The main hypothesis of this study is that gene expression profiles (GEPs) integrating both tumor antigenicity and a pre-existing adaptive immune response can be used to generate distinct immune-related signatures of BRAF mutant colorectal cancers (BRAF-CRCs) to identify actionable biomarkers predicting response to immunotherapy. GEPs of 89 immunotherapy-naïve BRAF-CRCs were generated using the Pan-Cancer IO 360 gene expression panel and the NanoString nCounter platform and were correlated with microsatellite instability (MSI) status and with CD8+ tumor-infiltrating lymphocyte (TIL) content. Hot/inflamed profiles were found in 52% of all cases, and high scores of Tumor Inflammation Signature were observed in 42% of the metastatic BRAF-CRCs. A subset of MSI tumors showed a cold profile. Antigen Processing Machinery (APM) signature was not differentially expressed in MSI tumors compared with MSS cases. By contrast, the APM signature was significantly upregulated in CD8+ BRAF-CRCs versus CD8- tumors. Our study demonstrates that a significant fraction of BRAF-CRCs may be a candidate for immunotherapy and that the simultaneous analysis of MSI status and CD8+ TIL content increases accuracy in identifying patients who can potentially benefit from immune checkpoint inhibitors. GEPs may be very useful in expanding the spectrum of patients with BRAF-CRCs who can benefit from immune checkpoint blockade.
Keywords: BRAF mutation; CD8+; MHC Class I antigen; MSI; NanoString; PanCancer IO 360; colorectal cancer; immune checkpoint inhibitors.