Mavacamten maintenance dose determination: insights into individualised therapy for hypertrophic cardiomyopathy

Introduction

Hypertrophic cardiomyopathy (HCM) is largely a sarcomeric disease characterised by increased left ventricular (LV) wall thickness without an apparent clinical cause of hypertrophy, such as aortic valve stenosis or arterial hypertension, which increase the afterload.1 2 The distribution and severity of hypertrophy can vary greatly in patients with HCM. About 70% of patients show the obstructive form of the disease with obstruction of the LV outflow tract (LVOT) at rest or on provocation.1 LVOT obstruction is an independent predictor of death related to HCM and progression of heart failure symptoms.3 According to current European guidelines, non-vasodilating ß-blockers are the first choice of medical therapy and should be titrated to effectiveness or maximally tolerated doses. For patients without ß-blocker effectiveness or intolerance, substitution with non-dihydropyridine calcium channel blockers (eg, verapamil, diltiazem) is recommended.2 If symptoms persist despite these first-line therapies, disopyramide or mavacamten is recommended for symptom management and reduction of LVOT obstruction. Although septal reduction therapies (alcohol septal ablation or septal myectomy) have level I evidence, they are only considered third-line options after pharmacological treatments have been explored.2

Mavacamten is a novel first-in-class drug, which specifically targets the underlying pathology of HCM by normalising myosin and actin interactions. This mechanism reduces hyperdynamic contraction, high energy consumption4 and LVOT gradients while even improving diastolic function.5 6 Mavacamten is supplied as oral gelatin capsules containing 2.5, 5, 10 or 15 mg with a variable terminal half life (t1/2) that depends mainly on cytochrome P450 2C19 enzyme (CYP2C19) metabolic status. CYP2C19 poor metabolisers have an increased risk of adverse events due to drug accumulation with increased plasma levels. Therefore, CYP2C19 genotyping is mandatory in Europe before prescribing mavacamten. In CYP2C19 poor metabolisers, the starting dose is 2.5 mg/day, with a maximum allowable dose of 5 mg/day. For all other phenotypes of CYP2C19 metaboliser (intermediate, normal, fast and ultrafast), the starting dose is 5 mg/day, with a maximum dose of 15 mg/day. Regular echocardiographic screening is compulsory to assess LVOT gradients and LV ejection fraction (LVEF) every 4 weeks in the first 3 months and every 12 weeks thereafter. Algorithms for initiation and maintenance dosing, patient monitoring schedules and guidance for treatment interruption or discontinuation are provided in the prescribing information.7

Evidence from the phase II and phase III clinical trials5 6 (EXPLORER-HCM and VALOR-HCM) also demonstrated that mavacamten improves symptoms, exercise capacity and biomarkers of cardiac wall stress and injury, thereby establishing its safety and efficacy. These findings led to the approval of mavacamten by the US Food and Drug Administration (FDA) in April 2022 and by the European Commission in June 2023 for the treatment of adult patients with symptomatic obstructive HCM (oHCM).

Here, we share our initial real-world experience using mavacamten for treatment of patients with symptomatic oHCM, particularly focusing on maintenance dose finding. Understanding the optimal maintenance dose is key to ensuring long-term symptom control and minimising adverse effects.

Methods

From our ongoing prospective registry, we included 36 patients with symptomatic oHCM with LVOT obstruction (resting or provoked gradient ≥50 mm Hg) who were started on mavacamten therapy between August 2023 and July 2024, according to the European Society of Cardiology guideline recommendations.2 All patients were on the maximum tolerated doses of ß-blockers or verapamil and underwent CYP2C19 genetic testing before initiating mavacamten therapy. Echocardiographic evaluations were conducted at baseline and after 4, 8 and 12 weeks of mavacamten therapy and continued further until the maintenance dose was achieved. Dose adjustments were applied during the initiating phase according to the manufacturer’s recommendations. Downtitration to 2.5 mg/day was implemented if resting and provoked LVOT gradients fell below 20 mm Hg. Uptitration was considered after 12 weeks for patients with a provoked LVOT gradient above 30 mm Hg and persisting symptoms. However, in contrast to the general recommendations, further uptitration was not pursued if the patient was classified as New York Heart Association (NYHA) class I and had a provoked LVOT gradient below 50 mm Hg, to minimise the risk of LV dysfunction.

