) of laboratories that indicated the possibility of type 2/2A/2B/2MVWD (thus fulfilling a need for further investigation by RIPA analysis). **** Percentage of laboratories identifying the type 2BVWD samples as ‘normal’, ‘equivocal’, or ‘type 1VWD’ respectively. NA = not available. 238 349 Downloaded by: University of Free State. Copyrighted material. 239 Favaloro et al. Early identification of type 2BVWD Phenotypic data returned by survey participants Data for the six 2B VWD samples is summarised in Table 2. As can been seen, all samples showed some evidence of functional VWF discordance with all functional VWF assays (i.e. VWF:Ag/VWF:functional [‘Ag/fun’] assay ratios >2.0 for each case), although the degree of discordance varied between samples and between functional VWF methods. Three samples (1998.4, 2000.2, 2006.1) gave normal median values for FVIII:C (i.e. >50%), and one sample (2006.1) gave normal median valu- es for VWF:Ag. In general, assay CVs were lowest for FVIII:C (usually <20%), followed by VWF:Ag (typically 10–25%), VWF:CB (typically 20–40%), and VWF:RCo (typically 25–50%). Assay CVs for the LIA based VWF:Act assay were generally comparable to these. Detailed comparative phenotypic data for the three pre- viously unpublished 2B VWD samples is provided in Figure 3. Detailed VWF assay ratio data for all samples are provided in Figures 4 and 5. This data is helpful in terms of interpreting error rates identified later. Most samples tended to show greater VWF functional discordance using the ratio of VWF:Ag/VWF:CB (‘Ag/CB’) or the reverse (VWF:CB/VWF:Ag or ‘CB/Ag’) com- pared to VWF:Ag/VWF:RCo (‘Ag/RCo’) or VWF:RCo/ VWF:Ag (‘RCo/Ag’). This was particularly evident for histori- cal samples (1998.4, 1998.8, 2000.2; Fig. 4), but notably less evi- dent with the new samples (2006.1, 2006.2, 2006.4; Fig. 5). Thus, for historical samples (Fig. 4), quite a few laboratories did not identify VWF functional discordance between VWF:Ag and VWF:RCo, whereas all but a few laboratories reported VWF functional discordance between VWF:Ag and VWF:CB. For the new samples (Fig. 5), a small number of laboratories did not identify VWF functional discordance using either approach, and there appeared less difference between VWF:RCo and VWF:CB data. To assess additional methodological reasons for such discre- pant findings, data was further analysed for the current data set according to sub-methodologies (Figs. 6 and 7). Users of VIDAS instrumentation (ELFA methodology) tended to report higher VWF:Ag data, and ELISA users tended to report the lowest data (Fig. 6A), although small VIDAS user numbers preclude defini- tive conclusions. Excluding obvious outliers, there was no ob- vious pattern of differential assay precision for VWF:Ag sub- methodologies. There was no consistent pattern of difference in VWF:RCo data between users of automated platelet aggluti- nation and platelet aggregometry based assays, although there was some evidence of better precision with automated methods (Fig. 6B). There was no consistent pattern of difference or preci- sion in VWF:CB data between users of commercial kit methods and in-house assays (Fig. 6C). For VWF:Act assays, results for users of ELISA methodology tended to be higher than those with Figure 3: Scatterplot ofVWF or FVIII:C level as reported by LIA methodology (Fig. 6D), although small ELISA user participants of the RCPA QAP for all assays assessed by the numbers again precludes definitive conclusions. Any methodo- QAP, and namelyVWF:Ag,VWF:RCo,VWF:CB,VWF:Act and logical differences noted above could lead to differences in FVIII:C, for the three new type 2BVWD plasma samples des- identification of functional discordance, and this was observed patched in 2006 (see alsoTable 2). Data shown as percent of nor- in some cases (Fig. 7).Thus, forVWF:RCo,VWF functional dis- mal (y-axis), with 100% reflecting the theoretical yield of a large-pooled cordance tended to be identified more often with automated pla- normal plasma. Small horizontal bars for each data set represent the median value. telet agglutination than with platelet aggregometry (Fig. 7A). For VWF:CB, there was a suggestion of better discordance ob- served for in-house assays compared to commercial kits (true for 239 350 Downloaded by: University of Free State. Copyrighted material. 240 Favaloro et al. Early identification of type 2BVWD Figure 4: Scatterplot ofVWF assay ratios using data forVWF assays as reported by participants of the RCPA QAP for VWF:Ag/VWF:function (left y-axis) orVWF:function/VWF:Ag (right y-axis), for the three historical type 2BVWD samples (1998.4, 1998.8, 2000.2) as perTable 2 and as in part previously reported (11,12). Data restricted to ratios ofVWF:Ag/VWF:RCo (‘Ag/RCo’) andVWF:Ag/VWF:CB (‘Ag/CB’) or the reverse ratios (as preferred by some workers; 1) As this is the data available. Small hori- zontal bars for each data set represent the median value. Shaded hori- Figure 5: Scatterplot ofVWF assay ratios using data forVWF zontal bars in each figure represent the region of interest for cut-off for assays as reported by participants of the RCPA QAP for identification of functionalVWF discordance. Current local cut-offs are VWF:Ag/VWF:function (left y-axis) orVWF:function/VWF:Ag 1.5 for Ag/Fun (or 0.7 for Fun/Ag; 18), but some workers use slightly dif- (right y-axis), for the three new type 2BVWD samples (2006.1, ferent cut-off values (e.g. cut-off of 2.0; 1), so a range of 1.5–2.0 for Ag/ 2006.2, 2006.4) as perTable 2 and as per current report. Data Fun (or 0.5–0.7 for Fun/Ag) is shown in figures. Results above the comprises ratios ofVWF:Ag/VWF:RCo (‘Ag/RCo’),VWF:Ag/VWF:CB shaded bar for Ag/Fun or below the bar for Fun/Ag would be acceptable (‘Ag/CB’) andVWF:Ag/VWF:‘Act’ (Ag/Act) or the reverse ratios (as pre- evidence of functionalVWF discordance (e.g. ?type 2A, 2B or 2MVWD) ferred by some workers; see Fig. 4 legend). Small horizontal bars for for further study, clarification and classification. Results below the each data set represent the median value. Shaded horizontal bars in each shaded bar for Ag/Fun or above the bar for Fun/Ag would be acceptable figure represent the region of interest for cut-off for identification of evidence for lack of functionalVWF discordance (e.g. normal or type 1 functionalVWF discordance. VWD depending on absoluteVWF levels determined). Results within the shaded bar would represent equivocal findings, but would still sug- gest some follow up, including repeat testing and possible further exten- sive testing. 240 351 Downloaded by: University of Free State. Copyrighted material. 241 Favaloro et al. Early identification of type 2BVWD A) C) B) D) Figure 6: Scatterplot ofVWF level as reported by participants of the RCPA QAP for allVWF assays assessed by the QAP, and namelyVWF:Ag,VWF:RCo,VWF:CB, andVWF:Act, and shown by sub-methodology (viz:VWF:Ag = LIA, ELISA orVIDAS/ELFA; VWF:RCo = automated platelet agglutination or classical platelet aggregometry;VWFCB = commercial kit or in-house method;VWF:‘Act’ = LIA or ELISA). Data shown as percent of normal (y-axis), with 100% reflecting the theoretical yield of a large pooled normal plasma, for the three new type 2BVWD plasma samples despatched in 2006 (see also Fig. 3). Small horizontal bars for each data set represent the median value. samples 2006.1 and 2006.2, but not 2006.3; Fig. 7B). For with the VWF panel of VWF:Ag, VWF:CB and VWF:Act VWF:Act, there was a similar suggestion of better discordance (8.3%, or 1/12 incorrect identifications). Interestingly, for observed for LIA based assays compared to ELISA-based kits VWF:Ag plus VWF:RCo performing laboratories, similar (Fig. 7C). numbers of laboratories performed automated analysis by pla- telet agglutination or by platelet aggregometry, and automation Diagnostic interpretive error rates for type 2BVWD did not protect laboratories from diagnostic identification errors This was assessed for the six samples with data summarised in (Fig. 8). Tables 2 and 3. There was a high error rate when laboratories A high proportion of identification errors were related to lab- only used a limited VWF test panel (i.e. VWF:Ag or VWF:RCo oratory interpretations rather than assay failures per se. Thus, for only). Indeed, an overall error rate of 76.9% misidentification as the misidentifications identified above, VWF functional assay normal or type 1VWD was obtained by these laboratories for the discordance was actually obtained in many cases, but was not three most recent type 2B VWD samples when an attempt at in- utilised by the laboratory in order to identify a possible type 2 terpretation was made. Such limited test panels obviously cannot VWD. In summary, for laboratories that misidentified the three identify functional VWF discordance. When laboratories used new cases of type 2B VWD as normal or type 1 VWD, functional VWF:Ag and VWF:RCo as their VWF test panel, an error rate of VWF assay discordance was actually obtained in 15/24 (62.5%) 25.7% was identified (comprising nine identifications of ‘nor- of test cases for VWF:RCo, 8/11 (72.7%) of test cases for mal’ or type 1 VWD from 35 submitted responses; Table 3, Fig. VWF:CB, and 1/3 (33.3%) of test cases for VWF:Act. Although 8). When laboratories added VWF:CB to the test panel of the split for VWF:RCo was fairly even between the two method- VWF:Ag and VWF:RCo, error rates fell to 10.9% (comprising 5 ologies in terms of misidentification (ie automated agglutination identifications of ‘normal’or type 1 VWD from 46 submitted re- vs. classical aggregation; Fig. 8), most cases of laboratory ‘self- sponses; Table 3, Fig. 8). Lower error rates were also associated interpretation errors’ belonged to users of automated aggluti- 241 352 Downloaded by: University of Free State. Copyrighted material. 242 Favaloro et al. Early identification of type 2BVWD nation (i.e. 6/7 [85.7%] test cases with misinterpretation errors actually showed evidence of VWF functional discordance). For VWF:CB, most cases of misidentification were in laboratories that utilised the commercial kit (i.e. 7/9 [77.8%]). Discussion In this report, focused on type 2B VWD identification, we have confirmed and extended the more general findings previously reported by us (11–16). In brief: – (a) Approximately 20% of laboratories participating in the A) RCPA currently use a core screening panel of VWF:Ag and FVIII:C only, or in one case VWF:RCo and FVIII:C only, which are insufficient to identify a qualitative VWF disorder such as 2B VWD. Understandably, these limited test panels led to a high error rate of misidentification (76.9%) of these samples as either normal or type 1 VWD. – (b) Addition of one functional VWF assay to VWF:Ag and FVIII:C reduces the misidentification rate, but the level of misidentifications is still significant. A VWF test panel com- prising VWF:Ag plus VWF:RCo was associated with a mis- identification error rate of over 25% (or one error for every four testing episodes; Fig. 8). Despite users of automated pla- telet agglutination tending to show better precision (Fig. 6B) and better evidence of functional VWF discordance (Fig. 7A), automation did not protect against VWD misidentifi- B) cation errors (Fig. 8). These findings are consistent with those previously reported by us for other type 2 VWD samples (15, 16). – (c) Performance of an additional functional VWF assay (i.e. VWF:CB) within the test panel was associated with a much lower error rate (Fig. 8). These findings are also consistent with those previously reported for other samples (15, 16). – (d) Particular functional VWF assay sub-methodologies tend to better identify VWF discordance (or possible type 2 VWD) than others (notably: (i) automated platelet aggluti- nation for VWF:RCo compared to classical platelet aggrego- metry, (ii) in-house VWF:CB assays compared to commer- cial kit methods, and (iii) automated LIA-based ‘VWF:Act’ assays compared to ELISA-based assays). This is a novel C) finding for the current report. – (e) Correct identification of functional VWF discordance (or possible type 2 VWD) differs according to the sample tested. Figure 7: Scatterplot ofVWF assay ratios using data forVWF For example, sample 2006.1 yielded normal median levels of assays as reported by participants of the RCPA QAP forVWF:Ag/VWF:function (left y-axis) orVWF:function/VWF:Ag VWF:Ag and FVIII:C (Table 1, Fig. 3), and many labora- (right y-axis), for the three new type 2BVWD samples (2006.1, tories (Fig. 5) did not observe VWF functional discordance, 2006.2, 2006.4) as perTable 2 and as per current report, but irrespective of whether they used VWF:RCo, VWF:CB, or with data separated according to sub-methodology as per Fig- VWF:Act. This sample also showed the lowest loss of HMW ure 6. Data comprises ratios ofVWF:Ag/VWF:RCo (‘Ag/RCo’), VWF in multimer analysis (Fig. 1). Accordingly, this sample VWF:Ag/VWF:CB (‘Ag/CB’) andVWF:Ag/VWF:‘Act’ (Ag/Act) or the re- resulted in many cases of misidentification (Table 2) of ‘nor- verse ratios (as preferred by some workers; see Fig. 4 legend). Smallhorizontal bars for each data set represent the median value. Shaded mal’ or ‘mild type 1 VWD’, as well as several reports of horizontal bars in each figure represent the region of interest for cut-off ‘equivocal’, and might represent a ‘worse-case’scenario type for identification of functionalVWF discordance. 2B VWD sample. The donor (aged 67) had been incorrectly diagnosed for many years as simply ‘thrombocytopenic’, and the current report (where 15/44 laboratories failed to report a discordance (Table 1, Fig. 5), and a greater loss of HMW type 2 VWD pattern) may help explain why. Her daughter’s VWF (Fig. 1). There were no reports of ‘equivocal’ and no results (sample 2006.2), yielded lower VWF and FVIII:C misinterpretations of ‘normal’for this case, but 18/43 labora- values (Table 2, Fig. 3), better evidence of functional VWF tories still misidentifed the sample as mild or severe type 1 242 353 Downloaded by: University of Free State. Copyrighted material. 243 Favaloro et al. Early identification of type 2BVWD Table 3: Summary of relatedVWF test panels versus diagnostic error rates for type 2BVWD samples assessed by the RCPA Hae- matology QAP. Sample No. ‘Incorrect’ diagnostic Number of incorrect diagnostic interpretations (n) according to the laboratoryVWF test panel ID* Labs interpretation used*** ** Total VWF:Ag & VWF:Ag, VWF:Ag & VWF:Ag or OtherVWF VWF:RCo VWF:RCo & VWF:CB VWF:RCo only panel com- VWF:CB binations 1998.4 25 ‘Equivocal’ 2 2 1 Type 1VWD 2 1 1998.8 25 Type 1VWD 6 6 2000.2 19 Type 1VWD 2 1 1 2006.1 52 ‘Equivocal’ 5 4 1 4 1 ‘Normal’ 6 2 1 2 Type 1VWD 5 1 2006.2 52 Type 1VWD 18 6 3 3 5 1 2006.4 52 Type 1VWD 6 2 2 2 * First three samples listed have been the subject of previous reports (11,12) and are listed here for comparative purposes only. Last three samples represent new 2BVWD sample data. ** Number of partici- pant laboratories per survey has varied, but equalled 52 in 2006. *** Blank entries are equal to zero. VWD. This occurred more often in laboratories using the more limited VWF assay panels (e.g. VWF:Ag and VWF:RCo only). The two main likely explanations are that these laboratories either do not understand the significance of functional VWF dis- cordance as suggesting type 2 VWD (i.e. educational issue) or else that these laboratories simply do not believe their own lab- oratory test data. We have previously reported that the VWF assay panel of VWF:Ag and VWF:RCo is associated with a higher rate of misidentification of normal or type 1 VWD samples as being type 2 than a more comprehensive panel in- cluding the VWF:CB (15, 16). That is, use of VWF:Ag and VWF:RCo is associated with a higher rate of false VWF func- tional discordance when type 1 VWD or normal samples are tested than test panels that incorporate the VWF:CB. Accord- ingly, it is likely that laboratories participating in the RCPA QAP Figure 8: Misidentification error rates for variousVWF assay panels. Data shown as composite percentage error (see Methods) for simply place more faith on the discordance when identified the new type 2BVWD samples when misidentified as either ‘normal’ or using the more comprehensive combinations. ‘type 1VWD’. Error rate (y-axis) shown separately for the following One significant novel finding in this report is the observed VWF test panels:VWF:Ag orVWF:RCo only (Ag or RCo),VWF:Ag plus difference between the various so-called ‘activity’ assays VWF:RCo (Ag & RCo),VWF:Ag plusVWF:RCo plusVWF:CB (Ag & (‘VWF:Act’). Better VWF functional discordance was observed RCo & CB). Data for laboratories performingVWF:Ag plusVWF:RCo using LIA based ‘VWF:activity’ assays compared to ELISA also shown according to sub-methodology (i.e. Ag & RCo (Auto) versus Ag & RCo (Agg)) (near equal numbers of laboratories). based. Findings need to be tempered, however, by the small number of current users, and need to be confirmed in subsequent studies. VWF:CB and VWF:RCo assays have previously been shown to better identify VWF functional discordance than ELISA based ‘VWF:Act’ assays (23, 24), but good comparative VWD. Sample 2006.4 was the least problematic of the cur- studies between VWF:CB, VWF:RCo and LIA based rent sample set, with the fewest incorrect identifications. Of ‘VWF:Act’assays are lacking. Neither of the previous reports on relevance, this sample showed the lowest levels of FVIII:C this methodology (25, 26) adequately deal with identification of and VWF parameters (Table 2, Fig. 3), the best evidence of functionalVWF discordance, nor are the comparisons with other functional VWF discordance (Table 2, Fig. 5) and the grea- functional VWF assays very comprehensive. Nevertheless, re- test loss of HMW VWF (Fig. 1). sults from the current report do suggest some promise of clinical utility for this assay. The test system uses a MAB that binds to It was of some interest to note that many laboratories reported VWF at a functional site, but the premise that this would confer numerical data that showed evidence of VWF functional discor- the assay some ability to measure VWF ‘activity’ (an argument dance, but then failed to identify the samples as probable type 2 previously forwarded by manufacturers of the ELISA based 243 354 Downloaded by: University of Free State. Copyrighted material. 244 Favaloro et al. Early identification of type 2BVWD MAB test systems [23, 24]) is still flawed. Nevertheless, it is pos- VWF:RCo assay, initially by replacing classical platelet aggre- sible that these tests show good correlation to the presence of gometry methods with automated methods, and more recently by HMW VWF, and that HMW VWF is preferentially ‘captured’by replacing the assay itself with VWF:CB or VWF:Act assays, or this assay system, so further study will help clarify their place in otherwise abandoning functional VWF assays altogether. We do VWD diagnostics. It also needs to be restated that neither of not support replacement of VWF:RCo with any other functional these VWF:Act assays (i.e. LIA or ELISA based) incorporate VWF assay (including VWF:CB), but emphasise that additional ristocetin or collagen within their test procedure. Thus, they assays such as VWF:CB are complementary to VWF:RCo. Al- would not be expected to behave identically to VWF:RCo nor though, by direct comparison, optimised VWF:CB assays do VWF:CB assays. Two versions of true ristocetin-based ELISA tend to better identify functional discordance in type 2A and 2B VWF:RCo assays have been reported (27–29), but these have not VWD compared to VWF:RCo, and are also less likely to mis- yet been taken up by general diagnostic laboratories because the identify false functional discordance (e.g. type 1 VWD or nor- assays are complex, and reagents are not yet readily available; mal samples as type 2VWD [15, 16]),VWF:RCo is still required thus confirmatory studies are still lacking. to identify some cases of type 2M VWD, and a dual functional Our overall findings also need to be considered in view of re- VWF panel of VWF:RCo and VWF:CB is likely to be diagnos- cent literature related to type 2B VWD. Baronciani et al. (30) tically more useful. Whether VWF:Act can replace VWF:RCo is have found in a large retrospective case series that the VWF:CB simply unknown, although also unlikely – as yet, no good com- better identifies functional VWF discordance than VWF:RCo, parative studies are available, and it must be remembered that both for 2B VWD cases with and without a lack of HMW VWF. VWF:Act assays do not incorporate ristocetin into their method- That is, there is some evidence that type 2B VWD reflects an in- ology. We are also concerned by the move by some laboratories herent collagen-binding defect in addition to its case-dependent to abandon VWF functional assays altogether (10/52 labora- (or case-inconsistent) loss of HMW VWF. Thus, the lowered tories participating in the RCPA QAP in 2006), since such assays relative VWF:CB binding in type 2B VWD may reflect a com- are required to identify type 2 VWD, and to discriminate type 2 bination of lowered plasma HMW VWF plus the presence of a from other VWD subtypes. We suspect that this may be occur- specific VWF collagen-binding defect. In contrast, VWF:RCo ring, because functional VWF assays tend to be more technically may be less likely to show assay discordance, because the assay demanding, time consuming and less precise than the simpler reflects a combination of reduced HMW VWF (leading to lower antigen assays. In addition, this may be a consequence of clinical VWF:RCo) together with enhanced ristocetin binding (leading pressures requesting ‘stat’ VWF testing, which can be accom- to potential ‘inflation’of VWF:RCo). Interestingly, the ability of modated using LIA testing for VWF:Ag, but less easily accom- VWF:CB assays to better detect type 2B VWD by either mech- modated by functional VWF assays. The general move towards anism (collagen binding defect or loss of HMW VWF) depends automation (i.e. performance of automated platelet aggluti- on the collagen source (type 1/3 mixture collagens work better nation assays rather than classical platelet aggregometry or auto- than purified type 3 collagen [30]). Although consistent with mated VWF:Act based assays instead of VWF:RCo assays) is previous published findings (4, 5, 19), it had always been as- also of interest, as is the move towards LIA technology for sumed that the ability of type 1/3-mixture collagens to better VWF:Ag, at the expense of current ELISA technology and pre- identify 2B VWD was related solely to better capture of HMW vious Laurell gel (EID) technology.This is also the experience of VWF. In any case, such findings support our assertion regarding other external QAPs (37–39). the added value of VWF:CB testing in VWD diagnostics, par- In conclusion, laboratories participating in the RCPA Hae- ticularly when investigating the possibility of type 2 VWD, and matology QAP were able in most cases to identify functional preferentially using a type 1/3 mixture collagen based assay.This VWF discordance. However, in a substantial number of cases, view is now supported by independent findings by many others functional VWF discordance was either not demonstrated, or (31–36). It is worth noting that in our geographic region, all in- when observed was not taken into account for the interpretation house and commercial VWF:CB assays used by RCPA QAP par- of possible VWD typing. We could also identify differential ticipants appear to be based on type 1/3-mixture collagens. power depending on the VWF functional panel used (e.g. fewer Nevertheless, in-house based VWF:CB assays appeared, in the errors when laboratories added theVWF:CB to the basic panel of current study, to more consistently identify VWF functional dis- VWF:Ag and VWF:RCo) or on the functional sub-methodology cordance in type 2B VWD than the commercial kit method. The (e.g. better discordance observed by: users of LIA based company producing this kit (Life Therapeutics; previously VWF:Act assay compared to ELISA based, users of automated Gradipore) has transferred manufacturing operations from Aus- VWF:RCo compared to classical methods, users of in-house tralia to the US, and it is possible that the current findings reflect VWF:CB assays compared to commercial methods). We thus some manufacturing process change. Alternatively, it might re- recommend that laboratories incorporate VWF:CB into their late to the level of experience in laboratories, with ‘in-house’ VWF test panel (using a type 1/3-collagen mixture-based assay) assay users being the most experienced. as this will better more consistently enable identification of func- Of additional interest is the fall in relative utilisation of tional VWF discordance present in type 2 VWD, or functional VWF:RCo assays. Currently, <60% of laboratories still perform VWF concordance in type 1 VWD and normal samples, when this assay, whereas >90% of laboratories previously did so (11, compared to more limited testing. We do not support replace- 12). In contrast, performance of VWF:CB over the years in our ment of VWF:RCo assays with another functional VWF assay geographic locality has been fairly consistent at around 50% such as VWF:CB or VWF:Act assays, since these assays may (11–16). Many laboratories appear to be replacing their miss some type 2M VWD. The role of VWF:Act assays in VWD 244 355 Downloaded by: University of Free State. Copyrighted material. 245 Favaloro et al. Early identification of type 2BVWD Figure 9: An algorithm containing one recommended approach criminate type 2NVWD/haemophilia A]; perform a DDAVP challenge to the current investigation ofVWD. The important considerations test if deemed appropriate [will help better identify some forms ofVWD are: perform a thorough clinical and family review; request the appropri- plus determine subsequent management]. This algorithm should be ate tests for investigation (at a minimum, for diagnostic-screening: pla- treated as a guide only; although it will help appropriately define the ma- telet count/morphology, FVIII:C,VWF:Ag,VWF:RCo,VWF:CB; if clinical jority of cases, some isolated cases may not fit this scheme, and it is im- history is strong, RIPA as part of a platelet function study should also be portant that all patients be treated and managed as individuals. Finally, considered); repeat tests for confirmation (when normal with strong his- please seek expert local opinion and also refer to other expert recom- tory or when abnormal for all test cases); follow up abnormal test re- mendations (1–5), as different geographic localities may require a slightly sults further as required (e.g.VWF:Multimers [e.g. if type 2M suspected different approach, and some experts may have a different working ap- and management indicates differential therapy], RIPA [particularly to help proach. Data using theVWF:Act assays described in the current report identify/discriminate type 2B/platelet-typeVWD, but also in some other is too preliminary to include in this figure, which has been updated and cases of suspectedVWD],VWF:FVIII binding assay [to help identify/dis- modified from reference 5. 245 356 Downloaded by: University of Free State. Copyrighted material. 246 Favaloro et al. Early identification of type 2BVWD diagnostics is by no means clear, but results with the LIA based be undertaken and considered in a broad sense when VWF levels assays suggest that continued assessment is warranted. are found to be around the reference range cut-off values, or if a We caution that our findings should not be construed as sup- conclusion of ‘mild type 1 VWD’ is being considered (18). A port for performing only the limited VWF test panel evaluated DDAVP-challenge test is also recommended as part of the pa- here; this panel simply reflects the current range of the RCPA tient work up wherever appropriate (2, 40). Finally, if labora- QAP’s proficiency-testing capacity and the participant testing tories can not provide a comprehensive testing process for inves- set. In addition to clinical criteria, the correct diagnosis of VWD tigations into VWD, then samples and/or patients should be re- requires a comprehensive and appropriate panel of tests. At a ferred to expert VWD testing centres for investigation. An algo- minimum diagnostic screening level, this panel must include rithm of a general and currently recommended laboratory test FVIII:C,VWF:Ag and at least one, but preferably more than one, approach is provided in Figure 9. functional VWF assay. Subsequent additional testing, including VWF:Multimer analysis, RIPA, and VWF:factor VIII binding Acknowledgments will provide supplementary or confirmatory information and/or The authors thank scientific and clinical personnel from laboratories that VWD-subclassification (1–5), and must be performed if it will participated in our surveys. We also thank other staff and haemostasis com- help guide clinical decisions and subsequent therapeutic sup- mittee members of the RCPA QAP for their ongoing support and assistance, port. In particular, RIPA is essential in order to identify and dif- and our VWD patients who volunteered blood for use in the surveys. ferentiate type 2B VWD from other VWD subtypes (1, 6), and RIPA mixing studies are also recommended to help distinguish Abbreviations type 2B VWD from pseudo-, or platelet-type, VWD (6, 17). Al- FVIII:C, factor VIII: coagulant (assay); MAB, monoclonal antibody; though not recommended as a general investigative procedure, NRR, normal reference range; QAP, QualityAssurance Program; RIPA, genetic studies should also be considered for difficult or interest- ristocetin-induced platelet agglutination /aggregation; VWD, von ing cases (6–9). Whether or not to take into consideration the pa- Willebrand disorder/disease; VWF, von Willebrand factor; VWF:Ag, tient’s blood group when investigating patients or diagnosing von Willebrand factor antigen (assay); VWF:CB, collagen binding ac-tivity/assay) for VWF; VWF:Rco, ristocetin cofactor assay for VWF. VWD is contentious, but we would recommend that such testing References 1. Sadler JE, Budde U, Eikenboom JCJ, et al. and the 9. Casais P, Carballo GA, Woods AI, et al. 2R924Q 16. Favaloro EJ, Bonar B, Kershaw G, et al., on behalf Working Party on von Willebrand Disease Classifi- substitution encoded within exon 21 of the von Wille- of the RCPA QAP in Haematology. Reducing errors in cation. Update on the pathophysiology and classifi- brand Factor gene related to mild bleeding phenotype. identification of von Willebrand disease: The experi- cation of von Willebrand disease: a report of the Sub- Thromb Haemost 2006; 96: 228–230. ence of the Royal college of Pathologists of Australasia committee on von Willebrand Factor. J Thromb Hae- 10. RCPA Haematology QAP contact and program de- Quality Assurance Program. Semin Thromb Hemost most 2006; 4: 2103–2114. tails available at http://www.rcpaqap.com.au/. Last ac- 2006; 32: 505–513. 2. Michiels JJ, Berneman Z, Gadisseur A, et al. Clas- cessed 14th September, 2006. 17. Favaloro EJ. 2B or not 2B? Differential identifica- sification and characterization of hereditary types 2A, 11. Favaloro EJ, Smith J, Petinos P, et al., on behalf of tion of type 2B, versus pseudo-von Willebrand disease. 2B, 2C, 2D, 2E, 2M, 2N, and 2U (unclassifiable) von the RCPA Quality Assurance Program (QAP) in Hae- Br J Haematol 2006; 135: 141–142. Willebrand disease. Clin Appl Thromb Hemost 2006; matology Haemostasis Scientific Advisory Panel. Lab- 18. Favaloro EJ, Soltani S, McDonald J, et al. Reassess- 12: 397–420. oratory testing for von Willebrand's disease: An assess- ment of ABO-blood group, gender and age on labora- 3. Budde U, PieconkaA, Will K, et al. Laboratory test- ment of current diagnostic practice and efficacy by tory parameters used to diagnose von Willebrand Dis- ing for von Willebrand disease: Contribution of means of a multi-laboratory survey. Thromb Haemost order (VWD): Potential influence on the diagnosis ver- multimer analysis to diagnosis and classification. 1999; 82: 1276–1282. sus the potential association with risk of thrombosis. Semin Thromb Hemost 2006; 32: 514–521. 12. Favaloro EJ, Thom J, Baker R, on behalf of the Aus- Am J Clin Pathol 2005; 124: 910–917. 4. Favaloro EJ. von Willebrand factor (VWF) collagen tralasian Society for Thrombosis and Haemostasis 19. Favaloro EJ. Collagen binding assay for von Wille- binding (activity) assay (VWF:CBA) in the diagnosis Emerging Technologies Group. Assessment of current brand Factor (VWF:CBA): Detection of von Wille- of von Willebrand's disorder (VWD): A 15-year jour- diagnostic practice and efficacy in testing for von brands Disease (VWD), and discrimination of VWD ney. Semin Thromb Hemost 2002; 28: 191–202. Willebrand's disease: Results from the second Austra- subtypes, depends on collagen source. Thromb Hae- 5. Favaloro EJ. Laboratory identification of von lasian laboratory survey. Blood Coagul Fibrinolysis most 2000; 83: 127–135. Willebrand Disease: Technical and scientific perspec- 2000; 11: 729–738. 20. Goodall AH, Jarvis J, Chand S, et al. An immuno- tives. Semin Thromb Hemost, 2006; 32: 456–471. 13. Favaloro EJ, Bonar R, Sioufi J, et al., on behalf of radiometric assay for human factor VIII von Wille- 6. Nurden P, Lanza F, Bonnafous-Faurie C, Nurden A. the RCPA QAP in Haematology Haemostasis Commit- brand factor (VIII:VWF) using a monoclonal antibody A second report of platelet-type von Willebrand dis- tee. Laboratory diagnosis of von Willebrand disorder: based that defines a functional epitope. Br J Haematol ease with a Gly233Ser mutation in the GPIBA gene. Current practice in the Southern Hemisphere. Am J 1985; 59: 565–577. Thromb Haemost 2007; 97: 319–321. Clin Path 2003; 119: 882–893. 21. Chand S, McCraw A, Hutton R, et al. A two-site, 7. Cumming A, Grundy P, Keeney Set al.; on behalf of 14. Favaloro EJ, Bonar R, Kershaw G, et al., on behalf monoclonal antibody based immunoassay for von the UK Haemophilia Centre Doctors’Organisation. An of the RCPA QAP in Haematology. Laboratory diag- Willebrand factor- Demonstration that VWF function investigation of the von Willebrand factor genotype in nosis of von Willebrand disorder: Quality and diag- resides in a conformational epitope. Thromb Haemost UK patients diagnosed to have type 1 von Willebrand nostic improvements driven by peer review in a multi- 1986; 55: 318–324. disease. Thromb Haemost 2006; 96: 630–641. laboratory test process. Haemophilia 2004, 10: 22. Murdock PJ, Woodhams BJ, Mathews KB, et al. 8. Hilbert L, Nurden P, Caron C, et al., and the IN- 232–242. von Willebrand factor activity detected in a monoclon- SERM Network on Molecular Abnormalities in von 15. Favaloro EJ, Bonar R, Kershaw G, et al., on behalf al antibody-based ELISA: an alternative to the Ristoce- Willebrand Disease. Type 2N von Willebrand disease of the RCPA QAP in Haematology. Laboratory diag- tin cofactor platelet agglutination assay for diagnostic due to compound heterozygosity for R854Q and a nosis of von Willebrand disorder: Use of multiple func- use. Thromb Haemost 1997; 78: 1272–1277. novel R763G mutation at the cleavage site of von Wille- tional assays reduces diagnostic error rates. Lab Hema- 23. Favaloro EJ, Henniker A, Facey D, et al. Discrimi- brand factor propeptide. Thromb Haemost 2006; 96: tol 2005; 11: 91–97. nation of von Willebrands disease (VWD) subtypes: 290–294. Direct comparison of von Willebrand factor:collagen 246 357 Downloaded by: University of Free State. Copyrighted material. 247 Favaloro et al. Early identification of type 2BVWD binding activity/assay (VWF:CBA) with monoclonal 29. Vanhoorelbeke K, Pareyn I, Schlammadinger A, et says with different binding affinities for low, medium antibody (MAB) based ELISA VWF-detection sys- al. Plasma glycocalicin as a source of GPIb alpha in the and high multimers of von Willebrand factor. Semin tems. Thromb Haemost 2000; 84: 541–547. von Willebrand factor in the ristocetin cofactor ELISA. Thromb Hemost 2002; 28: 139–147. 24. Favaloro EJ. Discrimination of von Willebrands Thromb Haemost 2005; 93: 165–171. 36. Casonato A, Pontara E, Bertomoro A, et al. von disease (VWD) subtypes: Direct comparison of com- 30. Baronciani L, Federici AB, Cozzi G, et al. von Willebrand factor collagen binding activity in the diag- mercial ELISA-based options used to detect qualitative Willebrand factor collagen binding assay in von Wille- nosis of von Willebrand disease: An alternative to ris- von Willebrand factor (VWF) defects. Am J Clin Path- brand disease type 2A, 2B, and 2M. J Thromb Haemost tocetin cofactor activity? Br J Haematol 2001; 112: ol 2000; 114: 608–618. 2006; 4: 2088–2090. 578–583. 25. De Vleeschauwer A, Devreese K. Comparison of a 31. Fischer BE, Thomas KB, Dorner F. von Willebrand 37. Hayes TE, Brandt JT, Chandler WL, et al. External new automated von Willebrand factor activity assay factor: measuring its antigen of function? Correlation peer review quality assurance testing in von Willebrand with an aggregation von Willebrand ristocetin cofactor between the level of antigen, activity, and multimer size disease: The recent experience of the United States activity assay for the diagnosis of von Willebrand dis- using various detection systems. Thromb Res 1998; 91: College of American Pathologists Proficiency Testing ease. Blood Coag Fibrinolysis 2006; 17: 353–358. 39–43. Program. Semin Thromb Hemost 2006; 32: 499–504. 26. Sucker C, Senft B, Scharf RE, et al. Determination 32. Hubbard AR, Sands D, Chang AC, et al. Standard- 38. Kitchen S, Jennings I, Woods TAL, et al. Labora- of von Willebrand factor activity: Evaluation of the isation of von Willebrand Factor in therapeutic concen- tory tests for measurement of von Willebrand factor HaemosIL™ assay in comparison with established pro- trates: calibration of the 1st International Standard for show poor agreement between different centres: Re- cedures. Clin Appl Thromb/Hemostas 2006; 12: von Willebrand Factor Concentrate (00/514). Thromb sults from the UK National External Quality Assess- 305–310. Haemost 2002; 88: 380–386. ment Scheme for Blood Coagulation. Semin Thromb 27. Vanhoorelbeke K, Cauwenberghs N, Vauterin S, et 33. Hubbard AR. von Willebrand factor standards for Hemost 2006; 32: 492–498. al. A reliable and reproducible ELISA method to plasma and concentrate testing. SeminThromb Hemost 39. Meijer P, Haverkate F. An external quality assess- measure ristocetin cofactor activity of von Willebrand 2006; 32: 522–528. ment programme for von Willebrand Factor laboratory factor. Thromb Haemost 2000; 83: 107–113. 34. Federici AB, Canciani MT, Forza I, et al. Ristocetin analysis. An overview from the European Concerted 28. Federici AB, Anciani MTC, Forza I, et al. A sensi- cofactor and collagen binding activities normalized to Action on Thrombosis and Disabilities Foundation. tive ristocetin co-factor activity assay with recom- antigen levels for a rapid diagnosis of type 2 von Wille- Semin Thromb Hemost 2006; 32: 485–491. binant glycoprotein Iba for the diagnosis of patients brand disease. Single center comparison of four differ- 40. Favaloro EJ. Laboratory monitoring of therapy in with low von Willebrand factor levels. Haematologica ent assays. Thromb Haemost 2000; 84: 1127–1128. von Willebrand Disease: Efficacy of the PFA-100® and 2004; 89: 77–85. 35. Neugebauer BM, Goy C, Budek I, et al. Compari- VWF:CB as coupled strategies. Semin Thromb He- son of two von Willebrand factor collagen-binding as- most 2006; 6: 566–576. 