Immune checkpoint expression and co-expression landscape in gastroesophageal adenocarcinoma.

Background: Addition of immunotherapy to standard chemotherapy marginally improved outcomes in gastroesophageal adenocarcinoma (GEAC). Recently, dual immunotherapy has shown efficacy in melanoma and non-small-cell lung cancer over single checkpoint inhibition. We sought to decipher the expression landscape of commonly targeted immune checkpoints in GEAC with a particular attention to their co-expression with programmed death-ligand 1 (PD-L1). Materials and methods: Targeted RNA sequencing was carried out on 65 metastatic GEAC tumors obtained from, and gene expression was measured for 394 immune transcripts. Co-expression analyses were conducted by calculating Pearson correlations for every possible pair of 15 checkpoint genes and clustering groups of similarly expressed genes. These analyses were validated using a 90-patient cohort identified from The Cancer Genome Atlas database. Results: The clinical cohort was primarily male (86.2%), Caucasian (93.9%), and had positive PD-L1 status (63.1%). The correlation matrix delineated two clusters of co-expression with the first including PD-L1, programmed cell death protein 1 ( Conclusions: We show that immune checkpoints are often co-expressed in GEAC, suggesting possible common underlying mechanisms for checkpoint expression. Strong correlation between checkpoints like