Transcription of human papillomavirus oncogenes in head and neck squamous cell carcinomas

Some head and neck cancers are caused by human papillomavirus (HPV). As HPV vaccination can prevent infection, an estimation of which HPV types have an active viral oncogene transcription in what proportion of tumors might allow estimation of the proportion of head & neck cancers preventable by HPV vaccination. We used all RNA sequencing data from primary tumors of head and neck squamous cell carcinomas from the Cancer Genome Atlas (n = 500 patients). We analysed 3.7 terabyte of sequencing data with the bioinformatics pipeline ViraPipe. Paired end reads were quality filtered using the original code and aligned to known HPV sequences. HPV transcripts were found in 113/500 specimens, with transcription of both the E6 and E7 viral oncogenes in 90 specimens. HPV16 had E6/E7 transcription in 67 cases, HPV33 in 14 cases, HPV18 in 6 cases and HPV35 in 5 cases. HPV oncogene transcription was most common in tumors from tonsils (34/40, 85%), followed by palate (4/5, 80%), base of tongue (10/20, 50%), oropharynx (4/10, 40%), and gum (4/11, 36%). Comparison to the cancer incidence statistics in the USA indicates that vaccine-preventable HPV16/18/33 oncogene transcription would be found in about 8.3% female and 20.2% male patients of head and neck cancers in the USA. Transcription of the HPV oncogenes is present in a large proportion of head and neck cancers in the TCGA database. If these cancers are caused by HPV, prevention of HPV16/18/33 infections would prevent ~49 300 annual head and neck cancer cases in the USA alone.