Integrating immunohistochemistry for biomarker detection in NSCLC: A step toward precision therapy

Background: Advances in NSCLC treatment increasingly rely on biomarker-driven strategies, with targeted therapies outperforming cytotoxic agents. However, rapid progress in precision medicine poses challenges for translating biomarker recommendations into clinical practice. Accurate characterization is essential to ensure timely and appropriate therapy selection. Methods: We analyzed 30 adult primary NSCLC FFPE tumors (n=25 unknown biomarker status; n=5 known biomarker controls) using immunohistochemistry (IHC) for actionable and exploratory targets: ALK, ROS1, BRAF, EGFR, c-Met, pan-TRK, HER2, and MEK1. Biomarker frequency, intensity, and specificity were assessed, and correlations between expression patterns were investigated. All slides were scored by a thoracic oncology pathologist per clinical standards or literature. Results: Among five known controls, concordance was observed for HER2, c-Met, and EGFR, while discrepancies occurred for ALK (focal staining requiring reflex FISH/next-generation sequencing (NGS) confirmation) and BRAF. Controls were originally identified by NGS, which is generally more sensitive than IHC. Negative staining for BRAF was consistent with a non-V600E mutation in the control specimen. Correlation trends were noted between c-Met and MEK1, and between EGFR and HER2. Frequencies of ALK, c-Met, and HER2 aligned with literature, while EGFR and MEK1 appeared higher, suggesting IHC may serve as an initial screen before reflex NGS testing. No positivity was observed for BRAF, ROS1, or pan-TRK, consistent with their rarity in NSCLC. Conclusions: IHC remains a valuable screening tool for detecting most actionable biomarkers, particularly in settings lacking NGS capabilities. Its integration into routine pathology may optimize patient selection for targeted therapies and clinical trials. This text has been revised with the assistance of Microsoft Copilot to comply with the specified character limit.