247 358 Downloaded by: University of Free State. Copyrighted material. S-28 248 Thrombosis Research 134 (2014) 393–403 Contents lists available at ScienceDirect Thrombosis Research j ourna l homepage: www.e lsev ie r .com/ locate / th romresRegular ArticleEvaluating errors in the laboratory identification of von Willebrand disease in the real worldEmmanuel J. Favaloro a,⁎, Roslyn A. Bonar b, Muriel Meiring c, Elizabeth Duncan d, Soma Mohammed a, John Siou bfi , Katherine Marsden b a Department of Haematology, Institute of Clinical Pathology and Medical Research (ICPMR), Pathology West, Westmead Hospital, Westmead, NSW, Australia b RCPAQAP Haematology, Suite 201, Level 2, 8 Herbert Street, St Leonards, NSW, 2065, Australia c Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa d Department of Haematology, South Australia Pathology, Royal Adelaide Hospital, Adelaide, AustraliaAbbreviations: AVWS, acquired von Willebrand syn immunosorbent assay; EQA, external quality assessment glycoprotein Ib; HMW, high molecular weight (VWF); L von Willebrand disease; VWF, von Willebrand factor; V [generic]); VWF:Ag, VWF antigen; VWF:CB, collagen bind ristocetin cofactor (assay for VWF). ⁎ Corresponding author at: Department of Haematolog and Medical Research (ICPMR), Pathology West, Westme 2145, Australia. Tel.: +61 2 9845 6618; fax: +61 2 9689 E-mail address: emmanuel.favaloro@health.nsw.gov.a http://dx.doi.org/10.1016/j.thromres.2014.05.020 0049-3848/Crown Copyright © 2014 Published by Elseviea b s t r a c ta r t i c l e i n f oArticle history: Introduction: von Willebrand disease (VWD), reportedly the most common bleeding disorder, arises from defi- Received 10 February 2014 ciency and/or defects of von Willebrand factor (VWF). Assessment requires a wide range of tests, including Received in revised form 6 May 2014 VWF activity and antigen. Appropriate diagnosis including differential identification of qualitative vs quantitative Accepted 13 May 2014 defects has important management implications, but remains problematic for many laboratories and clinicians. Available online 21 May 2014 Methods:Data using a large set (n=29) of varied plasma samples comprising both ‘quantitative’VWF deficiency (‘Type 1 and 3’VWD) vs ‘qualitative’ defects (‘Type 2VWD’) tested in a cross-laboratory setting has been evaluated Keywords: von Willebrand disease to assess the ability of real world laboratories to differentially identify these sample types. von Willebrand factor Results: Different VWF assays and activity/antigen ratios show different utility in VWD and type identification. laboratory testing VWD identification errors were often linked to high inter-laboratory test variation and result misinterpretation interpretation (i.e., laboratories failed to correctly interpret their own test panel findings). Thus, moderate quantitative VWF diagnosic errors deficient samples were misinterpreted as qualitative defects on 30/334 occasions (9% error rate); 17% of these assay variability errors were due to laboratories misinterpreting their own data, which was instead consistent with quantitativedeficiencies. Conversely, whilst qualitative VWF defects were misinterpreted as quantitative deficiencies at a similar error rate (~9%), this was more often due to laboratories misinterpreting their data (~50% of errors). For test-associated errors, ristocetin cofactor was associated with the highest variability and error rate, which was at least twice that using collagen binding. Conclusion: These findings in part explain the high rate of errors associated with VWD diagnosis.Crown Copyright © 2014 Published by Elsevier Ltd. All rights reserved.Introduction vonWillebrand disease (VWD) is reportedly themost common con- genital bleeding disorder; nevertheless, its differential identification and clinical diagnosis remains problematic for many laboratories and clinicians [1,2]. VWD arises from deficiency and/or defects of von Willebrand factor (VWF), an adhesive plasmaprotein essential for effec- tive primary haemostasis. VWF possesses many functional properties, including binding to platelets via several receptors, most notablydrome; ELISA, enzyme linked ; FVIII:C, FVIII coagulant; GPIb, IA, latex-immuno-assay; VWD, WF:Act, VWF activity (assays ing (assay for VWF); VWF:RCo, y, Institute of Clinical Pathology ad Hospital, WESTMEAD, NSW, 2331. u (E.J. Favaloro). r Ltd. All rights reserved.glycoprotein Ib (GPIb), binding to sub-endothelial matrix components (most notably collagen), and binding and protecting factor VIII (FVIII) function [3]. The first two noted functions are essential for overall pri- mary haemostasis and are respectively assessed in the laboratory by surrogate tests such as VWF ristocetin cofactor (VWF:RCo) and collagen binding (VWF:CB) [1,2,4,5], although alternative assays may also be employed. Themost recent classification scheme from the International Society on Thrombosis and Haemostasis (ISTH) recognizes six different types of VWD [5]. Type 1 represents a partial quantitative VWF deficiency, with VWF essentially functionally normal, but produced in lowered quantity. Type 3 VWD represents ‘complete’ deficiency of VWF. Type 2 VWD rep- resents a heterogeneous group of qualitative VWFdefects that comprise (i) 2A VWD (loss of high molecular weight (HMW) VWF), (ii) 2B VWD (enhanced functional binding of VWF that leads to loss of HMW VWF and typically mild thrombocytopenia), (iii) 2N VWD (loss of VWF- FVIII binding), and (iv) 2M VWF (VWF dysfunction not associated with loss of HMWVWF). The proper identification of VWDand differen- tiating its type is important for therapeutic management [6].248 249 394 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403In normal practice, VWD and VWD type can be determined by labo- ratory testing that encompasses a broad panel of different tests [1,2,4,5]. Virtually all laboratories perform VWF antigen (VWF:Ag) and FVIII coagulant (FVIII:C) [1], respectively measuring VWF protein and FVIII activity. VWF:Ag is most usually assessed using either ELISA (enzyme linked immunosorbent assay) or LIA (latex-immuno-assay) tech- nologies (Table 1). The most common activity based test is VWF:RCo [1,2,4,5], usually performed as a platelet agglutination assay using aggregometry or automatedmethodswith standard coagulation instru- ments. VWF:CB is assessed by a proportion of laboratories, most usually by ELISA [1,2,4,5]. Additional ‘activity’ assays include newly released commercial assays, many based on LIA, although representing distinct assays, using distinct reagents, and detecting VWF differently [7–10]. Two such assays do not use ristocetin [7,8], so cannot claim to be VWF:RCo assays, although many laboratories may use these instead of VWF:RCo. VWF:CB and/or VWF:RCo can also be detected by other methodologies including LIA, fluorescence and flow cytometry [9–12], and VWF can also be assessed by other methods including multimer analysis to determine loss of HMWVWF aswell as structural abnormal- ities [13]. In brief, type 1 VWD is identified as a quantitative deficiency of VWF, with the level of deficiency correlatingwith the severity of the disorder, so that terms such as ‘mild’, ‘moderate’ or ‘severe’ type 1VWDare some- times employed. A new concept related to ‘low VWF’ as a risk factor for bleeding, without a formal diagnosis of VWD, has also recently emerged [14]. Importantly, for all these cases, low levels of VWF:Ag, VWF:RCo, VWF:CB and other VWF activity assays (generically abbreviated ‘VWF: Act’ in this paper) will be identified by laboratories, but since VWF is functionally normal, similar (‘concordant’) levels will be identified using all assays, so the ratio of any assay to another will be close to one. In practice, these cases are identifiedby a low level of VWF together with a ratio of VWF activity (VWF:RCo, VWF:CB and/or VWF:Act) to VWF:Ag above 0.7 [1,2,4,5,14]. In contrast, in qualitative defects (i.e., type 2 VWD), activity-based assays identify some VWF defect, with this defect helping to character- ize VWD type. Thus, loss of HMW VWF (e.g., type 2A and 2B VWD) is identified by a relatively larger reduction in VWF:RCo, VWF:CB and VWF:Act compared to VWF:Ag; this VWF ‘discordance’ is expressed by a ratio of VWF activity to VWF:Ag typically below around 0.6 or 0.7 [1,2,4,5,14]. Type 2M reflects a variety of potential functional defects, with most common forms representing platelet GPIb binding defects; hence, VWF:RCo/Ag will be low, but VWF:CB/Ag may be normal. Although we and others have previously reported on variability of cross-laboratory results for VWD, as well as associated errors in diag- nostic interpretations from composite test panels [15–25], our last major report in this area was in 2007 [21]. We have therefore assessed the contemporary ability of real world laboratories to differentially identify VWF deficiency (potentially repre- sentative of type 3 VWD, type 1 VWD or low VWF) and qualitativeTable 1 Summary of VWF test methods employed in VWD diagnostic practice. VWF assay Description VWF:Ag Assessment of VWF protein level using an 'antigen' assay. Typ flow cytometry utilised infrequently. VWF:Rco Assessment of VWF activity level utilising ristocetin and an 'a methods. For platelet agglutination, tests can be performed u (automated method). Alternate methods including flow cyto VWF:CB Assessment of VWF activity level utilising collagen. Typically p Werfen-IL 'activity' assay Assessment of VWF activity level utilising a monoclonal antib binding region of VWF. Performed by LIA based method. An E antibody and concept was previously utilised. Siemens Innovance Assessment of VWF activity level utilising a Glycoprotein Ib b 'activity' assay within a recombinant molecule that facilitates VWF binding. (but not from Siemens) have also been previously utilised. Abrreviations: ELISA, enzyme linked immunosorbant assay; LIA, latex-particle immunoassay; Vdeficiency (potentially representative of types 2A, 2B or 2M VWD) using a large composite set (n= 29) of plasma samples. The differential utility of different VWF assays and different VWF activity to antigen ratios, the variability of laboratory test data, identification errors related to VWD and VWD ‘type’, and their cause have all been evaluated. Methods Setting This study derives from the Royal College of Pathologists of Australasia Quality Assurance Program (RCPAQAP) in Haematology (bhttp://www.rcpaqap.com.auN). Table 2 details the samples (n = 75) distributed to the external quality assessment (EQA) participants from the start of data collection (in 1998) to current (spanning 15 years in total). The principal author’s laboratory acts as a ‘host-laboratory’, including sample preparation and homogeneity/stability testing prior participant dispatch. Participants to this VWF/VWD EQA and therefore of the current study (current n = 55) derive from Australia (n = 34), New Zealand (n = 5), Asia (n = 10), South Africa (n = 4), India (n = 1) and France (n= 1). The majority of these laboratories perform VWF testing for the investigation of VWD for clinical diagnosis and management and most would not be considered expert laboratories. Thus, the composite of laboratories would be generally representative of ‘real-world’ laboratories, defined here as the realm of practical or actual experience, as opposed to the abstract, theoretical, or idealized experience. Samples The current report details results using 29 samples from 2006 to 2013 inclusive (Tables 2 and 3), comprising a mixture of true patient and in-house prepared samples. Although patient samples are preferred in research evaluations, within an EQA setting it is important to balance this desire against ethical and logistical challenges around regular collections from VWD-affected patients. In-house prepared material also carries specific design advantages related to desired specifications. The 29 samples distributed comprised:(i) Four pooled normal samples (two different batches; target VWF ~100U/dL); (ii) One sample from a heterozygous type 3 VWD carrier (‘2008- 3het’); VWF ~60U/dL; (iii) Seven samples mildly to moderately deficient in VWF (one patient sample and three different batches of in-house prepared material; targets VWF 25-40 U/dL); (iv) Three samples severely deficient in VWF (two different batches of in-house prepared material; target VWF ~ 10 U/dL);ically performed by either ELISA or LIA based methods. Alternative methods including gglutination' assay. Typically performed by either platelet agglutination or LIA based sing an aggregometer (semi-automated method) or a coagulation based instrument metry utilised infrequently. erformed by ELISA. Alternative methods including flow cytometry utilised infrequently. ody binding assay, where the antibody is directed against the platelet Glycoprotein Ib LISA based method from a different manufacturer (i.e., not Werfen-IL) using a similar inding method. The system employs two gain of function Glycoprotein Ib mutations Performed by LIA based method. ELISA based methods employing a similar concept WF, von Willebrand factor; VWD, von Willebrand disease. 249 250 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 395 Table 2 Summary of surveys undertaken and samples distributed. Year No. No. Sample types Data published in samples laboratories Normal Equivocal VWD or VWD-like samples Other Other Detail 1 2A 2B 2M 2N 3 1998 10 25 3 2 1 1 2 1 [15] 2000 7 19 1 1 1 1 1 1 1 [16] 2002 8 44 4 2 1 1 [17] 2003 8 44 1 4 1 1 1 [18] 2004 6 45 2 1 1 2 [VWF] [19] 2005 4 49 2 1 1 [20] 2006 4 52 3 [21] 1 current report 2007 4 55 1 1 1 1 AVWS current report 2008 4 56 1 1 1 1 AVWS current report 2009 4 55 2 1 1 current report 2010 4 56 1 1 1 1 AVWS current report 2011 4 58 2 1 1 AVWS current report 2012 4 59 1 1 1 1 AVWS current report 2013 4 57 1 2 1 current report Totals: 75 17 12 14 5 9 2 1 8 7 Current report 29 4 1 9 2 3 1 0 4 5 Abbreviations: AVWS, acquired von Willebrand Syndrome; VWF, von Willebrand factor; VWD, von Willebrand disease [VWF], VWF concentrate.(v) Three samples totally deficient in VWF (commercially obtained; VWF = 0 U/dL; purchased from Banksia Scientific (Bulimba, Australia) and prepared by Affinity Biologicals (Ontario, Canada)); (vi) Five samples selectively deficient in HMW VWF (two different batches of type 2B VWD patient plasma plus two different batches of type 2A VWD-like plasma prepared in house; one of the latter represented a small pool of cryosupernatant plasma, the other was prepared using a process similar to that previous identified [27]); (vii) One type 2M VWD (patient) sample;Table 3 Summary of samples utilised for current study. Sample Type Sample represents or mimics Sample ID⁎ Normal pool Normal 2006-N, 2008-N, 2010-N Normal pool Normal 2013-N Equivocal Borderline normal VWF 2008-3het Mildly VWF deficient ‘Low VWF’ 2007-1 m Moderately VWF deficient ‘mild/moderate VWD type 1' 2010-1 m, 2011a-1 m Moderately VWF deficient ‘mild/moderate VWD type 1’ 2009-1 m, 2011b-1 m, 2012-1 m Moderately VWF deficient ‘mild/moderate VWD type 1’ 2013-1 m Severely VWF deficient ‘severe VWD type 1’ 2009-1 s, 2013-1 s Severely VWF deficient ‘severe VWD type 1’ 2012-1 s Totally VWF deficient ‘Type 3 VWD’ 2007-3, 2009-3, 2010-3 HMW deficient ‘2A VWD’ 2008-2A HMW deficient ‘2A VWD’ 2013-2A HMW deficient 2B VWD 2007-2B HMW deficient 2B VWD 2009-2B, 2011-2B Platelet binding defect VWD 2M VWD 2012-2M AVWS (2A-like) AVWS (2A like) 2007-AVWS, 2008-AVWS AVWS (1/2A-like) AVWS (1/2A-like) 2010-AVWS, 2011-AVWS, 2012-AVWS Abbreviations: Ag, Antigen; AVWS, acquired vonWillebrand Syndrome;HMW,highmolecularw CB, collagen binding. ⁎ samples listed in the same cell indicate multiple despatches of same plasma sample in diff(viii) Five samples from two separate cases of acquired vonWillebrand syndrome (AVWS); both showed substantial loss of VWF and some selective loss of HMW VWF. In-house prepared VWF deficient and HMW deficient samples represents material with similar test matrix to true patient VWD samples as evidenced by similar reactivity profiles using standard VWF tests (Supplementary Fig. 1), and multimer analysis (Fig. 1). Normal pool samples were primarily derived from fresh frozen blood bank plasma from the Australian Red Cross Blood Bank. Samples severely to moderately deficient in VWF were prepared as mixtures ofSample Source Details Prepared in house target ~100U/dL VWF Prepared in house target ~100U/dL VWF Patient sample Mother of Type 3 VWD daughter; obligate carrier type 3 (heterozygous); VWF:Ag ~60U/dL; VWF activity/Ag N0.7 Patient sample VWF:Ag ~40U/dL; no mutation in exon 28; mild personal history of bleeding/bruising Prepared in house target ~30U/dL VWF; VWF activity/Ag N0.7 Prepared in house target ~30U/dL VWF; VWF activity/Ag N0.7 Prepared in house target ~25U/dL VWF; VWF activity/Ag N0.7 Prepared in house target ~10U/dL VWF; VWF activity/Ag N0.7 Prepared in house target ~10U/dL VWF; VWF activity/Ag N0.7 Commercial VWF actual 0U/dL VWF deficient Plasma Prepared in house Type 2A Mimic (Cryosupernate); target ~20U/dL VWF:Ag; with RCo/Ag and CB/Ag b 0.5 Prepared in house Type 2A Mimic; target ~30U/dL VWF:Ag; with RCo/Ag and CB/Ag b 0.5 Patient sample Well characterised 2BVWD;VWFmutation (R1306W); VWF:Ag ~30U/dL; RCo/Ag and CB/Ag b 0.3 Patient sample Well characterised 2B VWD; VWFmutation (R1306W) VWF:Ag ~40U/dL; RCo/Ag and CB/Ag b 0.3 Patient sample Well characterised 2MVWD;VWFmutation (H1322P); VWF:Ag ~50U/dL; RCo/Ag b 0.3 and CB/Ag N 0.7 Patient sample Well characterisedAVWS case; noVWFmutation in Exon 28; no evidence of inhibitory antibodies. Patient sample Well characterised AVWS case secondary to multiple myeloma; increased clearance post VWF concentrate infusion; no evidence of inhibitory antibodies. eight; VWD, vonWillebranddisease; VWF, vonwillebrand factor; RCo, ristocetin cofactor; erent years. 250 251 396 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 Fig. 1. Representative VWFmultimer gels (A) and densitometry (B toG) results for representative samples investigated in current report, including normal, VWF deficient, highmolecular weight (HMW) VWF deficient, and acquired vonWillebrand syndrome (AVWS). For densitometry figures, relative density shown on y-axis, and relative front on x-axis, showing approx- imate regions of HMW, intermediate molecular weight (IMW) and low molecular weight (LMW) VWF, as well as the dye front. Normal samples are shown as reference samples; quan- titative VWF deficient samples and the type 2M sample retain all molecular weight forms of VWF, but at a reduced level compared to normal plasma; in contrast, the HMWVWF deficient samples each show a reduction in HMW forms.normal pool plasma with VWF deficient plasma to generate different samples with VWF ranging between 10 and 40 U/dL (Table 3). All samples sent to participants were lyophilized plasma, typically 0.5 or 1.0 ml/vial. All samples passed stability and homogeneity testing to ensure sample integrity (data not shown). Strategy of Sample Dispatches and Aims of the EQA Program The overall aim of the program is to provide laboratories with awide selection of test material that reflects a similar heterogeneity to that seen in VWD diagnostic practice. Samples (2x twice a year) are dispatched in semi-random rotation, with different samples tested in each year, and repeat samples tested in separate years (permits assessment of repeatability and provides additional evidence of sample stability). Although FVIII testing is also performed, this report focuses on VWF assays, given the large amount of data generated, and because of the type of VWF samples dispatched over the period of analysis (viz, no type 2N VWD samples). Data Analysis All laboratories report numerical values to the program for the tests performed, and these are analyzed using descriptive statistics (median, ranges, intra-method or inter-laboratory coefficient of variation [CV%], etc). Laboratories are also encouraged to interpret overall test findings for the likelihood or otherwise of ‘VWD’, as well as potential ‘VWD type’ if applicable. It is well recognized that diagnosis of VWD requires more than laboratory testing, and some samples used were not derivedfrom VWD patients. Nevertheless, such interpretation exercises are valid within the context of EQA, and in-house prepared VWF-deficient and HMW-VWF-deficient samples cannot be otherwise distinguished from true VWD patient samples using the tests performed by these laboratories. Laboratory interpretations were assessed in the context of the provided sample, for both the variety of different interpretations as well as for ‘errors’ in interpretation. An error was defined to occur when the laboratory assigned (i) an ‘incorrect VWD type’ to the sample (e.g., qualitative defect assigned to a quantitative VWF deficiency, or visa versa), (ii) a VWD interpretation assigned to a clearly normal plasma sample, or (iii) a normal interpretation assigned to a moderate/ severe deficient VWF sample or qualitative defect. Results Participant Numbers and Test Methodologies Numbers of program participants have varied only slightly during the past seven years (n = 52-59; Table 2). All participants (i.e., 100%) have always performed FVIII:C testing, and N95% have always also per- formed VWF:Ag testing. VWF activity-based testing is performed by a large proportion of (but not all) laboratories, with themakeup changing appreciably over time (Fig. 2). Almost all laboratories (N80%) performed VWF:RCo testing in the past, but this has declined to just over 50% cur- rently. VWF:CB testing has been performed by ~50% of laboratories throughout the past 15 year period, with this increasing slightly to ~60% currently. Other VWF activity (‘VWF:Act’) assays are performed by a smaller proportion (initially b10%, but currently nearly a third) of251 252 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 397 Fig. 2. A. Proportion of program participants performing various VWF activity assays over the past 15 years. VWF:RCo, ristocetin cofactor; VWF:CB, collagen binding; VWF:Act, other VWF activity assays. Note: N95% of laboratories have consistently performed VWF:Ag testing during this period, and all perform FVIII:C testing. B.Methodological detail for VWF:Ag and VWF activity assays as per A, but shown as percentage of methods in approximate 4 year steps.laboratories. Thus, the use of VWF:CB has been fairly stable over all years, but VWF:RCo usage has fallen, in many cases replaced by other VWF:Act assays. VWF:Ag assays used to be primarily performed by electro-immuno diffusion (EID) and ELISA, but EID has completely disappeared, and ELISA methods have recently reduced in favor of LIA. VWF:RCo used to be performed exclusively by platelet aggregometry (‘agg’), but this is reducing in favor of automated (‘auto’) methodologies. VWF:CB is performed exclusively by ELISA, with similar numbers of in-house and commercial (‘com’) methods originally, although the latter now domi- nates. VWF:Act assays originated as monoclonal antibody ELISA-based methods, with these giving way to the Werfen-IL LIA monoclonal anti- body based method, and only recently in 2013 has the Siemens Innovance rGPIb-binding LIA method emerged (Table 1; Fig. 2). Inter-laboratory and Intra-method Variability Considerable inter-laboratory and intra-method variability was observed with all samples (Fig. 3), and CVs generally increase as VWF falls (Fig. 4A). Moreover, CVs are lowest for VWF:Ag, highest for VWF: RCo, and intermediate for VWF:CB and VWF:Act (Fig. 4B). Occasional identified outlier data (Fig. 3) reflect random error events, transcription errors, and occasional systematic error. Irrespective, these reflect real events that identify errors in laboratory reporting of numerical data that then impact on test result interpretation. Occasionally, laboratories reported outliers for several tests perhaps reflecting problems with a common assay calibrant standard. Lower Limit of VWF Sensitivity This was explored by assessment of samples totally (=0 U/dL VWF; equivalent to type 3 VWD) or severely deficient in VWF (~10 U/dL VWF) (Fig. 3). Even when testing samples that do not contain anyVWF, most laboratories in fact report values above 0 U/dL for most VWF assays. This is comparatively worse for VWF:RCo than other assays, approaching 20U/dL in some laboratories; however, limits are close to 5-10U/dL for many laboratories for all VWF assays. Thus, many cannot distinguish total from severe VWF deficiencies. Sensitivity limits will also lead to problems in type 2 VWD identification. Utility and Variability of VWF Activity/Ag Ratios This was assessed by comparison of assay ratios derived using normal plasma and mild/moderate VWF deficient samples, vs those with qualitative defects or HMW VWF deficient (Fig. 5). In normal and VWF deficient (equivalent to low VWF or type 1 VWD) samples, there is expected concordance of VWF activity and VWF:Ag. In practice, dif- ferent laboratories assign different cut-off values that in general range from 0.5-0.7. Conversely, samples that express qualitative defects and/ or are specifically deficient in HMW VWF (i.e., type 2A, 2B and 2M VWD) would be expected to yield VWF activity/Ag ratios below these cut-off values. As identified in Fig. 5, best discrimination between normal/VWF deficient vs qualitative/HMW VWF deficient samples was obtained using VWF:CB/Ag ratios. In contrast, considerable overlap of data was observed with VWF:RCo/Ag ratios, and an intermediate over- lap was observed with VWF:Act/Ag ratios. For the type 2M VWD case, functional discordance was observed using VWF:RCo/Ag but not with VWF:CB/Ag; VWF:Act/Ag ratios were again intermediate. Variability in Test Result Interpretation Considerable inter-laboratory variability in test result interpretation was also observed, which on occasion reflected subtle differences regarding the perception of ‘severity’ in VWF deficiency (mild, moder- ate, severe vs total), and on other occasions ‘diagnostic errors’ related to incorrect VWD type assignment. For example, for the five mildly/252 253 398 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 253 254 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 399 Fig. 4. A and B: Inter-laboratory variability for all performed tests expressed as coefficient of variation (CV%). FVIII:C data also shown here as comparator. A. CV% (y-axis) vs median reported assay values (x-axis). B. range of CV values (y-axis) expressed as 25th/median/75th percentiles for all test data (left portion of figure) and for select test data (median VWF levels N 15U/dL; right portion of figure). 15U/dL chosen since CVs for values below 15U/dL begin to increase exponentially (refer to Figure A). Median CVs for VWF:RCo are highest, and for VWF:Ag are lowest. Those for VWF:CB and VWF:Act are intermediate. Note the larger spread of CV data for VWF:Act where VWF N 15U/dL. Some of this variability is driven by data from acquired von Willebrand syndrome (AVWS) as partly shown in Fig. 3J. C: Variability in laboratory provided interpretations for severe VWF deficient (~10U/dL VWF) samples, totally VWFdeficient samples (VWF= 0U/dL) and for AVWS samples. Data shown as percentage of responses (y-axis); samples identified on x-axis. MiT1,mild VWF deficiency/mild type 1 VWD; MoT1, moderate VWF deficiency/moderate type 1 VWD; ST1, severe VWF deficiency/severe type 1 VWD; T3, total VWF deficiency/type 3 VWD; 2A/B/M, high molecular weight (HMW) VWF deficient/‘qualitative VWF defect’/type 2A/2B/2M VWD.moderately deficient samples (VWF25-40U/mL), laboratories provided a variable mixture of interpretations of primarily moderate and mild type 1 VWD (data not shown).More interesting perhaps,were the com- parative interpretations for severely deficient vs totally deficient VWF samples, which challenged many laboratories. While most identified these appropriately, several identified the former as ‘type 3 VWD’ and the latter as ‘severely’ VWF deficient (Fig. 4C). This partially relates to the lower limit of VWF sensitivity issue noted previously (Fig. 3). Also interesting was the wide range of interpretations provided for the AVWS samples. As these showed a specific loss of HMW VWF inFig. 3.Numerical data returned by participants for representative samples fromeach sample cate two of seven samples with mild/moderate VWF deficiency (C [2010-1 m] and D [2011a-1 m]) AVWS]), two of three samples of severe VWF deficiency (F [2012-1 s] and G [2013-1 s]), two of with highmolecularweight [HMW]VWFdeficiency (K [2008-2A], L [2013-2A]),M [2007-2B] an (antigen), VWF:RCo (ristocetin cofactor), VWF:CB (collagen binding), VWF:Act (other VWF act representative outlier values. For example, A 2008-N: a-d= four outlier points for each VWF a given as normal despite VWF:RCo/Ag=0.4; c. interpretation given as normal; d. Interpretation g for VWF:RCo, but also onoccasion for other assays, aswell as low outlier points. Thesewere ofte discordance in normal and quantitative VWF deficient samples, and false concordance in HMW sample shown in Figure J; apart from one or two participant points, these values reflect data f AVWS samples was also evident.addition to severe loss of VWF and VWF activity (Figs. 1 & 3), interpre- tations of type 2A/2B/2MVWDwere expected; however,most laborato- ries interpreted these as quantitative deficiencies of VWF ranging from severe (most laboratories) to moderate to even mild (Fig. 4C).Reproducibility of Test Results Using Repeat Test Samples As identified in Table 3, several identical samples were dispatched in different years. In general, returned numerical data showed similargory shown in Table 3; including two of four normal samples (A [2008-N] andB [2010-N]), , two of five samples of acquired von Willebrand Syndrome (E [2007-AVWS] and J [2011- three samples with total VWF deficiency (H [2007-3] and I [2010-3]), four of five samples dN [2011-2B]), and the type 2MVWD sample (O [2012-2M]), for all VWF assays (VWF:Ag ivity assays)). VWF level shown on y-axis; assays shown on x-axis. Circled values indicate ssay as reported by four different labs; a. interpretation given as normal; b. interpretation iven as 2A/2BVWDwith VWF:Act/Ag=0.2.Allfigures:notehighoutlier points, especially n associatedwith VWF/VWD identification errors, and yielded false VWF activity to antigen VWF deficient samples. Note also the high scatter of points for VWF:Act for the AVWS rom the Werfen-IL VWF activity assay. Similar high variability for this assay for the other 254 255 400 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 Fig. 5.VWFActivity/Ag ratios.A&B.Normal plasma (A) andmild/moderate VWFdeficiency (B) samples using a representative subset of four plasma samples in each category. C&D. high molecular weight (HMW) VWF deficient samples (‘2A VWD like’ [x2] and 2B VWD [x3] respectively); E. The 2M VWD case as shown in Fig. 3. Circled values in each case reflect false (dis- cordant) low ratios (b0.7) for normal and VWF deficient samples (A andB) and false (concordant) high ratios (N0.7) for HMWdeficient samples, or areas of overlap, and therefore causing difficulties and errors in quantitative vs qualitative VWD type assignment. Apart fromnoted (circled) outliers, the type 2Mcase (E) showedVWF:RCo/Ag ratios b0.5 and VWF:CB/Ag ratios N0.5; the VWF:Act/Ag ratios were intermediate between those for VWF:RCo/Ag and VWF:CB/Ag.median and ranges of VWF test data (examples shown in Fig. 3), and similar but not identical overall interpretations (examples in Fig. 4C). Errors in Result Interpretation Please also refer to highlighted outlier data points in Fig. 3. Normal Samples For the four normal samples (Table 3), acceptable interpretations were reported in all but four cases; one laboratory reported moderate type 1 VWD based on a normal VWF:Ag (86U/dL) but low VWF:RCo (39U/dL) for a VWF:RCo/Ag ratio of 0.5; three laboratories reported normal samples as ‘type 2A/2B VWD’ – two of these based on false discordancewith VWF:RCo/Ag ratios of 0.4 and 0.5 and one on false dis- cordance with VWF:Act/Ag (0.2). Mild/moderate/severe Deficient VWF Samples In total, there were 30 misinterpretations of qualitative defects (i.e., type 2A/2B/2M VWD) and false VWF Activity/Ag discordance reported, with errors reported for all seven samples (Table 4). In 5/30 (16.7%) cases these reflected laboratory misinterpretation of their own test data since VWF assay concordance (ratios N 0.7) was actually reflected in returned test data. In the majority of other cases (n = 24), false discordance was reported (i.e., VWF Activity/Ag ratios b 0.7). False low VWF:RCo/Ag ratios were reported in 19/24 (79.2%) of cases, false low VWF:CB/Ag ratios in 8/24 (33.3%), and false low VWF:Act/Ag ratios in 2/24 (8.3%). Adjusted error rates (taking into account numbers of laboratories performing testingwith each type of activity assay)were 9.5%, 4.6% and 2.3% respectively for VWF:RCo, VWF:CB and VWF:Act.Severe Deficient VWF Samples Misinterpretations of qualitative defects (i.e., type 2A/2B/2M VWD) were reported with each of three samples for a total 14 occasions (Table 4). In 5/14 (35.7%) cases these reflected laboratory misinterpre- tation of test data since assay concordance was actually obtained. In the majority of other cases, ‘false discordance’ in various activity/Ag ratios was ‘identified’ by laboratories. Again, most errors related to use of VWF:RCo and VWF:RCo/Ag ratios. HMW VWF Deficient Samples Misinterpretations of type 1 VWD were reported with each the five HMW VWF deficient samples for a total of 26 occasions (Table 4). In most cases (13/26 = 50%) this reflected a misinterpretation of laboratory test data, which was instead consistent with the test sample (i.e., evidencing VWF activity/Ag ratios b 0.7). In many other occasions (5/26 = 19.2%) a limited test panel (i.e., lacking VWF:Ag or VWF activ- ity assays) was identified. On other occasions, false VWF activity/Ag concordance was reported; this occurred more frequently for VWF: RCo/Ag ratios than VWF:CB/Ag or VWF:Act/Ag. Misidentification of Type 2M VWD One sample dispatched in this study was a type 2M VWD sample with a novel VWF mutation (Table 3). Consistent with expectations of a GPIb binding defect, multimer analysis showed retention of HMW VWF (Fig. 1), and this sample expressed a relative reduction in VWF:RCo/Ag but essentially normal VWF:CB/Ag ratio (Fig. 5). Even in the absence of multimer analysis, this pattern is highly suggestive of a platelet GPIb binding defect type 2M VWD. However, only 21/48 (43.8%) laboratories identified type 2M VWD and a higher proportion255 256 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 401 Table 4 Interpretation Error Rates for select study samples. Sample type Sample sets Total Total Error rate Lab Insuf False VWF activity / Ag concordance or discordancef Adjusted error ratesg errorsa interpsb (%)c misinterpd panele All VWF:RCo VWF:CB VWF:Act VWF:RCo events VWF:CB events VWF:Act events (n/N; %) (n/N; %) (n/N; %) (n/N; %) (n/N; %) (n/N; %) (n/N; %) Mild/ moderate 2007-1 m, 2009-1 m, 30 334 9.0 5/30 (16.7%) 1/30 (3.3%) 24/30 (80.0%) 19/24 (79.2%) 8/24 (33.3%) 2/24 (8.3%) 19/200 (9.5%) 8/175 (4.6%) 2/86 (2.3%) VWF deficient 2010-1 m, 2011a-1 m, 2011b-1 m, 2012-1 m, 2013-1 m Severe VWF deficient 2009-1 s, 2012-1 s, 2013-1 s 14 146 9.6 5/14 (35.7%) 1/14 (7.1%) 8/14 (57.1%) 5/8 (62.5%) 2/8 (25.0%) 1/8 (12.5%) 5/97 (5.2%) 2/96 (2.1%) 1/46 (2.2%) Qualitative defect/ 2008-2A, 2013-2A, 2007-2B, 26 233 11.2 13/26 (50.0%) 5/26 (19.2%) 8/26 (33.3%) 6/8 (75.0%) 1/8 (12.5%) 2/8 (25.0%) 6/136 (4.4%) 1/108 (0.9%) 2/55 (3.6%) HMW VWF deficient 2009-2B, 2011-2B All samples above All samples above 70 713 9.8 23/70 (32.9%) 7/70 (10.%) 40/70 (57.1%) 30/40 (75.0%) 11/40 (27.5%) 5/40 (12.5%) 30/433 (6.9%) 11/379 (2.9%) 5/187 (2.7%) Notes: a. Errors defined as perMaterials andMethods; essentially, when a laboratory identified a qualitative defect/specific loss of HMWVWF (or types 2A/2B/2M VWD) in quantitative VWF deficient samples, or failed to identify qualitative defect/specific loss of HMW VWF in HMW deficient samples (e.g. false identification of type 1 or 3 VWD). b. Total interps = total number of interpretations provided by participants for the combined sample set for that sample group. c. Error rate = total errors/total interps as percentage. d. Lab misinterp = misinterpretation of laboratory's own data set, which was actually consistent with the sample type presented. e. Insuf panel - laboratory had an insufficient test panel to make an appropriate interpretation (generally, these labs ran (in addition to FVIII:C) only VWF:Ag or VWF:RCo. f. False VWF activity / Ag concordance or discordance = false low ratios (i.e. b0.7) in mild/mod VWF deficient samples or false high ratios (i.e. N0.7) in HMW VWF deficient samples. g. Adjusted error rates = False VWF activity / Ag concordance or discordance / total number of possible events for each test group (i.e., adjusted for numbers of laboratories reporting test results for these tests for each sample set included in data set). Abbreviations: HMW, high molecular weight; VWD, von Willebrand disease; VWF, von willebrand factor; VWF:RCo, ristocetin cofactor assay for VWF; VWF:CB, collagen binding assay for VWF; VWF:Act, other 'activity' assays for VWF.(23/48= 47.9%) identified type 2A or 2B VWD. Three participants iden- tified mild type 1 VWD, and one identified the sample as normal. Discussion In this study, a large number of test samples (n = 29) of diverse nature have been cross-laboratory tested by ‘real world’ laboratories. Despite the large assay variability and occasional outlier results (Fig. 3) most participants reassuringly provided both numerical and in- terpretation results consistent with the nature of the sample tested. However, some sample types challenged some laboratories, in particu- lar the identification of HMWVWFdeficient plasma samples. In our lab- oratory cohort, only 2/57 (3.5%) laboratories performmultimer analysis, and therefore loss of HMWVWF in these exercises, and likely also in di- agnostic practice, is identified using surrogate laboratory markers, most notably the ratio of VWF activity/Ag. Thus, for loss of HMW VWF, all VWF activity assays as normally performed by laboratories (VWF:RCo, VWF:CB, VWF:Act) would be lower than VWF:Ag and yield VWF Activity/Ag ratios below 0.5 to 0.7. Consistent with previous reports from us [15–21,28–30], the VWF:RCo/Ag ratio was in general less able to discriminate HMW VWF deficient plasma samples (i.e., types 2A and 2B VWD) from normal or quantitative VWF deficient (i.e., type 1 VWD) plasma samples compared to the VWF:CB/Ag ratio. False VWF discordance using quantitative VWF deficient samples, and false VWF concordance usingHMWVWFdeficient samples, waspreviously associ- ated with use of the VWF:RCo/Ag ratio at rates at least three times that of VWF:CB/Ag [19,20], and this continues currently. The association of higher relative error rates using VWF:RCo compared with VWF:CB was also most recently reported by NASCOLA [25], where overall diag- nostic interpretation error rates ranged from 3% for normal samples, to 28% for type 1 VWD, to a staggering 60% for type 2 VWD. Notably, the type 2 VWD samples were identified correctly by all laboratories using VWF:CB/Ag ratios, but by only one-third using VWF:RCo/Ag or VWF:Act/Ag ratios. Much of the error rate associated with use of VWF: RCo seemed to be related to the higher assay variability of VWF:RCo testing compared to other VWF activity assays. Although some of the reported errors may be circumvented should laboratories undertake multimer analysis,multimer analysis per semaynot protect laboratories frommaking errors. For example, in the NASCOLA study, 27% of labora- tories performed multimer analysis, but VWD diagnostic and VWD typing error rateswere higher overall than those reported in the current study. Reasons for this are not entirely clear, but may reflect poor performance in multimer assays or misinterpretation of overall test patterns even when multimers are performed. Other recent comparative data using various VWF activity assays for identification of VWD is also available [7–10,31], but not in a cross- laboratory setting. Although these previous studies all individually ‘validate’ various VWF activity assays for assisting the diagnosis of VWD, comparative evaluation in the real world setting is best achieved by cross-laboratory studies. In another recent exercise using samples selectively and sequentially deficient in HMW VWF, the VWF:CB and VWF:RCo were better able to discriminate loss of HMW VWF than the Werfen-IL activity assay [27]. Unfortunately, the higher discriminating ability for VWF:RCo at the ‘median’ level is sometimes lost due to the high assay variability between laboratories, and therefore occasional failures occur. Cross laboratory data using the newest VWF activity assay, the Siemens Innovance assay [8], is currently lacking, as this assay is only just making inroads into diagnostic practice. An important aspect of our report is the data using the type 2M VWD sample, which is a difficult form of VWD to identify. Most lab- oratories in most geographies misidentify type 2M VWD as either 2A or type 1 VWD [1,30–36]. Type 2M VWD is consequently a signifi- cantly under-diagnosed disorder. Identification requires evidence of VWF dysfunction not associated with loss of HMWVWF. However, most VWD testing laboratories do not perform multimer analysis [37], leaving the diagnosis to be achieved using surrogate markers.256 257 402 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403Even when multimer analysis is performed, false concordance of VWF:RCo/Ag ratios may still lead to misidentification of 2M VWD as type 1 VWD (even in expert laboratories in upwards of 20% of type 1 VWD cases [35]). Most forms of type 2M VWD are platelet binding defect cases identified by relative loss of VWF:RCo (i.e., VWF:RCo/Ag ratio b0.7), but without a major effect on collagen binding (i.e., VWF:CB/Ag ratio N0.7). Despite this pattern being ob- served with our case (Figs. 3 and 5) most participants still failed to identify this as type 2M VWD. In most cases, this was simply a failure to recognize this pattern, and perhaps reflects a lack of general rec- ognition of type 2M VWD. Thus, most laboratories instead identified the sample as type 2A VWD, of which they are more aware. Low level sensitivity for VWFdetection as reported here is consistent with an earlier report [26]. Although this largely reflects a difficulty in differentially identifying type 3 vs severe type 1 VWD, there are also ad- verse effects on identification of type 2A/2B/2M VWD. For example, when tested in some laboratories, a type 2A, 2B or 2M VWD sample could yield ~20U/dL for both VWF:Ag and VWF:RCo, and therefore be identified as a (moderate) type 1 VWD. Conclusion We report on assay and laboratory variability in numerical and inter- pretative reporting for VWF and VWD using a large number of samples of varying type. The data reflects on the challenges in laboratory identi- fication of VWD and VWD type in real world laboratories, and this will subsequently reflect on the clinical diagnosis of VWD. Inmost cases, in- terpretive errors (‘misdiagnosis’ of VWD or VWD type) are due to performance of insufficient test panels or to misinterpretation of the laboratory’s own test data, which may instead be consistent with the sample tested. In regards to a sufficient ‘test panel’, laboratories should employ a comprehensive panel that comprises FVIII, VWF:Ag, and pref- erably two independent activity assays. For the latter, the additional combination of VWF:RCo andVWF:CBwill in general permit differential identification of types 2A/2B vs 2M vs 1. Performance of VWFmultimers will also assist identification of type 2 VWD. Someworkers believe that multimers may be replaced in a screening process by the combination of VWF:Ag, VWF:RCo and VWF:CB [38,39]. This is important as in fact, few laboratories worldwide perform multimer analysis [37], and only a small proportion of laboratories can generate good multimers results. In addition, performance of multimers per se will not protect laborato- ries from diagnostic errors [25,35]. It is hoped that this report highlights the challenges to accurate diagnosis of VWD, and that laboratories and clinicians adopt comprehensive test panels and best test practices, as well as repeat test all patients investigated using a fresh sample to con- firm original test findings. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.thromres.2014.05.020. Conflict of Interest Statement The authors have no conflicts of interest in relation to this work. Acknowledgements The authors would like to thank participants of the RCPAQAP Haematology. We also thank Jane McDonald and Ella Grezchnik from theWestmead ICPMR laboratory for technical assistance related to sam- ple testing including homogeneity and stability testing. References [1] Favaloro EJ. Von Willebrand disease: local diagnosis and management of a globally distributed bleeding disorder. Semin Thromb Hemost 2011;37:440–55. [2] Flood VH. Perils, problems, and progress in laboratory diagnosis of von Willebrand disease. Semin Thromb Hemost 2014;40:41–8.[3] Yee A, Kretz CA. von Willebrand Factor: Form for Function. Semin Thromb Hemost 2014;40:17–27. [4] Favaloro EJ. Rethinking the diagnosis of von Willebrand disease. Thromb Res 2011;127(Suppl. 