Gene Therapy · asia
Low-Dose Promise Collides With AAV Immune Risk: One Death Reported in First-in-Human DMD Gene-Editing Trial
A participant in the HG302 high-dose cohort developed a severe complement and cytokine response after systemic administration and ultimately died of acute respiratory distress; the company did not disclose the event until nearly a year later, bringing dose escalation, regulation, and information transparency into focus.
A gene-editing therapy for Duchenne muscular dystrophy (DMD), originally intended to achieve durable efficacy with a lower viral dose, has again encountered the most challenging safety boundary in systemic AAV delivery. HuidaGene Therapeutics confirmed that a participant in the high-dose cohort of the first-in-human HG302 trial died after receiving treatment in August 2025; the company did not disclose the event until August 2026.
According to the company’s investigation, the participant experienced severe complement and cytokine activation following systemic administration of a high dose of an adeno-associated virus (AAV) vector and developed acute respiratory distress syndrome. The company said it integrated laboratory, immunological, pathological, and autopsy analyses and submitted the complete findings for peer review in January 2026. Until the paper is formally published, the event timeline, dose, treatment interventions, and causality assessment cannot be independently examined.
HG302 packages the high-fidelity CRISPR-hfCas12Max system into a single AAV vector and targets the splice donor site of exon 51 of the DMD gene, seeking to restore dystrophin expression through exon skipping. When HuidaGene launched the trial, it said the system could achieve editing at a dose lower than those used by existing AAV therapies for muscle diseases, potentially reducing immune-related risks. This death indicates that, as the dose was increased, this premise had not yet been supported by human data.
The MUSCLE study, registered as NCT06594094, is an open-label, early Phase 1 dose-escalation study conducted in Shanghai that enrolled four ambulatory boys with DMD between the ages of 4 and 8. The company said the participant who died was the final participant under the planned trial protocol. The other three did not develop the same severe clinical syndrome and remain under long-term follow-up. However, such a small sample cannot establish the incidence of the reaction and is insufficient to rule out delayed risks at lower doses.
How the trial progressed to the high dose has become another key question. An investigative report by STAT indicated that efficacy signals in the first two participants were inconclusive, yet the research team subsequently proceeded to a higher-dose cohort. Publicly available information is currently insufficient to assess the basis for the escalation decision, including the safety data accumulated at the time, the extent of dystrophin restoration, and whether complement or cytokine monitoring was adequate to provide an early warning.
The delay in disclosure has also deepened concerns. STAT reported that the company issued no press release for as long as 15 months, during which its chief executive officer and chief technology officer departed. The study was also initially conducted through an investigator-initiated trial pathway that did not require prior review by China’s central drug regulator. HuidaGene, meanwhile, said the hospital submitted initial and follow-up reports on the event to ethics and research oversight procedures within the applicable time limits. The two accounts are not mutually exclusive, but they underscore that reporting to an institution’s internal oversight bodies is not the same as promptly providing an explanation to the participant community and the public.
The HG302 event does not yet demonstrate that all DMD gene-editing or AAV therapies carry the same risk. Whether the fatal reaction was driven primarily by the vector dose, pre-existing immune status, tissue damage, or a combination of factors also remains to be clarified by the full paper. But for a one-time, irreversible therapy delivered broadly throughout the body, subsequent evaluation cannot ask only whether editing can be achieved successfully. It must also determine how narrow the safe dose window is, which biomarkers can provide an early warning, and how quickly data sufficient for external scrutiny should be disclosed after a serious event.