Existing risk scores not enough to predict AF outcomes after pulmonary vein isolation

Currently available risk scores are unreliable for predicting recurrent atrial fibrillation (AF) after initial radiofrequency pulmonary vein isolation (PVI), a new study has found.

The study included 482 AF patients (mean age 62±9 years, 66 percent men) who had undergone initial PVI and were followed for a median of 16 months. Early recurrence, the primary outcome, was defined as any documented episode of AF, atrial flutter, or atrial tachycardia lasting >30 seconds during the first 90 days after ablation.

The following risk scores were assessed for their ability to predict AF ablation outcomes: APPLE, ATLAS, BASE-AF₂, CAAP-AF, CHADS₂, CHA₂DS₂-VASc, DR-FLASH, HATCH, LAGO, and MB-LATER.

A total of 102 participants experienced early recurrent AF events, yielding a rate of 11 percent. After the initial 90-day follow-up period, 199 more patients experienced recurrent AF, corresponding to a rate of 41 percent. Periprocedural complications occurred in 6 percent of participants, with a slight but significance predominance in women (p=0.048) and in older adults (p=0.04).

Receiver operating characteristic (ROC) curve analysis showed that all risk scores tested, except for HATCH, demonstrated significant predictive value for overall AF recurrence. However, such predictive ability remained weak, with area under the curve (AUC) values ranging from 0.533 to 0.669. Of all risk scores, CAAP-AF, BASE-AF₂, and MB-LATER seemed to have the best performance.

Moreover, only the CHA₂DS₂-VASc and CAAP-AF achieved significant predictive ability for periprocedural complications, with AUCs of 0.616 and 0.615, respectively.

“To improve prediction of AF ablation outcomes, further efforts should be taken to establish novel risk factors, and to improve reproducibility of achieving durable PVI,” the researchers said.