Incidence and genomic characteristics of gene fusions in a large Chinese colorectal cancer cohort.

Abstract

192 Background: Gene fusion is rare in colorectal cancer (CRC). With the advent of fusion-targeted therapeutics, such as entrectinib for NTRK fusion and selpercatinib for RET fusion, the identification of gene fusions holds clinical significance. Reports providing a comprehensive description of fusions in CRC are limited, and understanding the molecular characteristics of CRC with fusions will contribute to efficient fusion screening. Here, we investigate the incidence and genomic features of gene fusions in a large Chinese CRC population. Methods: We systematically analyzed next-generation sequencing results of tumor tissue from 5534 CRC patients between January 2020 and August 2023. The sequencing panel covers 769 cancer-related genes and can detect fusions, single nucleotide variants and indels, and microsatellite instability (MSI). Statistical significance was determined using chi-square test. Results: Overall, the median age of patients was 62 years (inter-quartile range [IQR]: 53-70), and 40% were female. Gene fusions were detected in 4.4% (242/5534) of all the patients. The incidence is 0.98% (54/5534) if only reported clinically actionable fusions are considered. The median age of the 242 fusion-positive patients was 64 years (IQR: 54-70), and 43% were female. The most highly detected potentially actionable gene fusions were NTRK (0.89%) and FGFR (0.81%), accounting for 39% of the fusion-positive patients. The other fusion genes were ERBB2 (0.52%), RET (0.38%), BRAF (0.31%), EGFR (0.27%), ALK (0.22%), MET (0.09%) and ROS1 (0.07%). Consistent with previous reports, patients with MSI-High were more likely to have fusions compared with those with microsatellite stability (MSS) or MSI-Low (9.9% vs 4.0%, p < 0.001). Besides, RAS or BRAF mutated patients accounted for 59% of the cohort, of whom 2.5% (83/3273) had fusions detected, while 7.0% (159/2261) of the RAS and BRAF wildtype patients had positive fusions ( p < 0.001). Moreover, 404 (7.3%) patients had RNF43 splice or truncating mutations, and gene fusions were detected in 12.9% of these patients. Combining these markers increased the detection rate of fusions, with 28% (33/118) of the RAS/ BRAF wildtype and MSI-H patients having fusions detected. Furthermore, in patients with wildtype RAS/BRAF and MSI-H and RNF43 splice or truncating mutations, the fusion detection rate was as high as 49% (32/65), including 24 patients with actionable fusions of definite clinical significance. Conclusions: Gene fusions were detected in 4.4% of the Chinese CRC population and tended to occur in patients with wildtype RAS/BRAF, MSI-H, or RNF43 mutations. Combining the mutational status of RAS, BRAF, and RNF43 along with MSI status can improve the fusion detection rate and help select candidates for fusion testing and targeted therapy.