Abstract B016: Utilizing gene expression profiles (GEP) to assess tumor microenvironment (TME) and response to immunotherapy

Abstract

Abstract Background: Despite the success of anti-PD-1/PD-L1 immunotherapies, outcomes for lung cancer patients remain suboptimal. There is a critical need to incorporate multi-omics data focused on the tumor microenvironment (TME) to enhance patient selection and treatment individualization. This study aimed to identify predictive and prognostic biomarkers of immunotherapy response by characterizing transcriptional profiles of lung adenocarcinoma patients who underwent genomic profiling at diagnosis. Methods: RNA sequencing data were obtained from 68 lung adenocarcinoma patients who underwent next-generation sequencing with CARIS at our institution. We analyzed gene expression profiles (GEP) comprising 18 inflammatory genes associated with antigen presentation, chemokine expression, cytolytic activity, and adaptive immune resistance, identified as potential biomarkers for immune checkpoint inhibitor (ICI) response (Cristescu et al., 2018). The genes included were CCL5, CD27, CD274 (PD-L1), CD276 (B7-H3), CD8A, CMKLR1, CXCL9, CXCR6, HLA-DQA1, HLA-DRB1, HLA-E, IDO1, LAG3, NKG7, PDCD1LG2 (PD-L2), PSMB10, STAT1, and TIGIT. GEP scores were calculated, and hierarchical clustering identified tumor subsets with similar GEPs. Gene set variation analysis was conducted to assess enrichment of cell-cell interaction activities of specific ligand-receptor pairs of immune checkpoint genes, partially overlapping with the 18 GEP genes. Results: Hierarchical clustering identified four distinct GEP clusters: "T cell-inflamed GEP," "T cell-non-inflamed GEP high," "T cell-non-inflamed GEP medium," and "T cell-non-inflamed low GEP," with average scores of 0.18 (N=9, 13%), 0.11 (N=29, 43%), 0.08 (N=19, 28%), and 0.04 (N=11, 16%), respectively. The gene scores among these clusters showed significant differences (P = 3.3e-09). Additionally, analysis of specific ligand-receptor pairs of immune checkpoint genes (PD1/PD-L1, PD-L2; CTLA4/CD28, CD80, CD86; ICOS/ICOSLG; TIM3/GAL9; SIRP1/CD47) showed consistent positive enrichment in the “T cell-inflamed” cluster, with no or negative enrichment in the “T cell-non-inflamed” clusters. This suggests active immune checkpoint interactions in clusters with high GEP scores. Conclusion: We identified four different clusters among tissues from patients with NSCLC based on GEP scoring highlighting differences in immune activation and TME makeup, which may play a role in predicting response to immunotherapies. Our results suggest that GEP may provide information on identifying optimal immunotherapeutic approaches based on TME. Our findings underscore the importance of using biomarker-guided strategies to improve patient selection for precision immunotherapy approaches in lung cancer. Citation Format: Changde Cheng, Doug Welsch, Kayla F. Goliwas, Jessy Deshane, Sameer Deshmukh, Sanad Ahlushki, Yanis Boumber, Darshan Shimoga Chandrashekar, Alexander C. Mackinnon, Aakash Desai. Utilizing gene expression profiles (GEP) to assess tumor microenvironment (TME) and response to immunotherapy [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor-body Interactions: The Roles of Micro- and Macroenvironment in Cancer; 2024 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(22_Suppl):Abstract nr B016.