JACC: Clinical Electrophysiology

Background. Durable mitral isthmus (MI) block remains challenging in persistent atrial fibrillation (PeAF) ablation. Recent epicardial vein of Marshall (VoM) recordings have shown incomplete MI transmurality and time-dependent conduction recovery after pulsed field ablation (PFA). Whether systematic VoM ethanol infusion (VoM-EI) followed by focal PFA provides stable acute MI block remains unknown. **Objectives.** To assess the incidence, timing, and procedural implications of early MI conduction recovery after systematic VoM-EI followed by focal Sphere-9 PFA. Methods.In this prospective single-center study, 55 consecutive patients undergoing first ablation for symptomatic PeAF with planned MI ablation were screened. VoM-EI was systematically attempted before left atrial access and successfully performed in 51 (92.7%), who constituted the study cohort. Pulmonary vein isolation, roof-line, and MI ablation were performed with the Sphere-9? lattice-tip catheter. After bidirectional MI block, conduction was systematically reassessed during a standardized 30-minute waiting period. Results.Mean age was 70.3 {+/-} 8.2 years, and 36 patients (70.6%) were men. Initial bidirectional MI block was achieved in 50/51 patients (98.0%). During the waiting period, conduction recovered in 9/50 (18.0%; 95% CI, 9.8%-30.8%), at a median of 16 minutes (IQR, 10-20; range, 8?23). Six of 9 patients with recovery (66.7%) required targeted coronary sinus (CS) ablation. Block was restored in all 9, yielding a final block rate of 50/51 (98.0%). Median procedure duration was 82 minutes (IQR, 73-95), with no major complications. Conclusions. Immediate bidirectional MI block was not synonymous with stable block. Despite systematic VoM-EI followed by focal Sphere-9 PFA, conduction recovered in approximately one in five patients, including beyond 20 minutes, and two thirds required targeted CS ablation. These findings support standardized 30-minute reassessment and targeted CS interrogation rather than reliance on immediate block. Chronic invasive remapping is required to determine whether this strategy improves long-term MI block durability.

BACKGROUND: Atrial fibrillation (AF) remains difficult to explain using a single focal driver or rotor-centered mechanism across disease stages. We tested whether progressive atrial substrate remodeling can drive a critical transition toward turbulence-like, decentralized multi wavelet electrical activity. METHODS: We constructed a controlled two-dimensional atrial reaction-diffusion model with six graded substrate-remodeling stages. We evaluated effective wavelength, theoretical wavelet capacity, AF inducibility, vulnerable-window dynamics, spatial randomness, temporal memory, spectral dispersion, nonlinear indices, virtual ablation response and ERP-prolongation reverse mechanistic testing. RESULTS: Progressive remodeling shortened effective wavelength from 12.0 to 2.4 cm and increased theoretical wavelet capacity from 0.69 to 17.36. Inducibility rose sigmoidally as wavelength shortened, with a model-derived transition near lambda50=4.5 cm. Advanced substrates showed increased wavebreak, spatial randomness, short-memory dynamics, broad spectral dispersion, positive nonlinear indices and resistance to random local ablation. Culprit atrial premature beats within the vulnerable window efficiently triggered AF, whereas counter pacing at 20 to 35 ms reduced inducibility from 52% to 11% in stage 2. CONCLUSIONS: In this controlled model, AF initiation and maintenance were linked to substrate-dependent wavelength, wavelet capacity and vulnerable-window triggering. The model-derived transition provides a testable framework for future high-density mapping, patient30 specific modeling and device-based studies. Key Words atrial fibrillation; turbulence-like electrical activity; substrate remodeling; critical wavelength; multi-wavelet re-entry; vulnerable window; culprit premature atrial beat; counter pacing

