Assays of Variant Effect and Automated Patch Clamping Improve KCNH2-LQTS Variant Classification and Cardiac Event Risk Stratification.

BACKGROUND: Long QT syndrome is a lethal arrhythmia syndrome, frequently caused by rare loss-of-function variants in the potassium channel encoded by KCNH2. Variant classification is difficult, often because of lack of functional data. Moreover, variant-based risk stratification is also complicated by heterogenous clinical data and incomplete penetrance. Here we sought to test whether variant-specific information, primarily from high-throughput functional assays, could improve both classification and cardiac event risk stratification in a large, harmonized cohort of KCNH2 missense variant heterozygotes. METHODS: We quantified cell-surface trafficking of 18 796 variants in KCNH2 using a multiplexed assay of variant effect (MAVE). We recorded KCNH2 current density for 533 variants by automated patch clamping. We calibrated the strength of evidence of MAVE data according to ClinGen guidelines. We deeply phenotyped 1458 patients with KCNH2 missense variants, including QTc, cardiac event history, and mortality. We correlated variant functional data and Bayesian long QT syndrome penetrance estimates with cohort phenotypes and assessed hazard ratios for cardiac events. RESULTS: Variant MAVE trafficking scores and automated patch clamping peak tail currents were highly correlated (Spearman rank-order ρ=0.69; n=433). The MAVE data were found to provide up to pathogenic very strong evidence for severe loss-of-function variants. In the cohort, both functional assays and Bayesian long QT syndrome penetrance estimates were significantly predictive of cardiac events when independently modeled with patient sex and adjusted QT interval (QTc); however, MAVE data became nonsignificant when peak tail current and penetrance estimates were also available. The area under the receiver operator characteristic curve for 20-year event outcomes based on patient-specific sex and QTc (area under the curve, 0.80 [0.76-0.83]) was improved with prospectively available penetrance scores conditioned on MAVE (area under the curve, 0.86 [0.83-0.89]) or attainable automated patch clamping peak tail current data (area under the curve, 0.84 [0.81-0.88]). CONCLUSIONS: High-throughput KCNH2 variant MAVE data meaningfully contribute to variant classification at scale, whereas long QT syndrome penetrance estimates and automated patch clamping peak tail current measurements meaningfully contribute to risk stratification of cardiac events in patients with heterozygous KCNH2 missense variants.

The impact of mavacamten dosing on wall thickness regression: an insight from longer term follow-up based on genetic profile.

INTRODUCTION: We have previously reported that genetically positive patients have a more profound early decrease in provocable left ventricular outflow tract gradient compared to genetically negative patients utilizing mavacamten in the first 12 weeks of therapy. METHODS AND RESULTS: In this current analysis, we found that genetically positive patients have less favorable remodeling as measured by left ventricular wall thickness regression when evaluated long-term as compared to genetically negative patients, despite an overall better early response to mavacamten. The majority of genetically positive patients were maintained on only 2.5 mg of mavacamten due to early robust response. CONCLUSION: We hypothesize that this lower dosing attenuated the long-term benefit of mavacamten in genetically positive patients. We believe that the long-term benefit of mavacamten on positive cardiac remodeling is dose-dependent and not solely related to the magnitude of left ventricular outflow gradient decrease.

Medical Therapy and Clinical Outcomes in Cardiac Sarcoidosis Patients With Systolic Heart Failure.

