Abstract 7103: p16 predictiveness for HPV-associated head and neck cancer in African Americans compared to Non-Hispanic Whites

Abstract This project addresses the need for precision biomarkers for HPV-associated oropharyngeal squamous cell carcinoma (OPSCC) in diverse populations. The immunohistochemistry (IHC) test for the protein p16 is the only clinically approved biomarker for HPV-status in OPSCC and is incorporated into the AJCC v8 staging for OPSCC. It is also used as inclusion criteria for clinical trials, determining the eligibility of patients to receive de-intensified treatment. Literature demonstrates that p16 IHC is estimated to have a sensitivity of 94% and a specificity of ∼85-90%. However, the predictive value is lower in populations with lower HPV prevalence, and for these reasons, it is important to study the predictive value of p16 in subpopulations. We investigated the reliability of p16 IHC as a surrogate for HPV+ in a cohort of OPSCC comprised of African Americans (AA) (n=58) and Non-Hispanic Whites (NHW) (n=88) from the Louisiana Tumor Registry (LTR) and University of Michigan (UM). HPV testing was performed on FFPE DNA using bulk RNA sequencing and mapping to high-risk HPV types. Twenty-nine percent (17/58) of AA and 88% (77/88) of NHW were HPV-positive based on RNA. However, 25/58 (43.1%) AA were p16+ and 77/88 (87.5%) of NHW were p16+. The sensitivity of the test for AA and NHW was estimated to be 94.1% and 94.8%, respectively. The specificity of the test for AA and NHW was estimated to be 78.0% and 63.6%, respectively. The positive predictive value (PPV) was significantly lower in AA patients (0.64) compared to NHW (0.948) (p-value = 0.00024, proportions test). In all, the correspondence of p16 and RNA HPV+ was 16/25 in AA OPSCC. Differential expression analysis confirmed that the expression of CDKN2A, the gene responsible for producing p16, was significantly different in HPV RNA positive status versus HPV RNA negative status of any race (logFC ∼2-fold, p < 0.05) and was not significantly different between NHW HPV RNA positive and AA HPV RNA positive groups (logFC = -0.49, p-value > 0.05). ROC analysis demonstrated that CDKN2A expression had predictive ability for HPV RNA status across all patients (AUC = 0.92), with subgroup analyses showing even higher predictiveness in AA patients (AUC = 0.96) compared to NHW patients (AUC = 0.90). These results indicate that use of p16 IHC for patient stratification in AA OPSCC cases could lead to p16+ HPV-negative cases inappropriately treated as HPV+ tumors, leading to adverse outcomes. This highlights the need for confirmatory HPV testing in OPSCC patients, especially in subpopulations that have a low attributable fraction of HPV, to avoid undertreatment. Our ongoing work will address the use of other HPV-related biomarkers and characterization of pathways that differ between AA and NHW OPSCCs. Prioritizing diversity in patient populations and outcomes research ensures advancements in cancer research reflects all populations, especially those historically underserved. Citation Format: Maria Christiene Lawas, Mei-Chin Hsieh, Siddhi Patil, Anju Duttargi, Aishani Kulshreshtha, Lucia Martinez Cruz, Bailey Garb, Shiting Li, Nisha D'Silva, Maureen Sartor, Laura Rozek. p16 predictiveness for HPV-associated head and neck cancer in African Americans compared to Non-Hispanic Whites [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7103.