Background: Mitral regurgitation (MR) is a major risk factor for the development of atrial fibrillation (AF), yet the molecular mechanisms linking volume overload to arrhythmogenic remodeling remain poorly understood. Although fibrosis has long been considered the primary substrate for AF, increasing evidence suggests that fibroblast heterogeneity and cell-cell interactions may play important roles in disease progression. Methods: MR was created endovascularly by chordal avulsion in 12 dogs with 6 controls. AF inducibility was assessed by transvenous burst pacing, left atrial volume by echocardiography, and collagen content by Masson trichrome and picrosirius red staining. Single-nucleus RNA sequencing (snRNA-seq) was performed on left atrial posterior wall tissue from control, 4-week, and 6-month MR animals. Fibroblast subpopulations and fibroblast-cardiomyocyte communication were analyzed and markers validated by RNA in situ hybridization in all 18 animals. Results: MR resulted in progressive left atrial dilation, but neither the change in left atrial volume from baseline nor total collagen burden correlated with the inducibility of AF (n=6 each). SnRNA-seq resolved seven major cardiac cell populations and identified four transcriptionally distinct fibroblast populations (NOX4/GRIA4, PCOLCE2, ADRB2/HCN1, PTX3/ICAM1). Fibroblast composition shifted markedly: matrix-associated PCOLCE2 fibroblasts starkly declined by 6 months, whereas inflammatory-associated PTX3/ICAM1 fibroblasts expanded stepwise over time. Cardiomyocyte-to-fibroblast signaling, dominated by PTPRM and LAMA2, was progressively redirected toward PTX3/ICAM1 fibroblasts. RNAscope confirmed a stepwise rise in ICAM1 transcripts and higher ICAM1 in AF-inducible than non-inducible animals. Conclusions: In a canine model of MR, the inducibility of AF was associated with fibroblast state remodeling rather than with atrial dilation or collagen burden. Progressive expansion of inflammatory-associated PTX3/ICAM1 fibroblasts, together with reorganized fibroblast-cardiomyocyte signaling, defines a candidate arrhythmogenic mechanism and therapeutic target in MR.

Background Substrate mapping using multielectrode catheters is increasingly used for post-myocardial infarction (MI) ventricular tachycardia (VT) avoiding repeated VT induction and mapping during VT. However, these catheters may mechanically induce ventricular arrhythmias with hemodynamic compromise. This study compares pro-arrhythmogenicity between single-tip and multi-spline catheters during functional substrate mapping. Methods Thirty post-MI patients (age 68{+/-}8 years, 97% male, LVEF 40% [IQR 33-46]) referred for VT ablation at two centers (2021-2024) underwent endocardial mapping during baseline rhythm in random order with both a multi-spline catheter (Octaray, n=4; Pentaray, n=26) and a single-tip QDOT catheter. The protocol was prematurely terminated if (i) two mechanically induced VTs required ECV, (ii) recurrent ATP-treated mechanical VTs caused hemodynamic compromise, or (iii) excessive mechanically induced ectopy impaired catheter contact. Mapping time, point density, and mechanically induced arrhythmias were assessed. Results Multi-spline catheters enabled faster mapping (26{+/-}9 vs 60{+/-}16 minutes, p<0.001) with more acquired points (p<0.001). VTs were more frequently mechanically induced with multi-spline catheters (median 2 [IQR 1-4] vs 0 [0-3], p<0.05) and these VTs were faster (304ms, IQR 292-320] vs 373ms, IQR [316-405], p=0.01) and degenerated more often into VF (3 vs. 0). Overall, 17 patients (57%) experienced at least one mechanically induced VT; seven (23%) required cardioversion, and mapping was prematurely terminated in eight (27%), all while using multi-spline catheters. Conclusion Multi-spline catheters allow rapid substrate mapping but with substantial risk of mechanically induced arrhythmias, requiring premature termination of substrate mapping because of safety concerns. Their use in post-MI VT ablation warrants careful risk?benefit assessment.

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and a major cause of cardioembolic stroke. Although polygenic risk scores (PRS) are well characterized to quantify inherited susceptibility for AF, they provide limited insight into the pathways and tissues underlying genetic risk, which are critical to uncover for individual risk prediction. In this study, we develop a pathway-level multi-omics representation learning framework that converts individual genetic profiles into interpretable biological features by integrating GWAS-derived pathway burden scores with tissue-specific transcriptomic pathway signals. We constructed machine learning models to assess population-level AF risk prediction performance across genomic and transcriptomic tissue contexts; the pathway-based global attention models substantially improved risk prediction performance over PRS and other baselines (AUROC improved from 0.601 to 0.738). Transformer and graph neural network frameworks then assessed individual-level pathway interpretability, revealing heterogeneous contributions from electrical signaling, cardiac development, and DNA repair pathways to AF risk. This added interpretability highlights the potential of this pathway approach to enable more mechanistically informed risk stratification than static PRS by capturing underlying heterogeneity. To independently assess whether prioritized pathways reflected cardiac regulatory biology, we compared pathway rankings with transcriptional effects predicted by the AlphaGenome foundation model. Variants in highly ranked pathways showed significantly greater predicted effects on expression in atrial and ventricular tissues (FDR = 0.032) relative to controls, providing orthogonal evidence that the model identifies biologically relevant mechanisms. Overall, this work reframes polygenic risk from a single measure of susceptibility to tissue-informed pathway mechanisms, providing a framework for interpretable genomic stratification in complex diseases.

