GDF8 and activin A are the key negative regulators of muscle mass in postmenopausal females: a randomized phase I trial

Abstract

Evolutionary pressures to protect against food scarcity likely resulted in highly-conserved pathways designed to minimize energy expenditure, one of which involves the minimization of muscle mass; these mechanisms may be counter-productive in a modern world suffering from obesity and sarcopenia. Growth differentiation factor 8 (GDF8)/myostatin, acting via ActRIIA/B receptors, is the best-characterized negative regulator of muscle mass, leading to therapeutic efforts to augment muscle growth by blocking GDF8 or ActRIIA/B. ActRIIA/B blockade approximately doubles the muscle increase of GDF8 blockade, and as ActRIIA/B responds to multiple other TGFβ-family members, this implies other ligands might also regulate muscle mass. Previously, we suggested that activin A (ActA) is the key second negative regulator acting via ActRIIA/B, as blockade of both GDF8 and ActA in mice/monkeys matches the muscle growth of ActRIIA/B blockade. Here, we extend these observations to humans in a two-part, randomized, placebo-controlled Phase 1 trial ( www.clinicaltrials.gov , NCT02943239) conducted at two sites in New Zealand. Eligible subjects included healthy postmenopausal females aged 45-70 years and males aged 35-60 years not intending to father children, with a body mass index of 18-32 kg/m². Part I tested single-dose administration of anti-GDF8 alone, anti-ActA alone, several dose combinations of anti-GDF8 + anti-ActA, or placebo in healthy postmenopausal females; part II tested multiple-dose administration of anti-ActA alone or placebo in healthy postmenopausal females, combination anti-GDF8 + anti-ActA or placebo in healthy postmenopausal females, and anti-ActA alone or placebo in healthy males. The primary outcome measure was the incidence and severity of treatment-emergent adverse events through week 16 for the single-dose part of the study and through week 40 for the multiple-dose part of the study. Secondary endpoints included percent and absolute change in thigh muscle volume, percent and absolute change in total and regional body composition, pharmacokinetic profiles of the GDF8 and ActA mAbs in serum over time, changes in serum total GDF8 and total ActA levels over time, and the presence of anti-drug antibodies against the GDF8 mAb or the ActA mAb. Magnetic resonance imaging was used to quantitate changes in thigh muscle volume and dual x-ray absorptiometry was used to quantitate changes in regional body composition (total lean mass, appendicular lean body mass, android fat mass, and total fat mass). A total of 82 subjects were enrolled (48 in the single-dose part and 34 in the multiple-dose part of the study). Baseline demographic and clinical characteristics were generally balanced across the single- and multiple-dose parts of the study. Combining GDF8 and ActA blocking antibodies led to greater muscle growth than either antibody alone; increases in muscle were accompanied by reductions in fat. The observed clinical effects on muscle and fat paralleled mAb exposure in serum. The combination was generally well tolerated, and no subjects tested positive for anti-drug antibodies post-treatment. These results suggest that GDF8 and ActA are the dominant negative regulators of muscle mass in humans, and that combined blockade may be a promising therapeutic approach in muscle atrophy and obesity settings.

Fig. 1. Study design.

Study design schematics for the single-dose (part I; panel A) and multiple-dose (part II; panel B) parts of the study. ^aTwo baseline MRIs and two baseline DXAs were obtained for each subject. ^bPlacebo subjects were pooled for analysis. ActA activin A, EOS end of study, DXA dual X-ray absorptiometry, GDF8 growth differentiation factor 8, MRI magnetic resonance imaging, Q2W every 2 weeks, Q4W every 4 weeks.

Fig. 2. Flow diagram.

Flow diagram for subjects randomized to the single-dose (part I) and multiple-dose (part II) parts of the study. All subjects in part I of the trial are postmenopausal females. ^aNumber of subjects screened and excluded before randomization includes those in both the single-dose (part I) and multiple-dose (part II) parts of the study. ^bPlacebo Q4W × 2 (male). ^cAnti-ActA 10 mg/kg Q4W × 2 (male); withdrew after completing all assigned doses of study drug due to new job demands and extra responsibilities. ^dOne subject receiving Anti-ActA 10 mg/kg Q4W × 4 had study drug withdrawn due to a treatment-emergent adverse event. ActA activin A, GDF8 growth differentiation factor 8, Q2W every 2 weeks, Q4W every 4 weeks.

Fig. 3. Single-dose blockade of GDF8 and ActA increased lean body mass and decreased fat mass.

Panels A–C present the percent change from baseline to week 8 in indicators of lean body mass following a single dose of IV GDF8 mAb, ActA mAb, both in combination, or placebo: thigh muscle volume as assessed by MRI (panel A), total lean mass as assessed by DXA (panel B) and appendicular lean mass assessed by DXA (panel C). Panels D, E present the percent change from baseline to week 8 in indicators of fat mass following a single dose of IV GDF8 mAb, ActA mAb, both in combination, or placebo: android fat mass assessed by DXA (panel D), and total fat mass as assessed by DXA (panel E). Panels F–H present the percent change from baseline over time in indicators of lean body mass following a single dose of IV GDF8 mAb, ActA mAb, both in combination, or placebo: thigh muscle volume as assessed by MRI (panel F), total lean mass as assessed by DXA (panel G) and appendicular lean mass assessed by DXA (panel H). Panels I, J present the percent change from baseline over time in indicators of fat mass following a single dose of IV GDF8 mAb, ActA mAb, both in combination, or placebo: android fat mass assessed by DXA (panel I) and total fat mass as assessed by DXA (panel J). Data were least squares mean ± standard error. In panels A–E, column data labels indicate difference from placebo. Nominal P values were calculated using a two-sided t-test; P values less than 0.05 were considered statistically significant. ^aP = 0.0335; ^bP = 0.0467; ^cP = 0.0012; ^dP = 0.0025; ^eP = 0.0056; ^fP < 0.0001; ^gP < 0.0001; ^hP = 0.0018; ⁱP = 0.0013; ^jP = 0.0039; ^kP = 0.0015; ^lP = 0.0017; ^mP = 0.0009; ⁿP = 0.0242; ^oP = 0.0003; ^pP < 0.0001; ^qP = 0.0004; ^rP < 0.0001; ^sP = 0.0488; ^tP = 0.0084; ^uP = 0.0358; ^vP = 0.0173. Full analysis set presented. Source data are provided as a Source Data file. ActA activin A, DXA dual X-ray absorptiometry, GDF8 growth differentiation factor 8, mAb monoclonal antibody, MRI magnetic resonance imaging.

Fig. 4. Pharmacokinetics and total ligand concentrations in serum of healthy postmenopausal females following single-dose administration of the GDF8 mAb (trevogrumab) and the ActA mAb (garetosmab).

Pharmacokinetics are shown in panels A, C, and concentration-time profiles of the total ligands are shown in panels B, D. For the anti-GDF8 6 mg/kg + anti-ActA 1 mg/kg treatment group, no data were presented for week 12 and week 16 in panel C nor for week 16 in panel D because these values were below the LLOQ. Data were mean ± standard deviation; mean serum concentrations reported as zero are not plotted. The pharmacokinetic analysis set is presented. Source data are provided as a Source Data file. ActA activin A, GDF8, growth differentiation factor 8, LLOQ lower limit of quantification, mAb monoclonal antibody; PK pharmacokinetics.