Mavacamten induces a clinical, hemodynamic, and biomarker response beyond the primary endpoint in EXPLORER-HCM: results from a post hoc machine learning analysis

Introduction Mavacamten, a first-in-class selective inhibitor of cardiac myosin, was demonstrated in EXPLORER-HCM (NCT03470545) to be superior to placebo in achieving a primary endpoint of either (1) a ≥1.5 mL/kg/min increase in peak oxygen consumption (pVO2) and at least one New York Heart Association (NYHA) class reduction, or (2) a ≥3.0 mL/kg/min pVO2 increase without NYHA class worsening, in adults with obstructive hypertrophic cardiomyopathy (oHCM). However, the observed benefits of mavacamten were broader than the primary endpoint, suggesting a complex effect of the drug beyond improvements in these two parameters. Purpose A post hoc investigation of mavacamten clinical effects beyond the primary endpoint of EXPLORER-HCM. Methods EXPLORER data at week 30 were analyzed to evaluate improvements from baseline in primary (specified above), secondary (e.g. postexercise left ventricular outflow tract gradient and Kansas City Cardiomyopathy Questionnaire-Clinical Summary Score) and exploratory endpoints (e.g. circulating NT-ProBNP and cardiac Troponin I). Responses were classified as improved or not improved based on published thresholds, clinical standards and analyses of the EXPLORER data. Patients were grouped according to their improvement status using unsupervised hierarchical clustering. Results The cluster analysis resulted in four main groups with the following trends (Table); Group 1 = patients who met the primary endpoint and showed improvement in secondary/exploratory endpoints; Group 2 = patients with improvement in secondary/exploratory endpoints who did not meet the primary endpoint; Group 3 = patients who met the primary endpoint without substantial secondary/exploratory endpoint responses; Group 4 = patients without appreciable improvement in any endpoint. A substantially larger proportion of patients in Group 1 received mavacamten compared with placebo (88% vs. 12%, respectively). A similar trend was observed in Group 2 patients who exhibited improvements in secondary/exploratory endpoints (85% mavacamten vs. 15% placebo). Group 3 consisted predominantly of placebo-treated patients who met the primary endpoint but had negligible responses to secondary/exploratory endpoints (5% mavacamten vs. 95% placebo). Group 4 consisted predominantly of placebo-treated patients without appreciable clinically relevant responses from this analysis (10% mavacamten vs. 90% placebo). Conclusions Mavacamten was associated with clinical improvements beyond the primary endpoint of EXPLORER-HCM and was predominantly accompanied by amelioration of other measures associated with oHCM pathophysiology. In contrast, most placebo-treated patients who met the primary endpoint did not exhibit improvement in the underlying pathophysiology. These findings suggest a potential underestimation of mavacamten clinical impact based on the primary endpoint and prompt a deeper examination of mavacamten efficacy in patients with oHCM based on other clinically relevant endpoints. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): MyoKardia, Inc., a wholly owned subsidiary of Bristol Myers Squibb