Identification of predictive biomarker for immunotherapy by associating with CD8+T cell Infiltration in lung adenocarcinoma.

Abstract

e21177 Background: Tumor immune microenvironment (TIME) has been proved associated with response to immunotherapy(I/O). We hypothesized that screening potential mutation pattern which could significantly impact the tumor infiltrating lymphocytes(TILs) can help us to identify predictive biomarkers for I/O in Lung adenocarcinoma(LUAD). Methods: Multiple-dimensional data from The Cancer Genome Atlas LUAD cohort (n = 514) was used for building a mathematical model beween mutation signature and CD8+TIL score (based on MCP-counter). An independent public validation cohort (cohort 1: LUAD, n = 598) were used to assess the immunotherapeutic predictive performance of the potential mutation patterns. Results: Top 100 gene associated with CD8+TIL score were selected based on MC+ model which can provides the minimum non-convexity of the penalized loss given the level of bias. Seven TIME genes (SPTA1 coef 0.09; MET coef 0.02; HSD3B1 coef -0.00; STAT4 coef -0.01; EGFR coef -0.08; PIK3CB coef -0.08; KEAP1 coef -0.24) were generated by taking the intersection of the top 100 mutant genes and FoundationOne (F1) CDx NGS 315 genes panel and verified in cohort 1. Survival analysis showed that SPTA1mt was the only one that associated with both significantly longer PFS (median PFS 3.15 vs 2.89 months; HR 0.65; 95% CI 0.45 to 0.93; p = 0.02) and OS (median PFS 15.08 vs 7.36 months; HR 0.59; 95% CI 0.40 to 0.88; p = 0.01) for patients who received I/O compared with chemotherapy(CT) among seven TIME genes. In order to test our hypothesis fully, a pooled analysis of SPTA1mt (a core positive predictors of CD8+TILs) and KEAP1mt (a core negative predictors for CD8+TILs ) were conducted and yielded that co occurrence of SPTA1mt and KEAP1mt had a compound effects for TIME. The validation showed that co mutation with SPTA1mt was accompanied by an decrease HR for I/O vs. CT in both PFS (HR S+K vs. K only 0.59 vs 1.56) and OS (HR S+K vs. K only 0.39 vs 0.80) for KEAP1mt patients. Conclusions: Our data show that it is feasible to identify individuals or groups of individual with specific mutations to immunotherapy responses from TIME view. SPTA1mt was a core predictors for higher CD8+ TILs and can be identified as a predictive biomarker for benefit from I/O compared with CT. Prospective studies are warranted for further investigation.