BACKGROUND: Cardiac sarcoidosis (CS) may result in systolic heart failure (heart failure with reduced ejection fraction [HFrEF]), but its response to guideline-directed medical therapy (GDMT) remains uncertain. METHODS AND RESULTS: We investigated 881 patients evaluated for CS to identify those with diagnosed CS, left ventricular ejection fraction (LVEF) ≤40% at diagnosis, and follow-up echocardiogram within 11-24 months. Demographics, LVEF, GDMT as quantified by Kansas City Medical Optimization (KCMO) score, and immunosuppressive treatment were recorded. The primary outcome was a composite of event-free survival (unplanned heart failure hospitalization, left ventricular assist device [LVAD]/heart transplant, or death). Seventy-nine (9%) CS patients met the inclusion criteria (35% female, median age 57 years, mean LVEF 30.9%, median New York Heart Association class II [46%], mean number of GDMT agents 1.7, and mean KCMO score 31.8). Most (87%) were treated with immunosuppressive treatment. At follow-up (median 16 months), the mean number of GDMT agents increased to 2.2 (P=0.02), and the mean KCMO score to 70.1 (P<0.001). Mean LVEF improved to 39.9% (excluding LVAD/transplant; P<0.001) and the change in LVEF was correlated with follow-up KCMO score (P<0.001). The primary outcome occurred in 13 (16%) patients and differed by KCMO score (log-rank P<0.001), but not by immunosuppressive treatment (log-rank P=0.36). CONCLUSIONS: GDMT optimization is associated with better cardiac remodeling and clinical outcomes in CS patients with HFrEF.

A multi-omics atlas of sex-specific differences in obstructive hypertrophic cardiomyopathy.

BACKGROUND: Hypertrophic cardiomyopathy (HCM) is a common genetic heart disease. Women with HCM tend to have a later onset but more severe disease course. However, the underlying pathobiological mechanisms for these differences remain unknown. METHODS: Myectomy samples from 97 patients (53 males/44 females) with symptomatic obstructive HCM and 23 control cardiac tissues were included in this study. RNA-sequencing was performed on all samples. Mass spectrometry-based proteomics and phosphoproteomics was performed on a representative subset of samples. RESULTS: The transcriptome, proteome, and phosphoproteome was similar between sexes and did not separate on PCA plotting. Overall, there were 482 differentially expressed genes (DEGs) between control females and control males while there were only 53 DEGs between HCM females and HCM males. There were 1983 DEGs between HCM females and control females compared to 1064 DEGs between HCM males and control males. Additionally, there was increased transcriptional downregulation of hypertrophy pathways in HCM females and in HCM males. HCM females had 119 differentially expressed proteins compared to control females while HCM males only had 27 compared to control males. Finally, the phosphoproteome showed females had 341 differentially phosphorylated proteins (DPPs) compared to controls while males only had 184. Interestingly, there was hypophosphorylation and inactivation of hypertrophy pathways in females but hyperphosphorylation and activation in males. CONCLUSION: There are subtle, but biologically relevant differences in the multi-omics profile of HCM. This study provides the most comprehensive atlas of sex-specific differences in the transcriptome, proteome, and phosphoproteome present at the time of surgical myectomy for obstructive HCM.

Arrhythmic prognosis according to left ventricular systolic dysfunction severity in cardiac sarcoidosis.

BACKGROUND: Current guidelines present varying classes of recommendations for implantable cardioverter-defibrillator (ICD) utilization in patients with cardiac sarcoidosis (CS) and left ventricular ejection fraction (LVEF) <50%. OBJECTIVE: The purpose of this study was to investigate the ventricular arrhythmia risk in CS patients with ICDs and varying degrees of left ventricular systolic dysfunction. METHODS: The study included CS patients with an ICD and LVEF <50% at index evaluation. The primary outcome was survival free of sustained ventricular tachycardia (VT)/ventricular fibrillation (VF) after ICD implantation and was assessed comparatively for LVEF ≤35% vs 36%-49% and for primary vs secondary prevention ICD indication. RESULTS: The study included 61 patients (median age 57 years; 61% male) with LVEF 36%-49% (n = 23) or LVEF ≤35% (n = 38). An ICD was implanted for secondary prevention in 24% and 44% of the LVEF ≤35% and 36%-49% groups, respectively (P = .11). The primary outcome did not differ between the 2 groups in univariable analysis (LVEF ≤35% vs 36%-49%: hazard ratio [HR] 0.85; 95% confidence interval [CI] 0.39-1.82; P = .67). In multivariable analysis, secondary prevention ICD indication was the only significant predictor of incident sustained VT/VF (HR 2.86; 95% CI 1.23-6.67; P = .015). Mean sustained VT/VF event burden was higher in the secondary compared with the primary prevention ICD patients (0.47 vs 0.11 events per patient-year; P = .005) but did not differ significantly between LVEF ≤35% and 36%-49% patients. CONCLUSION: CS patients with ICD indications and LVEF 36%-49% carry similarly high arrhythmic risk as those with LVEF ≤35%. Patients with secondary prevention ICDs have the highest overall risk.

