Cardiac Safety Outcomes in Delandistrogene Moxeparvovec Clinical Trials for Duchenne Muscular Dystrophy with Up to 5 Years of Follow-up

Abstract

Introduction

In patients with Duchenne muscular dystrophy (DMD), the gene transfer therapy delandistrogene moxeparvovec delivers a functional form of dystrophin, which has been shown to stabilize or slow disease progression.

Methods

We assessed cardiac safety of delandistrogene moxeparvovec in clinical trials with ≤ 5 years of follow-up. Data were collected from clinical trials 101 (NCT03375164, n = 4), 102 (NCT03769116, n = 41), ENDEAVOR (NCT04626674, n = 48), and EMBARK (NCT05096221, n = 125), which excluded patients with left ventricular ejection fraction (LVEF) < 40%. Adverse events and cardiac echocardiography were assessed regularly in all trials. Troponin I was assessed regularly in ENDEAVOR and EMBARK. Cardiac magnetic resonance imaging (MRI; without gadolinium enhancement) was assessed within an EMBARK substudy.

Results

Of 218 patients (baseline mean age [range], 6.4 [3.2–20.2] years; mean LVEF [range], 63.8% [48.9–78.0%]), 210 (96%) were ambulatory; 216 received delandistrogene moxeparvovec treatment. Two myocarditis cases were reported within 4 days after delandistrogene moxeparvovec infusion; both resolved within 3 weeks. Except in the two myocarditis cases, troponin I fluctuations were asymptomatic. Thirteen patients with baseline and postbaseline echocardiography data had elevated troponin I at baseline; 1 year post infusion, only one of these patients had LVEF < 50%. LVEF in all four patients with 5-year follow-up remained > 50%. Although cardiac MRI without gadolinium revealed no relevant differences in heart function between patients 1 or 2 years after delandistrogene moxeparvovec versus patients 1 year after placebo infusion, subclinical fibrosis cannot be ruled out.

Conclusion

Results from delandistrogene moxeparvovec trials with 1 to 5 years of follow-up suggest a manageable cardiac safety profile in this study population of predominantly younger, ambulatory patients with DMD who had no signs of persistent treatment-related cardiac injury.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40119-026-00457-5.

Key Summary Points

Why carry out this study?

Delandistrogene moxeparvovec, a gene transfer therapy approved for the treatment of Duchenne muscular dystrophy, delivers a functional form of dystrophin to the skeletal muscles, diaphragm, and the heart.

Cardiorespiratory failure is the leading cause of mortality in Duchenne muscular dystrophy; there is an unmet need to differentiate cardiac injury inherent to disease progression versus a potential injury due to gene therapy.

To evaluate cardiac safety of delandistrogene moxeparvovec, a comprehensive analysis of clinical trials with up to 5 years of follow-up was performed using echocardiography, magnetic resonance imaging, troponin I levels, and adverse event monitoring.

What was learned from the study?

Evaluation of 218 patients, with a mean age of 6.4 years (range 3.2–20.2) at study baseline, and with up to 5 years of follow-up after delandistrogene moxeparvovec treatment, identified two cases of treatment-related myocarditis, which resolved within 3 weeks; troponin fluctuations were mostly asymptomatic; left ventricular ejection fraction was stable over time.

Overall, in the evaluated clinical trial population consisting of younger and predominantly ambulatory patients, delandistrogene moxeparvovec has a manageable cardiac safety profile with no signs of persistent treatment-related cardiac injury.

Introduction

Duchenne muscular dystrophy (DMD) is a rare, X-linked neuromuscular disease caused by a pathogenic variant in the DMD gene encoding dystrophin protein, resulting in progressive muscle weakness, cardiorespiratory insufficiency, and premature death [1]. Cardiorespiratory failure is the most common cause of death in patients with DMD [2–4]. Therefore, any disease-modifying DMD therapy must be able to target skeletal, respiratory, and cardiac muscles [5].

Delandistrogene moxeparvovec is a recombinant adeno-associated vector (rAAVrh74)-based gene transfer therapy approved for the treatment of ambulatory patients ≥ 4 years of age with DMD and a confirmed mutation in the DMD gene in the USA and other select countries [6, 7]. The AAVrh74 serotype, which has a high tropism for skeletal and cardiac muscle [8], has been used to design a vector that can drive a robust expression of a micro-dystrophin protein in both muscle types via the MHCK7 promoter, which itself contains an enhancer that is highly active in the cardiac muscle [9].

