Safety and Efficacy of Apitegromab in Patients With Spinal Muscular Atrophy Types 2 and 3

Abstract

Background and Objectives

Currently approved therapies for spinal muscular atrophy (SMA) reverse the degenerative course, leading to better functional outcome, but they do not address the impairment arising from preexisting neurodegeneration. Apitegromab, an investigational, fully human monoclonal antibody, inhibits activation of myostatin (a negative regulator of skeletal muscle growth), thereby preserving muscle mass. The phase 2 TOPAZ trial assessed the safety and efficacy of apitegromab in individuals with later-onset type 2 and type 3 SMA.

Methods

In this study, designed to investigate potential meaningful combinations of eligibility and treatment regimen for future studies, participants aged 2–21 years received IV apitegromab infusions every 4 weeks for 12 months in 1 of 3 cohorts. Cohort 1 stratified ambulatory participants aged 5–21 years into 2 arms (apitegromab 20 mg/kg alone or in combination with nusinersen); cohort 2 evaluated apitegromab 20 mg/kg combined with nusinersen in nonambulatory participants aged 5–21 years; and cohort 3 blindly evaluated 2 randomized apitegromab doses (2 and 20 mg/kg) combined with nusinersen in younger participants ≥2 years of age. The primary efficacy measure was mean change from baseline using the Hammersmith Functional Motor Scale version appropriate for each cohort. Data were analyzed using a paired t test with 2-sided 5% type 1 error for the mean change from baseline for predefined cohort-specific primary efficacy end points.

Results

Fifty-eight participants (mean age 9.4 years) were enrolled at 16 trial sites in the United States and Europe. Participants had been treated with nusinersen for a mean of 25.9 months before enrollment in any of the 3 trial cohorts. At month 12, the mean change from baseline in Hammersmith scale score was −0.3 points (95% CI −2.1 to 1.4) in cohort 1 (n = 23), 0.6 points (−1.4 to 2.7) in cohort 2 (n = 15), and in cohort 3 (n = 20), the mean scores were 5.3 (−1.5 to 12.2) and 7.1 (1.8 to 12.5) for the 2-mg/kg (n = 8) and 20-mg/kg (n = 9) arms, respectively. The 5 most frequently reported treatment-emergent adverse events were headache (24.1%), pyrexia (22.4%), upper respiratory tract infection (22.4%), cough (22.4%), and nasopharyngitis (20.7%). No deaths or serious adverse reactions were reported.

Discussion

Apitegromab led to improved motor function in participants with later-onset types 2 and 3 SMA. These results support a randomized, placebo-controlled phase 3 trial of apitegromab in participants with SMA.

Trial Registration Information

This trial is registered with ClinicalTrials.gov (NCT03921528).

Classification of Evidence

This study provides Class III evidence that apitegromab improves motor function in later-onset types 2 and 3 spinal muscular atrophy.

Introduction

Spinal muscular atrophy (SMA) is an autosomal recessive neuromuscular disorder of motor neurons that manifests with weakness caused by widespread skeletal muscle denervation atrophy.¹ The genetic cause of SMA involves an interaction of 2 genes: the pairing of biallelic disabling variations of the survival motor neuron 1 (SMN1) gene with variable copy number of the paralogous SMN2 gene that partly compensates for the loss of SMN1. Three SMN-targeted therapies have reversed a poor prognosis in infants and children with SMA, but there are many individuals with SMA treated after the onset of symptoms who manifest persistent denervation atrophy and weakness.^2-4 These individuals have a range of impaired motor function with substantial unmet medical need. A treatment that increases the functional capacity of intact muscle fibers within partially denervated muscle may address this unmet medical need.^5-7

Myostatin is a member of the transforming growth factor β (TGFβ) superfamily and functions as a negative regulator of skeletal muscle mass.⁸ Myostatin induction is associated with muscle wasting,^9,10 while naturally occurring and experimental conditions of diminished myostatin signaling are associated with muscle hypertrophy and increased strength. In preclinical studies, induced reduction of myostatin was associated with improved outcome measures of physical function,¹¹ and a human case study of the homozygous loss-of-function genetic variant of the myostatin gene was associated with increased muscle mass and strength.¹²

Apitegromab (SRK-015) is an investigational, fully human monoclonal antibody that specifically binds to proforms of myostatin, including promyostatin and latent myostatin, thereby inhibiting release of active myostatin.¹³ Consequently, apitegromab prevents myostatin binding to its skeletal muscle cell surface receptor, activin receptor type IIB (ActRIIB). ActRIIB signaling in myofibers has been reported to limit protein synthesis and drive muscle protein turnover.¹⁴ Previous unsuccessful therapeutic initiatives to inhibit myostatin signaling targeted active myostatin itself or its target ActRIIB^15-20 but encountered difficulties associated with undesirable off-target effects relating to binding of other TGFβ family members with close sequence homology to active myostatin that also bind the ActRIIB receptor.^21,22 Targeting an epitope within promyostatin that has high sequence divergence from the proforms of other members of the TGFβ family provides a specificity that is key to the therapeutic intent. Apitegromab binding to promyostatin and latent myostatin inhibits tolloid-mediated cleavage of latent myostatin and release of the activated isolated myostatin molecule. An important effect, however, is that apitegromab-bound uncleaved, and thus inactive, latent myostatin is released into the circulation where its concentration can be measured with antimyostatin antibodies. After apitegromab treatment, an increase in total serum myostatin levels is largely in the inactive latent myostatin form and may therefore be a marker of target engagement associated with diminished myostatin signaling.^13,23

The primary objectives of this trial were to assess safety, tolerability, and efficacy of apitegromab by changes in SMA-specific ordinal scales of motor function in participants with SMA. As a first-in-disease trial, 3 combinations of participant eligibility and dosing regimens were used to best inform the design of subsequent clinical trials. Important biochemical and pharmacokinetic (PK) data were also generated. At the time of the start of the study, only nusinersen was licensed for treatment of SMA and hence was the only SMN-enhancing therapy studied in combination with apitegromab.

