A landscape analysis of immunotherapy-related biomarkers across solid tumors.

Abstract

e14268 Background: Checkpoint inhibitor-based immunotherapy has varying success among tumor types and patients. Predictive biomarkers are in high demand to assist with patient selection. Responses to these agents are correlated with programmed cell death ligand 1 (PD-L1) expression, microsatellite instability (MSI), and tumor mutational burden (TMB). Here we evaluated PD-L1 expression, MSI status and TMB in a variety of solid tumors to determine their relationships. Methods: A total of 109 specimens were identified in patients diagnosed with solid tumors that underwent a Paradigm Diagnostic Cancer Test. MSI scores were determined by the number of indels present in runs of 6 homopolymer bases per megabase, with a cut-off of 6 to distinguish between MSI-high (MSI-H) and MSI-stable (MSI-S). TMB scores greater than or equal to 10 muts/Mb were designated as high (TMB-H). Results: Of all tumors, 19.3% were MSI-high. PD-L1 testing was performed in 71.5% of all samples; of these, 21.8% were PD-L1 positive. TMB scores of PD-L1 positive tumors (mean = 10.3 muts/Mb) and PD-L1 negative tumors (mean = 8.6 muts/Mb) did not differ significantly (p = 0.57). All MSI-H tumors were TMB-H, and MSI-H tumors had higher mean TMB scores than MSI-S tumors (57 muts/Mb versus 7.6 muts/Mb; p = 0.06). Among TMB-H tumors, MSI-H status was associated with higher TMB scores compared to MSI-S tumors (55.2 muts/Mb versus 22.1 muts/Mb, p = 0.18). Of the both MSI-S and TMB-H tumors, 68.8% were lung cancers and 18.8% were breast cancers. Tumors both MSI-H and TMB-H were 47.6% gastrointestinal carcinomas and 33.3% endometrial carcinomas. Conclusions: In our analysis, TMB was independent of PD-L1 status. All MSI-H tumors were also TMB-H; these were primarily gastrointestinal and endometrial carcinomas whereas TMB-H tumors that were MSI-S consisted predominantly of lung cancers. This difference may be due to the different mechanisms of acquiring mutations during carcinogenesis in these two tumor types. These data also highlight the need for integrated PD-L1, MSI and TMB testing to identify potential responders to immunotherapy.