Validation of AJCC-8 for oropharyngeal squamous cell carcinoma in Appalachians with multiple confounders

Presentation: A103
Topic: Oropharynx / HPV Related Disease
Type: Poster
Authors: Clayton Burruss, BS1; Christine Sharrer, MS1; J. Zachary Porterfield, MD, PhD2; Alexandra E Kejner, MD2
Institution(s): 1University of Kentucky College of Medicine; 2University of Kentucky Department of Otolaryngology - Head and Neck Surgery


In contrast to the American Joint Committee on Cancer’s 7th edition staging criteria (AJCC-7), the 8th edition (AJCC-8) separates oropharyngeal squamous cell carcinomas (OPSCCs) into human papillomavirus-positive (HPV+) tumors and HPV-negative (HPV-) tumors.  While AJCC-8 offers better predictions for overall survival, various factors may modulate survival and patient outcomes.


To validate the AJCC-8 as a better metric of survivability over AJCC-7 and confirm AJCC-8 applicability among possible confounding variables common in Appalachia and other regions.

Study design: Retrospective chart review at an academic tertiary care institution from 2010 to 2020.

Methods: With IRB approval, a cohort of 1621 patients were obtained from medical records of patients admitted to the University of Kentucky healthcare system.  479 patients met the inclusion criteria consisting of an oropharyngeal primary with known p16 immunohistochemistry status.  Other data collected included demographics, insurance information, rurality, Appalachian status, Charlson Comorbidity Indices (CCI), recurrence, metastases, and mortality data. A Cox proportional hazard regression and multivariate logistic regression were performed.

Results: Within this cohort, the incidence of HPV+ and HPV- OPSCC was 377 (78.7%) and 102 (21.3%), respectively. AJCC-8 had a higher odds of predicting mortality of stage IV HPV+ OPSCCs compared to stages I - III (OR = 5.04, 95%CI: 1.98 - 14.12; p < 0.001) compared to AJCC-7 (OR = 1.41, 95%CI: 0.72 - 2.87; p = 0.32). Additionally, AJCC-8 appropriately stratified all stages of HPV+ OPSCC in a survival analysis (p < 0.001). AJCC-7 exhibited significant overlapping of the survival curves. Multivariate analysis demonstrated several factors that could have confounded the accuracy of AJCC-8. Extranodal extension (ENE) is associated with higher staging for HPV- OPSCCs under AJCC-8; it does not alter HPV+ staging. However, HPV+ OPSCCs with ENE had a higher odds of overall mortality (OR = 1.79, 95%CI: 1.11 - 2.90; p = 0.02) compared to HPV+ OPSCCs without ENE. Additionally, patients in this cohort demonstrate numerous additional comorbidities that may alter OS. HPV+ OPSCC patients with a CCI greater than or equal to 3 had a higher odds of mortality (OR = 1.80, 95%CI: 1.12 - 2.93; p = 0.02) compared to those with a score less than 3. Patients with HPV+ OPSCCs with Medicaid or self-pay-status had higher odds of overall mortality (OR = 1.62, 95%CI; 1.06 - 2.47; p = 0.03) compared to those with private insurance or Medicare. Finally, HPV+ patients from rural populations had a higher odds of presenting with stage IV OPSCC (OR = 2.68, 95%CI: 1.03 - 7.17; p = 0.04). Within this cohort, 32% of patients diagnosed with OPSCC were from rural populations. 

Conclusions/Relevance: AJCC-8 provides enhanced guidelines for staging OPSCCs based on p16 status. However, confounding variables may influence AJCC-8’s ability to accurately stratify patients into stages and predict overall survival. These results validate the improvements of AJCC-8 over AJCC-7 and confirm AJCC-8’s staging capabilities in the context of a population with statistically significant confounding variables.