Methods

TOPAZ (SRK-015-002) was a multicenter, phase 2, active treatment study to evaluate the safety and efficacy of apitegromab in participants (age 2–21 years) with types 2 and 3 SMA at 16 sites across the United States and Europe. Participants were divided into 3 cohorts: 2 open-label cohorts of participants with ambulatory type 3 SMA (cohort 1) and type 2 SMA or nonambulatory type 3 SMA (cohort 2) and 1 randomized double-blind, low–high dose assignment cohort in participants with type 2 SMA (cohort 3) (Table 1).

Table 1.

Treatment Cohorts and Trial Design

	Cohort 1	Cohort 2	Cohort 3a
	Ambulatory participants Age, 5–21 y	Nonambulatory participants Age, 5–21 y	Nonambulatory participants Age, ≥2 y
Design	Open-label, single-arm Equal arms, apitegromab alone or with nusinersen IV 20 mg/kg apitegromab every 4 wk 12-mo treatment period	Open-label, single-arm IV 20 mg/kg apitegromab every 4 wk 12-mo treatment period	Open-label, blinded randomization (1:1), to 2 mg/kg or 20 mg/kg apitegromab every 4 wk 12-mo treatment period
Participants	Ambulatory type 3 SMA Age 5–21 y (n = 23) RHS scores ≤63	Type 2 or nonambulatory type 3 SMA Concomitant nusinersen started ≥5 y old Age 5–21 y (n = 15) HFMSE scores ≥10	Type 2 SMA Concomitant nusinersen initiated before age 5 y Age ≥2 y (n = 20) HFMSE scores ≥10
Primary objectives	Safety Mean change in baseline RHS scores	Safety Mean change from baseline in HFMSE	Safety Mean change from baseline in HFMSE

Abbreviations: HFMSE = Hammersmith Functional Motor Scale Expanded; RHS, Revised Hammersmith Scale; SMA = spinal muscular atrophy.

^aFor cohort 3, randomization was performed using an interactive web-based randomization system.

Eligible participants needed an estimated life expectancy >2 years from screening and were aged 5–21 years for cohorts 1 and 2 and ≥2 years for cohort 3 at the time of screening. A documented diagnosis of 5q SMA and of type 2 or type 3 SMA before receiving any treatment with SMA-approved therapy was required. For cohort 1, a participant must have had a Revised Hammersmith Scale (RHS) score no greater than 63 at screening, and for cohorts 2 and 3, a Hammersmith Functional Motor Scale Expanded (HFMSE) score no less than 10 at screening was required. All participants received the same background SMA therapy (e.g., the approved SMN upregulator therapy nusinersen or no SMA therapy) for ≥6 months before screening and were anticipated to remain on that therapy throughout the duration of the study.

Participants were excluded from participation if they had a tracheostomy with positive pressure, used chronic daytime noninvasive ventilatory support for >16 hours daily in the 2 weeks before dosing, were anticipated regularly receiving daytime ventilator support over the duration of the study, had severe scoliosis or contractures at screening, or had any use of systemic corticosteroids within 60 days before screening. Inhaled or topical steroids were allowed. Additional selection criteria are provided in the supplemental study protocol and statistical analysis plan (eSAP 1, links.lww.com/WNL/D420 and eSAP 2, links.lww.com/WNL/D421, respectively).

Interventions

Apitegromab solution was diluted in normal saline and infused over 2 hours (±10 minutes) via peripheral IV line. If there were no acute reactions after the first 2 apitegromab doses, the infusion duration was shortened to a minimum of 1 hour. Participants receiving nusinersen were required to receive their maintenance dose ≥24 hours after receiving apitegromab or ≥14 days before a scheduled apitegromab dose. Apitegromab was administered every 28 days (±7 days after day 14) until day 364. The sponsor, participants, caregivers, investigators, and site personnel (except for the pharmacist) were blinded to the dose-level assignments of Cohort 3.

Efficacy End Points

The primary efficacy outcomes for all 3 cohorts were scores on 1 of 2 SMA-specific motor function scales; HFMSE²⁴ was used to assess individuals in the nonambulatory cohort and RHS²⁵ for the ambulatory cohort. The efficacy end point for cohort 1 was the change from baseline in the RHS total score at day 364 (visit 15) and for cohorts 2 and 3 was the change from baseline in HFMSE total score at day 364 (visit 15) (Table 1). Each item on the HFMSE can be scored with a 0, 1, or 2; a 3-point total increase in HFMSE indicates a measurable improvement in 2 or 3 scored motor item skills, and a 6-point total increase indicates improvements in 3–6 motor skills.²⁶ The RHS has high interrater and intrarater reliability from a statistical perspective and anchors this to the clinical interpretation of agreement (precision of between/within raters scoring) as ±2 points, which ensures minimal detectable change within both interrater and intrarater reliability.²⁵ The Revised Upper Limb Module (RULM) is a 20-item assessment of upper limb function in nonambulatory patients with SMA.²⁷ The 19 scored items test functions related to everyday life, such as reaching at shoulder height, placing hands from lap to the table, hands to mouth, and picking up tokens. The items are scored 0 (unable), 1 (able with modification), or 2 (able with no difficulty), and the maximum score achievable is 37. Nonambulant participant subgroups (cohorts 2 and 3) performed the RULM, which was completed by participants who were age 30 months or older at the time of the baseline assessment. Full descriptions of the HFMSE, RULM, and RHS are provided in eSAP 1 (links.lww.com/WNL/D420).

