Removing CD33 From the Graft First: CRISPR and a Targeted Drug Join Forces Against High-Risk Blood Cancers

Acute myeloid leukemia treatment presents a difficult challenge: a target that identifies cancer cells may also be present on healthy blood-forming cells. Now, an early phase 1/2 human trial has moved CRISPR gene editing to the pre-transplant stage, first removing CD33 from the surface of donor stem cells and then using a CD33-targeting drug to pursue residual cancer cells, in an attempt to uncouple efficacy from bone marrow toxicity.

This multicenter study enrolled 30 adults aged 18 to 70 with high-risk acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). The research team used CRISPR–Cas9 to process hematopoietic stem cells from HLA-matched donors, creating trem-cel, a graft lacking CD33; the trial registration also refers to this investigational product as VOR33. After transplantation, patients received the antibody-drug conjugate gemtuzumab ozogamicin (brand name Mylotarg) as maintenance therapy when appropriate.

The rationale behind the design was to cloak the newly established blood-forming system in a “target-invisibility cloak.” Gemtuzumab ozogamicin recognizes CD33 and delivers a cytotoxic drug into cells, but CD33 is found not only on leukemia cells but also on normal myeloid cells. If donor stem cells lose CD33 in advance, the drug could theoretically continue attacking residual tumors while reducing harm to newly formed blood cells.

The study first addressed whether the grafts could engraft successfully. All 30 participants achieved neutrophil engraftment by day 28, with a median time of 10 days, meeting the trial’s primary endpoint. Nineteen subsequently received gemtuzumab ozogamicin maintenance therapy, and the study observed no persistent high-grade cytopenias. These findings provide preliminary support that the edited stem cells can reconstitute blood formation and withstand subsequent anti-CD33 treatment.

But “able to engraft” is not the same as “able to prolong life.” The trial was small, had no randomized control group, and included a total of three transplant-related deaths. The available data are insufficient to determine whether this strategy can reduce relapse or improve relapse-free survival or overall survival. The study was also terminated early for financial reasons, limiting enrollment and the depth of follow-up, while rare or delayed risks from gene editing have likewise not yet been ruled out.

The significance of this study, therefore, lies not merely in adding another CRISPR therapy, but in using gene editing as protective engineering for targeted treatment: rather than directly modifying cancer cells, it reshapes normal tissue to potentially enable patients to withstand more sustained anticancer pressure. The participating Canadian team said that all trial grafts were manufactured in Montreal and that this was also the first time hematopoietic stem cells modified in a Canadian laboratory were administered to humans.

The next step requires larger studies with longer follow-up to compare relapse rates, survival, graft-versus-host disease, and long-term hematopoietic function, while clarifying manufacturing costs and cross-center quality control. For high-risk AML and MDS, this approach has demonstrated technical feasibility; whether it can change clinical outcomes still depends on efficacy and safety evidence that the early trial has not yet adequately provided.