Factors associated with long-term recurrence of hemoptysis after bronchial artery embolization for bronchiectasis

To identify factors associated with long-term recurrence of hemoptysis after bronchial artery embolization (BAE) for bronchiectasis and to develop a prediction model among patients who remained free of hemoptysis recurrence during the first 30 days after BAE. This study conducted a retrospective cohort analysis of patients who underwent BAE for bronchiectasis between June 2020 and September 2025. Patients with hemoptysis recurrence within 30 days after BAE were excluded. The primary endpoint was long-term recurrence of hemoptysis more than 30 days after BAE. Spearman correlation analysis was used to evaluate intervariable associations. To select predictive factors, we applied LASSO regression to all candidate clinical, radiological, and inflammatory variables. Predictor selection was performed using LASSO regression, and the selected variables were entered into a multivariable logistic regression model fitted with Firth’s penalized maximum-likelihood estimation. Of the 398 initially screened patients, 284 were excluded: 225 with hemoptysis unrelated to bronchiectasis, 4 with technical failure(failure to successfully embolize target artery), 3 with short-term recurrence, 13 with missing clinical data, and 39 lost to follow-up. After these exclusions, 114 patients who remained free of recurrence during the first 30 days were enrolled, with a long-term recurrence rate of 37.7% (43/114). Multivariate logistic regression revealed that non-bronchial systemic arteries (NBSAs) (OR = 2.703, 95% CI 1.033–7.226), number of affected lobes (OR = 1.513, 95% CI 1.023–2.280), white blood cell (WBC) count (on admission) (OR = 1.314, 95% CI 1.072–1.652), and neutrophil–monocyte–lymphocyte ratio (NMLR) (on admission) (OR = 1.251, 95% CI 1.012–1.613) were independent risk factors for long-term recurrence. The model achieved an area under the ROC curve (AUC) of 0.777 (95% CI 0.688–0.865), with an optimism-corrected AUC of 0.704 after bootstrap internal validation. Among patients with bronchiectasis-related hemoptysis who remained free of recurrence during the first 30 days after BAE, NBSAs, number of affected lobes, WBC and NMLR are the main risk factors and the model combined those risk factors provides a better prediction of long-term recurrence of hemoptysis after BAE for bronchiectasis.