May 14, 2026 —
Gemma Biotherapeutics, or GEMMABio, announced new preclinical data at the ASGCT 2026 Annual Meeting supporting two next-generation AAV gene therapy programs: GB703 for Duchenne muscular dystrophy (DMD) and GB221 for spinal muscular atrophy type 1 (SMA1). The data support candidate declaration for GB703 and further support the ongoing CHARISMA Phase 1/2 clinical trial of GB221 in pediatric SMA1 patients.
GB703 is GEMMABio’s newly named investigational gene therapy candidate for DMD, designed to address two major challenges in the field: improving muscle delivery at lower doses and broadening mutation coverage beyond current microdystrophin-based approaches. DMD is a severe X-linked neuromuscular disease caused by loss-of-function mutations in the dystrophin gene, leading to progressive skeletal, respiratory, and cardiac muscle deterioration.
GB703 uses an engineered myotropic AAV capsid, GCap104, designed to target skeletal and cardiac muscle after intravenous administration. The therapy carries a de-immunized, miniaturized hybrid micro-utro/dystrophin transgene, incorporating a self-utrophin-derived exon 8–11 region. This design may potentially expand eligibility to patients with N-terminal dystrophin deletions who may not be suitable candidates for some existing or investigational microdystrophin approaches due to immune response concerns.
In preclinical studies, GCap104 supported hybrid micro-utro/dystrophin expression in non-human primates at 3×10¹³ GC/kg by IV administration and was well tolerated over a 60-day study. In a mouse model of DMD, the hybrid transgene showed measurable biological effects, including restoration of the β-dystroglycan complex, nNOS sarcolemmal binding, and reductions in markers of muscle damage. GEMMABio is advancing GB703 through IND-enabling studies to support future clinical development.
GEMMABio also presented comprehensive preclinical data for GB221, a next-generation investigational AAV gene therapy for SMA1. GB221 uses AAVhu68 to deliver a codon-optimized SMN1 gene under a ubiquitous, moderate-strength human promoter. The therapy is designed for one-time delivery directly to the cerebrospinal fluid via intra-cisterna magna injection.
A key design feature of GB221 is its use of microRNA target site technology to suppress SMN1 transgene expression in dorsal root ganglia neurons, with the goal of reducing the risk of peripheral sensory neuropathy and overexpression-related toxicity. In a severe SMA mouse model, GB221 improved survival, with more than 50% of treated mice surviving to 120 days compared with less than 15 days in untreated animals. Treated mice also showed improved strength and trunk control.
In non-human primates, GB221 achieved SMN1 expression in 20–90% of motor neurons without observed dorsal root ganglia toxicity, sensory neuropathy, liver toxicity, cardiac toxicity, or other major toxicities at the maximum feasible dose. These findings support the ongoing CHARISMA Phase 1/2 trial in Brazil, which is evaluating safety, tolerability, and early efficacy signals in pediatric participants under 12 months of age.
Together, the ASGCT data highlight GEMMABio’s focus on next-generation AAV gene therapy design, including improved tissue targeting, de-immunized payloads, microRNA-regulated expression, and safety-driven vector engineering. The DMD and SMA1 programs reflect the company’s broader strategy to develop genetic medicines that prioritize efficacy, safety, and broader patient access.