In a rat model of DMD, treatment with delandistrogene moxeparvovec has been associated with an improvement in cardiac parameters and with reductions in the levels of troponin [10, 11]. Primary analyses of delandistrogene moxeparvovec clinical trials indicated a manageable cardiac safety profile [9, 12–15]. Despite the relatively low immunogenicity of rAAV-based vectors [16] (with 86% of individuals with DMD not having pre-existing antibodies against rAAVrh74 [17]), innate or adaptive immune-mediated damage to the heart muscle could potentially occur in patients with DMD treated with rAAV-based gene transfer therapy [18]. In a murine model of DMD, long-term follow-up of animals treated with rAAV9-based gene transfer therapy showed improved cardiac function and survival, but it also led to thickening of the intraventricular septum due to inflammation [19]. However, interactions with the immune system vary between the adeno-associated vector (AAV) serotypes [20]. Recently, a panel of pediatric cardiologists who treat patients with DMD (and which included two coauthors of this paper) issued a consensus statement advocating for the assessment of baseline cardiac function before treating patients with rAAV-based gene transfer therapy [21]. This statement also suggested a potential follow-up schema for patients who underwent gene transfer therapy, while clearly stating that these recommendations were based on expert consensus, given the lack of published data [21]. Therefore, a comprehensive, longer follow-up characterization of cardiac safety of delandistrogene moxeparvovec across the entire range of disease severity is needed.

In clinical studies, the overall safety profile of delandistrogene moxeparvovec is generally manageable with appropriate monitoring. Reported serious adverse events (AEs) include nausea, vomiting, rhabdomyolysis, myocarditis, increased transaminases, immune-mediated myositis, and acute liver injury [9, 12–15, 22, 23]. Patients with pre-existing liver impairment or active hepatic viral infection may be at increased risk for serious hepatic AEs and warrant careful evaluation prior to infusion [6, 24]. The goal of this analysis was to assess cardiac safety outcomes from clinical trials of delandistrogene moxeparvovec with 1 to 5 years of post-treatment follow-up.

Methods

Study Design and Participants

An overview of delandistrogene moxeparvovec trials in patients with DMD included in this analysis is shown in Fig. S1. Data were collected from trials 101 (NCT03375164; single-arm, open-label, phase 1/2a trial; data cutoff date, June 8, 2023; n = 4) [13], 102 (NCT03769116; randomized, double-blind, placebo-controlled, crossover trial in two parts at 48 weeks each, followed by an open-label extension phase [total observation period: 260 weeks after receiving infusion of delandistrogene moxeparvovec]; data cutoff date, September 29, 2023; n = 41) [14], ENDEAVOR (NCT04626674; a 52- to 156-week, single-arm, open-label, 7-cohort, phase 1b trial of patients with varying age, ambulatory status, and dystrophin variant type; data cutoff dates, July 2 and August 15, 2024; n = 48 [data from cohorts 1 to 5 reported here]) [15], and EMBARK, parts 1 and 2 (NCT05096221; randomized, double-blind, placebo-controlled, crossover trial in two parts at 52 weeks each; data cutoff date, October 25, 2024; n = 125) [12], with patient characteristics and duration of follow-up as indicated in Table 1.

Table 1.

Baseline characteristics of patients with DMD across pooled trials

Parameter	Study 101 [13] All patients received active treatment n = 4	Study 102 [14] (crossover trial) n = 41		Study 103 (ENDEAVOR) [15] All patients received active treatment n = 48						Study 301 (EMBARK) [12] (crossover trial) n = 125
		Received active treatment in part 1 and placebo in part 2 n = 20	Received active treatment in part 2 and placebo in part 1 n = 21	Cohort 1 n = 20	Cohort 2 n = 7	Cohort 3 n = 6	Cohort 4 n = 7	Cohort 5a n = 6	Cohort 5b n = 2	Received active treatment in part 1 and placebo in part 2 n = 63	Received active treatment in part 2 and placebo in part 1 n = 62
Ambulatory	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	No	Yes	Yes
DMD exons affected	Pathogenic varianta with frame shift (deletion or duplication) or premature stop codon mutation between exons 18 and 58			Pathogenic varianta fully contained between exons 18 and 79b				Pathogenic varianta partially or fully contained between exons 1 and 17b,c		Pathogenic varianta fully contained between exons 18 and 79d
Age at infusion, median (range), years	5.3 (4.1–6.1)	6.6 (4.5–7.9)	6.1 (4.4–8.0)	6.0 (4.4–8.0)	9.6 (8.0–12.1)	15.4 (9.9–20.3)	3.5 (3.3–4.0)	7.2 (4.7–8.7)	13.5 (12.3–14.6)	6.2 (4.1–7.9)	6.1 (4.1–8.1)
Age < 8 years at infusion, n  (%)	4 (100)	20 (100)	20 (95)	19 (95)	0	0	7 (100)	5 (83)	0	63 (100)	59 (95)
Weight range, kg	13.7–20.7	17.8–34.5	15.0–29.3	15.2–33.1	28.0–50.5	36.1–80.1	12.5–16.5	19.1–47.4	43.4–59.0	13.5–38.5	14.4–41.6
LVEF, mean (range), %	60.7e (57.0–65.0)	63.9 (54.5–70.0)	63.5 (55.0–70.0)	63.8 (53.0–69.0)	58.6 (53.0–62.6)	55.3 (48.9–62.2)	63.9 (56.4–72.0)	61.9 (55.1–68.0)	64.3 (64.0–64.6)	64.9 (55.0–77.0)	64.4 (52.0–78.0)
Troponin I, mean (range), µg/L	–	–	–	0.02f (0–0.23)	0.05 (0–0.22)	0.13 (0.01–0.47)	0.02 (0–0.05)	0.02 (0–0.11)	0.00 (0–0.01)	0.03g (0–0.59)	0.03h (0–0.81)
Follow-up from study start, mean (range), yearsi	5.0 (5.0–5.1)	4.0 (3.2–4.6)	4.0 (3.4–4.5)	3.0 (2.9–3.1)	3.0 (2.7–3.1)	3.0 (2.8–3.1)	2.0 (1.9–2.1)	1.0 (1.0–1.0)	1.0 (1.0–1.1)	2.1 (2.0–2.2)	2.1 (1.2–2.3)

