AINEWS
A 33 patient phase I trial at Children's National trained T cells against three tumor proteins and left four children alive years later, but the trial's design can't say why the other 29 didn't.
Three protein targets, one IV bag, and four children who should not be alive.
A phase I trial at Children's National Hospital in Washington, DC trained a patient's own T cells, the white blood cells the immune system uses to recognize and attack diseased tissue, to recognize three tumor-associated antigens: WT1, PRAME, and Survivin. Each protein is over-expressed in pediatric brain tumors, each is rare enough on healthy tissue that an immune cell trained against it should be able to tell tumor from normal brain, and each is being targeted simultaneously so the immune response has more than one foothold. Of the 33 children and young adults enrolled, three (one with recurrent glioblastoma, one with astroblastoma, one with medulloblastoma) have no evidence of disease two to five years after treatment. A fourth, with DIPG, a diffuse and inoperable brainstem tumor with a historical median survival of 11 months, is alive more than two years out.
The therapy, called a multi-antigen TAA-T product, was infused intravenously rather than injected into the cerebrospinal fluid or directly into the tumor itself, the routes most experimental brain-tumor immunotherapies still use. Children's National's co-senior authors Bollard and Hwang, and an IV bag is the same logistics as a blood transfusion.
The bet held in four cases. It did not hold in the other 29. The trial was a phase I dose-escalation study: the kind designed primarily to test whether a new treatment is safe rather than whether it works, and the kind that runs without a control group, a comparison arm of patients who did not receive the therapy, so the result cannot by itself prove the therapy caused the survival. Patients had already failed up to 17 rounds of chemotherapy and radiation before enrolling, and the responders are now durable and off-therapy. The mechanism for why those four responded, and the other 29 did not, is the open question the Nature Medicine paper leaves for the next trial to answer.
"We don't yet have a control group, so we can't say with certainty that the T cells caused the survival," said Hassall. "What we can say is that the signal in those four children is striking enough to justify a larger, controlled trial that does."
The next controlled trial is the design the Children's National team and outside collaborators have to deliver. They will have to decide what it must test: whether the three-antigen target list is the right one, whether the IV route is durable across more patients, and whether the responders share a biological feature, a specific T-cell receptor, a particular antigen density, a preconditioning regimen, that explains why the therapy landed for them. The current paper, on its own, cannot answer any of those questions. It can only point the next study in a direction.
Wire and aggregator coverage, including the New Scientist's writeup, has framed this as "four children saved." The source is explicit that the language does not fit.
The published cohort includes patients with newly diagnosed DIPG and recurrent central nervous system tumors. The trial, registered on ClinicalTrials.gov as NCT03652545, was small on purpose. So is the cohort of long-term survivors. The next controlled study, and what it chooses to test about the four responders, is the question that will tell the field whether the survival signal is real.