Additional analyses included safety, PK and pharmacodynamics (PD), dose-related changes in HFMSE, and time to achieve a 1-point, 3-point, or 5-point increase in HFMSE scores. A full list of objectives and efficacy analyses can be found in eSAP 1 (links.lww.com/WNL/D420), and additional publications will include results from secondary and exploratory end points.

Safety Assessments

Safety reviews were prepared approximately every 12 weeks by an independent contract research organization for review by a safety surveillance team comprising a sponsor physician, the medical monitor, an independent biostatistician, and a physician who was not a trial investigator. To assist in safety assessments, the core surveillance team could request other individuals, such as SMA experts, to review the data and participate in discussions. The safety surveillance team did not review efficacy data.

For PK, intermediate PD, and immunogenicity assessments, blood samples for determining apitegromab concentrations, serum latent myostatin concentrations, and antiapitegromab neutralizing antibodies were obtained every 28 days before each drug infusion by peripheral venipuncture; these samples could be used for additional immunogenicity testing as needed.

A physical examination was performed every 28 days before each drug infusion. A 12-lead electrocardiogram was obtained after the first 28 days and every 56 days thereafter. Comprehensive clinical laboratory assessments were made every 2 weeks for 6 weeks and every 28 days thereafter. These included clinical hematology, serum chemistry, and urinalysis. All safety laboratory analyses were performed at a central laboratory. Vital signs were assessed every 15 minutes (±5 minutes) during the infusion and 1 and 2 hours (±15 minutes) after the infusion.

Participants or their guardian were queried at each clinic visit in an open-ended manner that encouraged reporting all possible between-visit adverse events (AEs). Reported AEs were monitored until resolution or stabilization of the event. Clinical and laboratory AEs were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events, version 5.0.²⁸ Safety assessments continued for 12 weeks after the last dose of apitegromab.

Statistical Analysis

Sample sizes for each cohort were based on practical considerations. Sample sizes for cohort 2 (n = 15) and cohort 3 (n = 20) each provide 80% power (with effect sizes of 0.778 and 0.66, respectively) to reject the null hypothesis (expressed as means) relative to a prespecified alternative hypothesis using a paired t test with 2-sided 5% type 1 error for the mean change from baseline for predefined cohort-specific primary efficacy end points.

For cohort 3 (n = 20), an effect size of 1.325 was required to provide 80% power to compare the 2- and 20-mg/kg apitegromab doses, with 10 participants in each dose group for a 1:1 randomization (performed using an interactive web-based randomization system). Descriptive statistics (number of participants, mean, SD, median, minimum, and maximum) were used to summarize continuous data. Discrete measures were summarized using counts and percentages. These were performed by the contract research organization, and the sponsor did not have access to the individual efficacy data before data lock. No formal statistical testing was applied for the 12-month analyses.

Standard Protocol Approvals, Registrations, and Patient Consents

Each participating trial site obtained approval from their local Institutional Review Board (IRB) or Independent Ethics Committee. Written informed consent was obtained from participants or a parent/legal guardian before participating in any study-related activity according to local IRB requirements. Participants could withdraw consent at any time and could be terminated from the trial at any time because of protocol violation, serious or intolerable AEs, or clinically significant changes in a laboratory parameter. The procedures in this trial protocol were designed to ensure that the sponsor and investigators adhered to the principles of the International Council for Harmonisation Guidelines on Good Clinical Practice,²⁹ the Declaration of Helsinki,³⁰ and in accordance with US Investigational New Drug Regulations.³¹ This trial is registered with ClinicalTrials.gov (NCT03921528).³²

Data Availability

Anonymized data may be accessible by request from qualified investigators on completion of the apitegromab phase 3 clinical development program and execution of a data sharing agreement.

Results

The trial start date was April 22, 2019, with the primary completion date of January 12, 2021. In total, 58 participants were enrolled (Figure 1). Demographics and baseline characteristics of participants are summarized in Table 2. Because of coronavirus disease 2019 access restrictions, 4 participants missed 3 or more doses of apitegromab consecutively during the 12-month treatment period (cohort 2, n = 1; cohort 3, n = 3) and were excluded from the primary analysis per prespecified protocol criteria. Participants receiving nusinersen had received a mean of 9 doses (mean duration of 25.9 months) before enrollment. Preenrollment mean duration of nusinersen treatment for those in cohort 1, cohort 2, and cohort 3, was 19.9, 24.2, and 24.0 months, respectively.

Figure 1. Open-Label Interventional Study (12 Months).

Participants received IV apitegromab infusions every 4 weeks for 12 months in 1 of 3 cohorts to investigate participant suitability and treatment regimens. Cohort 1 assessed apitegromab 20 mg/kg in ambulatory type 3 SMA participants age 5–21 years, who were stratified into 2 equal arms receiving chronic nusinersen initiated at age ≥5 years, or who had not received SMN-targeted therapy for ≥6 months before screening. Cohort 2 assessed apitegromab 20 mg/kg in nonambulatory participants age 5–21 years who were receiving chronic nusinersen treatment initiated at age ≥5 years. Cohort 3 assessed 2 and 20 mg/kg apitegromab administered in a blinded manner to younger, earlier treated nonambulatory participants age 2 years or older who were receiving chronic nusinersen treatment initiated before age 5 years. ^aParticipants stratified based on previous treatment with approved SMN therapy. ^bParticipants randomized to receive 2 or 20 mg/kg apitegromab. ^cNumber of participants who completed the 12-month treatment period. ^dFor the primary analysis, if participants missed 3 consecutive doses because of site restrictions caused by COVID-19, records after dose skipping were excluded from analysis. This resulted in 14 participants in cohort 2, 9 participants in the cohort 3 2-mg/kg, and 8 participants in the cohort 3 20-mg/kg subcohorts having results for the 12-month end point. LOCF was used for data missing from non-COVID-19–related reasons. ^eFor this cohort 1 subset, 11 participants completed the treatment period; however, the analysis population remained at n = 12 because data from the discontinued participant was included in 12-month primary analysis. COVID-19 = coronavirus disease 2019; LOCF = last observation carried forward; SMA = spinal muscular atrophy.