Data in this table do not comprehensively represent all ongoing trials

DMD Duchenne muscular dystrophy, DMD gene dystrophin gene, LVEF left ventricular ejection fraction

^aExpected to lead to absent dystrophin

^bInitial inclusion criteria allowed for any mutations in DMD exons 1 through 79; however, an immune-mediated myositis event in a patient with a large deletion in the exon 1 to 17 region of the DMD gene prompted an update to the inclusion criteria

^cExcludes deletions that fully include exons 9–13

^dExcludes mutations fully contained within exon 45

^en = 3

^fn = 19

^gn = 62

^hn = 61

ⁱFollow-up duration = (date of censoring – infusion date + 1)/365.25

The construction and production of delandistrogene moxeparvovec (rAAVrh74.MHCK7.micro-dystrophin) have been described previously [9, 10, 15]. The therapy was administered as a single intravenous infusion at a dose of 1.33 × 10¹⁴ vg/kg, as measured using a linear-standard quantitative polymerase chain reaction (qPCR) method (equivalent to 2.0 × 10¹⁴ vg/kg with supercoiled standard qPCR) [9, 12, 14, 15].

Patients had to be on a stable dose of corticosteroids for ≥ 12 weeks prior to screening with the exception of ENDEAVOR cohort 4 (n = 7) [15], which included only patients ≥ 3 to < 4 years old who had not reached the stage of chronic steroid use and who were not receiving steroids at the time of screening. All trials excluded patients with signs of significant cardiomyopathy (including an echocardiogram [ECHO] that indicated left ventricular ejection fraction [LVEF] < 40%), as well as the patients with clinically significant laboratory values, or those with antibody titers > 1:400 against AAV8 (study 101 only [13]) or rAAVrh74 (EMBARK: ≥ 1:400 [12]).

All trials were performed in accordance with the ethical and scientific principles governing clinical research as set out in the Declaration of Helsinki, the principles of the International Council for Harmonisation’s Guideline for Good Clinical Practice, and applicable laws and regulations. All trial protocols were reviewed and approved by the appropriate institutional review boards or independent ethics committees before trial initiation; appropriate informed consent was obtained.

Procedures

Troponin I samples were collected regularly in ENDEAVOR (baseline, day 4, and weeks 1, 4, 8, 12, 24, 36, 52, 78, 104, 130, and 156) and EMBARK (baseline, day 4, and weeks 1, 4, 8, 12, 24, 36, and 52 in each of the two consecutive study periods). Troponin I levels were measured using the high-sensitivity assay, with 0.058 μg/L as the upper limit of normal (ULN).

ECHO data were captured in studies 101 (at baseline and 30 days, 1 year, and 2, 3, 4, and 5 years after infusion), 102 (part 1: baseline and weeks 4 and 48 after the first infusion; part 2: day − 1 and weeks 4 and 48 after the second infusion; part 3: weeks 78, 130, and 182 after the second infusion), ENDEAVOR (baseline and weeks 52, 104, and 156 after infusion), and EMBARK (part 1 and part 2: baseline and week 52 after infusion). A standard two-dimensional ECHO was obtained at a consistent time of day and before any invasive procedures (e.g., blood sampling, study drug infusion, or muscle biopsy). Review and interpretation were performed by local, medically qualified personnel; LVEF was calculated and noted.

Cardiac magnetic resonance imaging (cMRI) data were captured in EMBARK (at baseline and 52 weeks after infusion in each of the two consecutive study periods), within a dedicated cMRI substudy that was conducted at study sites qualified to conduct cMRI assessments. All participants at those sites (n = 39) underwent cMRI assessments. The measurements were performed using a Philips or Siemens 3T system, without gadolinium contrast. Prior to each scanning session, patients were carefully screened for MRI compatibility and safety, following local screening procedures. The captured parameters included LVEF, left ventricular end-diastolic volume, left ventricular mass, and left ventricular circumferential strain.

