Atrial Fibrillation Complexity Parameters Derived From Surface ECGs Predict Procedural Outcome and Long-Term Follow-Up of Stepwise Catheter Ablation for Atrial Fibrillation.

Background: The success rate of catheter ablation for persistent atrial fibrillation (AF) is still far from satisfactory. Identification of patients who will benefit from ablation is highly desirable. We investigated the predictive value of noninvasive AF complexity parameters derived from standard 12-lead ECGs for AF termination and long-term success of catheter ablation and compared them with clinical predictors.
Methods and Results: The study included a training (93 patients) and a validation set (81 patients) of patients with persistent AF undergoing stepwise radiofrequency ablation. In the training set AF terminated in 81% during catheter ablation, 77% were in sinus rhythm after 6 years and multiple ablations. ECG-derived complexity parameters were determined from a baseline 10-s 12-lead ECG. Prediction of AF termination was similar using only ECG (cross-validated mean area under the curve [AUC], 0.76±0.15) or only clinical parameters (mean AUC, 0.75±0.16). The combination improved prediction to a mean AUC of 0.79±0.13. Using a combined model of ECG and clinical parameters, sinus rhythm at long-term follow-up could be predicted with a mean AUC of 0.71±0.12. In the validation set AF terminated in 57%, 61% were in sinus rhythm after 4.6 years. The combined models predicted termination with an AUC of 0.70 and sinus rhythm at long-term follow-up with an AUC of 0.61. Overall, fibrillation-wave amplitude provided the best rhythm prediction.
Conclusions: The predictive performance of ECG-derived AF complexity parameters for AF termination and long-term success of catheter ablation in patients with persistent AF is at least as good as known clinical predictive parameters, with fibrillation-wave amplitude as the best predictor.
(© 2016 American Heart Association, Inc.)