A 6-year-old girl with CHD3 mutation (Snijders Blok-Campeau syndrome) died after receiving single-patient base editing therapy.
Treatment involved dual AAV9 vectors injected intrathecally; death occurred within 7 days from severe immune reaction.
Preclinical studies in nonhuman primates showed moderate to severe liver and kidney toxicity, but dosing was not adjusted.
The trial was not publicly registered; death was not reported in clinical trial databases; only a fine was imposed.
International experts have condemned the lack of safety oversight and called for regulatory reform.
What happened
In late March 2025, a 6-year-old girl with a rare genetic disorder known as Snijders Blok-Campeau syndrome (caused by a mutation in the CHD3 gene) received an experimental, single-patient base editing therapy at a hospital in Shanghai, China. The treatment, developed by a team from the Shanghai Jiao Tong University Brain Science Research Center, was funded by the child's parents at a cost of approximately $860,000. Seven days after intrathecal injection of trillions of dual AAV9 vectors carrying the base editor components, the girl died from a severe immune response triggered by the therapy. The case was never publicly reported in clinical trial registries, and only an administrative fine was levied against the hospital. The story was first disclosed in an exclusive investigation by Science and Retraction Watch in early 2026.
Background: Snijders Blok-Campeau syndrome and base editing
Snijders Blok-Campeau syndrome is a rare neurodevelopmental disorder caused by heterozygous dominant mutations in the CHD3 gene. It is characterized by intellectual disability, speech delay, motor difficulties, and characteristic facial features. There is currently no approved therapy that corrects the underlying genetic cause. Base editing is a precise gene-editing technique that can convert one DNA base into another without creating double-strand breaks. In this case, a personalized adenine base editor was designed to correct the specific CHD3 mutation. The approach had shown promise in animal models, but questions about its safety and efficacy in humans remained unanswered.
Neuroscientist Zilong Qiu, speaking here in 2015 at a science fiction convention, was confident in the safety of a gene-editing treatment he developed, but it led to a death that was not made public.LI YIBO/XINHUA/ALAMY
How the treatment was conducted
The research team delivered the base editor using a dual-vector adeno-associated virus serotype 9 (AAV9) system injected directly into the patient's cerebrospinal fluid (intrathecal route). The procedure was performed on a named-patient basis, outside of a formal clinical trial registration. According to the report, the parents self-funded the entire cost. Preclinical studies included experiments in nonhuman primates, but experts later noted that those studies had revealed moderate to severe liver and kidney toxicity in some animals. Nevertheless, the team proceeded with human dosing at the original level without first establishing a maximum tolerated dose.
Fatal outcome and lack of disclosure
Within a week of the injection, the child developed a severe immune reaction, which the source material describes as likely an innate immune response to the high-dose AAV9 vectors or to the base editing components. The girl died despite medical intervention. After the death, the hospital paid a minor fine to local health authorities, but no public disclosure was made. The investigators did not face formal sanctions, and the death was not recorded in ClinicalTrials.gov or the Chinese clinical trial registry. In early 2026, the same team published a proof-of-concept study in Nature describing the animal model work, but the paper was edited to remove any mention of the patient or family funding.
Expert reactions to the trial's design
Multiple international experts in genetics, bioethics, and gene therapy have expressed strong criticism. Dr. Steven Gray, a gene therapy specialist at UT Southwestern Medical Center, stated that this trial should never have entered human testing. Dr. David Sanders, a biochemist at Purdue University, questioned the reliability of the primate data and the editing efficiency of the dual-vector system in the brain. Dr. James Wilson, formerly of the University of Pennsylvania and a leading gene therapy researcher, noted that the animal toxicity data already showed moderate to severe liver and kidney damage in monkeys, which should have prompted dose de-escalation studies before human exposure. Bioethicists Hank Greely (Stanford) and Joy Zhang (University of Kent) highlighted the ethical failure, with Zhang pointing to a gap between regulatory rules and real-world enforcement. The family has demanded a full investigation and systemic reforms to prevent similar tragedies.
Regulatory and ethical implications
The case exposes critical weaknesses in the oversight of highly experimental, single-patient (n-of-1) gene therapy studies. In China, as in many countries, there are pathways for compassionate use or individual patient access to investigational therapies, but these pathways still require ethics committee approval, informed consent, and safety reporting. According to the investigation, none of these safeguards were properly followed. The child's death was never reported to any registry, and the team continued research activities without interruption. Experts have called for mandatory registration of all interventional procedures, independent safety monitoring boards for first-in-human gene editing studies, and transparent reporting of serious adverse events regardless of funding source.
MedXY commentary on AAV-associated TMA
MedXY has noted that thrombotic microangiopathy (TMA) is a recognized complication of AAV gene therapy. TMA resembles atypical hemolytic uremic syndrome (aHUS), presenting with acute thrombocytopenia, microangiopathic hemolytic anemia, and acute kidney injury, usually within two weeks of treatment. Although the source material does not confirm TMA as the cause of death in this case, the temporal pattern and the known risk of AAV9-associated immune responses underscore the need for meticulous dose titration and monitoring. The very high vector dose administered in this trial may have greatly increased the risk of a life-threatening immune reaction and TMA, a risk that should have been anticipated from the available animal data.
Conclusion
The death of a 6-year-old child in an unregistered, single-patient gene editing trial highlights the dangers of bypassing established safety steps. The case has prompted renewed calls for stricter oversight of personalized gene therapies, particularly when they are funded by families and performed outside formal clinical trials. While the hope of curing a devastating disease is powerful, the mechanisms that protect human subjects must not be set aside. The international community—including regulators, researchers, and funders—must work to ensure that such a tragedy is not repeated.
References
Cohen J, et al. Exclusive: Death of a girl in a Chinese gene-editing trial was never made public. Science. 2026. Available at: https://www.science.org/content/article/exclusive-death-girl-chinese-gene-editing-trial-was-never-made-public
MedXY. Available at: https://medxy.ai/share/5420f6d613f81d652cb14d72126af2fc
