A23-17 A Novel Model for Determination of Procedural Failure in Bronchoscopic Lung Volume Reduction Using Medical Comorbidities

Abstract Rationale Chronic obstructive pulmonary disease (COPD) is a major cause of worldwide morbidity and mortality with limited treatment options to improve lung function. Bronchoscopic lung volume reduction (BLVR) has demonstrated promise in restoring lung function and improving quality of life, however, results have been variable despite patients meeting typical accepted parameters of pulmonary function testing and quantitative lung imaging. Previously, we demonstrated that medical comorbidities may influence procedural outcomes. In this study, we have expanded to include an international cohort and describe a model to help guide clinical decision making for patients who may undergo BLVR. Methods This is a retrospective analysis of patients who have undergone BLVR at the University of Chicago and University Medical Center Gronigen between 2018 and 2025. Procedural success was defined as a sustained symptomatic improvement at least 6 months following procedure without downstream complication necessitating permanent valve removal or mortality. Patients who underwent intentional bilateral BLVR or BLVR for post-transplant native lung hyperinflation were excluded to preserve generalizability. A logistic regression model was used to determine the effects of multiple comorbidities on the outcome of procedural failure. Model fit and utility were tested using a receiver operator curve (ROC), calibration plot, and decision curve analysis. Results A total of 259 patients were included as observations in the final model, having available data for all parameters included. Adjusting for age, race, sex and Body-Mass Index (BMI), failure was associated with anemia (Odds Ratio (OR) 2.73, p = 0.023, 95% Confidence Interval (CI) 1.15, 6.48) and Body-Mass-Index, Obstruction, Dyspnea and Exercise (BODE) index (OR 1.41, p = 0.000, 95% CI 1.20, 1.67). Estimated Glomerular Filtration Rate (EGFR) (OR 0.98, p = 0.023, 95% CI 0.957, 0.997) and obstructive sleep apnea (OSA) (OR 0.269, p = 0.024, 95% CI 0.085, 0.841) were found to be factors associated with procedural success. The area under the ROC curve was 0.6911 and a calibration plot for model failure was acceptable. The decision curve analysis demonstrates that our model outperforms a “treat all strategy” at typical ranges of acceptable patient risk given the severity of COPD (0-50%). The decision curve analysis suggests that our model will likely successfully discourage some high-risk patients from pursuing BLVR, resulting in improved outcomes. Conclusion Medical comorbidities such as anemia, BODE, EGFR and OSA influence rates of success and failure in BLVR. A predictive logistic model using these patient factors may have clinical utility in pre-operative risk stratification and patient selection. This abstract is funded by: None