Open Access Research article

Human Papillomavirus-associated oropharyngeal cancer: an observational study of diagnosis, prevalence and prognosis in a UK population

Mererid Evans1, Robert Newcombe2, Alison Fiander3, James Powell1, Martin Rolles4, Selvam Thavaraj5, Max Robinson6 and Ned Powell7*

Author Affiliations

1 Velindre Cancer Centre, Whitchurch, Cardiff, CF14 2TL, UK

2 Department of Primary Care and Public Health, School of Medicine, Cardiff University, Heath Park, Cardiff, CF14 4XN, UK

3 Department of Obstetrics and Gynaecology, School of Medicine, Cardiff University, Heath Park, Cardiff, CF14 4XN, UK

4 Singleton Hospital, Sketty Lane, Swansea,, SA2 8QA, UK

5 Department of Clinical and Diagnostic Sciences, King’s College London Dental Institute, Strand, London, WC2R 2LS, UK

6 Centre for Oral Health Research, Newcastle University, Framlington Place, Newcastle Upon Tyne, NE2 4BW, UK

7 HPV Oncology Group, Institute of Cancer and Genetics, School of Medicine, Cardiff University, Heath Park, Cardiff, CF14 4XN, UK

For all author emails, please log on.

BMC Cancer 2013, 13:220  doi:10.1186/1471-2407-13-220

Published: 1 May 2013



The incidence of Human Papillomavirus (HPV) associated oropharyngeal cancer (OPC) is increasing. HPV-associated OPC appear to have better prognosis than HPV-negative OPC. The aim of this study was to robustly determine the prevalence of HPV-positive OPC in an unselected UK population and correlate HPV positivity with clinical outcome.


HPV testing by GP5+/6+ PCR, In Situ Hybridisation (ISH) and p16 immunohistochemistry (IHC) was performed on 138 OPCs diagnosed in South Wales (UK) between 2001–06. Kaplan-Meier analysis was used to correlate HPV status with clinical outcome.


Using a composite definition of HPV positivity (HPV DNA and p16 overexpression), HPV was detected in 46/83 (55%) samples where DNA quality was assured. Five year overall survival was 75.4% (95% CI: 65.2 to 85.5) in HPV-positives vs 25.3% (95% CI: 14.2 to 36.4) in HPV negatives, corresponding to a 78% reduction in death rate (HR 0.22, p < 0.001). HPV-positives had less locoregional recurrence but second HPV-positive Head and Neck primaries occurred. Poor quality DNA in fixed pathological specimens reduced both HPV prevalence estimates and the prognostic utility of DNA-based HPV testing methods. As a single marker, p16 was least affected by sample quality and correlated well with prognosis, although was not sufficient on its own for accurate HPV prevalence reporting.


This study highlights the significant burden of OPC associated with HPV infection. HPV positive cases are clinically distinct from other OPC, and are associated with significantly better clinical outcomes. A composite definition of HPV positivity should be used for accurate prevalence reporting and up-front DNA quality assessment is recommended for any DNA-based HPV detection strategy.

Oropharyngeal; Oropharynx; Cancer; HPV; Papillomavirus; Prognosis; Tonsil