Landscape of novel gene fusions in gastrointestinal tumors.

Abstract

268 Background: Gene fusions are recognized as critical drivers of cancer and targets for precision therapies. The 2022 ASCO provisional clinical opinion on tumor genomic testing includes recommendations for fusion testing and recognizes the strengths of RNA-based approaches. Targeted DNA/RNA sequencing panels can be used to detect well characterized, clinically actionable fusions, but their ability to detect rare novel fusions (i.e., those with uncharacterized fusion partners) can be limited by assay design. Whole-exome, whole-transcriptome sequencing (WES/WTS) may improve the identification of actionable gene fusions by detecting both known and novel fusions. This work describes clinically actionable novel fusions detected in gastrointestinal (GI) cancers. Methods: The study included samples from GI cancers that were analyzed using the OncoExTra test between May 2018 and May 2024. The assay uses tumor-normal WES/WTS to identify somatic alterations, including gene fusions with known and unknown partners and breakpoints.Clinical data were queried foractionable fusions with novel fusion partners that were unique in the entire laboratory database and have not, to our knowledge, been previously reported in the existing scientific literature or public databases. Actionable fusions were defined as those associated with FDA-approved targeted therapies or those that made patients eligible for a clinical trial. Results: Tumor samples from 2,776 patients with GI cancers were evaluated with a total of 75 actionable fusions identified in samples from 74 patients. Of these, 25.3% (n=19) were novel, and 74.7% (n=56) were previously described or recurrent fusions. Novel fusions were detected in small bowel (2 of 81), colorectal (4 of 1,621), gastric/gastroesophageal junction (GEJ) (2 of 259), pancreatic (7 of 466) and biliary (4 of 131) cancers. Thirteen (63.2%) of the novel fusions involved genes associated with FDA-approved therapies, including FGFR2 (n=6), BRAF (n=3), MET (n=2), FGFR1 (n=1) and NTRK2 (n=1). Of the fusions associated with FDA-approved therapies, 20% (8/40) of recurrent fusions and 38.5% (5/13) of novel fusions were detected only by RNA sequencing. No fusions were identified in appendiceal or esophageal cancers. Conclusions: Gene fusions with novel fusion partners represented one fourth of all actionable fusions detected in GI tumors. Use of a comprehensive WES/WTS testing approach that detects fusions agnostic of the fusion partner may affect clinical care in GI cancers by increasing the number of patients eligible for targeted therapies.