Left ventricular mass by three-dimensional echocardiography is associated with myocardial replacement fibrosis and ventricular arrhythmias in hypertrophic cardiomyopathy

Abstract Background In hypertrophic cardiomyopathy (HCM) patients, quantification of left ventricular (LV) mass carries important prognostic implications. Two-dimensional echocardiography (2DE) has limited accuracy for LV mass calculation, due to plane position errors, geometric assumptions and asymmetric distribution of LV hypertrophy in HCM. Purpose We aimed to explore: (1) the accuracy of three-dimensional echocardiography (3DE) vs 2DE to quantify LV mass in HCM compared to cardiac magnetic resonance (CMR); (2) the relationship of 3DE LV mass with non-sustained ventricular tachycardia (NSVT) and late gadolinium enhancement (LGE)≥15% by CMR. Methods In consecutive HCM patients referred to our Cardiomyopathy Clinic between 2020 and 2023, 2DE and 3DE were used to assess LV mass. LV systolic function was assessed by 3DE ejection fraction (LVEF) and peak global 2D longitudinal strain (2DGLS). Clinical, 24h ECG Holter and CMR data were collected. Results A total of 180 HCM patients (pts, age 58±18 years, 55% men) were enrolled. Apical HCM was present in 56 pts (31%) and obstructive HCM in 69 pts (38%). Maximal LV wall thickness (MWT) by 2DE was 19.5±4.6 mm. LV mass was 150±51 g/m2 by 2DE, 80±25 g/m2 by 3DE, and 79±26 g/m2 by CMR. Fifty-seven pts (32%) had evidence of NSVT at ECG Holter monitoring. LGE≥15% was present in 32% pts. Aim #1: In a subset of 63 pts who underwent CMR, 3DE LV mass was strongly correlated with CMR LV mass (r=0.85, p<0.001), while 2DE LV mass was not (p=0.38). LV mass by 3DE showed a better agreement with LV mass (bias 3.8 g/m2, LOA -25 to 32 g/m2) by CMR than 2DE (bias 68 g/m2, LOA -35 to 172 g/m2). Aim #2: In the entire cohort, 3DE LV mass had a stronger association compared to 2DE LV mass with the presence of LGE≥15% (AUC 0.68 for 3DE versus 0.56 for 2DE, p=0.08) and NSVT (AUC 0.65 for 3DE versus 0.54 for 2DE, p=0.06). By multivariable analysis, LV mass by 3DE was an independent predictor of LGE≥15% (HR 1.03) and of NSVT (HR 1.03), outperforming 2DGLS and MWT in the latter regression. Using the Youden Index from ROC curve, the optimal cutoff for predicting LGE≥15% using 3DE mass was 87 g/m² (sensitivity 47%, specificity 91%). The addition of 3DE LV mass to a model including 2DE MWT, 2DE LV mass and 2DGLS had a significant incremental value for the prediction of LGE≥15% (Figure 1). Conclusions In HCM patients, LV mass by 3DE was strongly correlated to LV mass by CMR and was an independent predictor of significant LV myocardial fibrosis and ventricular arrhythmias. In centers with low access to CMR, implementation of 3DE to measure LV mass in HCM patients may improve arrhythmic risk stratification compared to 2DE.