Prevalence and detection methodology for preliminary exploration of NTRK fusion in gastric cancer from a single-center retrospective cohort

The fusion of neurotrophic tyrosine receptor kinase (NTRK) is a novel target for cancer therapy and offers hope for patients with gastric cancer (GC). However, there are few studies on the prevalence and detection methods of NTRK fusions in GC. In this study, we used immunohistochemistry (IHC) as a screening method to select cases for molecular testing and evaluated the effectiveness of IHC, fluorescence in situ hybridization (FISH), and next-generation sequencing (NGS). We retrospectively collected 1970 patients with GC. Pan-TRK IHC was conducted in all cases, and three cases were positive: one with strong and diffuse cytoplasmic staining, while two with weak cytoplasmic staining. All three cases were validated using NTRK1/2/3 FISH. FISH results revealed a single 3' signal of NTRK1 in 95% of the tumor cells in the first case, while the remaining two cases were negative. NGS confirmed LMNA-NTRK1 fusion in the first case, with no gene fusion detected in the other two cases. Out of 46 negative controls, one had a non-functional fusion of IGR-NTRK1, and four had point mutations. The case with LMNA-NTRK1 fusion were negative for pMMR, EBV, HER2, and AFP. The pan-TRK IHC showed a 33.33% (1/3) concordance rate with RNA-based NGS. If the criterion for positivity was 3+ cytoplasmic staining, the agreement between IHC and RNA-based NGS was 100% (1/1). In conclusion, the incidence of NTRK fusion in GC is extremely low (0.05%). If the criteria are strict, pan-TRK IHC is highly effective for screening NTRK fusions. FISH could complement NGS detection, particularly when NTRK fusion is detected by DNA sequencing. NTRK fusion in GC may not be limited to specific subtypes.