1351250 - Clinical characteristics and CMR-derived cardiac phenotypes in relation to recurrence of atrial fibrillation after pulmonary vein isolation : a real world population study

Pulmonary vein isolation (PVI) is a frequently applied invasive catheter treatment for atrial fibrillation (AF). AF recurrence after PVI however, remains common. A small number of previous studies reported conflicting results regarding the effect of the clinical characteristics on AF recurrence following PVI. This potential relationship has not been tested in a real-word setting.

Objective:  This study focuses on the relationship of clinical characteristics and cardiac phenotype assessed by cardiac magnetic resonance (CMR) with AF recurrence, in real-world PVI-patients.

Methods:

In 110 consecutive patients scheduled for primo pulmonary vein ablation isolation (PVI), CMR cine imaging was performed to determine left and right atrial volume and global function measures. Left atrial volumes were indexed to body surface area. Clinical characteristics collected prior to PVI included demographics, baseline clinical physical examination, stroke risk factors, and medication.

Results:

Recurrence of AF was detected in 39 patients (35%) after a 90-day blanking period.

No statistically significant difference was observed for any clinical parameter between patients with and without AF recurrence (p-value >0.05 for all, table 1).

Larger LA minimum and maximum volumes and lower LA total ejection fraction, however, were observed in patients with AF recurrence (p-value< 0.01 for all, figure 1). RA minimum and maximum volume and RA ejection fraction did not show statistically significant difference between patients with and without AF recurrence (p-value >0.05 for all, figure 2)

The ROC curves of LA minimum and maximum volumes and LA total ejection fraction showed an area under the curve of 0.70, 0.66, and 0.67 for prediction of AF recurrence, respectively (p< 0.01, for all). The optimal cut-off level for LA minimum and maximum volumes and LA total ejection fraction to predict AF recurrence were 20ml/m², 50ml/m², 51%.

 

Conclusion:

In a real-world AF patient population that underwent PVI, no association was detected between clinical characteristics and AF recurrence following PVI. LA volume-function characteristics, however, are closely related to AF recurrence. The LA maximum volume demonstrated the highest predictive value.

These results highlight the importance of LA volume-function characteristics in prediction of the success of PVI procedures in AF patients.