Leveraging Big Data of Immune Checkpoint Blockade Response Identifies Novel Potential Targets

Abstract

We here report a large-scale comparative meta-analysis of genomics and gene expression signatures of immune checkpoint blockade (ICB) responses in over 3,500 patients across 12 tumor types. Non-synonymous tumor mutational burden (nsTMB), as well as 22 out of 39 reported gene expression signatures, were significantly associated with pan-cancer ICB responses. Strikingly, the predictive value of a de novo gene expression signature (PredictIO_100), composed of the top 100 genes most significantly associated with ICB responses at pan-cancer level, was superior to nsTMB and other gene expression signatures. Two genes within PredictIO_100, F2RL1 (encoding protease-activated receptor-2) and RBFOX2 (encoding RNA binding motif protein 9), were concurrently associated with worse ICB clinical outcomes, T cell dysfunction in ICB-naive patients and resistance to dual PD-1/CTLA-4 blockade in preclinical models. Our study underlines the relative impact of candidate biomarkers of ICB responses and identifies F2RL1 and RBFOX2 as potential therapeutic targets to overcome ICB resistance.