Accurate and safe mediastinal restaging by combined endobronchial and endoscopic ultrasound-guided needle aspiration performed by single ultrasound bronchoscope

Abstract

The aim of this prospective trial was to assess the diagnostic utility of combined endobronchial (EBUS) and endoscopic (EUS) ultrasound-guided needle aspiration by use of a single ultrasound bronchoscope (CUSb-NA) in non-small-cell lung cancer (NSCLC) restaging in patients after induction therapy. In a consecutive group of NSCLC patients with pathologically confirmed N2 disease (clinical stage IIIa and IIIb) who underwent induction chemotherapy, CUSb-NA was performed. All of the patients with negative or suspected for metastases (uncertain) diagnosed by endoscopy underwent subsequently transcervical extended mediastinal lymphadenectomy (TEMLA) as a confirmatory test. From January 2009 to December 2012, 106 patients met the inclusion criteria and underwent restaging CUSb-NA under mild sedation, in whom 286 (mean 2.7, range 2-5) lymph node stations were biopsied, 127 (mean 1.2, range 1-3) by EBUS-transbronchial needle aspiration (TBNA) and 159 (mean 1.5, range 1-4) by EUS-fine needle aspiration (FNA). The CUSb-NA revealed metastatic lymph node involvement in 37/106 patients (34.9%). In 69 (65.1%) patients with negative and uncertain CUSb-NA in 4 (3.8%) out of them, who underwent subsequent TEMLA metastatic nodes were found in 18 patients (17.0%) and there were single lymph nodes found only in one mediastinal station (minimal N2) in 10 (9.4%) out of them. False-positive results were found in 2 (1.9%) patients. In 9 (8.5%) patients CUSb-NA occurred to be false negative in Stations 2R and 4R (only accessible for EBUS), exclusively in small nodes and in 4 (3.8%) patients in Station 5-not accessible for CUSb-NA. The prevalence of mediastinal lymph node metastases in the present study was 51.9%. Diagnostic sensitivity, specificity, total accuracy, positive predictive value and negative predictive value (NPV) of the restaging CUSb-NA were 67.3% (95% CI [confidence interval]-53-79), 96.0% (95% CI-86-99), 81.0% (95% CI-73-87), 95.0% (95% CI-83-99) and 73.0% (95% CI-61-83), respectively. The sensitivity, accuracy and NPV of CUSb-NA were higher compared with EBUS-TBNA and EUS-FNA alone. No complications of CUSb-NA were observed. The CUSb-NA is a reasonable and safe technique in mediastinal restaging in NSCLC patients after induction therapy. Following our data, in patients with negative result of CUSb-NA, a surgical restaging of the mediastinum should be considered.