Background: The His-Purkinje network drives rapid ventricular activation and is a major substrate for ventricular arrhythmias, yet it is among the least molecularly characterized cardiac compartments. Markers validated in rodents transfer poorly across species, few are confirmed at the protein level in large mammals or humans, and most lack the stability and surface accessibility that demanding applications require. Methods: We combined histology-guided laser-capture microdissection with low-input, cell-type-resolved RNA-sequencing to profile Purkinje fibers, left-ventricular cardiomyocytes and peri-Purkinje stroma from adult sheep. Differentially expressed genes were ranked by a transparent composite framework weighting expression specificity, cross-individual stability and predicted subcellular accessibility; leading candidates were validated by RT-qPCR and immunolabelling in sheep and by RT-qPCR in human myocardium. Results: RNA-sequencing resolved a Purkinje transcriptome distinct from cardiomyocytes and stroma and defined 331 concordantly enriched genes, which the composite framework ranked into stable, specific candidates spanning intracellular and cell-surface compartments. By RT-qPCR, the canonical conduction markers connexin-40/GJA5, HCN4, NEFM and MYL4 were strongly enriched in Purkinje fibers, whereas the rodent gold-standard contactin-2 was not, underscoring species divergence. Thirteen of sixteen prioritized candidates were confirmed by RT-qPCR, and immunolabelling localized MYL4, CNN1, TAGLN and DKK3 to Purkinje fibers; contactin-5 emerged as a novel transcript- and protein-validated Purkinje marker. In human myocardium, a defined subset - MYL4, connexin-40/GJA5, contactin-5 and TAGLN - was conserved, while several markers proved species-restricted. Conclusions: We provide the first genome-wide, cell-type-resolved molecular portrait of the Purkinje fiber in a large-animal model and a generalizable strategy that selects markers for specificity, stability and accessibility. The resulting resource - including the cross-species marker contactin-5 and compartment-matched candidates - supplies validated tools to identify, isolate and target Purkinje cells and demonstrates the necessity of cross-species validation.

Conventional electrocardiography is highly effective for waveform and rhythm diagnosis, but it is less suited to showing how the internal shape of hundreds or thousands of consecutive heartbeats changes over time. We introduce FOXTAIL, a complementary view that represents each cardiac cycle as an ordered sequence of changes in signal direction. Overlaying these sequences in a fixed visual field makes beat-to-beat organization visible and allows the density, size, stability, and scale persistence of those changes to be measured. We evaluated the representation in recordings containing normal sinus rhythm, paroxysmal atrial fibrillation, severe heart failure, ventricular tachyarrhythmia, and controlled electrode-motion noise. Paired recordings showed that FOXTAIL descriptors can reveal within-person state changes that are not conveyed by a single average beat. The noise and pre-fibrillation analyses also showed that a dense event pattern is not automatically equivalent to physiological complexity, measurement artifact, or impending disease. FOXTAIL is therefore not proposed as a replacement for the diagnostic ECG or as a new classifier, but as an observation and measurement domain for asking a more basic question: how is the electrical organization of the heart changing from one beat to the next, and which of those changes persist across scale?

Transverse (t)-tubules enable rapid, synchronous Ca release required for efficient cardiac contraction by bringing L-type Ca channels into close apposition with ryanodine receptors. In heart failure with reduced ejection fraction (HFrEF), t-tubule disorganisation and loss occur alongside cardiac microtubule remodelling, contributing to impaired Ca handling and contractile dysfunction. Despite their canonical function in contraction, how t-tubules develop is unknown. Microtubules support delivery of L-type Ca channels to t-tubules via Amphiphysin-II/BIN1, yet whether microtubules directly regulate t-tubule formation and maintenance is unclear. Here, we investigated a role for microtubules in t-tubule development and homeostasis. Neonatal rat ventricular myocytes (NRVMs), which lack endogenous t-tubules, were used as a reductionist model in which BIN1 overexpression induces nascent membrane tubules. Microtubule depolymerisation with nocodazole before BIN1 overexpression impaired BIN1-driven tubule formation, reducing tubule density and length. Dynein inhibition with EHNA produced similar effects, indicating a requirement for microtubule-based motor activity during tubule elongation. Knockdown of the microtubule +TIP tracking protein CLIP-170 also reduced BIN1-driven tubule density, implicating BIN1-CLIP-170-dependent microtubule capture in tubule initiation. Microtubules were also required to maintain existing tubules. In NRVMs with established BIN1-driven tubules, microtubule depolymerisation, microtubule stabilisation or dynein inhibition each reduced tubule density and length. Consistent with this, acute microtubule depolymerisation or stabilisation disrupted native t-tubule networks in isolated adult sheep left atrial myocytes. Together, these findings identify cardiac microtubules as active regulators of t-tubule architecture. We propose that BIN1-dependent tubule formation requires CLIP-170-mediated microtubule plus-end capture and dynein-dependent elongation, while ongoing microtubule dynamics are necessary to preserve mature t-tubule structure.