Clinical Utility of Protein Language Models in Resolution of Variants of Uncertain Significance in KCNQ1, KCNH2, and SCN5A Compared With Patch-Clamp Functional Characterization.

BACKGROUND: Genetic testing for cardiac channelopathies is the standard of care. However, many rare genetic variants remain classified as variants of uncertain significance (VUS) due to lack of epidemiological and functional data. Whether deep protein language models may aid in VUS resolution remains unknown. Here, we set out to compare how 2 deep protein language models perform at VUS resolution in the 3 most common long-QT syndrome-causative genes compared with the gold-standard patch clamp. METHODS: A total of 72 rare nonsynonymous VUS (9 KCNQ1, 19 KCNH2, and 50 SCN5A) were engineered by site-directed mutagenesis and expressed in either HEK293 cells or TSA201 cells. Whole-cell patch-clamp technique was used to functionally characterize these variants. The protein language models, ESM1b and AlphaMissense, were used to predict the variant effect of missense variants and compared with patch clamp. RESULTS: Considering variants in all 3 genes, the ESM1b model had a receiver operator curve-area under the curve of 0.75 (P=0.0003). It had a sensitivity of 88% and a specificity of 50%. AlphaMissense performed well compared with patch-clamp with an receiver operator curve-area under the curve of 0.85 (P<0.0001), sensitivity of 80%, and specificity of 76%. CONCLUSIONS: Deep protein language models aid in VUS resolution with high sensitivity but lower specificity. Thus, these tools cannot fully replace functional characterization but can aid in reducing the number of variants that may require functional analysis.

A Clinical Diagnostic Test for Calcium Release Deficiency Syndrome.

Importance: Sudden death and cardiac arrest frequently occur without explanation, even after a thorough clinical evaluation. Calcium release deficiency syndrome (CRDS), a life-threatening genetic arrhythmia syndrome, is undetectable with standard testing and leads to unexplained cardiac arrest. Objective: To explore the cardiac repolarization response on an electrocardiogram after brief tachycardia and a pause as a clinical diagnostic test for CRDS. Design, Setting, and Participants: An international, multicenter, case-control study including individual cases of CRDS, 3 patient control groups (individuals with suspected supraventricular tachycardia; survivors of unexplained cardiac arrest [UCA]; and individuals with genotype-positive catecholaminergic polymorphic ventricular tachycardia [CPVT]), and genetic mouse models (CRDS, wild type, and CPVT were used to define the cellular mechanism) conducted at 10 centers in 7 countries. Patient tracings were recorded between June 2005 and December 2023, and the analyses were performed from April 2023 to December 2023. Intervention: Brief tachycardia and a subsequent pause (either spontaneous or mediated through cardiac pacing). Main Outcomes and Measures: Change in QT interval and change in T-wave amplitude (defined as the difference between their absolute values on the postpause sinus beat and the last beat prior to tachycardia). Results: Among 10 case patients with CRDS, 45 control patients with suspected supraventricular tachycardia, 10 control patients who experienced UCA, and 3 control patients with genotype-positive CPVT, the median change in T-wave amplitude on the postpause sinus beat (after brief ventricular tachycardia at ≥150 beats/min) was higher in patients with CRDS (P < .001). The smallest change in T-wave amplitude was 0.250 mV for a CRDS case patient compared with the largest change in T-wave amplitude of 0.160 mV for a control patient, indicating 100% discrimination. Although the median change in QT interval was longer in CRDS cases (P = .002), an overlap between the cases and controls was present. The genetic mouse models recapitulated the findings observed in humans and suggested the repolarization response was secondary to a pathologically large systolic release of calcium from the sarcoplasmic reticulum. Conclusions and Relevance: There is a unique repolarization response on an electrocardiogram after provocation with brief tachycardia and a subsequent pause in CRDS cases and mouse models, which is absent from the controls. If these findings are confirmed in larger studies, this easy to perform maneuver may serve as an effective clinical diagnostic test for CRDS and become an important part of the evaluation of cardiac arrest.

