Most NTRK fusions occur at very low frequencies in various common cancers. Recent recommendations on NTRK testing recommend immunohistochemistry (IHC) as the initial test for tumor types with a low frequency of NTRK fusions. This study investigated the accuracy of an IHC assay to detect NTRK fusions and characterize the clinicopathological and molecular features of NTRK-rearranged tumors. This retrospective study was conducted on 1113 solid tumor samples known to harbor no oncogenic driver alterations, including 510 non-small cell lung cancers (NSCLC), 503 colorectal cancers (CRC), and 79 inflammatory myofibroblastic tumors (IMT). Additionally, 21 ALK expression-positive cases were included. TRK expression was evaluated using a pan-Trk IHC assay, and positive cases were validated using NGS. TRK expression was observed in three NSCLCs (0.6%), six CRCs (1.2%), and six IMTs (6%). NTRK fusions were finally detected in two NSCLCs (0.4%), six CRCs (1.2%), and one IMT (1%). In NSCLC and CRC, the majority of NTRK fusions were readily discernible due to diffuse moderate-to-strong cytoplasmic staining on pan-Trk IHC. In IMT, focal weak nuclear staining indicated the presence of NTRK fusion. Therefore, the utility of pan-Trk IHC should be assessed considering that the difference in performance depends on tumor type.
Keywords: NTRK fusion; TRK immunohistochemistry; TRK inhibitors; colon cancer; inflammatory myofibroblastic tumor; lung cancer; next-generation sequencing.