Table 2.

Demographics and Baseline Characteristics

	Cohort 1		Cohort 2	Cohort 3
	Ambulatory participants Age, 5–21 y, RHS scores ≤63		Nonambulatory participants Age, 5–21 y, HFMSE scores ≥10	Nonambulatory participants Age, ≥2 y, HFMSE scores ≥10
	Apitegromab 20 mg/kg	Apitegromab 20 mg/kg + nusinersen	Apitegromab 20 mg/kg + nusinersen	Apitegromab 2 mg/kg + nusinersen	Apitegromab 20 mg/kg + nusinersen
n (baseline population)	11	12	15	10	10
Mean age, y (min, max)	12.1 (7, 19)	13.1 (7, 21)	11.7 (8, 19)	4.1 (2, 6)	3.8 (2, 6)
Mean age at diagnosis, y (min, max)	5.9 (2, 15)	4.5 (2, 15)	3.1 (1, 16)	1.2 (1, 2)	1.2 (1, 3)
Mean age at symptom onset, y (min, max)	3.7 (0.8, 11)	3.0 (0.7, 14)	1.4 (0.5, 2)	0.9 (0.5, 1.2)	1.0 (0.5, 3.5)
Female (%)	73	58	53	30	50
SMN2 gene copies, n (%)a
2	1 (9)	0	0	1 (10)	1 (10)
3	4 (36)	9 (75)	11 (73)	8 (80)	8 (80)
4	4 (36)	1 (8)	2 (13)	1 (10)	0
Mean nusinersen maintenance doses at baseline (min, max)b	N/A	3.9 (2, 6)	4.8 (2, 9)	4.8 (1, 7)
Total duration nusinersen treatment at baseline, mo (min, max)	N/A	19.9 (12, 28)	24.2 (12, 39)	24.0 (10, 34)
Discontinued	0	1c	0	0	0
Mean baseline RHS score (min, max)	47.6 (26, 63)	51.3 (43, 62)	N/A	N/A	N/A
Mean baseline HFMSE score (min, max)	N/A	N/A	22.7 (13, 39)	26.1 (12, 44)	23.5 (14, 42)
Mean baseline RULM score (min, max)	N/A	N/A	26.6 (19, 34)	25.0 (18, 34)	22.6 (15, 33)

Abbreviations: HFMSE = Hammersmith Functional Motor Scale Expanded; N/A = not applicable; RHS = Revised Hammersmith Scale; RULM = Revised Upper Limb Module.

^aData not available for all participants.

^bMaintenance dose was used as a surrogate for duration of nusinersen exposure at screening.

^cParticipant discontinued the trial for reasons unrelated to study drug.

Primary Efficacy Analysis of the Intent-to-Treat Population

The observed mean change in RHS score for cohort 1 (ambulatory type 3 SMA; n = 23) was −0.3 points from baseline. Participants receiving apitegromab without concomitant nusinersen had a mean change of −0.4 (95% CI −3.9 to 3.1) and those receiving apitegromab and nusinersen had a mean change of −0.3 (95% CI −2.0 to 1.4). Among participants receiving apitegromab alone, the numbers of participants achieving 3-point or 5-point increases in RHS scores were 3 (27.3%) and 1 (9.1%), respectively, whereas for participants receiving concomitant nusinersen, the numbers of participants achieving 3-point or 5-point increases in RHS scores were 2 (16.7%) and 0 (0.0%), respectively (Table 3).

Table 3.

Primary Efficacy Analysis, ITT Population

	Cohort 1		Cohort 2	Cohort 3
	Ambulatory participants Age, 5–21 y (RHS)		Nonambulatory participants Age, 5–21 y (HFMSE)	Nonambulatory participants Age, ≥2 y (HFMSE)
	Apitegromab 20 mg/kg	Apitegromab 20 mg/kg + nusinersen	Apitegromab 20 mg/kg + nusinersen	Apitegromab 2 mg/kg + nusinersen	Apitegromab 20 mg/kg + nusinersen
n (ITT population)	11	12a	8b	9b	8b
Primary efficacy end point	Mean change in baseline RHS scores		Mean change in baseline HFMSE scores	Mean change in baseline HFMSE scores
Mean change in baseline score (SD) (95% CI)c	−0.4 (5.20) (−3.9 to 3.1)	−0.3 (2.67) (−2.0 to 1.4)	0.6 (3.50) (−1.4 to 2.7)	5.3 (8.93) (−1.5 to 12.2)	7.1 (6.42) (1.8 to 12.5)
Participants achieving ≥1-point increase, n (%) (95% CI)	4 (36.4) (10.93 to 69.21)	5 (41.7) (15.17 to 72.33)	9 (64.3) (35.14 to 87.24)	7 (77.8) (39.99 to 97.19)	7 (87.5) (47.35 to 99.68)
Participants achieving ≥3-point increase, n (%) (95% CI)	3 (27.3) (6.02 to 60.97)	2 (16.7) (2.09 to 48.41)	4 (28.6) (8.39 to 58.10)	5 (55.6) (21.20 to 86.30)	5 (62.5) (24.49 to 91.48)
Participants achieving ≥5-point increase, n (%) (95% CI)	1 (9.1) (0.23 to 41.28)	0 (0.00 to 26.46)	2 (14.3) (1.78 to 42.81)	5 (55.6) (21.20 to 86.30)	5 (62.5) (24.49 to 91.48)

Abbreviations: COVID-19 = coronavirus disease 2019; HFMSE = Hammersmith Functional Motor Scale Expanded; ITT = intent-to-treat; RHS = Revised Hammersmith Scale.