Cardiac imaging enables quantitative assessment of cardiac structure and function but remains constrained by cost, infrastructure and specialist expertise. In contrast, electrocardiogram (ECG) is widely accessible yet underexploited, despite encoding latent information about cardiac physiology. Here we introduce visionECG, a conditional flow matching framework that learns a probabilistic mapping between two biological distributions - the space of cardiac electrical signals and the space of cardiac geometries. Using 71,132 paired ECG and cardiac mesh sequence datasets from the UK Biobank, with external assessment in 5,000 patients with ECG-echocardiogram pairs, the model reconstructs quantitatively accurate spatiotemporal representations of the left ventricle using ECG inputs and basic demographic information alone. These reconstructions enable discrimination of structural abnormalities and disease labels, provide visualisations of functional abnormalities, and support flexible quantification of both global and regional parameters. By reframing the ECG as a generative source of patient-specific left ventricular geometry and motion, this work establishes a scalable framework for translating low-dimensional signals into high-dimensional, physiologically grounded structured representations.

BackgroundAdipose tissue surrounding the heart and vasculature plays critical roles in cardiovascular homeostasis and disease, yet the cellular and molecular milieu of these depots at single-cell resolution remains incompletely characterized. Understanding how regional adipocytes differ transcriptionally and communicate with neighboring cardiovascular cells is essential for developing targeted therapeutic strategies. MethodsWe performed single-nucleus RNA sequencing (snRNA-seq) on human adipose tissue from four anatomically distinct depots: ascending aorta, left atrium, right coronary artery, and subcutaneous fat. We characterized cellular composition, adipocyte and progenitor heterogeneity, depot-specific transcriptional programs, and intercellular communication networks. We further examined signaling remodeling in disease contexts, including atrial fibrillation and aortic aneurysm. ResultsWe identified six transcriptionally distinct adipocyte subpopulations and six adipocyte stromal and progenitor cell (ASPC) subpopulations were shared across depots but showed marked differences in abundance and gene expression reflecting developmental imprinting, including HOX family genes and anterior-posterior patterning programs. Intercellular communication analysis revealed depot-specific ligand-receptor interactions, with EPHA signaling identified as selectively enriched in the left atrial adipose depot. Disease-state analyses demonstrated extensive change in cell-cell communication in atrial fibrillation and aortic aneurysm, with differential regulation of FN1, EGF, SLIT, NOTCH, and CD46 signaling pathways. ConclusionsOur study reveals that cardiac and vascular adipose depots harbor transcriptionally specialized adipocytes and progenitors with distinct intercellular communication programs that are remodeled in atrial fibrillation and aortic aneurysm.

RationaleHypertrophic cardiomyopathy (HCM) is a common inherited cardiac disorder characterized by cardiac hypertrophy, fibrosis, arrhythmias, and sudden cardiac death (SCD). Although current therapies primarily target sarcomere dysfunction, the contribution of epigenetic dysregulation to HCM pathogenesis and its therapeutic potential remain poorly understood. ObjectiveTo determine whether pharmacological inhibition of histone lysine demethylases (KDMs) with JIB-04 can prevent or reverse HCM progression and to identify the underlying epigenetic mechanisms. Methods and ResultsWe evaluated the pan-KDM inhibitor JIB-04 in Myh6R403Q/+ mice carrying the murine equivalent of the pathogenic human MYH7 R403Q mutation. JIB-04 prevented disease progression, reduced cardiac hypertrophy and fibrosis, preserved cardiac function, and completely prevented SCD in cyclosporin A- accelerated HCM. JIB-04 also reversed established disease, produced sustained therapeutic benefits after drug withdrawal, and improved cardiac function in aged mice with spontaneous HCM. Bulk RNA sequencing and ATAC-seq demonstrated partial restoration of disease-associated transcriptional programs and chromatin accessibility. Proteomic analyses identified PHF2 (KDM7C) as a candidate target of JIB-04 in both mouse and human HCM hearts. PHF2 knockdown suppressed hypertrophic, inflammatory, and fibrotic gene expression in cardiomyocytes, macrophages, and fibroblasts, respectively. Human HCM hearts exhibited increased expression of multiple JIB-04-sensitive KDMs, including PHF2. In MYH7 R403Q induced pluripotent stem cell- derived cardiomyocytes, JIB-04 normalized disease-associated gene expression, restored connexin-43 membrane localization, and improved mitochondrial respiration. Although prolonged treatment induced reversible hepatomegaly with hepatic lipid accumulation, co-administration of the antioxidant N-acetylcysteine mitigated liver toxicity while preserving the therapeutic efficacy of JIB-04. ConclusionsPharmacological KDM inhibition prevents and reverses HCM through epigenetic remodeling of disease-associated transcriptional and chromatin programs. These findings identify KDM inhibition as a promising therapeutic strategy for HCM, establish PHF2 as a candidate mediator of disease pathogenesis, and support further development of KDM-targeted therapies.

