Abstract 4893: Combined immunohistochemistry and NGS-based patient profiling for predicting anti-PD-1/PD-L1 therapy response

Abstract

Abstract Introduction: There are several different modalities of predictive tests which support response to anti-PD-1/PD-L1 inhibitors therapy, including PD-L1 expression by immunohistochemistry (PD-L1 IHC), mismatch repair deficiency (dMMR), microsatellite instability (MSI), and recently emerging tumor mutation burden (TMB), and Gene Expression Panels (GEP). Each of these methods capture different facets of the immune system: TMB and MSI evaluates represents mutational/neoantigen load which can stimulate the immune system; GEP establishes a profile of immune response, and whereas PD-L1 IHC directly evaluates the state of checkpoint inhibition in the tumor and tumor microenvironment (TME). We constructed a compound testing paradigm for immune system monitoring called PredicineX, which combines genome analysis which relies on tissue or blood-derived nucleic acids and advanced tissue context analytics based on PD-L1 IHC in solid tissue biopsies to create a comprehensive patient profile to support anti-PD-1/PD-L1 therapy decision making. Methods: Using Predicine’s GeneRADAR technology, we developed a tissue- and blood-based NGS assay to capture genomic alterations in cancer genes including tumor mutation burden (TMB) and microsatellite instability (MSI). Using Flagship Biosciences cTA® tissue image analysis technology, we created an artificial intelligence (AI) based PD-L1 IHC scoring platform which provides both current PD-L1 IHC scoring paradigms and novel computational scores from the rich tissue context data profile created from PD-L1 IHC slides using this technology. We compared biomarker profiles in a cohort of patients using the two technologies and created an integrative data profile to evaluate its association with anti-PD-1/PD-L1 therapy response. Results: We demonstrate the synergistic value of combining genomic based TMB, MSI, and GEP immune profiles with contextual information from PD-L1 IHC slides in patient biopsies. The high complexity gene profile, combined with the rich tissue context data, provided a novel means to stratify patients into 4 categories: 1)mutation high/immune competent; 2)mutation low/immune deficient; 3)mutation high/immune deficient; and 4) mutation low/immune competent. Notably, a strong association between the stratified groups and anti-PD-1/PD-L1 therapy was identified. Conclusion: In this study, we have demonstrated the feasibility of a compound test relying on tissue- and blood-based NGS assay and PD-L1 IHC which can ascribe differential phenotypes to patients for clinically relevant and actionable decisions about response to anti-PD-1/PD-L1 therapy. Citation Format: Jianjun Yu, Jianguo Dong, Amy Wang, Joseph S. Krueger. Combined immunohistochemistry and NGS-based patient profiling for predicting anti-PD-1/PD-L1 therapy response [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4893.