Left Atrial Appendage Morphology Predicts Atrial Fibrillation Recurrence: The Hidden Risks of Windsock Anatomy.

Background/

Objectives

Left atrial appendage (LAA) morphology has been implicated in atrial fibrillation (AF) recurrence following catheter ablation. However, the predictive value of specific anatomical shapes remains unclear. We aimed to evaluate the association between distinct LAA morphologies and AF recurrence post-ablation.

Methods

In this retrospective, single-center study, 463 patients with AF undergoing first-time catheter ablation were included. Pre-ablation contrast-enhanced cardiac computed tomography was performed to classify LAA morphology into chicken-wing, windsock, cauliflower, and cactus types. Patients were followed for one year, with AF recurrence defined as documented atrial tachyarrhythmia episodes lasting more than 30 s occurring between 3 and 12 months post-procedure. Clinical, anatomical, and procedural factors were analyzed using multivariable logistic regression to identify independent predictors of recurrence.

Results

Among the four morphologies, the windsock-type LAA had the highest recurrence rate at 48.3%, significantly greater than chicken-wing (25.2%), cauliflower (20.8%), and cactus (18.2%) types (p = 0.017). Multivariable analysis confirmed windsock morphology as an independent predictor for AF recurrence (adjusted OR = 2.720, 95% CI: 1.209-6.118; p = 0.016). Additionally, persistent AF (adjusted OR = 1.748, 95% CI: 1.075-2.842; p = 0.024) and antiarrhythmic drug use in the blanking period (adjusted OR = 2.862, 95% CI: 1.689-4.849; p < 0.001) independently increased the risk of recurrence.

Conclusions

Windsock-type LAA morphology significantly predicts increased AF recurrence following catheter ablation, underscoring the importance of morphological assessment in ablation planning. Individualized strategies targeting high-risk LAA morphologies may enhance procedural success and reduce AF recurrence. Future prospective studies are warranted to validate these findings.