Abstract
NTRK fusions are rare but recurrent driver alterations in papillary thyroid carcinoma (PTC), with therapeutic significance due to the availability of targeted TRK inhibitors. Pan-TRK immunohistochemistry (IHC) provides a practical approach for the identification of NTRK fusions; however, its application and reliability in routine pathology require further exploration. This study is aimed at evaluating the diagnostic utility of pan-TRK IHC for detecting NTRK fusions in PTC, assessing its correlation with histopathologic features, and developing a diagnostic algorithm. We analyzed 107 BRAF p.V600E-negative PTC cases using pan-TRK IHC, correlating staining patterns with molecular data and histopathologic features. RNA-based targeted sequencing confirmed gene fusions. NTRK fusion-positive tumors were enriched in distinct histopathologic features, including BRAF-like PTC with predominant follicular architecture, clear cells, and secretory-like cells. Findings such as tumor cell stratification, glomeruloid structures, and papillae with subfollicle formation (microfollicles within papillary structures) were associated with both NTRK and RET fusion-positive PTCs. Correlation of pan-TRK IHC and molecular testing results identified non-specific reactivity or false positivity in 62% of pan-TRK IHC-positive PTCs, including cases with RET fusions, BRAF fusion, or no detectable fusion. However, pan-TRK IHC with high H-scores (≥ 110) was observed exclusively in cases with NTRK fusions. For cases with lower H-scores (< 110), integrating histopathologic features improved the identification of fusion-driven PTCs. While our series further supports the limitations of pan-TRK IHC, a diagnostic algorithm that combines pan-TRK IHC H-scores and histopathologic patterns improved the triaging of NTRK molecular testing of BRAF p.V600E-negative PTCs when a stepwise approach is undertaken. This study also demonstrated that TRK protein localization may vary with tumor progression and dedifferentiation.
Published Version
Join us for a 30 min session where you can share your feedback and ask us any queries you have