Predictivity of human papillomavirus positivity in advanced oral cancer

Human papillomavirus (HPV) is a known prognostic factor world over in patients of carcinoma oropharynx. The role of HPV in oral cancers has not been investigated adequately. We tried to identify standard clinicopathological features in oral cancer, which would predict HPV-positivity.

This was a retrospective analysis of 124 cases of T4 oral cancer patients at our center. HPV-positive was defined in accordance with positive p16 immunohistochemistry done on pretreatment local tumor site biopsy. Age, sex, habits (smoking history and oral tobacco), Eastern Cooperative Oncology Group performance status (ECOG PS), T stage, N stage, grade, and site were selected, for testing of prediction for HPV-positivity. The analysis was performed by R studio version 3.1.1. Two-sample test for equality of proportions with continuity correction was used to identify factors predicting for HPV-positivity. P = 0.05 was considered as significant.

Of 124 patients, 16 patients (12.9%) were HPV-positive. The median age of the whole cohort was 43 years (interquartile range 37-52 years) with 15 females (12.1%). All had squamous cell carcinoma (100%). The grade of the tumor was well differentiated in 9 patients (7.2%), moderately differentiated in 98 patients (79.1%), and poorly differentiated in 17 patients (13.7%). The ECOG PS 0 in 19 patients (15.3%), 1 in 104 patients (83.9%), and 2 in 1 patient (0.8%). The subsite of the tumor was buccal mucosa in 74 patients (59.7%), anterior two-third of tongue in 33 patients (26.6%), and others in 17 patients (13.7%). None of the tested factors except the use of oral tobacco were statistically significantly associated with HPV-positivity. History of tobacco usage had a statistical trend toward ability to predict HPV-positivity. The proportion of patients with HPV-positive oral cancer in patients without history usage of oral tobacco was 31.3% while it was 10.2% in patients with previous history of tobacco use (P = 0.03).

Standard clinicopathological variables could not predict for HPV-positivity. Negative history of tobacco (smokeless) usage showed statistical trends toward ability to predict HPV-positivity in oral cancer patients.