Comparison of diagnostic yield and safety of three endobronchial ultrasound-guided transbronchial biopsy techniques in diagnosing patients with mediastinal/hilar lymphadenopathy (COFIDENCE Study)

COFIDENCE

Patient / Population Intervention / Exposition Vergleich Endpunkt

In 972 patients with mediastinal or hilar lymphadenopathy undergoing endobronchial ultrasound, transbronchial mediastinal cryobiopsy improved diagnostic yield compared with needle aspiration or forceps biopsy.

N

972Patienten

Design

Open-label, multicentre, three-arm randomised trial at 11 Chinese centres (NCT06262620); adults needing tissue diagnosis of mediastinal or hilar lymphadenopathy were randomised 1:1:1 to EBUS-guided transbronchial needle aspiration, transbronchial mediastinal cryobiopsy or transbronchial forceps biopsy, the latter two through a puncture-dilation tunnel between airway wall and node

Endpunkt

Diagnostic yield; secondary: sensitivity for malignant and benign disease, complications and bleeding

Relevanz

2Wichtig — eine von mehreren Säulen.

ErgebnisDiagnostic yield 86.1% with cryobiopsy vs 72.8% with needle aspiration (P < 0.001) and 77.2% with forceps biopsy (P = 0.003); forceps vs needle 77.2% vs 72.8% (P = 0.20). Sensitivity with cryobiopsy 89.0% for malignant and 93.0% for benign disease. Grade 2 bleeding 11.4% with needle aspiration, 15.7% with cryobiopsy, 12.4% with forceps; no deaths. 972 patients at 11 Chinese centres, all biopsied as allocated.

Hou G, Deng M, Zheng Z. Presented at ERS Congress 2026, ALERT 1, 6 September. Not yet published.

Diskussion & Kritik

The largest randomised comparison of EBUS sampling techniques, and cryobiopsy wins on yield by 13 points over the needle that has been the standard for two decades, with the higher sensitivity for benign disease where the needle is weakest. The likely price is bleeding: grade 2 bleeding ran at 15.7% against 11.4%, a difference the abstract reports without a test and nonetheless folds into a claim of superior safety. Open-label at 11 Chinese centres, with 972 randomised against a registered target of 552, and the recommendation the authors draw is explicitly for high-volume centres with cryobiopsy capability, not for every EBUS suite.