Demographic-Based Personalized Left Ventricular Hypertrophy Thresholds for Hypertrophic Cardiomyopathy Diagnosis.
Shiwani H., Davies RH., Topriceanu C-C., Ditaranto R., Owens A., Raman B., Augusto J., Hughes RK., Torlasco C., Dowsing B., Artico J., Joy G., Miranda I., Witschey W., Rodriguez-Palomares JF., Badia-Molins C., Crotti L., Cortina-Borja M., Chuang ML., Kwong RY., Kramer CM., Manning W., Ho CY., Kellman P., Hughes AD., Biagini E., Mohiddin S., Lopes L., Litt H., Ferrari VA., Captur G., Moon JC., PRECISION-HCM Collaborative None.
BACKGROUND: Hypertrophic cardiomyopathy (HCM) is a leading cause of sudden cardiac death. Current diagnosis emphasizes the detection of left ventricular hypertrophy (LVH) using a fixed threshold of ≥15-mm maximum wall thickness (MWT). This study proposes a method that considers individual demographics to adjust LVH thresholds as an alternative to a 1-size-fits-all approach. METHODS: Left ventricular MWT was measured in 3 cohorts: a Reference Cohort of healthy adults (n = 5,067, no comorbidities), a Population Cohort (n = 43,239, with comorbidities), and an HCM Cohort from 6 international centers (n = 2,424). Measurement used cardiovascular magnetic resonance (CMR) and a validated artificial intelligence algorithm. The Reference Cohort was used to developed demographically adjusted LVH thresholds, and individualized z-scores based on age, sex, and body surface area (BSA), which were used to explore the other cohorts. RESULTS: The traditional ≥15-mm threshold classified 4.3% (n = 1,854) of the Population Cohort as hypertrophic, with a significant sex skew (89% male). Demographic-adjusted LVH thresholds (range: 10-17 mm) reduced ascertainment to 2.2% (n = 945), reducing the sex skew (56% male). Similar reductions in bias with height, weight, and age also occurred. The HCM cohort was found to have a 2:1 male-to-female ratio. A significant proportion of patients received diagnoses of HCM despite having MWT below the traditional LVH threshold (<15 mm): 27% of female individuals and 18% of male individuals. Using demographic-adjusted LVH thresholds reduced these proportions to 7% of female individuals and 15% of male individuals (P < 0.0001). Female patients had lower absolute MWT (18 mm vs 19 mm; P < 0.001) but higher MWT z-scores (5.1 vs 4.5; P = 0.05). CONCLUSIONS: Age, sex, and body size influence the normal heart MWT. Using a fixed LVH threshold ≥15 mm biases LVH ascertainment in both population and HCM cohorts. A demographic-adjusted approach for LVH improves ascertainment and diagnostic accuracy.