Electrocardiography (ECG) data were collected in studies 101 (prior to infusion, and 30 days, 1 year, and 2, 3, 4, and 5 years after infusion), 102 (part 1: prior to the first infusion [screening] and 4 and 48 weeks after the first infusion; part 2: 4 and 48 weeks after the second infusion), ENDEAVOR (at screening, on the day of infusion, and 4, 52, 104, and 156 weeks after infusion), and EMBARK (at screening, on the day of infusion, and 4 and 52 weeks after infusion within each of the two consecutive study periods).

Outcomes

Cardiac monitoring included reporting of cardiac AEs, troponin I measurements, cMRI, ECHO, pulse rate, and QTc intervals. Cardiac AEs included those classified as “cardiac disorders” within the system organ class—preferred terms system (Medical Dictionary for Regulatory Activities, versions 24.1 [studies 101 and 102] and 27.0 [ENDEAVOR and EMBARK]), plus AEs from other categories that indicate cardiac involvement (e.g., investigations: troponin I elevated). Cases of myocarditis were identified using the previously proposed Brighton collaboration criteria (Table S1) [25].

Data Analysis

Data are presented using descriptive statistics (i.e., without hypothesis testing) and safety narratives.

Role of the Funding Source

Studies 101 and 102 were funded by Sarepta. ENDEAVOR and EMBARK were co-funded by Sarepta and F. Hoffmann-La Roche. The sponsors led the collection and analysis of data, which were interpreted jointly by the sponsors and academic authors.

Results

Baseline Characteristics

Baseline characteristics of patients from individual studies have been described previously [9, 12, 14, 15]. Briefly, data were collected from 218 patients across the four studies. Of those, 210 (96%) were ambulatory and 8 (4%), from ENDEAVOR cohorts 3 and 5b, non-ambulatory (Table 1). At infusion, participants’ ages ranged from 3.3 to 20.3 years (median 6.1 years), with 197 patients (90%) being under the age of 8 years. Overall, LVEF values at baseline ranged from 48.9% to 78.0% (median 63.7%), and the median (range) follow-up was 2.1 (1.0–5.1) years (Table 1). Two patients from EMBARK who were randomized to receive placebo in part 1 of the study did not receive delandistrogene moxeparvovec after the crossover to part 2; therefore, there were 216 patients who received active treatment and in whom cardiac safety was assessed (Table 2).

Table 2.

Summary of cardiac treatment-emergent adverse events

Parameter, n (%)	Study 101 [13] n = 4	Study 102 [14] (crossover trial) n = 41				Study 103 (ENDEAVOR) [15] n = 48						EMBARK [12] (crossover trial) n = 125a				Patients on active treatment n = 216a
		Part 1: Active treatment Part 2: Placebo		Part 1: Placebo Part 2: Active treatment		Cohort 1 n = 20	Cohort 2 n = 7	Cohort 3 n = 6	Cohort 4 n = 7	Cohort 5a n = 6	Cohort 5b n = 2	Part 1: Active treatment Part 2: Placebo		Part 1: Placebo Part 2: Active treatment
		Part 1 W1–W48 n = 20	Parts 2/3b W49–W260 n = 20	Part 1 W1–W48 n = 21	Parts 2/3b W49–W260 n = 21							Part 1 W1–W52 n = 63	Part 2 W53–W104 n = 63	Part 1 W1–W52 n = 62	Part 2 W53–W104 n = 60
At least 1 cardiac TEAE	1 (25)	0	5 (25)	0	3 (14.3)	8 (40)	1 (14.3)	2 (33.3)	1 (14.3)	3 (50)	0	6 (9.5)	12 (19)	3 (4.8)	12 (20)	52 (24.1)
Bradycardia	0	0	0	0	0	0	1 (14.3)	0	0	0	0	0	0	0	0	1 (0.5)
Bundle branch block	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	1 (0.5)
Cardiac murmur	0	0	0	0	0	0	0	0	0	0	0	1 (1.6)	0	0	0	1 (0.5)
Cardiac dysfunction	0	0	0	0	0	0	0	0	0	0	0	0	1 (1.6)	0	0	1 (0.5)
Cardiomyopathy	1 (25)	0	4 (20)	0	2 (9.5)	0	1 (14.3)	0	0	0	0	0	1 (1.6)	0	0	9 (4.2)
ECHO abnormalc	0	0	0	0	0	0	0	0	0	0	0	0	1 (1.6)	0	0	1 (0.5)
EF decreased	0	0	0	0	0	0	0	0	0	0	0	0	0	0	2 (3.3)	2 (0.9)
Electrocardiogram QT prolonged	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1 (1.7)	1 (0.5)
LV dysfunction	0	0	0	0	0	0	1 (14.3)	0	0	0	0	0	1 (1.6)	1 (1.6)	0	2 (0.9)
Myocarditis	0	0	0	0	0	0	1 (14.3)	0	0	0	0	1 (1.6)	1 (1.6)	0	0	3 (1.4)
Palpitations	0	0	0	0	1 (4.8)	0	0	0	0	1 (16.7)	0	0	0	0	0	2 (0.9)
Sinus tachycardia	0	0	0	0	0	0	0	0	0	0	0	1 (1.6)	0	1 (1.6)	0	1 (0.5)
Tachycardiad	0	0	1 (5)	0	0	0	0	0	0	0	0	2 (3.2)	2 (3.2)	0	2 (3.3)	6 (2.8)
Troponin increased	0	0	0	0	0	1 (5)	0	0	0	0	0	1 (1.6)	3 (4.8)	0	2 (3.3)	7 (3.2)
Troponin I increased	0	0	0	0	0	7 (35)	0	2 (33.3)	1 (14.3)	2 (33.3)	0	2 (3.2)	6 (9.5)	2 (3.2)	5 (8.3)	25 (11.6)

