Abstract
Background/Objectives: Robot-assisted bronchoscopy (RAB) is a novel platform for sampling peripheral pulmonary nodules (PPNs). To further clarify the role robot-assisted platforms have in diagnosing PPNs, we performed a review of the recent literature. Methods: A systematic review was performed in Medline from 2019 to 2024 using the search terms "robotic bronchoscopy", "diagnostic yield", "sensitivity", and "positive predictive value", alone and in combination. Studies that focused on earlier electromagnetic bronchoscopies were excluded. The patient demographic information, nodule characteristics, intra-procedure imaging modality, biopsy methods, diagnostic yield, sensitivity for malignancy, and adverse outcomes were analyzed. A total of 22 studies were available for the analyses. Results: The diagnostic yield was variable and ranged from 69 to 93%, with a median of 86%. The sensitivity ranged from 69% to 91.7%, with a median of 85%. The effect of the nodule size on the diagnostic yield was variable across the literature. Obtaining an eccentric or concentric view on a radial endobronchial ultrasound (rEBUS) was associated with a higher diagnostic yield than obtaining no view. A nodule appearance on CT imaging and the location were not definitively associated with a higher diagnostic yield. Fine needle aspiration usage ranged from 93.5 to 100%, with a median of 96.95%, while the use of biopsy forceps ranged from 2.7 to 96%, with a median of 69.9%. The most common complication was a pneumothorax, which occurred in 1-5.7% of cases, with a median of 1.6%. Conclusions: Robot-assisted transbronchial biopsies produce diagnostic yields that approach those of transthoracic needle aspirations. The nodule location and appearance may not affect the diagnostic yield. Obtaining a concentric or eccentric view on rEBUS is likely associated with an increased diagnostic yield. Additional prospective studies would better inform practitioners as this technology becomes more widespread.