Gene-edited stem cell transplant lets doctors safely attack leftover leukemia

Patients with high-risk acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS), types of blood cancer, are often treated with a stem cell transplant from a donor as their best chance at a cure. Even so, most of these patients relapse afterward, and long-term survival remains poor. Doctors have tried giving follow-up drugs after transplant to catch any leftover cancer cells, but a major obstacle has stood in the way: the drugs that target leukemia cells often also attack the newly transplanted, healthy blood-forming cells, because both carry the same marker protein on their surface.

A phase 1/2 clinical trial run by researchers at Washington University School of Medicine in St. Louis, together with 14 other sites in the U.S. and Canada, tested a way around that problem. Instead of leaving the donor's stem cells unchanged, researchers used CRISPR-Cas9 gene editing to delete a protein called CD33 from those cells before transplanting them into 30 adult patients with AML or MDS at high risk of relapse. CD33 is present on most AML cells and on normal blood-forming cells, but not on other body tissues. People who are naturally born without it have no apparent health problems, which is part of why researchers picked it as a target. The edited product is called tremtelectogene empogeditemcel, or trem-cel for short, and was manufactured by Vor Biopharma, which funded the study.

The idea is that once the transplanted, CD33-free cells take over blood production, doctors can safely give a CD33-targeting drug afterward. Any remaining cells in the body that still carry CD33 should mostly be cancer, since the new donor blood supply lacks the target. The results were published in Nature Medicine on May 12.

All 30 patients who received the CD33-deleted stem cells achieved neutrophil engraftment, meaning their new donor cells began producing infection-fighting white blood cells, by day 28 after transplant. That was the trial's primary safety goal, and every patient met it, with a median time to engraftment of 10 days. Platelet recovery, which restores the blood's ability to clot, occurred in 29 of the 30 patients, at a median of 16 days. One patient developed an immune reaction against platelets and needed additional treatment before recovering.

Gene editing efficiency at the time of manufacturing was high: a median of 90% of the CD34+ donor cells (the type of cell that gives rise to all blood cell types) had the CD33 gene successfully disrupted. After transplant, doctors tracked how much of the patient's blood was coming from donor cells versus their own, a measure called chimerism, alongside how well the CD33 edit persisted across different blood cell types over time. Nineteen patients went on to receive at least one cycle of the follow-up drug, gemtuzumab ozogamicin, which targets CD33 and carries an attached cancer-killing agent. It is already approved by U.S. regulators to treat CD33-positive AML.

Normally, gemtuzumab ozogamicin causes severe and prolonged drops in blood cell counts, because it damages the same donor stem cells it is meant to spare, on top of killing cancer cells. In this trial, the drug was given in escalating doses, from 0.5 mg per square meter of body surface up to 2 mg per square meter, once every 28 days for up to eight cycles. No dose-limiting toxicities occurred during the escalation, and researchers established 2 mg per square meter as the recommended dose to carry forward. Two patients had their dose lowered because of low platelet counts, but overall, patients maintained blood cell counts across all doses tested — the outcome the CD33 deletion was designed to produce.

Seven patients died during the study: four from their cancer progressing, and three from transplant-related causes, including kidney failure, liver toxicity, and sepsis, all occurring after day 100 post-transplant. Side effects that did occur, such as anemia, low platelets, fever, infections, and graft-versus-host disease (where donor immune cells attack the patient's own tissues), were similar in kind to those seen with standard transplants. Median survival and relapse-free survival were both about 14 months across all 30 patients who received the edited cells, based on a median follow-up of roughly 8 months among patients who were still alive. All relapses observed still carried CD33 on the cancer cells, meaning the target for follow-up therapy remained present in the disease that returned.

The trial was stopped early for financial reasons, so this report serves as the final analysis and focuses mainly on the completed first phase of dose escalation. Only four patients were treated in the expanded second phase at the recommended dose, and five more patients had received the transplant but had not yet reached the point of starting the follow-up drug when data was collected. That limits how confidently the approach's benefits can be generalized beyond the patients described here.

Researchers see this work as a first step toward pairing CD33-deleted transplants with more powerful CD33-targeted immune therapies, such as CAR-T cells, which use a patient's or donor's own immune cells re-engineered to hunt a specific target. A separate case report, published in October 2025, described one patient with an aggressive form of AML who received a CD33-deleted transplant and later, after relapsing, was given CD33-targeted CAR-T cells made from the same donor; that patient reached complete remission and remained cancer-free more than a year later, with normal blood cell production restored entirely from CD33-lacking donor cells. That is a single case, not a controlled trial result, and the researchers describe the current findings as groundwork for future combination approaches rather than a finished treatment.