Cardiac Myosin Inhibition in Heart Failure With Normal and Supranormal Ejection Fraction

Importance Patients with heart failure with preserved ejection fraction (HFpEF) who have left ventricular ejection fraction (LVEF) of 60% or greater have limited treatment options. Objective To examine the effects of cardiac myosin inhibition with mavacamten in patients with HFpEF with LVEF of 60% or greater. Design, Setting, and Participants The EMBARK-HFpEF trial was a phase 2a, open-label, single-arm, multicenter trial conducted from November 6, 2020, to February 26, 2024, at 20 sites in the US and Canada. Patients with symptomatic HFpEF (defined as a New York Heart Association [NYHA] functional class II or III), LVEF of 60% or greater, elevated N-terminal pro-B-type natriuretic peptide (NTproBNP), and left ventricular hypertrophy were eligible for inclusion. Intervention Mavacamten treatment for 26 weeks, starting at 2.5 mg and potentially titrated up to 5 mg at week 14 based on prespecified LVEF and NTproBNP criteria. Main Outcomes and Measures Primary efficacy end points (measured as the change from baseline to week 26) included NTproBNP and high-sensitivity troponin T (hsTnT); additional efficacy end points included changes in high-sensitivity troponin I (hsTnI), NYHA functional class, and echocardiographic parameters (resting and peak exercise). Safety end points included treatment-emergent adverse events and reductions in LVEF to less than 30%. Results A total of 30 patients were enrolled and treated with mavacamten. Median (IQR) patient age was 76 (70-80) years, and 16 patients (53.3%) were female. From baseline to week 26, mavacamten was associated with reductions in NTproBNP (mean reduction, −26%; 95% CI, −44% to −4%; P = .04), hsTnT (mean reduction, −13%; 95% CI, −23% to −3%; P = .02), and hsTnI (mean reduction, −20%; 95% CI, −32% to −6%; P = .01). Cardiac biomarker values returned toward baseline levels 8 weeks after drug discontinuation. NYHA class improved in 10 of 24 patients (41.7%) who had evaluable NYHA class data at the end of treatment, and improvements in echocardiographic markers of LV diastolic function were observed. Mean LVEF decreased by 3.2 absolute percentage points (95% CI, 1.1-5.4; P = .005) during treatment. Mavacamten was interrupted in 3 patients (10% of the study population; 95% CI, 2.1%-26.5%) due to protocol prespecified criteria of LVEF less than 50% (n = 2) or a more than 20% relative decrease from baseline (n = 1; nadir LVEF, 58%), with LVEF recovery observed in all 3 patients. There were no deaths or instances of LVEF less than 30%; 1 patient had worsening heart failure deemed unrelated to the study drug. Conclusions and Relevance In an open-label trial in patients with HFpEF with LVEF of 60% or greater, mavacamten was associated with improvements in biomarkers of cardiac wall stress and injury, with no sustained reductions in LVEF observed. Trial Registration ClinicalTrials.gov Identifier: NCT04766892