Preclinical Efficacy vs. Human Response: First CRISPR Gene Therapy Clinical Trial Reveals Pediatric Patient Death and Low Correction Rate
Background
Duchenne Muscular Dystrophy (DMD) is a rare genetic disease where muscle cells undergo necrosis due to a deficiency in dystrophin. Most patients experience gait disturbances from childhood and pass away in their 20s or 30s due to cardiomyopathy or respiratory failure. Current standard treatments, such as corticosteroids or exon-skipping agents, only partially slow disease progression and fail to fundamentally correct the damaged gene.
The dystrophin gene is among the longest in the human genome, making it difficult to fully load into existing viral vectors. Consequently, CRISPR gene-editing technology, which can precisely fix defective regions, has emerged as an alternative. 'HG302', developed by the Chinese biotech firm HuidaGene Therapeutics, was a highly anticipated candidate after showing excellent restoration of dystrophin expression in preclinical mouse models.
Researchers at HuidaGene chose a method to deliver Cas12 protein and guide RNA via Adeno-Associated Virus (AAV) systemically. Based on the safety and efficacy confirmed in animal experiments, they entered an investigator-initiated clinical trial targeting pediatric patients. While the goal was to treat a fatal disease early, the hurdle of human administration concealed unexpected pitfalls.
Key Findings
According to clinical analysis reported in Nature, a pediatric patient in the high-dose HG302 group died due to systemic immune toxicity. Immediately after receiving an intravenous injection of a high concentration of AAV vector per unit body weight, the neonate experienced a rapid onset of high fever and a systemic inflammatory response. This was caused by the body's immune system recognizing the foreign viral capsid as an invader, triggering massive complement activation and cytokine secretion.
The immune overreaction progressed to acute respiratory distress syndrome (ARDS), and despite intensive care, the patient ultimately failed to recover. The innate immune system surge feared upon high-dose AAV administration directly struck the child's respiratory system, leading to fatal outcomes. The aspect that further froze the clinical field was the muscle tissue biopsy analysis indicators of deceased patients and surviving participants.
Contrary to the researchers' expectations, the target DNA editing efficiency in patients' skeletal muscle tissue was extremely low. The expression of functional dystrophin protein within the tissue was also insufficient to expect a therapeutic effect, providing evidence that the extensive muscle regeneration efficacy observed in animal models was not reproduced in humans.
The extreme interspecies gap between mice and humans was clearly demonstrated. Vector concentrations that were effective in small-bodied animal models with simpler immune systems acted as a fatal toxicity barrier rather than providing efficacy in humans. This result confirms the limitations of in vivo delivery efficiency and the risks of excessive viral vector administration, as evidenced by clinical data.
Implications and Outlook
These results serve as a warning about the danger of blindly trusting animal experiment data during the clinical entry process of gene-editing technology. The fact that preclinical success does not guarantee human efficacy and safety has been demonstrated by the sacrifice of a child and low correction rates. The immune toxicity caused by high-dose systemic AAV administration is a top priority that must be re-examined across the entire field of gene therapy.
Regulatory agencies and academia are calling for stronger safety oversight standards for investigator-initiated clinical trials. It is urgent to redefine the maximum tolerated dose of in vivo-administered vectors and establish immune control protocols to suppress complement activation. There is a growing trend toward focusing on the discovery of next-generation capsids with enhanced targeting specificity or lipid nanoparticle (LNP) delivery systems as alternatives to AAV.
Emphasis is being placed on the fundamentals of increasing the precision of platform technologies rather than the speed of drug development. A technological leap is required to deliver CRISPR/Cas9 gene-editing tools only to target tissues while lowering immunogenicity through non-viral delivery technologies. It is time to restore scientific rigor so that technology intended to save patients does not threaten them.
Nature, Published online: 27 October 2026; doi:10.1038/d41586-026-03449-6Therapy pioneered by biotechnology Chinese firm HuidaGene showed promise in animals but caused a fatal reaction in a child.
This clinical result demands immediate changes in the design of future systemic gene therapy clinical trials. Safety standards involving real-time monitoring of complement cascade reactions and inflammatory cytokine levels before and after high-dose AAV administration are expected to be essential in clinical settings. Protocols for the preventive co-administration of complement inhibitors, such as eculizumab, are being discussed as concrete alternatives to mitigate systemic toxicity.
The industry is expected to shift its R&D focus toward non-viral delivery platforms, reducing dependence on viral vectors. Modified capsids that enhance muscle tissue targeting, allowing the dosage to be reduced to one-tenth or less, and lipid nanoparticles (LNPs) that evade immune responses are gaining attention as next-generation delivery technologies. Rather than relying solely on laboratory animal data, the practice of thoroughly verifying in vivo distribution and immune responses using human organoids or primate models is likely to become the new standard for drug development.