Cardiac treatment-emergent adverse events (TEAEs) of interest includes the following preferred terms across the pooled trials: bradycardia, bundle branch block, cardiac murmur, cardiac dysfunction, cardiomyopathy, electrocardiogram QT prolonged, echocardiogram abnormal, ejection fraction decreased, left ventricular dysfunction, myocarditis, palpitations, sinus tachycardia, tachycardia, troponin increased, and troponin I increased

AEs were coded using MedDRA Version 24.1 (studies 101 and 102) or Version 27.0 (ENDEAVOR and EMBARK). Data cutoff dates were June 8, 2023 (study 101), September 29, 2023 (study 102), July 02, 2024 (ENDEAVOR, cohorts 1, 4, and 5), August 15, 2024 (ENDEAVOR, cohorts 2 and 3), and October 25, 2024 (EMBARK)

ECHO echocardiogram, EF ejection fraction, LV left ventricular, TEAE treatment-emergent adverse event, W weeks

^a123 of 125 patients received active treatment (delandistrogene moxeparvovec): part 1, 63 patients and part 2, 60 patients. Two patients received only placebo in part 2 and did not continue in part 2 to receive active treatment

^bStudy 102 part 3: open-label extension

^cECHO assessments deemed clinically significant at the discretion of the treating physician were reported as abnormal

^dOne patient in EMBARK study experienced tachycardia in both part 1 and part 2 and was therefore counted twice in this table. The total number of patients who experienced this AE was 6. For details, see Supplementary Table S3

Cardiac treatment-emergent adverse events (TEAEs)

Across all four trials (n = 216), 52 patients (24%) experienced cardiac TEAEs. The most common were increased troponin I (n = 25 [12%]) or troponin (n = 7 [3%]) (coded separately), cardiomyopathy (n = 9 [4%], as reported by treating physicians), and tachycardia (n = 6 [3%]) (Table 2). Two patients (1%) had decreased ejection fraction; cardiac murmur, abnormal echocardiogram, and prolonged QT interval were each observed in one patient (0.5%; Table 2). Cardiac TEAEs by patient are described further in Table S2.

In total, three patients (1%) experienced myocarditis (Table 2). Of those, one patient from ENDEAVOR cohort 2 and one from EMBARK had myocarditis that was deemed treatment-related by the treating physicians (Table S3). Both cases were identified incidentally, within 4 days of delandistrogene moxeparvovec infusion, after the patients presented for treatment of symptoms that were not cardiac specific. Some details about the EMBARK participant were published previously [12]. For both patients, myocarditis resolved within 3 weeks; a mineralocorticoid drug (spironolactone) and a beta-blocker (carvedilol) were added to the treatment regimen of the patient who had history of DMD-related cardiomyopathy (Table S3). The third case of myocarditis was reported for another patient in EMBARK, 605 days after delandistrogene moxeparvovec infusion in part 1, and approximately 8 months after crossover to placebo infusion. This event was adjudicated as not related to study treatment.

Across all participants and all visits, troponin I values in ENDEAVOR and EMBARK trials ranged from 0.06 to 0.4 μg/L and from 0.06 to 36 μg/L, respectively. In the two patients with treatment-related myocarditis, cardiac-related symptoms such as chest discomfort (patient from ENDEAVOR study; day 6 post-infusion) or tachycardia and fever (patient from EMBARK study; day 1 post-infusion) were followed by elevations in troponin I levels, which ultimately resolved (Table S3). In addition, the patient from ENDEAVOR had cMRI changes consistent with myocarditis. Notably, except in the two myocarditis cases, troponin I fluctuations were asymptomatic and without an apparent trend (Fig. 1), consistent with observations in the natural history of DMD [26].

Fig. 1.

