BACKGROUND

The left atrial appendage (LAA) is thought to be the main source of thrombi in patients with atrial fibrillation (AF). The purpose of this study was to describe the LAA orifice diameter, LAA length, and morphologic type of the LAA in Chinese patients with AF as well as to evaluate whether these LAA parameters are associated with a history of stroke in patients with AF from a single center in China.

METHODS

The study population consisted of 219 consecutive patients with drug-refractory, symptomatic paroxysmal, or persistent AF scheduled to undergo radiofrequency catheter ablation in our single center. All patients underwent extensive clinical assessment and multidetector computed tomography to fully explore the anatomy of the LAA.

RESULTS

Of the 219 patients who underwent catheter ablation procedures, chicken wing LAA morphology was found in 114 patients (52.2%), windsock in 52 (23.9%), cauliflower in 29 (13.0%), and cactus in 24 (10.9%). Compared with the windsock LAA morphology, cactus had a larger left atrial diameter ((42.40 ± 3.68) and (37.91 ± 4.32) mm, P = 0.005) and LAA orifice diameter ((27.38 ± 3.70) and (24.14 ± 3.58) mm, P = 0.048). The LAA length was significantly larger in the chicken wing morphology than in the windsock ((37.50 ± 6.74) and (31.33 ± 3.92) mm, P = 0.015) and cauliflower morphologies ((37.50 ± 6.74) and (31.33 ± 3.92) mm, P = 0.015). According to their medical records, 26 patients (11.9%) had suffered a prior stroke. Compared with patients who had no history of stroke, the prior-stroke patients were older (62.04 ± 8.07 and 58.24 ± 9.24, P = 0.047) and there were fewer patients with chicken wing (23.1% and 59.1%, P = 0.001) and more patients with cauliflower (26.9% and 9.8%, P = 0.046). Multivariate Logistic regression analysis demonstrated that age (odds ratio (OR) 1.26; 95% confidence interval (CI) 1.08-1.47; P = 0.003), non-chicken wing morphology (OR 5.82; 95% CI 1.61-21.03; P = 0.007), and LAA orifice diameter (OR 1.25; 95% CI 1.05-1.49; P = 0.014) were independent predictors of stroke after adjusting for all parameters that emerged as potential confounders with univariate analysis.

CONCLUSIONS

LAA analysis can potentially be used to inform guidance on the implication for stroke risk assessment.