Faculty Bibliography

BACKGROUND:Genetic testing is a cornerstone in the assessment of many cardiac diseases. However, variants are frequently classified as variants of unknown significance, limiting the utility of testing. Recently, the DeepMind group (Google) developed AlphaMissense, a unique artificial intelligence-based model, based on language model principles, for the prediction of missense variant pathogenicity. We aimed to report on the performance of AlphaMissense, accessed by VarCardio, an open web-based variant annotation engine, in a real-world cardiovascular genetics center. METHODS AND RESULTS/RESULTS:<0.001). Genotype-phenotype concordance was highly aligned using VarCard.io predictions, at 95.9% (95% CI, 92.8-97.9) concordance rate. For 109 variants classified as pathogenic, likely pathogenic, benign, or likely benign by ClinVar, concordance with VarCard.io was high (90.5%). CONCLUSIONS:AlphaMissense, accessed via VarCard.io, may be a highly efficient tool for cardiac genetic variant interpretation. The engine's notable performance in assessing variants that are classified as variants of unknown significance in ClinVar demonstrates its potential to enhance cardiac genetic testing.

Ventricular arrhythmias are commonly associated with hypertrophic cardiomyopathy with and without midventricular obstruction. Although the overall prognosis is relatively good with an annual mortality rate <1%, the propensity to potentially fatal ventricular arrhythmias (ventricular tachycardia) is the most feared complication. Electrical storms are a severe manifestation of ventricular arrhythmias, with poor outcomes. In this report, we present a case of a young patient with non-obstructive hypertrophic cardiomyopathy who presents after a syncopal episode and is found to have an electric storm that is refractory to medical therapy.

Ongoing research attempts to benchmark large language models (LLM) against physicians' fund of knowledge by assessing LLM performance on medical examinations. No prior study has assessed LLM performance on internal medicine (IM) board examination questions. Limited data exists on how knowledge supplied to the models, derived from medical texts improves LLM performance. The performance of GPT-3.5, GPT-4.0, LaMDA and Llama 2, with and without additional model input augmentation, was assessed on 240 randomly selected IM board-style questions. Questions were sourced from the Medical Knowledge Self-Assessment Program released by the American College of Physicians with each question serving as part of the LLM prompt. When available, LLMs were accessed both through their application programming interface (API) and their corresponding chatbot. Mode inputs were augmented with Harrison's Principles of Internal Medicine using the method of Retrieval Augmented Generation. LLM-generated explanations to 25 correctly answered questions were presented in a blinded fashion alongside the MKSAP explanation to an IM board-certified physician tasked with selecting the human generated response. GPT-4.0, accessed either through Bing Chat or its API, scored 77.5-80.7% outperforming GPT-3.5, human respondents, LaMDA and Llama 2 in that order. GPT-4.0 outperformed human MKSAP users on every tested IM subject with its highest and lowest percentile scores in Infectious Disease (80th) and Rheumatology (99.7th), respectively. There is a 3.2-5.3% decrease in performance of both GPT-3.5 and GPT-4.0 when accessing the LLM through its API instead of its online chatbot. There is 4.5-7.5% increase in performance of both GPT-3.5 and GPT-4.0 accessed through their APIs after additional input augmentation. The blinded reviewer correctly identified the human generated MKSAP response in 72% of the 25-question sample set. GPT-4.0 performed best on IM board-style questions outperforming human respondents. Augmenting with domain-specific information improved performance rendering Retrieval Augmented Generation a possible technique for improving accuracy in medical examination LLM responses.

BACKGROUND/UNASSIGNED:Catheter ablation (CA) for ventricular tachycardia (VT) can be a useful treatment strategy, however, few studies have compared CA to medical therapy (MT) in the sarcoidosis population. OBJECTIVE/UNASSIGNED:To assess in-hospital outcomes and unplanned readmissions following CA for VT compared to MT in patients with sarcoidosis. METHODS/UNASSIGNED:Data was obtained from the Nationwide Readmissions Database between 2010 and 2019 to identify patients with sarcoidosis admitted for VT either undergoing CA or MT during elective and non-elective admission. Primary endpoints were a composite endpoint of inpatient mortality, cardiogenic shock, cardiac arrest and 30-day hospital readmissions. Procedural complications at index admission and causes of readmission were also identified. RESULTS/UNASSIGNED: = 0.343). The most common cause of readmission were ventricular arrhythmias (VA) in both groups, however, those undergoing elective CA were less likely to be readmitted for VA compared to non-elective CA. The most common complication in the CA group was cardiac tamponade (4.8 %). CONCLUSION/UNASSIGNED:VT ablation is associated with similar rates of 30-day readmission compared to MT and does not confer increased risk of harm with respect to inpatient mortality, cardiogenic shock or cardiac arrest. Further research is warranted to determine if a subgroup of sarcoidosis patients admitted with VT are better served with an initial conservative management strategy followed by VT ablation.

