Catheter Ablation -> Atrial Fibrillation & Atrial Flutter: -> Clinical Trials / Outcomes D-PO03 - Poster Session III (ID 48) Poster

D-PO03-183 - The Challenge Of Predicting Recurrence After Atrial Fibrillation Ablation: Validation Of Seven Risk Scores In A Prospective And Independent Cohort (ID 1145)


Background: Several predictive scores for recurrence of atrial fibrillation (AF) after ablation have been developed, but only few have been validated prospectively.
Objective: We aimed to prospectively compare the predictive value of previously published scores for 3- month and 12-month recurrence of AF after AF ablation.
Methods: Using prospectively collected data from our AF ablation registry, we compared seven previously developed risk scores (HATCH, APPLE, CAAP-AF, BASE-AF2, MB-LATER, CHA2DS2 and CHA2DS2-VASC). The patients underwent radio-frequency ablation (RFA) as either de-novo or redo treatment or cryoablation as de-novo treatment. The dataset consisted of n=665 consecutive patients (pts) for 3-month and n=324 pts for 12-month follow-up (FU).
The HATCH, APPLE, CAAP-AF as well as CHA2DS2 and CHA2DS2-VASC scores were used for prediction of 3-month outcomes. All seven scores were analyzed for their 12-month predictive value.
Results: Of the initial population (n=665, 62% male, mean age 65.6 SD ± 11 years, BMI 27.6 SD ± 4.6 kg/m2, LA 41.1 SD ± 6.7 mm), 54.9% had paroxysmal AF, and 57.9% (n=385) received RFA vs cryoablation. Recurrence rates after 3 and 12 months were 36,2% (n=221) and 40,8% (n=95).
For 3-month FU, only the CAAP-AF Score (AUC 0.559, p=0.015) had a significant but low predictive value. The CHA2DS2 (AUC 0.472, p=0.242), CHA2DS2-VASC (AUC 0.502, p=0,938), HATCH (AUC 0.455, p=0.064) and APPLE (AUC 0.543, p=0.075) scores showed no predictive value.
For 12-month recurrence, the BASE-AF2 (AUC 0.623, p=0.001), MB-LATER (AUC 0.637, p<0.001) and APPLE (AUC 0.591, p=0.019) scores were of significant but modest predictive value. None of the scores was superior. The CHA2DS2 (AUC 0.542, p=0.278), CHA2DS2-VASC (AUC 0.560, p=0,121) HATCH (AUC 0.524, p=0.535), CAAP-AF (AUC 0.567, p=0.083) scores showed no predictive value.
Conclusion: Only the CAAP-AF score had a significant but modest predictive value in predicting 3-month recurrences after AF ablation. For 12-month outcomes, the BASE-AF2, MB-LATER and APPLE scores had a significant but low predictive value. Other previously published scores were of no predictive value.
Predicting recurrences after AF ablation remains a challenge. New simple and robust predictors are needed.