A Prospective, Multicenter Evaluation of Safety and Diagnostic Outcomes With Robotic-Assisted Bronchoscopy

It remains challenging to safely and reliably biopsy peripheral pulmonary lesions (PPLs). Robotic-assisted bronchoscopy (RAB) is gaining adoption for navigation to PPLs. However, evidence from large studies remains limited. What is the clinical safety, navigational success, and diagnostic yield of RAB for biopsy of PPLs in a broad range of patients in a real-world setting? This multicenter, prospective, single-arm study enrolled patients >21 years old with 8-50 mm lung lesions requiring bronchoscopic diagnosis. The primary endpoint was the incidence of the following device- or procedure-related events: (1) pneumothorax requiring intervention; (2) bleeding requiring intervention; or (3) respiratory failure. Secondary endpoints included individual components of the primary endpoint, procedure time, pneumothoraces, radial probe endobronchial ultrasound (R-EBUS) confirmation, conversion to an alternative biopsy procedure, complications, and diagnostic yield (DY). Among 715 patients at 21 sites, 679 met study criteria and underwent RAB (mean age 68.7, 55.4% female, 86.5% White, 77.5% current/past tobacco users). Mean (range) lesion size was 20.9 (7.0-63.0) mm; median (IQR) distance from pleural surface was 5 (0-16) mm. Most lesions were solid (n=587; 86.6%) and within the outer two-thirds of the lung (n=593; 87.5%). The primary endpoint was observed in 26 (3.8%) patients (19 pneumothorax, 7 bleeding, 0 respiratory failure). Users reported that RAB reached the lesion in 670/679 (98.7%) cases, and lesion location was confirmed with R-EBUS in 607/662 (91.7%) cases; sampling through the bronchoscope was performed in 675/679 (99.4%) cases. Prevalence of malignancy was 64.1% through 12 months. Adjudicated DY was 61.6% when calculated with the American Thoracic Society (ATS)/American College of Chest Physicians (ACCP) definition for strict reporting criteria. Sensitivity for malignancy was 78.8%. This largest multicenter prospective study of RAB to date demonstrated that RAB-guided sampling of PPLs is safe and compares favorably to results from sizable non-robotic bronchoscopy studies. NCT04182815.