← Back to List

Program Number: AHNS01
Session Name: Scientific Session 1 - HPV-positive Oropharyngeal Disease
Session Date: Wednesday, May 14, 2025
Session Time: 11:15 AM - 12:00 PM

HPV+ OPSCC recurrence risk after TORS is associated with a smaller fraction of tumor cells containing detectable E6/E7 RNA

Jean-Nicolas Gallant, MD, PhD¹; Malay K Sannigrahi, PhD¹; Lovely Raghav, PhD¹; David Shimunov, MD²; Robert M Brody, MD¹; Phyllis A Gimotty, PhD¹; Jalal Jalaly, MD¹; Devraj Basu, MD, PhD¹; ¹The University of Pennsylvania; ²SUNY Stonybrook

Background/objectives: Presence of high-risk human papillomavirus (HPV) types other than HPV16 has been associated with worse outcomes in p16+ oropharyngeal squamous cell carcinomas (OPSCCs), and not finding any HPV by various direct detection assays indicated markedly worse prognosis in a large multinational cohort of p16+ OPSCCs. In addition, therapy resistance has been proposed to arise from expansion of a fraction of tumor cells that has undetectable HPV and distinct host gene expression by single-cell mRNA sequencing. We evaluated a case-control cohort of p16+ OPSCCs receiving transoral robotic surgery (TORS) for relationships between recurrence and three features: (1) presence of non-HPV16 types, (2) lack of directly detectable HPV mRNA , and (3) decreased size of the HPV mRNA-expressing tumor cell fraction.

Methods: A single institution retrospective cohort of 851 treatment-naïve p16+ OPSCC patients receiving TORS-based therapy was used to curate 50 pT1/2 tumors that later recurred (cases) and 50 cured tumors (controls) that were matched for 8th edition AJCC pathologic TNM stage, smoking history, and adjuvant therapy. HPV type and mRNA levels were defined by bulk mRNA sequencing (RNAseq). A sensitive, clinically validated in situ hybridization (ISH) assay probing for E6/E7 mRNAs encoded by HPV16, 18, 31, and 33 was used to define HPV positivity as a binary readout. All E6/E7 ISH-negative cells in tumor areas and their subset of CD45-IHC-positive tumor infiltrating leukocytes (TILs) were digitally quantified and used to estimate the percent ISH-positive tumor cells by calculating (100 – (%ISH-negative cells - %TILs)).

Results: 100% of tumors expressed mRNAs from high-risk HPV types by RNAseq, with 9% (n=9) expressing mRNAs from non-HPV16 types (18, 33, 35, 59). There was no significant difference in non-HPV16+ tumors between cases (10%) and controls (8%) (p=1.0). HPV was undetectable by E6/E7 RNA-ISH in six tumors (6%), and three of the six were explained by presence of HPV types not in the ISH probe set. The six false negatives were equally distributed between cases and controls, and the two groups also did not differ significantly in bulk RNAseq reads of total HPV mRNA or E6/E7. However, the percentage of E6/E7 ISH-positive tumor cells only weakly correlated with overall RNAseq E6/E7 levels (r=0.22, p=0.04) and was markedly reduced in cases (p<0.001) (Figure 1A). Cases often contained large tumor regions devoid of any ISH-positive cells (Figure 1B), and the %ISH-positive tumor cells provided favorable discrimination between cases and controls based on an area under the ROC curve of 0.77 (OR=67.5, p<0.001, 95% CI=11.1-547.1) (Figure 1C). 

Conclusions: In typical p16+ OPSCC patients receiving TORS-based therapy, ISH-based HPV mRNA detection is generally unlikely to identify HPV false positives based on the p16 surrogate marker. In addition, recurrence prediction in this population was not aided by evaluating for presence of non-HPV16 types, HPV status by RNA-ISH, or overall HPV mRNA levels. By contrast, a reduced percentage of tumor cells reaching the threshold of HPV E6/E7 detection by RNA-ISH appears predictive of recurrence. This feature merits evaluation for generalizability as a biomarker and mechanistic interrogation as an etiology for therapy resistance.

 

← Back to List