Background: The data on iatrogenic atrial septal defect (iASD) after left atrial appendage closure (LAAC), especially intracardiac echocardiography (ICE)-guided LAAC, are limited. Compared with transesophageal echocardiography (TEE)- or digital subtraction angiography (DSA)-guided LAAC, the transseptal puncture (TP) ICE-guided LAAC is more complicated. Whether or not ICE-guided TP increases the chances of iASD is controversial. We investigate the incidence, size, and clinical outcomes of iASD after ICE-guided LAAC. Methods: A total of 177 patients who underwent LAAC were enrolled in this study and were assigned to the ICE-guided group (group 1) and the TEE- or DSA-guided group (group 2). Echocardiography results and clinical performances at months 2 and 12 post-procedure were collected from the electronic outpatient records. Results: A total of 112 and 65 patients were assigned to group 1 and group 2, respectively. The incidence of iASD at follow-up (FU) month 2 was comparable between the groups (21.4% in group 1 vs. 15.4% in group 2, p = 0.429). At month 12 of FU, the closure rate of iASD was comparable to that of group 2 (70.6% vs. 71.4%, p = 1.000). No right-to-left (RL) shunt was observed among the iASD patients during the FU. Numerically larger iASD were observed in group 1 patients at month 2 of FU (2.8 ± 0.9 mm vs. 2.2 ± 0.8 mm, p = 0.065). No new-onset of pulmonary hypertension and iASD-related adverse events were observed. Univariable and multivariable logistic regression analysis showed that ICE-guided LAAC was not associated with the development of iASD (adjusted OR = 1.681; 95%CI, 0.634–4.455; p = 0.296). Conclusions: The ICE-guided LAAC procedure does not increase the risk of iASD. Despite the numerically large size of the iASD, it did not increase the risk of developing adverse complications.

BackgroundA one-stop procedure involving catheter ablation and left atrial appendage occlusion (LAAO) is an option for high-risk atrial fibrillation patients. Few studies have reported the efficacy and safety of cryoballoon ablation (CBA) combined with LAAO, and no studies have compared the combination of LAAO with CBA or radiofrequency ablation (RFA).MethodsA total of 112 patients were enrolled in the present study; 45 patients received CBA combined with LAAO (group 1), and 67 patients received RFA combined with LAAO (group 2). Patient follow-up was performed for 1 year to detect peri-device leaks (PDLs) and safety outcomes (defined as a composite of peri-procedural and follow-up adverse events).ResultsThe number of PDLs at the median 59 days follow-up was comparable between the two groups (33.3% in group 1 vs. 37.3% in group 2, p = 0.693). Safety outcomes were also comparable between the two groups (6.7% in group 1 vs. 7.5% in group 2, p = 1.000). Multivariable regression showed that PDLs risk and safety outcomes were all similar between the two groups. Subgroup analysis of PDLs indicated no significant differences. Follow-up safety outcomes were related to anticoagulant medication, and patients without PDLs were more likely to discontinue antithrombotic therapy. The total procedure and ablation times were all significantly shorter for group 1.ConclusionWhen compared with left atrial appendage occlusion combined with radiofrequency, left atrial appendage occlusion combined with cryoballoon ablation has the same risk of peri-device leaks and safety outcomes, but the procedure time was significantly reduced.

Background A number of models have been reported for predicting atrial fibrillation (AF) recurrence after catheter ablation. Although many machine learning (ML) models were developed among them, black-box effect existed widely. It was always difficult to explain how variables affect model output. We sought to implement an explainable ML model and then reveal its decision-making process in identifying patients with paroxysmal AF at high risk for recurrence after catheter ablation. Methods Between January 2018 and December 2020, 471 consecutive patients with paroxysmal AF who had their first catheter ablation procedure were retrospectively enrolled. Patients were randomly assigned into training cohort (70%) and testing cohort (30%). The explainable ML model based on Random Forest (RF) algorithm was developed and modified on training cohort, and tested on testing cohort. In order to gain insight into the association between observed values and model output, Shapley additive explanations (SHAP) analysis was used to visualize the ML model. Results In this cohort, 135 patients experienced tachycardias recurrences. With hyperparameters adjusted, the ML model predicted AF recurrence with an area under the curve of 66.7% in the testing cohort. Summary plots listed the top 15 features in descending order and preliminary showed the association between features and outcome prediction. Early recurrence of AF showed the most positive impact on model output. Dependence plots combined with force plots showed the impact of single feature on model output, and helped determine high risk cut-off points. The thresholds of CHA2DS2-VASc score, systolic blood pressure, AF duration, HAS-BLED score, left atrial diameter and age were 2, 130 mmHg, 48 months, 2, 40 mm and 70 years, respectively. Decision plot recognized significant outliers. Conclusion An explainable ML model effectively revealed its decision-making process in identifying patients with paroxysmal atrial fibrillation at high risk for recurrence after catheter ablation by listing important features, showing the impact of every feature on model output, determining appropriate thresholds and identifying significant outliers. Physicians can combine model output, visualization of model and clinical experience to make better decision.