Faculty Bibliography

AIMS/UNASSIGNED:We aimed to develop and externally validate a convolutional neural network (CNN) using sinus rhythm electrocardiograms (ECGs) and CHARGE-AF features to predict incident paroxysmal atrial fibrillation (AF), benchmarking its performance against the CHARGE-AF score. METHODS AND RESULTS/UNASSIGNED:We curated 157 192 sinus ECGs from 76 986 patients within the New York University (NYU) Langone Health system, splitting data into training, validation, and test sets. Two cohorts, from suburban US outpatient practices and Greek tertiary hospitals, were used for external validation. The model utilizing the sinus ECG signal and all CHARGE-AF features achieved the highest test set area under the receiver operator curve (AUC) (0.89) and area under the precision recall curve (AUPRC) (0.69), outperforming the CHARGE-AF score alone. Model robustness was maintained in the external US cohort (AUC 0.90, AUPRC 0.67) and the European cohort (AUC 0.85, AUPRC 0.78). Subgroup analyses confirmed consistent performance across age, sex, and race strata. A CNN using ECG signals alone retained strong predictive ability, particularly when simulating missing or inaccurate clinical data. CONCLUSION/UNASSIGNED:Our CNN integrating sinus rhythm ECGs and CHARGE-AF features demonstrated superior predictive performance over traditional risk scoring alone for detecting incident paroxysmal AF. The model maintained accuracy across geographically and clinically diverse external validation cohorts, supporting its potential for broad implementation in AF screening strategies.

• LAAO devices are used for patients with AF intolerant of anticoagulation. • Periprocedural multimodality imaging is crucial for successful device deployment. • LAAs with bifid morphologies are uncommon and pose procedural challenges. • We describe closure of a bifid LAA using 2 side-by-side lobe-and-disk devices.

BACKGROUND:Pulmonary vein (PV) isolation is the cornerstone of radiofrequency (RF) ablation for atrial fibrillation (AF) and PV reconnection is a common cause of recurrent AF. The relationship between PV ostial wall thickness (WT) and durable PV isolation is a matter of ongoing investigation. Additionally, the relationship between catheter impedance and WT is not well understood. We studied the relationship between PV ostial WT, ablation lesion metrics, and PV reconnection. METHODS:16 patients were identified who underwent an initial and redo AF ablation procedure and had a cardiac CTA analyzed using ADAS-3D imaging software performed prior to the initial ablation. Ablation lesion metrics from the initial ablation procedure were collected from the electroanatomic mapping software. Reconnected and isolated PV were identified based on electroanatomic mapping data collected at the redo AF ablation procedure. Patients with reconnected PV exhibited thicker left atrial walls (1.4 mm vs 1.2 mm, P < 0.05) and reconnected veins exhibited thicker ostial walls (1.7 mm, vs 1.5 mm, P < 0.05). LA volume, number of ablation lesions, and ablation lesion time were not significantly different between reconnected and isolated PV. Impedance drop during ablation was greater in patients with reconnected PV compared to patients with isolated PV (- 9.0 Ω vs - 6.6 Ω, P < 0.05). There was no correlation between PV ostial WT and ablation lesion impedance drop. CONCLUSION/CONCLUSIONS:PV reconnection was associated with thicker LA and PV ostial WT. Future studies will examine whether targeting thicker PV ostial tissue with more aggressive lesion metrics or different ablation technology can improve PV isolation and ablationoutcomes.

BACKGROUND:Individuals who report meeting weekly moderate to vigorous physical activity (MVPA) guidelines have lower risk of atrial fibrillation (AF). However existing studies have relied on subjective questionnaires or short-duration (<1 week) objective assessments using accelerometry. The objective of this research was to investigate an association between MVPA levels and the incidence of AF, utilizing long-term, free-living accelerometry data. METHODS:1-year Fitbit data, in addition to survey and electronic health record (EHR) data, were extracted from the NIH All of Us (AoU) research database. Cox proportional hazards regression was used to model the association of average MVPA and incident AF over a five-year follow-up period. RESULTS:, 41±12 complete weeks of Fitbit wear). 97 individuals (0.6%) experienced incident AF in the five-year follow-up period. Every additional hour of MVPA was associated with 8% lower AF risk (HR = 0.92 [0.86,0.99], p=0.02). In a subset of 10533 participants with genomic data, this association persisted after adjustment for AF genetic risk score. CONCLUSIONS:Higher amounts of objectively measured MVPA, measured using free-living, long-term accelerometry data, were inversely associated with risk of incident AF, independent of clinical and genetic risk factors.

