PDL
In 51 cases (60.7%) there was no PDL detectable by CCTA. A leak with width of <3 mm was present in 31 cases (36.9%), and a leak width of 3 to 4.9 mm was present in 2 cases (2.4%). None were found to have a severe leak >5 mm in width. Baseline characteristics according to the presence or absence of PDL are listed in Table 3.
Preoperative troponin, maximum diameter of LAA orifice, and LAAV differed between with and without PDL groups, while age, sex, body mass index (BMI), presence of hypertension/diabetes, AF pattern, preoperative NT-ProBNP, fibrinogen, D-dimer, LVEF, LAV, or device diameter did not. On univariate logistic regression analysis, predictors of PDL were maximum diameter of LAA orifice, preoperative LAAV and device size, while age, sex, BMI, presence of hypertension/diabetes, AF pattern, NT-ProBNP, troponin, fibrinogen, D-dimer, LVEF, or LAV were not (Table 4).
Further multivariate logistic regression analysis indicated predictor of PDL was maximum diameter of LAA orifice (Table 5), which was not related with LAAV and device size. Mean maximum diameter of LAA orifice for the 84 cases was 27.8 ± 5.2 mm. Maximum diameter of LAA orifice was greater in cases with PDL compared to those without PDL (25.8 mm vs. 30.9 mm, P = 0.009). Increasing maximum diameter of LAA orifice was positively correlated with PDL (OR, 1.31; 95% CI, 1.11-1.55; P = 0.002). On ROC analysis (Figure 3), AUC for the presence versus absence of PDL was 0.766 (95% CI: 0.659 to 0.872, P < 0.001). The Youden index (0.501) was the largest when maximum diameter of LAA orifice was 28.2 mm (sensitivity of 69.7% and specificity of 80.4%), suggesting that maximum diameter of LAA orifice >28.2 mm is more likely to demonstrate PDL.