Confirmatory Mediastinoscopy after Negative Endobronchial Ultrasound-guided Transbronchial Needle Aspiration for Mediastinal Staging of Lung Cancer: Systematic Review and Meta-analysis
José Sanz-Santos, Pere Almagro, Komal Malik, Pablo Martinez-Camblor, Conxi Caro, Ramón Rami-Porta
Abstract
Abstract Rationale Current guidelines for non–small cell lung cancer (NSCLC) mediastinal staging recommend starting invasive staging with endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA). However, the indication to confirm a negative result of EBUS-TBNA by means of video-assisted mediastinoscopy (VAM) before resection differs in every guideline. Objectives Our aim was to evaluate the current evidence regarding the added value of confirmatory VAM after a negative EBUS-TBNA result for mediastinal staging in patients with NSCLC. Methods Systematic searches of studies on EBUS-TBNA for NSCLC mediastinal staging with or without confirmatory VAM but with surgical confirmation of negative results were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis statement in PubMed, SCOPUS, the Cochrane Library, and guidelines from 2005 through November 2021. In the meta-analysis, the sensitivity of confirmatory VAM after a negative EBUS-TBNA result, as well as the sensitivity and negative predictive value of the combination EBUS-TBNA plus confirmatory VAM, alongside the number of confirmatory VAMs required to detect additional N2/3 disease (number needed to treat [NNT]), in patients with a previous negative EBUS-TBNA result were estimated. Results A total of 5,412 articles were found, of which 29 studies were included. Random effects meta-analysis showed a sensitivity of 66.9% (95% confidence interval [CI], 55.8–77.1%) for confirmatory VAM, and 96.7% (95% CI, 95.1–98%) for the combination EBUS-TBNA plus confirmatory VAM. Negative predictive value in studies with confirmatory VAM increased of 79.2% (95% CI, 71.4–86.1%) for EBUS-TBNA alone to 91.8% (95% CI, 87.1–95.5%) for EBUS-TBNA plus confirmatory VAM. The NNT of confirmatory VAM in patients with a previous negative EBUS-TBNA result was 23.8 (95% CI, 19.3–31.2). Conclusions Confirmatory VAM after negative EBUS-TBNA reduces the rate of unforeseen N2/3 disease, but with a high NNT, and it should be recommended only for certain cases yet to be defined.