Standard four, three and two-chamber views were recorded to evaluate LV function, and LVOT peak gradient was documented at rest and under provocation with the Valsalva manoeuvre. For patients with a resting LVOT gradient >100 mm Hg, provocation manoeuvres were deemed unnecessary, and in cases of LVOT gradient <50 mm Hg, further provocation was done with 15–20 squats. To evaluate changes in LV filling pressures, the left atrial volume index and E/E’ were recorded at baseline and after 3 months. Safety outcomes were defined as a drop in LVEF below 50%, cardiac hospitalisation or death. Cardiac hospitalisation was defined as an unplanned hospital admission due to worsening heart failure, arrhythmias (atrial fibrillation, ventricular tachycardia or ventricular fibrillation), cardiac interventions (alcohol septal ablation, septal myectomy, pacemaker implantation) or cardiogenic shock. All patients gave informed consent before taking part in the study.

Results

Between August 2023 and July 2024, thirty-six patients were initiated on therapy with mavacamten. Baseline characteristics are presented in table 1. Patients had a mean age of 60.6±12.1 years (range: 35–87 years), and 17 were female (47.2%). None of the patients had CYP2C19 poor metabolic status. 11 patients underwent genetic testing; pathogenic or likely pathogenic variants in HCM-related genes were identified in six patients. All patients were symptomatic, with 19 patients (52.8%) classified as NYHA class II and 17 patients (47.2%) as NYHA class III. Baseline medication consisted of 25 patients on ß-blockers, six patients on verapamil and five patients who received no medication due to intolerance or hypotension. Baseline medication was continued after initiating mavacamten therapy, and none of the patients were on strong CYP2C19 or CYP3A4 inhibitors that could interfere with mavacamten metabolism. Five patients had a history of septal reduction therapy (alcohol septal ablation >12 months prior) with persistent or recurrent LVOT obstruction, and nine patients had an implanted cardioverter defibrillator.

At baseline, the mean resting LVOT gradient was 67.5±36.7 mm Hg, and the provoked LVOT gradient was 112.2±31.7 mm Hg. After just 4 weeks of treatment with mavacamten, a significant reduction of both resting and provoked gradient was observed, and these reductions persisted throughout subsequent follow-ups during the initiating phase (figure 1A,B). The LV wall thickness measured at the interventricular basal septum at baseline was 21.6±4.4 mm and decreased significantly to 19.6±3.8 mm at follow-up. At baseline, mild mitral regurgitation was present in 29 patients, while five patients had moderate regurgitation. An improvement by one grade was observed in 10 patients (27.8%), while the remaining 26 patients showed no change. The LVEF decreased slightly from 66.9±5.6% at baseline to 64.0±5.1% at 4 weeks and further to 62.8±5.6% at 8 weeks and remained almost stable at 12 weeks with 62.7±5.1% (figure 1C). Though this drop in LVEF was statistically significant, no patient experienced a drop in LVEF below 50%. Hence, no temporary discontinuation of mavacamten was necessary. However, dose adjustments were required in more than 50% of patients. Only in 17 patients the initiating dose of 5 mg matched the maintenance dose. Three patients (8.3%) required a dose reduction to 2.5 mg/day, as recommended by the company, due to a provoked gradient falling below 20 mm Hg (two at 8 weeks, one at 12 weeks). 15 patients had a persistent provoked LVOT gradient >30 mm Hg at 12 weeks, necessitating an increase in mavacamten dose to 10 mg/day; of these, three required a further dose increase to 15 mg/day due to persistently high LVOT gradients. Only one of these three patients achieved significant LVOT gradient reduction and remained on the high (15 mg/day) dose, while the other two patients were discontinued from the medication after 8 weeks on the maximum dose due to lack of response. Both non-responder patients were classified as NYHA class II at baseline and exhibited no clinical improvement, with persistently elevated LVOT gradients exceeding 50 mm Hg even at rest. One additional patient initially had a dose reduction and was ultimately discontinued from mavacamten therapy at the end of the 12-week treatment period due to QT prolongation. This patient subsequently underwent alcohol septal ablation. The distribution of mavacamten dosage is shown in figure 2. Mean time to reach maintenance dose was 14 weeks (range: 8–20 weeks).

Clinically, significant improvements to mavacamten therapy were observed already after 3 months, with 19 patients (52.8%) classified as NYHA class I, 15 patients (41.7%) as NYHA class II and only two patients (5.5%) remained at NYHA class III (figure 3A). In accordance, the Kansas City Cardiomyopathy Questionnaire (KCCQ) score increased significantly from 52±18 at baseline to 73±20 points at 3 months (p<0.0001), and heart failure marker N-terminal pro B-type natriuretic peptide decreased significantly from 1707±2660 ng/mL to 558±737 ng/mL (p=0.0005) (figure 3B,C).