OBJECTIVES/OBJECTIVE:Assess the characteristics and management of patients with LAA thrombus despite compliance with oral anticoagulation (OAC). BACKGROUND:Atrial fibrillation guidelines consider 4 weeks of uninterrupted OAC sufficient to avoid transesophageal echocardiography to rule out left atrial appendage thrombus. However, some patients may exhibit persistent thrombus despite compliance with OAC. METHODS:Clinical history, management, and outcomes were reviewed for patients with LAA thrombus on preprocedural TEE presenting for an AF related procedure between 2021 and 2024. RESULTS:Sixty-five (1.8%) of 3653 preprocedural TEEs exhibited LAA thrombus. OAC compliance of at least 4 week was documented in 39 (60%) of these patients, including Apixaban 64%, Rivaroxaban 23%, Warfarin 8%, Dabigatran 5%. Two of these patients (3%) experienced an embolic event and 8 (12%) died during the follow up period. Resolution of LAA thrombus was documented in 12/32 patients, 6 who switched to Dabigatran, 2 to Eliquis, 1 to Warfarin, and 3 remained on Eliquis. LAA-occlusion was successfully performed in seven patients with persistent LAA thrombus. CHADS-VASc 3 or greater, HFrEF, or valvular AF were present in 37/39 of these patients. CONCLUSION/CONCLUSIONS:For 3653 patients who underwent Preprocedural TEE, 39 exhibited LAA thrombus despite compliance with OAC. Switching OAC or maintaining the same agent for longer period of time resolved the thrombus in 31% of cases. LAA-O was effective in cases where the thrombus did not resolve. Patients with non-valvular AF, compliance with OAC > 4 weeks, CHADS-VASc ≤ 2, and normal EF exhibited the lowest probability for not having a thrombus on TEE.

BACKGROUND:Arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited cardiomyopathy associated with a high risk of ventricular arrhythmia (VA). Several animal models have been used to postulate a therapeutic role of the inhibition of the ryanodine 2 receptor via the use of flecainide for this disease. Clinical data describing its use are scarce, however, especially in patients without implantable cardioverter-defibrillators or with left ventricular (LV) involvement. OBJECTIVES/OBJECTIVE:This study sought to report safety and effectiveness long-term, multicenter data on the impact of flecainide therapy on arrhythmic outcomes in patients with a definite diagnosis of ARVC. METHODS:Patients with definite ARVC receiving flecainide at 12 academic institutions were enrolled in the study. Baseline was defined as the time of flecainide initiation. Premature ventricular complex burdens, nonsustained ventricular tachycardia (NSVT) rates, and sustained VA yearly/rates were collected and compared while on and off flecainide. Side effects and flecainide discontinuation were tracked. Analyses were performed in the overall cohort as well as stratifying for genotype (gene positive vs negative; plakohpillin-2 [PKP-2] vs non PKP-2) and for LV involvement. RESULTS:; LV ejection fraction 55.9 ± 7.3%; right ventricular ejection fraction 44.5 ± 10.5% at baseline) were enrolled, with 66 patients (34.6%) showing LV involvement. The median dose of flecainide was 200 mg/d [150-200 mg/d], with 166 patients (86.9%) also taking a beta-blocker. The median follow-up time on flecainide was 4.2 years [1.9-6.3 years]. Flecainide was well tolerated, with a low (7.9%) discontinuation rate. After flecainide initiation, a significant reduction in the 24-hour premature ventricular complex burden and in the rate of nonsustained ventricular tachycardia was observed (2,190 vs 418; P < 0.001; 35.1% vs 21.5%; P = 0.003). For patients with prior VA events, a significant reduction in the amount of VA episodes/y (1.1 [0.4-1.6] episodes/y vs 0 [0-0.3] episodes/y; P < 0.001) was observed. These safety and effectiveness findings were consistent across genotype subgroups, as well as in patients with and without LV involvement. CONCLUSIONS:Flecainide use had a favorable safety profile and was associated with an observed to a significant reduction in arrhythmic burden in patients with ARVC, irrespective of the underlying genotype or LV involvement.

