Objectives: Human papillomavirus (HPV) infection status is an important consideration in defining oropharyngeal squamous cell carcinoma (OPSCC) treatment and prognosis. HPV-positive OPSCC is associated with significantly better oncologic outcomes compared to HPV-negative OPSCC, and recent management strategies have focused on treatment de-intensification. While the vast majority of cases experience excellent oncologic outcomes, there is a subset of patients who develop persistent or recurrent disease.  The purpose of this study is to assess differences in clinical, demographic, and pathologic characteristics between HPV-positive OPSCC treatment failures and responders.

Study Design: Single institutional retrospective chart review

Methods: A retrospective chart review was performed of consecutive OPSCC patients seeking treatment at the University of North Carolina (UNC), a tertiary care cancer center, between 2007 and 2015. All tumor samples underwent p16 immunohistochemistry as a surrogate for HPV status. Differences in covariate distributions by treatment response status were tested by Pearson Chi-square test or the Fisher Exact test for sparse data.

Results: 217 p16-positive OPSCC cases were identified. The majority of patients 127 (58.7%) were treated with chemoradiation, 19 (8.76%) with radiation alone, and 28 (13.0%) with primary surgery and radiation.  Thirty-seven (17.1%) patients failed treatment with curative intent, of which 9 (24.3%) had biopsy-proven persistent disease and 28 (75.7%) recurrence. Treatment failures were less likely to be male (73.0% vs. 87.8%, failures vs. responders, respectively, p=0.021), less likely to have a tonsil subsite (38.9% vs. 47.5%, failures vs. responders, respectively, p=0.020), more likely to have advanced T-stage (43.2% vs. 26.3%, treatment vs, responders, respectively, p=0.012), and less likely to have poorly differentiated tumors (12.9% vs. 37.7%, failures vs. responders, respectively, p=0.010).  There was no significant difference between OPSCC treatment failures and responders with respect to age at diagnosis, race, marital status, N-stage, M staging, ever smoking status, heavy smoking history (greater than 10 pack years), ever drinking status, and heavy alcohol history (greater than 5 drinks per day). 

Conclusions: Our findings demonstrate a distinct clinical, demographic, and pathologic profile associated with HPV-positive OPSCC treatment failures. Advanced T-stage, female sex, non-tonsil subsite, and moderate/well differentiation were all statistically significantly associated with treatment failure.  Importantly, advanced nodal status, smoking, and alcohol use did not materially impact treatment response. These criteria may contribute to risk stratification, defining high-risk HPV-positive OPSCC, and defining criteria for treatment de-intensification.  Future studies aimed at identifying molecular characteristics of HPV-positive OPSCC treatment failures are currently underway.