Abstract 345: An integrative approach of image analysis and transcriptome profiling to explore potential predictive biomarkers for TGF-beta blockade therapy

Abstract

Abstract Background: The pattern of tumor infiltrating lymphocytes (TILs) in the tumor microenvironment (TME) can be classified as immune-inflamed, -excluded and -desert and may be a critical factor for immunotherapy efficacy and disease course of solid malignancies. Recent studies have identified TGFβ activity and signaling in the TME as a determinant of cytotoxic T cell exclusion and poor response to PD-1/PD-L1 blockade. Our goal was to identify tumors with immune-excluded phenotype with high levels of TGFβ activity, as predictors of response to TGFβ inhibition in checkpoint refractory tumors to make “cold” (immune excluded and desert) tumors “hot”. We applied an integrative approach using image analysis together with gene expression profiling to quantify distribution of TILs and characterize the associated TGFβ pathway activity. Methods: A multiplex fluorescent IHC assay was developed for PanCK/CD3/CD8 by Indivumed GmbH (Germany) and applied to a collection of archival biopsies including Bladder (n=20), CRC (n=29), HNSCC (n=19), Gastric (n=18) and Ovarian (n=19) cancers. We utilized digital pathology image analysis to classify the immune phenotypes based on annotation of the images by a pathologist. Computational imaging methods were used to localize the CD3+ and CD8+ lymphocytes and discern their proximity to tumor cells. RNAseq was performed on the 29 CRC tumors to investigate the prominent biological features associated with infiltration phenotypes, as well as the correlation between these phenotypes and a TGFβ activation signature derived from TGFβ-stimulated versus naïve cancer cell lines. Results: Analysis of T cell spatial distribution in the solid tumor biopsies (N=105) revealed substantial differences in the distribution of the infiltration phenotypes by tumor type. Computational assessments achieved 100% concordance with the pathologist assessment for tumors with the inflamed and desert phenotype but just 70% concordance with the excluded phenotype. Additionally, the distribution of topographies differs between cancer types, i.e., CRC and bladder cancer tend to be immune-excluded. The TIL score was negatively correlated with TGFβ pathway activation, together with elevated TGFβ signaling activity observed in “cold” tumors. Conclusion: Our image analysis platform was able to identify these T cell topographies in TME in the five tumor types. The robust association between TGFβ gene signature and immune phenotypes further demonstrates their potential as predictive biomarkers to identify appropriate patients that may benefit from TGFβ blockade. Citation Format: Robert Pomponio, Qi Tang, Anthony Mei, Anne Caron, Bema Coulibaly, Joachim Theihaber, Maximilian Rogers-Grazado, Tun Tun Lin, Rui Wang. An integrative approach of image analysis and transcriptome profiling to explore potential predictive biomarkers for TGF-beta blockade therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 345.