Existing deep learning algorithms for electrocardiogram (ECG) classification rely on supervised training approaches requiring large volumes of reliably labeled data. This limits their applicability to rare cardiac diseases like Brugada syndrome (BrS), often lacking accurately labeled ECG examples. To address labeled data constraints and the resulting limitations of supervised training approaches, we developed a novel deep learning model for BrS ECG classification using the Variance-Invariance-Covariance Regularization (VICReg) architecture for self-supervised pre-training. The VICReg model outperformed a state-of-the-art neural network in all calculated metrics, achieving an area under the receiver operating and precision-recall curves of 0.88 and 0.82, respectively. We used the VICReg model to identify missed BrS cases and hence refine the previously underestimated institutional BrS prevalence and patient outcomes. Our results provide a novel approach to rare cardiac disease identification and challenge existing BrS prevalence estimates offering a framework for other rare cardiac conditions.

BACKGROUND:Left atrial appendage occlusion (LAA-O) with Amulet and Watchman FLX are approved for reducing stroke risk in patients with atrial fibrillation when oral anticoagulation is not tolerated. Real world clinical outcomes reported along with imaging data are needed to help clinicians choose between these two technologies and manage device-related complications. METHODS:The study retrospectively analyzed clinical, transesophageal (TEE), and available computed tomography (CT) data from 364 FLX and 292 Amulet procedures performed at an academic medical center over a 4-year period. RESULTS:were included. TTE LAA-orifice area correlated with CT-derived measurements. There were more late pericardial effusions for Amulet (3.1%) compared to FLX (0.3%), though the majority were conservatively managed. Mean procedure times were similar (FLX 64 ± 24, Amulet 65 ± 21 min) as were the rates of device related thrombus (FLX 1% and Amulet 1.4%). Clinically relevant peridevice leak (PDL) on follow-up TEE imaging was greater for FLX (16%) compared to Amulet (10%). Combined AF ablation-LAA-occlusion procedures exhibited lower rates of PDL and late pericardial effusions compared to solo procedures. CONCLUSIONS:Based on retrospective analysis, an initial strategy with Watchman FLX in patients with favorable LAA anatomy would reduce the risk of late pericardial effusions at the expense of a higher rate of clinically relevant PDL compared to Amulet. Combined AF ablation and LAA-O procedures exhibit less PDL.

BACKGROUND:Recent evidence suggests atrial fibrillation (AF) causes cardiomyopathy due to remodeling driven by both irregular rate and rhythm. Atrial fibrillation (AF) ablation in patients with reduced ejection fraction (EF) ≤ 35% has been shown to improve EF and mortality. It is unknown whether the benefits of AF ablation among patients with reduced EF are affected by the degree of pre-ablation rate control. OBJECTIVES/OBJECTIVE:To evaluate AF ablation echocardiographic outcomes for patients who have EF ≤ 35% with varying degrees of pre-ablation rate control. METHODS:Single-center, retrospective study of patients with EF ≤ 35% undergoing first-time ablation of persistent AF. Primary analyses evaluated the degree to which pre-ablation rate control impacted echocardiographic outcomes. Rates of EF recovery to > 35% were compared at three different cutoffs: 110 bpm, 90 bpm, and 70 bpm. A linear regression analysis was then performed to evaluate whether baseline heart rate (HR) predicted change in EF. RESULTS: = 0.05, p = 0.07). CONCLUSIONS:Catheter ablation of persistent AF in patients with reduced EF frequently resulted in recovery in EF > 35%, irrespective of pre-ablation achieved rate control. While patients with HR > 70 bpm experienced a greater improvement in EF compared to those ≤ 70 bpm, patients with baseline HR below this target still experienced significant EF improvements. Further investigation into irregularity-mediated cardiomyopathy is warranted.