Analysis of actionable gene fusions in a large cohort of Chinese patients with colorectal cancer

Fu-Rong Kou,Jian Li,Zheng-Hang Wang,Ting Xu,Juan-Juan Qian,En-Li Zhang,Li-Jun Zhang,Lin Shen,Xi-Cheng Wang
DOI: https://doi.org/10.1093/gastro/goae092
2024-10-09
Abstract:Background: The prevalence of gene fusion is extremely low in unselected patients with colorectal cancer (CRC). Published data on gene fusions are limited by relatively small sample sizes, with a primary focus on Western populations. This study aimed to analyse actionable gene fusions in a large consecutive Chinese CRC population. Methods: This study included 5,534 consecutive CRC patients from the Genecast database. Genomic profiling was performed using a panel of 769 cancer-related genes. Data for 34 CRC patients with actionable gene fusions were also collected from cBioPortal and ChimerSeq. Results: Among 5,534 CRC patients, 54 (0.98%) had actionable gene fusions, with NTRK1/2/3 being the most common fusion (0.38%), accounting for 38.9% (21/54) of those with fusions. Actionable gene fusion enrichment was higher in patients with microsatellite instability-high (MSI-H) (6.7% vs. 0.5%, P < 0.001), RAS/BRAF wildtype (2.0% vs. 0.2%, P < 0.001) and RNF43 mutation (7.7% vs. 0.4%, P < 0.001) than in patients with microsatellite stability/MSI-low, RAS/BRAF mutation and RNF43 wildtype, respectively. When these markers were combined, the fusion detection rate increased. Among patients with RAS/BRAF wildtype and MSI-H, fusions were detected in 20.3% of patients. The fusion detection rate further increased to 37.5% when RNF43 mutation was added. The fusion detection rate was also higher in colon cancer than in rectal cancer. No significant differences in clinical or molecular features were found in patients with actionable gene fusions between the Genecast, cBioPortal, and ChimerSeq databases. Conclusions: Approximately 1% of the unselected Chinese CRC population carries actionable gene fusions, mostly involving NTRK. Actionable gene fusions are more prevalent in MSI-H, RAS/BRAF wildtype, or RNF43-mutated CRC, as well as in colon cancer. Mapping of these molecular markers can markedly increase the fusion detection rate, which can help clinicians select candidates for fusion testing and targeted therapy.
What problem does this paper attempt to address?