Background and AimsMutations in the desmin (DES) gene cause a variety of cardiomyopathies associated with arrhythmias, yet the electrophysiological consequences of these variants remain largely uncharacterized. The aim of this study was to investigate the pathogenic mechanisms of the de novo DES p.R406W variant, which was identified in a 9-year-old patient who suffered from severe ventricular arrhythmias and sudden cardiac death without overt structural heart disease. MethodsHuman induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) carrying the DES p.R406W variant (including the patients line) were compared to isogenic controls. Action potentials (AP) of hiPSC-CMs were recorded using patch-clamp. Furthermore, 3D engineered heart tissues (EHTs) were generated from hiPSC-CMs and their APs were recorded with sharp microelectrodes. Analytical techniques also included transmission electron microscopy (TEM) and integrated transcriptomic and proteomic profiling. Finally, a heterozygous knock-in (KI) mouse model carrying the Des p.R405W ortholog was evaluated through surface ECG, echocardiography and ex vivo cardiac optical mapping. ResultsThe DES p.R406W mutation prolonged AP duration in IM-R406W hiPSC-CMs and EHTs vs Control ones. Multi-omics analysis of EHTs revealed a dysregulation of genes and proteins involved in contractile function, cell adhesion, and electrical activity. TEM imaging revealed changes in Z-disc architecture in mutant tissues. Twenty-week-old Des p.R405W KI mice exhibited ventricular conduction slowing (prolonged QRS) and a high susceptibility to ventricular tachyarrhythmias, likely due to reentrant mechanisms. Mild hypertrophy was also observed, but only in females. ConclusionThe DES p.R406W variant is highly pathogenic, causing electrical and structural remodeling of the myocardium. This study highlights the effectiveness of hiPSC-CMs and EHTs in recapitulating the clinical phenotype of desminopathy, providing a platform for investigating the mechanisms of early-onset cardiac arrhythmias and SCD.

Background: Bailout cardiac surgery during transcatheter aortic valve replacement (TAVR) is uncommon but remains associated with substantial morbidity and mortality. Although registries have described its incidence and major causes, they often provide limited detail regarding device-related failure mechanisms, attempted transcatheter rescue, and the clinical pathway leading to surgical conversion. We aimed at analyzing post-marketing safety reports from the U.S. Food and Drug Administration (FDA) Manufacturer and User Facility Device Experience (MAUDE) database to characterize the mechanisms, management strategies, and reported outcomes of bailout surgery during or shortly after TAVR. Methods: We retrospectively analyzed FDA MAUDE reports received from July 1, 2016, through June 30, 2026. Eligible reports described unplanned urgent or emergent open cardiac surgery during or immediately after TAVR. Candidate reports were screened, adjudicated, and deduplicated at the clinical-event level. Events were classified by precipitating complication, transcatheter rescue, operative pathway, and reported outcome. Associations were evaluated using permutation tests, Fisher exact tests with Benjamini?Hochberg correction, adjusted regression models, and sensitivity analyses. Results: After screening 43,239 initial reports, we identified 376 bailout-surgery events, with survival status was documented in 254, including 104 deaths and 150 survivors, corresponding to 40.9% reported mortality. Valve embolization, migration, or malposition was the most frequent complication phenotype (32.4%), whereas ventricular perforation or laceration was associated with the highest mortality (74.1%; OR, 4.86; 95% CI, 1.97?11.99). Mortality differed across complication phenotypes (p<0.001) and operative pathways (p<0.001), but not across transcatheter rescue pathways (p=0.355). Valve explantation with SAVR was associated with lower reported mortality (18.9%; OR, 0.29; 95% CI, 0.12?0.69), whereas unspecified surgery or access/support alone was associated with higher mortality (56.9%; OR, 3.04; 95% CI, 1.80?5.12). Ancillary analyses identified potential platform-specific differences in complication and management patterns, while bailout timing was not independently associated with mortality after adjustment. Conclusions: In this MAUDE analysis, bailout cardiac surgery after TAVR was most commonly precipitated by valve embolization, migration, or malposition, whereas ventricular perforation or laceration was associated with the highest reported mortality. Outcomes differed across complication and operative pathways but not across transcatheter rescue strategies or bailout timing after adjustment. These findings identify clinically relevant post-marketing safety signals but should not be interpreted as incidence estimates, comparative device risks, or causal treatment effects.