2):s17–21. [5] Sadler JE, Budde U, Eikenboom JC, Favaloro EJ, Hill FG, Holmberg L, et al. Working Party on von Willebrand Disease Classification. Update on the pathophysiology and classification of von Willebrand disease: a report of the Subcommittee on von Willebrand Factor. J Thromb Haemost 2006;4:2103–14. [6] Favaloro EJ, Franchini M, Lippi G. Biological therapies for von Willebrand Disease. Expert Opin Biol Ther 2012;12:551–64. [7] Chen D, Tange JI,Meyers BJ, Pruthi RK, NicholsWL, Heit JA. Validation of an automated latex particle-enhanced immunoturbidimetric vonWillebrand factor activity assay. J Thromb Haemost 2011;9:1993–2002. [8] Lawrie AS, Stufano F, Canciani MT, Mackie IJ, Machin SJ, Peyvandi F. A comparative evaluation of a new automated assay for vonWillebrand factor activity. Haemophilia 2013;19:338–42. [9] Cabrera N, Moret A, Caunedo P, Cid AR, Vila V, Espana F, et al. Comparison of a new chemiluminescent immunoassay for von Willebrand factor activity with the ristocetin cofactor induced platelet agglutination method. Haemophilia 2013;19: 920–5. [10] Stufano F, Lawrie AS, La Marca S, Berbenni C, Baronciani L, Peyvandi F. A two-centre comparative evaluation of new automated assays for von Willebrand factor ristocetin cofactor activity and antigen. Haemophilia 2014;20:147–53. [11] Mina A, Favaloro EJ, Koutts J. A novel flow cytometry single tube bead assay for quantitation of von Willebrand factor antigen and collagen-binding. Thromb Haemost 2012;108:999–1005. [12] Chen D, Daigh CA, Hendricksen JI, Pruthi RK, Nichols WL, Heit JA, et al. A highly- sensitive plasma von Willebrand factor ristocetin cofactor (VWF:RCo) activity assay by flow cytometry. J Thromb Haemost 2008;6:323–30. [13] Budde U, Pieconka A, Will K, Schneppenheim R. Laboratory testing for von Willebrand disease: contribution ofmultimer analysis to diagnosis and classification. Semin Thromb Hemost 2006;32:514–21. [14] Nichols WL, Hultin MB, James AH, Manco-Johnson MJ, Montgomery RR, Ortel TL, et al.Willebrand disease (VWD): evidence-based diagnosis andmanagement guide- lines, the National Heart, Lung, and Blood Institute (NHLBI) Expert Panel report (USA). Haemophilia 2008;14:171–232. [15] Favaloro EJ, Smith J, Petinos P, Hertzberg M, Koutts J, on behalf of the RCPA Quality Assurance Program (QAP) in Hematology Scientific Hemostasis Advisory Panel. Laboratory testing for vonWillebrand's disease: An assessment of current diagnostic practice and efficacy by means of a multi-laboratory survey. Thromb Haemost 1999;82:1276–82. [16] Favaloro EJ, Thom J, Baker R, on behalf of the Australasian Society for Thrombosis and Haemostasis (ASTH) Emerging Technologies Group. Assessment of current diagnostic practice and efficacy in testing for von Willebrand's disorder: Results from the second Australasian multi-laboratory survey. Blood Coagul Fibrinolysis 2000;11:729–38. [17] Favaloro EJ, Bonar R, Sioufi J, HertzbergM, Street A, Lloyd J, et al. Laboratory diagnosis of vonWillebrand Disorder: Current practice in the Southern Hemisphere. Am J Clin Pathol 2003;119:882–93. [18] Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Hertzberg M, Street A, et al. Laboratory diagnosis of von Willebrand Disorder: Quality and diagnostic improvements driven by peer review in a multi-laboratory test process. Haemophilia 2004;10:232–42. [19] Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Thom J, Baker R, et al. Laboratory diagnosis of von Willebrand Disorder: Use of multiple functional assays reduces diagnostic error rates. Lab Hematol 2005;11:91–7. [20] Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Baker R, HertzbergM, et al. Reducing errors in identification of von Willebrand disease: The experience of the Royal college of Pathologists of Australasia Quality Assurance Program. Semin Thromb Hemost 2006;32:505–13. [21] Favaloro EJ, Bonar R, Meiring M, Street A, Marsden K, on behalf of the RCPA QAP in Haematology. 2B or not 2B? Disparate discrimination of functional VWF discordance using different assay panels or methodologies may lead to success or failure in the early identification of Type 2B VWD. Thromb Haemost 2007;98:346–58. [22] Kitchen S, Jennings I, Woods TA, Kitchen DP, Walker ID, Preston FE. Laboratory tests for measurement of von Willebrand factor show poor agreement among different centers: results from the United Kingdom National External Quality Assessment Scheme for Blood Coagulation. Semin Thromb Hemost 2006;32:492–8. [23] Hayes TE, Brandt JT, Chandler WL, Eby CS, Kottke-Marchant K, Krishnan J, et al. External peer review quality assurance testing in vonWillebrand disease: the recent experience of the United States College of American Pathologists proficiency testing program. Semin Thromb Hemost 2006;32:499–504. [24] Meijer P, Haverkate F. An external quality assessment program for von Willebrand factor laboratory analysis: an overview from the European concerted action on thrombosis and disabilities foundation. Semin Thromb Hemost 2006;32:485–91. [25] Chandler WL, Peerschke EIB, Castellone DD, Meijer P, on behalf of the NASCOLA Proficiency Testing Committee. von Willebrand factor assay proficiency testing. The North American specialized coagulation laboratory association experience. Am J Clin Pathol 2011;135:862–9. [26] Favaloro EJ, Bonar R, Marsden K, on behalf of the RCPA QAP Haemostasis Committee. Lower limit of assay sensitivity: an under-recognised and significant problem in von Willebrand disease identification and classification. Clin Lab Sci 2008;21: 178–85. [27] Favaloro EJ, Bonar R, Chapman K, Meiring M, Adcock DF. Differential sensitivity of von Willebrand factor 'activity' assays to large and small VWF molecular weight forms: a cross-laboratory study comparing ristocetin cofactor, collagen binding and monoclonal antibody based assays. J Thromb Haemost 2012;10:1043–54.257 258 E.J. Favaloro et al. / Thrombosis Research 134 (2014) 393–403 403[28] Favaloro EJ. Evaluation of commercial vonWillebrand factor collagen binding assays to assist the discrimination of types 1 and 2 von Willebrand disease. Thromb Haemost 2010;104:1009–21. [29] Favaloro EJ, Grispo L, Exner T, Koutts J. Development of a simple collagen binding assay aids in the diagnosis of, and permits sensitive discrimination between, Type I and Type II von Willebrand's disease. Blood Coagul Fibrinolysis 1991;2:285–91. [30] Favaloro EJ, Koutts J. Laboratory assays for vonWillebrand Factor: Relative contribu- tion to the diagnosis of von Willebrand's disease. Pathology 1997;29:385–91. [31] Rodgers SE, Lloyd JV, Mangos HM, Duncan EM, McRae SJ. Diagnosis and manage- ment of adult patients with von Willebrand disease in South Australia. Semin Thromb Hemost 2011;37:535–41. [32] Favaloro EJ, Bonar R, Favaloro J, Koutts J. Diagnosis and management of von Willebrand disease in Australia. Semin Thromb Hemost 2011;37:542–54. [33] Favaloro EJ. Diagnosis of type 1 versus 2A and 2M von Willebrand disease. Haemophilia 2012;18:e9-11. [34] Favaloro EJ, Forsyth C, Koutts J. Distinguishing types 1 and 2M von Willebrand disease. Int J Lab Hematol 2012;34:102–5.[35] Favaloro EJ. Detailed von Willebrand factor multimer analysis in patients with von Willebrand disease in the European study, molecular and clinical markers for the diagnosis and management of type 1 von Willebrand disease (MCMDM-1VWD) – a rebuttal. J Thromb Haemost 2008;6:1999–2001. [36] Federici AB, Bucciarelli P, Castaman G, Baronciani L, Canciani MT, Mazzucconi MG, et al. Management of inherited von Willebrand disease in Italy: results from the retrospective study on 1234 patients. Semin Thromb Hemost 2011;37: 511–21. [37] Favaloro EJ, Plebani M, Lippi G. Regulation in hemostasis and thrombosis: Part I-in vitro diagnostics. Semin Thromb Hemost Apr 2013;39(3):235–49. [38] Adcock DM, Bethel M, Valcour A. Diagnosing von Willebrand disease: a large reference laboratory's perspective. Semin Thromb Hemost Jul 2006;32(5): 472–9. [39] Flood VH, Gill JC, Friedman KD, Christopherson PA, Jacobi PM, Hoffmann RG, et al. Collagen binding provides a sensitive screen for variant von Willebrand disease. Clin Chem Apr 2013;59(4):684–91.258 S-29 259 Haemophilia (2016), 22, e145–e155 DOI: 10.1111/hae.12903 ORIGINAL ARTICLE Von Willebrand disease Type 2M von Willebrand disease – more often misidentified than correctly identified E. J. FAVALORO,* R. A. BONAR,† S. MOHAMMED,* A. ARBELAEZ,‡ J . NIEMANN,‡ R. FRENEY,‡ M. MEIRING,§ J. SIOUFI† and K. MARSDEN† *Department of Haematology, Sydney Centres for Thrombosis and Haemostasis, Institute of Clinical Pathology and Medical Research (ICPMR), Pathology West, NSW Health Pathology, Westmead Hospital, Westmead; †RCPAQAP Haematology, St Leonards, NSW; ‡Mater Pathology, Mater Health Services, Brisbane, Qld, Australia; and §Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa Introduction: Appropriate diagnosis of von Willebrand disease (VWD), including differential identification of qualitative vs. quantitative von Willebrand factor (VWF) defects has important management implications, but remains problematic. Aim: The aim of the study was to assess whether 2M VWD, defining qualitative defects not associated with loss of high molecular weight (HMW) VWF, is often misidentified, given highly variable reported frequency ranging from 0 to ~60% of all type 2 VWD. Methods: A comparative evaluation of laboratory ability to appropriately identify 2M VWD (n = 4) vs. HMW VWF reduction (n = 4), as sent to participants of an international external quality assessment programme. Results: Laboratories had considerably greater difficulty identifying type 2M VWD, correctly identifying these on average only 29.4% of occasions, with the 70.6% error rate representing use of insufficient test panels (41.7%), misinterpretation of test results (10.0%) and analytical errors (13.3%). One type 2M case, giving a median of 49 U dL1 VWF:Ag, was more often misidentified as type 2A/2B VWD (46.7%) than 2M (34.8%). Another 2M case, giving a median of 189 U dL1 VWF:Ag, was instead often misidentified as being normal (non-VWD) (36.4%), with identifications of type 2A/2B VWD (13.6%) also represented. In comparison, errors in identification of HMW VWF reduced samples only averaged 11.5%, primarily driven by use of insufficient test panels (6.3%) or misinterpretation of results (4.2%) and infrequently analytical errors (1.0%). Conclusion: Type 2M VWD is more often misidentified (70.6%) than correctly identified as 2M VWD (29.4%), and potentially explaining the relative under-reported incidence of 2M VWD in the literature. Keywords: 2M VWD, diagnosic, errors, interpretation, laboratory testing, von Willebrand disease, von Willebrand fac- tor VWF possesses many functional properties, including Background binding to platelets via several receptors, most notable von Willebrand disease (VWD) is reportedly the most glycoprotein Ib, binding to sub-endothelial matrix common congenital bleeding disorder; nevertheless, its components (most notably collagen) and binding and differential identification and clinical diagnosis protecting factor VIII (FVIII) function [3]. The first remains problematic for many laboratories and clini- two noted functions are essential for overall primary cians [1,2]. VWD arises from deficiency and/or defects haemostasis and are, respectively, assessed in the labo- of von Willebrand factor (VWF), an adhesive plasma ratory by surrogate tests such as VWF ristocetin cofac- protein essential for effective primary haemostasis. tor (VWF:RCo) and collagen binding (VWF:CB) [1,2,4,5], although alternative assays may also be employed [6]. Correspondence: Emmanuel J. Favaloro, Department of Haema- The most recent classification scheme from the Inter- tology, Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, Westmead, NSW 2145, Aus- national Society on Thrombosis and Haemostasis tralia. (ISTH) recognizes six different types of VWD [5]. Type Tel.: (612) 9845 6618; fax: (612) 9689 2331; 1 represents a partial quantitative VWF deficiency, e-mail: emmanuel.favaloro@health.nsw.gov.au with VWF essentially functionally normal, but pro- Accepted after revision 21 December 2015 duced in lowered quantity. Type 3 VWD represents © 2016 John Wiley & Sons Ltd 25e9145 260 e146 E. J. FAVALORO et al. ‘complete’ deficiency of VWF. Type 2 VWD represents (n = 1), and generally represent ‘real-world’ laborato- a heterogeneous group of qualitative VWF defects that ries investigating VWD for clinical diagnosis and man- comprise (i) 2A VWD (loss of high molecular weight agement; however, most would not be considered (HMW) VWF), (ii) 2B VWD (enhanced functional ‘expert laboratories’ (defined here as a laboratory binding of VWF that leads to loss of HMW VWF and regarded by others as having a particular expertise in typically mild thrombocytopenia), (iii) 2N VWD (loss the field of VWF/VWD testing; generally, these labora- of VWF-FVIII binding) and (iv) 2M VWF (VWF dys- tories are associated with academic teaching hospitals function not associated with loss of HMW VWF). The or haemophilia centres). proper identification of VWD and differentiating its type is important for therapeutic management [7]. Samples Of particular relevance to the current report, type 2A VWD is often represented as being the most common Table 1 details the samples (n = 8) distributed to type 2 VWD type, and 2M alternatively represented as EQA participants for the current report. These sam- being a relatively uncommon form of VWD [1]. ples comprised a set of four type 2M VWD samples Indeed, depending on the centre reporting data, type derived from two separate well characterized patients 2M VWD has been variously reported in the literature with type 2M VWD (sent as duplicates in separate as representing between 0 and nearly 60% of all type 2 years), and a comparator set of four HMW VWF VWD cases [1]. Rather than this representing a geo- reduced samples. One of the type 2M VWD samples graphically variable incidence, we believe this simply (a c.3965A>C VWF mutation, with p.His1322Pro) represents a relative difficulty in correctly identifying presented with VWF:Ag ~50 U dL1 (plus VWF:RCo/ type 2M VWD in some centres, with this instead being Ag <0.3, but with no loss of HMW VWF), whereas often misidentified as either type 2A or type 1 VWD, the other (a c.3974C>T VWF mutation, with depending in part on the test panel being used as well p.Ser1325Phe, homozygous), presented with a VWF: as the expertise of the reporting laboratory [8]. We Ag level of ~190 U dL1 (again plus VWF:RCo/Ag have tested this hypothesis in this study by assessing <0.3, and with no loss of HMW VWF). These two the relative correctness of identification of test samples cases would therefore reflect different challenges to representing type 2M VWD (n = 4), vs. a comparator correct VWD subtype; we hypothesized that the first set of test samples showing HMW VWF reduction (2A 2M case, reflecting a ‘text-book’ 2M VWD-case, or 2B VWD cases; n = 4), and as sent to participant would be often misidentified as a type 2A-VWD (or laboratories of an international external quality assess- similar) sample. On the other hand, the latter 2M ment (EQA) programme, over the past 6 years. VWD case might instead be missed as a VWD case, given the high VWF:Ag level, and depending on the participant test panels performed, plus the resultant Methods laboratory test panel interpretations. The comparator set comprised four HMW VWF Setting reduced samples, including one well characterized type This study derives from the Royal College of Patholo- 2B VWD case (VWF mutation (c.3916C>T; gists of Australasia (RCPAQAP) Haematology EQA p.Arg1306Trp), and two samples that were purpose programme (). Partici- generated to express reduction in HMW VWF, and pants to this VWF/VWD EQA (current n = 60) derive similar to those previously reported by this EQA from Australia (n = 33), New Zealand (n = 5), Asia [9,10]. These samples are produced by a propriety (n = 13; Malaysia=9, Hong Kong=4), South Africa process of disulphide bond reduction with N-acetyl- (n = 6), India (n = 1), France (n = 1) and Spain cysteine, similar to that recently described by Chen Table 1. Summary of samples utilized for current study. Sample Sample type represents Sample ID* Sample source Details HMW 2B VWD 2009-2B, 2011-2B Patient sample Well-characterized 2B VWD; VWF mutation (c.3916C>T (p.Arg1306Trp)); deficient VWF:Ag ~40 U dL1; with loss of HMW VWF and RCo/Ag and CB/Ag <0.3 2A VWD’ 2013-2A Prepared in Type 2A mimic; target ~30 U dL1 VWF:Ag; with loss of HMW VWF and house RCo/Ag and CB/Ag <0.3 2015-2A Prepared in Type 2A mimic (‘pregnancy’); target ~80 U dL1 VWF:Ag; with loss of HMW house VWF and RCo/Ag and CB/Ag <0.3 Platelet 2M VWD 2012-2M, 2014-2M(a) Patient sample Well characterized 2M VWD; VWF mutation (c.3965A>C (p.His1322Pro)); with binding no loss of HMW VWF and VWF:Ag ~50 U dL1, RCo/Ag <0.3 and CB/Ag >0.7 defect 2014-2M(b), 2015-2M Patient sample Well characterized 2M VWD; VWF mutation (c.3974C>T (p.Ser1325Phe)); with VWD no loss of HMW VWF and VWF:Ag ~190 U dL1, RCo/Ag <0.3 and CB/Ag >0.7 HMW, high molecular weight; VWD, von Willebrand disease; VWF, von Willebrand factor; RCo, ristocetin cofactor; CB, collagen binding. *Samples listed in the same cell indicate repeated despatches of the same plasma sample in different years. Haemophilia (2016), 22, e145--e155 © 2016 John Wiley & So2n6s0Ltd 261 TYPE 2M VWD IS OFTEN MISIDENTIFIED e147 et al. [11]. Such samples generate VWF and FVIII test All samples sent to participants were buffered lyo- results and VWF multimer patterns that are very simi- philized plasma, and tested blind by laboratories. All lar to true type 2A VWD samples. Although patient samples passed local stability and homogeneity testing samples are preferred in research evaluations, within to ensure sample integrity (data not shown). Informa- an EQA setting it is important to balance this desire tion on stability and homogeneity testing is available against ethical and logistical challenges centred on reg- on request. VWF multimers and densitometry patterns ular collections of VWD-affected patients. In-house for most of the samples are shown in Fig. 1, and were prepared material also carries specific design advan- consistent with expectations (i.e. no HMW-VWF tages related to preferred specifications. Thus, for this reduction in the four dispatched 2M VWD samples, study, one such sample was prepared with a VWF:Ag but clear HMW-VWF reduction in the comparator level of ~30 U dL1 and was meant to mimic a stan- four 2A/2B VWD cases). dard type 2A VWD-like sample, whereas another was prepared with a VWF:Ag level of ~80 U dL1, and Data analysis was meant to mimic a type 2A VWD-like sample as might be derived from a pregnant woman [12]. In All laboratories report numerical values to the RCPA- pregnancy, VWF:Ag levels can normalize in type 2A QAP Haematology for the tests performed, and these VWD, but HMW VWF and VWF activity by assays are analysed using descriptive statistics [median, such as VWF:RCo, do not rise in unison [12]. As ranges, intra-method or inter-laboratory coefficient of noted, the two other HMW VWF reduced samples variation (CV%), etc.]. Although participants to this derived from a single well-characterized type 2B VWD EQA also perform FVIII testing, this report focuses on patient, sent as a duplicate sample in separate years. VWF assays and test results, given the type of VWF These four samples would also potentially reflect dif- samples assessed herein. In this study, we have not ferent challenges to correct VWD (2A/2B) subtype distinguished between different collagen binding or identification, given the different levels of VWF:Ag. ristocetin cofactor methods, since there was no Fig. 1. Sample von Willebrand factor (VWF) multimer gels (a–c) and densitometry (d, e) results for representative samples reported in this study, including normal, high molecular weight (HMW) VWF reduced (2009-2B, 2015-2A; a, c, e), and type 2M VWD (2014-2M(b), 2015-2M; b, c, d, e). For densitometry figures, relative density is shown on the y-axis, and relative front on x-axis, including the dye front at the extreme right. Normal samples are shown as refer- ence samples; the type 2M VWD samples retain all molecular weight forms of VWF; the HMW VWF reduced samples show a reduction in HMW forms. © 2016 John Wiley & Sons Ltd Haemophilia (2016), 22, e1452--6e1155 262 e148 E. J. FAVALORO et al. Fig. 2. Percentage of RCPAQAP Haematology participants currently performing various von Willebrand factor (VWF) assays and test panels. Note that all (i.e. 100% of) laboratories perform VWF:Ag (Ag) and factor VIII:C (FVIII) testing, but that performance of additional (‘+’) VWF activity assays and consequent test panels varies. RCo, ristocetin cofactor; CB, collagen binding; Ab, IL VWF activity assay; GPIbM, Siemens Inno- vance activity assay. Asterisks identify test panel combinations showing fewest errors in identifica- tion for the sample set reported in this study. evidence of any differential patterns based on these oratory interpretations reported to our EQA were different methods. We have, however, differentially assessed in the context of the sample provided, for assessed data reported for the Werfen-IL latex- both the variety of different interpretations as well as immuno assay (LIA) monoclonal antibody based for ‘errors’ in interpretation. For the current report, method (‘VWF:Ab’), or (from 2013) the Siemens Inno- an error in VWD type identification was defined to vance recombinant platelet glyoprotein I (rGPIb)-bind- occur when the laboratory failed to identify a HMW- ing LIA method (‘VWF:GPIbM’) (using the recent VWF reduced sample as either a type 2A or 2B VWD nomenclature proposed by the International Society sample (from the choices available as above) or when on Thrombosis and Haemostasis (ISTH) Scientific and a laboratory failed to identify the type 2M VWD cases Standardization Committee (SSC); [13]). as being 2M VWD. It is recognized that misidentifica- Laboratories are also encouraged to interpret overall tion of type 2M VWD cases as 2A VWD may reflect a test findings for the likelihood or otherwise of ‘VWD’, relatively non-serious misidentification in clinical prac- as well as potential ‘VWD type’ if applicable. It is well tice, given that these cases are managed similarly recognized that diagnosis of VWD requires more than [8,14]; however, we were interested in identifying laboratory testing, and some samples used are not how often laboratories correctly identified 2M VWD derived from true VWD patients. Nevertheless, such vs. how often this diagnosis was missed. interpretation exercises are valid within the context of EQA, and in-house prepared HMW-VWF reduced Results samples provide similar FVIII and VWF test results to true 2A VWD patient samples, and thus should be Participant numbers and test methodologies identified as representing HMW-VWF reduction (or ‘2A/2B VWD’) using the tests performed by these lab- Numbers of RCPAQAP participants have varied only oratories. The selection choices provided to laborato- slightly during the period of analysis (n = 52–60). All ries for interpretation comprise: normal/not VWD, participants (i.e. 100%) perform VWF:Ag and FVIII:C type 1 VWD, type 2A or 2B VWD, type 2M VWD, testing, but thereafter the test makeup differs substan- type 2N VWD, type 3 VWD and other (specify). Lab- tially between laboratories (Fig. 2). A large proportion Fig. 3. Numerical data returned by participants for all samples from current study, and as summarized in Tables 1 and 2. HMW-VWF reduced samples shown in (a–d), and type 2M VWD samples shown in (e–h). Figures show data for all VWF assays (VWF:Ag (antigen), VWF:RCo (ristocetin cofactor), VWF: CB (collagen binding), VWF:Ab (IL VWF activity assay), VWF:GPIbM (Siemens Innovance VWF activity assay), with VWF level shown on y-axis (U dL1); assays shown on x-axis and activity/Ag ratios on right y-axis. The shaded area defines the cut-off region for discrimination of type 2 (qualitative) defects, which may differ according to laboratory and test methodologies used, but which tends to be between 0.5 and 0.7. Arrowed values indicate representative outlier values, representing analytical error, potential transcription error, or incorrect assay type reported to EQA. For example, (a, b, d, e): high outlier VWF:RCo, VWF:CB and VWF:Ab values here likely reflect analytical error; in contrast, (g, h) arrowed outliers for VWF:Ab and VWF:GPIbM may reflect participants incorrectly identifying their VWF activity assays to our EQA. Haemophilia (2016), 22, e145--e155 © 2016 John Wiley & So2n6s2Ltd 263 TYPE 2M VWD IS OFTEN MISIDENTIFIED e149 (a) (c) (b) (d) (e) (g) (f) (h) © 2016 John Wiley & Sons Ltd Haemophilia (2016), 22, e1452--6e3155 264 e150 E. J. FAVALORO et al. of laboratories (~60%) still perform VWF:RCo test- 2M VWD cases lead to different interpretation chal- ing, although many have now replaced this test with lenges. The case giving a median of ~50 U dL1 alternate VWF activity assays (in particular either the VWF:Ag was more often misidentified as a type 2A/ Werfen-IL LIA monoclonal antibody based method 2B VWD sample (46.7%) than 2M (34.8%), with (‘VWF:Ab’), or (from 2013) the Siemens Innovance type 1 VWD (14.1%) also represented. The other 2M rGPIb-binding LIA method (‘VWF:GPIbM’)). VWF: VWD case, giving a median of ~190 U dL1 VWF:Ag, CB testing is also performed by ~60% of participant was instead often misidentified as being normal (non- laboratories. However, as shown in Fig. 2, the actual VWD) (36.4%), with identifications of type 2A/2B composite test panels currently used vary widely. VWD (13.6%) less often represented. Overall, fewer errors in interpretation were identified Numerically reported values from laboratories using comprehensive test panels including GPIb and collagen binding compared to This is shown in Fig. 3 and summarized in Table 2. those using only one or other of GPIb and collagen Occasional identified outlier data may reflect random binding. For the 2M VWD case with ~50 U dL1 error events, transcription errors, and occasionally sys- VWF:Ag, there were a total of 92 interpretative events, tematic error (e.g. potentially reflecting problems with with near equal numbers of events for those perform- a common assay calibrant standard if a laboratory ing both types of tests (n = 45) vs. those only perform- reports outliers for several tests). Nevertheless, in gen- ing one type of test (n = 47). Of these, 54.8% of those eral, the reported test findings were consistent with performing both types of tests correctly identified 2M the sample type being presented to laboratories. Thus, VWD, compared to only 8.9% of those performing the HMW-VWF reduced samples showed a general only one type of test. For the 2M VWD case with reduction in all VWF activity/VWF:Ag ratios (all gen- ~190 U dL1 VWF:Ag, there were a total of 88 inter- erally <0.5), reflecting similar sensitivities of all VWF pretative events, again with near equal numbers of activity assays to HMW-VWF-reduction. However, events for those performing both types of tests the type 2M VWD cases showed a different test pat- (n = 45) vs. those only performing one type of test tern, with low RCo/Ag ratios, but more variable test (n = 43). Of these, 66.8% of those performing both patterns with other VWF activity/Ag ratios. Thus, CB/ types of tests identified 2M VWD, compared to only Ag ratios were normal (i.e. >0.7) for one 2M VWD 25.6% of those performing only one type of test. For case, and in general higher and closer to normal val- this sample, participants performed better with the sec- ues than RCo/Ag for the other 2M VWD case. The ond sample dispatch, perhaps learning from the first Siemens Innovance VWF:GPIbM/Ag ratios followed dispatch and subsequent EQA report. Thus, for the those for RCo/Ag, whereas the IL VWF:Ab/Ag ratios first dispatch, 50.0% of those performing both types of showed discrepancies between the two represented tests correctly identified 2M VWD, compared to only 2M VWD cases. 15.8% of those performing only one type of test, whereas for the second dispatch the respective values Test result interpretation and VWD type were both higher at 82.6% and 33.3%. identification Reproducibility of test results using repeat test This is highlighted in Figs 4 and 5. Most participants samples correctly identified the HMW-VWF reduced samples as being 2A/2B-VWD-like (viz, correctly identified on As shown in Table 2 and Fig. 3, there was good 88.5% of occasions; Table 2). Overall, then, errors in reproducibility in test data for identical samples dis- identification of these HMW VWF reduced samples patched in different years, both for HMW-VWF averaged 11.5% (Table 2), and were primarily driven reduced (type 2B VWD) samples and type 2M VWD by laboratories utilizing insufficient test panels (6.3% samples. Thus, returned numerical data showed simi- contributory error rate) or misidentification of their lar median and ranges of VWF test data, again high- own test panel results (which as reported were actu- lighting some robustness in analytical test results. ally consistent with a HMW-VWF reduced sample; Interestingly, however, although the interpretations 4.2%), and infrequently due to analytical errors and VWD type identification were similar in different (1.0%). In contrast, laboratories had considerably dispatches, they were not identical (Fig. 4). For one greater difficulty in identifying type 2M VWD. In type 2M VWD case (VWF:Ag ~50 U dL1), the initial total, laboratories correctly identified type 2M VWD sample yielded a high level of misidentification as a in only 29.4% of occasions, with the average 70.6% HMW-VWF reduced (2A/2B VWD) sample, whereas error rate represented by insufficient test panels the repeat testing showed a similar error rate but a (41.7% contributory error rate), misidentification of higher level of misidentification of this samples as type participant’s own test panel results (10.0%) and ana- 1 VWD. The second 2M VWD case (VWF:Ag ~190 lytical errors (13.3%). As expected, the two different U dL1) was initially misidentified primarily as a non- Haemophilia (2016), 22, e145--e155 © 2016 John Wiley & So2n6s4Ltd 265 TYPE 2M VWD IS OFTEN MISIDENTIFIED e151 © 2016 John Wiley & Sons Ltd Haemophilia (2016), 22, e1452--6e5155 Table 2. Median test values and summary error rates for study samples. Median Median Median Median Median Analytical Sample VWF: VWF: VWF: VWF: VWF: Median Median Median Median Total Total Error issue Insufficient Participant Sample type Sample sets information Ag RCo CB Ab GPIbM RCo/Ag CB/Ag Ab/Ag GPIbM/Ag errors* interps† rate (%)‡ (n; %)§ panel¶ misinterpretation** HMW VWF 2009-2B 2B VWD; VWF 40 8 12 13 N/A 0.19 0.29 0.32 N/A 2 46 4.4 0 1 (2.2) 1 (2.2) deficient mutation samples (c.3916C>T; p.Arg1306Trp) 2011-2B Repeat of 38 10 11 11 N/A 0.26 0.29 0.32 N/A 3 49 6.1 2 (4.1) 0 1 (2.0) above sample 2013-2A HMW reduced 30 6 5 11.5 5 0.21 0.17 0.36 0.16 4 48 8.4 0 3 (6.3) 1 (2.1) 2A-VWD like 2015-2A HMW reduced; 83 9 4 18 5.5 0.11 0.04 0.25 0.04 13 49 26.5 0 8 (16.3) 5 (10.2) ‘pregnant’ 2A-VWD like All HMW VWF reduced samples above 22 192 11.5 2 (1.0) 12 (6.3) 8 (4.2) Type 2M 2012-2M 2M VWD; 48 12 43 19 NA 0.28 0.89 0.39 NA 27 49 56.0 3 (6.1) 18 (36.7) 6 (12.2) VWD VWF mutation samples (c.3965A>C; p.His1322Pro) 2014-2M (a) Repeat of above 50 14 37 19 11 0.3 0.74 0.41 0.22 35 43 81.4 7 (16.3) 21 (48.8) 7 (16.3) 2M VWD (c.3965A>C; p.His1322Pro) samples above combined 62 92 67.4 10 (10.9) 39 (42.4) 13 (14.1) 2014-2M (b) 2M VWD; 187 7 167 178.5 9.5 0.04 0.94 1.29 0.07 29 41 70.7 7 (17.1) 18 (43.9) 4 (9.8) VWF mutation (c.3974C>T; p.Ser1325Phe; homozygous) 2015-2M Repeat of above 192 9 166 202.5 14 0.05 0.97 1.00 0.08 26 47 55.3 7 (14.9) 18 (38.3) 1 (2.1) 2M VWD (c.3974C>T; p.Ser1325Phe; homozygous) samples above combined 65 88 73.9 14 (15.9) 36 (40.9) 5 (5.7) All 2M-VWD 127 180 70.6 24 (13.3) 75 (41.7) 18 (10.0) samples above HMW, high molecular weight; VWD, von Willebrand disease; VWF, von willebrand factor; Ag, antigen; RCo, ristocetin cofactor; CB, collagen binding; Ab, IL Activity assay; GPIbM, Siemens Innovance activity assay. *Errors defined as per Methods; essentially, when a laboratory failed to identify qualitative VWF defects correctly; viz, failed to identify a 2M VWD as 2M VWD or failed to identify a HMW VWF reduced sam- ple as either a 2A or 2B VWD-like sample. †Total interps = total number of interpretations provided by participants for the (combined) sample set for that sample group. ‡Error rate = total errors/total interpretations as percentage. §Analytical problem = when the test result for a given assay yielded a false normal result (when the median reported result was abnormal) or yielded a false abnormal result (when the median reported result was normal). ¶Laboratory performed an insufficient test panel to make an appropriate interpretation (generally, these labs only performed VWF:Ag and FVIII:C assays, or VWF:Ag and FVIII:C and VWF:RCo assays, or VWF: Ag and FVIII:C and VWF:CB assays). **Participant misinterpretation of their own data set; i.e. the data set reported by the participant was actually consistent with the sample type provided, and not the interpretation they reported. Bold text for median test values indicate abnormal test results. Bold text for summary errors indicates data for combined sample sets (either HMW VWF deficient samples or 2M VWD samples), with italic bold indicating data for combined sample set pairs (i.e., same sample sent in different surveys). 266 e152 E. J. FAVALORO et al. Fig. 4. Participant interpretations (‘VWD diag- noses’) for the sample set reported in this study. Other comprised occasional responses of ‘equivocal’, haemophilia, or type 2N VWD. Fig. 5. Error rates for the sample set reported in this study, according to error type. Insufficient test panels and participant misinterpretations accounted for far more errors in VWD type identification than analytical errors. VWD sample, and although the repeat testing identi- reduced’) or not VWD. In our EQA only 2/60 (3.3%) fied fewer overall errors, these were still predomi- laboratories currently perform multimer analysis; thus, nantly that of a failure to identify VWD. HMW-VWF reduction (or identification of samples as 2A/2B-VWD like), is achieved using surrogate labora- tory markers, most notably the ratio of VWF activity/ Discussion Ag. This is likely to also be the case worldwide, as This study was focused on a comparative evaluation more laboratories do not perform multimer analysis of HMW-VWF reduced samples (‘type 2A or 2B-like than do perform multimer analysis [15]. Thus, data VWD’) vs. type 2M VWD to address the question of from other EQAs indicate VWF multimers are only whether 2M VWD is a rare type of VWD, as largely performed by 1–20% of VWF test laboratories, reported in the literature [1], or alternatively (as we depending on the reporting EQA [15]. hypothesized) might just not be identified correctly by In HMW-VWF reduced samples (i.e. 2A or 2B many laboratories [8]. Four plasma samples represent- VWD), a low VWF activity/Ag ratio would be expected ing repeat dispatches of two different cases of 2M using all standard VWF activity assays (i.e. VWF:RCo, VWD identified a high error rate of VWD misidentifi- VWF:CB, VWF:Ab and VWF:GPIbM), and reflecting cation, as either being 2A/2B-like (i.e. ‘HMW-VWF similar sensitivity of these assays for HMW-VWF. This Haemophilia (2016), 22, e145--e155 © 2016 John Wiley & So2n6s6Ltd 267 TYPE 2M VWD IS OFTEN MISIDENTIFIED e153 was in general observed in our study, although the most centres (including some expert centres) report RCo/Ag, CB/Ag and GPIbM/Ag ratios appeared to type 2M VWD as being a relatively uncommon Type show better sensitivity to HMW-VWF reduction than 2 VWD disorder [1], instead ‘preferentially’ identify- the Ab/Ag ratio (Table 2 and Fig. 3). That the VWF:CB ing a high proportion (>80%) of type 2 cases as being and VWF:RCo are better able to discriminate HMW- type 2A. We believe that this simply reflects a ‘report- VWF reduction than the Werfen-IL VWF:Ab assay has ing bias’ to identification of qualitative (type 2) VWD also been previously reported by us [9,10]. cases reflecting low VWF activity/Ag ratios as being For type 2M VWD, the relative sensitivity of differ- due to reduction in HMW-VWF, whether or not ent VWF activity assays (i.e. VWF:RCo, VWF:CB, HMW VWF is actually reduced, and because most VWF:Ab and VWF:GPIbM) would instead reflect their laboratories do not perform multimer analysis. As ability to identify structural changes in VWF as repre- VWF:RCo is the most commonly employed VWF sented by the specific VWF mutation. For the type 2M activity assay, and this will yield low RCo/Ag ratios VWD case representing a heterozygous c.3965A>C with most 2A, 2B and 2M VWD cases, many labora- VWF mutation (p.His1322Pro), the VWF:CB (and tories may simply assume that if a low RCo/Ag ratio thus the CB/Ag ratio) was least sensitive, and would is obtained, that this reflects HMW-VWF reduction thus identify this case as a mutation that primarily (rather than a VWF dysfunction per se). A similar affects the VWF-platelet GPIb binding domain (and finding would be expected if laboratories performed not VWF-collagen binding). Nevertheless, the IL the Siemens Innovance activity assay (VWF:GPIbM), VWF:Ab (and thus the Ab/Ag ratio) was also rela- either instead of VWF:RCo, or in addition to VWF: tively less sensitive to this mutation than the VWF: RCo. RCo (and RCo/Ag) and Siemens VWF:GPIbM (and Although many might therefore argue that this then GPIbM/Ag ratio) (Fig. 3). For the type 2M VWD case validates the need for greater performance of VWF:- representing a homozygous c.3974C>T mutation multimers, we do not believe that this would provide (p.Ser1325Phe) [16], the VWF:CB (and CB/Ag ratio) a ‘panacea’ to this issue, given the high error rate was also again insensitive, and thus also identifying otherwise associated with this test. Thus, in one this case as a mutation that primarily affects the recent study, NASCOLA (North American Specialized VWF-platelet GPIb binding domain (and not the Coagulation Laboratory Association) reported an VWF-collagen binding domain). In this case, both overall average error rate of 15% associated with VWF:RCo (and RCo/Ag) and VWF:GPIbM (and VWF multimer performance, with 5% of laboratories GPIbM/Ag) were similarly sensitive to the mutational reporting a loss of HMW VWF in normal samples, effect (Fig. 3). However, the IL VWF:Ab assay failed 18% reporting a loss of HMW VWF in type 1 VWD to identify the mutational change, and yielded normal samples, and 18% reporting a normal multimer pat- Act/Ag ratios with this sample. We believe that this is tern in HMW reduced type 2 VWD [24]. More con- at least the third reported type 2M VWD case that cerning was an ECAT (External quality Control of this assay has failed to identify [17,18]. One of these diagnostic Assays and Tests) report [25], which asso- previous cases was not specifically reported, but ciated VWF multimers with a 23% error rate for nor- merely shown as a discrepancy in VWF:RCo (nearly mal samples and an up to 52% error rate in type 1 0 U dL1) vs. the IL VWF:Ab at ~70 U dL1 [17]. VWD. The other ‘case’ was that of a family expressing a The experience of the RCPA QAP, and of the West- p.Gly1324Ala change in VWF, where the VWF:Ab mead laboratory, therefore reiterates the value of the was either similar to VWF:Ag or where this resulted a VWF:CB in helping to identify and discriminate type normal value [18,19]; here, the cases of 2M VWD 2 VWD. Essentially, a low RCo/Ag plus low CB/Ag might have been missed as VWD (where VWF levels ratio, reflective of combined discordant VWF activity were normal), or else identified as type 1 VWD (where and antigen levels, most likely reflects a loss of HMW VWF levels were low). VWF. In contrast, low RCo/Ag but normal CB/Ag In conclusion, this study has identified that for the ratio is incompatible with a loss of HMW VWF and cases reported here, type 2M VWD is more often must be a type 2M VWD. This strategy presumes both misidentified (70.6% of interpretations) than correctly the VWF:RCo and VWF:CB assays in use to be sensi- identified as 2M VWD (29.4%), and this therefore tive to loss of HMW VWF, and laboratories should potentially explains the relative under-reported inci- confirm this to utilize this strategy. The Westmead dence of type 2M VWD in the literature. In the West- laboratory knows this to be the case for our method- mead-based laboratory, we have previously identified ology, based on extensive assay characterization [26– type 2M VWD as potentially being as common as 28]. In our extended experience, this differential is type 2A VWD [20]. A few other laboratories, either in further accentuated in DDAVP testing [29]. Moreover, Australia or Italy, have reported a similar incidence of addition of the VWF:CB to the VWD diagnostic panel Type 2M VWD (at least 50% of type 2VWD cases) consistently reduces diagnostic errors according to based on their own experience [21–23]. In contrast, previous data from the RCPA QAP [10,30,31]. The © 2016 John Wiley & Sons Ltd Haemophilia (2016), 22, e1452--6e7155 268 e154 E. J. FAVALORO et al. Siemens Innovance VWF:GPIbM assay provides data Most importantly, although both 2M and 2A VWD that is broadly similar to VWF:RCo [32]. are currently treated similarly, current practice may We also caution on the use of the IL monoclonal not reflect the best possible clinical practice. There is antibody based VWF activity assay (VWF:Ab), as it some evidence, for example, that 2A and 2M VWD may fail to identify some type 2M VWF mutations, may reflect different severity disorders, just as haemo- with the case reported here being at least the third philia A and B are now reported to reflect [8]. Also, such case reported in the literature. We do acknowl- with future formulations of recombinant (r) VWF, it edge, however, that this may be a relatively rare event, may be possible to treat patients according to the con- given that others reporting their experience with this cept of personalized therapy, and this will only pro- assay did not report such discrepant behaviour [33– gress if laboratories and clinicians better recognize the 35]. On the other hand, the IL VWF activity assay differential that 2A and 2M VWD may represent. (VWF:Ab) also appears to be relatively less sensitive Thus, for 2A VWD, an enriched HMW rVWF may to HMW-VWF reduction than the other commonly become available, whereas for 2M VWD, a different utilized VWF activity assays (VWF:CB, VWF:RCo, approach may eventuate [37]. Better differential diag- VWF:GPIbM) (current report and [9,10]). nosis of 2A vs. 2M VWD will also enable develop- We acknowledge some limitations with our study. ment of evaluable patient cohorts that will in turn In particular, although this study comprises a compar- permit better identification of differential clinical ative cross laboratory evaluation of two distinct and symptoms, as well as justifying future advances in well characterized type 2M VWD cases against HMW rVWF development, therapy and personalizing patient VWF reduced samples, including a well characterized therapy. At the moment, the lack of distinction of 2M type 2B case sent twice, the ‘type 2A’ comparators VWD by most laboratories will thwart such possible comprised artificially prepared samples. This permits future therapy landscapes, and currently causes a us to circumvent ethical concerns regarding unneces- ‘melding’ of VWD types that in itself obscures the sary collection of patient samples, recognizing the vol- clinical differentials. ume of plasma required for EQA use. In addition, Finally, the United Kingdom Haemophilia Centre generation of 2A VWD ‘mimic’ samples facilitates Doctors Organization/British Committee for Standards specific design characteristics, such as a ‘pregnancy’ in Haematology recommend that diagnosis of VWD like sample in the current report. Moreover, such sam- be undertaken using both GPIb binding and collagen ples can be shown to provide laboratory test data for binding assays [38], useful for the discrimination of VWF:Ag and activity, as well as FVIII and VWF mul- type 2M VWD, but also for minimizing errors in diag- timers, that is similar to real 2A VWD samples. Nev- nosis of other VWD types, including 2A, 2B, 1 and 3 ertheless, these are not real cases and this can raise VWD [10,30,31]. Our findings are in full support of concerns over conclusions. On the other hand, the these recommendations. error rates in identifying these artificial samples as 2A- like were comparatively low in comparison with the Acknowledgements error rates associated with identifying 2M VWD sam- ples as 2M VWD (Table 2), and the ‘text book’ 2 The authors thank participants of the RCPAQAP Haematology. They VWD like sample (2013-2A) yielded similar error also thank Jane McDonald, Ella Grezchnik, Monica Ahuja and Yifang rates to the 2B VWD cases reported here, as well as Zhang from the Westmead ICPMR laboratory for technical assistance related to sample testing including homogeneity and stability testing. reported historically [10,31]. Some may also question the ultimate significance of some of our findings; in particular, what is the clinical Author contributions significance if 2M VWD is instead identified as either E. J. Favaloro prepared some samples for this study, performed sample 2A (or even type 1 VWD for that matter)? In essence, testing, undertook data analysis and wrote the original draft of the paper. all such VWD cases may be given a DDAVP trial and R. A. Bonar undertook some data analysis, and assisted with sample dis- may or may not show sufficient responsiveness for a tribution logistics. M. Meiring performed sample testing including multi- particular clinical purpose (e.g. minor surgery); should mer analysis and densitometry. S. Mohammed performed sample testing and some data analysis. J. Sioufi and K. Marsden are involved in the DDAVP responsiveness not be sufficient, then all sub- logistical and supportive aspects of the RCPAQAP Haematology VWF/ types will be given VWF factor concentrates, essen- VWD EQA programme. A. Arbelaez, J. Niemann and R Freney were tially at the same ‘dose’, to achieve the ‘same’ target involved in the work up for, and provided, one of the 2M VWD cases utilized in this study. All authors were involved in manuscript preparation levels of FVIII and VWF:RCo. Moreover, additional and approved the final version. therapeutic interventions (e.g. tranexamic acid) will also be similarly applied to both types of VWD [14,36]. On the other hand, the current classification Disclosures scheme [5] distinguishes 2M VWD as a distinct quali- The authors stated that they had no interests which might be perceived as tative subtype, and there are good reasons for this. posing a conflict or bias. Haemophilia (2016), 22, e145--e155 © 2016 John Wiley & So2n6s8Ltd 269 TYPE 2M VWD IS OFTEN MISIDENTIFIED e155 References nes, the National Heart, Lung, and Blood 27 Favaloro EJ. An update on the von Wille- Institute (NHLBI) Expert Panel report brand factor collagen binding (VWF:CB) 1 Favaloro EJ. Von Willebrand disease: local (USA). Haemophilia 2008; 14: 171–232. assay: 21 years of age and beyond adoles- diagnosis and management of a globally 15 Favaloro EJ, Plebani M, Lippi G. Regula- cence, but not yet a mature adult. Semin distributed bleeding disorder. Semin tion in hemostasis and thrombosis: part I- Thromb Hemost 2007; 33: 727–44. Thromb Hemost 2011; 37: 440–55. in vitro diagnostics. Semin Thromb Hemost 28 Favaloro EJ. Evaluation of commercial von 2 Flood VH. Perils, problems, and progress in 2013; 39: 235–49. Willebrand factor collagen binding assays laboratory diagnosis of von Willebrand dis- 16 Arbelaez A, Niemann J, Freney R et al. to assist the discrimination of types 1 and 2 ease. Semin Thromb Hemost 2014; 40: 41–8. Bleeding in the jungle. Am J Hematol von Willebrand disease. Thromb Haemost 3 Yee A, Kretz CA. von Willebrand factor: 2015; 90: 843–6. 