Human papillomavirus expression in oral mucosa, premalignant conditions, and squamous cell carcinoma: a retrospective review of the literature.

OBJECTIVES: The literature of human papillomavirus detection in normal oral mucosa and oral lesions associated with the dysplastic progression of epithelium was reviewed to help define the role of this virus in the development of oral squamous cell carcinoma. STUDY
DESIGN: All available data from published studies were analyzed by chi-square test for association between the human papillomavirus and age, gender, race, DNA type, location, lesional dysplastic progression, method of detection, tissue preservation, and use of tobacco and alcohol.
RESULTS: Human papillomavirus was identified with increasing frequency in normal oral mucosa (13.5%), benign leukoplakia (14.8%), intraepithelial neoplasia (18.5%), squamous carcinoma (26.2%), and verrucous carcinoma (27%). It was detected in oral squamous cell carcinoma significantly (p < 0.005) more often (37.1%; 122 of 329) in studies that used a high sensitivity assay (polymerase chain reaction) than studies that used moderate sensitivity assays (25.2%; 84 of 334) (e.g., Southern blot hybridization) and low sensitivity assays (16.9%; 61 of 362) (e.g., immunohistochemistry, in situ hybridization). Human papillomavirus DNA was detected significantly more often (p < 0.001) in frozen oral squamous cell carcinoma (51.6%; 115 of 223) than paraffin-embedded tissue (21.7%; 136 of 628). High-risk human papillomaviruses (2, 16, 18) were detected in 81.4% of OSCCs that contained the virus compared with low-risk human papillomavirus genotypes (6, 11) in 17.9% of oral squamous cell carcinoma that contained the human papillomavirus (p < 0.001). In studies that analyzed the use of chemical cofactors, the use of tobacco and alcohol (87.3%) was associated more often with oral squamous cell carcinoma than the presence of human papillomavirus (51.3%), however, the difference was not significant.
CONCLUSION: High-risk human papillomavirus genotypes have a significant association with oral squamous cell carcinoma. However, the presence of this virus in a high proportion of oral normal mucosa makes the virus alone a poor prognosticator of progression to malignancy.