On July 29, 2026, the journal Nature appended a note to a paper it had published five months earlier: “concerns have been raised regarding this article, including with the data presented.” The paper, from a team led by neuroscientist Zilong Qiu of Shanghai Jiao Tong University School of Medicine, described a preclinical gene-editing treatment for a rare neurodevelopmental disorder. It did not mention that a six-year-old girl had already been given a version of that treatment, or that she was dead.

She died in late March 2025, seven days after a medical team at Xinhua Hospital in Shanghai infused hundreds of trillions of viral particles carrying a CRISPR base editor into her spinal fluid. A hospital panel attributed the death to a severe immune reaction to the therapy. The reporting that surfaced the case, a joint investigation by Science and Retraction Watch, calls the child Mei. Diagnosed in 2013 with Snijders Blok-Campeau syndrome, tied to a mutation in the CHD3 gene, she was on the milder end of it by the accounts collected there: a girl in academic decline whose parents worried about whether she would ever live on her own. The same reporting describes her as the first person in the world to receive a gene-editing therapy aimed at the brain.

The entire official penalty for that death is a fine. In September 2025, local health authorities fined Xinhua Hospital roughly 24,000 yuan, about $3,500, for inadequate oversight. Qiu faced no public sanction. There the matter sat for the better part of a year, a five-figure line in a local administrative file, until reporters found it.

Set that fine against the money moving the other way. The parents put more than $860,000 toward developing, testing, and dosing the therapy, a sum that, according to the reporting, included informal payments sent directly to individual members of the research team, more than $130,000 to one of them, money the recipient returned only after Mei was dead. The people who dosed the child were, in part, paid by her family. The hospital’s $3,500 penalty is less than half of one percent of what that family had already spent before the infusion that killed her.

Then there is the paper. Published in Nature in February 2026, it presented the work as preclinical and animal-stage. An earlier preprint had thanked families of children with the syndrome for their “participation and support” and carried genetic data from Mei’s own family. The published version deleted the acknowledgment and removed the family. What survived was a single line noting that “bridging the gap between preclinical research and clinical translation remains a significant challenge.” A child had already died crossing that gap. The sentence calling it a challenge still ahead ran eleven months after she was buried, from a team the family had already asked to withdraw the work.

The documents establish a set of facts and stop short of a motive. A child died. A hospital was fined a nominal sum for failures of oversight. Her family paid a large amount, some of it to individual researchers. The published account left out the death, the money, and the withdrawal request. They do not establish that the payment structure caused the concealment, or that any one person set out to bury the result. What they establish is a shape: money moving from patient to investigators, an experimental infusion with no published safety record, a fatal outcome, a $3,500 fine, and a literature scrubbed of the person at its center. The shape is too specific to wave off, and too incomplete to name a single actor beyond what each document supports.

This is not the first time the same system has run past its own oversight and been caught only afterward. He Jiankui announced the birth of gene-edited babies in China in 2018 and was later imprisoned; in the years after, researchers argued in print that the field needed a formal international mechanism to report exactly this kind of unsanctioned human experiment before it reached a patient. No such mechanism caught Mei’s case. Two publications reading paperwork did. Steven Gray, a gene-therapy researcher at UT Southwestern, gave the plainest verdict on the decision to dose her at all: “This shouldn’t have gone to trial.”

So account for who has and has not acted. Nature has posted its note and says it is investigating under its integrity policy; it has not ruled, retracted, or said what it knew and when. Shanghai Jiao Tong University has convened a task force and restated its commitment to research integrity; it has published no findings. China’s national regulators, past the one local fine, have announced nothing. The records that would close the gap are the ones no institution has produced: the trial’s ethics approval and registration file, the consent form the parents signed, and the raw safety data from the dose that killed her. Those records sit with a hospital, a university task force, and a national regulator, none of which has put them on the table. Until one of them does, the fullest official accounting China has offered for the death of a six-year-old in the first gene-editing therapy aimed at a human brain is a fine of $3,500.

Sources

  1. Retraction Watch – A couple paid more than $800,000 for a gene-editing therapy for their daughter. She died, and it wasn’t made public (July 23, 2026)
  2. Science (AAAS) – Exclusive: Death of girl in Chinese gene-editing trial was never made public
  3. Retraction Watch – Journal, university investigations launched into gene-editing death following Science–Retraction Watch reporting (July 30, 2026)
  4. Science (AAAS) – Chinese university launches probe into girl’s gene-editing death
  5. South China Morning Post – Chinese scientist faces probe after girl, 6, dies following experimental gene therapy
  6. Gizmodo – Researcher faces investigation for concealing 6-year-old girl’s death in gene-editing trial
  7. Trends in Biotechnology – International Reporting Mechanism for Unethical Germline Gene Editing Experiments Is Needed (2021)