C110-02 Impact of Apneic Oxygenation on Diagnostic Yield and Safety During Robotic-assisted Bronchoscopy

Abstract Rationale During robotic-assisted bronchoscopy (RAB), atelectasis and motion cause CT-to-body divergence (CTBD), reducing peripheral pulmonary lesion (PPL) localization accuracy. Anesthesia protocols minimize CTBD. Our group previously described using high tidal volumes (VT) and moderate positive end expiratory pressure (PEEP) to reduce atelectasis with apneic oxygenation (AO) to minimize motion. We evaluated whether AO impacted diagnostic yield or complications and identified lesion or patient characteristics most likely to benefit. Methods We analyzed 1,188 lesions from 952 patients who underwent RAB PPL biopsy between April 2022 and December 2023 using an anesthesia protocol with TV 10 mL/kg non-ideal body weight (500-850 mL), PEEP 10-12 cm H2O, and 100% oxygen. AO was used per proceduralist preference via adjustable pressure-limiting (APL) valve set at 20-25 cm H2O with inspiratory hold and 4 L/min oxygen flow. Bayesian logistic regression models were fitted within a causal inference framework to estimate the average and conditional average treatment effects (ATE and CATE) of AO, while adjusting for age, sex, body mass index (BMI), lesion size, location, bronchus sign, nodule shape or composition, pleural or fissure adjacency, cone beam CT (CBCT) imaging use, R-EBUS view, proceduralist, and year. Diagnostic yield was modeled at the lesion level and complications at the procedural level. Results AO demonstrated comparable diagnostic yield (ATE= 0.014 -0.055 to 0.090) and complications (ATE= -0.021 -0.072 to 0.018) compared with procedures without AO. The probability of diagnosis was 0.825 0.789 to 0.854 with AO and 0.810 0.744 to 0.865 without AO. Complication probabilities were low for both AO (0.043 0.028 to 0.062) and without AO (0.063 0.035 to 0.11). AO yielded higher diagnostic probability 65.7% of the time and lower complications 83.8% of the time compared to procedures without AO. Posterior CATE distributions overlapped zero across all subgroups. Modest increases in diagnostic probability with AO were observed for lesions in the left upper lobe (Pr(CATE0)=0.80), right lower (0.67), right upper (0.68), adjacent to fissure (0.71) or pleura (0.76), with eccentric R-EBUS views (0.78), when CBCT was used (0.73). Conclusions Within a standardized anesthesia protocol, adding AO during RAB PPL biopsy did not change overall diagnostic yield or complications. Bayesian subgroup modeling suggests AO may be advantageous for selected PPLs, particularly when an eccentric r-EBUS view is obtained, or if CBCT is used. These results identify procedural contexts where AO may be beneficial and warrant further evaluation. Figure: Comparable diagnostic yield and complications with and without apneic oxygenation This abstract is funded by: None