P16 and HPV-DNA Tests Discordance in Human Papilloma Virus (HPV)-Associated Oropharyngeal Cancer: Results From a Case-matched Study

Abstract

HPV-associated oropharyngeal cancer (OPC) has been associated with improved prognosis compared to HPV-negative OPC. Concomitant HPV DNA and p16 testing has been shown to improve sensitivity and specificity; however, the prognostic significance of discordant testing (i.e., p16+/HPV DNA or p16-/HPV DNA+) in OPC patients has not been quantitatively studied due to inadequate sample sizes. We aim to determine whether there is a prognostic significance for discordant p16/HPV-DNA test results for OPC patients receiving curative intent radiation treatment. A case-control study design was used to retrospectively analyze 136 patients with pathologically proven oropharyngeal cancer (OPC) treated with curative intent radiation therapy at our institution between the years 2005 and 2012. HPV DNA status was evaluated using in-situ hybridization (ISH) methods, and p16 status was identified via immunohistochemistry (IHC). Two subgroups were identified: concordant (p16+/HPV DNA+) and discordant (p16+/HPV DNA- or p16-/HPV DNA+) subgroups, with the former serving as control. Characteristics including sex, age at diagnosis, ethnicity, smoking status, cancer stage, and treatment modality were identified. Kaplan-Meier survival analysis and univariate Cox proportional hazard analysis were performed. Patient characteristics, including age, sex, smoking status, stage, pathological grade, tumor subsite within oropharynx, treatment modality, and follow-up duration were well-matched between the concordant and discordant subgroups. Survival analysis failed to show any statistically significant difference between the concordant and discordant subgroups in terms of overall survival, local, locoregional, or distant control. The discordant subgroup was comprised entirely of p16+/HPV DNA- patients; no patients in the discordant subgroup were found to be p16-/HPV DNA+. Univariate analysis demonstrated statistically significant correlation between overall survival and N-category in the discordant group (P = .024 by Effect Wald Test), in contrast to the concordant group (P = .18). This study suggests that HPV test discordance is not correlated with differential treatment outcomes as compared to HPV test concordance. p16 testing by IHC has been previously shown to be more sensitive and specific than HPV DNA testing by ISH. This study further validates this observation and suggests that p16 may be used as a standalone test in the diagnosis of HPV in HNC patients.