Troponin I fluctuations in individual participants in a EMBARK, parts 1 and 2 (n = 93)^a,b and b ENDEAVOR, cohorts 1–5 (n = 46).^{a a}Each line corresponds to the troponin I levels from a single participant. For clarity, data from patients with troponin I values > 1.3 µg/L are omitted from the enlarged image. Since normal levels of troponin I in patients with DMD have not been established, the upper limit of troponin I used in both EMBARK and ENDEAVOR (0.058 µg/L) was based on observed values. ^bEMBARK was a crossover trial, in which patients were randomized 1:1 to receive delandistrogene moxeparvovec or placebo infusion at the beginning of a 52-week part 1. The treatment assignment was switched at the beginning of a 52-week part 2. Dotted lines indicate participants with more than one elevated troponin I value. DMD Duchenne muscular dystrophy, ULN upper limit of normal

Of the 109 treated patients in ENDEAVOR and EMBARK part 1 who had both troponin I and LVEF data at baseline and post-baseline, 13 had baseline troponin I values exceeding the ULN (ENDEAVOR, n = 7; EMBARK, n = 6). However, in all of these patients, troponin I levels returned to normal values post-baseline. One year after delandistrogene moxeparvovec infusion, ECHO assessments showed normal LVEF (≥ 50%) in 12 out of 13 patients who had elevated troponin values at baseline, while 1 patient had LVEF < 50% (Table S4). Assessment of left ventricular end-diastolic diameter yielded similar findings (Table S5).

Left Ventricular Parameters

The EMBARK cMRI substudy (n = 39) revealed no relevant differences in LVEF%, volume, mass, or circumferential strain between patients who received delandistrogene moxeparvovec and were followed for up to 104 weeks and patients who received either delandistrogene moxeparvovec or placebo and were followed for 52 weeks (Fig. 2). These findings were consistent with the ECHO data, obtained from the entire study sample (n = 125; Fig. 3c).

Fig. 2.

a–d Measures of left ventricular function, volume, and mass over time: EMBARK cMRI substudy (n = 39).^{a a}Point estimates at the end of part 1 (week 52) were very similar but not identical to the point estimates for the baseline of part 2. For clarity, only baseline part 2 values are shown. ^bNumbers of patients at visits indicate those with ECHO LVEF data available both at baseline and at that visit. cMRI cardiac magnetic resonance imaging, LV left ventricle, LVEF left ventricular ejection fraction

Fig. 3.

Change in LVEF (by ECHO) over time^a in a study 102 (n = 41),^b b ENDEAVOR cohorts 2, 3, and 5b^c (n = 15), and c EMBARK (n = 125). ^aNumbers of patients at visits indicate those with ECHO LVEF data available both at baseline and at that visit. ^bIn study 102, parts 1 and 2 were 48 weeks in duration each. In phase 2, participants received the double-blinded infusion at variable times following the completion of 48 weeks of phase 1. In this graph, for simplicity, phase 1 LVEF values at week 48 were considered to be identical to the values obtained 1 day before the phase 2 infusion. For the same reason, week 4 data for both phase 1 and phase 2 are not shown. ^cFor clarity, only ENDEAVOR cohorts 2, 3, and 5b are shown here, because they include late-ambulatory or non-ambulatory patients (Table 1), in whom cardiac issues would be expected. Patients with LVEF < 50% at any visit are indicated in red, purple, or blue color. The bar chart was used to enhance presentation of individual data. ECHO echocardiogram, LVEF left ventricular ejection fraction

In studies 102 and ENDEAVOR (Fig. 3), ECHO revealed stable LVEF values over 3.5 years (182 weeks) and 2 years (104 weeks) of follow-up, respectively. In ENDEAVOR, four patients from cohorts 1 and 3 showed post-baseline decline in LVEF. The patient from cohort 1 (aged 7 years at baseline) had LVEF of 58% at baseline, with values of 65% (week 52), 46% (week 104) and 42% (week 156) during follow-up. From cohort 3, patient 1 (age at baseline 11 years) had LVEF of 58% at baseline and values of 44% and 54% at weeks 52 and 104, respectively; patient 2 (age at baseline 15 years) had baseline LVEF of 49%, followed by 49% at week 52 and 44% at week 156; patient 3 (age at baseline 20 years) had LVEF of 52% at baseline and the post-baseline values of 48% (week 52), 45% (week 104), and 49% (week 156). In study 101, LVEF (assessed by ECHO) for the four patients 5 years after delandistrogene moxeparvovec infusion ranged from 51% to 65%.

ECG Parameters

Except for the case of tachycardia in a patient from EMBARK part 2 who experienced myocarditis (Table S3), there was no evidence that delandistrogene moxeparvovec infusion was associated with clinically significant changes in heart rate or rhythm abnormalities (data not shown).

Discussion

Extended follow-up of younger, predominantly ambulatory patients treated with delandistrogene moxeparvovec across four clinical trials (N = 218) showed that there were no signs of persistent treatment-related cardiac injury in the period of 1 to 5 years following treatment, although subclinical fibrosis cannot be ruled out. A key strength of this analysis is a thorough cardiac assessment of a relatively large number of patients with DMD, with a majority being evaluated for longer than 2 years.