2010; 104: 1009–21. form for function. Semin Thromb Hemost 17 Sucker C, Senft B, Scharf RE, Zotz RB. Deter- 29 Favaloro EJ, Dean M, Grispo L, Exner T, 2014; 40: 17–27. mination of von Willebrand factor activity: Koutts J. von Willebrand’s disease: use of 4 Favaloro EJ. Diagnosis and classification of evaluation of the HaemosIL assay in compar- collagen binding assay provides potential von Willebrand disease: a review of the dif- ison with established procedures. Clin Appl improvement to laboratory monitoring of ferential utility of various functional von Thromb Hemost 2006; 12: 305–10. desmopressin (DDAVP) therapy. Am J Willebrand factor assays. Blood Coagul 18 Trossaert M, Ternisien C, Lefrancois A Hematol 1994; 45: 205–11. Fibrinolysis 2011; 22: 553–64. et al. Evaluation of an automated von 30 Favaloro EJ, Bonar R, Kershaw G et al.; on 5 Sadler JE, Budde U, Eikenboom JC et al.; Willebrand factor activity assay in von behalf of the RCPA QAP in Haematology. Working Party on von Willebrand Disease Willebrand disease. Clin Appl Thromb Reducing errors in identification of von Classification. Update on the pathophysiol- Hemost 2011; 17: E25–9. Willebrand disease: the experience of the ogy and classification of von Willebrand 19 Hilbert L, Fressinaud E, Ribba AS, Meyer Royal college of Pathologists of Australasia disease: a report of the Subcommittee on D, Mazurier C; INSERM network on Quality Assurance Program. Semin Thromb von Willebrand factor. J Thromb Haemost molecular abnormalities in von Willebrand Hemost 2006; 32: 505–13. 2006; 4: 2103–14. disease. Identification of a new type 2M 31 Favaloro EJ, Bonar R, Meiring M, Street A, 6 Favaloro EJ. Recent advances in labora- von Willebrand disease mutation also at Marsden K; on behalf of the RCPA QAP in tory-aided diagnosis of von Willebrand dis- position 1324 of von Willebrand factor. Haematology. 2B or not 2B? Disparate dis- ease. Expert Opin Orphan Drugs 2015; 3: Thromb Haemost 2002; 87: 635–40. crimination of functional VWF discordance 975–95. 20 Favaloro EJ, Bonar R, Favaloro J, Koutts J. using different assay panels or methodolo- 7 Favaloro EJ, Franchini M, Lippi G. Biologi- Diagnosis and management of von Wille- gies may lead to success or failure in the cal therapies for von Willebrand disease. brand disease in Australia. Semin Thromb early identification of Type 2B VWD. Expert Opin Biol Ther 2012; 12: 551–64. Hemost 2011; 37: 542–54. Thromb Haemost 2007; 98: 346–58. 8 Favaloro EJ, Pasalic L, Curnow J. Type 2M 21 Federici AB, Bucciarelli P, Castaman G 32 Favaloro EJ, Mohammed S. Towards and type 2A von Willebrand disease: simi- et al. Management of inherited von Wille- improved diagnosis of von Willebrand dis- lar but different. Semin Thromb Hemost brand disease in Italy: results from the ret- ease: comparative evaluations of several 2016. in press. rospective study on 1234 patients. Semin automated von Willebrand factor antigen 9 Favaloro EJ, Bonar R, Chapman K, Meir- Thromb Hemost 2011; 37: 511–21. and activity assays. Thromb Res 2014; 134: ing M, Adcock DF. Differential sensitivity 22 Rodgers SE, Lloyd JV, Mangos HM, Dun- 1292–300. of von Willebrand factor ‘activity’ assays to can EM, McRae SJ. Diagnosis and manage- 33 Salem RO, Van Cott EM. A new auto- large and small VWF molecular weight ment of adult patients with von Willebrand mated screening assay for the diagnosis of forms: a cross-laboratory study comparing disease in South Australia. Semin Thromb von Willebrand disease. Am J Clin Pathol ristocetin cofactor, collagen binding and Hemost 2011; 37: 535–41. 2007; 127: 730–5. monoclonal antibody based assays. J 23 Iorio A, Oliovecchio E, Morfini M, Man- 34 Pinol M, Sales M, Costa M et al. Evalua- Thromb Haemost 2012; 10: 1043–54. nucci PM; Association of Italian Hemophi- tion of a new turbidimetric assay for von 10 Favaloro EJ, Bonar RA, Meiring M et al. lia Centres Directors. Italian Registry of Willebrand factor activity useful in the gen- Evaluating errors in the laboratory identifi- Haemophilia and Allied Disorders. Objec- eral screening of von Willebrand disease. cation of von Willebrand disease in the real tives, methodology and data analysis. Hae- Haematologica 2007; 92: 712–3. world. Thromb Res 2014; 134: 393–403. mophilia 2008; 14: 444–53. 35 Chen D, Tange JI, Meyers BJ et al. Valida- 11 Chen J, Reheman A, Gushiken FC et al. N- 24 Chandler WL, Peerschke EIB, Castellone tion of an automated latex particle acetylcysteine reduces the size and activity of DD, Meijer P; on behalf of the NASCOLA enhanced immunoturbidimetric von Wille- von Willebrand factor in human plasma and Proficiency Testing Committee. von Wille- brand factor activity assay. J Thromb Hae- mice. J Clin Invest 2011; 121: 593–603. brand factor assay proficiency testing. The most 2011; 9: 1993–2002. 12 Castaman G. Changes of von Willebrand North American specialized coagulation 36 Curnow J, Pasalic L, Favaloro E. Treat- factor during pregnancy in women with and laboratory association experience. Am J ment of von Willebrand disease. Semin without von Willebrand disease. Mediterr J Clin Pathol 2011; 135: 862–9. Thromb Haemost 2016. doi: 10.1055/s- Hematol Infect Dis 2013; 5: e2013052. 25 Meijer P, Haverkate F. An external quality 0035-1569070. 13 Bodo I, Eikenboom J, Montgomery R, assessment program for von Willebrand 37 Favaloro EJ. Towards personalized therapy Patzke J, Schneppenheim R, Di Paola J; on factor laboratory analysis: an overview for von Willebrand disease: future role of behalf of Subcommittee on von Willebrand from the European concerted action on recombinant products. Blood Transfus Factor. Platelet-dependent von Willebrand thrombosis and disabilities foundation. 2016. in press. factor activity. Nomenclature and method- Semin Thromb Hemost 2006; 32: 485–91. 38 Laffan MA, Lester W, O’Donnell JS et al. ology: communication from the SSC of the 26 Favaloro EJ, Grispo L, Exner T, Koutts J. The diagnosis and management of von ISTH. J Thromb Haemost 2015; 13: 134– Development of a simple collagen binding Willebrand disease: a United Kingdom 150. doi: 10.1111/jth.12964. assay aids in the diagnosis of, and permits Haemophilia Centre Doctors Organization 14 Nichols WL, Hultin MB, James AH et al. sensitive discrimination between, Type I guideline approved by the British Commit- von Willebrand disease (VWD): evidence- and Type II von Willebrand’s disease. tee for Standards in Haematology. Br J based diagnosis and management guideli- Blood Coag Fibrinolysis 1991; 2: 285–91. Haematol 2014; 167: 453–65. © 2016 John Wiley & Sons Ltd Haemophilia (2016), 22, e1452--6e9155 270 Copyright of Haemophilia is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. 270 S-30 271 Technical report 475 C Potential supplementary utility of combined PFA-100 and functional von Willebrand factor testing for the laboratory assessment of desmopressin and factor concentrate therapy in von Willebrand disease Emmanuel J. Favaloroa, Jim Thomb, David Pattersonc, Sarah Justd, Maria Baccalab, Tracy Dixone, Muriel Meiringf, Jerry Kouttsa, John Rowelld and Ross BakerbWe performed a retrospective audit of cross-laboratory testing of desmopressin and factor concentrate therapy to assess the potential utility of supplementary testing using the PFA-100 with functional von Willebrand factor (VWF) activity testing. Data were evaluated for a large number of patients with von Willebrand disease of type 1, type 2A or type 2M, as well as a comparative subset of individuals with haemophilia or carriers of haemophilia. Laboratory testing comprised pre and postdesmopressin, or pre and postconcentrate, evaluation of factor VIII, VWF antigen (VWF:Ag) and VWF ristocetin cofactor activity as traditionally performed, supplemented with collagen-binding (VWF:CB) testing and PFA-100 closure times. In brief, both therapies tended to normalize VWF test parameters and closure times in individuals with type 1 von Willebrand disease, with the level of correction in closure times related to the level of normalization of VWF, particularly the VWF:CB. However, although occasional correction of closure times was observed in patients with type 2A or type 2M von Willebrand disease, these did not in general normalize PFA-100 closure times either with desmopressin or factor concentrate therapy. In these patients, improvement in closure times was more likely in those in whom VWF:CB values normalized or when VWF:CB/VWF:Ag ratios normalized. This study confirms that there is a strong relationship between the presenting levels of plasma VWF and PFA-1000957-5235  2009 Wolters Kluwer Health | Lippincott Williams & Wilkins opyright © Lippincott Williams & Wilkins. Unauthclosure times, and that the supplementary combination of PFA-100 and VWF:CB testing might provide added clinical utility to current broadly applied testing strategies limited primarily to VWF:Ag, VWF ristocetin cofactor and factor VIII:coagulant. Future prospective investigations are warranted to validate these relationships and to investigate their therapeutic implications. Blood Coagul Fibrinolysis 20:475–483  2009 Wolters Kluwer Health | Lippincott Williams & Wilkins. Blood Coagulation and Fibrinolysis 2009, 20:475–483 Keywords: concentrates, DDAVP, desmopressin, PFA-100, therapy, von Willebrand disease, von Willebrand factor aDepartment of Haematology, Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, New South Wales, bDepartment of Haematology, Royal Perth Hospital, Western Australia, Australia, cDepartment of Haematology, Canterbury Health Laboratories, Christchurch, New Zealand, dDepartment of Haematology, Pathology Queensland, Royal Brisbane Hospital, Queensland, eDepartment of Haematology, Fremantle Hospital, Western Australia, Australia and fDepartment of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa Correspondence to Dr E.J. Favaloro, Department of Haematology, Institute of Clinical Pathology and Medical Research (ICPMR), Westmead Hospital, SWAHS, Westmead, NSW 2145, Australia Tel: +612 98456618; fax: +612 96892331; e-mail: emmanuel.favaloro@swahs.health.nsw.gov.au Received 12 January 2009 Revised 12 March 2009 Accepted 5 April 2009Introduction Von Willebrand disease (VWD) is the most common inherited bleeding disorder and is characterized by low levels, or abnormal function, of the plasma protein von Willebrand factor (VWF). Six types of VWD have been defined, namely types 1, 2A, 2B, 2M, 2N and 3 [1–3]. A number of recent guidelines [4–7] have been published on the diagnosis and management of VWD. Management typically comprises either desmopressin (DDAVP; Aven- tis Pharmaceuticals, Inc., Bridgewater, New Jersey, USA) or VWF/factor VIII (FVIII) concentrate therapy depend- ing on the VWD type and the individual patient response. In brief, DDAVP therapy is typically utilized for most cases of type 1 VWD and for a subset of patients with type 2A, 2M or 2N VWD. As it is difficult to predict individual DDAVP responsiveness, it is a standard practice toundertake a DDAVP trial. VWF/FVIII concentrate therapy is used for patients unresponsive to DDAVP, or in whom DDAVP is thought to be contraindicated, or in patients with prolonged therapeutic need. Irrespective of the type of therapy, current recommendations suggest monitoring of therapy primarily using the ristocetin cofactor (VWF:RCo) assay and/or the FVIII coagulant (FVIII:C) assay [3–7], sometimes supplemented by test- ing of VWF antigen (VWF:Ag). Recent international attention has begun to focus on the collagen-binding assay (VWF:CB) as a supplementary test of VWF activity [2,3] and also on the PFA-100 (Dade-Behring, Sydney, Australia), a simple point of care instrument utilized to assess primary haemostasis and sensitive to both VWD and its associated therapy [8].DOI:10.1097/MBC.0b013e32832da1ad 271 orized reproduction of this article is prohibited. 476 B2l7o2od Coagulation and Fibrinolysis 2009, Vol 20 No 6 C Some of us have previously reported on the potential added utility of the VWF:CB and the PFA-100 for monitoring therapy in VWD using a small number of patients [9]. To our knowledge, this has only been effectively evaluated by one other group [10–13], which most recently reported a large study in VWD, with differential findings according to VWD type. In the current study, we have further and independently assessed the question of supplementary utility, with a similarly large case series of VWD patients in a cross- laboratory audit process. Materials and methods Study aims and setting We report a cross-laboratory, retrospective audit of DDAVP and VWF/FVIII concentrate usage and effect. The main aim of this study was to assess whether supple- mentary laboratory testing using PFA-100 closure times together with VWF activity using the VWF:CB would provide any potential improvement in the therapeutic monitoring of VWD compared with the international laboratory standard test panel of VWF:Ag, VWF:RCo and FVIII:C. Although some expert laboratories perform the supplementary test of VWF multimers, this test is no longer routinely available within our geographic locality and indeed to most ‘real-world’ pathology laboratories [2,3]. Furthermore, VWF multimer analysis, when avail- able and as applied to typing of VWD, is often poorly applied in clinical practice, with high-associated error rates [14]. Accordingly, we wished to observe whether we could identify differential patterns according to type of VWD that would be potentially useful for most testing laboratories without the requirement for additional more complex or availability-limited testing. The original study population comprised over 300 indi- viduals treated with either DDAVP or VWF/FVIII con- centrates for treatment of various bleeding disorders, but the current report is limited to individuals with VWD (type 1, 2A or 2M) or with haemophilia or carriers of haemophilia for comparison. Laboratory testing com- prised pre and post-DDAVP or VWF/FVIII concentrate evaluation of the standard recommended test panel of FVIII:C, VWF:Ag and VWF:RCo [4–7], supplemented by VWF:CB and PFA-100 closure times using collagen/ epinephrine (C/Epi) and/or collagen/ADP (C/ADP) test cartridges. Data were derived from five diagnostic and treatment centres broadly distributed within our geo- graphic region (i.e. New Zealand as well as differential east, west and north parts of Australia). Data collection was from local hospital databases and included the period since 1991 (Westmead Hospital), 2003 (Royal Perth Hospital), 2002 (Canterbury Health) or more recently (Royal Brisbane Hospital and Fremantle Hospital). Patients were initially diagnosed by the respective diag- nostic/treatment centre, although the specific type ofopyright © Lippincott Williams & Wilkins. UnauthoVWD was occasionally revised following a re-review of overall data. Study population and patient von Willebrand disease typing The current study is an extension of a previous study performed by us that looked at the potential combined benefit of performing VWF:CB testing with VWF:RCo testing, together with DDAVP responsiveness, to help better identify and characterize different VWD patients [15]. That study incorporated data from a total of 208 patients comprising severe type 1 VWD (‘VWD- 1s’, n¼ 17), mild/moderate type 1 VWD (‘VWD-1m’, n¼ 30), ‘possible mild’ type 1 VWD (‘VWD-1p’, n¼ 113), haemophilia A/carriers (‘HA’, n¼ 19), type 2A VWD (‘VWD-2A’, n¼ 19) and type 2M VWD (‘VWD- 2M’, n¼ 10). The current study draws from that pool and incorporates patients where PFA-100 closure times were also available. Patient categories were as per our previous report [15]. Thus, type 1 VWD was defined by consist- ently low but concordant test values for all VWF tests performed and evident on initial and repeat testing. In practice, concordance is an evidence of VWF ‘activity’/ VWF:Ag ratios [i.e. both VWF:RCo/VWF:Ag (RCo/Ag) and VWF:CB/VWF:Ag (CB/Ag)] above 0.7 [1–3,14]. Alternatively, type 2 VWD was defined by low levels of VWF and consistent discordance in VWF test results (i.e. VWF ‘activity’/VWF:Ag ratios of 0.7), with the type of discordance helping to identify the probable type of VWD. In this and the previous [15] study, the VWF:CB was used by us as a surrogate marker of high-molecular weight (HMW) VWF, as previously validated and reported [2,3]. Thus, a consistently reduced CB/Ag as well as a consistently reduced RCo/Ag ratio would suggest either VWD type 2A or type 2B, whereas con- sistently reduced RCo/Ag but consistently normal CB/Ag would suggest a VWD-2M (i.e. platelet function discor- dant type [1–3,15]). These patterns and classifications are both consistent with the current classification scheme [1] and with previous observations regarding the ability of optimized VWF:CB assays to most appropriately identify HMW VWF [3] more reproducibly than interlaboratory VWF:RCo testing. Type 2B VWD patients identified by heightened responsiveness to ristocetin in a ristocetin- induced platelet agglutination assay were excluded from the current study. Accordingly, a consistent test pattern comprising reduced RCo/Ag but normal CB/Ag would by definition comprise platelet function discordant VWD- 2M and exclude VWD-2A phenotype, and a consistent test pattern comprising reduced RCo/Ag and reduced CB/Ag would most likely reflect a specific loss of HMW VWF, and thus a VWD-2A phenotype. Type 1 VWD was further separated according to ‘severity’ as previously identified [15], with VWD-1s defined in those with VWF:Ag of 15 U/dl or less, VWD-1m in those with VWF:Ag between 16 and 35 U/dl and VWD-1p in those with VWF:Ag between 36 and 65 U/dl.272 rized reproduction of this article is prohibited. 273 Laboratory assessment of therapy for VWD Favaloro et al. 477 C Laboratory test methods Methodologies for phenotypic VWF and FVIII:C assays have been previously reported [15]. In brief, all labo- ratories tested for FVIII:C using a standard automated one-stage clot-based assay and for VWF:Ag using either a standard in-house ELISA or a commercial immunotur- bidimetric (‘latex-immuno-assay’) procedure. All labo- ratories performed VWF:CB (by ELISA) using an in- house or commercial assay, with all methods based on a type I/III collagen mixture (inline with current recom- mendations [3]) and validated to be sensitive to HMW VWF [2,3,15]. VWF:RCo was performed by aggluti- nation either using a platelet aggregometer or automated analyser [15]. The interlaboratory variation for these VWF tests is approximately 5–10% for FVIII:C, 10– 15% for VWF:Ag, 15–20% for VWF:CB and 20–30% for VWF:RCo [2,3]. All laboratories are specialized testing referral centres affiliated with tertiary level teaching hospitals and undertake internal assessments and validity of assays prior to incorporation into diagnostic practice. All laboratories also participate in ongoing external quality assurance processes and are accredited by local regulatory authorities for suitability of testing processes. Cross-validation of test processes for labora- tories within the current study was evaluated using retrospective data from the Royal College of Pathologists of Australasia (RCPA) haematology external quality assurance program (QAP). PFA-100 closure times were performed according to manufacturer (Dade-Behring) instructions. Desmopressin and von Willebrand factor/factor VIII concentrate sampling data This was a retrospective sampling of data from the hospital databases of several tertiary level institutions. To help standardize data analysis, we have standardized post-DDAVP and post-VWF/FVIII concentrate sam- plings to a single time-point, namely the first time-point after therapy (median 1 h after; range 30–90 min after). Current guidelines recommend the use of 0.3mg/kg desmopressin infused intravenously over a period of 30 min [4–6]. The institutions to which the authors belong all follow current guidelines. The VWF/FVIII concentrate currently used in Australia is Biostate (CSL Limited, Melbourne, Australia) having replaced antihaemophilic factor (AHF) (high purity) recently; the composition and pharmacokinetic profiles of both of these have previously been reported [16]. It is a recommended practice to trial DDAVP in certain VWD patients [4–6] and to utilize VWF/FVIII concen- trate for patients or clinical needs in whom DDAVP would be ineffective. As the institutions to which the authors belong all follow current guidelines, retro- spective patient data for DDAVP vs. VWF/FVIII con- centrate would comprise distinct sets in which either both or only one or other were utilized for any given patient.opyright © Lippincott Williams & Wilkins. UnauthResults Cross-laboratory comparisons Given that we are reporting a cross-laboratory study in which laboratories are using different test processes, we also assessed the potential for cross-laboratory variance, as this might influence ‘VWD diagnostic accuracy’ on a laboratory-to-laboratory (or patient-by-patient) basis. Accordingly, cross-laboratory testing data, available through ongoing laboratory performance in an external QAP, were evaluated for comparability, as shown in Fig. 1. Multimer profiles for these cross-tested samples are shown in Fig. 2. Using a cut-off value of 0.7, all laboratories were consistently able to identify the loss of HMW VWF in a series of type 2B VWD samples using both CB/Ag and RCo/Ag ratios, whereas testing of normal samples or those from moderate/mild type 1 VWD patients (i.e. without loss of HMW VWF) typically yielded normal assay ratios. Consistent with previously published findings, the use of the CB/Ag ratio showed greater consistency in this discrimination. Laboratory-identified responses to desmopressin Data for VWF and FVIII test responses to DDAVP for differential patient groupings have previously been reported [15]. The salient points to report are: type 1 VWD patients: good rises in all VWF test parameters (e.g. VWF:Ag, VWF:RCo, and VWF:CB) and FVIII:C post- DDAVP for each VWD subgroup. Rises in VWF:CB tend to exceed those of VWF:RCo, and hence the increase in CB/Ag ratios tend to exceed those of RCo/Ag; Haemo- philia A/carriers: similar to type 1 VWD, except that lower initial FVIII:C levels and moderate FVIII:C response. Type 2A VWD: good rises in VWF:Ag and FVIII:C but modest rises in VWF:CB and VWF:RCo, so that although CB/Ag and RCo/Ag both rise slightly, they both tend to remain below 0.7. Type 2M VWD: good rises in VWF:Ag, VWF:CB and FVIII:C but less rise in VWF:RCo, so that although CB/Ag rises, RCo/Ag tends to remain below 0.7. Summary of data for PFA-100 C/Epi closure times is shown in Fig. 3. Salient points to note are: Pre-DDAVP closure times are generally normal in haemophilia but prolonged in all VWD test groups, although in type 1 VWD, the extent of prolongation is related to the severity of the disorder. DDAVP results in normalization of closure times in type 1 VWD (with the extent of normal- ization related to the severity of VWD), but no normal- ization occurs in types 2A and 2M VWD (some minimal correction is observed in type 2M VWD). Data using the PFA-100 C/ADP test yielded similar findings (data not shown). The relationship between VWF assays and FVIII:C vs. PFA closure times for C/Epi for type 1 VWD groups (combined data) is shown in Fig. 4 and for types 2A and 2M VWD in Fig. 5. Note the strong inverse relationship in type 1 VWD, whereas type 2 VWD patterns are different. DDAVP resulted in modest shortening of273 orized reproduction of this article is prohibited. 478 B2l7o4od Coagulation and Fibrinolysis 2009, Vol 20 No 6 Fig. 1 Comparative retrospective cross-laboratory data for a series of samples cross-tested (via the RCPA Haematology QAP) by laboratories participating in the current study. A different symbol has been arbitrarily assigned to each laboratory. Data are shown for RCo/Ag (a) and CB/Ag (b) ratios using four type 2B VWD samples (6a-2B, 6b-2B, 6c-2B and 7-2B), two moderate/mild type 1 VWD samples (5-MT1 and 7-MT1) and three normal samples (5a-NM, 6-NM and 5b-NM). The dashed horizontal line indicates the cutoff value of 0.7. Ratio values below this value indicate functional VWF discordance, whereas ratio values above this indicate functional VWF concordance. All laboratories were able to identify the loss of HMW VWF in each type 2B VWD samples (i.e. ratios of RCo/Ag and CB/Ag below 0.7), although a single outlier data point (circled; above 0.7) was noted for RCo/Ag for sample 7-2B. Similarly, all laboratories obtained normal assay ratios for normal samples or those from moderate/mild type 1 VWD (i.e. ratios of RCo/Ag and CB/Ag above 0.7), although some outlier data points (circled; below 0.7) were occasionally obtained. Consistent with previously published findings, the use of the CB/Ag ratio (b) showed greater consistency in this discrimination compared with RCo/Ag (a). CB/Ag, VWF:CB/VWF:Ag; HMW, high molecular weight; RCo, ristocetin cofactor; RCo/Ag, VWF:RCo/VWF:Ag; VWD, von Willebrand disease; VWF, von Willebrand factor. Fig. 3 C closure times in some patients only (primarily 2M); in most cases, closure times were virtually unchanged, despite normalization of VWF:Ag (Fig. 5a) and FVIII:C (Fig. 5d). Notably, VWF:CB and VWF:RCo did not tend to normalize in these patients (Fig. 5b and c). Although modest shortening of closure times was occasionallyFig. 2 VWF multimer patterns for the samples cross-tested in Fig. 1, showing the loss of HMW VWF in the type 2B VWD samples. Test data are retrospective. Accordingly, test samples were run at different times, and a new normal control (NM con) sample was included with each test run. HMW, high molecular weight; VWD, von Willebrand disease; VWF, von Willebrand factor. opyright © Lippincott Williams & Wilkins. Unauthoobserved in type 2 VWD, this was often associated with normalization of VWF:CB and CB/Ag ratios (Fig. 5b and e) but not with normalization of VWF:RCo or RCo/Ag ratios (Fig. 5c and f).PFA-100 C/Epi closure times (y-axis) before and after DDAVP (¼ ) for patient groupings as identified in Materials and methods; notably, VWD-1s, n¼7/7; VWD-1m, n¼18/17; VWD-1p, n¼69/66; HA, n¼19/17; VWD-2A, n¼10/10; VWD-2M, n¼6/6. The dotted horizontal line indicates the upper limit of normal cutoff value. DDAVP, desmopressin; VWD, von Willebrand disease. 274 rized reproduction of this article is prohibited. 275 Laboratory assessment of therapy for VWD Favaloro et al. 479 Fig. 4 Comparative data for PFA-100 C/Epi closure times (x-axis) versus plasma FVIII:C and various VWF test parameters [on y-axis, U/dl; graphs (a) VWF:Ag, (b) VWF:CB, (c) VWF:RCo, (d) FVIII:C], before and after DDAVP for combined VWD-1 patient group data set (i.e. VWD-1s and VWD-1m and VWD-1p). Dotted horizontal lines indicate the nominal VWF and FVIII:C cutoff value of 50 U/dl; values for VWF or FVIII:C below this would be considered low (or abnormal) and values above this would be considered normal. Dotted vertical lines indicate the nominal (manufacturer) PFA-100 closure time cutoff value of 165 s. Thus, all pre-DDAVP VWF:Ag values would be lower than 65 U/dl (depending on the derived patient set) with most data points falling below the 50 U/dl cutoff line, and most corresponding PFA-100 closure times would be abnormal or fall above the PFA-100 closure time 165 s cutoff value. Post-DDAVP, VWF and FVIII:C test values would generally normalize (i.e. rise above the horizontal 50 U/dl cutoff line) as would PFA-100 closure times (i.e. fall below the vertical 165 s cutoff line). Note the inverse relationship between closure times and VWF test parameters. DDAVP, desmopressin; FVIII, factor VIII; RCo, ristocetin cofactor; VWD, von Willebrand disease; VWF, von Willebrand factor. C Laboratory-identified responses to von Willebrand factor/factor concentrate The relationship between VWF and FVIII:C vs. PFA- 100 C/Epi closure times for type 1 VWD groups (com- bined data) is shown in Fig. 6. Note once again the good inverse relationship observed. In contrast, closure times essentially failed to normalize or correct in type 2 VWD, despite occasional correction of VWF or FVIII (data not shown). Discussion Although a retrospective audit, this is the largest case series of VWD patients to our knowledge assessed with the panel of tests identified (e.g. FVIII:C, VWF:Ag, VWF:CB, VWF:RCo, and PFA-100 closure times), fol- lowing DDAVP treatment or VWF/FVIII concentrate therapy. In particular, several differential test patternsopyright © Lippincott Williams & Wilkins. Unauthcould be identified, as also in part previously reported by us [15]. For type 1 VWD, there was also a noticeable inverse relationship between VWF test parameters and PFA-100 closure times (Fig. 4), with VWF test increases associated with normalization or correction of PFA closure times. VWF/FVIII concentrate therapy provided a similar conclusion regarding this relationship (Fig. 6). For types 2A and 2M VWD, differential test patterns could be identified compared with each other and with type 1 VWD, both in relation to functional VWF assays and assay ratios, and neither DDAVP nor VWF/ FVIII concentrate therapy was able to substantially correct or normalize PFA-100 closure times. As data for C/ADP were not disparate to that using C/Epi, a case can be made for evaluating therapeutic respon- siveness using only a single test (i.e. C/Epi) cartridge type.275 orized reproduction of this article is prohibited. 480 B2l7o6od Coagulation and Fibrinolysis 2009, Vol 20 No 6 Fig. 5 Comparative data for PFA-100 C/Epi closure times (x-axis) versus plasma FVIII:C and various VWF test parameters [on y-axis, U/dl; graphs: (a) VWF:Ag, (b) VWF:CB, (c) VWF:RCo, (d) FVIII:C], before and after DDAVP for combined VWD-2A and 2M VWD patient group data set. Dotted horizontal and vertical lines show cutoff values as per Fig. 4. Graphs (e) and (f), respectively, show CB/Ag and RCo/Ag ratios vs. closure times. Dotted horizontal and vertical lines, respectively, identify ratios of 1.0 and PFA-100 cutoff values. DDAVP failed to normalize any closure time, and partially corrected closure times only occasionally. There was no obvious relationship between closure times and any test parameter, excepting that partial correction of closure times was more likely to occur if there was correction of VWF:CB or CB/Ag. DDAVP, desmopressin; FVIII, factor VIII; RCo, ristocetin cofactor; VWD, von Willebrand disease; VWF, von Willebrand factor. C Thus, to summarize for DDAVP, we could identify differential CB/Ag and RCo/Ag patterns among VWD groups, with elevation of both CB/Ag and RCo/Ag in type 1 VWD, selective elevation only in CB/Ag in type 2Mopyright © Lippincott Williams & Wilkins. UnauthoVWD and no substantial elevation of either CB/Ag or RCo/Ag in type 2A VWD, in which ratios tended to remain below 0.7 [15]. Similar observations regarding comparative effects with FVIII:C, VWF:Ag, VWF:RCo276 rized reproduction of this article is prohibited. 277 Laboratory assessment of therapy for VWD Favaloro et al. 481 Fig. 6 Comparative data for PFA-100 C/Epi closure times (x-axis) versus plasma FVIII:C and various VWF test parameters [on y-axis, U/dl; graphs: (a) VWF:Ag, (b) VWF:CB, (c) VWF:RCo, (d) FVIII:C], before and after VWF/FVIII concentrate therapy for combined VWD-1 patient group data set (i.e. VWD-1s and VWD-1m and VWD-1p). Dotted horizontal and vertical lines show cutoff values as per Fig. 4. Note again the good inverse relationship between closure times and VWF test parameters, similar to the case for DDAVP therapy. DDAVP, desmopressin; FVIII, factor VIII; RCo, ristocetin cofactor; VWD, von Willebrand disease; VWF, von Willebrand factor. C and VWF:CB post-DDAVP were reported by some of us for type 1 VWD using small numbers of patient sample within a single institution over 10 years ago, with test correlations to PFA-100 closure times reported sub- sequently [9]. The working hypothesis at that time was that patterns showing higher relative VWF:CB increases reflected the initial release of (ultra) HMW VWF post- DDAVP, together with the higher comparative sensitivity of VWF:CB (e.g. compared with VWF:Ag and VWF:RCo) for detection of HMW VWF [3] and the similar effect of such HMW VWF on PFA-100 closure times. That this has been validated in a current larger cross-laboratory study ([15] and current report) is, we believe, a significant find- ing. The patterns with types 2A and 2M VWD are distinct, both in terms of relative increases in VWF post-DDAVP and in PFA-100 closure times. VWF/FVIII concentrate therapy provided similar find- ings to that of DDAVP therapy for type 1 VWD, but foropyright © Lippincott Williams & Wilkins. Unauthtype 2 VWD, correction in PFA-100 closure times was an even rarer event than that observed with DDAVP, and there was only a minor relationship between closure time correction and correction in VWF:CB or CB/Ag. This latter finding was again similar to that previously reported for a small single institution pilot study [9]. Notably, those type 2M patients, in whom some shortening of closure times was observed, were those in whom DDAVP managed to increase CB/Ag ratios above 1.0. VWF/FVIII concentrate therapy generally failed to shorten closure times, and in no patient did CB/Ag ratios rise above 1.0. There have been a large number of studies that have assessed the potential added utility of the PFA-100 in monitoring therapy in VWD and a smaller number of studies assessing the potential added utility of the VWF:CB in this setting, as recently reviewed and reported [9]. There are, however, very few studies that have assessed both the PFA-100 and the VWF:CB in this277 orized reproduction of this article is prohibited. 482 B2l7o8od Coagulation and Fibrinolysis 2009, Vol 20 No 6 C setting, and data from only one other group provide good comparative data [10–13]. Thus, although Giannini et al. [17] and Hanebutt et al. [18] have recently reported on DDAVP therapy in VWD and included both VWF:CB and PFA-100 closure time testing in their respective studies, neither group provided detailed data on com- parative responses within their patient groups. Overall, our data for type 1 VWD are similar to that most recently reported by van Vliet et al. [13], but data are somewhat disparate for type 2 VWD, as these workers saw better correction of closure times for both DDAVP and VWF/FVIII concentrate therapy (i.e. a greater num- ber of their type 2 VWD patients showed closure time corrections). Further, they identified that patients with type 2 VWD showed correction only if the CB/Ag ratio also normalized. We observed occasional cases in which correction of VWF:CB or CB/Ag ratios were associated with improved closure times, but overall data indicated that this was a rare event and essentially restricted to cases in which CB/Ag ratios exceeded 1.0. There are two plausible explanations for the discrepancy between these two studies for types 2A and 2M VWD, and namely that either the therapy was more effective in their study than in ours and/or that they utilized a milder type 2 VWD patient group. van Vliet et al. [13] utilized a different VWF/factor concentrate (i.e. Haemate P) to those that we used, but published data [16] would suggest that this concentrate is in fact of a similar composition to one of the concentrates assessed within the current study. Accord- ingly, our conclusion remains that van Vliet et al. [13] utilized a ‘milder’ set of types 2A and 2M VWD patients compared with our patient set, and that in general, PFA- 100 closure times are unlikely to correct in type 2A or 2M VWD post-DDAVP or post-VWF/FVIII concentrate. We believe that the current report supports our belief regarding the potential for additional clinical utility of supplementary PFA-100 and VWF:CB testing within context of DDAVP and VWF/factor concentrate in VWD. Thus, in type 1 VWD, and following either DDAVP or VWF/factor concentrate therapy, one should generally see correction of VWF and FVIII:C as well as PFA-100 closure times, except in very severe cases of type 1 VWD. Moreover, there is a strong inverse relation- ship between closure times and VWF test parameters, with the advantage that the PFA-100 can provide data in real time, taking only minutes to perform, whereas results using a comprehensive panel of VWF tests would typi- cally not be available for some time. The situation with types 2A and 2M VWD is somewhat different. In our experience, PFA-100 closure times do not correct in these VWD types either using DDAVP or VWF/factor concen- trate therapy. Nevertheless, testing of PFA-100 closure times after therapy might be useful in patients in whom VWD types are unclear and require definition (i.e. cor- rection, likely type 1). In any case, the benefits ofopyright © Lippincott Williams & Wilkins. Unauthoadditional post-therapy testing with the PFA-100 are generally restricted to patient test cases in whom baseline VWF:Ag exceeds around 10%. Conclusion We report findings of a retrospective cross-laboratory evaluation of DDAVP and VWF/FVIII concentrate therapy in patients with VWD and haemophilia and incorporating testing with the standard panel of FVIII:C, VWF:Ag and VWF:RCo, supplemented with PFA-100 closure times and the VWF:CB to derive CB/Ag and RCo/ Ag ratios. We recognize several limitations to our study, including its retrospective nature, the lack of genetic testing to explore genetic–phenotypic correlations and the lack of multimer analysis to confirm specific VWD types based on phenotypic testing. Nevertheless, we observed differential plasma VWF test result patterns in different patient groups (i.e. haemophilia/carriers and types 1, 2A and 2M VWD) and using different laboratory assays and in particular using comparative data with activity-based VWF assays (i.e. VWF:CB and VWF: RCo). We also noted inverse relationships between all plasma VWF test results and PFA-100 closure times following either DDAVP or VWF/factor concentrate therapy in type 1 VWD but not in types 2A or 2M VWD, with correction of closure times also requiring correction of VWF (and notably VWF:CB). Our study findings warrant appropriate validation with prospective studies to definitively address the role of these tests in monitoring therapy and to investigate genetic associ- ations and therapeutic implications. Acknowledgements The following staff members are thanked for providing technical support: Soma Soltani, Jane McDonald, Ella Grezchnik, Sandra Harvey and Celeste Bell. References 1 Sadler JE, Budde U, Eikenboom JCJ, Favaloro EJ, Hill FGH, Holmberg L, et al., and the Working Party on von Willebrand Disease Classification. Update on the pathophysiology and classification of von Willebrand disease: a report of the Subcommittee on von Willebrand Factor. J Thromb Haemost 2006; 4:2103–2114. 2 Favaloro EJ. Laboratory identification of von Willebrand Disease: technical and scientific perspectives. Semin Thromb Hemost 2006; 32:456–471. 3 Favaloro EJ. An update on the von Willebrand factor collagen binding (VWF:CB) assay: 21 years of age and beyond adolescence, but not yet a mature adult. Semin Thromb Hemost 2007; 33:727–744. 4 Nichols WL, Hultin MB, James AH, Manco-Johnson MJ, Montgomery RR, Ortel TL, et al. von Willebrand disease (VWD): evidence-based diagnosis and management guidelines, the National Heart, Lung, and Blood Institute (NHLBI) Expert Panel report (USA). Haemophilia 2008; 14:171–232. 5 Pasi KJ, Collins PW, Keeling DM, Brown SA, Cumming AM, Dolan GC, et al. Management of von Willebrand disease: a guideline from the UK Haemophilia Centre Doctors’ Organization. Haemophilia 2004; 10:218– 231. 6 Federici AB, Castaman G, Mannucci PM, Italian Association of Hemophilia Centers (AICE). Guidelines for the diagnosis and management of von Willebrand disease in Italy. Haemophilia 2002; 8:607–621. 7 Laffan M, Brown SA, Collins PW, Cumming AM, Hill FG, Keeling D, et al. The diagnosis of von Willebrand disease: a guideline from the UK Haemophilia Centre Doctors’ Organization. Haemophilia 2004; 10:199–217. 8 Favaloro EJ. Clinical utility of the PFA-100. Semin Thromb Hemost 2008; 34:709–733.278 rized reproduction of this article is prohibited. 279 Laboratory assessment of therapy for VWD Favaloro et al. 483 C 9 Favaloro EJ. Laboratory monitoring of therapy in von Willebrand Disease: efficacy of the PFA-100 and VWF:CB as coupled strategies. Semin Thromb Hemost 2006; 6:566–576. 