Abstract

Background: Microsatellite instability (MSI) is a well-established biomarker in colorectal cancer. Gene fusions may also influence prognosis, but the clinical significance of their co-occurrence with MSI remains poorly understood.

Methods: We analyzed a cohort of colorectal (N = 29,099) cancer patients, all profiled with Tempus AIs xT platform. Fishers tests assessed the enrichment of pathogenic and likely pathogenic gene fusions within microsatellite stability groups. Cox models evaluated the independent and interaction effects of MSI status and gene fusions on real-world overall survival from at biopsy collection, adjusting for age, stage, line of therapy, tumor purity, tumor mutation burden, and PD-L1 combined positive score. We tested for enrichment of all fusions, kinase fusions (NTRK1, NTRK3, RET, ALK, FGFR2, BRAF), and NTRK fusions (NTRK1, NTRK3).

Results: The proportion of clinically relevant gene fusions was higher in MSI versus MSS colorectal cancer tumors (5.9% vs 1.8%): NTRK1 (2.2% vs. 0.048%), NTRK3 (1.2% vs. 0.14%), RET (0.87% vs. 0.096%), and TPM3 (1.0% vs. 0.0074%). Survival analyses revealed that all fusions and kinase fusions were both associated with significantly worse overall survival (Table 1). These trends persisted in sub-cohorts. MSI was associated with increased overall survival in models including all fusions, kinase fusions, and NTRK fusions. We did not detect a significant interaction between MSI and gene fusions across any of the models, likely due to smaller sample sizes. However, we confirmed fusions remained associated with poor overall survival when restricting to MSS tumors for all gene fusions and kinase fusions. Additional models fit to patients treated with specific therapies, such as immunotherapy, did not converge.

Conclusion: Gene fusion rates differ between MSI and MSS colorectal tumors. While most fusions are generally associated with worse survival, this effect may be reversed in MSI cases.

VIEW THE PUBLICATION