Fluorescent in situ hybridization is a deceptive method for screening NRG1 gene rearrangements

Abstract

Abstract Background NRG1 rearrangement has been identified in many tumors and is considered an important treatment target. However, the prevalence of NRG1 fusion is extremely rare, and there are no universal testing algorithms for genetic testing. Methods A total of 3008 cases of various kinds of tumors were included in this study. Fluorescence in situ hybridization (FISH) and immunohistochemistry (IHC) were used to screen NTRK translocation and p-HER3 expression. FISH translocation or p-HER3 IHC-positive cases were further subjected to next-generation sequencing (NGS) testing. Results No cases were p-HER3 positive by IHC. Twenty-nine cases (0.96%, 29/3008) with NTRK translocation were found by FISH, and there were three different signal types: (A) break-apart signal (three cases) with or without a high copy number of the 3’-end of the gene; (B) low copy number of the 5’-end of the gene with respect to the 3’-end of the gene, with fusion signals (12 cases); and (C) low copy number of the 5’-end of the gene with respect to the 3’-end of the gene, without fusion signals. Through NGS, only eight of the 29 cases were confirmed to carry NRG1 fusion. The FISH type C group was completely consistent with the NGS results. For clinical characteristics, all of these NGS NRG1 fusion tumors were adenocarcinomas, and the majority of these tumors (7/8, 87.5%) were female. In addition to NRG1 fusion-enriched breast cancer and lung cancer, we also found cholangiocarcinoma and colorectal carcinoma with NRG1 fusion. Conclusions Although FISH is a deceptive method for screening NRG1 gene rearrangements, signals showed low copy number of the 5’-end of the gene with respect to the 3’-end of the gene, without fusion signals were reliable for NTRK fusions. Because of the high false negativity and high cost of NGS, FISH is still a good method for screening NRG1 fusions across cancers.