Deep learning volumetric imaging revealed MYH7 and MYBPC3 HCM variants drive concentric cellular hypertrophy, whereas SARC- wall thickening is primarily driven by extracellular matrix expansion.
Observational (n=65)
Deep learning 3D imaging reveals that HCM wall thickening is driven by both cellular and extracellular expansion in sarcomere variant patients, but primarily by extracellular expansion in non-sarcomere patients.
Abstract Introduction Hypertrophic cardiomyopathy (HCM) is a prevalent genetic heart disease characterized by asymmetric left ventricular (LV) wall thickening, myocyte disarray, and increased extracellular matrix (ECM) deposition. While cellular hypertrophy has long been cited as a defining feature of HCM, precise morphological assessment is limited by the constraints of 2D histology. Consequently, the relative contributions of cellular and extracellular remodeling to asymmetric wall thickening, and the influence of genotype and disease stage on these processes, remain poorly defined. Purpose This study investigates the relationship between genotype, cardiomyocyte (CM) morphology, and ECM remodeling in HCM. Methods We analyzed cardiac tissues from 24 obstructive HCM patients who underwent myectomy, including those with sarcomere variants (MYH7: n=8; MYBPC3: n=7) and those without sarcomere variants (SARC-: n=9), as well as 11 non-failing donor hearts. To study disease progression, a MYH7 R403Q pig model (WT: n=14, R403Q: n=16) was examined longitudinally from 2 to 13 months. Confocal microscopy and a deep learning-based image analysis pipeline enabled high-resolution 3D quantification of CM volume, eccentricity (length-to-width ratio), alignment, and extracellular volume. Single-nucleus RNA sequencing of porcine CMs and fibroblasts was integrated with morphologic data to identify gene correlates strongly associated with microarchitectural features. Linear regression models were used to assess the relative contributions of cellular and extracellular remodeling to LV wall thickening. Results Volumetric analysis of 13,000 CMs revealed that MYH7 and MYBPC3 HCM myectomy samples exhibited the most pronounced concentric cellular hypertrophy and disarray, with MYH7 hearts showing the most severe phenotype compared to control septum (Figure 1). Notably, cell volumes from SARC- myectomy were not significantly different from controls. Increased LV wall thickness in MYH7 and MYBPC3 was driven by nearly equal contributions from cellular and extracellular volume expansion, whereas SARC- wall thickening was primarily driven by ECM expansion (Figure 2). Analyses of MYH7 pig hearts confirmed progressive concentric cellular hypertrophy and revealed early fibrotic remodeling preceding CM enlargement. Integration of transcriptomic and morphometric data identified CCN2 and BAG3 as top genes correlated with CM hypertrophy, while COL1A1 was strongly associated with ECM expansion. Conclusion By leveraging advanced microscopy and deep learning-based 3D image analysis of human and pig HCM hearts across various disease stages, we uncovered previously unrecognized genotype-specific differences in CM morphology and ECM composition. These findings highlight distinct underlying biological mechanisms and underscore the need for a refined clinical classification of HCM based on genotype and remodeling patterns.Figure 1 Figure 2
Wei et al. (2025) conducted an observational in Hypertrophic cardiomyopathy (n=65). Deep learning volumetric imaging and single-nucleus RNA sequencing vs. Non-failing donor hearts and wild-type pigs was evaluated on Cardiomyocyte volume, eccentricity, alignment, and extracellular volume. Deep learning volumetric imaging revealed MYH7 and MYBPC3 HCM variants drive concentric cellular hypertrophy, whereas SARC- wall thickening is primarily driven by extracellular matrix expansion.