Five-variable nomogram including PR interval and left atrial appendage flow velocity predicts atrial fibrillation recurrence after cryoballoon ablation

TL;DR

This study creates a new tool to predict atrial fibrillation recurrence after a specific heart procedure, using a combination of clinical and imaging data.

Contribution

A novel five-variable nomogram incorporating PR interval and left atrial appendage flow velocity improves AF recurrence prediction after cryoballoon ablation.

Findings

The nomogram outperformed existing models with AUCs of 0.81–0.83 in training and 0.82–0.80 in validation.

Risk groups showed 24-month recurrence-free survival rates of 80%, 45%, and 20% for low, intermediate, and high risk.

The model integrates clinical, electrocardiographic, and echocardiographic features for enhanced prediction.

Abstract

Cryoballoon ablation for atrial fibrillation (AF) carries a non-negligible 1-year recurrence risk, and existing risk models often omit electrocardiographic and echocardiographic features. We aimed to develop and internally validate an AF recurrence prediction model that integrates the PR interval and echocardiographic measures—left atrial dimension (LAD) and left atrial appendage flow velocity (LAAFV)—with clinical characteristics. In this single-center retrospective cohort of 757 patients who underwent first-time cryoballoon ablation (2017–2023), participants were randomly divided into training (70%) and validation (30%) sets. Candidate predictors were selected via least absolute shrinkage and selection operator (LASSO) and random forest. Model performance was evaluated at prespecified 12- and 24-month horizons with time-dependent receiver operating characteristic (ROC) curves, calibration, decision curve analysis, and Kaplan–Meier estimates. Fiveindependently associated predictors were identified: female sex, persistent AF, prolonged PR interval, increased LAD, and reduced LAAFV. A multivariable Cox proportional hazards model was used to construct the nomogram. Head-to-head benchmarking against APPLE, SUCCESS, PAT2C2H, HATCH, BASE-AF2, and CHA₂DS₂-VASc was performed within the same participants via two-sided paired DeLong tests with Holm correction. The nomogram showed strong discrimination (AUC 0.81 and 0.83 in training; 0.82 and 0.80 in validation at 12 and 24 months, respectively) and outperformed all comparators at both horizons (Holm-adjusted P < 0.01). Risk stratification separated the low-, intermediate-, and high-risk groups, with 24-month recurrence-free survival rates of approximately 80%, 45%, and 20%, respectively. By combining structural and functional atrial parameters with clinical features, this tool enhances the prediction of AF recurrence after cryoballoon ablation and may inform individualized postablation management. Prospective multicenter external validation is warranted.

The online version contains supplementary material available at 10.1038/s41598-026-35653-9.

Figures