^aOne participant in cohort 1 received concomitant treatment with an acetylcholinesterase inhibitor before and during the trial and was excluded from the per-protocol analysis because of this protocol violation.

^bFour participants missed 3 doses of apitegromab during the 12-month treatment period (cohort 2, n = 1; cohort 3, n = 3) because of COVID-19–related site access restrictions and were not included in the primary analysis.

^cMean change from baseline presented is to the month 12 end point.

Cohort 2 (type 2 or nonambulatory type 3 SMA; n = 15) had a mean increase in HFMSE scores of 0.6 points (95% CI −1.4 to 2.7) (Table 3). Most (64.3%) participants either stabilized or increased their baseline HFMSE scores during the 12 months of treatment with apitegromab. The numbers of participants achieving 3-point or 5-point increases in HFMSE scores were 4 (28.6%) and 2 (14.3%), respectively. Five participants (35.7%) had ≥2-point change from baseline on the RULM scale, achieving a significant improvement.

In cohort 3 (type 2 SMA; n = 20), the mean change from baseline at 12 months in HFMSE total score was 6.2 (95% CI 2.2 to 10.1). Participants in cohort 3 who received 20 mg/kg apitegromab (n = 8) demonstrated a mean 7.1-point improvement in HFMSE score; 5 (62.5%) achieved a ≥5-point increase in HFMSE score (Table 3), and 3 (37.5%) achieved a ≥10-point increase in HFMSE. In the 2-mg/kg arm (n = 9), the mean change from baseline in HFMSE score was a 5.3-point improvement: 5 (55.6%) participants achieved a ≥3-point increase and 5 (55.6%) achieved a ≥5-point increase. The improvement in HFMSE in both dose groups was evident at 8 weeks. Combining both dose arms permitted sufficient sample size to identify increases in the RULM post hoc, with 31.3% (5/16) of pooled participants achieving a ≥2-point increase over baseline. Participants also displayed gains in World Health Organization (WHO) motor milestones. Among the 20 participants, 1 in the 20-mg/kg arm gained 3 new WHO motor milestones, whereas another in the 20-mg/kg arm and 2 in the 2-mg/kg arm gained 1 new WHO motor milestone each.

This difference in clinical response to the 2-mg/kg and 20-mg/kg dose of apitegromab was mirrored by graded elevation of latent serum myostatin levels, a surrogate PD measure of target engagement. Participants receiving either 2 or 20 mg/kg apitegromab quickly reached a dose-proportional, high steady level of measured myostatin (Figure 2A). The elevated latent myostatin levels of the higher dose were similar across all 3 cohorts of different ages and functional abilities (Figure 2B), indicating linearity of dose to weight and age.

Figure 2. Intermediate Pharmacodynamics (A) and Pharmacokinetics (B).

Pharmacokinetic and intermediate pharmacodynamic data support clinically observed dose response and sustained drug exposure after administration of apitegromab. Both 2-mg/kg and 20-mg/kg doses yielded high levels of target engagement (>100-fold increase from baseline) assessed by serum latent myostatin levels. The 20-mg/kg dose offers higher levels of target engagement than the 2-mg/kg dose. The ambulatory cohort had the highest mean baseline latent myostatin concentrations and the highest increase in response to apitegromab; however, the percent changes from baseline were the same in both nonambulatory and ambulatory cohorts. ^‡Cohort 1: ambulatory (type 3), age 5–21 years. Participants enrolled equally into cohorts either receiving chronic nusinersen or not. Unblinded. ^†Cohort 2: nonambulatory (types 2 and 3), age 5–21 years. Unblinded. *Cohort 3: nonambulatory (type 2) participants (age ≥2 years). Blinded, randomized 1:1.

A post hoc analysis evaluated differences within this nonambulatory cohort by age, dividing by a presumption of puberty around age 13 years; those aged 5–12 years had a higher mean HFMSE increase from baseline at month 12 compared with participants aged 13–21 years (4.7 points vs −0.7 points). A higher percentage of this prepubertal subpopulation also achieved improved HFMSE scores, with 56.0% experiencing a >3-point increase in HFMSE (vs 0% for participants aged 13–21 years). One participant within this cohort also made substantial gains with 2 new WHO motor milestones; meanwhile, 2 others gained 1 additional motor milestone each.

At the time of enrollment, any existing scoliosis or contractures had to have been stable over the previous 6 months and, based on clinical judgment, thought to remain stable for the entire trial. These restrictions were intended to limit confounders of a measurable apitegromab effect during the trial. In all 3 cohorts combined, 65.5% of participants had contractures and 50.0% had scoliosis. Participants in the nonambulatory SMA cohorts 2 and 3 who did not have scoliosis at the outset of the trial (48.6% at baseline) approached mean changes of motor function improvements of 5.6 points compared with 2.1 points among those participants with scoliosis and 7.1 points compared with 2.3 points for those without and with contractures, respectively, at the time of enrollment. Participants in cohort 1 (the ambulatory type 3 SMA; pooled RHS) without scoliosis (52.2% at baseline) had increases of nearly 1 point in RHS while those with scoliosis or with contractures at the time of enrollment had a decrease of 1 point. Changes of motor function in participants with contractures at baseline followed a similar pattern.

