Risk factors and predictive nomogram model for late lumen restenosis following coronary drug-coated balloon angioplasty: An ambispective analysis

Abstract

Objective: To determine predictors of late lumen restenosis (LLR) after coronary drug-coated balloon (DCB) angioplasty and to construct a nomogram-based clinical model to enhance prediction accuracy for LLR.

Methods: This a misperceived observational study included patients who underwent coronary DCB angioplasty at Changzhou Peoples Hospital (Changzhou, China) between January 2021 and December 2024. Data were collected retrospectively from medical records, and patients were prospectively followed for 6-12 months postoperatively (median follow-up9 months). They were categorized into the LLR group (n = 44) and the non-LLR (NLLR) group (n = 92). Potential risk factors were identified through univariate analysis, and independent predictors were determined using multivariate logistic regression analysis. A nomogram model was developed based on the independent predictors. Internal validation was performed via bootstrap resampling, and the predictive performance and clinical utility of the model were comprehensively evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

Results: Univariate analysis identified several potential risk factors significantly associated with LLR, including preoperative target lesion stenosis severity, history of hypertension, immediate postoperative residual lumen stenosis Greater-Than 30%, and major adverse cardiovascular events (all p vales Less-than 0.05). Logistic regression analysis revealed that preoperative target lesion stenosis severity, history of hypertension, and immediate postoperative residual lumen stenosis Greater-Than 30% were independent predictors of LLR (all p vales P Less-than 0.05). The nomogram model constructed based on these independent predictors demonstrated an area under the curve (AUC) of 0.792 (95% confidence interval: 0.71-0.875) in the bootstrap internal validation. The calibration curve indicated a high concordance between predicted and observed probabilities, and DCA confirmed the good clinical applicability of the model.

Conclusion: Preoperative target lesion stenosis severity, history of hypertension, and immediate postoperative residual lumen stenosis Greater-than 30% emerged as robust predictors of late lumen restenosis following coronary DCB angioplasty, enabling the development of a nomogram model that demonstrated strong discriminatory power (AUC=0.792) and clinical utility through comprehensive validation, thereby providing clinicians with an effective tool for individualized risk stratification and optimized postprocedural management to mitigate LLR incidence in clinical practice.

Keywords: Coronary drug-coated balloon angioplasty, late lumen restenosis, predictor, nomogram, prognostic model.