Detection of human papillomavirus types in head and neck squamous cell carcinoma

Abstract

Human papillomaviruses (HPVs) are the aetiologic agents for diverse clinical conditions ranging from benign lesions to cervical cancer. HPVs have also been associated with head and neck squamous cell carcinomas (HNSCC), in both men and women. HNSCC refers to carcinomas that occur in different sub-sites of the head and neck including the larynx, oral cavity, oropharynx and sinonasal cavity. Studies have shown that the prevalence of HPV infection varies in different population groups from geographically distinct regions, with much higher burden documented in developed countries compared to less developed countries. For instance, HPV-associated HNSCC age‐standardized incidence rates (ASR) of over 1.25 per 100,000 were reported in countries in Europe and Northern America. However, the prevalence in less developed countries could be an underestimation due to limited studies. South Africa has a high incidence of oral squamous cell carcinoma, however not much is known about the prevalence of HNSCC cases due to HPV, or the HPV types associated with HNSCC. HPV associated HNSCCs have a better prognosis compared with HNSCCs caused by other agents such as smoking and alcohol. Currently there are limited tests and biomarkers for early diagnosis and prognosis of the cancer. Therefore, a better understanding of the molecular pathways of carcinogenesis is essential in order to select accurate predictive biomarkers and for the development of effective treatment. Technological advancements in molecular science have improved the discovery of potential biomarkers in cancer, including non-coding RNAs (ncRNAs) which are divided into small non-coding (<200 bp) and long non-coding (>200 bp). Most studies on biomarkers for application in diagnosis of cancers have focused on small non-coding RNAs using tissue samples. While little is known about the association of long non-coding RNAs such as antisense noncoding RNA in the INK4 locus (ANRIL), HOXA transcript at the distal tip (HOTTIP) and HOX transcript antisense intergenic RNA (HOTAIR) with use of less invasive samples such as blood. Therefore, the aims of the present study were firstly to determine the prevalence of HPV in head and neck cancer using archived specimens from patients with histologically confirmed HNSCCs submitted to the Department of Anatomical Pathology, Universitas Academic Laboratory in Bloemfontein, South Africa, between January 2004 and December 2014. HPV DNA was detected using conventional nested multiplex PCR targeting the L1 region and confirmed using in house type specific E6 primer pairs and a heminested PCR. A total of 780/994 samples tested were found to have intact DNA based on amplification of a partial reference gene. The 780 samples were further tested for the presence of HPV DNA. In total 57/780 (7.3%) were positive for the HPV DNA. High risk (HR) HPV types were more frequently detected than low risk (LR) HPV types. HPV16 was the most frequently detected HR type, and was detected in 26/57 (45.6%) HPV DNA positive biospies. Secondly, the study determined the phylogenetic relatedness and single nucleotide polymorphisms (SNPs) among the high risk HPV isolates using a Maximum Likelihood method for construction of phylogenetic trees. To determine the usefulness of partial L sequence data for identifying HPV lineages, a comparison was made using sequence data retrieved from GenBank for complete HPV genomes, complete L1 genes and partial L1 genes. Based on the outcome the ability to differentiate HPV lineages/sub-lineages was determined. HPV31 and 52 were classified with the L1 partial sequence data. With regards to HPV18 isolates, the partial L1 region was not diverse enough to identify sub-lineages. For HPV45, data for the partial L1 gene was able to discriminate sub-lineages within lineage B but not clearly within the A lineage. Similarly, for HPV16 and 33 clear discrimination of sublineages was not possible. However overall, although the partial L1 sequence data was too conserved for sub-lineage discrimination, it was possible to resolve lineages for some of the HPV types, particularly HPV31 and 52. The identification of biomarkers for early detection of cancers, including HNSCC, would have benefit for any patient. Hence a previously described potential biomarker, HOTAIR was investigated. The expression levels of HOTAIR in laryngeal squamous cell carcinoma (LSCC) patients and healthy volunteers were determined using real time quantitative PCR (RT-qPCR) assay. Patient biopsies were also screened for HPV DNA. Total RNA was extracted from whole blood of patients presenting to Universitas Academic Hospital with histologically confirmed LSCC and healthy volunteers, while DNA was extracted from the LSCC biopsies. Most of the patients, 80% (4/5) had an advanced tumour (T3 and T4), with advanced stage cancer (stages III and IV). The majority of the patients 80% (4/5) did not show lymph node invasion nor metastasis. Histopathology revealed moderately differentiated squamous cell carcinoma in one patient while well differentiated histopathology was seen in the rest of the patients. The mean expression level of the biomarker was found to be significantly higher (p=0.03) in blood collected from patients with laryngeal cancer compared to controls (6.39± 2.13 versus 10.19± 2.24, respectively), with a 13.9-fold overexpression in laryngeal cancer patients’ blood as compared to expression in blood from healthy participants. HPV DNA was detected in 1/5 biopsy samples. The results suggest that further investigation of expression levels of HOTAIR as a potential biomarker is warranted.