PhD Scientific Days 2026

Cardiovascular Medicine and Research 1.

A Novel LEAP-AF Score for Predicting Recurrence Following Atrial Fibrillation Ablation: Development and Validation in a Large Single-Center Cohort

Name of the presenter

Orbán, Gábor

Institute/workplace of the presenter

Heart and Vascular Center, Semmelweis University

Authors

Gábor Orbán¹, Márton Boga¹, Nándor Szegedi¹, Zoltán Salló¹, Klaudia Vivien Nagy¹, István Osztheimer¹, Péter Perge¹, Szilvia Herczeg¹, Ferenc Komlósi¹, Patrik Tóth¹, Edit Tanai¹, Anna Padisák¹, Béla Merkely¹, László Gellér¹
1: Heart and Vascular Center, Semmelweis University

Text of the abstract

Introduction: Pulmonary vein isolation (PVI) represents the most effective strategy for rhythm control in atrial fibrillation (AF). However, despite ongoing advances in ablation techniques, long-term outcomes following PVI remain inconsistent. Currently available risk scores for predicting post-ablation AF recurrence demonstrate only modest discriminatory performance.
Aims: To determine clinical and procedural predictors of recurrence after ablation and to develop a simple, validated model for individualized risk assessment.
Method: We retrospectively analyzed 2,455 consecutive AF patients who underwent catheter ablation at our center between 2006 and 2023. Clinical, echocardiographic, and procedural data were collected. AF recurrence was defined as any atrial arrhythmia lasting >30 s after a 90-day blanking period, documented on ECG or Holter monitoring. Predictors were identified using LASSO Cox regression with multiple imputation, followed by stepwise backward elimination. Regression coefficients (β) were multiplied by 10 and rounded to derive the LEAP-AF score. The model was internally validated using 10,000 bootstrap samples. Predictive performance was assessed at 36 months; discrimination by ROC AUC and calibration by the observed-to-expected (O:E) ratio. For visualization, patients were divided into two groups based on predicted recurrence probabilities, using the median as the cut-off (low- vs high-risk groups).
Result: AF recurred in 985 patients (40%) after a median follow-up of 321 days (IQR 146–575). Six independent predictors constituted the LEAP-AF score: female sex (2 points), age ≥60 years (1), persistent AF (2), AF at the start of the procedure (4), left atrial diameter ≥40 mm (2), and E-wave velocity ≥80 cm/s (1). The LEAP-AF score showed moderate predictive performance (C-statistic = 0.62, O:E = 1.16), with 36-month arrhythmia-free survival of 63.7% in the low-risk group (0–4 points) compared with 45.2% in the high-risk group (≥5 points).
Conclusion: The LEAP-AF score provides a simple, reproducible tool for assessing post-ablation recurrence risk. Although its predictive performance is moderate, it enhances risk stratification compared with existing models and may aid individualized follow-up after AF ablation.
Funding: Predoctoral Scholarship; Gedeon Richter Excellence PhD Scholarship of Gedeon Richter; Kerpel Scholarship 2025