Genotype-Phenotype Taxonomy of Hypertrophic Cardiomyopathy

Area of Science:

• Cardiology
• Genetics
• Medical Imaging

Background:

• Hypertrophic cardiomyopathy (HCM) is a significant cause of sudden cardiac death, characterized by diverse phenotypes but lacking a systematic classification framework for morphology and risk assessment.
• Understanding genotype-phenotype associations is crucial for a data-driven approach to HCM classification.

Purpose of the Study:

• To develop a data-driven taxonomy of hypertrophic cardiomyopathy (HCM) expression by quantitatively surveying genotype-phenotype associations.
• To create a systematic framework for classifying HCM morphology and assessing associated risks.

Main Methods:

• Utilized machine learning to analyze 3D left ventricular structure from cardiac MRI in 436 HCM patients.
• Built a tree-based classification of HCM phenotypes, projecting genotype and mortality risk distributions onto the tree.
• Validated the model's generalizability on an independent cohort.

Main Results:

• Identified four main phenotypic branches of HCM using unsupervised learning based on 3D shape.
• Found that carriers of pathogenic variants had lower left ventricular mass, greater basal septal hypertrophy, and reduced lifespan.
• Demonstrated associations between polygenic risk and distinct patterns/degrees of disease expression.

Conclusions:

• A data-driven taxonomy for HCM has been developed, enabling identification of patient groups with similar morphology while maintaining a continuum of disease severity, genetic risk, and outcomes.
• This novel approach offers value in comprehending the causes and consequences of HCM's diverse disease expression.