Nivolumab, a PD-1 inhibitor, has variable response rates in patients and is not fully characterized in treatment-naïve patients. Here, we describe exploratory analyses of immune-related gene expression in squamous cell carcinoma of the head and neck (SCCHN) as predictors of response to nivolumab, administered for 4 weeks before planned surgical resection in the context of a window-of-opportunity trial.

RNA-seq was performed on pre- and post-treatment tumor tissues from 39 patients with SCCHN of any stage, who were candidates for complete surgical resection (NCT03238365). Expression was quantified using kallisto. Noncoding, histone-coding and mitochondrial transcripts were removed. HPV-specific reads were filtered using pathseq and aligned to HPV reference using viGEN. TME signature scores were quantified using ssGSEA.

While HPV-status did not predict clinical responses to nivolumab, the expression of individual HPV16 genes in post-treatment samples reflected response in this patient cohort. Patients who responded to nivolumab showed lower levels of E6 expression, while non-responsive patients had significantly higher E6 expression levels. In contrast, E4 expression was found to be increased in patients who responded to nivolumab treatment. As expected, HPV-positive tumors presented with a more inflamed TME compared to HPV-negative tumors.

A tumor microenvironment (TME) classification platform developed by BostonGene was used to analyze several gene expression signatures. Four distinct microenvironment subtypes were reproduced, consistent with pooled analysis. In addition, a high proliferation rate signature in baseline samples was closely associated with non-response to immunotherapy regardless of HPV status.

HPV-negative tumors with significantly higher CD8, FoxP3, and CD163 as measured by IHC, had a greater response to immunotherapy in contrast to HPV-positive tumors, where no association was found. This finding was supported by single gene expression and the BostonGene-developed deconvolution algorithm to reconstruct the TME composition of the tumors using RNA sequencing; predicted CD8+ T-cells and Macrophages were significantly enriched in HPV-negative responders. Interestingly, patients with HPV-negative tumors disproportionately responded better to treatment when they presented with an immune-inflamed TME; however, this pattern was not observed in HPV-positive tumors.

Our analysis indicates that HPV16 gene transcription appears to impact response to treatment. In addition, the immune-inflamed TME impacts response to nivolumab treatment in HPV-negative tumors. These properties can be assessed in a multiparametric analysis and could lead to the identification of predictors of response to immunotherapy in head and neck cancer.

Citation Format: Nikita Kotlov, Zoya Antysheva, Dina Antonova, Naira Samarina, Sandrine Degryse, Alban Linnenbach, Sanket Shukla, Larry Harshyne, Andrew P. South, Jennifer M. Johnson, Young Kim, Alexander Bagaev, Nathan H. Fowler, Adam Luginbuhl. Predictors of treatment response in a preoperative window of opportunity trial of nivolumab in resectable squamous cell carcinoma of the head and neck [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1441.