Machine Learning-based Prediction of Pacing-Induced Cardiomyopathy in Patients with Atrioventricular Block Underlying Heart Failure with Mildly Reduced or Preserved Ejection Fraction
- 1Chung-Ang University Hospital, Chung-Ang University College of Medicine
Abstract
BackgroundPacing-induced cardiomyopathy (PICM) is a major complication of chronic right ventricular pacing and develops in up to 25% of patients. Reliable prediction of PICM may facilitate individualized pacing strategy selection; however, currently available approaches demonstrate limited discriminative performance. This study aimed to develop and validate machine learning (ML) models for predicting PICM in patients with atrioventricular block (AVB) underlying heart failure with mildly reduced or preserved ejection fraction.
MethodsIn this multicenter prospective study derived from the RISK-PICM registry (January 2002 to January 2026), 125 patients with AVB who underwent pacemaker implantation were analyzed. PICM was defined as a [≥]10% reduction in left ventricular ejection fraction (LVEF) resulting in an LVEF <50%. Four ML algorithms, including logistic regression, support vector machine, random forest, and XGBoost, were trained using stratified 4-fold cross-validation with hyperparameter optimization and model discrimination was evaluated.
ResultsPICM developed in 41 patients (32.8%) during a median follow-up of 4.8 years. Male sex, pre-implant minimum QTc interval, and post-implant maximum QRS duration were significantly associated with PICM in univariable Cox regression. XGBoost achieved the highest discriminative performance (AUROC 0.831; sensitivity 0.805; specificity 0.845). Feature importance analysis identified baseline LVEF and post-implant maximum QRS duration as the most influential features. Model-derived risk stratification significantly differentiated high-risk and low-risk groups on Kaplan-Meier analysis (P < 0.001).
ConclusionsML-based models, particularly XGBoost, demonstrated clinically meaningful performance for PICM and may support individualized pacing strategy selection in patients with AVB and underlying heart failure with mildly reduced or preserved ejection fraction.
Clinical PerspectiveO_ST_ABSWhat Is NewC_ST_ABSO_LIIn pacing-dependent AVB patients with HFmrEF or HFpEF, ensemble ML (XGBoost) predicted PICM with an AUROC of 0.831. C_LIO_LIModel-derived risk stratification separated PICM-free survival (log-rank P<0.001). C_LI
What Are the Clinical ImplicationsO_LIAn ML model combining baseline LVEF, paced QRS duration, and the full parameter set may guide individualized choice between right ventricular and conduction system pacing in AVB patients with HFmrEF or HFpEF. C_LIO_LIUsing paced QRS duration combined with baseline LVEF during temporary pacing in AVB patients with HFmrEF or HFpEF, this risk estimate could inform pacing-strategy selection before permanent implantation, enabling pre-emptive use of conduction system pacing in high-risk patients. C_LI
Related
- Left Bundle-Branch Area Pacing vs Right Ventricular Pacing for Atrioventricular Block and Preserved Left Ventricular Function
- Effect of dapagliflozin on left ventricular longitudinal strain and atrioventricular coupling in heart failure with preserved ejection fraction and chronic kidney disease
- Early Prediction of Cardiomyopathy in Pediatric Cancer Survivors Using Echocardiographic Video-Based Deep Learning.
- Artificial Intelligence-Enhanced Cardiac Magnetic Resonance in Hypertrophic Cardiomyopathy: Diagnosis, Phenotypic Characterization, and Risk Prediction
- Artificial Intelligence-Enhanced Electrocardiography for Detection and Prediction of Hypertrophic Cardiomyopathy across Monogenic and Polygenic Susceptibility
BibTeX
@misc{lee2026machine,
title = {{Machine Learning-based Prediction of Pacing-Induced Cardiomyopathy in Patients with Atrioventricular Block Underlying Heart Failure with Mildly Reduced or Preserved Ejection Fraction}},
author = {J.-W. Lee and U. Do and H. M. Kim and M. J. Choi and S. Y. Shin and S. Park and K.-W. Kang},
howpublished = {medRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.09.26365179},
url = {https://www.medrxiv.org/content/10.64898/2026.10.09.26365179v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Machine Learning-based Prediction of Pacing-Induced Cardiomyopathy in Patients with Atrioventricular Block Underlying Heart Failure with Mildly Reduced or Preserved Ejection Fraction
AU - J.-W. Lee
AU - U. Do
AU - H. M. Kim
AU - M. J. Choi
AU - S. Y. Shin
AU - S. Park
AU - K.-W. Kang
PY - 2026
DA - 2026/10/09
PB - medRxiv
DO - 10.64898/2026.10.09.26365179
UR - https://www.medrxiv.org/content/10.64898/2026.10.09.26365179v1
AB - BackgroundPacing-induced cardiomyopathy (PICM) is a major complication of chronic right ventricular pacing and develops in up to 25% of patients. Reliable prediction of PICM may facilitate individualized pacing strategy selection; however, currently available approaches demonstrate limited discriminative performance. This study aimed to develop and validate machine learning (ML) models for predicting PICM in patients with atrioventricular block (AVB) underlying heart failure with mildly reduced or preserved ejection fraction. MethodsIn this multicenter prospective study derived from the RISK-PICM registry (January 2002 to January 2026), 125 patients with AVB who underwent pacemaker implantation were analyzed. PICM was defined as a [≥]10% reduction in left ventricular ejection fraction (LVEF) resulting in an LVEF <50%. Four ML algorithms, including logistic regression, support vector machine, random forest, and XGBoost, were trained using stratified 4-fold cross-validation with hyperparameter optimization and model discrimination was evaluated. ResultsPICM developed in 41 patients (32.8%) during a median follow-up of 4.8 years. Male sex, pre-implant minimum QTc interval, and post-implant maximum QRS duration were significantly associated with PICM in univariable Cox regression. XGBoost achieved the highest discriminative performance (AUROC 0.831; sensitivity 0.805; specificity 0.845). Feature importance analysis identified baseline LVEF and post-implant maximum QRS duration as the most influential features. Model-derived risk stratification significantly differentiated high-risk and low-risk groups on Kaplan-Meier analysis (P < 0.001). ConclusionsML-based models, particularly XGBoost, demonstrated clinically meaningful performance for PICM and may support individualized pacing strategy selection in patients with AVB and underlying heart failure with mildly reduced or preserved ejection fraction. Clinical PerspectiveO_ST_ABSWhat Is NewC_ST_ABSO_LIIn pacing-dependent AVB patients with HFmrEF or HFpEF, ensemble ML (XGBoost) predicted PICM with an AUROC of 0.831. C_LIO_LIModel-derived risk stratification separated PICM-free survival (log-rank P<0.001). C_LI What Are the Clinical ImplicationsO_LIAn ML model combining baseline LVEF, paced QRS duration, and the full parameter set may guide individualized choice between right ventricular and conduction system pacing in AVB patients with HFmrEF or HFpEF. C_LIO_LIUsing paced QRS duration combined with baseline LVEF during temporary pacing in AVB patients with HFmrEF or HFpEF, this risk estimate could inform pacing-strategy selection before permanent implantation, enabling pre-emptive use of conduction system pacing in high-risk patients. C_LI
ER -