10 Michiels JJ, van de Velde A, van Vliet HH, van der Planken M, Schroyens W, Berneman Z. Response of von Willebrand factor parameters to desmopressin in patients with type 1 and type 2 congenital von Willebrand disease: diagnostic and therapeutic implications. Semin Thromb Hemost 2002; 28:111–132. 11 Michiels JJ, Berneman Z, Gadisseur A, van der Planken M, Schroyens W, van de Velde A, van Vliet H. Characterization of recessive severe type 1 and 3 von Willebrand Disease (VWD), asymptomatic heterozygous carriers versus bloodgroup O-related von Willebrand factor deficiency, and dominant type 1 VWD. Clin Appl Thromb Hemost 2006; 12:277–295. 12 Michiels JJ, van Vliet HH, Berneman Z, Gadisseur A, van der Planken M, Schroyens W, et al. Intravenous DDAVP and factor VIII-von Willebrand factor concentrate for the treatment and prophylaxis of bleedings in patients With von Willebrand disease type 1, 2 and 3. Clin Appl Thromb Hemost 2007; 13:14–34. 13 van Vliet HHDM, Kappers-Klunne MC, Leebeek FWG, et al. PFA-100 monitoring of von Willebrand factor (VWF) responses to DDAVP and FVIII/ VWF concentrate substitution in von Willebrand disease type 1 and 2. Thromb Haemost 2008; 100:462–468. 14 Meijer P, Haverkate F. An external quality assessment program for von Willebrand factor laboratory analysis: an overview from the European concerted action on thrombosis and disabilities foundation. Semin Thromb Hemost 2006; 32:485–491. 15 Favaloro EJ, Thom J, Patterson D, Just S, Dixon T, Koutts J, et al. Desmopressin therapy to assist the functional identification and characterisation of von Willebrand disease: differential utility from combining two (VWF:CB and VWF:RCo) von Willebrand factor activity assays? Thromb Res 2009; 123:862–868. 16 Favaloro EJ, Kershaw G, McLachlan AJ, Lloyd J. Time to think outside the box? Proposals for a new approach to future pharmacokinetic studies of von Willebrand factor concentrates in people with von Willebrand disease. Semin Thromb Hemost 2007; 33:745–758. 17 Giannini S, Mezzasoma AM, Leone M, Gresele P. Laboratory diagnosis and monitoring of desmopressin treatment of von Willebrand’s disease by flow cytometry. Haematologica 2007; 92:1647–1654. 18 Hanebutt FL, Rolf N, Loesel A, Kuhlisch E, Siegert G, Knoefler R. Evaluation of desmopressin effects on haemostasis in children with congenital bleeding disorders. Haemophilia 2008; 14:524–530.279 opyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. 280 S-31 International Journal of Cardiology 168 (2013) 4638–4642 Contents lists available at ScienceDirect International Journal of Cardiology j ourna l homepage: www.e lsev ie r .com/ locate / i j ca rdTestosterone and acute stress are associatedwith fibrinogen and vonWillebrand factor in African men: The SABPA study Nicolaas T. Malan a,⁎,1, Roland von Känel b,c,1, Alta E. Schutte a,1, Hugo W. Huisman a,1, Rudolph Schutte a,1, Wayne Smith a,1, Carina M. Mels a,1, Ruan Kruger a,1, Muriel Meiring d,1, Johannes M. van Rooyen a,1, Leoné Malan a,1 a Hypertension in Africa Research Team (HART), Faculty of Health Sciences, North West University, Potchefstroom, South Africa b Department of General Internal Medicine, Division of Psychosomatic Medicine, Inselspital, Bern University Hospital, Switzerland c Department of Clinical Research, University of Bern, Switzerland d Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, South Africa⁎ Corresponding author at: Hypertension of Africa Res Health Sciences, North-West University, PotchefstroomCam Street, Potchefstroom, 2520, South Africa. Tel.: +27 18 299 E-mail address: nico.malan@nwu.ac.za (N.T. Malan). 1 This author takes responsibility for all aspects of the re of the data presented and their discussed interpretation. 0167-5273/$ – see front matter © 2013 Elsevier Ireland L http://dx.doi.org/10.1016/j.ijcard.2013.07.191a b s t r a c ta r t i c l e i n f oArticle history: Background: Low testosterone, acute and chronic stress and hypercoagulation are all associated with hyperten- Received 4 June 2013 sion and hypertension-related diseases. The interaction between these factors and future risk for coronary artery Accepted 20 July 2013 disease in Africans has not been fully elucidated. In this study, associations of testosterone, acute cardiovascular Available online 30 July 2013 and coagulation stress responses with fibrinogen and von Willebrand factor in African and Caucasian men in a South African cohort were investigated. Keywords: Methods: Cardiovascular variables were studied by means of beat-to-beat and ambulatory blood pressure mon- Testosterone Haemostasis itoring. Fasting serum-, salivary testosterone and citrate coagulation markers were obtained from venous blood von Willebrand factor samples. Acute mental stress responses were evoked with the Stroop test. Fibrinogen Results: The African group demonstrated a higher cardiovascular risk compared to Caucasian men with elevated D-dimer blood pressure, low-grade inflammation, chronic hyperglycemia (HbA1c), lower testosterone levels, and elevat- Acute stress ed vonWillebrand factor (VWF) and fibrinogen levels. Blunted testosterone acute mental stress responses weredemonstrated in African males. In multiple regression analyses, higher circulating levels of fibrinogen and VWF in Africans were associated with a low T environment (R2 0.24–0.28; p ≤ 0.01), but only circulating fibrinogen in Caucasians. Regarding endothelial function, a low testosterone environment and a profile of augmented α-adrenergic acute mental stress responses (diastolic BP, D-dimer and testosterone) were associated with circulating VWF levels in Africans (Adj R2 0.24; p b 0.05). Conclusions: An interdependence between acute mental stress, salivary testosterone, D-dimer and vascular responses existed in African males in their association with circulating VWF but no interdependence of the independent variables occurred with fibrinogen levels.© 2013 Elsevier Ireland Ltd. All rights reserved.1. Introduction Black SouthAfricans are facing an epidemic of hypertension and vas- cular disease but there is still inadequate information on the physiolog- ical factors contributing to this process [1–3]. Low testosterone (T) levels are associated with endothelial dysfunction, increased arterial stiffness and hypertension [4,5], atherosclerosis [6], and target organ damage like renal impairment [7]. Endothelial dysfunction and damageearch Team (HART), Faculty of pus, Private BagX6001,Hoffman 2436; fax: +27 18 2991053. liability and freedom from bias td. All rights reserved.are involved in a number of these conditions suggesting that a low T environment may be involved in the aetiology of these conditions. Testosterone deficiency is also associated with an increase in haemostatic factors like fibrinogen and plasminogen activator inhibitor type 1 which, together with endothelial dysfunction, may increase the risk for thromboembolic events [6] and the development of hyperten- sion related target organ damage [8]. Chronic and acute stress which are major health problems of our modern world, are also associated with low T [9] and increases in vari- ous coagulation factors (fibrinogen, von Willebrand factor (VWF) and D-dimer) [10,11]. Von Känel et al. [12] demonstrated that changes in VWF and D-dimer during acute mental stress are positively correlated as markers of coagulation activation VWF is almost exclusively pro- duced by endothelial cells and is an important factor in the coagulation process as it mediates platelet adhesion and stabilization of clotting fac- tor VIII. Alpha-adrenergic activity (rather than β-adrenergic activity)280 281 N.T. Malan et al. / International Journal of Cardiology 168 (2013) 4638–4642 4639results in an increase in D-dimer [13], a covalently cross-linked product of fibrin degradation, indicating increased coagulation activation [14]. Together with D-dimer, VWF is regarded as a useful marker of endothe- lial damage and the risk of coronary artery disease (CAD) [15]. In addi- tion, Spiel et al. proposed that VWF is pathogenically and directly involved as a causative agent in the development of CAD and acute myocardial infarction [16]. Both fibrinogen and D-dimer are indepen- dently associated with increased risk for future myocardial infarction, stroke and peripheral vascular disease [17,18]. The occurrence of target organ damage is, therefore, under the direct influence of the coagulation system. The aim of this study is to determine whether a low testosterone environment and acute stress responses are associated with hypercoagulation markers in a cohort of African and Caucasian men. 2. Methods 2.1. Design and participants Participants were recruited as part of the Sympathetic Activity and Ambulatory Blood Pressure in Africans (SABPA) comparative cohort population study conducted between February and May of 2008 and 2009. The study sample comprised of urban African and Caucasian teachers working in the Kenneth Kaunda Education district in the North West Province of South Africa. The motivation for this selection was to obtain a homogenous sample from a similar working environment and socio-economic status although cultural differences could not be excluded. We invited all eligible participants between the ages of 25 and 65 years to partic- ipate. Exclusion criteria were an ear temperature above 37.5 °C, chronic use of α- and/or β-adrenergic blockers, psychotropic substance dependence or abuse, blooddonors and in- dividuals vaccinated in the past 3 months. A total of 200menwere initially included in the study. For the current sub-study participants whowere infected with HIV (N = 16), clin- ically diagnosed with diabetes (N = 8), antidepressant users (N = 2) and/or diagnosed with renal dysfunction (N = 1) were excluded. A total of 75 Africans and 98 Caucasians were thus included in the study. Participants were fully informed about the objectives and procedures of the study prior to their recruitment. All participants provided written, informed consent. The study conformed to the Helsinki Declaration and was approved by The Ethics Review Board of the North-West University, Potchefstroom Campus: Approval number 0003607S6. The protocol and methodology used in this study have been described extensively elsewhere [2]. Ambulatory blood pressure monitors (ABPM), electrocardiograms (ECG), as well as physical activity apparatus were fitted to the participants before 07 h00 min after which they commenced with their normal daily activities. Participants were transported at 16 h30 min to the Metabolic Unit Research Facility of the North-West University and were encouraged to go to bed at around 22 h00 min. They were woken at 05 h45 min the following morning to undergo a battery of clinical assessments. 2.2. Assessment of health behaviours Physical activity was assessed by means of Actical® accelerometers (Mini Mitter, Bend OR, Montréal, Québec). The device was initialised using 15-s epochs and converted to 1-min epochs for data analysis [19]. Smoking status was assessed using serum cotinine [20,21]. Serum gamma-glutamyltransferase (γ-GT) was measured and considered as a marker of alcohol abuse [22]. 2.3. Anthropometric measurements All anthropometric measurements were performed in triplicate by registered level II anthropometrists according to standardised procedures [23]. Body surface area (BSA) was calculated based on the Mosteller formula [24]. 2.4. Ambulatory blood pressure measurements On the morning of the first clinical assessment day, ABPM and 2-lead electrocardio- graph were attached to participants on the non-dominant arm at their workplace (Meditech CE120 CardioTens®; Meditech, Budapest, Hungary). This apparatus was vali- dated and approved by the British Hypertension Society. The ABPM was programmed to measure BP at 30-min intervals during the day (07 h00 min–22 h00 min) and every hour during night time (22 h00 min–06 h00 min) [25]. The successful inflation rate over this period was 75.2% (±9.8) in Africans and 84.7% (±9.1) in Caucasians. Hyperten- sive status was classified from 24 h ABPM as SBP N 125 mm Hg and/or DBP N 80 mm Hg [26]. Apparatus was removed after the last BP measurement at 06 h00 min. 2.5. Acute mental stress testing Participants were to avoid exercising, smoking and consuming alcohol and caffeine 8 h prior to saliva sampling commencement. Furthermore, they were requested to takecare with dental hygiene, to prevent bleeding gums, and not to brush their teeth before saliva sampling [27]. Application of the Stroop colour-word-conflict (CWC) testwas given. At 06 h45 min the next morning participants were in semi-recumbent position for 10 min followed by beat-to-beat BP, saliva and blood sampling. A small amount of anti- clotting solution (0.5 ml of a Heparin Sodium-Fresenius 5000 IU/ml in 50 ml normal saline solution) to prevent blood clotting was left in situ. A 10 minute resting period followed to ensure resting BP values before the Stroop colour-word-conflict (CWC) test was applied. Successive series offive colourwordswritten in incongruent colours in a ran- domorder on a cardboardwere shown to theparticipant. The participantswere instructed to recognize and verbally confirm colours [28]. The task lasted for 1 min with added time pressure and amonetary incentive on completion of the task. Thiswas followed by 10 min post stress blood samplingwhere the infusion set was thoroughly flushedwith 2–3 ml sa- line. A 2 ml blood sampling was discarded before post stress sampling was obtained. Sali- va sampling followed 30 min after CWC application. Beat-to-beat bloodpressurewas continuously assessed throughout CWC testingusing the Finometer device validated for relative changes (FinapresMedical Systems®, Amster- dam) [29]. The last 15 s of the CWC test, Finometer measurements resembling maximum BP plateau valueswere used for analyses. Data was processedwith the Beatscope 1.1 soft- ware (Finapres Medical Systems, Amsterdam, The Netherlands). The BeatScope® version 1.1a software further calculates an integrated age dependent aortic flow curve from the surface area beneath the pressure/volume curve determining β-adrenergic central cardiac responses including SBP, stroke volume (SV) and cardiac output (CO) while DBP and TPR were used to represent α-adrenergic vascular responses [2,30]. 2.6. Biochemical analyses Serum samples were analysed for total testosterone (T) using an electro chemilumi- nescence immunoassay on the Elecsys 2010 apparatus (Roche, Basel, Switzerland). Both the intra- and inter-assay coefficients of variation (CV) for all the assays were less than 10%. Saliva samples were obtained through chewing a Salivette (Sarstedt Inc., Leicester, UK) cotton swab for 1–2 min, which was snap-frozen immediately in liquid nitrogen and stored at−80 °C until analyses. Saliva samples were analysed with an immunoassay kit with chemiluminescence detection (CLIA; IBL Hamburg, Germany). Intra-assay and inter-assay coefficients of variation were less than 10%. High sensitivity C-reactive protein (CRP), cotinine and γ-GT were analysed in serum (Unicel DXC 800, Beckman and Coulter, USA; Modular ROCHE Automized, Switzerland and Konelab™ 20I Sequential Multiple Analyser Computer, ThermoScientific, Vantaa, Finland respectively). Citrate coagulation markers, fibrinogen and D-dimer were analysed with a viscosity-based Immuno- turbimetric clotting method. Instrument: STA compact; Manufacturer: TAGO diagnostic, France). The citrate VWF antigen levels were measured with a “sandwich” ELISA assay. A polyclonal rabbit anti-VWF antibody and a rabbit anti-VWF-HRP antibody (DAKO, South Africa) were used to form the assay. The 6th International Standard for VWF/FVIII was used to set the standard curve against which the samples were measured [31]. 2.7. Statistical methods Data was analysed using Statistica® software version 11.0 (Statsoft Inc., Tulsa, USA, 2012). Skewness of datawas tested andonlyγ-GT valueswere logarithmically transformed. Independent T-tests determined participant characteristic differences. Confounders which are implicated in higher sympathetic activity and cardiovascular risk were identified [age, body surface area (BSA), physical activity and log γ-GT] [2,26]. Chi-square (χ2) statistics calculated and compared proportions. General linear model analyses, independent of co- variates, were computed to test interactions within race and coagulation markers. ANCOVA's determined significant differences by comparing ethnic male groups from least square means analyses, while adjusting for confounders and resting values. Cardiovascular and coagulation responses (indicated by Δ) during the mental stressors were determined, as the ratio of the variable reaction and the resting value, by using the formula (stressor value/resting value). Multiple linear regression analyses were computed. Forward stepwise multiple re- gression analyses were performed for various models from two perspectives. Both per- spectives used circulating fibrinogen and VWF as the dependent variables. The first perspective included covariates: resting serum testosterone, Δ cardiovascular and Δ coag- ulation responses during the CWC. The second perspective included covariates: Δ salivary testosterone, Δ cardiovascular and Δ coagulation responses during the CWC. Covariates included in all the models were age, BSA, physical activity, log γ-GT and anticoagulants (aspirin and warfarin). Sensitivity analyses were computed for multiple linear regression analyses excluding one African and 9 Caucasians using oral anticoagulants. Significance was noted as p ≤ 0.05 and tendencies were noted as p b 0.1. 3. Results Table 1 shows baseline characteristics of African and Caucasianmen. The African men displayed (p ≤ 0.05) lower physical activity and BSA than the Caucasian men (Table 1). After adjusting for age, BSA, physical activity and log γ-GT, the African men, despite a smaller stature (height), displayed a higher body mass and waist circumference than the Caucasians. Furthermore, Africans demonstrated higher resting levels of CRP, VWF, fibrinogen and D-dimer than Caucasians. Sensitivity281 282 4640 N.T. Malan et al. / International Journal of Cardiology 168 (2013) 4638–4642 Table 1 3 African Caucasian Comparing adjusted baseline characteristics between African and Caucasian men. 2.3 2.5 Africans Caucasians P 2.01 (N = 75) (N = 98) 1.79 2 Unadjusted values 1.22, Age (years) 42.7 ± 8.39 44.7 ± 11.2 0.18 1.311.12 p=0.01 Body surface area (BSA) (m2) 1.94 ± 0.23 2.18 ± 0.21 b0.0001 1.5 1.14 1.02 Physical activity (kcal/24 h) 2719.2 ± 827.7 3675.8 ± 2088.4 b0.001 0.92 0.92 γ-GT, U/L (geometrical mean) 58.9 (23.4, 182.0) 26.9 (10.0, 89.1) b0.0001 1 Adjusted values: Lifestyle and biochemical variables Body mass (kg) 89.8 (89.0, 90.7) 86.0 (85.2, 86.8) b0.0001 0.5 Height (cm) 172.7 (171.2, 174.1) 179.1 (177.8, 180.4) b0.0001 Waist circumference (cm) 99.6 (97.7, 101.5) 95.1 (93.4, 96.9) 0.002 0 BMI (weight kg/length m2) 30.0 (29.3, 30.7) 26.6 (26.0, 27.3) b0.0001 Δ DBP Δ D-Dimer Δ Fibrinogen Δ vWF Δ Salivary T Cotinine (μg/L) 19.0 (−1.78, 39.9) 37.3 (17.4, 57.3) 0.27 C-reactive protein (mg/L) 5.45 (4.04, 6.86) 2.01 (1.16, 2.86) 0.003 Fig. 1. Cardiovascular and coagulation acute responses (Δ) in African vs. Caucasian Resting blood glucose 6.11 (5.71, 6.51) 5.89 (5.52, 6.25) 0.47 men where Δ values were calculated as Stroop value/Resting value. Data were adjusted (mmol/L) for age, body surface area, physical activity, log γ glutamyl transferase, and use of HbA1C (%) 6.30 (6.06, 6.54) 5.59 (5.36, 5.81) b0.001 anticoagulants. Endocrine variables Serum testosterone (nmol/L) 15.3 (13.9, 16.7) 18.6 (17.3, 19.9) b0.01 Salivary testosterone 0.47 (0.42, 0.53) 0.64 (0.59, 0.69) 0.0001 (nmol/L) Cardiovascular and coagulation variables 24 h SBP (mm Hg) 139 (136, 142) 127 (124, 129) b0.0001 24 h DBP (mm Hg) 88 (86, 90) 79 (77, 81) b0.0001 24 h heart rate (beats/min) 77 (75, 80) 73 (71, 76) 0.05 D-dimer rest (mg/L) 0.67 (0.38, 0.96) 0.25 (−0.02, 0.52) 0.06 von Willebrand factor rest (%) 85.8 (80.5, 91.1) 63.7 (58.6, 68.7) b0.0001 Fibrinogen rest (g/L) 3.23 (3.09, 3,38) 2.90 (2.77, 3.03) 0.003 TPR resting (mm Hg/ml/sec) 1.09 (0.94, 1.23) 1.05 (0.92, 1.11) 0.77 Resting SBP (Finometer) 147 (143, 152) 133 (129, 137) b0.0001 (mm Hg) Resting DBP (Finometer) 87 (85, 90) 79 (77, 81) b0.0001 (mm Hg) Stroke volume (ml/min) 103.0 (97.2, 108.9) 101.1 (95.6, 106.7) 0.68 Cardiac output (L/min) 6.79 (6.36, 7.22) 6.48 (6.08, 6.89) 0.38 Hypertensive, n (%) 59 (78.7) 67 (68.4) 0.12 Coagulation medication, n (%) 1 (1.33) 9 (9.18) Variables adjusted for age, body surface area, physical activity, γ-GT and anticoagulant users. Mean (95% confidence interval); p values ≤ 0.05 regarded as statistically significant. Where: BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; HR, heart rate; TPR, total peripheral resistance; HT, hypertension; γ-GT, γ-glutamyltransferase; HbA1C, glycated hemoglobin.analyses after excluding one African and 9 Caucasians using oral antico- agulants demonstrated no difference in the outcome (data not shown). Significant interactions, independent of covariates, demonstrated an ethnic difference for resting coagulation markers, VWF (%) [F (1, 180), 17.70; p = 0000.4] and fibrinogen [F (1, 180), 8.16; p = 0.005]. Group differences maintained significance after controlling for covari- ates age, BSA, physical activity and log γ-GT. Higher ambulatory blood pressure (BP) as well as heart rate (HR) values were evident in African men compared to the Whites. Hypertension, according to the ESH guidelines [26] was prevalent in 78.7% of Africans and 69.0% of the Whites in this study. In terms of stress responses (Fig. 1), no significant differences be- tween groups, except for blunted Δ salivary testosterone responses were evident in Africanmales compared to their Caucasian counterparts in general linear models (p = 0.01). Forward stepwise linear regression analyses demonstrated associa- tions between resting dependent variables fibrinogen and VWF includ- ing covariates (resting testosterone, Δ cardiovascular and Δ coagulation markers and confounders) following our first perspective (Table 2: models 1–4). D-dimer acute mental stress responses (ΔD-dimer) were positively associated with circulating fibrinogen (β = 0.33, p b 0.02) and VWF (β = 0.24, p = 0.03) in Africans only. Also only in Africans, Δ-DBP was positively associated with circulating VWF (β = 0.23, p = 0.04). In Caucasians,Δ-DBP showed a negative association with circulat- ing fibrinogen (β = −0.36, p = 0.0001). Resting T was negativelyassociated with circulating fibrinogen in Africans (β = −0.49, p = 0.0001) with only a tendency in Caucasians (β = −0.17, p = 0.09), as well as with circulating VWF in Africans (β = −0.41, p b 0.01). To pursue our secondperspective of assessing acutemental stress re- sponses we repeated linear stepwise forward regression analyses. We therefore aimed to demonstrate associations between circulating de- pendent variables fibrinogen and VWF including covariates (Δ salivary testosterone,Δ cardiovascular,Δ coagulationmarkers and confounders) following our second perspective (Table 3: models 5–8). ΔD-dimer remained positively associated with both circulating fibrinogen and VWF in the Africans and ΔDBP with VWF in the Africans only. In Caucasians, Δ-DBP showed a negative association with circulating fibrinogen (beta = −0.32, p = 0.001). ΔT was positively associated withVWF in theAfricans (beta = 0.27, p = 0.01) but notwithfibrinogen. When either of T or the acute mental stress response of D-dimer was left out of the model, the association of the remaining variable with fi- brinogen did not change indicating independent associations. However, when anyof the acute stress responses (ΔD-dimer,Δ salivary T orΔDBP) was left out of themodel with VWF (dependent variable), the other var- iables remained in the model (p b 0.05) with unchanged β-values but with greatly reduced R2-values. This indicates that ΔD-dimer, Δ salivary T andΔDBP are interdependent in their associationwithVWF. Therefore, in models 3 and 7, interdependent associations between the indepen- dent variables were evident for VWF % in the African only. This may indicate endothelial risk. Augmented Δ-DBP, ΔD-Dimer and Δ salivary T acute stress responses and a low resting T environment were asso- ciated with elevated VWF % (p b 0.05). 4. Discussion The aim of this study was to investigate the possible association between testosterone, acute stress and hypercoagulation in a bi- ethnic male cohort. Main findings revealed a negative association between serum T and circulating VWF as well as a positive associa- tion between salivary T stress responses and circulating VWF. We also found interdependence between Δ salivary testosterone, ΔD-dimer and Δ vascular responses in their association with VWF, a marker of endothelial dysfunction and an increased risk for stroke events and car- diovascular disease in Africans. 4.1. Cardiovascular risk in Africans Although the African and Caucasian groups differ in cultural envi- ronment, they were similar in socio-economic status (similar income, educational level and employment) and age. The African group's cardiometabolic health was in a poorer condition than that of the Caucasians with mean hypertensive blood pressure [2], low grade282 283 N.T. Malan et al. / International Journal of Cardiology 168 (2013) 4638–4642 4641 Table 2 Forward stepwise regression analyses predicting relationships between coagulation, tes- tosterone and acute cardiovascular STROOP responses. Resting serum fibrinogen Model 1: Africans Model 2: Caucasians Adjusted R2 0.28 0.35 β (±95% CI) β (±95% CI) ΔD-dimer 0.33 (0.13, 0.54), p b 0.02 – ΔDBP – −0.36 (−0.53, −0.36), p b 0.0001 Testosterone (nmol/L) −0.49 (−0.70, −0.28), −0.17 (−0.38, 0.03), p b 0.0001 p = 0.09 Resting von Willebrand factor Model 3: Africans Model 4: Caucasians Adjusted R2 0.24 0.15 β (±95% CI) β (±95% CI) ΔD-dimer 0.24 (0.03, 0.46), p = 0.03 – ΔDBP (mm Hg) 0.23 (0.01, 0.45), p = 0.04 – Testosterone (nmol/L) −0.41 (−0.66, −0.15) p b 0.01 – β denotes standardised regression coefficient. Acutemental stress changes (delta,Δ) were calculated as STROOP values/Resting values. Covariates considered for models 1–4: Age, body surface area, physical activity, log γ-GT, and anticoagulants.inflammation and pre-diabetic (HbA1c N 5.7%) status values. It is known that BP as well as coagulation factors like fibrinogen [11] and VWF [32] increase with age. However, after adjusting for age these values remained unchanged (data not shown). In accordance with Pieper et al., we found that the levels of the coagulation factors, VWF, fibrinogen and D-dimer were race related with higher values in Blacks than in Caucasians [33]. It is also known that increased BP as well as coagulation factors are risk factors for coronary artery disease (CAD) and other target organ damage (TOD) like renal impairment [8]. Sechi et al. also found a strong and independent association be- tween fibrinogen as well as D-dimer and the presence and severity of hypertension-related TOD [8]. The fact that these variables were higher in the African group may indicate a higher risk in this group for the development of CAD and TOD.Table 3 Forward stepwise regression analyses predicting relationships between coagulation and acute testosterone and cardiovascular STROOP responses. Resting serum fibrinogen Model 5: Africans Model 6: Caucasians Adjusted R2 0.17 0.25 β (±95% CI) β (±95% CI) ΔD-dimer 0.30 (0.08, 0.52), p = 0.03 – ΔDBP – −0.32 (−0.51, −0.14), p = 0.001 Δsaliva testosterone – – Resting von Willebrand factor Model 7: Africans Model 8: Caucasians Adjusted R2 0.29 0.14 β (±95% CI) β (±95% CI) ΔD-dimer 0.24 (0.03, 0.44), p = 0.03 – ΔDBP 0.37 (0.16, 0.57), p = 0.001 0.23 (0.03, 0.44), p = 0.03 Δsaliva testosterone 0.27 (0.07, 0.47), p = 0.01 – β denotes standardised regression coefficient. Acutemental stress changes (delta,Δ) were calculated as STROOP values/Resting values. Covariates considered for models 1–4: Age, body surface area, physical activity, log γ-GT, and anticoagulants.4.2. Low testosterone environment and cardiovascular risk Serum T decreases with age and a low T environment is associated with increasedBP and TOD in Caucasians aswell as in Africans [6,7]. Fur- thermore, a low T environment will induce higher levels of coagulation factors like fibrinogen [6]. The serum levels of coagulation factors like fi- brinogen and VWF will also increase with age [32]. Here, we showed that higher VWF is directly associated with a low T environment. The age adjusted T values were lower in Africans than in Caucasians and that may have contributed to their higher BP and coagulation factor values. The potential role of a low T environment on the BP of the Africans is supported by another SABPA sub-study where T was nega- tively associated with nocturnal SBP as well as daytime and nocturnal DBP [5]. 4.3. Cardiovascular and coagulation acute mental stress responses Acute and chronic mental or psychosocial stress triggers sympathetic nervous system hyperactivity and have been shown to be associated with increasedα-adrenergic vascular responsiveness, pressure overload, hypertension andmyocardial ischemic risk in Africans [2,34]. Previously, we have demonstrated that 24 h SBP predicted structural vascular changes and remodelling in lowTAfricans [5]. Interdependence between associated VWF, augmented ΔT, ΔD-dimer, Δ vascular DBP responses and a low T environment underscores the impact of α-adrenergic vaso- constrictive responses on the endothelium. For Caucasians, there was only a negative association between coag- ulation markers and DBP stress responses suggesting less vasoconstric- tion [2] improving endothelial function [35]. Conversely, the increased DBP in the Africans associated with low T [5] strengthens the notion that a low T environment will enhance VWF levels and, therefore, also blood coagulability. The interdependent asso- ciation betweenΔD-dimer andΔDBP in the Africans is an indication that acute stress may exert joint influences on coagulation factors and blood pressure. These results are in line with the findings of Wakabayashi and Masuda who found significant correlations of pulse pressure with blood fibrinolysismarkers suchasD-dimer [36]. It iswell known that acute and chronic stress is positively associatedwith coagulation factors [10]. How- ever in our study, acute stress responses (ΔD-dimer, ΔDBP and ΔT) re- vealed associations with VWF and fibrinogen only in the Africans and not in the Caucasians. Our previously reported result of an increased vul- nerability of the Africans to a low T environment was again confirmed [5,7]. As the CWC test provokes an acute stress reaction which may sim- ulate everyday stress situations there is a possibility thatAfricans are also more susceptible to the stressful situation in every-day life such as teachers in a very demanding environment [Jackson et al., 2006]. Job stress is one of the most important types of chronic stress associated with increased plasma fibrinogen and, therefore, a hypercoagulable state [37]. It may seem that the Africans find their working environment more stressful than their Caucasian counterparts [38]. It might translate into a less favourable cardiovascular health profile in general and in- creased VWF in particular. There exists an independent association between fibrinogen,ΔT and ΔD-dimer responses but that is not the case with VWF. This indicates that ΔD-dimer, ΔT and ΔDBP are interdependent in their association with VWF and ultimately there may be a synergistic action on VWF when two or all three of these variables are present. To our knowledge, this is novel data where a low T environment will directly affect the en- dothelium as shown by VWF although it is well known that an associa- tion do exist between low T and fibrinogen [6]. Several limitations should be noted. The sex hormoneswere sampled only once, thus circadian patterns could not be considered. The cross sec- tional design of the current study prevents us from being inferring cau- sality. Larger sample sizes are needed and should include interactions between sex hormone profiles and sympathetic drive. In addition, fur- ther data on autonomic and endothelial function are needed to delineate283 284 4642 N.T. Malan et al. / International Journal of Cardiology 168 (2013) 4638–4642possible physiological mechanisms at play. Direct immunoassays used in this study could have resulted in higher values in both the ethnic groups. Nevertheless, the assays used in the present study had intra- and inter- assay CVs below 10%. We conclude that in Caucasians a low T environment is associated with elevated fibrinogen levels as has been reported previously, while no association between low T, ΔT and ΔD-dimer with circulating VWF could be observed. Blood coagulability and, therefore, the risk of cardio- vascular thrombotic events and stroke in black Africans, may increase by the effect of a low T environment. If the individual is challenged during acute stress episodes the risk seems to increase as augmented ΔDBP, ΔD-dimer and ΔT responses predict increased levels of VWF, a marker of endothelial function. Indications were found that ΔT, ΔDBP and ΔD-dimer were interdependently associated with VWF but not with fibrinogen. These associations may set the stage for endothelial dysfunction, coronary artery disease and stroke risk in the African population. Acknowledgement of grant support The ethics on publishing scientific articles were followed. We grate- fully acknowledge the voluntary collaboration of the participants. The SABPA studywould not have been possible without the valuable contri- butions of co-investigators and technical staff. The present work was partially supported by the North-West Department of Education, Medical Research Council, National Research Foundation, North-West University, ROCHE Diagnostics, South Africa, and the Metabolic Syn- drome Institute, France. References [1] Opie LH, Seedat YK. Hypertension in Sub-Saharan populations. Circulation 2005;112: 3562–8. [2] Malan L, HamerM, SchlaichMP, et al. Facilitated defensive coping, silent ischaemia and ECG left-ventricular hypertrophy: the SABPA study. J Hypertens 2012;30(3):543–50. [3] Schutte AE, Schutte R, Huisman H, et al. Are behavioural risk factors to blame for the conversion from optimal blood pressure to hypertensive status in black South Africans? A five-year prospective study. Int J Epidemiol 2012;41(4):1114–23. [4] Daka B, Rosen T, Jansson PA, Larsson CA, Råstam L, Lindblad U. Low sex hormone-binding globulin is associated with hypertension: a cross-sectional study in a Swedish population. BMC Cardiovasc Disord 2013;13:30. [5] Malan NT, Hamer M, Schutte AE, et al. Low testosterone and hyperkinetic blood pressure responses in a cohort of South African men: the SABPA study. Clin Exp Hypertens 2013;35(3):228–35. [6] Jones TH. Testosterone deficiency: a risk factor for cardiovascular disease? Trends Endocrinol Metab 2010;21:496–503. [7] MalanNT, HamerM, Lambert GW, et al. Sex hormones associatedwith subclinical kid- ney damage and atherosclerosis in South African men: the SABPA study. J Hypertens 2012;12:2387–94. [8] Sechi LA, Zingaro L, Catena C, Casaccio D, de Marchi S. Relationship of fibrinogen levels and hemostatic abnormalities with organ damage in hypertension. Hyperten- sion 2000;36:978–85. [9] Yasuda M, Furuya K, Yoshii T, Ide H, Muto S, Horie S. Low testosterone level of middle-aged Japanese men — the association between low testosterone levels and quality-of-life. J Mens Health and Gend 2007;4(2):149–55. [10] Von Känel R, Mills PJ, Fainman C, Dimsdale JE. Effect of psychological stress and psy- chiatric disorders on blood coagulation and fibrinolysis: biobehavioral pathway to coronary artery disease? Psychosom Med 2001;63:531–44.[11] Wirtz PH, Redwine LS, Baertschi C, Spillman M, Ehlert U, Von Känel R. Coagulation activity before and after psychosocial stress increases with age. Psychosom Med 2008;70:476–81. [12] von Känel R, Kudielka BM, ABD-el-Razik A, Gander M-L, Frey K, Fischer JE. Relation- ship between overnight neuroendocrine activity and morning haemostasis in work- ing men. Clin Sci 2004;107:89–95. [13] Wirtz PH, Ehlert U, Emini L, et al. Anticipatory cognitive stress appraisal and the acute procoagulant stress response in men. Psychosom Med 2006;68:851–8. [14] Thrall G, Lane D, Carroll D, Lip GY. A systematic review of the effects of acute psycho- logical stress and physical activity on haemorheology, coagulation, fibrinolysis and platelet reactivity: implications for the pathogenesis of acute coronary syndromes. Thromb Res 2007;120:819–47. [15] Blann AD. Plasma von Willebrand factor, thrombosis, and the endothelium: the first 30 years. Thromb Haemost 2006;95:49–55. [16] Spiel AO, Gilbert JC, Jilma B. Von Willebrand factor in cardiovascular disease: focus on acute coronary syndromes. Circulation 2008;117:1449–59. [17] Danesh J, Whincup P, Walker M, et al. Fibrin D-dimer and coronary heart disease: prospective study and meta-analysis. Circulation 2001;103:2323–7. [18] Danesh J, Lewington S, Thompson SG, et al. Plasmafibrinogen level and the risk ofmajor cardiovascular diseases and nonvascular mortality. J Am Med Assoc 2005;294(14): 1799–809. [19] Klippel NJ, Heil DP. Validation of energy expenditure prediction algorithms in adults using the Actical electronic activity monitor. Med Sci Sports Exerc 2003;35:S284. [20] Jarvis MJ, Tunstall-Pedoe H, Feyerabend C, Vesey C, Saloojee Y. Comparison of tests used to distinguish smokers from nonsmokers. Am J Public Health 1987;77:1435–8. [21] Jarvis MJ, Feyerabend C, Bryant A, Hedges B, Primatesta P. Passive smoking in the home: plasma cotinine concentrations in non-smokers with smoking partners. Tob Control 2001;10:368–74. [22] Tsai J, Ford ES, Li C, Zhao G. Past and current alcohol consumption patterns and elevations in serum hepatic enzymes among US adults. Addict Behav 2012;37: 78–84. [23] Marfell-Jones M, Olds T, Steward A, Carter JEL. International standards for anthropo- metric assessment. New Zealand: ISAK; 2006 137. [24] Mosteller RD. Simplified calculation of body-surface area. N Eng J Med 1987;317: 1098. [25] Kohara K, Nishida W, Maguchi M, Hiwida K. Autonomic nervous function in non-dipper essential hypertensive participants: evaluation by power spectral analy- sis of heart rate variability. Hypertension 1995;26:808. [26] Mancia G, De Backer G, Dominiczak A, et al. Guidelines for the management of arte- rial hypertension: the task force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens 2007;25:1105–87. [27] Earle TL, Linden W, Weinberg J. Differential effects of harassment on cardiovascular and salivary cortisol stress reactivity and recovery in women and men. J Psychosom Res 1999;46:125–41. [28] Stroop JP. Studies in interference in verbal reactions. J Exp Psychol 1935;18:643–62. [29] Wesseling KH, Jansen JRC, Settels JJ, Schreuder JJ. Computation of aortic flow from pressure in humans using a nonlinear, three-element model. J Appl Physiol 1993;74: 2566–73. [30] Opie LH. Heart physiology — from cell to circulation. 4th ed. USA: Lippincott Williams & Wilkins; 2004. p. 421. [31] Meiring SM, Kelderman M, Coetzee MJ, Badenhorst PN. Performance and utility of a cost-effective collagen-binding assay for the laboratory diagnosis of von Willebrand disease. Clin Chem Lab Med 2007;45:1068–72. [32] Franchini M. Hemostasis and aging. Crit Rev Oncol Hematol 2006;116:1307–12. [33] Pieper CF, Rao KM, Currie MS, Harris TB, Cohen HJ. Age, functional status, and racial differences in plasma D-dimer levels in community-dwelling elderly persons. J Gerontol A Biol Med Sci 2000;55(110):M649–57. [34] Seedat Y. Perspectives on research in hypertension. Cardiovasc J Afr 2009;1:39–42. [35] Malan L, HamerM, SchlaichMP, et al. Defensive coping facilitates higher blood pres- sure and early sub-clinical structural vascular disease via alterations in heart rate variability: the SABPA study. Atherosclerosis 2013;227:391–7. [36] Wakabayashi I, Masuda H. Association of pulse pressure with fibrinolysis in patients with type 2 diabetes. Thromb Res 2007;121:95–102. [37] Dragos D, Tanasescu MD. Stress and inflammation. Rev Med Romana 2011;58(1): 31–6. [38] Jackson L, Rothmann S. Occupational stress, organizational commitment, and ill-health of educators in the North West Province. S Afr J Educ 2006;26(1):75–95.284 S-32 285 Blood Coagulation, Fibrinolysis and Cellular Haemostasis 977 Procoagulant reactivity to laboratory acute mental stress in Africans and Caucasians, and its relation to depressive symptoms: The SABPA Study Roland von Känel1,2; Mark Hamer3,4; Nico T. Malan4; Kobus Scheepers4; Muriel Meiring5; Leoné Malan4 1Division of Psychosomatic Medicine, Department of General Internal Medicine, Inselspital, Bern University Hospital and University of Bern, Switzerland; 2Department of Clinical Research, University of Bern, Switzerland; 3Department of Epidemiology and Public Health, University College London, UK; 4Hypertension in Africa Research Team (HART), Faculty of Health Sciences North West University, Potchefstroom, South Africa; 5Department of Haematology and Cell Biology, University of the Free State, Bloemfontein, South Africa Summary reactivity in VWF (p<0.001) and fibrinogen (p=0.016), but not in The risk of cardiovascular disease is dramatically increasing in Africans D-dimer (p=0.27), were decreased in Africans relative to Caucasians (black). The prothrombotic stress response contributes to athero- with Africans showing greater reactivity of total peripheral resistance thrombotic disease and is modulated by depressive symptoms. We (p=0.017). Depressive symptoms, but not general psychological dis- examined coagulation reactivity to acute mental stress and its relation tress, were associated with greater VWF increase (p=0.029) and to psychological well-being in Africans relative to Caucasians (white). greater fibrinogen decrease (p=0.030) in Africans relative to Cau- A total of 102 African and 165 Caucasian school teachers underwent casians. In conclusion, Africans showed greater hypercoagulability at the Stroop Color-Word Conflict test. Circulating levels of von Wille- rest but diminished procoagulant reactivity to acute mental stress brand factor (VWF) antigen, fibrinogen, and D-dimer were measured when compared with Caucasians. Ethnic differences in the vascular before and after the Stroop. Cardiovascular reactivity measures were adrenergic stress response might partially explain this finding. De- also obtained. All participants completed the Patient Health Question- pressive symptoms were associated with exaggerated VWF reactivity naire-9 and the General Health Questionnaire-28 for the assessment in Africans relative to Caucasians. The clinical implications of these of depressive symptoms and total psychological distress, respectively. findings for Africans need further study. After controlling for covariates, resting levels of VWF, fibrinogen, and D-dimer were higher in Africans than in Caucasians (all p-values Keywords ≤0.006). Depressive symptoms and psychological distress were not Cardiovascular disease, coagulation, depression, ethnicity, psychologi- significantly associated with resting coagulation measures. Stress cal stress Correspondence to: Received: May 11, 2013 Roland von Känel, MD Accepted after major revision: July 20, 2013 Professor of Psychosomatic and Psychosocial Medicine Prepublished online: August 22, 2013 Department of General Internal Medicine doi:10.1160/TH13-05-0383 Inselspital, Bern University Hospital Thromb Haemost 2013; 110: 977–986 CH-3010 Bern, Switzerland Tel.: +41 31 632 20 19, Fax: +41 31 382 11 84 E-mail: roland.vonkaenel@insel.ch Introduction dicted an increase in ambulatory systolic blood pressure (SBP) over three years in healthy individuals (7). Owing to impaired en- Enhanced coagulation, impaired fibrinolysis, endothelial acti- dothelial anticoagulant function, patients with atherothrombotic vation, and hyperactive platelets play an important role in the de- diseases showed greater platelet activation (8) and D-dimer in- velopment of atherogenesis, atherothrombosis, and acute coronary crease (9) than controls free of CVD. Moreover, patients with emo- syndromes (ACS) (1). During acute mental stress, healthy individ- tional triggers like depressive feelings in the 2 hours before ACS uals show a prothrombotic state that is viewed as an adaptive fight- onset, showed greater platelet aggregation during laboratory men- flight response protecting the organism from excessive bleeding tal stress than patients reporting non-emotional triggers (10). De- should injury occur (for review see [2, 3]). Fibrinogen, von Wille- pressive symptoms have also been associated with stress-induced brand factor antigen (VWF:Ag) and D-dimer, the latter indicating elevation in platelet reactivity (11) and D-dimer (12) in elderly fibrin turnover, are particularly responsive to acute mental stress subjects, thereby supporting the notion that hypercoagulability is (4, 5). Excessive stress procoagulant changes might potentially in- one mechanism linking depression with an increased CVD risk crease the risk of incident cardiovascular disease (CVD) and re- (13). current CVD events (3). For instance, stress-induced fibrinogen Potential ethnical differences in the acute procoagulant re- increase was associated with carotid artery stiffness (6) and pre- sponse to stress and their relation to depressive symptoms have © Schattauer 2013 Thrombosis and Haemostasis 110.5/2013 285 Downloaded by: University of Free State. Copyrighted material. 