Objective(s)Tricuspid valve (TV) remodeling and functional tricuspid regurgitation (FTR) progression during right ventricular (RV) pressure overload and reverse remodeling after resolution of RV afterload is poorly understood. We set out to investigate tricuspid leaflet tissue response to induction and subsequent alleviation of pressure overload in a large animal model of RV failure with FTR. MethodsFifteen healthy adult male Dorset sheep (72{+/-}4 kg) underwent pulmonary artery banding (PAB) to induce RV failure and FTR. After 8 weeks, 7 sheep (PAB, n=7) were terminated, and remaining 8 had the PAB removed (rPAB, n=8) and were followed for another 8 weeks before termination. Both groups underwent epicardial echocardiography and hemodynamic assessment during banding surgery and at terminal operation. Ten healthy sheep served as a control group (CTL, n=10) and underwent terminal procedure only. In all animals, TV leaflets and right ventricular (RV) tissue were harvested at terminal procedure and analyzed histologically and transcriptionally. ResultsTV leaflets in PAB animals showed increased cross-sectional area and ECM alterations, some of which persisted after resolution of RV pressure overload. rPAB valves exhibited distinct ECM composition, with notably altered mucin and fibrin content, suggesting a shift toward matrix stabilization, dissimilar to control and PAB. RNA sequencing uncovered a unique molecular state in rPAB valves, with persistent changes in PRG4, PDE3A, CXCL8, and HLA transcripts. RV tissue also demonstrated a separate remodeling trajectory, with sustained expression of stress-related genes including PDE3A, NAV2, ANFB, and ACTS. These findings indicate that both valve and ventricular tissues retain a persistent remodeled phenotype post-unloading. ConclusionsTV leaflets actively remodel in response to hemodynamic stress and do not fully revert to a normal state after relief of pressure overload. This persistent altered phenotype may represent a biological contribution of the TV leaflets to recurrent TR with implications for long-term outcomes following treatment of FTR. Clinical Perspective What is new?O_LIRelief of right ventricular pressure overload, in a large animal model, resulted in substantial reverse remodeling of the right heart and reduction of tricuspid regurgitation severity, but tricuspid valve leaflets did not return to a normal state. C_LIO_LIReverse remodeled leaflets remained enlarged despite normalization of hemodynamics with an altered extracellular matrix. C_LIO_LICellular proliferation and immune cell infiltration observed during pressure overload resolved after unloading, yet transcriptomic analysis identified a distinct molecular phenotype that differed from both healthy and diseased valves. C_LIO_LITricuspid valve leaflets are active biological participants in the remodeling process and exhibit persistent adaptation or maladaptation after resolution of the initiating hemodynamic stress. C_LI What Are the Clinical Implications?O_LISecondary tricuspid regurgitation should be considered a disease involving both right heart geometry and leaflet biology. C_LIO_LIResolution of the underlying cause of tricuspid regurgitation may not restore leaflet structure and molecular homeostasis. C_LIO_LIPersistent leaflet remodeling may contribute to residual or recurrent tricuspid regurgitation despite successful treatment of pulmonary hypertension or other inciting conditions. C_LIO_LITherapies directed at leaflet remodeling may ultimately complement surgical and transcatheter strategies currently focused on annular and ventricular geometry. C_LI

Objective: Chronic atrial and intestinal dysrhythmia (CAID) syndrome is a rare autosomal recessive cohesinopathy classically defined by sick sinus syndrome and chronic intestinal pseudo-obstruction; however, emerging evidence suggests an association with cerebral small vessel disease (CSVD). We aimed to characterize the neurological and neuroimaging spectrum of CSVD in CAID syndrome. Methods: We conducted a cross-sectional, retrospective study of 16 French-Canadians with genetically confirmed CAID syndrome. All patients underwent comprehensive neurological assessment. Brain MRI was performed in 14 patients, with CSVD markers evaluated by an expert neuroradiologist according to the STRIVE-2 criteria. Results: The median age at last evaluation was 34 years (range, 19-60); 62.5% were women. Neurological manifestations included migraines (44.4%), mild cerebellar signs (16.7%), and ischemic or hemorrhagic cerebrovascular events (12.5%). MRI showed white matter hyperintensities (92.9%), lacunes (50%) and cerebral microbleeds (85.7%), affecting deep, lobar, and infratentorial regions, with marked cerebellar predominance (11/12; 91.7%); five patients exhibited innumerable microbleeds. Despite the young cohort, moderate-to-severe CSVD was common (median SVD score 1.5, IQR 0-4). Patients with countless microbleeds were older than those with discrete lesions (46.2 vs. 31.2 years; p=0.043). Management of atrial fibrillation required individualized strategies, including left atrial appendage closure, balancing ischemic and hemorrhagic risks. Interpretation: CAID syndrome represents a novel monogenic cause of CSVD, characterized by early, extensive cerebral microbleeds with mixed distribution and distinctive cerebellar predominance. Coexisting congenital cardiac disease and arrhythmias place patients at dual ischemic and hemorrhagic risk. Systematic neurological evaluation and MRI are warranted, particularly prior to antithrombotic therapy.