Cardiac management is an essential part of care for patients with DMD, in whom cardiorespiratory failure is the leading cause of mortality [2]. There were three cases of myocarditis during the follow-up period, of which two were adjudicated by the treating physician as treatment-related. All resolved with treatment, suggesting that emerging cardiac issues attributed to delandistrogene moxeparvovec by investigators can be effectively managed with adequate care. Outside the follow-up period of this study (on day 400, approximately 13 months post infusion), one patient from ENDEAVOR cohort 5a experienced myocarditis following weaning of immunosuppression in the context of immune-mediated myositis [27]. The immune-mediated myositis symptoms (pain, weakness, and frequent falls) and cardiac-related events (chest pain, elevated troponin I, and decline of LVEF to 41%) were resolved and the patient became hemodynamically stable with 2 weeks of treatment [7].

Myocarditis in DMD should be considered a disease-related phenomenon, because it can occur in individuals who have not received gene transfer therapy [28–30], but an untoward effect of gene therapy cannot be excluded. For example, the two patients with myocarditis close to the infusion date in our studies (Fig. 1, Table S3) had troponin elevation at levels that generally exceed those observed in clinical practice and that would be attributed to underlying disease, especially in patients under 10 years of age, in whom cardiac complications are not expected [31, 32]. However, it needs to be pointed out that, to date, troponin I levels in asymptotic patients have not been studied systematically or to such a frequent sampling schedule during routine follow-up, which may be a source of bias. In addition, DMD-related troponin I elevations that occur after the screening but before gene therapy could be misattributed to treatment. Although current evidence indicates that rAAV-based gene transfer therapies for DMD (including delandistrogene moxeparvovec) are associated with a risk of myocarditis, the absolute risk of myocarditis following gene transfer therapy with rAAV-based vectors—and whether that risk varies between rAAV serotypes—remains to be established.

Troponin I fluctuations, which were assessed as part of the regular safety monitoring, were common in both ENDEAVOR and EMBARK, including in patients who received placebo in EMBARK part 1. Troponin I fluctuations observed in these trials are consistent with the natural history of DMD [26], and their clinical consequences, if any, remain to be determined. Although elevated troponin I is recognized as a biomarker for cardiac damage, in patients with DMD, its levels can fluctuate between normal and elevated values, with or without symptoms consistent with cardiac injury [26, 33].

The 13 patients with elevated troponin prior to infusion of delandistrogene moxeparvovec had no adverse cardiac events after infusion and no apparent difference in the safety profile compared with patients who had normal pre-infusion troponin I levels. While overall this is reassuring and suggests that patients with elevated troponin levels can be infused safely, the small sample size limits data interpretation. Troponin I levels can vary due to several factors (e.g., disease stage, physiological stress, measurement assay) and should be interpreted within the context of MRI findings and symptomatology rather than as an isolated marker. The relationship, if one exists, between troponin I, LVEF, and age in patients with DMD is unclear; an expert panel proposed conducting both retrospective and prospective analyses of troponin I values, as well as the use of high-sensitivity assays [33].

Only patients with LVEF > 40% were included in the studies analyzed here. One year after receiving delandistrogene moxeparvovec, nearly all participants had LVEF > 50%, and LVEF did not deteriorate further in patients with baseline values between 40% and 50%. Of note, EMBARK recruited patients from 4 to 7 years of age—an age range typically associated with normal cMRI values. In the EMBARK cMRI substudy (which underscores the ECHO findings), there was no evidence that viral load correlated with LVEF changes 1 year post infusion. ECG parameters were overall stable, with no apparent association with troponin I elevations or LVEF values.

Several limitations of this analysis were related to enrollment criteria of individual studies. Exclusion of patients with LVEF < 40% and the fact that nearly 90% of patients were below 8 years of age at baseline limit the generalizability of the findings, as cardiac dysfunction in DMD tends to occur in patients aged 10 or older [32]. Protocol-related limitations include the lack of cMRI follow-up at 1 to 3 months of treatment, which precluded capturing potential early myocardial changes, as indicators of inflammation following gene therapy. In addition, gadolinium-enhanced images were not obtained in the EMBARK cMRI substudy, which may have reduced the sensitivity of detecting early changes in myocardial fibrosis. This may be somewhat compensated for by the assessment of circumferential strain, which is also considered a finding that precedes left ventricular dysfunction [34]. Furthermore, differences in ECHO measurement techniques across study sites may have influenced the ECHO data. Cardiac MRI assessments with gadolinium enhancement and more frequent cMRI assessments are warranted in future evaluations of long-term cardiac safety in patients treated with delandistrogene moxeparvovec. Some of these features are incorporated into an ongoing phase 3 trial of delandistrogene moxeparvovec in ambulatory and non-ambulatory patients with DMD (ENVISION, NCT05881408).

