Left ventricular dysfunction during mavacamten therapy: risk profiling from a multicenter cohort
J Lopes, I Pereira De Miranda, D Correia, M Marques Antunes, I Cardoso, P Garcia Bras, J Miguel Viegas, I Gracio Almeida, J Bicho Augusto, B Rocha, R Cruz Ferreira, S Aguiar RosaAbstract
Introduction
Mavacamten, currently approved to treat patients with obstructive hypertrophic cardiomyopathy (oHCM), reduces left ventricular outflow tract (LVOT) gradient through inhibition of cardiac contractility. Owing to its mechanism of action, its main adverse effect is reduced left ventricular ejection fraction (LVEF), often requiring temporary treatment suspension.
Aim
To evaluate predictors of reduced LVEF in a cohort of patients treated with mavacamten.
Methods
We conducted a prospective, observational, multicenter study of symptomatic oHCM patients enrolled in an Early Access Program for mavacamten at three centers. Baseline demographic, clinical, echocardiographic, cardiac magnetic resonance (CMR), cardiopulmonary exercise testing, and electrocardiographic data were collected and analyzed as potential predictors of "systolic dysfunction" during treatment, defined as an LVEF decline to <50%, using a logistic regression model. Given the anticipated low number of events, variables with a p-value <0.100 were considered for inclusion in the multivariable analysis to assess their strength as independent predictors.
Results
73 patients were included (median age 67 years [IQR 59–72], 32.9% male). The mean resting LVOT gradient at baseline was 62 ± 37 mmHg, and 102 ± 43 mmHg with Valsalva. All patients had been receiving mavacamten for at least one month. Nine patients (10.8%) required treatment suspension due to reduced LVEF, of whom two discontinued therapy permanently. In univariate analysis, variables associated with the outcome included atrial fibrillation (p=0.054) or pacemaker rhythm (p=0.019), age (p=0.094), LVEF on baseline CMR (p=0.048), and percentage of late gadolinium enhancement (p=0.065). Mavacamten dosage was not associated with the outcome (p=0.805). In multivariable analysis, atrial fibrillation was the strongest independent predictor (p=0.020; OR 32.8). Pacemaker rhythm (p=0.066), baseline LVEF (p=0.073), and percentage of late gadolinium enhancement (LGE) (p=0.084) were borderline predictors, whereas age was not independently associated with the outcome (p=0.120).
Conclusion
In our cohort, atrial fibrillation independently predicted LVEF dysfunction during mavacamten therapy, while lower baseline LVEF, greater LGE burden on CMR, and pacemaker rhythm demonstrating borderline associations with the primary outcome.