Preclinical DMD gene therapy GB703 debuts as GEMMABio advances SMA program

Gemma Biotherapeutics reported new preclinical data for 2 investigational gene therapies at the American Society of Gene and Cell Therapy 2026 Annual Meeting: GB703 for Duchenne muscular dystrophy (DMD) and GB221 for spinal muscular atrophy type 1 (SMA1). The company said the DMD data support candidate declaration for GB703, while GB221 remains in early clinical testing in the phase 1/2 CHARISMA study in infants with SMA1.¹

“We are excited to announce Duchenne muscular dystrophy as our latest named program,” James M. Wilson, MD, PhD, chief executive officer of Gemma Biotherapeutics, said in the company announcement.¹ He added that the company is pursuing investigational new drug–enabling studies for GB703. For clinicians, the update is notable less as a practice-changing development than as an early signal of where next-generation neuromuscular gene therapy is heading: lower-dose muscle targeting in DMD and strategies to reduce dorsal root ganglia toxicity in SMA.

According to the company, GB703 is an adeno-associated virus (AAV) gene therapy designed to deliver a hybrid micro-utro/dystrophin payload using a myotropic capsid, GCap104.¹ In mouse and nonhuman primate studies, the vector targeted skeletal and cardiac muscle, and expression of the hybrid transgene was observed in nonhuman primates after intravenous administration at 3 × 10^13 genome copies/kg over 60 days, with tolerability reported in that study period.¹ In a mouse model of DMD, transgene expression was associated with restoration of the β-dystroglycan complex, recovery of sarcolemmal neuronal nitric oxide synthase binding, and reductions in markers of muscle damage.¹

That approach attempts to address 2 continuing problems in DMD gene therapy. First, approved and investigational microdystrophin constructs are constrained by AAV packaging limits and may not suit all mutation subtypes.² Second, systemic AAV delivery has raised concerns about dose-related toxic effects, including hepatotoxicity, thrombocytopenia, and myocarditis in some gene therapy settings.² Delandistrogene moxeparvovec, the first US Food and Drug Administration–approved gene therapy for DMD, received accelerated approval in 2023 and expanded labeling in 2024, but questions remain about durability, patient selection, and comparative performance across constructs.³

Gemma said GB703 incorporates a self-utrophin–derived exon 8-11 region intended to broaden eligibility, including for some patients with N-terminal dystrophin deletions who may be poor candidates for other microdystrophin strategies.¹ That rationale is biologically plausible, but the current evidence remains preclinical, and no human efficacy or safety data for GB703 were presented in the announcement.

For SMA1, the company presented preclinical findings supporting GB221, an investigational AAVhu68-based therapy carrying codon-optimized SMN1 under a moderate-strength promoter with microRNA target sequences intended to suppress expression in dorsal root ganglia neurons.1 Sensory ganglion toxicity has been an important issue in systemic AAV gene transfer, including in preclinical and clinical development programs, making tissue-selective expression a relevant design goal.⁴

In a severe SMA mouse model, Gemma reported survival beyond 120 days in more than 50% of treated animals compared with less than 15 days in untreated mice; righting reflex improved by day 14 after treatment, and surviving animals reportedly had normal strength and trunk control by weaning.¹ In nonhuman primates, intra-cisterna magna administration led to transgene expression in 20% to 90% of motor neurons without observed dorsal root ganglia toxicity, sensory neuropathy, or liver or cardiac toxic effects at the maximum feasible dose, according to the company.1

The SMA therapeutic landscape is already crowded, with onasemnogene abeparvovec, nusinersen, and risdiplam changing survival and motor outcomes when used early, especially in presymptomatic infants.⁴ GB221 therefore enters a field in which any new therapy will need to show either a clearer safety advantage, improved biodistribution, or practical benefits related to route of administration or access. The ongoing first-in-human CHARISMA trial (NCT07070999) is enrolling children younger than 12 months in Brazil to assess safety, tolerability, and early efficacy signals.¹

The main limitation of the current update is that it is based largely on conference-stage preclinical findings described in a company press release rather than peer-reviewed full datasets. Duration of follow-up, reproducibility across models, and translation to human benefit remain unknown for both programs. Near-term attention will likely focus on whether GB703 reaches the clinic and whether CHARISMA yields a safety profile for GB221 that supports continued development.

References

1. Gemma Biotherapeutics. GEMMABio unveils Duchenne muscular dystrophy (DMD) program and presents preclinical data for next-generation gene therapies for DMD and spinal muscular atrophy type 1 (SMA1) at the ASGCT 2026 annual meeting. PR Newswire. May 14, 2026. Accessed May 14, 2026. https://www.prnewswire.com/news-releases/gemmabio-unveils-duchenne-muscular-dystrophy-dmd-program-and-presents-preclinical-data-for-dmd-and-spinal-muscular-atrophy-type-1-sma1-at-the-asgct-2026-annual-meeting-302771279.html

2. Duan D. Systemic AAV Micro-dystrophin Gene Therapy for Duchenne Muscular Dystrophy. Mol Ther. 2018;26(10):2337-2356. doi:10.1016/j.ymthe.2018.07.011

3. Elevidys. Prescribing information. Sarepta Therapeutics; 2024. Accessed May 14, 2026. https://www.fda.gov/files/vaccines,%20blood%20&%20biologics/published/Package-Insert-ELEVIDYS_1.pdf

4. Strauss KA, Farrar MA, Muntoni F, et al. Onasemnogene abeparvovec for presymptomatic infants with two copies of SMN2 at risk for spinal muscular atrophy type 1: the Phase III SPR1NT trial. Nat Med. 2022;28(7):1381-1389. doi:10.1038/s41591-022-01866-4