Original Abstract
BACKGROUND: Real-world data examining structured modality selection for mediastinal/hilar adenopathy in mixed malignant and benign cohorts is limited. We evaluated the performance of a multidisciplinary team (MDT)-guided strategy incorporating EBUS/EUS-B and EUS for both benign and malignant disease. METHODS: We conducted a retrospective study, including consecutive adults undergoing mediastinal/hilar lymph node sampling by endoscopy after MDT discussion. All cases were discussed at a dedicated thoracic MDT before the procedure, where the indication for sampling and the modality were predetermined based on anatomic accessibility on CT/PET-CT, clinical data, and anticipated tissue requirements. The primary outcome was diagnostic yield, defined as the proportion of procedures providing a definitive pathologic diagnosis consistent with the clinical and radiologic context, with concordance for benign diagnoses confirmed by a minimum of 3 months of clinical and radiologic follow-up. Secondary outcomes included yield by indication, modality, and adequacy of molecular testing. RESULTS: Among 213 patients (64.3% male; mean age 56±14.5 y), the overall diagnostic yield was 92%. Indications for biopsy comprised 134 suspected carcinomas and 79 suspected benign or lymphoproliferative cases. Diagnostic yield was 91.8% (n=123/134) in suspected carcinoma and 92.4% (n=73/79) in suspected benign or lymphoproliferative disease. Molecular testing was processed in 57 malignant cases, with 89.5% successful completion. CONCLUSION: A structured, MDT-driven endosonographic strategy was associated with high diagnostic and molecular adequacy in this cohort of malignant and benign mediastinal/hilar lymphadenopathy. Given the nonrandomized design and small sample size, findings should be interpreted cautiously. Prospective studies are warranted to further define the optimal role of each modality.