Comparison Between Endobronchial Ultrasound-Guided Transbronchial Node Biopsy and Transbronchial Needle Aspiration: A Meta-Analysis

Introduction: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) can be limited by the inadequacy of intact tissues, especially in patients with lymphoma, sarcoidosis, and lymph node tuberculosis. A novel technique called transbronchial node biopsy (TBNB) by forceps or cryoprobe has been proposed and studied to improve specimen quality and diagnostic yield. We performed a systematic review of studies describing the safety and sensitivity of EBUS-TBNB versus EBUS-TBNA in diagnosing intrathoracic lymphadenopathy/masses. Methods: We systematically searched MEDLINE, Embase, Cochrane, and China National Knowledge Infrastructure to identify studies focusing on the application of EBUS-TBNB for diagnosis of intrathoracic lymphadenopathy. The quality of each study was evaluated using the QUADAS-2 tool. Using inverse-variance (I-V) weighting, we performed a meta-analysis of diagnostic yield estimations. We also reviewed the complications related to the procedure. Results: Thirteen studies were included in the final analysis. The meta-analysis yielded a pooled overall diagnostic yield of 77.80% (939/1,207) for EBUS-TBNA and 86.01% (834/958) for EBUS-TBNB, with an inverse-variance-weighted odds ratio of 3.13 (95% confidence interval [CI], 1.61–6.01; p = 0.0008) and I2 of 82%. The pooled diagnostic yield of EBUS-TBNB versus EBUS-TBNA for the diagnosis of malignancy (including primary lung cancer and extrapulmonary malignancy) was 84.53% (590/698) for EBUS-TBNA and 90.84% (476/524) for EBUS-TBNB, with an I-V-weighted OR of 2.33 (95% CI, 1.15–4.74; p = 0.02) and I2 of 64%. The pooled diagnostic yield of EBUS-TBNB versus EBUS-TBNA for the diagnosis of benignancy was 71.19% (252/354) for EBUS-TBNA and 86.62% (233/269) for EBUS-TBNB, with an I-V-weighted OR of 4.39 (95% CI, 2.00–9.65; p = 0.002) and I2 of 59%. The overall complications included bleeding (n = 11, 0.90%), pneumomediastinum (n = 6, 0.49%), pneumothorax (n = 6, 0.49%), pneumonia (n = 4, 0.33%), respiratory failure (n = 1, 0.08%), and haemoptysis (n = 1, 0.08%). The funnel plot analysis illustrated no major publication bias. Conclusions: EBUS-TBNB improves the overall diagnostic yield of sampling intrathoracic lymphadenopathy and mass lesions relative to EBUS-TBNA. The complication rate of EBUS-TBNB is higher than that of EBUS-TBNA but reportedly lower than that of surgical biopsies.