286 978 von Känel et al. Procoagulant stress reactivity in Africans not previously been explored. Such bio-behavioural research district in the North West Province, South Africa. This selection seems important for urbanised Africans in whom a concerning in- assured a homogenous sample from a similar socio-economic crease in CVD can be observed, much of it being a consequence of class. Exclusion criteria included pregnancy, lactation, and vacci- their transition from a traditional African to a modernised “West- nation or blood donation within three months prior to partici- ern” lifestyle (14). For instance, in Africans, the CVD risk pre- pation. dicted by fibrinogen becomes stronger with an increasing degree For purposes of the present study, we excluded participants of urbanisation (15). Moreover, subclinical atherosclerosis, as with a history of atherothrombotic disease (myocardial infarction, measured by increased carotid intima media thickness (CIMT), n=4; stroke, n=1), atrial fibrillation (n=16), and HIV-positive was greater in Africans than in Caucasians (16) as well as in those status (n=18), all of which are associated with hypercoagulability with more severe depressive symptoms of both these ethnicities at rest and/or in response to acute stress (10, 24, 25), as well as par- (17). ticipants with an ear temperature ≥37.5°C (n=3) because of a The autonomic nervous system and hypothalamic pituitary ad- probable acute phase response that might increase coagulation fac- renal axis are the two major physiologic stress response systems in tors like fibrinogen. Users of oral anticoagulants (n=2), aspirin humans. As cortisol peaks between 15 and 30 minutes (min) after (n=12), cortisone (n=3), oral contraceptives (n=21), antidepres- stress onset, prothrombotic changes within a few minutes of acute sants (n=3), anxiolytics (n=1), beta blockers (n=4), and calcium stress are largely governed by the sympathetic nervous system antagonists (n=12) were also excluded, as all of these might affect through release of catecholamines and adrenergic receptor stimu- resting coagulation activity and/or coagulation reactivity with lation (18, 19). Stress-hormone mediated stimulation of vascular acute mental stress and/or mood (21, 26-29). To allow full multi- beta2-adrenergic receptors releases VWF from endothelial cells variate statistical analyses, we also excluded subjects who missed (19-21). Stress-induced increases in D-dimer and in alpha-recep- data for coagulation measures (n=22), haemodynamic reactivity tor activating norepinephrine are correlated with each other (5), (n=10), psychological questionnaires (n=1), haemoglobin A1c but beta-adrenergic stimulation did not result in D-dimer increase (HbA1c) (n=1), creatinine (n=2), alcohol consumption (n=1), and (20). Compared to Caucasians, Africans show a greater alpha-ad- physical activity (n=1). Finally, we excluded four participants be- renergic vascular response, which is reflected by a greater total pe- cause of excessively high D-dimer levels after stress (≥10,000 ng/ ripheral resistance (TPR) (22). One specific hypothesis following ml), yielding a final study sample of 267 subjects. Excluded partici- from the above literature is that Africans would show greater pants (n=142) were significantly more often Africans, women, and D-dimer reactivity but lower VWF reactivity to acute stress than antihypertensive drug users, and they also had significantly higher Caucasians. More severe depressive symptoms would hypotheti- screening systolic blood pressure (SBP), HbA1c, and gamma glu- cally relate to greater procoagulant activity in both Africans and tamyl transferase (γ-GT) levels than those included. Caucasians. The participants were admitted at 4:30 PM to the multi-bed- The primary aim of this study was to investigate procoagulant roomed Metabolic Unit Research Facility of the North-West Uni- reactivity between Africans and Caucasians through measure- versity on Monday through Thursday. All received a standardised ments of changes in circulating levels of VWF, fibrinogen and dinner at 06:00 PM after which they completed psychological D-dimer to acute standardised laboratory stress. A secondary aim questionnaires. They had their last beverages (tea/coffee) and bis- was to explore the association of depressive symptoms both con- cuits at 8:30 PM followed by recreational activities such as reading, tinuously and categorically (i.e. based on a clinical cut-off level, watching television, or interacting socially. We also obtained infor- with stress-induced coagulation changes. We further examined mation on demographic data, general health, including medi- whether procoagulant reactivity would specifically relate to de- cation, and health behaviours. All participants went to bed at pressive mood or would rather be associated with elevated levels of around 10:00 PM and were woken at 05:45 AM. After the comple- general psychological distress. tion of anthropometric and manual BP measurements, a registered nurse obtained fasting blood samples which were handled accord- ing to standardised procedures. A sterile winged infusion set was Materials and methods left in situ with a heparin block (0.5 ml of a Heparin Sodium- Fresenius 5,000 IU/ml in 50 ml normal saline solution; Fresenius Study participants and design Kabi, Port Elizabeth, South Africa) to prevent clotting. Thereafter, The participants of this blood coagulation reactivity study were re- all participants underwent the laboratory stress. cruited as part of the Sympathetic Activity and Ambulatory Blood Pressure in Africans (SABPA) study conducted between February Acute laboratory stress testing 2008 and May 2009. The study protocol was conducted in accord- ance with the Declaration of Helsinki (23) and was approved by Mental stress test the Ethics Review Board of the North-West University, Potchef- stroom Campus (0003607S6). All participants provided written in- All participants underwent the Stroop Color-Word Conflict test formed consent before participation. The SABPA study has a tar- for 1 min. The Stroop is a standardised laboratory stressor show- get population comparative design and recruited 409 teachers, ing reproducibility on cardiovascular reactivity (30). The Stroop aged 25–65 years, working in the Dr Kenneth Kaunda Education requires identification of the ink colour of the word rather than the Thrombosis and Haemostasis 110.5/2013 © Schattauer 2013 286 Downloaded by: University of Free State. Copyrighted material. 287 von Känel et al. Procoagulant stress reactivity in Africans 979 name of the colour spelled by the word, under time pressure. The Blood lipids participants received a monetary motivation reward in line with performance. Perceived stress triggered by the Stroop was rated on Total cholesterol (T-C) and high-density lipoprotein cholesterol a Likert scale (1=”not at all stressful”, 7=”very stressful”). (HDL-C) levels were measured in serum with the Konelab 20i (Thermo Fisher Scientific, Vantaa, Finland). For statistical analy- sis, we computed the T-C/HDL-C ratio. Cardiovascular data collection With participants in a semi-recumbent position a registered nurse Haemoglobin A1c took a 5-min continuous measurement of resting cardiovascular parameters with the validated Finometer® device (Finapres Medi- HbA1c levels were determined by a turbidometric inhibition im- cal Systems, Amsterdam, the Netherlands) (31). The Finometer munoassay method from EDTA plasma (Integra 400, Roche, provides beat-by-beat BP and thus detects the full contour of the Basel, Switzerland). cardiovascular responses. The Finometer recorded the SBP, dias- tolic blood pressure (DBP) and heart rate (HR) and computed an Renal function integrated age dependent aortic flow curve from the surface area beneath the pressure/volume curve determining cardiac output We used the Modification of Diet in Renal Disease (MDRD) Study (CO), stroke volume (SV), and total peripheral resistance (TPR) equation to estimate glomerular filtration rate (eGFR) from creati- online (32); data were stored in the results files. Calculations of SV, nine levels, age, sex, and ethnicity (36). Creatinine was measured CO, and TPR yield reliable measures (33, 34). We stabilised BP to in serum using an enzymatic colorimetric test (Cobas Integra 400 resting state before participants underwent the Stroop. We ob- plus, Roche, Basel, Switzerland). tained beat-to-beat BP responses throughout mental stress and the 5 min recovery phase. For analyses, we used the average of the last Health behaviours 2 min of the resting recordings and the averages of the last 15 sec- onds (s) of the stressor recordings as well as of the recovery record- Smoking ings at 1, 3, and 5 min after the Stroop had ended. Cardiovascular reactivity was calculated as the area under the curve (AUC) with Participants who indicated that they currently smoked and/or had respect to increase from rest (35) across these five time points for smoked at least one cigarette per day during one year in the past SBP, DBP, HR, CO, SV, and TPR. were categorised as ever smokers. Blood collection Alcohol consumption Citrated blood samples were taken before the Stroop to determine Serum levels of γ-GT activity were used as a marker of alcohol resting levels of coagulation measures. At 10 min post-stress, the abuse (37) and measured with an enzymatic colorimetric assay infusion set was thoroughly flushed with 2-3 ml of saline, and the (Cobas Integra 400 plus). first 2 ml of blood were discarded before sampling was done for coagulation measures. Coagulation reactivity was calculated for Physical activity each participant as the percentage change from the resting value. We used the Actical® accelerometers (Montréal, Québec) to quan- tify and index physical activity with 3 (vigorous intensity), 2 (mod- Cardiometabolic risk factors erate intensity) or 1 (light intensity) (38). Body mass index Coagulation measures With participants in their underwear, we measured height and weight to the nearest 0.1 cm and 0.1 kg to calculate the body mass Blood processing index (BMI, kg/m2). Citrated blood samples were centrifuged at 1,500 rpm (2,000 g) for 15 min at room temperature. Citrated plasma samples were ali- Screening blood pressure quoted and frozen at -80°C until analysis. Fibrinogen and D-dimer After 5 min of rest, using a stethoscope and a mercury sphygmom- were determined by an accredited laboratory. anometer (auscultatory method), duplicate BP readings were taken 5 min apart and measurements averaged to obtain screening Von Willebrand factor SBP and DBP. Plasma VWF:Ag (%) were measured with a "sandwich" ELISA assay. A polyclonal rabbit anti-VWF antibody and a rabbit anti -VWF-HRP antibody (DAKO, Cape Town, South Africa) were © Schattauer 2013 Thrombosis and Haemostasis 110.5/2013 287 Downloaded by: University of Free State. Copyrighted material. 288 980 von Känel et al. Procoagulant stress reactivity in Africans used to form the assay. The 6th International Standard for VWF/ variate analysis of variance (MANOVA) and of covariance (MAN- FVIII was used to set the standard curve against which the COVA) to test for a significant association of ethnicity, PHQ-9 samples were measured (39). measures, GHQ-28 measures, and cardiovascular reactivity measures with coagulation measures. We only deemed results of individual coagulation measures to be significant if the Fibrinogen MAN(C)OVA test statistic was significant, as this procedure pro- Plasma fibrinogen levels (g/l) were determined using a viscosity- tects against inflated Type I errors due to multiple tests of (likely) based method (STA Compact, STAGO Diagnostic, Roche, As- correlated dependent variables. Covariates were selected a priori nieres, France). based on the literature about demographic, cardiometabolic and life style factors affecting circulating levels of coagulation measures (42-44). Effect sizes are expressed as partial eta squared (ηp2) or D-dimer partial correlation coefficients. Plasma D-dimer levels (ng/ml) were determined with an immuno- based method (STA Compact, STAGO). There were 47.6% and 39.7%, respectively, of D-dimer values at rest and after stress below Results the limit of detection (<220 ng/ml). Undetectable values were sub- Participant characteristics stituted using the maximum likelihood method that generates values from the lognormal distribution of uncensored data under ▶ Table 1 shows the demographic and health characteristics of the the assumption that censored values follow a lognormal distribu- 267 study participants per ethnic group. The percentage of men tion. was higher in Africans than in Caucasians. In terms of cardio- metabolic risk factors, Africans showed higher BMI, HbA1c, and screening BP – despite using more antihypertensives – than Cau- Psychological questionnaires casians, while Caucasians had a higher T-C/HDL-C ratio than Af- Depressive symptoms ricans. Of the 20 subjects who took antihypertensive medications, nine had thiazide diuretics, six had angiotensin-converting We used the 9-item Patient Health Questionnaire (PHQ-9) to enzyme (ACE) inhibitors, two had angiotensin receptor blockers measure the frequency of depressive symptoms during the prior (ARB), and three had combination therapy (i.e. thiazide diuretic two weeks corresponding to criteria in the Diagnostic and Statisti- plus ACE inhibitor or ARB). Regarding health behaviours, Afri- cal Manual of Mental Disorders, Fourth Edition (40). Each item is cans were physically less active and had higher γ-GT indicating rated on a 4-point scale (0=“not at all“, 1=“several days“, 2= “more more alcohol abuse than Caucasians. Africans were significantly than half the days”, 3=“nearly every day”) giving a possible global more depressed and they also perceived more distress than Cau- score of 0-27. PHQ-9 scores ≥10 have 88% sensitivity and specifi- casians. city for major depression (40). In our sample, Cronbach’s α was 0.82 for Africans and 0.84 for Caucasians for the total scale. Coagulation measures at rest Psychological distress Ethnicity With the 28-item General Health Questionnaire (GHQ-28), we Both MANOVA (F3,263=43.28, ηp 2=0.330, p<0.001) and MANCO- quantified perceived psychological distress in general related to VA (F3,252=36.31, ηp 2=0.302, p<0.001) showed a significant associ- symptoms of depression, anxiety, somatic complaints and social ation between ethnicity and resting levels of the three coagulation withdrawal over the past few weeks applying the binary scoring measures. ▶ Table 2 shows that compared with Caucasians, Afri- method. The global score provides a severity measure of total psy- cans had significantly higher resting levels of VWF, fibrinogen, chological distress. If exceeding the threshold of 4, subjects are and D-dimer in the unadjusted as well as in the adjusted analyses. classified as achieving "psychiatric caseness" (41). In our sample, Cronbach’s α for the total scale was 0.91 for Africans and 0.90 for Psychological questionnaires Caucasians. To test for an association of PHQ-9 and GHQ-28 measures with resting levels of coagulation measures, we reran the above MAN- Statistical analysis COVA with continuous and categorical PHQ-9 and GHQ-28 We used SPSS version 21.0 for Windows to analyse the data with measures as additional covariates (four separate models). Ethnicity significance level at p<0.05 (two-tailed). In case of a non-normal remained associated with coagulation measures in all of these distribution, dependent variables were log transformed; for clarity, models (all p-values <0.001). However, coagulation measures all data are given in original units. Chi-square test and Student’s showed no significant association with continuous PHQ-9 scores t-test with Levene's correction in the case of unequal variances (p=0.46), categorical PHQ-9 scores (p=0.071), continuous were applied for data comparison between groups. We used multi- GHQ-28 scores (p=0.59), and GHQ-28 caseness (p=0.47). Thrombosis and Haemostasis 110.5/2013 © Schattauer 2013 288 Downloaded by: University of Free State. Copyrighted material. 289 von Känel et al. Procoagulant stress reactivity in Africans 981 Interaction between ethnicity and psychological coagulation reactivity in MANOVA (F 23,263=13.73, ηp =0.135, questionnaires p<0.001) and MANCOVA (F3,251=16.86, ηp 2=0.168, p<0.001) tests. ▶ Table 1 shows that relative to Caucasians, Africans experienced Taking into account main effects of ethnicity and PHQ-9 and lower procoagulant reactivity with less of an increase in VWF GHQ-28 scores, the interaction terms between ethnicity and the reactivity and more of a decrease in fibrinogen reactivity; D-dimer continuous as well as categorical PHQ-9 and GHQ-28 scores were reactivity showed no significant difference between Africans and not significant (all p-values >0.10). Caucasians. Stress-induced changes across all participants Haemodynamic reactivity The mean level of perceived stress from the Stroop was 3.70 ± 1.64 MANCOVA showed a significant association between ethnicity (range 0-7). Mental stress provoked significant responses in coagu- and AUC measures of haemodynamic variables (F6,248=3.18, lation and haemodynamic measures. While VWF levels increased ηp2=0.072, p=0.005). Relative to Caucasians, Africans had lower from 74.9 ± 26.1% to 99.9 ± 40.1% (p<0.001), fibrinogen levels de- SBP AUC (29.3 ± 5.8 vs 45.1 ± 4.3, ηp2=0.016, p=0.043), and lower creased from 3.18 ± 0.69 g/l to 2.88 ± 0.68 g/l (p<0.001). D-dimer SV AUC (–32.3 ± 8.1 vs 5.4 ± 6.1, ηp2=0.049, p<0.001), but greater levels showed no significant change (320 ± 339 ng/ml vs 376 ± 503 TPR AUC (0.40 ± 0.27 vs –0.30 ± 0.20, ηp2=0.029, p=0.007). Eth- ng/ml, p=0.44). Moreover, SBP (134.4 ± 16.2 mmHg vs 154.2 ± nicity was not related to AUC measures of DBP (p=0.46), HR 20.8 mmHg, p<0.001), DBP (78.6 ± 9.3 mmHg vs 88.6 ± 11.4 (p=0.12), and CO (p>0.09). Ethnicity did not significantly interact mmHg, p<0.001), HR (66.8 ± 10.7 bpm vs 87.4 ± 16.6 bpm, with AUC measures of any cardiovascular parameter to determine p<0.001), and CO (6.59 ± 1.81 l/min vs 8.12 ± 2.26 l/min, p<0.001) coagulation reactivity (all p-values >0.18). all increased and SV (99.6 ± 23.9 ml vs 94.2 ± 23.1 ml, p<0.001) and TPR (1.02 ± 0.46 mmHg/ml/s vs 0.99 ± 0.51 mmHg/ml/s, Table 1: Characteristics of 267 study participants. p=0.013) both decreased. Variable Africans Caucasians P-value (n=102) (n=165) Association between haemodynamic and coagulation Gender (% men) 62.7 49.7 0.037 reactivity Age (years) 43.4 ± 8.2 44.1 ± 11.0 0.580 Reactivity measures of VWF:Ag, fibrinogen, and D-dimer were Body mass index (kg/m2) 29.9 ± 7.8 27.6 ± 5.6 0.008 not significantly correlated with each other (all p-values >0.25). Total cholesterol (mmol/l) 4.76 ± 1.17 5.50 ± 1.22 <0.001 Separate MANOVA models for AUC measures of haemodynamic variables showed that reactivity in DBP (F3,263=3.83, ηp 2=0.042, HDL-cholesterol (mmol/l) 1.15 ± 0.36 1.20 ± 0.41 0.339 p=0.010), SV (F =3.35, ηp2=0.037, p=0.020), and TPR Total cholesterol/HDL-cholesterol 4.51 ± 1.86 4.98 ± 1.60 0.0283,263 (F3,263=2.74, ηp 2=0.030, p=0.044) were significantly associated ratio with coagulation reactivity, whereas reactivity in SBP (p=0.44), HR Systolic blood pressure (mmHg) 134.4 ± 19.6 126.7 ± 13.5 <0.001 (p=0.83), and CO (p=0.52) were not. Correlation analysis on indi- Diastolic blood pressure (mmHg) 88.4 ± 12.6 83.1 ± 9.6 <0.001 vidual measures revealed that VWF reactivity increased with Hemoglobin A1c (%) 6.05 ± 1.12 5.48 ± 0.41 <0.001 greater SV reactivity (r=0.18, p=0.003), but decreased with greater reactivity in DBP (r=-0.18, p=0.003) and TPR (r=-0.14, p=0.022). Creatinine (µmol/l) 77.6 ± 13.5 74.2 ± 15.3 0.069 Fibrinogen and D-dimer reactivity were not significantly associ- Estimated GFR (ml/min/1.73 m2) 114.7 ± 29.3 112.0 ± 27.2 0.453 ated with DBP, SV, and TPR reactivity (all p-values >0.12). Ever smoker (%) 25.5 21.2 0.419 Gamma glutamyl transferase (U/l) 64.6 ± 75.8 26.9 ± 36.2 <0.001 Stress-induced changes and ethnicity Physical activity index 1.33 ± 0.57 1.62 ± 0.70 0.001 Perceived stress Antihypertensive drugs (%) 12.7 4.2 0.010 Patient Health Questionnaire-9 (scores) Africans (3.69 ± 1.87) and Caucasians (3.70 ± 1.49) perceived the Total depressive symptoms 9.40 ± 5.60 5.70 ± 4.73 <0.001 Stroop protocol as similarly stressful (p=0.94). Moderate depression (cut-off 46.1 17.6 <0.001 ≥10) (%) Coagulation reactivity General Health Questionnaire-28 (scores) Using MANOVA (p=0.19) and MANCOVA (p=0.62) tests, there Total psychological distress 8.50 ± 6.71 3.79 ± 4.80 <0.001 were no significant associations between ethnicity and post-stress Cases (cut-off >4) (%) 64.7 33.3 <0.001 levels of coagulation measures (see ▶ Table 2 for unadjusted and adjusted post-stress levels of VWF, fibrinogen, and D-dimer). Data are given as means ± SD or percentage values. P-value refers to group differences. GFR, glomerular filtration rate; HDL, high-density lipoprotein. However, there were significant association between ethnicity and © Schattauer 2013 Thrombosis and Haemostasis 110.5/2013 289 Downloaded by: University of Free State. Copyrighted material. 290 982 von Känel et al. Procoagulant stress reactivity in Africans Psychological questionnaires nogen reactivity in Africans (r=-0.187), but directly so in Cau- casians (r=0.064). After controlling for ethnicity MANCOVA tests showed no signifi- We performed six complementary univariate analyses of co- cant associations of continuous PHQ-9 scores (p=0.71), categori- variance to test whether sympathetic activity as indicated by hae- cal PHQ-9 scores (p=0.94), continuous GHQ-28 scores (p=0.18), modynamic reactivity might relate to the direct association be- and GHQ-28 caseness (p=0.91) with coagulation reactivity. How- tween continuous PHQ-9 scores and VWF reactivity in Africans ever, there were significant interactions of ethnicity with continu- vs Caucasians. However, all six three-way-interactions between ous PHQ-9 scores (F3,249=3.03, ηp 2=0.035, p=0.030) and categori- ethnicity, continuous PHQ-9 scores and AUC measures of any cal PHQ-9 scores (F3,249=3.80, ηp 2=0.044, p=0.011), but not with haemodynamic parameter were non-significant (all p-values continuous GHQ-28 scores (p=0.24) and GHQ-28 caseness ≥0.12). (p=0.62). Between-subject analysis revealed significant associations be- tween ethnicity and continuous PHQ-9 scores for reactivity of Discussion VWF (ηp2=0.016, p=0.044) and fibrinogen (ηp2=0.020, p=0.026), but not for D-dimer reactivity (p=0.54). Partial correlation coeffi- The main finding from our study is that Africans showed lower cients differed significantly between ethnic groups for VWF reac- procoagulant reactivity in response to standardised acute labora- tivity (p=0.029) and for fibrinogen reactivity (p=0.030). With tory mental stress when compared with Caucasians, controlling greater continuously scaled PHQ-9 scores, indicating greater se- for a range of potentially confounding variables. Relative to Cau- verity of depressive symptoms, VWF increased more in Africans casians, VWF levels had increased less and fibrinogen levels had (r=0.202) than in Caucasians (r=-0.074), whereas fibrinogen de- decreased more in Africans after stress relative to resting levels. creased more in Africans (r=-0.215) than in Caucasians (r=0.058). One explanation for the diminished increase in VWF in Africans Similarly, ethnicity interacted with categorical PHQ-9 scores in could be a “ceiling effect” as Africans started out with higher rest- determining VWF reactivity (ηp2=0.023, p=0.015) and fibrinogen ing levels of VWF than Caucasians. For instance, previous studies reactivity (ηp2=0.018, p=0.035), but not D-dimer reactivity in apparently healthy individuals reported stress-induced increases (p=0.28). ▶ Figure 1 illustrates these interactions. Partial corre- of VWF of 6%, 9%, and 30% from resting levels which were 91%, lation coefficients significantly differed between groups for VWF 98%, and 95%, respectively, in these studies (4, 45, 46). This reactivity (p=0.035) and for fibrinogen reactivity (p=0.047). amount of an increase in VWF is clearly lower than in our Cau- PHQ-9 scores ≥ 10 were directly correlated with VWF reactivity in casian subjects who showed an increase in VWF of 60% starting Africans (r=0.149), but inversely so in Caucasians (r=-0.119). out from average resting values of 64%, but similar to our African Moreover, PHQ-9 scores ≥ 10 were inversely correlated with fibri- subjects who showed an increase in VWF of 18% from average Table 2: Coagulation measurements per ethnic group at rest and after mental stress. Condition Adj. von Willebrand factor Fibrinogen D-dimer Africans Cau- ES P Africans Cau- ES P Africans Cau- ES P casians casians casians Rest – 93.2 ± 63.6 ± 0.330 <0.001 3.35 ± 3.07 ± 0.030 0.004 354 ± 298 ± 0.015 0.044 28.4% 16.5% 0.82 g/l 0.57 g/l 340 ng/ml 338 ng/ml + 94.2 ± 62.9 ± 0.291 <0.001 3.33 ± 3.09 ± 0.029 0.006 379 ± 283 ± 28 0.030 0.005 2.4% 1.8% 0.06 g/l 0.05 g/l 3 7 ng/ml ng/ml Post-stress – 101.8 ± 98.7 ± 0.010 0.101 2.95 ± 2.84 ± 0.006 0.208 398 ± 363 ± 0.004 0.302 34.1% 43.4% 0.71 g/l 0.65 g/l 515 ng/ml 496 ng/ml + 97.5 ± 101.4 ± <0.001 0.817 2.91 ± 2.86 ± 0.001 0.593 426 ± 345 ± 42 0.006 0.217 4.4% 3.3% 0.07 g/l 0.05 g/l 56 ng/ml ng/ml Reactivity – 18.1 ± 60.1 ± 0.110 <0.001 –11.2 ± –7.4 ± 0.022 0.016 64.5 ± 74.5 ± 0.002 0.499 53.6% 70.2% 11.6% 12.6% 241.9% 319.9% + 10.2 ± 65.0 ± 0.140 <0.001 –11.7 ± –7.2 ± 0.023 0.016 48.1 ± 84.7 ± 0.005 0.270 7.0% 5.2% 1.4% 1.0% 32.6% 24.6% Data are given as means ± SD for unadjusted values (–) and as means ± SEM for adjusted values (+) with p-values for group comparisons. A djustment was made for gender, age, body mass index, total cholesterol/high-density cholesterol ratio, screening systolic blood pressure, haemoglobin A1c, estimated glomerular filtration rate, smoking, gamma glutamyl transferase, physical activity, and antihypertensive drugs for resting values and additionally for perceived stress with the Stroop test for post-stress and reactivity values. Reactivity measures are expressed as percentage change of post-stress values from resting values. Adj., adjustment; ES, effect sizes (partial eta-squared). Thrombosis and Haemostasis 110.5/2013 © Schattauer 2013 290 Downloaded by: University of Free State. Copyrighted material. 291 von Känel et al. Procoagulant stress reactivity in Africans 983 resting levels of 93%. In contrast, a “floor effect” in Caucasians who started out on relatively lower resting fibrinogen levels might account for the greater decrease in post-stress fibrinogen levels relative to resting levels in Africans. Another explanation for the blunted VWF reactivity in Africans might be their alpha-adrener- gic vascular stress response, as was suggested by their increased TPR reactivity (22). Across all participants, lower VWF reactivity correlated with both greater TPR reactivity and lower SV reactiv- ity, the latter likely reflecting increased afterload due to elevated TPR (22). Moreover, in Caucasians, the systemic vasodilation re- sponse during the Stroop test is largely mediated by the beta2-ad- renergic receptor (47) whose stimulation through stress hormones will result in endothelial release of VWF into the circulation (19-21). A decrease in fibrinogen levels between rest and 10 min after the Stroop was seen across all study participants, and to an even greater extent in Africans than Caucasians. This is contrary to ex- pectations, as fibrinogen levels were shown to increase in response to acute mental stress in several previous studies (2-4). The Stroop protocol provoked an average level of psychological distress that A compares to similar protocols (10). However, previous studies sug- gest that the Stroop alone might be less effective in provoking a significant fibrinogen response (48) than combinations of the Stroop with other stressors (e.g. mental arithmetic) (45) or speech stressors (4). As there was no correlation of VWF reactivity with both fibrinogen and D-dimer reactivity in the present study, but in a previous one (49), it is also possible that the 1-min stressor was too short to evoke a significant fibrinogen increase with fibrin formation and degradation further downstream. Nevertheless, previous studies showed coagulation and fibrinolysis activation within 2-5 min of acute mental stress (50, 51) and infusion with the stress hormone epinephrine (52). D-dimer did not increase in a previous study that combined the Stroop with mental arithmetic (53), but in studies that applied speech stressors to inflict social evaluative threat (5, 9). The sensitivity of the various D-dimer as- says used in previous stress studies might also explain heterogen- eous results (54). Saying that, previous stress reactivity studies also used different methods to measure fibrinogen (e.g. the functional Clauss method) (4) and VWF (e.g. enzyme-linked immunosor- bent assays) (46). To our knowledge, a direct comparison of the sensitivity between different assays to detect reactivity in pro- B thrombotic measures has not been performed, but would seem Figure 1: Coagulation reactivity related to ethnicity and depression. important to reconcile heterogeneous study findings. Depicted are multivariate-adjusted interaction effects between ethnicity and Coagulation reactivity was specifically associated with depress- categorical depression (i.e. score ≥10 on the PSQ-9 total depressive symptom ive symptoms (i.e. PHQ-9 scores) as opposed to more general psy- scale) on von Willebrand factor (VWF) and fibrinogen reactivity (i.e. relative chological distress (i.e. GHQ-28 scores). More severe depressive change in levels of coagulation measures in % from resting levels to 10 min symptoms were associated with greater VWF reactivity in Africans after the Stroop test). Relative to non-depressed individuals (Dep –) of the re- compared with Caucasians. This concurs with the notion that de- spective ethnicity, VWF reactivity was significantly greater in Africans vs Cau- pressive mood along a continuum of severity is associated with an casians with depression (Dep +). Fibrinogen reactivity showed a significantly increased risk of incident CVD and recurrent cardiac events (55). greater decrease in depressed Africans than in depressed Caucasians (see text for detailed statistics). Africans/Dep + (n=47), Africans/Dep – (n=55), However, post-stress levels of VWF and fibrinogen were similar in Caucasians/Dep + (n=29), Caucasians/Dep – (n=136). Africans and Caucasians, fibrinogen decreased more in depressed Africans than depressed Caucasians, and VWF reactivity was, on the whole, lower in Africans than in Caucasians; therefore, a de- pression-associated increase in CVD risk in Africans through a © Schattauer 2013 Thrombosis and Haemostasis 110.5/2013 291 Downloaded by: University of Free State. Copyrighted material. 292 984 von Känel et al. Procoagulant stress reactivity in Africans pathway of stress-induced VWF increase would need to be shown BP and HR, but there is an ongoing debate as to whether cardio- in prospective studies. vascular reactivity data gained from standardised lab stressor are Ethnical differences in coagulation are still poorly understood generalisable to real-life situations (62). We controlled our analysis with studies showing enhanced clotting but also increased bleed- for antihypertensive medications as a group but due to insufficient ing tendency in various clinical settings in Africans compared with statistical power could not take into account potential class effects Caucasians (56). We found resting levels of VWF:Ag, fibrinogen, of the different antihypertensives on outcomes. We did not include and D-dimer to be higher in Africans than Caucasians, which is in a non-stress control group. Therefore we were unable to account agreement with some previous studies (57-59). Theoretically, this for circadian changes in coagulation measures, particularly a de- potential predisposition to hypercoagulability might contribute to crease in plasma fibrinogen levels during the morning hours. We the CVD risk in Africans. In contrast, lower increase in VWF lev- are unable to account for the possibility that stress-related changes els and greater decrease in fibrinogen levels both with acute men- in clearance impacted plasma concentrations of coagulation tal stress could be viewed a hypoactive fight-flight response (60). measures. For instance, glycosylation of VWF influences its clear- As opposed to constitutively released VWF, the VWF released ance by the liver, which, moreover, occurs more rapidly in individ- from the endothelium after stimulation consists of large and hae- uals with blood group O than in non-O individuals (63); unfortu- mostatically highly active multimers (61). Moreover, although not nately, information on ABO blood group was not available in our investigated in our study, acute stress also activates fibrinolysis (2, study. The difference in liver function between Africans and Cau- 45, 50). Therefore, the inability to mount physiologic hypercoagu- casians seems not an apparent explanation for the ethnic differ- lability might subject Africans to an increased risk of bleeding in ence in VWF reactivity because the relation between ethnicity and fight-flight situations either upon injury or even spontaneously. procoagulant reactivity persisted after adjustment for γ-GT activ- Therefore, the almost absent VWF response in the non-depressed ity. Africans might be as harmful as the increased VWF reactivity in Taken together, Africans seem to have lower procoagulant reac- the depressed Africans for the maintenance of the haemostatic bal- tivity with acute mental stress than Caucasians. This observation ance between thrombosis and haemorrhage during acute mental might partially be explained by ethnic differences in the vascular stress. stress response. Depressive symptoms may modulate coagulation Rigorous selection of participants on medical characteristics to reactivity against an ethnic background. The potential clinical im- minimise confounding of coagulation measures was a strength of plications of our findings for CVD risk and bleeding disorders in our study, but also might reduce generalisability of study findings Africans need further studies. to the larger African population and those with established CVD. We selected three coagulation measures that were previously Acknowledgements shown to be stress-responsive. However, they do not cover the en- The authors are grateful to Péter Szabolcs, Chrissie Lessing, and tire dynamics of the coagulation and fibrinolysis pathways during Tina Scholtz for their technical assistance and support. The study acute mental stress. A speech stressor inflicting social evaluative was partly funded by The Metabolic Syndrome Institute, France; threat and measuring coagulation measures longer into the re- the Medical Research Council, National Research Foundation, covery period from stress might have provided additional in- North-West University, and North-West Department of Edu- formative data. The Stroop test provoked substantial increases in cation, South Africa. Conflicts of interest None declared. What is known about this topic? • Acute mental stress induces a hypercoagulable state that may contribute to atherothrombotic diseases. References • Depressive symptoms are associated with the procoagulant stress 1. Borissoff JI, Spronk HM, ten Cate H. The haemostatic system as a modulator of response. atherosclerosis. N Engl J Med 2011; 364: 1746-1760. • There are ethnic differences in coagulation between Africans and 2. von Känel R, Mills PJ, Fainman C, et al. Effects of psychological stress and psy- Caucasians. chiatric disorders on blood coagulation and fibrinolysis: a biobehavioural path- way to coronary artery disease? Psychosom Med 2001; 63: 531-544. What does this paper add? 3. Thrall G, Lane D, Carroll D, et al. A systematic review of the effects of acute psy- chological stress and physical activity on haemorheology, coagulation, fibrinoly- • Africans show lower procoagulant reactivity in response to acute sis and platelet reactivity: Implications for the pathogenesis of acute coronary laboratory mental stress than Caucasians. syndromes. Thromb Res 2007; 120: 819-847. 4. von Känel R, Preckel D, Zgraggen L, et al. The effect of natural habituation on • Explanation might be higher resting levels of coagulation factors coagulation responses to acute mental stress and recovery in men. Thromb Hae- in Africans than Caucasians and ethnic differences in the vascular most 2004; 92: 1327-1335. stress response. 5. Wirtz PH, Ehlert U, Emini L, et al. Anticipatory cognitive stress appraisal and the acute procoagulant stress response in men. Psychosom Med 2006; 68: • Depressive symptoms are associated with greater stress reactivity 851-858. of von Willebrand factor in Africans relative to Caucasians. 6. Ellins E, Halcox J, Donald A, et al. Arterial stiffness and inflammatory response to psychophysiological stress. Brain Behav Immun 2008; 22: 941-948. Thrombosis and Haemostasis 110.5/2013 © Schattauer 2013 292 Downloaded by: University of Free State. Copyrighted material. 293 von Känel et al. Procoagulant stress reactivity in Africans 985 7. Brydon L, Steptoe A. Stress-induced increases in interleukin-6 and fibrinogen 30. Falaschi P, Proietti A, De Angelis C, et al. Effects of mental stress on cardiovas- predict ambulatory blood pressure at 3-year follow-up. J Hypertens 2005; 23: cular and endocrine response in Air Force Academy cadets. Neuro Endocrinol 1001-1007. Lett 2003; 24: 197-202. 8. Strike PC, Magid K, Brydon L, et al. Exaggerated platelet and haemodynamic 31. Imholz BP, Wieling W, van Montfrans GA, et al. Fifteen years experience with reactivity to mental stress in men with coronary artery disease. Psychosom Med finger arterial pressure monitoring: assessment of the technology. Cardiovasc 2004; 66: 492-500. Res 1998; 38: 605- 616. 9. von Känel R, Dimsdale JE, Ziegler MG, et al. Effect of acute psychological stress 32. Wesseling KH, Jansen LR, Settels JJ, Schreuder JJ. Computation of aortic flow on the hypercoagulable state in subjects (spousal caregivers of patients with Alz- from pressure in humans using a nonlinear, three-element model. J Appl Physiol heimer's disease with coronary or cerebrovascular disease and/or systemic hy- 1993; 74:2 566-2573. pertension. Am J Cardiol 2001; 87: 1405-1408. 33. Harms MP, Wesseling KH, Pott F, et al. Continuous stroke volume monitoring 10. Strike PC, Magid K, Whitehead DL, et al. Pathophysiological processes underly- by modelling flow from non-invasive measurement of arterial pressure in hu- ing emotional triggering of acute cardiac events. Proc Natl Acad Sci USA 2006; mans under orthostatic stress. Clin Sci 1999; 97: 291-301. 103: 4322-4327. 34. Jansen JR, Schreuder JJ, Mulier JP, et al. A comparison of cardiac output derived 11. Aschbacher K, Roepke SK, von Känel R, et al. Persistent versus transient de- from the arterial pressure wave against thermodilution in cardiac surgery pa- pressive s ymptoms in relation to platelet hyperactivation: a longitudinal analysis tients. Br J Anaesth 2001; 87: 212-222. of dementia caregivers. J Affect Disord 2009; 116: 80-87. 35. Pruessner JC, Kirschbaum C, Meinlschmid G, et al. Two formulas for compu- 12. von Känel R, Dimsdale JE, Adler KA, et al. Effects of depressive symptoms and tation of the area under the curve represent measures of total hormone concen- anxiety on haemostatic responses to acute mental stress and recovery in the tration versus time dependent change. Psychoneuroendocrinology 2003; 28: elderly. Psychiatry Res 2004; 126: 253-264. 916-931. 13. Lippi G, Montagnana M, Favaloro EJ, et al. Mental depression and cardiovascu- 36. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glo- lar disease: a multifaceted, bidirectional association. Semin Thromb Haemost merular filtration rate from serum creatinine: a new prediction equation. Modi- 2009; 35: 325-336. fication of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 14. Sliwa K, Wilkinson D, Hansen C, et al. Spectrum of heart disease and risk fac- 461-470. tors in a black urban population in South Africa (the Heart of Soweto Study): a 37. Sharpe PC. Biochemical detection and monitoring of alcohol abuse and absti- cohort study. Lancet 2008; 371: 915-922. nence. Ann Clin Biochem 2001; 38: 652-664. 15. Pieters M, de Maat MP, Jerling JC, et al. Fibrinogen concentration and its role in 38. Heil DP. Predicting activity energy expenditure using the Actical® Activity CVD risk in black South Africans-effect of urbanisation. Thromb Haemost Monitor. Res Q Exercise Sport 2006; 77: 64-80. 2011; 106: 448-456. 39. Meiring M, Badenhorst PN, Kelderman M. Performance and utility of a cost-ef- 16. Hamer M, Malan L, Schutte AE, et al. Conventional and behavioural risk factors fective collagen-binding assay for the laboratory diagnosis of Von Willebrand explain differences in sub-clinical vascular disease between black and Caucasian disease. Clin Chem L ab Med 2007; 45: 1068-1072. South Africans:t he SABPA study. Atherosclerosis 2011; 215: 237-242. 40. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression 17. Hamer M, Malan NT, Harvey BH, et al. Depressive symptoms and sub-clinical severity measure. J Gen Intern Med 2001; 16: 606-613. atherosclerosis in Africans: role of metabolic syndrome, inflammation and sym- 41. Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. pathoadrenal function. Physiol Behav 2011; 104: 744-748. Psychol Med 1979; 9: 139-145. 18. Preckel D, von Känel R. Regulation of haemostasis by the sympathetic nervous 42. Fibrinogen Studies Collaboration, Kaptoge S, White IR, Thompson SG, et al. system: any contribution to coronary artery disease? Heartdrug 2004; 4: Associations of plasma fibrinogen levels with established cardiovascular disease 123-130. risk factors, inflammatory markers, and other characteristics: individual partici- 19. von Känel R, Dimsdale JE. Effects of sympathetic activation by adrenergic infu- pant meta-analysis of 154,211 adults in 31 prospective studies: the fibrinogen sions on h aemostasis in vivo. Eur J Haematol 2000; 65: 357-369. studies collaboration.A m J Epidemiol 2007; 166: 867-879. 20. von Känel R, Dimsdale JE, Adler KA, et al. Effects of nonspecific beta-adrener- 43. Lee KW, Lip GY. Effects of lifestyle on haemostasis, fibrinolysis, and platelet gic stimulation and blockade on blood coagulation in hypertension. J Appl Phy- reactivity: a systematic review. Arch Intern Med 2003; 163: 2368-2392. siol 2003; 94:1 455-1459. 44. Dubin R, Cushman M, Folsom AR, et al. Kidney function and multiple haemos- 21. von Känel R, Kudielka BM, Helfricht S, et al. The effects of aspirin and nonselec- tatic markers: cross sectional associations in the multi-ethnic study of athero- tive beta blockade on the acute prothrombotic response to psychosocial stress in sclerosis. BMC Nephrol 2011; 12: 3. apparently healthy subjects. J Cardiovasc Pharmacol 2008; 51: 231-238. 45. Jern C, Eriksson E, Tengborn L, et al. Changes of plasma coagulation and fibri- 22. Huisman HW, Schutte AE, Schutte R, et al. Exploring the link between cardio- nolysis in response to mental stress. Thromb Haemost 1989; 62: 767-771. vascular reactivity and end-organ damage in African and Caucasian men: the 46. Hamer M, Gibson EL, Vuononvirta R, et al. Inflammatory and haemostatic re- SABPA study. Am J Hypertens 2013; 26: 68-75. sponses to repeated mental stress: individual stability and habituation over time. 23. World Medical Association. Declaration of Helsinki. Ethical principles of medi- Brain Behav Immun 2006; 20: 456-569. cal research involving human subjects. J Indian Med Assoc 2009; 107: 403-405. 47. Freyschuss U, Hjemdahl P, Juhlin-Dannfelt A, et al. Cardiovascular and sympa- 24. Lip GY, Lowe GD, Rumley A, et al. Increased markers of thrombogenesis in thoadrenal responses to mental stress: influence of beta-blockade. Am J Physiol chronic atrial fibrillation: effects of warfarin treatment. Br Heart J 1995; 73: 1988; 255: H1443-H1451. 527-533. 48. Muldoon MF, Herbert TB, Patterson SM, et al. Effects of acute psychological 25. Baker J, Ayenew W, Quick H, et al. High-density lipoprotein particles and stress on serum lipid levels, haemoconcentration, and blood viscosity. Arch In- markers of inflammation and thrombotic activity in patients with untreated tern Med 1995; 155: 615-620. HIV infection. J Infect Dis 2010; 201: 285-292. 49. Zgraggen L, Fischer JE, Mischler K, et al. Relationship between haemoconcen- 26. Brotman DJ, Girod JP, Posch A, et al. Effects of short-term glucocorticoids on tration and blood coagulation responses to acute mental stress. Thromb Res haemostatic factors in healthy volunteers. Thromb Res 2006; 118: 247-252. 2005; 115: 175-183. 27. Prasad RN, Koh SC, Viegas OA, et al. Effects on haemostasis after two-year use 50. Palermo A, Bertalero P, Pizza N, et al. Decreased fibrinolytic response to adre- of low dose combined oral contraceptives with gestodene or levonorgestrel. Clin nergic stimulation in hypertensive patients. J Hypertens Suppl 1989; 7: Appl ThrombH aemost 1999; 5: 60-70. S162-S163. 28. Gebara OC, Jimenez AH, McKenna C, et al. Stress-induced haemodynamic and 51. de Boer D, Ring C, Wood M, et al. Time course and mechanisms of mental haemostatic changes in patients with systemic hypertension: effect of verapamil. stress-induced changes and their recovery: hematocrit, colloid osmotic pressure, Clin Cardiol 1996; 19: 205-211. whole blood viscosity, coagulation times, and haemodynamic activity. Psycho- 29. Geiser F, Conrad R, Imbierowicz K, et al. Coagulation activation and fibrinoly- physiology 2007; 44: 639-649. sis impairment are reduced in patients with anxiety and depression when medi- 52. Cash JD, Lind AR, McNicol GW, et al. Fibrinolytic and forearm blood flow re- cated with serotonergic antidepressants. Psychiatry Clin Neurosci 2011; 65: sponses to intravenous adrenaline in healthy subjects. Lif Sci 1969; 8: 207-213. 518-525. © Schattauer 2013 Thrombosis and Haemostasis 110.5/2013 293 Downloaded by: University of Free State. Copyrighted material. 294 986 von Känel et al. Procoagulant stress reactivity in Africans 53. Steptoe A, Kunz-Ebrecht S, Rumley A, et al. Prolonged elevations in haemostatic 59. Sukhu K, Poovalingam V, Mahomed R, et al. Ethnic variation in von Willebrand and rheological responses following psychological stress in low socioeconomic factor levels can influence the diagnosis of von Willebrand disease. Clin Lab status men and women. Thromb Haemost 2003; 89: 83-90. Haematol 2003;2 5: 247-249. 54. Righini M, Perrier A, De Moerloose P, et al. D-Dimer for venous thromboem- 60. Lovallo WR. Do low levels of stress reactivity signal poor states of health? Biol bolism diagnosis: 20 years later. J Thromb Haemost 2008; 6: 1059-1071. Psychol 2011; 86: 121-128. 55. Frasure-Smith N, Lespérance F. Depression and cardiac risk: present status and 61. Sporn LA, Marder VJ, Wagner DD. Inducible secretion of large, biologically po- future directions. Heart 2010; 96: 173-176. tent von Willebrand factor multimers. Cell 1986; 46: 185-190. 56. Mayr FB, Spiel AO, Leitner JM, et al. Racial differences in endotoxin-induced 62. Kamarck TW, Lovallo WR. Cardiovascular reactivity to psychological challenge: tissue f actor-triggered coagulation. J Thromb Haemost 2009; 7: 634-640. conceptual and measurement considerations. Psychosom Med 2003; 65: 9-21. 57. Lutsey PL, Cushman M, Steffen LM, et al. Plasma haemostatic factors and en- 63. Lenting PJ, Van Schooten CJ, Denis CV. Clearance mechanisms of von Wille- dothelial markers in four racial/ethnic groups: the MESA study. J Thromb Hae- brand factor and factor VIII. J Thromb Haemost 2007; 7: 1753-1760. most 2006; 4: 2629-2 635. 58. Gader A, Bahakim H, Awadalla S, et al. Ethnic variations in the haemostatic sys- tem: comparison between Arabs, Westerners (Europeans and Americans), Asians andA fricans. Blood Coagul Fibrinolysis 1995; 6: 537-542. Thrombosis and Haemostasis 110.5/2013 © Schattauer 2013 294 Downloaded by: University of Free State. Copyrighted material. Meiring_A4_2011 25/05/2012 14:56 Page 89 295 Original Contribution S-33 HIV-associated Thrombotic Thrombocytopenic Purpura – What We Know So Far Murie l Meir ing, 1 Mike Webb, 2 Dominique Goedhals 3 and Vernon Louw 4 1. Associate Professor and Specialist Scientist, Department of Haematology and Cell Biology; 2. Senior Specialist, Division of Clinical Haematology, Department of Internal Medicine; 3. Pathologist and Senior Lecturer, Department of Medical Microbiology and Virology; 4. Associate Professor and Principal Specialist, Division of Clinical Haematology, Department of Internal Medicine, University of the Free State, Bloemfontein, South Africa Abstract Thrombotic thrombocytopenic purpura (TTP) is a life-threatening disease characterised by microvascular platelet deposition and thrombus formation in selected organs, resulting in microangiopathic haemolytic anaemia, thrombocytopenia, neurological symptoms and renal failure. Typically a very rare disorder, TTP is being seen with increased frequency in patients infected with the human immunodeficiency virus (HIV). Deficiency of the von Willebrand factor cleavage protease, ADAMTS13, has been implicated as the cause of TTP. However, the pathophysiology of HIV-associated TTP and the thrombotic potential in these patients are not known. This article provides not only an overview of the literature regarding HIV-associated TTP, but also presents new data on this disease. We propose a mechanism for the initial onset of HIV-associated TTP that includes the release of extreme amounts of von Willebrand factor and the downregulation of ADAMTS13 and/or the production of autoantibodies to ADAMTS13. Keywords Thrombotic thrombocytopenic purpura, HIV, ADAMTS13, anti-ADAMTS13 antibodies, von Willebrand factor Disclosure: The authors have no conflicts of interest to declare. Received: 5 March 2012 Accepted: 9 April 2012 Citation: European Oncology & Haematology, 2012;8(2):89–91 DOI: 10.17925/EOH.2012.08.02.89 Correspondence: Muriel Meiring, Associate Professor and Specialist Scientist, Department of Haematology and Cell Biology, University of the Free State, P.O. Box 339, Bloemfontein 9300, South Africa. E: GNHMSMM@ufs.ac.za Thrombotic thrombocytopenic purpura (TTP) is an acute prothrombotic of HIV-associated TTP will continue to rise.9 HIV-associated TTP cases disorder resulting from a deficiency of the von Willebrand factor were also described in other countries, such as the UK and Italy.10–12 cleavage protease ADAMTS13.1 The enzyme ADAMTS13 (a disintegrin and metalloprotease with thrombospondin type 1 repeats) cleaves the Features and Treatment of HIV-associated peptide bond between Tyr-842 and Met-843 of the mature subunit of Thrombotic Thrombocytopenic Purpura von Willebrand factor (VWF) and prevents the interactions of the HIV-associated TTP is a haematologic disorder observed in patients largest VWF multimers with platelets. In the plasma of patients with infected with HIV. As with congenital non-HIV-associated TTP, this TTP, ultra large VWF multimers (ULVWF) have been observed that syndrome is characterised by microangiopathic haemolytic anaemia, provoke widespread microvascular thrombosis.2 TTP is characterised thrombocytopenia, renal dysfunction, fluctuating neurological by microvascular platelet aggregation and thrombus formation abnormalities and fever. Treatment regimens include highly active resulting in thrombocytopenia, microangiopathic haemolytic anaemia, antiretroviral therapy (HAART), infusion of fresh frozen plasma, variable renal and neurologic dysfunction and fever.3 Infection with the plasma exchange, steroids and immunomodulatory agents. human immunodeficiency virus (HIV) is postulated as a direct precipitant of TTP, presumably through infection of vascular One case study stressed the importance of CD4 count and viral loads endothelial cells resulting in dysfunction, localised thrombin in HIV-associated TTP.11 The patient presented with a normal CD4 generation and consumption of ADAMTS13.4 HIV-associated TTP was count, a high viral load and a suboptimal response to plasma exchange first described in 1987 by Jokela et al.5 Since then, several case studies while remission was obtained with combined antiretroviral therapy have been reported. The occurrence of TTP in association with HIV (cART). Other case studies confirmed that there may be a causative infection is now well recognised.6,7 link between TTP and advanced HIV infection, since all patients presented with extremely low CD4 and CD8 counts.12 This is not always Prevalence of HIV-associated Thrombotic the case in our experience (unpublished data). The treatment strategy Thrombocytopenic Purpura in the case described by Miller et al.11 included fresh plasma and a Although congenital TTP is a very rare disease, the incidence of corticosteroid, but only after plasma exchange did the platelet count HIV-associated TTP is much higher. A group in South Africa estimated start to rise and the steroid was tapered; cART treatment was started the incidence of TTP in HIV-infected individuals to be 15–40 times that afterwards. Another case study described a patient with myocardial in non-infected individuals.8 Also, more than 80 % of TTP cases in South injury in HIV-associated TTP who responded poorly to plasma infusion Africa are found to be HIV-related and it is expected that the incidence and steroid therapy, but recovered fully following plasma exchange.9 © T O U C H B R I E F I N G S 2 0 1 2 295 89 Meiring_A4_2011 28/05/2012 16:07 Page 90 Origin29a6 l Contribution Table 1: ADAMTS13 Levels, Activities and Autoantibodies, VWF Levels and Tissue Factor Levels of Citrated Plasma from Patients with TTP, HIV-positive People on cART and HIV-positive People not on cART Test Unit TTP (Mean±SD) HIV+ on cART (Mean±SD) HIV+ not on cART (Mean±SD) Normal Range ADAMTS13 levels ng/ml 144±109 934±245 627±109 520–1,060 ADAMTS13 activity ng/ml 366±97 735±230 421±260 481–785 ADAMTS13 autoantibodies % of positive patients 50 50 83 none VWF antigen levels % 494±130 317±187 767±378 50–150 Tissue factor levels % of patients with 36 50 87 none increased levels ADAMTS13 = a disintegrin and metalloprotease with thrombospondin type 1 repeats; cART = combination antiretroviral therapy; SD = standard deviation; TTP = thrombotic thrombocytopenic purpura; VWF = von Willebrand factor. Figure 1: Proposed Mechanism for the Initial Onset of microcirculation. Only limited data are available regarding HIV-associated Thrombotic Thrombocytopenic Purpura the presence of autoantibodies in HIV-associated TTP patients. In one study, only 38 % of patients with reduced ADAMTS13 activity had HIV infection inhibitory antibodies. This was measured by using the residual collagen binding assay and performing mixing studies with normal plasma.14 The same group also found only a mild reduction of Increased cytokine Micronutrient, e.g., Production of autoantibodies ADAMTS13 levels in patients with HIV-associated TTP and an increase release zink deficiency to ADAMTS13 in VWF levels in all patients. Increased VWF synthesis Decreased ADAMTS13 Production of inhibitory We measured ADAMTS13 levels, activity and autoantibodies and release by synthesis and release by antibodies to endothelium endothelium ADAMTS13 in 40 patients diagnosed with HIV-associated TTP at the Haematology Clinic of the Universitas Hospital in Bloemfontein, South Africa. As a control group, we used the plasma of 104 Microvascular HIV-positive persons visiting the local primary healthcare clinic. The thrombosis control HIV group was divided into two groups, depending on whether they were on cART or not. The ADAMTS13 levels and Haemolytic anaemia autoantibodies were measured using the respective enzyme-linked and thrombocytopenia immunosorbent assay (ELISA) kits from American Diagnostica (US). The ADAMTS13 activities were measured using the Actifluor™ ADAMTS13 fluorescence resonance energy transfer (FRET) assay HIV-associated TTP (also from American Diagnostica). We also measured the VWF levels using antibodies from Dako (South Africa) and tissue factor levels in ADAMTS13 = a disintegrin and metalloprotease with thrombospondin type 1 repeats; TTP = thrombotic thrombocytopenic purpura; VWF = von Willebrand factor. plasma with an ELISA kit from American Diagnostica. The results are summarised in Table 1. A recent study of 24 patients with HIV-associated TTP emphasises the importance of the prompt initiation/re-initiation of cART in parallel We found that the ADAMTS13 levels extensively decreased in TTP with plasma exchange and steroid treatment that lead to prompt patients. HIV infection on its own did not seem to have a significant remission.10 Adjunct immunomodulatory agents such as rituximab influence on ADAMTS13 levels. However, HIV-positive persons who have been used with some success in refractory cases.10 It has been were not on cART seemed to have slightly lower ADAMTS13 levels than reported that patients with HIV-associated TTP seem to be far those who were on cART, although the levels were still in the normal more responsive to plasma infusion treatment regimens than their range. It is possible that less metalloproteases such as ADAMTS13 HIV-negative counterparts, suggesting a different pathophysiology.13 are synthesised in these patients, since micronutrient deficiencies are widespread and compound the effects of HIV disease.16 Our findings Pathophysiology of HIV-associated Thrombotic regarding ADAMTS13 levels differed from those of another group in Thrombocytopenic Purpura South Africa. They detect normal ADAMTS13 levels in one-third of adult The pathophysiology and initial onset of TTP in HIV-positive patients cases with HIV-associated TTP.8 The activities of ADAMTS13 that we are not fully understood. In 1994, Cruccu et al. suggested that the measured using the FRET assay did not correlate with the ADAMTS13 aetiology of thrombocytopenia in HIV-infected patients may be due to levels in all groups. However, the lowest activities were found in the TTP direct infection of megakaryocytes by the virus, immune-mediated patients. The assay might not be as suitable as the ADAMTS13 levels to destruction, impaired haematopoiesis, toxic effects from medications diagnose the presence of HIV-associated TTP. We also found that only and microangiopathic anaemia syndromes.12 It is now generally 50 % of our TTP patients presented with autoantibodies to ADAMTS13. accepted that HIV-associated TTP is an acquired form of ADAMTS13 This correlates with what is reported in the literature.14 An interesting deficiency. Acquired ADAMTS13 deficiency is often due to the finding was that most of the HIV-positive persons who were not presence of autoantibodies directed against ADAMTS13.14,15 These on cART also presented with high amounts of autoantibodies autoantibodies to ADAMTS13 play a pivotal role in the pathogenesis of against ADAMTS13. We thus suggest that antibody formation against acquired TTP. By decreasing the function of ADAMTS13, the ADAMTS13 is a result of the weakened immune system in HIV. autoantibodies impair the cleavage of ULVWF into smaller sizes, Similar to what is reported in the literature, we found increased VWF leading to the formation of platelet-VWF thrombi in the levels in patients with HIV-associated TTP.8 The VWF antigen levels were 90 E U R O P E A N O N C O L O G Y & H A2E9M6A T O L O G Y Meiring_A4_2011 25/05/2012 14:57 Page 91 297 HIV-associated Thrombotic Thrombocytopenic Purpura – What We Know So Far greatly increased in HIV-positive patients who were not on cART. This disease. A schistocyte count of 1 % or more in conjunction with could be due to an HIV-associated inflammatory state or concomitant thrombocytopenia suggests TTP if other causes of MAHA have been infection. Similar to the VWF levels, we found that most (87 %) of the excluded.20 We suggest the use of ADAMTS13 levels as a confirmatory HIV-positive persons who were not on cART had increased tissue factor test of HIV-associated TTP. The measurement of ADAMTS13 activity levels, which is known to be a risk factor for thrombosis. We found that levels using the FRET assay is expensive and does not necessarily only 36 % of the TTP patients presented with increased tissue factor confirm the disease. The exact role of inhibitory autoantibodies to levels. This is probably because these patients were all on cART. ADAMTS13 remains uncertain and further investigation is required. We propose a mechanism for the initial onset of HIV-associated TTP Conclusion (see Figure 1). HIV type 1 (HIV-1) infection is associated with HIV-associated TTP is a heterogeneous disorder and its initial onset is inflammation and elevated levels of inflammatory cytokines such as still not clear. We propose that infection with HIV might trigger the gamma interferon and tumour necrosis factor alpha (TNF-α) and disease through the inflammatory process. Inflammatory cytokines beta (TNF-β).17 It is known that inflammatory cytokines, such as TNF-α stimulate the release of extreme amounts of VWF.17 Extremely high and interleukin-1, 6 and 8, have profound stimulatory effects on levels of VWF were found in the plasma of HIV-positive persons who the endothelial release of ULVWF18 In addition, the synthesis of are not on HAART as well as in TTP plasma. The cytokines also ADAMTS13 that cleaves the ULVWF is inhibited.19 This might ultimately downregulate the release of the VWF-cleaving protease ADAMTS13.19 lead to the deficiency of ADAMTS-13 and the overexpression of We found very low levels of ADAMTS13 in TTP plasma. This might be ULVWF, resulting in the initiation of TTP. Furthermore, HIV-infection is sufficient to cause thrombotic microangiopathy and precipitate an associated with widespread micronutrient deficiencies that might acute episode of TTP even in the presence of normal cleaving cause decreased ADAMTS13 synthesis.16 Autoantibodies to protease activity. Furthermore, the release of VWF may be sufficient ADAMTS13 are also present in HIV-positive patients as a result of the to overwhelm the capacity of the cleaving protease and result in a weakened immune system. An inhibitory antibody might also cause ‘consumptive deficiency’, as speculated by Gunther et al.14 An the initial onset of HIV-associated TTP. important finding in HIV-associated TTP that differs from acquired TTP is the fact that not all patients present with autoantibodies to Laboratory Diagnosis of HIV-associated ADAMTS13. More specific methods might be needed to identify Thrombotic Thrombocytopenic Purpura autoantibodies, especially inhibitory autoantibodies, in these The diagnosis of TTP rests on evidence of microangiopathic patients. Inhibitory antibodies to ADAMTS13 might also trigger the haemolytic anaemia (MAHA) and thrombocytopenia in the absence disease. We could finally show that neither HIV-infection per se, nor of disseminated intravascular coagulation and other known causes of the use of HAART, has any effect on ADAMTS13 levels, making it a thrombotic microangiopathy. Highly specific diagnostic tools such as useful diagnostic tool to diagnose HIV-positive patients with TTP. The plasma levels of ADAMTS13 are recommended, but are not always increased tissue factor levels in HIV-positive persons who are not on routinely available for immediate clinical diagnosis. The presence of cART indicate the thrombotic potential in HIV that might also schistocytes on a blood smear is the morphologic hallmark of the contribute to the initial onset of HIV-associated TTP. n 1. Levy GG, Nichols WC, Lian EC, et al., Mutations in a member different? ISBT Science Series, 2006;1:246–50. distribution of anti-ADAMTS13 antibodies in patients of the ADAMTS gene family cause thrombocytopenic 9. Visagie GJ, Louw VJ, Myocardial injury in HIV-associated with acquired thrombotic thrombocytopenic purpura, purpura, Nature, 2001;413(6855):488–94. thrombotic thrombocytopenic purpura (TTP), Transfus Med, J Thromb Haemost, 2009;7(10):1703–10. 2. Moake JL, Rudy CK, Troll JH, et al., Unsually large plasma 2010;20(4):258–64. 16. Kaiser JD, Campa AM, Ondercin JP, et al., Micronutrient factor VIII: von Willebrand factor multimers in chronic 10. Hart D, Sayer R, Miller R, et al., Human immunodeficiency supplementation increases CD4 count in HIV-infected relapsing thrombotic thrombocytopenic purpura, virus associated thrombotic thrombocytopenic purpura – individuals on higly active antiretroviral therapy: a N Engl J Med, 1982;307(23):1432–5. favourable outcome with plasma exchange and prompt prospective, double-blinded, placebo-controlled trial, 3. Moake JL, Thrombotic microangiopaties, N Engl J Med, initiation of highly active antiretroviral therapy, J Acquir Immune Defic Syndr, 2006;42(5):523–8. 2002;347(8):589–600. Brit J Haematol, 2011;153(4):515–9. 17. Jassoy C, Harrer T, Rosenthal T, et al., Human 4. Brecher ME, Hay SN, Park YA, Is it HIV TTP or 11. Miller R, Scully M, Cohen H, et al., Thrombotic immunodeficiency virus type-1 specific cytotoxic T HIV-associated thrombotic microangiopathy? thrombocytopaenic purpura in HIV-infected patients, lymphocytes release gamma interferon, tumour necrosis J Clin Apher, 2008;23(6):186–90. Int J STD AIDS, 2005;16(8):538–42. factor alpha (TNF-alpha), and TNF-beta when they 5. Jokela J, Flynn T, Henry K, Thrombotic thrombocytopenic 12. Cruccu V, Parisio E, Pedretti D, et al., HIV-related encounter their target antigens, J Virol, 1993;67(5):2844–52. purpura in a human immunodefediency virus thrombotic thrombocytopenic purpura (TTP) as first clinical 18. Paleolog EM, Crossman DC, McVey JH, Pearson JD, (HIV)-seropositive homosexual man, Am J Hematol, manifestation of infection, Haematologica, 1994;79(3):277–9. Differential regulation by cytokines of constitutive and 1987;25(3):341–3. 13. Novitzky N, Thomson J, Abrahams L, et al., Thrombotic stimulated secretion of von Willebrand factor from 6. Hymes KB, Karpatkin S, Human immunodeficiency virus thrombocytopenic purpura in patients with retroviral endothelial cells, Blood, 1990;75(3):688–95. infection and thrombotic microangiopathy, Semin Hematol, infection is highly responsive to plasma infusion therapy, 19. Bernardo A, Ball C, Nolaso L, et al., Effects of inflammatory 1997;34(2):117–25. Br J Haematol, 2005;128(3):373–9. cytokines on the release and cleavage of the endothelial 7. Blazes DL, Decker CF, Thrombotic thrombocytopenic 14. Gunther K, Garizio D, Nesara P, ADAMTS13 activity and the cell-derived ultralarge von Willebrand factor multimers purpura in HIV-infected patients, Infect Dis Clin Pract, presence of acquired inhibitors in human immunodeficiency under flow, Blood, 2004;104(1):100–6. 2004;12(2):99–106. virus-related thrombotic thrombocytopenic purpura, 20. Burns ER, Lou Y, Pathak A, Morphologic diagnosis of 8. Gunther K, Garizio D, Dhlamini B, The pathogenesis of Transfusion, 2007;47(9):1710–6. thrombotic thrombocytopenic purpura, Am J Hemat, HIV-related thrombotic thrombocytopaenic purpura – is it 15. Ferrari S, Mudde GC, Rieger M, et al., IgG subclass 2004;75(1):18–21. E U R O P E A N O N C O L O G Y & H A E M A T O L O G Y 297 91 S-34 298 Transfusion and Apheresis Science 58 (2019) 72–78 Contents lists available at ScienceDirect Transfusion and Apheresis Science journal homepage: www.elsevier.com/locate/transci Comparison of ADAMTS13 and Von Willebrand factor levels and activities, T and plasminogen levels, in plasma products currently available for the treatment of thrombotic thrombocytopenic purpura in South Africa A.C. van Marlea,b, J. Jouberta,b,⁎, S.M. Meiringa,b a Department of Haematology and Cell Biology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa bNational Health Laboratory Service (NHLS), Universitas Hospital Haematology Academic Laboratories, Bloemfontein, South Africa A R T I C L E I N F O A B S T R A C T Keywords: Objective: Thrombotic thrombocytopenic purpura (TTP) results from a deficiency in the Von Willebrand factor Thrombotic thrombocytopenic purpura (VWF) cleaving protease, ADAMTS13. Treatment involves plasma exchange (PEX) therapy with either fresh Von Willebrand factor frozen plasma (FFP), cryosupernatant (CSP) or solvent/detergent-treated plasma (SDP), available in South Africa ADAMTS13 as Bioplasma FDP. The aim of the study was to generate in vitro data on these products, and to explore possible Plasminogen differences between the products that may offer treatment advantages. Fresh frozen plasma Cryosupernatant Methods: Twenty samples per product (FFP, CSP and Bioplasma FDP) were analysed for levels and activities of Solvent/detergent-treated plasma ADAMTS13 and VWF. Plasminogen levels, a proposed physiological back-up system for ADAMTS13, were also determined. FFP and CSP samples were subanalysed according to ABO blood group. Samples were analysed by means of commercially available ELISA assays. Results: All samples had normal/high ADAMTS13 activity (Median activity for SDP=94.0%, CSP=80.5%, FFP=122.0%) and plasminogen levels. The VWF content was mostly normal for Bioplasma FDP, typically deficient for CSP and mostly deficient for FFP, which was an unexpected finding. Depending on the parameter, Bioplasma FDP was the most standardised, with coefficients of variation (CV) from 14.1% to 27.3%, while FFP showed great inter-individual variation (CV 24.6% to 208.6%). Statistically significant differences were found across products (P≤ 0.0095), and ABO blood groups (P = 0.0001). Conclusion: All three products can be used for the treatment of TTP. The choice of product depends on the need for additional viral safety, costs, product availability and the perceived impact of within-product variations. 1. Introduction identified as a possible back-up mechanism for ADAMTS13, and the potential therapeutic importance of fibrinolysis was recently demon- Thrombotic thrombocytopenic purpura (TTP) is a life-threatening strated in a murine model of acquired TTP [4]. condition that belongs to a diverse group of disorders, collectively Acute idiopathic TTP is the most common form of TTP. It is char- known as thrombotic microangiopathy (TMA) syndromes [1]. Despite acterised by autoantibodies, usually IgG, directed against ADAMTS13 their diversity, these disorders are all characterised by widespread [2]. HIV-associated TTP differs from acute idiopathic TTP in that au- microvascular occlusion in several organs [1]. toantibodies to ADAMTS13 are often not detected [5]. The incidence of Central to the pathophysiology of TTP is a deficiency of ADAMTS13 TTP in HIV-positive individuals is considered to be up to 40 times (a disintegrin and metalloprotease with thrombospondin Type I repeats, greater than that in the general population [6]. As a country with one of 13th member), the plasma protease that functions to cleave ultra-large the world’s highest HIV prevalence rates, HIV-associated TTP is of Von Willebrand factor (ULVWF) multimers into smaller, less adhesive particular importance in South Africa [7]. Although the association forms [2]. Unusually large VWF multimers accumulate and result in between HIV and TTP has been evident since the 1980s, the exact pa- abnormal platelet aggregation and thrombus formation, with sub- thogenesis remains unclear and is considered to be a multifactorial sequent thrombocytopenia and an associated fragmentation haemolysis process [6]. Gunther et al. [8] and Meiring et al. [5] summarised the [3]. Plasmin, the major enzyme of the fibrinolytic system, has been suspected pathogenesis of HIV-associated TTP to be a combination of ⁎ Corresponding author at: Department of Haematology and Cell Biology (G2), Faculty of Health Sciences, University of the Free State, 205 Nelson Mandela Drive, Bloemfontein 9300, South Africa. E-mail address: joubertj@ufs.ac.za (J. Joubert). https://doi.org/10.1016/j.transci.2018.11.005 Received 10 July 2018; Received in revised form 12 November 2018; Accepted 26 November 2018 1473-0502/ © 2018 Elsevier Ltd. All rights reserved. 298 299 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 endothelial cell damage with loss of its inherent anticoagulant prop- Von Willebrand factor testing included two ELISA-based assays erties, excessive VWF release, and reduction in ADAMTS13 activity, to determine VWF antigen (VWF:Ag) level and collagen-binding whether it be secondary to consumption, inhibition or proteolytic in- (VWB:CB) activity. These assays were performed as previously de- activation [8]. scribed by Meiring et al. [24]. The ristocetin co-factor assay Daily plasma exchange (PEX) or, alternatively, large volume plasma (VWF:RCo), based on ristocetin-induced platelet agglutination, was infusion, forms the mainstay of therapy in TTP [2,9] and has reduced done on the PAP-8E aggregometer (Bio/Data Corporation; Horsham, mortality rates by more than 70% [2]. Proposed explanations for the PA, USA). Multimeric analysis of VWF involved electrophoretic se- efficacy of PEX include removal or dilution of inhibitor autoantibodies paration of VWF multimers on a 0.65% agarose gel, followed by rabbit and ULVWF multimers, together with replacement of the deficient anti-human VWF/horseradish peroxidase binding, and finally lumino- ADAMTS13 [10]. graphic visualisation with the aid of x-ray film and cassettes [24]. Plasma products used for this purpose in South Africa include fresh Densitograms of the multimer patterns were plotted using a gel doc- frozen plasma (FFP), cryosupernatant (CSP) and solvent/detergent- umentation system (Syngene G Box, Syngene; Frederick, MD USA). treated plasma (SDP), locally available as Bioplasma FDP [2]. Many international laboratory and clinical studies have compared and as- 2.3. Data analysis sessed the content and in vivo effectiveness of the different products [10–20]. The general conclusion drawn from these comparative studies Data were analysed by the Department of Biostatistics, University of was that despite documented differences among the different plasma the Free State. The results of the various tests were described using products, no one product has been proven superior in the treatment of ranges, medians, means, standard deviations (SDs), coefficients of TTP [11]. variation (CVs) and analysis of variance (ANOVA). No data have been published on the ADAMTS13, VWF or plasmi- For each plasma product, the various measurements – ADAMTS13 nogen content of the products locally available in South Africa. In ad- antigen, ADAMTS13 activity, Glu-plasminogen, VWF:Ag, VWF:RCo, dition, no clinical trials comparing the efficacy of these plasma products VWF:CB and VWF multimers – were reported as proportions below/ in the setting of HIV-associated TTP have been conducted. Considering within/above the recommended reference ranges supplied with the the lack of empirical data on the composition of products available package inserts [21–23]. These proportions of below-/within-/above- locally, and keeping in mind that South Africa has a different disease range readings were then compared across the various plasma products profile with most of its TTP cases being associated with HIV [6,7], an by means of chi-square cross tabulation with Fisher’s exact P-value. The ideal treatment regimen is yet to be defined. coefficients of variance (CVs) were calculated and compared to avail- The aim of this study was to generate empirical in vitro data on the able biological variation database values [25–28]. products available locally (FFP, Bioplasma FDP and CSP) to confirm their in vitro equivalence and utility for local use, and also to explore 2.4. Ethical considerations any possible differences in the products that may offer treatment ad- vantages in the setting of HIV-associated TTP. This study specifically Approval to conduct the research was obtained from the Ethics compared the levels and activities of the plasma proteins and enzymes Committees of both the Faculty of Health Sciences, University of the relevant in the treatment of HIV-associated TTP, namely ADAMTS13, Free State (ECUFS NR 191/2015) and the South African National Blood VWF and plasminogen. Service (SANBS) (2015/25). 2. Materials and methods 3. Results 2.1. Selection and grouping of samples The basic descriptive statistics of the comparison between the dif- ferent plasma products are shown in Table 1 and Fig. 1. Table 2 sum- Testing-segments of the FFP and CSP units issued in December marises the comparison of the proportions of below-/within-/above- 2015, still frozen at below −18 °C, were collected from the Universitas range readings across the various plasma products for the different Academic Hospital blood bank. In order to maintain the cold-chain, the parameters tested and Table 3 shows the P-values of these comparisons. samples were subsequently stored at −80 °C in the Department of Fig. 1 illustrates the deficiency in ADAMTS13 antigen levels Haematology and Cell Biology until batch testing commenced in (Fig. 1a) contrasted to the high ADAMTS13 activity levels (Fig. 1b) for January 2016. FFP. CSP showed deficiency in VWF antigen and activity levels The segments were subdivided according to the different ABO blood (Fig. 1c–e). FFP, which still contained the cryoprecipitate fraction with groups. Five samples each of blood groups A and B, and 10 blood group VWF, showed deficiency of VWF activities (Fig. 1d and e). The dis- O samples were included for each of the blood products, giving a total tribution of Glu-plasminogen is illustrated in Fig. 1f. of 10 Group O and 10 non-group O samples per product. Due to cost Table 2 shows that Bioplasma FDP samples were nearly consistently constraints and limited availability, samples from group AB products within the reference ranges for most of the parameters tested. CSP had were not included in the study. The SDP tested in the study was predominantly normal ADAMTS13 levels and activities, with an ab- Bioplasma FDP, a product of the National Bioproducts Institute (NBI). sence of VWF levels and activities. FFP samples showed variation, while The NBI provided 40 200mL bottles of Bioplasma FDP from separate normal Glu-plasminogen levels were observed in all three plasma pro- lots for analysis, which included two bottles per lot from 20 different ducts. Table 3 demonstrates statistically significant differences in at lots. least one plasma product for almost every parameter tested, particu- larly when Bioplasma FDP was compared with each of the other two 2.2. Sample testing plasma products. The proportions of below-/within-/above-range readings of the The ADAMTS13 antigen levels and activities and plasminogen levels blood group O and non-O CSP and FFP samples for the various para- were measured using the applicable commercially available ELISA kits meters tested, are summarised in Table 4. Table 5 shows the P-values of (Technozym ADAMTS13 Antigen, Technozym ADAMST13 Activity and the blood group comparisons for the different parameters. When all the Technozym Glu-Plasminogen; Technoclone; Vienna, Austria) according results for a particular parameter fell within the same category (below/ to the manufacturer’s instructions [21–23]. within/above the reference range), no P-values were calculated for the 73 299 300 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 Table 1 Comparison of the minimum (min) and maximum (max) values, medians, means, lower and upper quartiles (LQ and UQ), standard deviations (SD) and coefficient of variance (CV%) for the levels and activities of ADAMTS13 and VWF, and the antigen levels of Glu-plasminogen in the different plasma products. Variable Normal range [21–23] Min LQ Median UQ Max Mean SD CV% Bioplasma FDP (n=20) ADAMTS13 Antigen 0.6–1.6 μg/mL 1.13 1.58 1.98 2.70 2.70 2.07 0.56 27.34 ADAMTS13 Activity 40–130 % 62.00 82.50 94.00 104.00 127.00 92.95 15.67 16.86 VWF:Ag 51–143 % 54.00 76.50 87.50 95.50 122.00 86.15 16.21 18.81 VWF:RCo 50–150 % 39.00 54.50 58.00 68.00 81.00 61.00 10.21 16.74 VWF:CB 49–157 % 38.00 57.50 75.00 83.00 100.00 71.30 18.82 26.39 Glu-plasminogen 60–250 μg/mL 92.00 98.00 113.00 122.00 143.00 122.45 15.85 14.10 CSP (n=20) ADAMTS13 Antigen 0.6–1.6 μg/mL 0.20 0.31 0.84 1.33 1.73 0.86 0.51 59.69 ADAMTS13 Activity 40–130 % 60.00 72.50 80.50 97.00 142.00 87.05 22.56 25.92 VWF:Ag 51–143 % 0.00 0.00 0.00 0.00 15.00 1.60 3.80 237.69 VWF:RCo 50–150 % 0.00 0.00 0.00 0.00 10.00 0.50 2.24 447.21 VWF:CB 49–157 % 0.00 0.00 0.00 0.00 17.00 1.00 3.80 379.75 Glu-plasminogen 60–250 μg/mL 71.00 91.50 100.50 138.50 194.00 113.35 32.34 28.53 FFP (n=20) ADAMTS13 Antigen 0.6–1.6 μg/mL 0.23 0.29 0.32 0.46 1.56 0.49 0.39 78.96 ADAMTS13 Activity 40–130 % 63.00 100.50 122.00 160.00 160.00 124.70 33.29 26.69 VWF:Ag 51–143 % 22.00 45.00 55.50 70.00 95.00 57.05 20.59 36.09 VWF:RCo 50–150 % 0.00 0.00 0.00 0.00 75.00 12.85 26.80 208.58 VWF:CB 49–157 % 0.00 0.00 0.00 0.00 84.00 13.35 27.78 208.09 Glu-plasminogen 60–250 μg/mL 73.00 87.50 95.00 117.50 163.00 105.70 26.01 24.61 CSP= cryosupernatant; FFP= fresh frozen plasma; VWF:Ag=Von Willebrand factor antigen; VWF:RCo=VWF ristocetin co-factor; VWF:CB=VWF col- lagenbinding activity. Fig. 1. The analysis of variance (ANOVA) procedure illustrating the distribution of the different test parameters for each plasma product. The normal reference range [21–23] for each test is indicated below the box plot. 74 300 301 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 Table 2 Number of below-/within-/above-range readings of the different plasma products for the various parameters tested. Plasma product Range reading Parameters tested ADAMTS13 Ag ADAMTS13 Act VWF:Ag VWF:RCo VWF:CB Glu-plasminogen Bioplasma FDP (n=20) Below 0 0 0 2 4 0 Within 5 20 20 18 16 20 Above 15 0 0 0 0 0 CSP (n=20) Below 7 0 20 20 20 0 Within 12 18 0 0 0 20 Above 1 2 0 0 0 0 FFP (n=20) Below 16 0 9 17 16 0 Within 4 11 11 3 4 20 Above 0 9 0 0 0 0 Ag= antigen; Act= activity; VWF:Ag=Von Willebrand factor antigen; VWF:RCo=VWF ristocetin co-factor; VWF:CB=VWF collagen-binding activity; CSP= cryosupernatant; FFP= fresh frozen plasma. data. No significant differences between blood group O and non-O CSP and ADAMTS13 [29–34]. The adequate levels and activities of samples were found. Although the differences in VWF:RCo and VWF:CB ADAMTS13 and normal levels of plasminogen found in this study levels between blood group O and non-O FFP samples were not statis- (Table 2) confirm that Bioplasma FDP is on par with international SDP, tically significant, it is worth noting that the four non-O samples with and would provide consistent replacement of these important enzymes normal values were all of blood group B. if used as replacement fluid. The VWF multimer patterns were Fig. 2 shows the VWF multimer patterns of the Bioplasma FDP equivalent to that of normal plasma controls, showing absence of only samples of which 18 (90.0%) yielded results that were comparable to the largest forms (Table 4; Fig. 2). Bearing in mind that only the ultra- normal plasma controls. The VWF multimer patterns of the CSP samples large forms are implicated in the pathogenesis of TTP [3], presence of are represented in Fig. 3. Three (15.0%) CSP samples contained inter- the large multimers would not be regarded as detrimental. mediate multimers, while in the remainder, only small multimers and a When considering the large volume of plasma required for the complete absence of the larger forms were observed. The VWF mul- treatment of an acute episode of TTP, the additional benefit of viral timer patterns of FFP samples, shown in Fig. 4, demonstrated great inactivation provided by the solvent/detergent process may be con- inter-individual variation across the different samples. The patterns that sidered essential. In fact, both the British Committee for Standards in were comparable to normal plasma all occurred in blood group B Haematology (BCSH) guidelines for the use of FFP, cryoprecipitate samples. and CSP [35], and South African clinical guidelines for the use of blood products [36], suggest the use of pathogen inactivated plasma in pa- tients likely to receive large or repeated doses of plasma. 