Regarding safety, no cardiac hospitalisations or deaths occurred during the initiating phase. The medication was generally well tolerated, with only four patients (11.1%) reporting dizziness, which resolved after a few weeks without requiring discontinuation. No syncope or atrial fibrillation occurred in this patient population, and no drop in LVEF <50% was observed. Efficacy and safety parameters are listed in table 2.

Discussion

This study presents an early real-world experience with mavacamten, the first available myosin inhibitor, in patients with oHCM from a specialised HCM centre in Europe. Consistent with findings from randomised controlled trials (RCTs), our data demonstrate that treatment with mavacamten significantly reduces LVOT obstruction and improves clinical symptoms without major adverse events. However, unlike RCTs, which usually include highly selected patient populations and adhere to very strict and rigorous protocols, real-world studies are essential for understanding the feasibility and broader applicability of this therapy in routine clinical use.

The clinical status (ie, characteristics and outcomes) observed in our cohort differed widely from those reported in the two main RCTs. In the EXPLORER-HCM trial, 72% of patients were classified as NYHA class II and only 28% as NYHA class III, with 50% of patients reaching NYHA class I after 30 weeks of treatment.5 Complete response, defined as LVOT gradient <30 mm Hg and NYHA class I, was achieved in 27% of patients. In contrast, in the VALOR-HCM trial, 93% of patients were classified as NYHA class III, of whom 37% showed no clinical response to mavacamten treatment.6 In our cohort, 53% of patients reached NYHA class I after 12 weeks of treatment, but complete response, as defined in EXPLORER-HCM, was only achieved in 16.7%. Currently, it remains unclear whether further dose increase is truly necessary to reduce the LVOT gradient below 30 mm Hg when patients already achieve clinical benefits, such as reaching NYHA class I. In the present study, the medication dose was not increased further if the patient was classified as NYHA class I and had an LVOT gradient below 50 mm Hg, to minimise the risk of systolic dysfunction. Indeed, none of the patients experienced an LVEF below 50% that necessitated temporary discontinuation of the drug. In contrast, 3.7% of patients in VALOR-HCM and 5.7% of patients in EXPLORER-HCM required discontinuation due to LV dysfunction, possibly reflecting the more rigorous treatment protocols in those trials.

A notable distinction between our study and the RCTs is the inclusion of CYP2C19 genetic testing, which is not part of the US prescription within the Risk Evaluation and Mitigation Strategy (REMS) programme where the standard starting dose of mavacamten is 5 mg/day. The European Medicines Agency mandates CYP2C19 genetic testing as an additional safety measure, requiring patients to be tested for specific genetic variants before starting mavacamten treatment.8 Variants in the CYP2C19 gene can lead to different metaboliser types: poor, intermediate, extensive, normal or ultrarapid metabolisers. If the test identifies a poor metaboliser status, characterised by two non-functional alleles, such as *2/*2, *2/*3 or *3/*3, the enzyme function is markedly reduced or absent, leading to increased drug levels and a higher risk of adverse effects, such as systolic dysfunction. Poor metabolisers require a starting dose of 2.5 mg/day, with a maximum dose of 5 mg/day, while all other phenotypes can tolerate up to 15 mg/day. In the present study, all 36 patients showed no evidence of being poor metabolisers, and the starting dose was 5 mg/day. However, the prevalence of poor metabolisers varies by ethnicity, being highest in East Asians (11.9%) and Asians (11.1%), followed by African Americans (3.4%), and lowest in Europeans (2.1%).9 This highlights the importance of personalised approaches in dose optimisation.

In the EXPLORER-HCM trial,5 dose adjustments were steered by echocardiographic LVOT gradient measurements and mavacamten plasma concentrations. In contrast, the VALOR-HCM trial6 used core lab measured echocardiographic LV function and LVOT gradients at rest and during Valsalva provocation, similar to the recommendations in the prescribing information for routine practice. However, the distribution of maintenance doses in the present study differed from that observed in VALOR-HCM. More than 80% of patients reached a maintenance dose of 5 or 10 mg/day after 16 weeks of treatment compared with 57.1% in VALOR-HCM. Only one patient in our group (2.7%) required the maximum dose of 15 mg/day, compared with 21.4% in VALOR-HCM. No patients were discontinued permanently from treatment in VALOR-HCM, whereas in our group we discontinued three patients, one due to QT prolongation and two due to lack of response. The lack of response in these two patients remains unclear but could potentially be attributed to an ultrarapid metabolic status, which induces an increased CYP2C19 enzyme activity, resulting in reduced plasma drug levels and reduced drug efficacy. In terms of cost-effectiveness, CYP2C19 genotyping usually does not always cover the full spectrum of alleles. For example, the LightMix kit (Roche Diagnostik Schweiz) used in our centre screens for *2 and *3 loss-of-function alleles to identify poor metabolisers but does not detect the gain-of-function allele *17, which may therefore explain the missed ultrarapid metaboliser status.

