Authors (including presenting author) :
Joyce WY Chan(1), Aliss TC Chang(1), Rainbow WH Lau(1), Jenny CL Ngai(2), WK Lam(3), Calvin SH Ng(1)
Affiliation :
(1) Division of Cardiothoracic Surgery, Department of Surgery, Prince of Wales Hospital (2) Department of Medicine and Therapeutics, Prince of Wales Hospital (3) Department of Medicine and Therapeutics, North District Hospital
Keyword 2: :
robotic bronchoscopy
Keyword 3: :
electromagnetic navigation bronchoscopy
Keyword 5: :
general anaesthesia
Introduction :
The demand for transbronchial lung nodule biopsy is rising due to increasing incidence of lung nodules discovered incidentally or via CT screening. Electromagnetic navigation bronchoscopy (ENB) biopsy performed under local anaesthesia (LA) or general anaesthesia (GA) provides reasonable diagnostic yield, but the recent development of robotic-assisted bronchoscopy (RAB) may further improve accuracy. There is a lack of local data regarding the diagnostic yield of each method.
Objectives :
The objective of this study is to examine where are any differences in diagnostic yield performing ENB biopsy under GA versus LA, and whether robotic bronchoscopy further improve the diagnostic accuracy.
Methodology :
This is a retrospective study involving 154 patients who underwent transbronchial lung nodule biopsy in two centres in the New Territory East Cluster of Hong Kong in 2024. The biopsy method is divided into 3 groups: ENB under LA, ENB under GA, or RAB under GA. Primary outcome was diagnostic yield using strict ACCP definition, with secondary outcomes of pathology sufficient to guide treatment and sufficiency of tissue for molecular testing. Group comparisons used chi square or z tests for proportions and t based confidence intervals for continuous variables, with multivariable logistic regression adjusting for nodule size.
Result & Outcome :
There were 84 ENB LA, 30 ENB GA, and 40 RAB GA procedures. All GA cases used intraoperative cone beam CT for confirmation, while 91.7% and 100% of ENB-LA cases used fluoroscopy and rEBUS as adjunct respectively. Mean nodule size was 28.5 mm (95% CI 26.1–30.8) for ENB LA, 15.2 mm (12.4–18.0) for ENB GA, and 20.0 mm (16.9–23.0) for RAB GA. Diagnostic yields were 48.1%, 30.0%, and 70.0% respectively (overall p=0.002), with RAB GA significantly higher than both ENB LA (p=0.0068) and ENB GA (p=0.0009). After adjustment for nodule size, RAB GA retained higher odds of definitive diagnosis versus ENB GA (OR 3.61, 95% CI 1.23–10.6), while ENB LA did not differ significantly from ENB GA. Pathology sufficient to guide treatment was achieved in 59.5% (ENB LA), 40.0% (ENB GA), and 76.9% (RAB GA), with size adjusted odds higher for RAB GA versus ENB GA (OR 3.81, 95% CI 1.27–11.4). Among definitive diagnoses of lung cancers, tissue for molecular testing was sufficient in 100% of ENB LA, 55.6% of ENB GA, and 69.2% of RAB GA cases, but size adjusted comparison was not possible due to small sample sizes. The need for additional invasive procedure for biopsy due to non-diagnostic result was 31% for ENB-LA, 8.6% for ENB-GA and 8% for RAB-GA. Procedure duration differed significantly between groups, with ENB-GA being longest (mean 124.8 min), followed by RAB-GA (95.8 min) and ENB-LA (83.0 min; p=0.0037). Within robotic cases, Noah Galaxy® and Auris Monarch® platforms had crude yields of 75.0% and 61.1% respectively (p=0.36), without a clear size adjusted difference. Conclusion RAB under GA demonstrated superior diagnostic yield compared with ENB under GA or LA, even after adjustment for nodule size, at the expense of slightly longer procedure duration. Differences in molecular adequacy and between robotic platforms were not statistically robust and require confirmation in larger, prospective cohorts.
