The landscape of gene fusions and the correlation with immunotherapy biomarkers in Chinese patients with advanced solid tumor.

Abstract

e15039 Background: Due to the U.S. Food and Drug Administration approval of larotrectinib, a tumor or tissue-agnostic cancer treatment, gene fusions and fusion-driven cancer were identified as indications. Exploration of the gene fusions might be essential in both target therapy and immunotherapy in advanced solid tumor. However, the landscape of gene fusions in Chinese solid tumor patients is poorly reported. Methods: Tissue samples were obtained from 7514 Chinese patients with advanced solid tumor. Genomic alterations were identified via the next generation sequencing (NGS) with a validated commercial 381-cancer-gene panel (panel size over 0.5 MB), including 100 MSI loci. TMB was calculated according to NGS results. Statistical analysis was performed by Fisher exact test using R 3.6.1. Results: There were 162 pathologic or very like pathologic gene fusions observed in 249 (3.31%) patients, covering 30 cancer types. Top frequent fusion genes were ALK (93), RET (32), FGFR2 (26), FGFR3 (13), ROS1 (10), TMPRSS2 (10) and NTRK1 (9). Their main partner genes were EML4 (66/ 93), KIF5B (15/32), BICC1 (5/26), TACC3 (13/13), CD47 (3/10), ERG (9/10) and TPM3 (5/9), respectively. Mostly, only one fusion was detected in one patient. Three patients harbored 3 gene fusions and 35 patients harbored 2 gene fusions, among which 21 patients were observed harboring both fusion and its reciprocal fusion. In the group of patients harboring gene fusions, the prevalence of MSI-H was 4.0% (10/ 249). Especially, the prevalence was significant high (41.2%, 7/ 17) in colorectal cancer patients (P < 0.001). In MSI-H colorectal cancer patients, 4 carried NTRK1 fusions, 2 RET fusions and 1 ALK fusion. And all of them had high tumor mutation burden (TMB). As another immunotherapy response predictor, TMB was low in most solid tumor patients in this study, except for the patients harboring gene fusions involving NTRK1, NTRK2 or STK11, whose TMB-H proportion were 67%, 50% and 33%, respectively. Conclusions: Gene fusions might be associated with MSI-H. In colorectal cancer patients, gene fusions involved either NTRK1 or NTRK2 might be associated with TMB-H, which revealed a potential biomarker for both target and immunotherapy.