Objective: The global rise of HPV(+) oropharyngeal squamous cell carcinoma (OPSCC) has generated considerable interest underlying its etiology and management. Relative to HPV(-) OPSCC, HPV(+) patients are younger and more responsive to chemoradiation, though non-responders within this sub-class of OPSCC indicate divergent molecular pathophysiology, and continue to challenge clinicians. Rather than an adaptive approach to what many consider distinct disease processes, treatment remain the same for all OPSCC. Improved molecular stratification would greatly enhance the clinician's ability to precisely tailor treatment without jeopardizing outcomes. Here, two HPV(+) OPSCC cohorts delineated based on treatment response are compared via a hybrid Data Dependent Acquisition/Data Independent Acquisition (DDA/DIA) approach via mass spectrometry, to detect low-abundance proteins and highlight pathways not detected through genomic platforms alone.

Methods: Fourteen HPV(+) OPSCC patients who underwent definitive chemoradiation were stratified as clinical responders (n=6) or nonresponders (n=8). Responders were defined as having no recurrence for at least three years after treatment. Formalin-fixed paraffin tissue (FFPE) slides of tissue at diagnosis were confirmed to be p16(+) OPSCC by an independent pathologist, microdissected to achieve >80% tumor purity, and subjected to heat and pressure cycling via barocycler to enhance protein yield and identification. Equal mass of protein was digested and run on a Sciex TripleTOF mass spectrometer. Peptides identified with <1% Protein FDR from DDA and DIA modes were combined into a high confidence assay library comprised of 1603 proteins, which were subsequently quantified using the openSWATH workflow. Identified proteins underwent normalization and statistical analysis using mapDIA and Ingenuity Pathway Analysis.

Results: Collectively, 1440 proteins were identified and quantified. Of those, 212 were found to be significantly different between responders and nonresponders (FDR<0.01 and fold change greater than ± 1.25). Responders showed relative enrichment in targets of the insulin receptor (INSR) (p=2.86 x 10-14) whose expression pattern predicted inhibition of this pathway (z-score=-3.52). Targets linked to insulin-like growth factor 1 (IGF1R) indicated this related pathway was also inhibited (z-score=-2.60). IGF1R and INSR are both known to have oncogenic functions and increase cell growth, proliferation, and survival, with the biofunction “proliferation of cells” found to be inhibited in the responders (z-score=-1.54). Overall, tumor sub-types from non-responder patients demonstrated molecular profiles consistent with elevated growth, survival, and proliferation signaling.

Conclusions: We demonstrate the feasibility of a modified barocycler approach to obtain robust proteomic data from FFPE tissues at a comparable rate to fresh frozen tissues. We further identify via a hybrid mass spectrometry approach targets of insulin receptor signaling that may be a driver of pathogenesis at the proteomic level. Such targets may stratify HPV(+) OPSCC patients at diagnosis and help tailor the intensity of therapy to mitigate toxicities based on recurrence risk.

Citation Format: Aaron Robinson, Sarah J. Parker, Shiao L. Stephen, Beatrice S. Knudsen, Zachary S. Zumsteg, Jennifer E. Van Eyk, Allen S. Ho. Proteomic stratification of HPV(+) oropharyngeal squamous cell carcinoma identifies targets in insulin receptor signaling within nonresponders [abstract]. In: Proceedings of the AACR-AHNS Head and Neck Cancer Conference: Optimizing Survival and Quality of Life through Basic, Clinical, and Translational Research; April 23-25, 2017; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2017;23(23_Suppl):Abstract nr 41.