Abstract
Immune checkpoint inhibitors (ICIs) significantly improve the survival outcomes of patients with advanced melanoma. However, response varies among from patient to patient and predictive biomarkers are urgently needed. We integrated mutational profiles from next-generation sequencing (NGS) data and clinicopathologic characteristics of melanoma patients to investigate whether tumor genomic profiling contribute to clinical benefit of ICIs treatment. The majority of genes identified with high mutation frequency have all been reported as well-known immunotherapy-related genes. Thirty-five patients (43.2%) had at least 1 BRAF/RAS/NF1 mutation. The other 46 (56.8%) melanomas without BRAF/RAS/NF1 mutation were classified as Triple-WT. We identified mutational signature 6 (known as associated with defective DNA mismatch repair) among cases in this cohort. Compared to patients with PD-L1 expression (TPS < 1%), patients with PD-L1 expression (TPS ≥ 1%) had significantly higher median progression-free survival (mPFS), but no significantly higher durable clinical benefit (DCB) rate. In contrast, FAT1, ATM, BRCA2, LRP1B, and PBRM1 mutations only occurred frequently in patients with DCB, irrespective of PD-L1 expression status. Our study explored molecular signatures of melanoma patients who respond to ICIs treatment and identified a series of mutated genes that might serve as predictive biomarker for ICIs responses in melanoma.