Conclusion

Clinical trials with up to 5 years of follow-up provide evidence that delandistrogene moxeparvovec has a manageable cardiac safety profile in this younger, predominantly ambulatory patient population, with no signs of persistent treatment-related cardiac injury. It is crucial to understand cardiac involvement in DMD natural history to differentiate between myocardial inflammation and injury that is inherent to disease progression versus due to gene therapy. The ongoing multinational, randomized, placebo-controlled trial ENVISION (NCT05881408) evaluating the safety and efficacy of delandistrogene moxeparvovec in patients with DMD for up to 128 weeks will provide additional cardiac safety information in both non-ambulatory and older (≥ 8 years of age) ambulatory patients with this disease.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conception or design of the work: Jerry R. Mendell, Stefanie Mason, Mark Vivien, Alexander P. Murphy, Christoph Wandel, and James Richardson; Acquisition of data: Aravindhan Veerapandiyan, John Day, Craig M. McDonald, Jerry R. Mendell, and Craig M. Zaidman; Analysis of data: Jianfeng Meng; Interpretation of data: all authors (Jerry R. Mendell, Stefanie Mason, Mark Vivien, Alexander P. Murphy, Christoph Wandel, James Richardson, Aravindhan Veerapandiyan, John Day, Craig M. McDonald, Craig M. Zaidman, Jianfeng Meng, Jonathan H. Soslow, John Bourke); Drafting of the work or critical review: all authors. All named authors meet the International Committee of Medical Journal Editors criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Funding

Studies 101, 102, ENDEAVOR, and EMBARK are sponsored and funded by Sarepta Therapeutics, Inc., Cambridge, MA, USA. ENDEAVOR and EMBARK are also funded by F. Hoffmann-La Roche Ltd, Basel, Switzerland. The journal’s Rapid Service Fee was funded by Sarepta Therapeutics, Inc.

Data Availability

The data supporting this study are available upon reasonable request to the corresponding author, with Sarepta’s approval.

Declarations

Conflict of Interest

Aravindhan Veerapandiyan has received compensation for ad hoc advisory boards or consulting for Avidity, Biogen, Catalyst, Entrada Therapeutics, Gruenenthal, Insmed, Italfarmaco, Keros, Lupin, Mesoblast, Novartis, Pfizer, PTC Therapeutics, Precisionbio, RegenxBio, Sarepta Therapeutics, Scholar Rock, and Solid Biosciences. John Bourke is a Data Monitoring Committee member for Sarepta gene-therapy trials and is a scientific advisor to EspeRare Foundation, Pfizer, Roche, and Sarepta. Craig M. McDonald reports grants from Capricor Therapeutics, Catabasis, Edgewise Therapeutics, Epirium Bio, Italfarmaco, Pfizer, PTC Therapeutics, Santhera Pharmaceuticals, Sarepta Therapeutics, and Solid Biosciences and has a consultancy/advisory role with BioMarin, Capricor Therapeutics, Catalyst, Edgewise Therapeutics, Italfarmaco, PTC Therapeutics, F. Hoffmann-La Roche Ltd, Santhera Pharmaceuticals, Sarepta Therapeutics, and Solid Biosciences; he has received honoraria from Edgewise Therapeutics, PTC Therapeutics, and Sarepta Therapeutics. Jerry R. Mendell received study funding from Sarepta Therapeutics while at Nationwide Children’s Hospital at the time of the trials and is currently an employee of Sarepta Therapeutics and may own stock and/or stock options in the company; he is also a co-inventor of rAAVrh74.MHCK7.micro-dys technology. Jonathan Soslow serves on the Data Monitoring Committee for Sardocor, receives research support from Ametris, is a consultant with Boehringer Ingelheim, Capricor, Catalyst, Dyne, Medpace, Sarepta, Secretome, Solid, and Wave, has served on advisory boards for Sarepta and Solid Biosciences, has received payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing, or educational events from NS Pharma, and may have Secretome stocks or stock options. John Day received grants or contracts from Edgewise Therapeutics, Novartis, Roche/Genentech Pharmaceuticals, RegenxBio, and Sarepta Therapeutics; he has received consulting fees from Avidity Biosciences, Kate Therapeutics, Novartis, Sarepta Therapeutics, Scholar Rock, and Solid Biosciences. Craig M. Zaidman has received research support from Biogen and Novartis, and consulting fees from Guidepoint, and has served on an advisory board for Sarepta Therapeutics. Stefanie Mason was a Sarepta employee during the data generation and drafting of this manuscript and may own stock and/or stock options in the company. Jianfeng Meng, Mark Vivien, and James Richardson are Sarepta employees and may own stock and/or stock options in the company. Alexander P. Murphy and Christoph Wandel are Roche employees and may own stock and/or stock options in the company.

Ethical Approval

All trials were performed in accordance with the ethical and scientific principles governing clinical research as set out in the Declaration of Helsinki, the principles of the International Council for Harmonisation’s Guideline for Good Clinical Practice, and applicable laws and regulations. All trial protocols were reviewed and approved by the appropriate institutional review boards or independent ethics committees before trial initiation; appropriate informed consent was obtained.