Comparison of NTRK fusion detection methods in microsatellite-instability-high metastatic colorectal cancer

Abstract

Tropomyosin receptor kinase (TRK) inhibitors have been approved for metastatic solid tumors harboring NTRK fusions, but the detection of NTRK fusions is challenging. International guidelines recommend pan-TRK immunohistochemistry (IHC) screening followed by next generation sequencing (NGS) in tumor types with low prevalence of NTRK fusions, including metastatic colorectal cancer (mCRC). RNA-based NGS is preferred, but is expensive, time-consuming, and extracting good-quality RNA from FFPE tissue is challenging. Alternatives in daily clinical practice are warranted. We assessed the diagnostic performance of RNA-NGS, FFPE-targeted locus capture (FFPE-TLC), fluorescence in situ hybridization (FISH), and the 5′/3′ imbalance quantitative RT-PCR (qRT-PCR) after IHC screening in 268 patients with microsatellite-instability-high mCRC, the subgroup in which NTRK fusions are most prevalent (1–5%). A consensus result was determined after review of all assay results. In 16 IHC positive tumors, 10 NTRK fusions were detected. In 33 IHC negative samples, no additional transcribed NTRK fusions were found, underscoring the high sensitivity of IHC. Sensitivity of RNA-NGS, FFPE-TLC, FISH, and qRT-PCR was 90%, 90%, 78%, and 100%, respectively. Specificity was 100% for all assays. Robustness, defined as the percentage of samples that provided an interpretable result in the first run, was 100% for FFPE-TLC, yet more limited for RNA-NGS (85%), FISH (70%), and qRT-PCR (70%). Overall, we do not recommend FISH for the detection of NTRK fusions in mCRC due to its low sensitivity and limited robustness. We conclude that RNA-NGS, FFPE-TLC, and qRT-PCR are appropriate assays for NTRK fusion detection, after enrichment with pan-TRK IHC, in routine clinical practice.

Cost-Efficient Detection of NTRK1/2/3 Gene Fusions: Single-Center Analysis of 8075 Tumor Samples

“…This validation would require RNA-based NGS analysis involving huge tumor collections, and, therefore, excessive costs. It is noteworthy that the frequencies of NTRK rearrangements presented in this dataset match the previously reported NTRK prevalence estimates, thus supporting the suitability of the 5′-3′ NTRK imbalance assay [ 25 , 28 , 35 , 36 , 37 , 39 , 40 ]. Furthermore, we have encouraging results from the NSCLC study, which focused on patients with a very high probability of the presence of actionable gene fusion in the tumor tissue, i.e., young-onset female non-smokers.…”

“…Most of the studies aimed at detecting gene rearrangements employ IHC, FISH, or NGS, while the comparison of the amount of transcripts of 5′- and 3′-portions of the genes is a less common technique [ 25 , 26 , 27 , 28 , 29 ]. An example of its successful implementation is the fully automated commercial Idylla GeneFusion assay (Biocartis, Mechelen, Belgium) [ 30 , 31 ].…”

Navigating through novelties concerning mCRC treatment—the role of immunotherapy, chemotherapy, and targeted therapy in mCRC

“…The NTRK genes (NTRK1/2/3) encode tropomyosin receptor kinase proteins (TRK A/B/C) primarily involved in neural development and cellular homeostasis (117). NTRK fusions are a rare occurrence in CRC patients, accounting for around 1% of cases (118)(119)(120). Previous reports have indicated that CRC characterized by NTRK positivity may constitute a distinctive subset exhibiting specific features e.g., high TMB and a higher likelihood of MSI (121).…”