MSI-High RAS-BRAF wild-type colorectal adenocarcinomas with MLH1 loss have a high frequency of targetable oncogenic gene fusions whose diagnoses are feasible using methods easy-to-implement in pathology laboratories

Abstract

Targetable kinase fusions are extremely rare (<1%) in colorectal cancers (CRCs), making their diagnosis challenging and often underinvestigated. They have been shown particularly frequently among MSI-High, BRAF/KRAS/NRAS wild-type CRCs with MLH1 loss (MLH1loss MSI-High wild-type). We searched for NTRK1, NTRK2, NTRK3, ALK, ROS1, BRAF, RET, and NRG1 kinase fusions in CRCs using methods easy-to-implement in pathology laboratories: immunohistochemistry (IHC), fluorescent in situ hybridization (FISH), and fully automated real-time PCR targeted analyses. RNA-sequencing analyses were used for confirmation. Among 84 selected MLH1 deficient (IHC) CRCs cases, MLH1loss MSI-High wild-type CRCs consisted first in 19 cases after Idylla™ analyses and finally in 18 cases (21%) after RNA-sequencing (detection of one additional KRASG12D mutation). FISH (and when relevant, IHC) analyses concluded in 5 NTRK1, 3 NTRK3, 1 ALK, 2 BRAF, and 2 RET FISH positive tumors. ALK and NTRK1 rearranged tumors were IHC positive, but pan-TRK IHC was negative in the 3 NTRK3 FISH positive tumors. RNA-sequencing analyses confirmed 12 of 13 fusions with only one false positive RET FISH result. Finally, 12/18 (67%) of MLH1loss MSI-High wild-type CRCs contained targetable kinase fusions. Our study demonstrates the feasibility, but also the cost-effectiveness, of a multistep but rapid diagnostic strategy based on nonsequencing methods to identify rare and targetable kinase fusions in patients with advanced CRCs, as well as the high prevalence of these kinase fusions in MLH1loss MSI-High wild-type CRCs. Nevertheless, confirmatory RNA-sequencing analyses are necessary in case of low FISH positive nuclei percentage to rule out FISH false-positive results.