Safety and Tolerability

The incidence and severity of AEs were consistent with the underlying SMA patient population.³³ The 5 most frequently reported treatment-emergent AEs were headache, pyrexia, upper respiratory tract infection, cough, and nasopharyngitis (Table 4). Five participants reported serious AEs that were unrelated to apitegromab: adenoidal and tonsillar hypertrophy requiring adenotonsillectomy, gait inability considered a disability (n = 2), postlumbar puncture syndrome (n = 1), viral upper respiratory tract infection (n = 1), and trial discontinuation because of muscle fatigue that began before initiation of apitegromab (n = 1). No deaths or suspected unexpected serious adverse reactions were reported. All participants tested negative for the presence of antiapitegromab antibodies. Consistent with the phase 1 healthy volunteer trial, no hypersensitivity reactions were identified.

Table 4.

Safety Results, All Participants

TEAE, n (%)	Apitegromab 2 mg/kg (n = 10)	Apitegromab 20 mg/kg (n = 48)	Total (N = 58)
Any TEAE	9 (90.0)	44 (91.7)	53 (91.4)
Serious TEAEsa	1 (10.0)	4 (8.3)	5 (8.6)
TEAEs leading to discontinuationa	0	1 (2.1)	1 (1.7)
≥Grade 3 TEAEsa	0	3 (6.2)	3 (5.2)

Preferred term, n (%)b,c	Grade 1	Grade 2	Totald	Grade 1	Grade 2	Totald	Grade 1	Grade 2	Totald
Headache	2 (20.0)	0	2 (20.0)	11 (22.9)	1 (2.1)	12 (25.0)	13 (22.4)	1 (1.7)	14 (24.1)
Upper respiratory tract infection	2 (20.0)	1 (10.0)	3 (30.0)	7 (14.6)	3 (6.3)	10 (20.8)	9 (15.5)	4 (6.9)	13 (22.4)
Pyrexia	1 (10.0)	2 (20.0)	3 (30.0)	6 (12.5)	4 (8.3)	10 (20.8)	7 (12.1)	6 (10.3)	13 (22.4)
Cough	0	3 (30.0)	3 (30.0)	8 (16.7)	2 (4.2)	10 (20.8)	8 (13.8)	5 (8.6)	13 (22.4)
Nasopharyngitis	1 (10.0)	2 (20.0)	3 (30.0)	8 (16.7)	1 (2.1)	9 (18.8)	9 (15.5)	3 (5.2)	12 (20.7)

Abbreviation: TEAE = treatment-emergent adverse event.

^aDetermined to be unrelated to treatment.

^bMedical Dictionary for Regulatory Activities.

^cMost frequently reported.

^dCombined, pooled, prespecified, and post hoc.

Additional Results

Additional results are available on ClinicalTrials.gov,³² including more demographic details and full listing of AEs. For cohort 1: change from baseline in RHS scores at prespecified time points and change from baseline in 6-minute walk test, 30-second sit-to-stand, 10-meter walk/run (from RHS), and timed rise from floor (from RHS). For cohorts 2 and 3: change from baseline in HFMSE total score at prespecified time points, proportion of participants achieving a new WHO motor development milestone relative to baseline, and proportion of participants achieving various magnitudes of change in RULM score from baseline.

Classification of Evidence

The primary objectives of this trial were to assess the safety and tolerability of apitegromab in participants with later-onset type 2 and type 3 SMA and the efficacy of apitegromab by changes in motor function outcome measures. Cohort 3 was a prospective matched cohort study with masked or objective outcome assessment in a representative population that included a primary outcome, inclusion/exclusion criteria, and adequate accounting of dropouts. Relevant baseline characteristics were presented and were considerably equivalent among treatment groups or there is appropriate statistical adjustment for differences. This study provides Class III evidence that apitegromab improves motor function in later-onset Types 2 and 3 spinal muscular dystrophy.

Discussion

There has been substantial interest and effort in developing treatments for the weakness associated with various neuromuscular disorders by inhibiting myostatin signaling. By targeting the release of active myostatin, apitegromab circumvents issues that may have undermined previous and more direct approaches having largely disappointing results.^6,7,34-37 The results of this pilot trial show the potential of selective inhibition of myostatin in SMA and suggest value in further investigations.

Multiple clinical, pathophysiologic, and epidemiologic characteristics of SMA suggested it as the ideal target for a pilot trial of apitegromab treatment of human neuromuscular disease. If demonstrable in SMA, the benefit of apitegromab treatment can be extended to other neuromuscular disorders. The biology of myostatin inhibition fits the known pathophysiology of SMA. The weakness of SMA results from denervation of skeletal muscle fibers after motor neuron degeneration; affected muscles combine normally innervated and functioning fibers with denervated atrophic fibers. Myostatin signally focuses on fast-twitch muscle fibers that are targeted in SMA.³⁸ Apitegromab substantially increases fast-twitch fiber mass in multiple animal species, including nonhuman primates. In a murine model of SMA, treatment with apitegromab substantially increased muscle strength.²³ Myostatin regulates muscle catabolism rather than anabolism, so a background of anabolic capacity is also important to drive the development of muscle-fiber size in the setting of myostatin blockade. Anabolic capacity is robust in healthy, younger individuals and diminishes with age, and there is currently a large population of children and young adults with SMA in whom this anabolic tendency may permit a more robust measurable effect from myostatin inhibition.⁶

In addition, SMA appears to be largely, if not wholly, a motor neuron cell–autonomous disorder and thus the remaining innervated muscle fibers responsible for generating muscle force are likely free of the potential confounding effects of the disease. The 3 recent successful SMN-targeted therapeutics appear to slow further motor neuron degeneration; improvements in motor function are most evident in the initial interval after initiation of treatment but slow thereafter, which permits the added beneficial effect of myostatin inhibition to be more easily identified. The presence of a wide range in the age and severity in those with residual weakness after treatment with nusinersen and specifically those who are treated after the onset of clinical symptoms enables clinical trial design that isolates these 2 factors to assess the efficacy of inhibiting myostatin.

