Endobronchial ultrasound-guided sampling: diagnostic yield of needle biopsy and cryobiopsy in addition to needle aspiration

RATIONALE: Endobronchial ultrasound (EBUS)-guided sampling for evaluation of benign conditions and lymphoma is not standardized. Sampling methods remain under study as new tools become available. METHODS: We performed a single-center study with protocolized sampling and processing to evaluate the diagnostic yield of transbronchial needle aspiration (TBNA), Franseen needle biopsy (transbronchial needle biopsy [TBNB]), and cryobiopsy in patients with mediastinal and hilar abnormality with concern for sarcoid, lymphoma, or undifferentiated lymphadenopathy. Following TBNA, exploratory sampling was performed. A tract was created with a TBNB needle pass followed by cryobiopsy. Sampling was subsequently alternated for a total of 3 passes per tool. The primary outcome was diagnostic yield, using a strict criterion modeled from recent American Thoracic Society/American College of Chest Physicians guidelines. Secondary outcomes included procedural outcomes and pathologic assessment of TBNB and cryobiopsy samples. RESULTS: Between January and December 2024, 56 nodes were sampled in 51 patients. Diagnostic yield per node was 86% (48/56). Regarding individual tools, the yield was 50% for TBNA, 73% for TBNB, and 82% for cryobiopsy (TBNA vs cryobiopsy, P = .0001; TBNA vs TBNB, P = .006; TBNB vs cryobiopsy, P = .13). In subgroup analysis, TBNB and cryobiopsy enhanced diagnostic yield in benign conditions (TBNA [59%], TBNB [88%], and cryobiopsy [100%]; TBNA vs cryobiopsy, P = .002; TBNA vs TBNB, P = .01; TBNB vs cryobiopsy, P = .11). For lymphoma, the diagnostic yield per node was 3/10 for TBNA, 5/10 for TBNB, and 6/10 for cryobiopsy (TBNA vs cryobiopsy, P = .001; TBNA vs TBNB, P = .1; TBNB vs cryobiopsy, P = .1). Upon pathology review, TBNB provided larger samples compared to cryobiopsy but TBNB samples contained a larger proportion of blood; these differences translated to similar resultant areas of diagnostic tissue for both tools. Sample usage for immunohistochemistry and special staining was higher for cryobiopsy (61%) vs TBNB (33%). Study sampling was feasible in all patients with a 2% complication rate (1 pneumothorax). CONCLUSIONS: Both TBNB and cryobiopsy enhance the diagnostic yield of EBUS-TBNA in benign conditions. Multiple passes with 3 biopsy tools at an individual lymph node is safe and feasible. Cryobiopsy provides superior sample quality relative to TBNB and superior yield vs TBNA in lymphoma.