Background: Unstructured biomedical data, such as echocardiography reports, are rich in information but time consuming to analyze at scale. Rule-based, regular expression-driven terminology mapping can only extract individual variables while large language models (LLMs) offer scalable and clinically meaningful interpretations of heterogeneous disease processes. Right ventricular dysfunction (RVD) is an example of a multifactorial disease state in which key structural and physiologic features are captured both narratively and in structured fields, making it an ideal test case for evaluating whether LLMs can recover complex phenotypes that rules based methods routinely miss. Purpose: To compare an LLM-based extraction method to a conventional rules-based schema for identifying and phenotyping echocardiographic features associated with RVD in a large TTE dataset. Methods: MIMIC-III NOTE2NUM echocardiography reports (n = 45,794) were analyzed using GPT-4o-based LLM extraction deployed within a secure health system enclave and were benchmarked against echocardiographic measurements defined in the MIMIC-III dictionary schema. In MIMIC-III, PH was recorded qualitatively (mild/moderate/severe) based on tricuspid regurgitant (TR) jet velocity and then re-coded as present vs. absent. LLM based extraction defined RVD as (1) RV structural abnormality (>= 1 of hypertrophy, dilation, or wall hypo-/akinesis) or (2) RV pressure/volume overload (>= 2 of the following: estimated right atrial pressure > 8 mmHg, TR jet velocity > 2.8 m/s, fractional area change < 35%, tricuspid annular planar systolic excursion < 17 mm, S' < 9.5 cm/s, or E/e' > 14), with PH defined as estimated pulmonary artery systolic pressure > 35 mmHg or qualitative documentation of PH. Results: LLM extraction identified PH in 15,394 (33.6%), RV pressure/volume overload in 14,449 (31.6%), and RV structural abnormalities in 11,955 (26.1%). Co-occurrence was common: overload + structural changes in 9,380 (20.5%), overload + PH in 9,756 (21.3%), structural changes + PH in 6,183 (13.5%), and all three in 5,620 (12.3%). Using the MIMIC-III dictionary schema, PH prevalence was similar (15,371; 33.6%), but RV overload fields were captured less often (pressure overload 1,357 [3.0%], volume overload 1,128 [2.5%], pressure + volume overload 1,093 [2.4%]; any overload field 3,578 [7.8%]), and RV pressure/volume overload with PH was identified in only 731 (1.6%). Conclusions: LLM-based extraction outperforms rules-based schemas for identifying complex disease states not defined by any single variable. By synthesizing multifactorial signals, LLMs can phenotype RVD with higher fidelity and support population-level assessment. Further validation using multimodality imaging, invasive hemodynamics, and clinical outcome data is needed.

Aims: Arrhythmogenic cardiomyopathy (ACM) is a genetic disease defined by arrhythmias and myocardial fibrosis with impaired cardiac function and increased risk of sudden cardiac death. Pathogenic variants are mostly identified in desmosomal genes such as desmoglein-2 (DSG2). We identified a novel disease phenotype in patients homozygous for the DSG2 variant c.523+2T>C (splice site of exon 5/intron 5), characterized by cardioembolic events in addition to classical ACM features. Here, we evaluate this new thromboembolic phenotype by comparing the clinical data to specific murine disease models. Methods and Results: We describe three unrelated patients presenting with an embolic event and/or left ventricular thrombus. Clinical evaluation revealed a shared right ventricular ACM phenotype characterized by arrhythmias, impaired function, and fibrotic remodeling. In addition, patients exhibited localized fibrotic changes of the left ventricular apex with formation of an aneurysm and predisposition to thrombus formation. Genetic analysis identified the DSG2 variant c.523+2T>C as a founder variant from the "Bernese Oberland". To elucidate the variant's functional impact, a mouse model deficient for Dsg2 exon 5 (Dsg2{Delta}ex5) was established and compared to a model carrying the adhesion-deficient Dsg2-W2A variant. Echocardiography, ECG, and histology in Dsg2{Delta}ex5 mice revealed similar disease patterns to patients and a loss of DSG2 expression. Importantly, these animals exhibited left apical fibrosis with aneurysm formation and left ventricular thrombus formation. In contrast, the Dsg2-W2A model presented with a biventricular ACM-phenotype but without left ventricular thrombi. Conclusions: We identified a novel ACM phenotype in patients homozygous for the DSG2 founder variant c.523+2T>C characterized by left ventricular apical fibrosis. Dsg2{Delta}ex5 mice recapitulate the patients' phenotype suggesting a causative link between left ventricular aneurysm due to DSG2 deficiency and thrombus formation with subsequent embolism. This highlights a novel pathological feature of ACM and the need for variant and phenotype-specific therapy.

