Ultrathin bronchoscopy with radial endobronchial ultrasound and rapid on-site evaluation for the diagnosis of peripheral pulmonary lesions: a multicentre randomized controlled factorial trial.

Background: The routine use of computed tomography and lung cancer screening has increased the incidence of peripheral pulmonary lesions (PPLs). Sampling may be needed for some nodules. Many new technologies are available to improve the diagnostic performance of bronchoscopy for the sampling of PPLs, but few comparative trials exist. The objective of this study was to compare the diagnostic performance of bronchoscopy with radial endobronchial ultrasound (rEBUS) using an ultrathin bronchoscope (BF-MP190F, Olympus) with a non-ultrathin bronchoscope, and to compare the diagnostic performance of bronchoscopy with and without rapid on-site evaluation (ROSE).

Objective: Does diagnostic performance differ between ultrathin and non-ultrathin bronchoscopes with rEBUS? What impact does ROSE have on diagnostic performance of rEBUS?

Methods: This pragmatic, multicentre, 2x2 factorial, randomized controlled trial involved adult patients with PPLs (mean diameter <5 cm) and radiographic N0 disease, referred for bronchoscopy. The study was powered to detect a 20% improvement in the primary outcome of difference in diagnostic yield (DY) between ultrathin and non-ultrathin bronchoscopes and between procedures with and without ROSE. Secondary outcomes included sensitivity for malignancy, complications and procedure duration.

Results: Of 215 patients assessed, 186 were randomized, and 181 analyzed. Malignancy prevalence was 84%. No significant differences in DY or sensitivity for malignancy were found between non-ultrathin and ultrathin bronchoscopes: 65.6% vs 58.2% (diff -7.3%, p=0.36) and 84.3% vs 74.3% (diff -10.0%, p=0.21). Similarly, no significant differences were observed with or without ROSE: 60.4% vs 63.5% (diff -3.1%, p=0.76) and 80.3% vs 78.3% (diff 2.0%, p=0.83).

Conclusions: We could not identify a difference in diagnostic yield or sensitivity for malignancy for the diagnosis of PPLs between an ultrathin bronchoscope and a non-ultrathin bronchoscope and between ROSE and no ROSE. The study was underpowered to detect smaller but potentially clinical meaningful differences.