Pan-tumor screening for NTRK gene fusions using pan-TRK immunohistochemistry and RNA NGS fusion panel testing

Abstract

Targetable NTRK gene fusions can be detected across tumor types using methodologies such as pan-TRK IHC, DNA or RNA NGS testing, or FISH. Challenges for implementation of clinical testing for NTRK fusions may arise due to the range in NTRK fusion prevalence across tumors, endogenous levels of TRK expression in tissues, and the large number of potential fusion partners. In this study, we examined our experience evaluating driver mutation negative lung, urothelial or cholangiocarcinoma cases, in addition to cases with positive, equivocal, or weak staining by pan-TRK IHC for NTRK fusions. 63/127 (49.6%) of these cases were positive for pan-TRK IHC, of which 71.4% showed weak or focal staining, potentially due to physiologic or non-specific TRK expression. Of these 127 cases, 4 harbored a NTRK fusion (1 fusion was seen in two separate samples from the same patient) as confirmed by RNA fusion panel testing. Pan-TRK IHC was positive in 1 case with TPM3-NTRK1 fusion, equivocal in 1 case with GOLGA4-NTRK3 fusion, and negative in 2 samples with ADAM19-NTRK3 fusion. Our findings show that we were able to successfully identify NTRK fusions that resulted in targeted therapy. However, our results suggest limited sensitivity of pan-TRK IHC for NTRK3 fusions, and that the reduced specificity for pan-TRK IHC in tumors with physiologic or non-specific TRK expression, results in false positive samples that require confirmatory testing by RNA based NGS.