A 6-year-old died after a first-of-its-kind gene therapy. No one had to tell the public.

A 6-year-old with a nonlethal neurodevelopmental condition was given a first-of-its-kind gene therapy inside a Shanghai teaching hospital in March 2025. She died within a week. No one was required to tell the public.

The girl, identified by her parents as "Mei" in a joint Retraction Watch and Science investigation published July 23, 2026, had been diagnosed in 2023 with Snijders Blok-Campeau syndrome, a rare neurodevelopmental condition caused by mutations in the CHD3 gene. Children with the syndrome can have intellectual disability, speech delay, and motor difficulties. With speech, occupational, and special-education support, most have a normal life expectancy.

Her parents, who spoke to investigators under the pseudonyms "Jason" and "Linda," said they hoped an experimental therapy could give their daughter more independence. The procedure, conducted at Xinhua Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, was an in vivo base-editing treatment: a delivery vehicle, in this case adeno-associated virus (AAV) vectors, carries a gene-editing tool into the patient's cells to rewrite a single letter of DNA. According to the joint reporting, it was the first time this particular base-editing approach had been used in a child.

A Live Science summary of the investigation noted that lead investigator Zilong Qiu, a neuroscientist who runs a brain-research center at Shanghai Jiao Tong, was contacted for comment. The published account does not record a response.

Within roughly a week of receiving the infusion, Mei was dead. The death was not publicly disclosed by the scientists, the hospital, or any trial sponsor. Sixteen months later, when the joint investigation ran, the case was still not on the public record.

The mechanism that allowed the silence is not obscure. It is the system working as designed. Under China's drug-trial adverse-event reporting rules, serious and fatal adverse events during an investigational therapy must be reported to regulators and to the trial's ethics committee, but no rule requires public disclosure of a single-patient death. The clinical trial registry entry for the program, NCT06860672, lists no public update matching the death. The result is a gap that is procedural, not conspiratorial: a fatal outcome in a first-in-child gene therapy trial can pass through the existing reporting chain without ever reaching a family considering the same therapy, a journalist, or a regulator outside the trial itself.

The preclinical basis for the therapy was published in Nature: "In vivo base editing of CHD3 rescues behavioural abnormalities in mice." The title is precise. The work was done in mice. The leap to a 6-year-old with a nonlethal syndrome is, by any standard in the field, a wide one. Companion coverage in the Straits Times and Times Higher Education noted that the family paid more than $860,000 for the experimental treatment, financed privately, outside any insurance or public coverage.

The Science editorial accompanying the investigation, headlined "Four takeaways from our investigation into a hidden gene-editing death," names the underlying fault line directly: a nonlethal condition, a first-in-child intervention, and a death that no existing rule required anyone to make public. China is not the only jurisdiction with a disclosure gap of this shape. But China is the country in which this particular child was given this particular therapy under this particular set of rules, and the parents are the ones who decided, more than a year later, to stop waiting for someone else to say it.

The trial registry entry for NCT06860672 still lists the program as enrolling. The Nature paper has not been retracted. The hospital has not published a public statement on the death. The parents' account, told to a science magazine and a research-integrity newsroom, is the public record that exists.