Finally, SMA is a neuromuscular disorder for which beneficial results would have relevance to a substantial population. We directed this pilot investigation to participants who were within the middle range of well-established ordinal scales of motor function to best identify and measure associated treatment positive or negative effects. Targeting this measurable intermediate group in a pilot study does not preclude the possibility that meaningful change in function may exist in those weaker or stronger or older and younger than this study's eligibility criteria.

Although all participants received apitegromab, blinding of 20-mg/kg vs 2-mg/kg dose arms in cohort 3 provides additional evidence of a specific apitegromab treatment benefit. Although cohorts were small, shorter time to achieve HFMSE scores of >3 or >5 was seen in participants receiving the 20-mg/kg dose, but it is also apparent that participants receiving the 2-mg/kg dose eventually saw improvement in HFMSE (Figure 3, A and B). Biochemical measures of latent myostatin as a surrogate for target engagement demonstrated rapid achievement of steady state in both cohorts, but difference in levels suggested that low dose had not fully saturated the target potential for best treatment effect. Overall, age and disease duration appeared to influence the magnitude of measurable treatment benefit: Those in cohort 3 receiving either 2-mg/kg or 20-mg/kg apitegromab doses fared better than those in the nonambulatory cohort 2 and the ambulatory cohort 1. Post hoc analysis of cohort 2, comparing older and younger participants using age 12 years as an approximate measure of pubertal status, identified greater gains in the younger subset. While this tendency may be true and expected—as some preclinical evidence suggests greater effect in younger animals—additional research to determine the benefits from treatment in older participants beyond stabilization of continued progressive decline is warranted.

Figure 3. Cohort 3 Dose-Related Changes in HFMSE (A) and Time to Achieve a 1-Point, 3-Point, or 5-Point Increase in HFMSE Score During the 12-Month Treatment Period (B).

Nonambulatory age ≥2 years cohort: Dose-related changes in HFMSE scores demonstrated and both dosage arms manifested early benefit; a greater latency of the low-dose arm supports effect attributable to apitegromab. Dose-responsive improvement correlates with time to reach HFMSE motor function benefit. Benefits in both dose arms manifested as early as 2 months of treatment initiation. HFMSE = Hammersmith Functional Motor Scale Expanded.

Musculoskeletal complications of SMA, such as contracture and scoliosis, often accumulate over time in a manner that further constrains functional ability. These functional losses due to subtle advancing contracture may progress during an interval of treatment even as muscle power increases, thus confounding motor function assessment. The cohort 1 ambulatory arm showed the least improvement in this trial, though most participants maintained or showed small increases in RHS scores from baseline. This was true of both groups in the cohort, participants receiving chronic nusinersen and those naive to or not currently on any SMN-targeted therapy. Given the small sample size, further investigation of a larger population, for a longer duration, will be needed to identify treatment effects in this population. However, in patients with ambulatory SMA, stability of functional status may be a meaningful goal for treatment.

The robust and consistent results of PK and intermediate PD assessments, negative incidence of antiapitegromab antibodies, and favorable safety profile of apitegromab suggest minimal or no immunogenicity or hypersensitivity reactions and is in agreement with the phase 1 human volunteers trial.³⁹ We expect a more thorough analysis of safety with additional drug exposure data after open-label extension of TOPAZ and completion of the next randomized apitegromab trial (SRK-015-003/SAPPHIRE). Apitegromab displayed an acceptable safety profile after 12 months of therapy at doses up to 20 mg/kg. Incidence and severity of AEs were consistent with the underlying patient population and nusinersen therapy.³³ This study focused on outcomes validated in SMA; additional studies concentrating on other more proximate biological effects are worthwhile and anticipated.

The open-label design of this trial, in which most participants were on a demonstrably beneficial treatment for SMA, raises concern that observed improvements are primarily related to the background therapy. However, several features of the trial suggest that some or much of the observed improvement from baseline is indeed due to apitegromab. Specifically, both blinded arms comparing 2-mg/kg or 20-mg/kg apitegromab arms in cohort 3 manifested early benefit, but the greater latency of the low-dose arm supports attribution of the benefit to apitegromab (Figure 3). These participants were receiving an SMN-targeted therapy for which the observed benefit is most robust in the youngest and least impaired, with age the most important factor in initial rate of improvement and overall treatment response.^40-43 After the initiation of SMN-targeted therapy, the change in HFMSE improvements slows as a function of treatment duration, suggesting that improvement after SMN-targeted therapies is largely a manifestation of normal development revealed by absence of further SMA-related neurodegeneration. Against this background, the mean 7.1-point improvement in HFMSE scores seen in the younger, earlier-treated nonambulatory participants receiving 20 mg/kg apitegromab stands well above the curves seen in longer-term nusinersen follow-up studies. Importantly, trial participants had received nusinersen for a mean of 2+ years, which is well into the interval where gains associated with nusinersen plateau.^44-46 Notably, there was no clear relationship between duration of nusinersen treatment and apitegromab treatment response, especially after taking age into consideration (eFigure 1, links.lww.com/WNL/D419). Moreover, there was no baseline without treatment, so it is harder to distinguish between normal development and development while on treatment.

