The landscape of novel and complementary targets for immunotherapy: an analysis of gene expression in the tumor microenvironment

// Stephen G. Gaffney 1 , * , Elizabeth B. Perry 1 , * , Ping-Min Chen 1 , Andrew Greenstein 2 , Susan M. Kaech 3 and Jeffrey P. Townsend 1 1 Department of Biostatistics, Yale University School of Public Health, New Haven, CT, USA 2 Gilead Sciences, Foster City, CA, USA 3 Department of Immunobiology, Yale University School of Medicine, New Haven, CT, USA * These authors contributed equally to this work Correspondence to: Jeffrey P. Townsend, email: Jeffrey.Townsend@Yale.edu Keywords: immunotherapy; tumor microenvironment; T lymphocyte; cancer genomics; RNA-seq Received: May 17, 2019     Accepted: May 29, 2019     Published: July 16, 2019 ABSTRACT Background: Immunotherapies targeting immune checkpoint proteins CTLA-4, PD-1, and PD-L1 have saved lives, but these therapies have only been effective in some patients. Patients positive for expression of immune checkpoint proteins in the tumor microenvironment show better response to immune checkpoint inhibitors. Consequently, knowledge of which genes are consistently expressed in lymphocytes within the tumor microenvironment can convey potentially effective and complementary new immunotherapy targets. Results: We identified 54 genes that have higher co-expression with the pan T-cell marker CD3E than CTLA4 or PDCD1 . In a dataset of 26 patients who received anti-PD-1 therapy, we observed that co-expression between CD3E and PDCD1 was higher among responders than non-responders, supporting our correlation-based approach. Conclusions: The genes highlighted in these analyses, which include CD6 , TIGIT , CD96 , and SLAMF6 , warrant further investigation of their therapeutic potential. Methods: We analyzed and ranked genes that were co-expressed with the pan T-cell marker CD3E in 9,601 human tumors, spanning 31 cancer types. To further identify targets that may be complementary to existing PD-1 therapy, we examined and ranked genes with high CD3E co-expression and relatively low PDCD1 co-expression.