Comparing diagnostic sensitivity of different needle sizes for lymph nodes suspected of lung cancer in endobronchial ultrasound transbronchial needle aspiration: Systematic review and meta‐analysis

BACKGROUND Endobronchial ultrasound transbronchial needle aspiration (EBUS-TBNA) is a safe and minimally invasive procedure for evaluating hilar and mediastinal lymph nodes. The reported sensitivity and specificity of EBUS-TBNA are 95% and 97%, respectively. A comparison of diagnostic sensitivity for lymph nodes suspected of lung cancer according to needle size in EBUS-TBNA is needed. OBJECTIVES To compare the diagnostic sensitivity of the 19-G, 21-G, 22-G and 25-G needles for lymph nodes suspected of non-small cell lung cancer (NSCLC) using EBUS-TBNA. METHODS A literature search from PubMed, EMBASE, LILACS, DOAJ and CENTRAL through October 2020 was performed by two reviewers. The extracted data were evaluated using STATA® and Open Meta Analyst software for meta-analysis with a binary method model to compare sensitivity, specificity and summary receiver operating characteristic curve for each needle size. RESULTS Fourteen studies including 1296 participants were considered for the analysis. The overall sensitivity of EBUS-TBNA was 88.2% (95% CI 84%, 91%) and 93% (95% CI 88%, 95%) for the 19-G needle, 87.6% (95% CI 79.6%, 92.8%) for the 21-G needle and 85% (95% CI 80%, 88%) for the 22-G needle. The overall sensitivity of EBUS-TBNA for diagnosing NSCLC was 88.3% (95% CI, 81%, 93%) and 92.9% (95% CI, 85%, 97%) for the 19-G needle, 89.4% (95% CI 79.4%, 94.8%) for the 21-G needle and 82.1% (95% CI 66%, 91%) for the 22-G needle. CONCLUSION The 19-G, 21-G and 22-G needles present a similarly high diagnostic sensitivity in EBUS-TBNA. The 19-G needle provided better sample adequacy for molecular and immunohistochemical testing, improving diagnostic yield in this subgroup.