P1056Left atrial and left atrial appendage ejection fraction assessed by preprocedural cardiac CT are useful parameters for predicting the outcome of pulmonary vein isolation

Abstract

Abstract Introduction Pulmonary vein isolation (PVI) is an effective but complex treatment option for atrial fibrillation (AF). Therefore preprocedural outcome prediction is of special interest. Left atrial volume (LAVmax) is a commonly used predictor for recurring arrhythmia. Several studies have investigated different parameters for the prediction of sustained rhythm control. We hypothesized that left atrial and left atrial appendage ejection fraction (LAEF & LAAEF) assessed by high-resolution CT-imaging are even more sensitive predictors of the outcome of PVI than LAVmax. Methods All patients who underwent PVI between 2015 and 2018 with available preinterventional cardiac CT were included in this retrospective study and separated into 2 groups: Group A comprises all patients with sinus rhythm (SR) at follow-up and group B all patients in whom recurrence of AF was observed. Volumetric analysis of the left atrium was performed in ventricular systole (30%-Phase of the cardiac cycle) and diastole (0%-Phase). Obtained data were used to calculate left atrial and left appendage ejection fractions (LAEF & LAAEF). Success of pulmonary vein isolation was defined as clinical freedom of symptomatic AF together with sinus rhythm on the available Holter-ECG recordings during follow-up examinations. Uni- and multivariate logistic regression models and NAMS RMLE tests were used to compare LAVmax to these functional parameters. Results In total 152 patients with symptomatic paroxsymal or persistent AF underwent PVI at our hospital from 2015 to 2018. Due to inconsistencies in archiving in our PACS-system, 98 patients had to be excluded from analysis. Four patients were excluded due to motion artifacts. 50 patients were included in the final study (41 patients without and 9 patients with recurring AF on average 254 days after ablation). Significant differences in means were found for all assessed parameters. LAEF (accuracy 94%, sensitivity 67%) and LAAEF (accuracy 90%, sensitivity 67%) had a higher sensitivity than LAVmax (accuracy 86%, sensitivity 33%), though not significant in this study population (p = 0.18). LAVmax and LAEF in combination improved sensitivity significantly from 33% to 78% (p = 0.046). Measurements on cardiac-CT showed an excellent interobserver-reliability. Conclusion Reduced LAEF and LAAEF were found to be significant predictors of the outcome of PVI. Furthermore we found a trend that these functional parameters might be more sensitive than LAVmax. Thus we propose that left atrial function, assessed with preprocedural cardiac CT, offers important prognostic information for successful PVI. Abstract Figure. Surface rendering of LA and LAA