BACKGROUND:Many patients with mildly to moderately reduced left ventricular ejection fraction (LVEF) who require permanent pacemaker (PPM) implantation do not have a concurrent indication for implantable cardioverter-defibrillator (ICD) therapy. However, the risk of ventricular tachycardia/ventricular fibrillation (VT/VF) in this population is unknown. OBJECTIVE:The aim of this study was to describe the risk of VT/VF after PPM implantation in patients with mildly to moderately reduced LVEF. METHODS:Retrospective analysis was performed of 243 patients with LVEF between 35% and 49% who underwent PPM placement and did not meet indications for an ICD. The primary end point was occurrence of sustained VT/VF. Competing risks regression was performed to calculate subhazard ratios for the primary end point. RESULTS:Median follow-up was 27 months; 73% of patients were male, average age was 79 ± 10 years, average LVEF was 42% ± 4%, and 70% were New York Heart Association class II or above. Most PPMs were implanted for sick sinus syndrome (34%) or atrioventricular block (50%). Of 243 total patients, 11 (4.5%) met the primary end point of VT/VF. Multivessel coronary artery disease (CAD) was associated with significantly higher rates of VT/VF, with a subhazard ratio of 5.4 (95% CI, 1.5-20.1; P = .01). Of patients with multivessel CAD, 8 of 82 (9.8%) patients met the primary end point for an annualized risk of 4.3% per year. CONCLUSION:Patients with mildly to moderately reduced LVEF and multivessel CAD undergoing PPM implantation are at increased risk for the development of malignant ventricular arrhythmias. Patients in this population may benefit from additional risk stratification for VT/VF and consideration for upfront ICD implantation.

BACKGROUND:There is limited information on the mode of arrhythmia initiation in idiopathic ventricular fibrillation (IVF). A non-pause-dependent mechanism has been suggested to be the rule. OBJECTIVES/OBJECTIVE:The aim of this study was to assess the mode and characteristics of initiation of polymorphic ventricular tachycardia (PVT) in patients with short or long-coupled PVT/IVF included in THESIS (THerapy Efficacy in Short or long-coupled idiopathic ventricular fibrillation: an International Survey), a multicenter study involving 287 IVF patients treated with drugs or radiofrequency ablation. METHODS:We reviewed the initiation of 410 episodes of ≥1 PVT triplet in 180 patients (58.3% females; age 39.6 ± 13.6 years) with IVF. The incidence of pause-dependency arrhythmia initiation (prolongation by >20 ms of the preceding cycle length) was assessed. RESULTS:Most arrhythmias (n = 295; 72%) occurred during baseline supraventricular rhythm without ambient premature ventricular complexes (PVCs), whereas 106 (25.9%) occurred during baseline rhythm including PVCs. Nine (2.2%) arrhythmias occurred during atrial/ventricular pacing and were excluded from further analysis. Mode of PVT initiation was pause-dependent in 45 (15.6%) and 64 (60.4%) of instances in the first and second settings, respectively, for a total of 109 of 401 (27.2%). More than one type of pause-dependent and/or non-pause-dependent initiation (mean: 2.6) occurred in 94.4% of patients with ≥4 events. Coupling intervals of initiating PVCs were <350 ms, 350-500 ms, and >500 ms in 76.6%, 20.72%, and 2.7% of arrhythmia initiations, respectively. CONCLUSIONS:Pause-dependent initiation occurred in more than a quarter of arrhythmic episodes in IVF patients. PVCs having long (between 350 and 500 ms) and very long (>500 ms) coupling intervals were observed at the initiation of nearly a quarter of PVT episodes.

BACKGROUND:Prediction of drug-induced long QT syndrome (diLQTS) is of critical importance given its association with torsades de pointes. There is no reliable method for the outpatient prediction of diLQTS. OBJECTIVES/OBJECTIVE:This study sought to evaluate the use of a convolutional neural network (CNN) applied to electrocardiograms (ECGs) to predict diLQTS in an outpatient population. METHODS:We identified all adult outpatients newly prescribed a QT-prolonging medication between January 1, 2003, and March 31, 2022, who had a 12-lead sinus ECG in the preceding 6 months. Using risk factor data and the ECG signal as inputs, the CNN QTNet was implemented in TensorFlow to predict diLQTS. RESULTS:Models were evaluated in a held-out test dataset of 44,386 patients (57% female) with a median age of 62 years. Compared with 3 other models relying on risk factors or ECG signal or baseline QTc alone, QTNet achieved the best (P < 0.001) performance with a mean area under the curve of 0.802 (95% CI: 0.786-0.818). In a survival analysis, QTNet also had the highest inverse probability of censorship-weighted area under the receiver-operating characteristic curve at day 2 (0.875; 95% CI: 0.848-0.904) and up to 6 months. In a subgroup analysis, QTNet performed best among males and patients ≤50 years or with baseline QTc <450 ms. In an external validation cohort of solely suburban outpatient practices, QTNet similarly maintained the highest predictive performance. CONCLUSIONS:An ECG-based CNN can accurately predict diLQTS in the outpatient setting while maintaining its predictive performance over time. In the outpatient setting, our model could identify higher-risk individuals who would benefit from closer monitoring.

OBJECTIVES/OBJECTIVE:We investigated characteristics of left atrial conduction in patients with HCM, paroxysmal AF and normal bipolar voltage. BACKGROUND:Patients with hypertrophic cardiomyopathy (HCM) exhibit abnormal cardiac tissue arrangement. The incidence of atrial fibrillation (AF) is increased fourfold in patients with HCM and confers a fourfold increased risk of death. Catheter ablation is less effective in HCM, with twofold increased risk of AF recurrence. The mechanisms of AF perpetuation in HCM are poorly understood. METHODS:We analyzed 20 patients with HCM and 20 controls presenting for radiofrequency ablation of paroxysmal AF normal left atrial voltage(> 0.5 mV). Intracardiac electrograms were extracted from the CARTO mapping system and analyzed using Matlab/Python code interfacing with Core OpenEP software. Conduction velocity maps were calculated using local activation time gradients. RESULTS: = 0.13, p = 0.03). CONCLUSIONS:Atrial conduction velocity is significantly reduced in patients with HCM and paroxysmal AF, possibly contributing to arrhythmia persistence after catheter ablation.