Real-world data on mavacamten therapy outside the USA are rare, where its prescription is governed by the REMS programme. At the John Hopkins HCM centre, 66 patients on mavacamten therapy showed results (ie, baseline and outcome data) consistent with our findings; however, 4.5% required temporary discontinuation due to LVEF dropping below 50%.10 In St Luke’s Medical Center (Milwaukee), 31 patients with similar baseline characteristics were treated with mavacamten. Surprisingly, 83.8% had a provoked LVOT gradient <30 mm Hg without requiring any dose adjustments, especially no uptitration.11 Data from Stanford University on 50 patients enrolled in the FDA-mandated REMS pathway for mavacamten therapy demonstrated significant improvements in wall thickness, mitral regurgitation, LVOT obstruction and clinical symptoms. Only two patients required temporary discontinuation of mavacamten due to an LVEF <50%, and at the most recent follow-up, mavacamten doses were distributed as follows: 2.5 mg in 20%, 5 mg in 38%, 10 mg in 29% and 15 mg in 13% of patients.12 The largest reported patient population was published from Cleveland Clinic facilities in Ohio and Florida, with 150 patients initiated on mavacamten. Results were similar, with improvements in LVOT gradients and clinical symptoms, with only 2% of patients requiring temporary interruption of mavacamten due to LVEF <50%. Interestingly, 40 patients reported no symptomatic improvement despite significant reductions in Valsalva LVOT gradients.13 Additionally, only 30% of patients had a Valsalva LVOT gradient >30 mm Hg at 12 weeks, which is much lower compared with our patient group, where 72.2% of patients had LVOT gradient >30 mm Hg, and requiring dose adjustments. A longer treatment duration may reduce this proportion of patients with higher LVOT gradients, as demonstrated in the MAVA-LTE study. At week 84, 85% of patients who were followed up at that time had Valsalva LVOT gradients <30 mm Hg.14

These differences in maintenance dose, clinical outcomes and haemodynamic responses underscore the importance of a personalised approach to improve outcomes by tailoring therapy to each patient’s tolerability and response. It is becoming evident that HCM is a complex and highly heterogeneous disease, and not all patients will respond to mavacamten. Furthermore, implementing mavacamten therapy is more demanding than first-line medications like ß-blockers or verapamil, requiring multiple echocardiograms for monitoring and dose titration. This places a significant strain on the healthcare system, particularly on outpatient clinics and private practice cardiologists. Aficamten, the second myosin inhibitor that is currently under investigation but yet not available, might address some of the limitations of mavacamten therapy. It has a wider therapeutic window, has a shorter half life of 3–4 days and reaches steady state within 2 weeks. Importantly, aficamten has no substantial cytochrome P450 induction or inhibition in preclinical assessment.15 The phase II REDWOOD-HCM study and the phase III SEQUOIA-HCM trial demonstrated promising results of aficamten in terms of LVOT gradient reduction and clinical improvement.16 17

It is currently unclear whether myosin inhibitors provide a definitive treatment for patients with HCM or merely delay the need for septal reduction therapies. Long-term data from randomised trials and real-world studies are necessary to resolve these unanswered questions. Nevertheless, cardiac myosin inhibitors have the potential to be game changer alternatives in the treatment cascade of HCM.

Conclusion

Based on these real-world data, we conclude that treatment of patients with oHCM with the only myosin inhibitor currently available, mavacamten, is safe and effective in the initiating phase. Implementing mavacamten therapy requires CYP2C19 genetic testing, frequent echocardiographic monitoring and dose adjustments to ensure patient safety. The challenge of non-responders to mavacamten is currently unresolved, and these patients will continue to require septal reduction therapies. Additionally, data on long-term safety and efficacy are necessary to determine mavacamten’s place in the treatment cascade of oHCM.