Sudden cardiac arrest occurring in temporal proximity to consumption of energy drinks.

BACKGROUND: Energy drinks potentially can trigger life-threatening cardiac arrhythmias. It has been postulated that the highly stimulating and unregulated ingredients alter heart rate, blood pressure, cardiac contractility, and cardiac repolarization in a potentially proarrhythmic manner. OBJECTIVE: The purpose of this study was to describe our experience regarding sudden cardiac arrest (SCA) occurring in proximity to energy drink consumption in patients with underlying genetic heart diseases. METHODS: The electronic medical records of all SCA survivors with proven arrhythmias referred to the Mayo Clinic Windland Smith Rice Genetic Heart Rhythm Clinic for evaluation were reviewed to identify those who consumed an energy drink before their event. Patient demographics, clinical characteristics, documented energy drink consumption, and temporal relationship of energy drink consumption to SCA were obtained. RESULTS: Among 144 SCA survivors, 7 (5%; 6 female; mean age at SCA 29 ± 8 years) experienced an unexplained SCA associated temporally with energy drink consumption. Of these individuals, 2 had long QT syndrome and 2 had catecholaminergic polymorphic ventricular tachycardia; the remaining 3 were diagnosed with idiopathic ventricular fibrillation. Three patients (43%) consumed energy drinks regularly. Six patients (86%) required a rescue shock, and 1 (14%) was resuscitated manually. All SCA survivors have quit consuming energy drinks and have been event-free since. CONCLUSION: Overall, 5% of SCA survivors experienced SCA in proximity to consuming an energy drink. Although larger cohort studies are needed to elucidate the incidence/prevalence and quantify its precise risk, it seems prudent to sound an early warning on this potential risk.

Histone Modifications and miRNA Perturbations Contribute to Transcriptional Dysregulation of Hypertrophy in Obstructive Hypertrophic Cardiomyopathy.

Background: Recently, we demonstrated transcriptional downregulation of hypertrophy pathways in myectomy tissue derived from patients with obstructive hypertrophic cardiomyopathy (HCM) despite translational activation of hypertrophy pathways. The mechanisms and modifiers of this transcriptional dysregulation in HCM remain unexplored. We hypothesized that miRNA and post-translational modifications of histones contribute to transcriptional dysregulation in HCM. Methods: First, miRNA-sequencing and chromatin immunoprecipitation sequencing (ChIP-seq) were performed on HCM myectomy tissue and control donor hearts to characterize miRNA and differential histone marks across the genome. Next, the differential miRNA and histone marks were integrated with RNA-sequencing (RNA-seq) data. Finally, the effects of miRNA and histones were removed in silico to determine their necessity for transcriptional dysregulation of pathways. Results: miRNA-analysis identified 19 differentially expressed miRNA. ChIP-seq analysis identified 2,912 (7%) differential H3K4me3 peaks, 23,339 (21%) differential H3K9ac peaks, 33 (0.05%) differential H3K9me3 peaks, 58,837 (42%) differential H3K27ac peaks, and 853 (3%) differential H3K27me3 peaks. Univariate analysis of concordance between H3K9ac with RNA-seq data showed activation of cardiac hypertrophy signaling, while H3K27me showed downregulation of cardiac hypertrophy signaling. Similarly, miRNAs were predicted to result in downregulation of cardiac hypertrophy signaling. In silico knock-out that effects either miRNA or histones attenuated transcriptional downregulation while knocking out both abolished downregulation of hypertrophy pathways completely. Conclusion: Myectomy tissue from patients with obstructive HCM shows transcriptional dysregulation, including transcriptional downregulation of hypertrophy pathways mediated by miRNA and post-translational modifications of histones. Cardiac hypertrophy loci showed activation via changes in H3K9ac and a mix of activation and repression via H3K27ac.

