Cryoballoon ablation of persistent atrial fibrillation: Who will benefit from additional posterior roof ablation?

Introduction: Pulmonary vein isolation (PVI) via cryoballoon ablation is widely used in patients with paroxysmal and persistent AF. The overall effectiveness of cryoablation is in the range of 70–80%. Long-term results in persistent AF (PersAF) are inferior to those reported in paroxysmal AF. For improving outcome in PersAF patients, additional posterior roof ablation (PRA) has been proposed. However, the optimal selection of patients in whom supplementary PRA may provide significant benefit has not yet been established. The aim of this study was to identify these patients based on an AF ablation cohort at our institution.

Methods: Consecutive patients with PersAF who underwent PVI with cryoballoon ablation at our institution since 2013 were included. All patients were interviewed upon admission, and atrial size and ventricular function (TAPSE for the right ventricle) were assessed by echocardiography. Left atrial size was recorded as left atrial index (LAI=left atrial area/body surface area). PRA was performed at the discretion of the investigator. Continuous parameters were dichotomized using ROC curves. Further, dichotomized parameter were analysed in a univariate Cox regression model, and, if found to be significant, were included in a multivariable analysis. Based on significant predictors, a prediction score was calculated. Finally, the impact of PRA was analysed for each score value. The study endpoint was the first documented recurrence of atrial tachyarrhythmia after PVI. The statistical analysis was performed using SPSS v.27.

Results: A total of 656 patients (male=453 [69.1%], age=65 [IQR =57-71], time since first diagnosis [TFD]=24 [6-60] months, LAI=8. 64 [7.20-9.39], LVEF=60 [57-62] %, TAPSE=22 [19-25] mm) were included. Reduced LVEF (<55%) was found in 147 patients and reduced TAPSE (<19 mm) in 210. PRA was performed in 122 patients. The cut-off of LAI was identified to be 7.85 and the cut-off of TFD was 9 months. Within the maximum follow-up of 5 years, a total of 308 (47%) patients reached the endpoint. Multivariable Cox regression analysis revealed TFD >9 months (HR=1.589 [1.222-2.066]; p=.001), LAI >7.85 (HR=1.574 [1.223-2.027]; p=.001), and TAPSE <19 mm (HR=1.329 [1.048-2.685]; p=0.019) to be significant predictors. Notably, reduced LVEF was not associated with outcome (HR=1.176 [0.907-1.525]; p=.222).Based on dichotomized TAPSE, LAI, and TFD values, a prediction score was calculated. The free of events time (months) is given in Table. Patients with a score of 1 and 2 derived the maximum benefit from PRA. In patients with a score of 0, i.e. TFD ≤9 months, LAI ≤7.85, and preserved TAPSE , PVI provided best results irrespective to PRA. The patients with long TFD, increased LAI, and reduced TAPSE (score=3) had poorest outcome and also did not benefit from PRA. In these patients new studies on optimal therapy should be performed. 

Conclusion. Left atrial size, time from first diagnosis, and TAPSE were revealed to be predictive of the outcome after cryoballoon PVI. Reduced LVEF was irrelevant for the prediction of outcome. The constructed score allowed preprocedural identification of the patients who possibly may benefit from additional PRA