The post hoc observation that showed an inverse relationship between 2 musculoskeletal complications, scoliosis and contracture, and change from baseline on the respective primary functional motor scales is notable. It suggests that these orthopedic limitations identify an unrecognized intermediate-term marker of worsening and that they present a primary impediment to apitegromab action, are a marker of some other factor(s) limiting apitegromab response, or act as a constraint on ordinal functional scale (HFMSE or RHS) ascertainment of a treatment effect. This should be further investigated. In addition, another limitation for the subanalysis according to contractures and scoliosis is likely to be colinear with the effect of age and baseline HFMSE.

Apitegromab, an investigational, fully human, monoclonal antibody that inhibits myostatin activation, in this study appears to provide substantial functional gains in participants with types 2 and 3 SMA. The magnitude of effect appears to be greater in younger participants, though potential modest effects in older individuals may not have been discerned in this small pilot trial. The results of this pilot trial suggest the potential for apitegromab to result in clinically meaningful benefits and value in further randomized studies and will assist in their design. A phase 3 randomized, placebo-controlled clinical trial (SRK-015-003/SAPPHIRE) is ongoing.

Glossary

ActRIIB

activin receptor type IIB

AE

adverse event

HFMSE

Hammersmith Functional Motor Scale Expanded

IRB

Institutional Review Board

PD

pharmacodynamic

PK

pharmacokinetic

RHS

Revised Hammersmith Scale

RULM

Revised Upper Limb Module

SMA

spinal muscular atrophy

SMN1

survival motor neuron 1

TGFβ

transforming growth factor β

WHO

World Health Organization

Appendix. Authors

Name	Location	Contribution
Thomas O. Crawford, MD	Department of Neurology, Johns Hopkins University, Baltimore, MD	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data
Basil T. Darras, MD	Department of Neurology, Boston Children's Hospital, Harvard Medical School, MA	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
John W. Day, MD, PhD	Department of Neurology, Stanford University, Palo Alto, CA	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Sally Dunaway Young, PT, DPT	Department of Neurology, Stanford University, Palo Alto, CA	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Tina Duong, MPT, PhD	Department of Neurology, Stanford University, Palo Alto, CA	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Leslie Nelson, MPT, PhD	Department of Physical Therapy, University of Texas Southwestern Medical Center, Dallas	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Doreen Barrett, MS	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data
Guochen Song, MS, DRPH	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Sanela Bilic, PharmD, MBA	Vanadro, LLC, Urbandale, IA	Drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data
Shaun Cote, PhD	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Mara Sadanowicz, MSW	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Ryan Iarrobino, BA	Scholar Rock, Inc., Cambridge, MA; Tourmaline Bio, Inc., New York, NY	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Tiina Xu, BA	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Janet O'Neil, MBA	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
José Rossello, MD, PhD, MHCM	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Amy Place, PhD	Scholar Rock, Inc., Cambridge, MA; Pfizer, Inc., New York, NY	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Nathalie Kertesz, PhD	Scholar Rock, Inc., Cambridge, MA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
George Nomikos, MD, PhD	Scholar Rock, Inc., Cambridge, MA; Harmony Biosciences, Plymouth Meeting, PA	Drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data
Yung Chyung, MD	Scholar Rock, Inc., Cambridge, MA; Stealth BioTherapeutics, Needham, MA	Drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data

Study Funding

Funding for this trial and support for this publication were provided by Scholar Rock, Inc. (Cambridge, MA) and in accordance with Good Publication Practice.

Disclosure

T.O. Crawford is the lead principal investigator of the TOPAZ trial; and a consultant and/or advisory board member for AveXis/Novartis, Biogen, Pfizer, and Roche/Genentech. B.T. Darras has served as an ad hoc scientific advisory board member for AveXis/Novartis Gene Therapies, Biogen, Pfizer, Sarepta Therapeutics, Vertex, and Roche/Genentech; steering committee chair for Roche FIREFISH and MANATEE studies and DSMB member for Amicus Inc. and Lexeo Therapeutics; he has no financial interests in these companies. He has received research support from the NIH/National Institute of Neurological Disorders and Stroke, the Slaney Family Fund for SMA, the Spinal Muscular Atrophy Foundation, CureSMA, and Working on Walking Fund; received grants from Ionis Pharmaceuticals, Inc. for the ENDEAR, CHERISH, CS2/CS12 studies; from Biogen for CS11; and from AveXis, Sarepta Pharmaceuticals, Novartis (AveXis), PTC Therapeutics, Roche, Scholar Rock, and Fibrogen; and has received royalties for books and online publications from Elsevier and UpToDate, Inc. J.W. Day has received consulting fees from Biogen, Cytokinetics, Ionis Pharmaceuticals, NGT, Pfizer, Roche, and Sarepta Therapeutics; license fees or royalty payments from Athena Diagnostics; and research funding from Biogen, Cytokinetics, NGT, Roche, Sanofi-Genzyme, and Sarepta Therapeutics. S. Dunaway Young has been a member of advisory boards for Biogen, Roche/Genentech, and Scholar Rock; received personal compensation for activities with Biogen, Roche/Genentech, Scholar Rock, and CureSMA as a consultant; and received research support from CureSMA. T. Duong is an advisory board member for Biogen, CureSMA, Novartis, Roche, and Scholar Rock; and a consultant for Astellas, Avidity, Biohaven, Dyne, Genentech, Novartis, Roche, and Sarepta Therapeutics. L. Nelson is an advisory board member for Scholar Rock. D. Barrett, G. Song, S. Cote, M. Sadanowicz, T. Xu, J. O'Neil, and J. Rossello are employees of Scholar Rock. S. Bilic and N. Kertesz were employed by Scholar Rock as paid consultants at the time of this study. R. Iarrobino, A. Place, G. Nomikos, and Y. Chyung were employees of Scholar Rock at the time of this study. Go to Neurology.org/N for full disclosures.