Magnetic Resonance Imaging Characterization and Clinical Outcomes of Dilated and Arrhythmogenic Left Ventricular Cardiomyopathies.

BACKGROUND: Nondilated left ventricular cardiomyopathy (NDLVC) has been recently differentiated from dilated cardiomyopathy (DCM). A comprehensive characterization of these 2 entities using cardiac magnetic resonance (CMR) and genetic testing has never been performed. OBJECTIVES: This study sought to provide a thorough characterization and assess clinical outcomes in a large multicenter cohort of patients with DCM and NDLVC. METHODS: A total of 462 patients with DCM (227) or NDLVC (235) with CMR data from 4 different referral centers were retrospectively analyzed. The study endpoint was a composite of sudden cardiac death or major ventricular arrhythmias. RESULTS: In comparison to DCM, NDLVC had a higher prevalence of pathogenic or likely pathogenic variants of arrhythmogenic genes (40% vs 23%; P < 0.001), higher left ventricular (LV) systolic function (LV ejection fraction: 51% ± 12% vs 36% ± 15%; P < 0.001) and higher prevalence of free-wall late gadolinium enhancement (LGE) (27% vs 14%; P < 0.001). Conversely, DCM showed higher prevalence of pathogenic or likely pathogenic variants of nonarrhythmogenic genes (23% vs 12%; P = 0.002) and septal LGE (45% vs 32%; P = 0.004). Over a median follow-up of 81 months (Q1-Q3: 40-132 months), the study outcome occurred in 98 (21%) patients. LGE with septal location (HR: 1.929; 95% CI: 1.033-3.601; P = 0.039) was independently associated with the risk of sudden cardiac death or major ventricular arrhythmias together with LV dilatation, older age, advanced NYHA functional class, frequent ventricular ectopic activity, and nonsustained ventricular tachycardia. CONCLUSIONS: In a multicenter cohort of patients with DCM and NDLVC, septal LGE together with LV dilatation, age, advanced disease, and frequent and repetitive ventricular arrhythmias were powerful predictors of major arrhythmic events.

Atrial Fibrillation Substrate and Catheter Ablation Outcomes in MYBPC3- and MYH7-Mediated Hypertrophic Cardiomyopathy.

