Utility of adding oesophageal to endobronchial endosonography when staging lung cancer: a randomised trial.

Introduction: Both combined endobronchial ultrasonography (EBUS) and transoesophageal bronchoscopic ultrasonography (EUS-B) and EBUS alone have been recommended for preoperative mediastinal staging of nonsmall cell lung cancer (NSCLC). However, no randomised study comparing these two methods has been published. The purpose of the present study was to compare the sensitivity of EBUS and that of combined EBUS and EUS-B (EBUS/EUS-B) in terms of detecting N2/N3 disease during staging of NSCLC. Methods: Patients with known or suspected, potentially operable NSCLC were recruited and randomised to undergo EBUS or EBUS/EUS-B under conscious sedation. The primary end-point was a comparison of the sensitivity of EBUS alone and EBUS/EUS-B. Results: A total of 240 patients were enrolled and randomised, among whom 219 (105 EBUS group; 114 EBUS/EUS-B group) were included in the analysis. The sensitivities of EBUS and EBUS/EUS-B in terms of detecting N2/N3 disease were 75.0% and 79.3% respectively (p=0.698). In the EBUS/EUS-B group, only EUS-B yielded diagnostic results in two patients; the sensitivity thus increased from 72.4% to 79.3% on addition of EUS-B to EBUS. Conclusions: The difference in the sensitivities of EBUS alone and EBUS/EUS-B in terms of diagnosing N2/N3 disease was not statistically significant. Although the increase in sensitivity with the addition of EUS-B is modest, it is maximised when EUS-B is used to sample lymph nodes not accessible by EBUS alone.