Predicting AFib recurrence after catheter ablation remains a challenge
Risk scores designed to predict atrial fibrillation (AFib) after catheter ablation are largely ineffective, according to new data published in Circulation: Arrhythmia and Electrophysiology.
"While several risk scores for predicting AFib recurrence post-ablation have been proposed, it is unknown which is most accurate as few have undergone external validation within a single cohort of patients with well-defined outcomes," wrote lead author Lisa Y.W. Tang, PhD, with the Data Science Institute at the University of British Columbia in Vancouver, and colleagues.
Tang et al. used data from the CIRCA-DOSE trial to track more than 300 patients given an implantable cardiac monitor before catheter ablation for arrhythmia detection. They used to 11 different models to predict post-ablation atrial tachyarrhythmia recurrence from 91 to 365 days after the ablation.
Overall, the group found, the models fell short in the EP lab.
The team did find that the SUCCESS and APPLE scores were highly specific for predicting recurrence, but they were insensitive. In contrast, CAAP-AF and CHADS2 were highly sensitive, but did not have high specificity.
The M-HATCH and ATLAS scores had the overall best performance, but Tang and colleagues still saw room for improvement.
“All models performed poorly and had objectively limited clinical utility," they wrote. "There remains a need to develop generalizable tools for prediction of post-ablation AFib recurrence."
The study did have limitations. For example, the study population had paroxysmal AFib and the ablation strategy was limited to pulmonary vein isolation (PVI).
“Our results cannot be generalized to patients with persistent AFib or those undergoing more extensive ablation beyond PVI," the authors wrote. "Moreover, despite the primary goal of ablation being to improve symptoms and quality-of-life, none of the evaluated scores predict symptomatic improvement.”
Read the research letter here.