Heart failure (HF) is a leading cause of morbidity and mortality. We conducted multi-ancestry genome-wide association studies of 345,687 HF cases (4,468,166 individuals), and 47,192 and 46,934 cases of HF with preserved (HFpEF) and reduced ejection fraction (HFrEF), respectively, integrating plasma proteomics and multi-tissue transcriptomics to identify druggable targets. Across HF, HFrEF, and HFpEF, we identified 383 loci (166 novel) and 568 genes (375 novel). Eleven novel genes are targets of approved or investigational cardiovascular therapies, supporting indication expansion of aldosterone synthase inhibitors (CYP11B2) and type-II activin receptor antagonists (ACVR2A) to HF. Six cardiomyopathy genes were novel for HF and associated with cardiac structure and function. We identified nearly 100 genes involved in food intake and energy expenditure; metabolism of fatty acids, glucose, and branched-chain amino acids; and mitochondrial proteome, sustaining myocardial energy production. Our findings highlight the primordial role of metabolic pathways and adipokines as therapeutic targets for HF management.

Background: Right ventricular (RV) function predicts survival in pulmonary hypertension (PH) and other cardiovascular diseases, yet echocardiographic AI has largely focused on the left ventricle (LV). Objectives: To develop and evaluate PH-ECHO-AI, a unified deep learning model performing four-chamber segmentation, landmark localisation, biventricular ejection fraction (EF) estimation, deformation analysis, and PH prediction from a single apical four-chamber (A4C) clip. Methods: We developed the model using 8,416 clips from four public datasets and no institutional data: EchoNet-Dynamic, CAMUS, RVENet (apical four-chamber clips paired with 3D-echocardiographic right ventricular ejection fraction, RVEF), and MIMIC-IV-ECHO. Evaluation used held-out, training-excluded data with expert-reviewed reference standards and a per-cohort audit of patient-level separation: 1,416 clips for segmentation; 600 clips for function and deformation (350 referenced to 3D-echocardiographic RVEF, 250 to the EchoNet LVEF); and 1,076 MIMIC-IV patients for PH prediction, with five-fold cross-validation. Performance measures were Dice, correlation, mean absolute error (MAE), Bland-Altman agreement, and area under the receiver operating characteristic curve (AUC). Results: Four-chamber segmentation generalised robustly across all datasets (pooled Dice: LV 0.925, RV 0.836, LA 0.910, RA 0.904). Left ventricular ejection fraction (LVEF) was estimated with r=0.845 (95% CI 0.786 to 0.886) and MAE 4.67%. RVEF, regressed directly from the clip by a supervised head trained on 3D-echocardiographic labels with no geometric assumption, reached r=0.754 (95% CI 0.690 to 0.806) and MAE 4.98%, matching published single-view RVEF ceilings and exceeding geometric RV fractional area change (RVFAC; r=0.278). Deformation and excursion metrics, namely RV free-wall and LV A4C longitudinal strain and tricuspid and mitral annular plane systolic excursion (TAPSE, MAPSE), proved physiologically coherent. Segmentation generalised to the external MIMIC-IV cohort, and PH prediction was developed and evaluated entirely within it; RVEF evaluation was clip-disjoint and same-source, so cross-centre RVEF validation remains outstanding. Using echocardiographic geometry alone, confirmed PH was detected with an AUC of 0.697 and strong calibration (Brier 0.061). Conclusions: A single, reproducible model provides comprehensive right-heart-focused interpretation from one A4C view. It achieves RVEF accuracy competitive with dedicated RV models while simultaneously delivering segmentation, deformation, annular excursion (TAPSE and MAPSE), and PH prediction. Registration: This retrospective study used existing datasets. Code is openly released, and trained model weights are available to credentialed investigators, for independent evaluation.

Scaling laws help determine the optimal data size for training large models but are established in domains where the target is deterministic. Physiological signals are different: heartbeat sequences are stochastic, so part of the error is irreducible even with large amounts of data. Metrics such as MAE do not account for non-deterministic behavior, and therefore assessing scaling requires evaluating distributional calibration (measuring how well predicted probability densities capture true conditional characteristics). We formulate a scaling law metric(n) = E + A n- and evaluate it with five metrics: accuracy (MAE, RMSE), distributional calibration (KS distance, goodness-of-fit), and training objective (negative log loss) using a neural temporal point process trained on a cohort of four-ECG datasets. The law fits all five metrics. While point accuracy is near saturation at n = 183, KS distance and goodness-of-fit improve by 6% and 12% respectively when extrapolated to 10,000 subjects, showing that scaling decisions in stochastic domains must be guided by distributional calibration rather than point accuracy.