Human papillomavirus and head and neck cancer: a systematic review and meta-analysis


Dr S.J. Thomas, Division of Maxillofacial Surgery, University of Bristol, Lower Maudlin Street, Bristol BS1 2LY, UK Tel.: +44 117 928 4500; fax: +44 117 925 2603; e-mail:


•It has been suggested that the link between human papillomavirus (HPV) and head and neck squamous cell carcinoma (HNSCC) is specific to carcinoma of the tonsil.

• We systematically reviewed studies that tested for HPV16 exposure in anatomically defined sites in the head and neck and a control group.

• The association between HPV16 and cancer was strongest for tonsil (OR: 15.1, 95% CI: 6.8–33.7), intermediate for oropharynx (OR: 4.3, 95% CI: 2.1–8.9) and weakest for oral (OR: 2.0, 95% CI: 1.2–3.4) and larynx (OR: 2.0, 95% CI: 1.0–4.2).

• To investigate heterogeneity, further stratification by method of HPV16 detection, suggested that variation in the magnitude of the HPV-cancer association with cancer site was restricted to studies using ELISA: among studies using PCR, the magnitude of the summary odds ratios was similar across the four sites.

• The association between HPV16 infection and HNSCC in specific sites suggests the strongest and most consistent association is with tonsil cancer, and the magnitude of this association is consistent with an infectious aetiology.

• However, the method of viral detection may be an important source of heterogeneity. Resolution of this issue will require further studies using both methods, examining associations separately in different sites.