4. Discussion CSP is FFP from which the precipitable cryoproteins (Factor VIII, VWF, Factor XIII, fibronectin and fibrinogen) have been removed [35]. The rationale behind the use of plasma therapy as the cornerstone in As expected, in 17 of the 20 samples tested, VWF multimer patterns the management of TTP, includes the removal or dilution of inhibitor revealed complete absence of large VWF multimers and intermediate autoantibodies and ULVWF multimers, and replacement of the deficient forms (Table 4; Fig. 4). Although these results could be anticipated, CSP ADAMTS13 [1]. Although the exact pathogenesis of HIV associated TTP should still contain ADAMTS13 [35]. Several studies confirmed the is still unclear, excessive VWF release and reduction in ADAMTS13 preservation of ADAMTS13, not only in CSP, but also FFP and SDP activity, whether secondary to consumption, inhibition, or proteolytic [10,20,31]. Seven of the CSP samples tested had deficient ADAMTS13 inactivation, are evident [8]. Consequently, the ideal plasma product to antigen levels, yet normal enzyme activity. This was also the case with use in PEX/plasma-infusion therapy would be one with normal/high 16 of the FFP samples tested. ADAMTS13 activity levels actually ex- ADAMTS13 levels and activity, and absent/low levels of the large and ceeded the upper limit of the assay's detection range in six of these 16 ULVWF multimers. In addition, adequate levels of plasminogen, the FFP samples (Fig. 1a and b). A possible explanation could be that partial circulating zymogen of plasmin that is a proposed back-up mechanism degradation of the enzyme may result in loss of the specific epitope for ADAMTS13, can also be considered beneficial [4]. required to bind to the anti-ADAMTS13 monoclonal antibody in the One of the major benefits of pooling plasma products is standardi- antigen assay without impairing enzyme function. The importance of sation of the content [29]. Not surprisingly, Bioplasma FDP, which is this finding is uncertain, as activity assays may have more clinical derived from pooled fresh human plasma, showed the least variation for significance than antigen assays [37]. all the parameters tested, with CVs ranging from 14.1% to 27.3% When analysing all 20 FFP samples as a single cohort, significant (Table 1). variation was found across the different FFP samples with regard to all Several studies have confirmed that SDP retains normal in vitro the parameters tested. To a certain extent, these variations may be levels and activities of selected plasma proteins, including plasminogen Table 3 Fisher’s exact P-values for the comparisons across the different plasma products for the various parameters tested. Comparison ADAMTS13 Ag ADAMTS13 Act VWF:Ag VWF:RCo VWF:CB Glu-plasminogen VWF multimer patterns Bioplasma FDP vs < 0.0001* 0.0012* < 0.0012* < 0.0001* 0.0004* 0.1085 <0.0001* FFP Bioplasma FDP vs < 0.0001* 0.4872 <0.0001* < 0.0001* <0.0001* 0.1085 <0.0001* CSP FFP vs CSP 0.0095* 0.0310* 0.0001* 0.2308 0.1060 0.1085 0.2308 *Statistically significant difference, P < 0.05; Ag = antigen; Act = activity; VWF:Ag = Von Willebrand factor antigen; VWF:RCo = VWF ristocetin co-factor; VWF:CB = VWF collagen-binding activity; CSP = cryosupernatant; FFP = fresh frozen plasma. 75 301 302 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 Table 4 Proportions of below-/within-/above-range readings of blood groups O and non-O CSP and FFP samples for the various parameters tested. Plasma product Blood group Range reading Parameters tested ADAMTS13 Ag ADAMTS13 Act VWF:Ag VWF:RCo VWF:CB Glu-plasminogen CSP O Below 1 0 10 10 10 0 Within 8 9 0 0 0 10 Above 1 1 0 0 0 0 Non-O Below 6 0 10 10 10 0 Within 4 9 0 0 0 10 Above 0 1 0 0 0 0 FFP O Below 10 0 10 10 10 0 Within 0 1 0 0 0 10 Above 0 9 0 0 0 0 Non-O Below 6 0 4 7 6 0 Within 4 10 6 3 4 10 Above 0 0 0 0 0 0 Ag= antigen; Act= activity; VWF:Ag=Von Willebrand factor antigen; VWF:RCo=VWF ristocetin co-factor; VWF:CB=VWF collagen-binding activity; CSP= cryosupernatant; FFP= fresh frozen plasma. Table 5 Fisher’s exact P-values for the comparison between blood groups O and non-O FFP and CSP samples for the various parameters tested. Comparison ADAMTS13 Ag ADAMTS13 Act VWF:Ag VWF:RCo VWF:CB Glu-plasminogen CSP: O vs non-O 0.0573 1.000 ND ND ND ND FFP: O vs non-O 0.0867 0.0001* 1.000 0.0867 0.2105 ND *Statistically significant difference, P < 0.05; ND = not determined; Ag = antigen; Act = activity; VWF:Ag = Von Willebrand factor antigen; VWF:RCo=VWF ristocetin co-factor; VWF:CB=VWF collagen-binding activity; CSP= cryosupernatant; FFP= fresh frozen plasma. Fig. 2. VWF multimer patterns of all 20 Bioplasma FDP samples. The arrows heads (▼) indicate normal plasma controls. Except for two samples, indicated by arrows (↓), all Bioplasma FDP samples had VWF multimer patterns comparable to normal plasma con- trols. NP=normal plasma. Fig. 3. VWF multimer patterns for CSP samples 1–20. The arrows heads (▼) indicate normal plasma controls. Only three samples, indicated with arrows (↓), still contained intermediate multimers, the remaining samples showed complete absence of the larger forms, with only small VWF multimers left. A= blood group A; B=blood group B; O=blood group O; NP=normal plasma. Fig. 4. VWF multimer patterns of all 20 FFP samples. The arrows heads (▼) indicate normal plasma controls. Note the great inter- individual variation across the different samples: some patterns (all from ABO blood group B), indicated by arrows (↓), are com- parable to that of the normal plasma controls, showing absence of only the largest multimer forms, while others show absence of the large and intermediate forms. A= blood group A; B= blood group B; O=blood group O; NP=normal plasma. attributed to the different ABO blood groups. However, with inter- unexpected, considering that FFP is prepared from single-donor whole- individual variations evident even among samples of the same blood blood units following application of physical separation methods [36]. group category (O versus non-O), additional factors should be con- Variations across the different FFP samples could therefore reflect sidered. biological variations in the donors. However, the between-subject This higher inter-individual variation was not completely variation (CVG) for almost all the parameters tested in this study still 76 302 303 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 Table 6 very limited supply and average storage times almost never exceed a Between-subject biological variation (CVG) of the different parameters tested in three month period. fresh frozen plasma (FFP) samples, compared to previously published results. Quality monitoring of standard FFP in the United Kingdom also Between-subject biological variation (CV ) entails measurement of factor VIII levels which need to be at least 0.7G IU/mL in at least 75% of units, immediately after being thawed [42]. Test Blood group Previously published Coagulation factor content of locally available plasma products is also findings tested for internal quality control purposes. According to the package All (n=20) Non-O O (n= 10) (n=10) insert, Bioplasma FDP® contains a minimum of 0.4 IU/mL of each coagulation factor [43]. Average coagulation factor levels for FFP are ADAMTS13 Act 26.69 8.19 20.50 9.63 available on the Western Province Blood Transfusion Service (WPBTS) (CVTG: 9.63) [27] website and are also published in the clinical guidelines for the use of ADAMTS13 Ag 78.96 22.30 7.82 6.28 (CV : 10.70) [27] blood products in South Africa, available from the South African Na-TG VWF:Ag 36.09 29.77 39.29 31.71 [25] tional Blood Service (SANBS) and WPBTS [36]. Average factor VIII 27.30 [26] levels are reported as 0.85 IU/mL [36]. As per the Standards of Practice 22.59 (CVTG: 22.87) for Blood Transfusion in South Africa, 7th ed.(2016), random sampling [27] VWF:RCo 208.58 131.93 ND 18.5 and quality control testing of blood products are routinely performed (CV : 19.83) [27] according to local establishments’ quality manuals [44]. The focus onTG VWF:CB 208.09 131.52 ND ND ADAMTS13, VWF and plasminogen levels in our study, was not to as- Glu-plasminogen 24.61 22.74 14.89 10.5 [28] certain quality in general, but rather to assess suitability of the different products for the treatment of TTP in South Africa. ND=not determined. exceeded that reported in the literature (Table 6) [25–28]. 6. Conclusions Unlike CSP, the cryoprecipitate fraction, rich in high molecular weight VWF multimers, is still included in the units of FFP. Yarranton All 60 samples had normal/high ADAMTS13 activity and normal et al. [10] confirmed that the whole spectrum of VWF multimers of all Glu-plasminogen levels. With a few exceptions, Bioplasma FDP sample weights are present in FFP. Therefore, the most surprising result in our results were consistently within the normal reference ranges for all the study was the absence of high, and even intermediate, molecular weight parameters tested. As expected, CSP had deficient/absent VWF levels multimers found in 85% (17/20) of all the FFP samples tested. We and activities, with absence of the large and intermediate multimer speculate that to an extent, these variations might be attributed to, forms. No significant difference between blood group O and non-O CSP among others, pre-analytical variables, including handling, transport samples were found. and storage of blood products after donation, or specimen handling, Surprisingly, the majority of FFP samples also showed reduced/ transport or storage. absent VWF activities with abnormal multimer patterns. Those FFP Finally, the choice of blood product is also determined by cost. samples with near-normal VWF activities and multimer patterns were Table 7 summarises the cost per unit and PEX of the different blood all of blood group B, while the FFP samples with above-normal products [38,39] at the time of finalising this report ADAMTS13 activities were all representative of blood group O. These (July 2018). On average, even when fewer units are required, FFP is findings suggest that non-O FFP may be more beneficial in the treat- still less economical at R14 944.90 (US$1 117.36) per exchange, ment of bleeding diatheses, whereas group O FFP may be preferable in compared to R12 228.96 (US$913.89) and R10 559.20 (US$790.42) per the treatment of TTP – a hypothesis that requires further investigation. exchange for Bioplasma FDP and CSP, respectively [40]. Theoretically, all three plasma products meet the minimum criteria for a replacement fluid, namely that it contributes to improved VWF 5. Limitations cleavage, whether secondary to replacement of ADAMTS13 or its pur- ported back-up mechanism, plasminogen, without adding additional Cost constraints limited the number of samples tested per studied pathogenic ULVWF multimers. One can therefore conclude that any one plasma product. As a result, the small sample sizes, and in particular, of these plasma products would be equally appropriate for the treat- the small numbers representing the different ABO blood groups, reduce ment of HIV-associated TTP. The choice of product would depend on the likelihood of statistically significant results and necessitate cautious factors such as the need for additional viral safety, financial ex- interpretation. penditure, product availability and the perceived practical impact of Plasma products in South Africa are stored at below −18 °C for up within-product variations. to one year [36]. The 2018 BCSH guidelines recommend storage at ≤ −25 °C for up to 36 months to preserve the activity of labile coagula- tion factors. (A) The International Society of Blood Transfusion (ISBT) Funding considers storage temperatures between −18 °C and −25 °C acceptable for up to three months [41]. Differences in South African practice may This research did not receive any specific grant from funding be explained by the fact that plasma products are in high demand with agencies in the public, commercial, or not-for-profit sectors. Table 7 Price per unit and estimated price per exchange (calculated for a 60 kg patient with an estimated plasma dose of 40mL/kg) for the different plasma products according to the 2018 South African public sector price lists [38,39]. Product Unit volume 2018 price including value-added tax Estimated price per PEX Bioplasma FDP® 200mL R1 019.08 (US$76.14)* R12 228.96 (US$913.89) FFP 240–300 mL R1 494.49 (US$111.74) R14 944.90 (US$1 117.36) CSP 190–310 mL R1 055.92 (US$79.03) R10 559.20 (US$790.42) *Calculation based on ZAR (South African rand) versus US$ (United States dollar) exchange rate of approximately R13.38 per dollar on 10 July 2018 [40]; PEX=plasma exchange; FFP= fresh frozen plasma; CSP= cryosupernatant. 77 303 304 A.C. van Marle et al. Transfusion and Apheresis Science 58 (2019) 72–78 Competing interests purpura? No answer yet. Br J Haematol 2005;129(1):79–86. [20] Yarranton H, Lawrie AS, Mackie IJ, Pinkoski L, Corash L, Machin SJ. Coagulation The authors have no conflict of interest to declare. factor levels in cryosupernatant prepared from plasma treated with amotosalenhydrochloride (S-59) and ultraviolet A light. Transfusion 2005;45(9):1453–8. [21] Technozym®. ADAMTS13 antigen ELISA package insert. Vienna: Technoclone; 2012 Acknowledgements https://diapharma.com › … › ADAMTS-13 › Technozym® ADAMTS-13 Antigen (accessed 3 July 2018). [22] Technozym®. ADAMTS13 activity ELISA package insert. Vienna: Technoclone; 2011 SANBS and NBI provided all the blood products used in this research https://diapharma.com › … › ADAMTS-13 › Technozym® ADAMTS-13 Activity project, free of charge, and provided technical and scientific advice. All (Accessed 3 July 2018). laboratory preparation and analyses were performed by the staff of the [23] Technozym®. Glu-Plasminogen ELISA package insert. Vienna: Technoclone; 2014https://diapharma.com › … › Plasminogen › Technozym® Glu-Plasminogen ELISA Specialised Haemostasis Laboratory in the Department of Haematology (Accessed 3 July 2018). and Cell Biology, Faculty of Health Sciences, University of the Free [24] Meiring M, Coetzee M, Kelderman M, Badenhorst P. Laboratory diagnosis and State and National Health Laboratory Service (NHLS), in Bloemfontein, management of von Willebrand disease in South Africa. Semin Thromb Hemost 2011;37(5):576–80. South Africa. Statistical analysis of data was performed by the [25] Shou W, Chen Q, Wu W, Cui W. Biological variations of lupus anticoagulant, an- Department of Biostatistics, Faculty of Health Sciences, University of tithrombin, protein C, protein S, and von Willebrand factor assays. Semin Thromb the Free State; and Dr. Daleen Struwig assisted with the technical and Hemost 2016;42(1):87–92. [26] Minchinela J, Ricós C, Perich C, Fernández-Calle P, Alvarez V, Domenech M, et al. editorial preparation of the manuscript. Desirable specifications for total error, imprecision, and bias, derived from intra- and inter-individual biologic variation. Annex I, Part I: within-subject and be- References tweensubject CV values of analytes and Desirable Analytical Quality Specifications for imprecision, bias and total error. 8th edition Westgard QC; 2014 (Accessed 4 July 2018). https://www.westgard.com/biodatabase1.htm. [1] George JN, Nester CM. Syndromes of thrombotic microangiopathy. N Engl J Med [27] Kilercik M, Coskun A, Serteser M, Inan D, Unsal I. Biological variations of 2014;371(7):654–66. ADAMTS13 and von Willebrand factor in human adults. Biochem Med (Zagreb) [2] Scully M, Hunt BJ, Benjamin S, Liesner R, Rose P, Peyvandi F, et al. Guidelines on 2014;24(1):138–45. the diagnosis and management of thrombotic thrombocytopenic purpura and other [28] Banfi G, Del Fabbro M. Biological variation in tests of hemostasis. Semin Thromb thrombotic microangiopathies. Br J Haematol 2012;158(3):323–35. Hemost 2008;34(7):635–41. [3] Moake JL. Thrombotic microangiopathies. N Engl J Med 2002;347(8):589–600. [29] Solheim BG, Chetty R, Flesland O. Indications for use and cost-effectiveness of [4] Tersteeg C, De Maat S, De Meyer SF, Smeets MWJ, Barendrecht AD, Roest M, et al. pathogen-reduced ABO-universal plasma. Curr Opin Hematol 2008;15(6):612–7. Plasmin cleavage of von Willebrand factor as an emergency bypass for ADAMTS13 [30] Chapanduka ZC, Fernandes-Costa FJ, Rochat C, Blyth DF. Comparative safety and deficiency in thrombotic microangiopathy. Circulation 2014;129(12):1320–31. efficacy of Bioplasma FDP versus single-donor fresh-dried plasma in cardio- [5] Meiring M, Webb M, Goedhals D, Louw V. HIV-associated thrombotic thrombocy- pulmonary bypass patients. S Afr Med J 2002;92(5):356–7. topenic purpura – what we know so far. Eur Oncol Haematol 2012;8(2):89–91. [31] Scott EA, Puca KE, Pietz BC, Duchateau BK, Friedman KD. Comparison and stability [6] Gunther K, Garizio D, Dhlamini B. The pathogenesis of HIV-related thrombotic of ADAMTS13 activity in therapeutic plasma products. Transfusion thrombocytopaenic purpura – is it different? ISBT Sci Series 2006;1(1):246–50. 2007;47(1):120–5. [7] Opie J. Haematological complications of HIV infection. S Afr Med J [32] Leebeek FW, Schipperus MR, van Vliet HH. Coagulation factor levels in solvent/ 2012;102(6):465–8. detergent-treated plasma. Transfusion 1999;39(10):1150–1. [8] Gunther K, Garizio D, Nesara P. ADAMTS13 activity and the presence of acquired [33] Beeck H, Hellstern P. In vitro characterization of solvent/detergent-treated human inhibitors in human immunodeficiency virus-related thrombotic thrombocytopenic plasma and of quarantine fresh frozen plasma. Vox Sang 1998;74(Suppl 1):219–23. purpura. Transfusion 2007;47(9):1710–6. [34] Liumbruno GM, Franchini M. Solvent/detergent plasma: pharmaceutical char- [9] George JN. How I treat patients with thrombotic thrombocytopenic purpura: 2010. acteristics and clinical experience. J Thromb Thrombolysis 2015;39(1):118–28. Blood 2010;116(20):4060–9. [35] O’Shaughnessy DF, Atterbury C, Bolton Maggs P, Murphy M, Thomas D, Yates S, [10] Yarranton H, Lawrie AS, Purdy G, Mackie IJ, Machin SJ. Comparison of von et al. Guidelines for the use of fresh-frozen plasma, cryoprecipitate and cryosu- Willebrand factor antigen, von Willebrand factor-cleaving protease and protein S in pernatant. Br J Haematol 2004;126(1):11–28. blood components used for treatment of thrombotic thrombocytopenic purpura. [36] South African National Blood Service (SANBS). Clinical guidelines for the use of Transfus Med 2004;14(1):39–44. blood products in South Africa. 5 ed Cape Town: SANBS and WP Blood Transfusion [11] Brunskill SJ, Tusold A, Benjamin S, Stanworth SJ, Murphy MF. A systematic review Service; 2014 https://www.wpblood.org.za/village/…/clinical_guidelines_5th of randomized controlled trials for plasma exchange in the treatment of thrombotic %20Edition_2014.pdf (Accessed 25 June 2018). thrombocytopenic purpura. Transfus Med 2007;17(1):17–35. [37] Starke R, Machin S, Scully M, Purdy G, Mackie I. The clinical utility of ADAMTS13 [12] Rock GA, Shumak KH, Buskard NA, Blanchette VS, Kelton JG, Nair RC, et al. activity, antigen and autoantibody assays in thrombotic thrombocytopenic purpura. Comparison of plasma exchange with plasma infusion in the treatment of throm- Br J Haematol 2007;136(4):649–55. botic thrombocytopenic purpura. Canadian Apheresis Study Group. N Engl J Med [38] Department of Health, Republic of South Africa. Contract Number HP10-2016BIO: 1991;325(6):393–7. Supply and delivery of biological preparations to the Department of Health for the [13] Harrison CN, Lawrie AS, Iqbal A, Hunter A, Machin SJ. Plasma exchange with period 01 October 2016 to 30 September 2018. Addendun 15: National Bioproducts solvent/detergent-treated plasma of resistant thrombotic thrombocytopenic pur- Institute – price amendment. Information provided per email (cpa@health.gov.za) pura. Br J Haematol 1996;94(4):756–8. on 6 July 2018. [14] Moake JL, Byrnes JJ, Troll JH, Rudy CK, Hong SL, Weinstein MJ, et al. Effects of [39] South African National Blood Service (SANBS). State patients pricelist. 1 April 2018 fresh-frozen plasma and its cryosupernatant fraction on von Willebrand factor – 31 March 2019. https://sanbs.org.za/wpcontent/uploads/2018/…/SANBS_State_ multimeric forms in chronic relapsing thrombotic thrombocytopenic purpura. Patients_Price_List.pdf (Accessed 4 July 2018). Blood 1985;65(5):1232–6. [40] XE Currency Converter. https://www.xe.com/currencyconverter/convert/? [15] Horowitz B, Bonomo R, Prince AM, Chin SN, Brotman B, Shulman RW. Solvent/ Amount=1&From=USD&To=ZAR (Accessed 10 July 2018). detergent-treated plasma: a virus-inactivated substitute for fresh frozen plasma. [41] Hardwick J. Section 12: blood storage and transportation. In Armstrong B (editor). Blood 1992;79(3):826–31. Introduction to blood transfusion technology. Wiley Online Library. ISBT Sci Series [16] Owens MR, Sweeney JD, Tahhan RH, Fortkolt P. Influence of type of exchange fluid 2008;3(2):177–97. on survival in therapeutic apheresis for thrombotic thrombocytopenic purpura. J [42] Green L, Bolton-Maggs P, Beattie C, Cardigan R, Kallis Y, Stanworth SJ. British Clin Apher 1995;10(4):178–82. Society of Haematology Guidelines on the spectrum of fresh frozen plasma and [17] Rock G, Shumak KH, Sutton DM, Buskard NA, Nair RC. Cryosupernatant as re- cryoprecipitate products: their handling and use in various patient groups in the placement fluid for plasma exchange in thrombotic thrombocytopenic purpura. absence of major bleeding. Br J Haematol 2018;181(1):54–67. Members of the Canadian Apheresis Group. Br J Haematol 1996;94(2):383–6. [43] Bioplasma FDP package insert. National Blood Products Institute; 2015. http:// [18] Zeigler ZR, Shadduck RK, Gryn JF, Rintels PB, George JN, Besa EC, et al. home.intekom.com/pharm/nbi/bioplasl.html (Accessed 12 November 2018). Cryoprecipitate poor plasma does not improve early response in primary adult [44] South African National Blood Services (SANBS). Standards for practice of blood thrombotic thrombocytopenic purpura (TTP). J Clin Apher 2001;16(1):19–22. transfusion in South Africa. www.satiba.org.za/documents/resources/standards- [19] Rock G, Anderson D, Clark W, Leblond P, Palmer D, Sternbach M, et al. Does for-blood-transfusion.pdf (Accessed 12 November 2018). cryosupernatant plasma improve outcome in thrombotic thrombocytopenic 78 304 305 Research High levels of von Willebrand factor and low levels of its cleaving protease, S-35 ADAMTS13, are associated with stroke in young HIV-infected patients Sameera Allie1, Alan Stanley1*, Alan Bryer1, Muriel Meiring2, and Marc I. Combrinck1,3 Background Stroke associated with human immunodeficiency Introduction virus infection may occur through a variety of mechanisms. Von Willebrand factor is a marker of endothelial dysfunction, and is elevated in human immunodeficiency virus infection. Stroke in the setting of HIV infection may occur through a High levels of von Willebrand factor, a protein involved in number of mechanisms. These include HIV-associated vascu- platelet adhesion and aggregation, and low levels of lopathy, opportunistic infections, cardio-embolism, and disorders ADAMTS13, a metalloproteinase that cleaves von Willebrand of hemostasis (1–3). factor, have been associated with an increased risk of Von Willebrand factor (VWF) and its cleaving protein, thrombosis. Aim To investigate the role of von Willebrand factor and ADAMTS13, are involved in hemostasis. VWF is vital for platelet ADAMTS13 in the pathogenesis of human immunodeficiency adhesion and initiates platelet plug formation at sites of vascular virus-related stroke in young patients. injury. It is secreted by platelets and damaged endothelial cells (4). Methods A case-control study (n = 100) comprising three par- ADAMTS13 is synthesized by the liver. It cleaves the ultra large ticipant groups: human immunodeficiency virus-positive anti- VWF multimers into smaller multimers. Uncleaved large multi- retroviral therapy-naïve young strokes (n = 20), human immunodeficiency virus-negative young strokes (n = 40), and mers may promote active thrombosis if present in excess (5,6). human immunodeficiency virus-positive antiretroviral High VWF and low ADAMTS13 levels may therefore be associ- therapy-naïve nonstroke controls (n = 40). von Willebrand ated with an increased risk for thrombosis and stroke (7,8). Low factor and ADAMTS13 levels were measured in plasma activity of ADAMTS13 due to blocking antibodies has been samples collected five- to seven-days poststroke. described in HIV-associated thrombotic thrombocytopenia Results Human immunodeficiency virus-positive stroke par- ticipants had higher von Willebrand factor levels than human purpura (TTP) (9,10). immunodeficiency virus-negative strokes (173·5% vs. 135%, Little is known about the role of these proteins in HIV-related P = 0·032). They tended to have higher levels of von Wille- stroke. We therefore sought to test the hypothesis that HIV- brand factor than human immunodeficiency virus-positive associated strokes would be related to elevated VWF and low nonstroke controls (173·5% vs. 129%, P = 0·061). Human ADAMTS13 levels. immunodeficiency virus-positive stroke participants had lower levels of ADAMTS13 than human immunodeficiency virus- positive nonstroke controls (0% vs. 23·5% P = 0·018) most Patients and methods likely due to the effect of the acute stroke. However, in the nonstroke group, these levels were significantly reduced com- Twenty consecutive ART-naïve HIV-positive (HIV+ve) ischemic pared with population norms. von Willebrand factor levels in stroke patients and 40 HIV-negative (HIV−ve) ischemic stroke all human immunodeficiency virus-positive participants were patients, matched for age, were enrolled from the Groote Schuur negatively correlated with CD4 counts. Conclusions Stroke in human immunodeficiency virus infec- young stroke cohort. Acute ischemic stroke participants in the tion is associated with a prothrombotic state, characterized by Groote Schuur young stroke cohort were recruited between elevated von Willebrand factor and low ADAMTS13 levels. August 2010 and August 2012 from the stroke unit and acute Key words: HIV infection, stroke, von Willebrand factor medical admissions of four hospitals in Cape Town, South Africa. Forty healthy HIV+ve controls matched for age were enrolled from a primary health care HIV clinic in the same community. All Correspondence: Alan Stanley*, E8 Division of Neurology, Groote stroke participants were investigated and treated according to Schuur Hospital, Cape Town 7925, South Africa. standard clinical protocols. The study was approved by the GSH/ E-mail: alanmstanley@yahoo.co.uk University of Cape Town Human Research Ethics Committee 1Division of Neurology, Department of Medicine, University of Cape (HREC ref: 178/2010), and informed consent was obtained from Town and Groote Schuur Hospital, Cape Town, South Africa 2Department of Haematology and Cell Biology, University of Free State, all participants. Bloemfontein, South Africa VWF and ADAMTS13 levels were measured in plasma samples 3Division of Geriatric Medicine, Department of Medicine, University of collected five- to seven-days after stroke onset in cases, or on the Cape Town and Groote Schuur Hospital, Cape Town, South Africa day of enrolment of controls. The timing of collection was stan- Received: 4 December 2014; Accepted: 8 April 2015; Published online 29 dardized to minimize fluctuations during the acute phase of June 2015 stroke. VWF and ADAMTS13 levels were assessed using an Conflict of interest: None declared. antigen-based enzyme-linked immunosorbent assay (ELISA) in the Haemostasis Laboratory at the University of the Free State, Funding: The study was funded from a South African National Research South Africa. VWF levels were measured using antibodies from Foundation (NRF) grant awarded to M. C. Dako (Glostrup, Denmark). VWF levels were calculated from a DOI: 10.1111/ijs.12550 standard curve using the WHO 6th FVII/VWF standard. 1294 Vol 10, December 2015, 1294–1296 © 2015 World Stroke Organi3z0at5ion 306 S. Allie et al. Research ADAMTS13 levels were measured using an Imubind ADAMTS13 was not significantly different from that of HIV+ve controls ELISA kit from American Diagnostica (Stamford, CT, USA), (P = 0·058) (Table 1). according to the manufacturer’s instructions. We used a normal No stroke etiology could be identified in nine (45%) of the range for both tests of 50–150%. CD4 T-lymphocyte counts HIV+ve and 17 (43%) of the HIV−ve stroke participants. Cardio- (CD4) were obtained for all HIV+ve participants. embolic disease, neurosyphilis, and hypertensive small vessel Mann–Whitney U tests were used for between-group compari- disease were identified as etiological factors in the remaining sons. Spearman’s Rank Order correlation was used to relate VWF cases. Six of the HIV+ve strokes were polymerase chain reaction levels to CD4 counts. (PCR) positive for varicella zoster (VZV) in the cerebrospinal fluid, but none had clinical features of encephalitis or meningitis. HIV+ve strokes had elevated VWF levels compared with Results HIV−ve strokes (P = 0·032). There was a trend to higher levels of VWF in HIV+ve strokes compared with HIV+ve nonstroke con- All three participant groups were comparable with respect to age trols (P = 0·061; see Fig. 1b). ADAMTS13 levels were low in both and gender. The median age of participants was 35 years (IQR 30 HIV+ve and HIV−ve stroke groups and were not significantly – 38). The median CD4 count of the HIV+ve stroke participants different between the two. ADAMTS13 levels were lower in the Table 1 Results HIV+ stroke HIV− stroke HIV+ controls HIV+ stroke vs. HIV− stroke HIV+ stroke vs. HIV+ controls n 20 40 40 20 vs. 40 20 vs. 40 Age (median) 35 36·5 33·5 P = 0·16 P = 0·86 Age IQR 29·5–36·8 30·5–41 29·3–38 Female (%) 60 52·5 55 P = 0·78 P = 0·93 Black African n (%) 16 (80) 24 (60) 40 (100) P = 0·21 P = 0·017 VWF median (%) 173·5 135 129 U = 263 U = 280·5 VWF IQR (%) 118·3–215·5 107·5–168 94·8–175 P = 0·032 P = 0·061 r = 0·28 r = 0·24 ADAMTS13 median (%) 0 0 23·5 P = 0·71 U = 262 ADAMTS13 IQR (%) 0–1·5 0–6 0–240·8 P = 0·018 r = 0·31 n 20 40 39 20 vs. 40 20 vs. 39 CD4 count median (cells/mm3) 234 NA 383 NA U = 271·5 CD4 count IQR (cells/mm3) 100·8–381·3 205–520 P = 0·058 r = 0·25 Fig. 1 (a) VWF levels and CD4 count correlation in all HIV-positive participants: VWF levels in HIV+ve participants were negatively correlated with CD4 count [Spearman’s Rank Order correlation rs(59) = 0·36, P = 0·006]. (b) Intergroup VWF levels: HIV+ve stroke participants had higher plasma levels of VWF compared with HIV−ve stroke participants (P = 0·032). There was a trend to higher levels of VWF in HIV+ve strokes compared with HIV+ve nonstroke controls (P = 0·061). © 2015 World Stroke Organization Vol 10, December 2015, 1294–1239066 1295 Re30s7earch S. Allie et al. HIV+ve stroke group compared with the HIV+ve nonstroke con- HIV+ve patients without stroke. These findings support the role trols (P = 0·018). However, the ADAMTS13 levels of the latter of both VWF and ADAMTS13 in the pathogenesis of HIV-related group were still lower than that of the defined normal range. stroke. These factors are likely to contribute to the increased cere- CD4 counts were negatively correlated with VWF levels in all brovascular risk associated with HIV infection. HIV+ve participants (strokes and nonstroke controls; P = 0·006; see Fig. 1a). CD4 counts were not correlated with ADAMTS13 Authors’ contributions levels in HIV+ve participants. S. A., M. C., and A. S. contributed to the conception and design of Discussion the study. S. A. and A. S. performed the main data collection. All authors contributed to writing and critical review of the manu- In this study, performed within a few days of an acute ischemic script, and gave their final approval of the version to be published. stroke, stroke in HIV-infected participants was associated with a prothrombotic state with high levels of VWF and low levels of Ethics approval ADAMTS13. The question is whether these changes occurred as a consequence of the stroke, as part of an acute phase reaction, or The Human Research Ethics Committee of the University of Cape whether they preceded the stroke and contributed to its cause. We Town/Groote Schuur Hospital approved this study (HREC ref: cannot definitively answer this question. However, the finding of 178/2010). a negative correlation between CD4 counts and VWF levels in the HIV+ve participants shows that VWF rises with advancing References 1 Benjamin LA, Bryer A, Emsley HC, Khoo S, Solomon T, Connor MD. immunosuppression. At a certain critical prothrombotic level, HIV infection and stroke: current perspectives and future directions. this may precipitate strokes in susceptible individuals. Endothelial Lancet Neurol 2012; 11:878–90. doi: 10.1016/S1474-4422(12)70205-3; dysfunction is an important feature of HIV infection and PubMed PMID: 22995692; PubMed Central PMCID: PMC3460367. becomes more severe in advanced HIV where it is associated with 2 Tipping B, de Villiers L, Wainwright H, Candy S, Bryer A. Stroke in increased cardiovascular risk (11). VWF is released by platelets patients with human immunodeficiency virus infection. J Neurol Neu- rosurg Psychiatry 2007; 78:1320–4. [Epub 2007 Apr 30]. and endothelial cells. It is likely that in HIV infection VWF levels 3 Ortiz G, Koch S, Romano JG, Forteza AM, Rabinstein AA. Mecha- are elevated as a result of both endothelial dysfunction and nisms of ischemic stroke in HIV-infected patients. Neurology 2007; chronic low grade activation of the procoagulant pathways. In our 68:1257–61. study, the median VWF levels were significantly higher in the 4 von Blann A. Willebrand factor and the endothelium in vascular HIV+ve stroke group compared with the HIV−ve group. There disease. Br J Biomed Sci 1993; 50:125–34. 5 Hassan MI, Saxena A, Ahmad F. Structure and function of von Wille- was a trend to higher VWF in HIV+ve strokes compared with brand factor. Blood Coagul Fibrinolysis 2012; 23:11–22. doi: 10.1097/ HIV+ve controls, but there was no significant difference between MBC.0b013e32834cb35d; PubMed PMID: 22089939. the HIV−ve strokes compared with HIV+ve controls. Therefore, 6 Denis CV, von Lenting PJ. Willebrand factor: at the crossroads of the elevated VWF is not solely the consequence of the stroke. bleeding and thrombosis. Int J Hematol 2012; 95:353–61. doi: 10.1007/ With regard to ADAMTS13 levels, these were markedly reduced s12185-012-1041-x; [Epub 2012 Apr 5]; PubMed PMID: 22477538. 7 Bongers TN, de Maat MP, van Goor ML et al. High von Willebrand in both HIV+ve and HIV−ve stroke participants. It is likely there- factor levels increase the risk of first ischemic stroke: influence of fore that this indicates a nonspecific poststroke effect. However, ADAMTS13, inflammation, and genetic variability. Stroke 2006; ADAMTS13 levels were reduced to 23% of the population mean 37:2672–7. [Epub 2006 Sep 21]. in the HIV+ve nonstroke controls (Table 1). This suggests 8 Lip GY, Blann AD, Farooqi IS, Zarifis J, Sagar G, Beevers DG. Sequen- ADAMTS13 is reduced with HIV infection. It may be that the low tial alterations in haemorheology, endothelial dysfunction, platelet activation and thrombogenesis in relation to prognosis following levels poststroke mask a low prestroke level due to a floor effect. acute stroke: the West Birmingham Stroke Project. Blood Coagul This, in combination with high VWF levels, may place patients at Fibrinolysis 2002; 13:339–47. higher risk of stroke. A prothrombotic state with auto-antibodies 9 Novitzky N, Thomson J, Abrahams L, du Toit C, McDonald A. Throm- to ADAMTS13 has previously been described in the condition of botic thrombocytopenic purpura in patients with retroviral infection HIV-associated TTP (10). is highly responsive to plasma infusion therapy. Br J Haematol 2005; 128:373–9. A potential limitation of this study is the heterogeneity of 10 Meiring M, Webb M, Goedhals D, Louw V. HIV–associated throm- stroke etiology. However, the strokes were most commonly idio- botic thrombocytopenic purpura – what we know so far. European pathic. Furthermore, the relevance of a positive VZV PCR in HIV Oncology and Haematology 2012; 8:89–91. stroke is still debated. Studies using measurements of these factors 11 Subbarao V, Lowe D, Aghamohammadzadeh R, Wilkinson RJ at several time points poststroke would help to differentiate causal Endothelial dysfunction in HIV; in Aghdassi E (ed): HIV Infection in the Era of Highly Active Antiretroviral Treatment and Some of Its from consequential effects. Associated Complications, 2011: 17–36 [Online]. InTech; Available at In summary, we have shown that VWF levels are elevated in http://www.intechopen.com/articles/show/title/endothelial- HIV+ve stroke patients and that ADAMTS13 levels are low in dysfunction-in-hiv [accessed 19 March 2012]. 1296 Vol 10, December 2015, 1294–1296 © 2015 World Stroke Organi3z0at7ion S-36 308 Case Report Valproic Acid Associated Platelet Dysfunction: Case Report Andrea Deborah Jafta, Muriel Meiring, Charmaine Conradie. Department of Haematology and Cell Biology, Faculty Of Health Sciences, University Of Free State, Bloemfontein, RSA. Corresponding author: Dr AD Jafta, Department of Haematology and Cell Biology, Faculty Of Health Sciences, University Of Free State, P.O. Box 339(G2), Bloemfontein, 9300, Republic Of South Africa. Email: JaftaAD@ufs.ac.za. SUMMARY Valproic acid is commonly used as an anticonvulsant. It has been shown to inhibit the secondary phase of platelet aggregation. This can be reflected in increased bleeding times and haemorrhage. We describe a case of a 56-year-old male with a history of bleeding during a previous operation. He had valproic acid associated platelet dysfunction. Keywords: Valproic acid; Platelet aggregation; Platelet function tests; Preoperative care; Haemorrhage. INTRODUCTION Preoperative evaluation of bleeding disorders partial thromboplastin time (aPTT). A prolonged relies heavily on history taking. The first step is to PFA-100 bleeding time of 15 minutes led to further establish if the patient has had a history of major investigations. Von Willebrand screening tests; surgery and/or trauma and whether there was any von Willebrand factor antigen (VWF:Antigen), significant bleeding associated with these events. If von Willebrand factor-collagen binding activity previous major surgery and/or trauma had not been (VWF:CBA) & von Willebrand factor ristocetin associated with significant bleeding then surgery cofactor (VWF:RCo) assay, factor VIII (FVIII) level, can proceed. When the patient has had no history and the multimer pattern were all normal. Platelet of major surgery or trauma, history still remains aggregometry on platelet-rich plasma showed lack important to establish the likelihood that the patient of aggregation with arachidonic acid (AA) but there has an acquired bleeding disorder or the likelihood was aggregation with adenosine diphosphate (ADP), of an inherited bleeding disorder in the patient and/or epinephrine (EPI), Collagen (Col), and ristocetin in close relatives. Drug history must be considered (Figure 1). These findings were suggestive of a drug when one is evaluating for an acquired bleeding induced platelet dysfunction. Since valproic acid disorder. was the only drug which the patient was taking, it was stopped for two weeks under close monitoring for seizures. The bleeding time and platelet function CASE REPORT tests were then repeated. The bleeding time normalised and the platelet function tests were all A 56-year-old male with recurrent meningioma normal. The patient proceeded with the operation was preoperatively screened for a bleeding which was uneventful. tendency after he gave a history of bleeding during a previous operation for the meningioma. He was DISCUSSION AND CONCLUSION on valproic acid for seizure prophylaxis. The full blood count was normal: white cell count (WCC) There are no standard guidelines on the testing and 9.0, haemoglobin (Hb) 14 and platelet count (Plt) interpretation of platelet function tests 1 but lack 213; as were prothrombin time (PT) and activated of aggregation with only AA acid is found in the 54 Afr J Haematol Oncol April - June 2010;1(2):54-56 www.afjho.com 308 309 JAFTA AD ET AL 55 Figure 1. Aggregometry results. (A and B) adenosine diphosphate (ADP); (A) is control and (B) is patient’s sample. Traces 1, 2 and 3 represent high, medium and low concentrations of ADP respectively. Graph (B) shows aggregation with ADP. (C) Arachidonic acid (AA); Trace 1 is control and Trace 2 is patient’s sample. There was no aggregation with AA. (D and E) Collagen (Col); (D) is control and (E) is patient’s sample. Traces 1 and 2 represent high and low concentrations of Col. (E) shows aggregation with Col. (F and G) epinephrine (EPI); (F) is control and (G) is the patient’s sample. Traces 1, 2 and 3 represent high, medium and low concentration of EPI. (G) shows aggregation with EPI. (H and I) ristocetin; (H) is control and (I) is patient’s sample. Traces 1, 2 and 3 represent high, medium and low concentrations of ristocetin respectively. (I) shows aggregation with ristocetin. Afr J Haematol Oncol April - June 2010;1(2):54-56 www.afjho.com 309 310 56 VALPROIC ASSOCIATED PLATELET DYSFUNCTION so called “aspirin-like defect” 1 of platelets which bleeding during surgery has been well described. is seen in patients taking aspirin and other drugs. 4-5 Valproic acid may affect both platelet count and The lack of aggregation with AA implies inability function (thereby prolonging bleeding time) and by the platelets to produce thromboxane A2 which coagulation factors such as fibrinogen and factor is required for the AA agonist to induce platelet VIII. 6-7 aggregation. Platelet aggregation testing with AA has been suggested as the ideal test to perform before all This case emphasises the role of a proper other aggregation tests to screen for induced platelet medication history in the workup of a patient with a dysfunction. 2 It should of course be noted that the bleeding diathesis. impedance-based whole blood platelet aggregation test may give different aggregometry patterns to the FOOTNOTES ones obtainable by the method we used (optical- based platelet-rich plasma test). 3 Conflicts of interest: The authors declare no competing conflicts of interest An association between valproic acid and excessive REFERENCES 1. Zhou L, Schmaier AH. Platelet Impedance-based Whole Blood acid? Analysis of bilateral femoral aggregation testing in platelet- Platelet Aggregation Methods. osteotomy in children with total rich plasma. Am J Clin Pathol Clin Appl Thrombosis/Hemostasis involvement cerebral palsy. J 2005;123:172-183. 2005;11(1):25-35. Pediatr Orthop 2005;25:283-5. 2. Sirridge MS, Shannon R. Laboratory 4. Pohlmann-Eden B, Peters CNA, 6. Winter SL, Kriel RL, Novacheck Evaluation of hemostasis and Wennberg R et al. Valproate TF et al. Perioperative blood loss: Thrombosis. Lea & Febiger, induces reversible factor the effect of valproate. Pediatr Philadelphia, p95, 1983. XIII deficiency with risk of Neurol 1996;15:19-22. 3. Dyszkiewicz-Korpanty AM, Frenkel perioperative bleeding. Acta 7. Gidal B, Spencer N, Maly M et al. EP, Sarode R. Approach to the Neurologica Scandinavica Valproate-mediated disturbances Assessment of Platelet Function: 2003;108:142-5. of hemostasis: relationship to Comparison between Optical- 5. Carney BT, Minter CL. Is operative dose and plasma concentration. based Platelet-rich Plasma and blood loss associated with valproic Neurology 1994;44:1418-22. Afr J Haematol Oncol April - June 2010;1(2):54-56 www.afjho.com 310