BACKGROUND: The effects of disease-causing MYBPC3 or MYH7 genetic variants on atrial myopathy, atrial fibrillation (AF) clinical course, and catheter ablation efficacy remain unclear. OBJECTIVES: The aim of this study was to characterize the atrial substrate of patients with MYBPC3- or MYH7-mediated hypertrophic cardiomyopathy (HCM) and its impact on catheter ablation outcomes. METHODS: A retrospective single-center study of patients with HCM who underwent genetic testing and catheter ablation for AF was performed. Patients with MYBPC3- or MYH7-mediated HCM formed the gene-positive cohort; those without disease-causative genetic variants formed the control cohort. High-density electroanatomical mapping was performed using a 3-dimensional mapping system, followed by radiofrequency ablation. RESULTS: Twelve patients were included in the gene-positive cohort (mean age 55.6 ± 9.9 years, 83% men, 50% MYBPC3, 50% MYH7, mean ejection fraction 59.3% ± 13.7%, mean left atrial [LA] volume index 51.7 ± 13.1 mL/m2, mean LA pressure 20.2 ± 5.4 mm Hg) and 15 patients in the control arm (mean age 61.5 ± 12.6 years, 60% men, mean ejection fraction 64.9% ± 5.1%, mean LA volume index 54.1 ± 12.8 mL/m2, mean LA pressure 19.6 ± 5.41 mm Hg). Electroanatomical mapping demonstrated normal voltage in 87.7% ± 5.03% of the LA in the gene-positive cohort and 94.3% ± 3.58% of the LA in the control cohort (P < 0.001). Of the abnormal regions, intermediate scar (0.1-0.5 mV) accounted for 6.33% ± 1.97% in the gene-positive cohort and 3.07% ± 2.46% in the control cohort (P < 0.01). Dense scar (<0.1 mV) accounted for 5.93% ± 3.20% in the gene-positive cohort and 2.61% ± 2.19% in the control cohort (P < 0.01). Freedom from AF at 12 months was similar between the gene-positive (75%) and control (73%) cohorts (P = 0.92), though a greater number of procedures were required in the gene-positive cohort. CONCLUSIONS: Patients with MYBPC3- or MYH7-mediated HCM undergoing AF ablation have appreciably more low-amplitude LA signals, suggestive of fibrosis. However, catheter ablation remains an effective rhythm-control strategy.

Arrhythmic manifestations and outcomes of definite and probable cardiac sarcoidosis.

BACKGROUND: The 2014 Heart Rhythm Society consensus statement defines histological (definite) and clinical (probable) diagnostic categories of cardiac sarcoidosis (CS), but few studies have compared their arrhythmic phenotypes and outcomes. OBJECTIVE: The purpose of this study was to evaluate the electrophysiological/arrhythmic phenotype and outcomes of patients with definite and probable CS. METHODS: We analyzed the arrhythmic/electrophysiological phenotype in a single-center North American cohort of 388 patients (median age 56 years; 39% female, n = 151) diagnosed with definite (n = 58) or probable (n = 330) CS (2000-2022). The primary composite outcome was survival to first ventricular tachycardia/fibrillation (VT/VF) event or sudden cardiac death. Key secondary outcomes were also assessed. RESULTS: At index evaluation, in situ cardiac implantable electronic devices and antiarrhythmic drug use were more common in definite CS. At a median follow-up of 3.1 years, the primary outcome occurred in 22 patients with definite CS (38%) and 127 patients with probable CS (38%) (log-rank, P = .55). In multivariable analysis, only a higher ratio of the 18F-fluorodeoxyglucose maximum standardized uptake value of the myocardium to the maximum standardized uptake value of the blood pool (hazard ratio 1.09; 95% confidence interval 1.03-1.15; P = .003, per 1 unit increase) was associated with the primary outcome. During follow-up, patients with definite CS had a higher burden of device-treated VT/VF events (mean 2.86 events per patient-year vs 1.56 events per patient-year) and a higher rate of progression to heart transplant/left ventricular assist device implantation but no difference in all-cause mortality compared with patients with probable CS. CONCLUSION: Patients with definite and probable CS had similarly high risks of first sustained VT/VF/sudden cardiac death and all-cause mortality, though patients with definite CS had a higher overall arrhythmia burden. Both CS diagnostic groups as defined by the 2014 Heart Rhythm Society criteria require an aggressive approach to prevent arrhythmic complications.

State of Gene Therapy for Monogenic Cardiovascular Diseases.

Over the past 2 decades, significant efforts have been made to advance gene therapy into clinical practice. Although successful examples exist in other fields, gene therapy for the treatment of monogenic cardiovascular diseases lags behind. In this review, we (1) highlight a brief history of gene therapy, (2) distinguish between gene silencing, gene replacement, and gene editing technologies, (3) discuss vector modalities used in the field with a special focus on adeno-associated viruses, (4) provide examples of gene therapy approaches in cardiomyopathies, channelopathies, and familial hypercholesterolemia, and (5) present current challenges and limitations in the gene therapy field.

