Abstract 1176: Immunogenomic profiling identifies a subgroup of squamous cell lung cancers with immunosuppressed tumor microenvironment and correlates TGF-beta and Wnt/beta-catenin signaling as predictive of low PD-L1 expression

Introduction: immune checkpoint inhibitors have revolutionized the treatment of non-small cell lung cancer (NSCLC), but only ~15% of patients will achieve durable clinical benefit. Our group and others have focused on characterizing the interaction between tumor genomic characteristics and immune microenvironment. Genes such as PTEN, CTNNB1, STK11 and KEAP1 have been associated with intra-tumor immune depletion, but a broader understanding of other key driver genes of NSCLC and their interplay with the immune microenvironment is currently lacking. To prospectively delineate the genomic and immune landscape of NSCLC, we launched the ICON Project (Immunogenomic Profiling of NSCLC). Methods: we prospectively collected tissue specimens from patients with stage I-IIIA NSCLC undergoing surgical resection. Samples were subjected to extensive immune-genomic profiling, including whole-exome and RNA sequencing, multiplex immunofluorescence (mIF), and flow cytometry. Results: from 2016-2018, 150 patients were accrued, with 78 samples having undergone genomic profiling. Most patients had stage I-II disease (74%) and non-squamous histology (70%). Most commonly mutated genes were TP53, EGFR, KRAS, CDKN2A, STK11, and mean TMB was 7.6 mut/Mb, which is comparable to TCGA (6.9mut/Mb, p=ns). Unsupervised clustering according to expression of CD8+ T cell related genes segregated tumors into immune-enriched (clusters 1 and 3) and immune-depleted (clusters 2 and 4) subgroups. We validated these findings through flow cytometry (CD8+GzB+ T cells) and mIF (CD8), with both methods showing CD8+ T cell infiltration to be enriched in clusters 1 and 3. Cluster 1 included squamous and non-squamous histology, tumors with epithelial and mesenchymal signatures, and high TMB. Cluster 2 included exclusively squamous histology with predominant mesenchymal tumor signature, high TMB, and was enriched for TP53 mutations. Cluster 3 had predominantly non-squamous histology, epithelial signature, and low TMB. Cluster 4 had exclusively non-squamous histology, predominant epithelial signature, low TMB, and was enriched for KEAP1 mutations. In addition, PD-L1 expression (by mIF) was negatively correlated with TGF-beta and WNT/Beta-catenin signaling pathway, as well as high EGFR and FGFR2 expression on malignant cells. Conclusion: CD8+ T-cell gene signature identifies immunosuppressed subgroups of lung cancer, especially among squamous cell carcinomas with a mesenchymal signature, that may be associated with resistance to immune-based therapies. TGF-beta and Wnt/beta-catenin pathways are associated with an immunosuppressed tumor microenvironment, and may be potential novel targets to restore anti-tumor immunity in lung cancers with low PD-L1 expression. Citation Format: Marcelo V. Negrao, Tatiana Karpinets, Jun Li, Alexandre Reuben, Cara Haymaker, Kyle G. Mitchell, Junya Fujimoto, Chi-Wan Chow, Edwin R. Parra, Lorenzo Federico, Jianhua Zhang, Ara A. Vaporciyan, Chantale Bernatchez, Tina Cascone, Boris Sepesi, Ignacio I. Wistuba, John V. Heymach, Jianjun Zhang, Don L. Gibbons, ICON Team. Immunogenomic profiling identifies a subgroup of squamous cell lung cancers with immunosuppressed tumor microenvironment and correlates TGF-beta and Wnt/beta-catenin signaling as predictive of low PD-L1 expression [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 1176.