NTRK oncogenic fusions are exclusively associated with the serrated neoplasia pathway in the colorectum and begin to occur in sessile serrated lesions

Abstract

Neurotrophic tropomyosin receptor kinase (NTRK) gene fusions are emerging tissue-agnostic drug targets in malignancies including colorectal carcinomas (CRCs), but their detailed landscape in the context of various colorectal carcinogenesis pathways remains to be investigated. In this study, pan-tropomyosin receptor kinase (TRK) protein expression was assessed by immunohistochemistry (IHC) in retrospectively collected colorectal epithelial tumor tissues, including 441 CRCs [133 microsatellite instability-high (MSI-high) and 308 microsatellite stable (MSS)] and 595 premalignant colorectal lesions (330 serrated lesions and 265 conventional adenomas). TRK-positive cases were then subjected to next-generation sequencing and/or fluorescence in situ hybridization to confirm NTRK rearrangements. TRK IHC positivity was not observed in any of the MSS CRCs, conventional adenomas, traditional serrated adenomas, or hyperplastic polyps, whereas TRK positivity was observed in 11 of 58 (19%) MLH1-methylated MSI-high CRCs, 4 of 23 (17%) sessile serrated lesions with dysplasia (SSLDs), and 5 of 132 (4%) sessile serrated lesions (SSLs). The 11 TRK-positive MSI-high CRCs commonly harbored CpG island methylator phenotype-high (CIMP-high), MLH1 methylation, BRAF/KRAS wild-type, and NTRK1 or NTRK3 fusion (TPM3-NTRK1, TPR-NTRK1, LMNA-NTRK1, SFPQ-NTRK1, ETV6-NTRK3, or EML4-NTRK3). Both NTRK1 or NTRK3 rearrangement and BRAF/KRAS wild-type were detected in all nine TRK-positive SSL(D)s, seven of which demonstrated MSS and/or CIMP-low. TRK expression was selectively observed in distorted serrated crypts within SSLs and was occasionally localized at the base of serrated crypts. NTRK fusions were detected only in SSLs of patients aged ≥50 years, whereas BRAF mutation was found in younger age-onset SSLs. In conclusion, NTRK-rearranged colorectal tumors develop exclusively through the serrated neoplasia pathway and can be initiated from non-dysplastic SSLs without BRAF/KRAS mutations prior to full occurrence of MSI-high/CIMP-high.

NTRK Fusions in 1113 Solid Tumors in a Single Institution

“…The six NTRK -rearranged CRCs were all MSI-H tumors, a significantly higher proportion than the 8% proportion of MSI-H in the entire CRC population. As in previous reports [ 19 , 20 , 24 ], we confirmed once again that NTRK -positive CRCs demonstrated a higher frequency of MSI. Therefore, a subset of CRCs harboring no known driver mutations and exhibiting MSI-H should be tested using pan-Trk IHC and further validated using RNA-based targeted NGS.…”

“…The ETV6–NTRK3 fusion case showed weak to intermediate nuclear staining. Weak pan-Trk IHC expression was commonly observed in various tumors, including CRCs harboring an ETV6–NTRK3 fusion, and it was recently demonstrated that the lower sensitivity of pan-Trk IHC was caused by NTRK3 fusion, especially for ETV6–NTRK3 fusion [ 20 , 21 , 22 , 23 ]. Notably, no false-positive CRC cases were identified in the pan-Trk IHC in this study.…”

Lynch syndrome‐related colorectal carcinomas are NTRK‐negative

“…In colorectal cancers (CRC), NTRK fusions are rare, occurring in less than 1% of cases, and ranging in the literature from 0.2% 2 to 2.4%. 3 This difference in prevalence among the different studies is probably related to the absence of selection of the tested lesions. In fact, when restricted to DNA mismatch repair deficient (MMRd)/microsatellite instable (MSI) tumours, 0.9% of cases are positive for NTRK compared to 0.17% of MMR proficient/microsatellite stable lesions.…”

“…5 Of note, a relatively high prevalence of NTRK fusions were also observed in their premalignant counterpart, such as sessile serrated lesions with or without dysplasia, which showed a positivity in NTRK in 17% and 4% of the cases, respectively. 3 In this context, we performed an exploratory evaluation of the incidence of NTRK alterations in Lynchrelated CRC, which are universally recognized as MMRd, BRAF wildtype, and MLH1-promoter unmethylated lesions, mostly characterized by germline pathogenic variants in the MMR genes.…”

And the story goes on: non-conventional dysplasia of the colorectum

“…As a matter of topical interest and therapeutic relevance, findings of a recent study by Kim et al [ 5 ] concluded that NTRK -rearranged colorectal carcinomas progress exclusively via the serrated pathway of neoplasia, thus expanding the molecular landscape of serrated colorectal lesions. To the best of our knowledge, oncogenic gene fusions including NTRK fusions have yet to be studied in serrated lesions in IBD, and would indeed form an interesting research question.…”