Imaging of Cardiac Sarcoidosis: An Update and Future Aspects.

Cardiac sarcoidosis (CS), an increasingly recognized disease of unknown etiology, is associated with significant morbidity and mortality. Given the limited diagnostic yield of traditional endomyocardial biopsy (EMB), there is increasing reliance on multimodality cardiovascular imaging in the diagnosis and management of CS, with EMB being largely supplanted by the use of 18F-fluorodeoxyglucose (FDG-PET) and cardiac magnetic resonance imaging (CMR). This article aims to provide a comprehensive review of imaging modalities currently utilized in the screening, diagnosis, and monitoring of CS, while highlighting the latest developments in each area.

A Multi-Omics Atlas of Sex-Specific Differences in Obstructive Hypertrophic Cardiomyopathy.

Background: Hypertrophic cardiomyopathy (HCM) is a common genetic heart disease. Women with HCM tend to have a later onset but more severe disease course. However, the underlying pathobiological mechanisms for these differences remain unknown. Methods: Myectomy samples from 97 patients (53 males/44 females) with symptomatic obstructive HCM and 23 control cardiac tissues were included in this study. RNA-sequencing was performed on all samples. Mass spectrometry-based proteomics and phosphoproteomics was performed on a representative subset of samples. Results: The transcriptome, proteome, and phosphoproteome was similar between sexes and did not separate on PCA plotting. Overall, there were 482 differentially expressed genes (DEGs) between control females and control males while there were only 53 DEGs between HCM females and HCM males. There were 1963 DEGs between HCM females and control females compared to 1064 DEGs between HCM males and control males. Additionally, there was increased transcriptional downregulation of hypertrophy pathways in HCM females and in HCM males. HCM females had 119 differentially expressed proteins compared to control females while HCM males only had 27 compared to control males. Finally, the phosphoproteome showed females had 341 differentially phosphorylated proteins (DPPs) compared to controls while males only had 184. Interestingly, there was hypophosphorylation and inactivation of hypertrophy pathways in females but hyperphosphorylation and activation in males. Conclusion: There are subtle, but biologically relevant differences in the multi-omics profile of HCM. This study provides the most comprehensive atlas of sex-specific differences in the transcriptome, proteome, and phosphoproteome present at the time of surgical myectomy for obstructive HCM.

Artificial intelligence-enhanced electrocardiogram for arrhythmogenic right ventricular cardiomyopathy detection.

Aims: ECG abnormalities are often the first signs of arrhythmogenic right ventricular cardiomyopathy (ARVC) and we hypothesized that an artificial intelligence (AI)-enhanced ECG could help identify patients with ARVC and serve as a valuable disease-detection tool. Methods and results: We created a convolutional neural network to detect ARVC using a 12-lead ECG. All patients with ARVC who met the 2010 task force criteria and had disease-causative genetic variants were included. All case ECGs were randomly assigned in an 8:1:1 ratio into training, validation, and testing groups. The case ECGs were age- and sex-matched with control ECGs at our institution in a 1:100 ratio. Seventy-seven patients (51% male; mean age 47.2 ± 19.9), including 56 patients with PKP2, 7 with DSG2, 6 with DSC2, 6 with DSP, and 2 with JUP were included. The model was trained using 61 case ECGs and 5009 control ECGs; validated with 7 case ECGs and 678 control ECGs and tested in 22 case ECGs and 1256 control ECGs. The sensitivity, specificity, positive and negative predictive values of the model were 77.3, 62.9, 3.32, and 99.4%, respectively. The area under the curve for rhythm ECG and median beat ECG was 0.75 and 0.76, respectively. Conclusion: Our study found that the model performed well in excluding ARVC and supports the concept that the AI ECG can serve as a biomarker for ARVC if a larger cohort were available for network training. A multicentre study including patients with ARVC from other centres would be the next step in refining, testing, and validating this algorithm.