CD33-Deleted Allograft Allows Subsequent Gemtuzumab Ozogamicin in AML

Allogeneic hematopoietic cell transplantation (alloHCT) using tremtelectogene empogeditemcel (trem-cel), a CRISPR–Cas9–edited, CD33-deleted stem cell product, achieved reliable engraftment in patients with acute myeloid leukemia (AML) and allowed subsequent maintenance dosing with the CD33-directed antibody-drug conjugate gemtuzumab ozogamicin (GO; Mylotarg) without the prolonged cytopenias that have historically limited posttransplant use of that agent, according to findings published in Nature Medicine.¹

“We are encouraged by the results of this study showing that a CD33-deleted stem cell transplant looks very similar to the outcomes of standard stem cell transplantation,” John DiPersio, MD, PhD, Virginia E. & Sam J. Golman Professor of Medicine and director of WashU Medicine’s Center for Gene and Cellular Immunotherapy, stated in a news release.²

The phase 1/2 multicenter VBP101 trial (NCT04849910) represents the first clinical report of a gene-edited allograft designed specifically to shield donor hematopoiesis from a targeted posttransplant therapy.

Background and Rationale

Allogeneic HCT remains the standard of care for patients with AML unlikely to achieve cure with chemotherapy alone, yet most patients with high-risk features including adverse cytogenetics, measurable residual disease (MRD), and secondary AML relapse after transplant, and long-term survival remains poor. Post-HCT maintenance strategies have been constrained by hematotoxicity against the engrafted graft, and no FDA-approved therapy exists specifically for this setting.

GO is a CD33-directed antibody–drug conjugate approved for certain AML indications, but its post-HCT use has been sharply limited by severe, prolonged cytopenias arising from on-target destruction of CD33-expressing donor myeloid cells and progenitors. Trem-cel addresses this by using CRISPR–Cas9 to delete CD33 from donor CD34+ hematopoietic stem and progenitor cells before transplant, preventing the graft cells from being targeted by subsequent CD33-directed agents while preserving multilineage hematopoietic function. CD33 was selected as the editing target because its expression is restricted to hematopoietic cells, it is present on the vast majority of AML cases, and prior studies suggest it is not essential for normal hematopoiesis.¹

Trial Design and Patient Population

An initial 59 adult patients with CD33-positive AML or myelodysplastic syndrome (MDS) at high risk of post-HCT relapse across 15 US and Canadian centers were screened, 30 of which had successful manufacturing of trem-cel from an 8/8 human leukocyte antigen–matched donor and were enrolled. Twenty-nine patients had AML and 1 had MDS; 55% carried adverse cytogenetic risk by European LeukemiaNet 2022 criteria, 30% had TP53 mutations, and 38% had secondary AML. The median age was 53 years (range, 22–68).

Patients received myeloablative conditioning followed by trem-cel infusion on day 0; no graft-vs-host disease (GVHD) prophylaxis was administered given the CD34-selected graft composition. Beginning approximately day 60 post-HCT, eligible patients received maintenance GO in a 3+3 dose-escalation design across cohorts of 0.5, 1.0, and 2.0 mg/m² per 28-day cycle for up to 8 cycles.

Key Results

The primary end point of neutrophil engraftment by day 28 was met with successful engraftment in all patients with a median time to engraftment of 10 days, comparable to historical data from unedited CD34-selected grafts. Platelet recovery was achieved in 29 of 30 patients, with a median of 16 days. Full donor myeloid chimerism was established by day 28 or 60 in all patients. Median CD33 gene editing efficiency at manufacture was 90% (range, 71%–94%), and CD33 negativity in peripheral blood myeloid cells was sustained at close to 100% through subsequent GO cycles.

Nineteen patients received at least 1 cycle of maintenance GO. No dose-limiting toxicities were observed across any dose cohort, and the recommended phase 2 dose was established at 2 mg/m². Critically, neutrophil and platelet counts were preserved across all GO dose levels without prolonged high-grade cytopenia, a marked departure from the hematologic toxicity profile reported with GO following standard alloHCT. GO pharmacokinetics in the CD33-depleted environment showed a proportionally larger increase in area under the curve relative to maximum concentration compared with standard relapsed/refractory AML populations.

Median relapse-free survival and overall survival for all 30 patients were both 14.1 months, with an approximately 2-year relapse-free survival rate of 40%. The authors characterized these figures as encouraging given the high-risk, heterogeneous population enrolled, including patients with active disease at transplant. Seven patients died during the study: 4 from disease progression and 3 from transplant-related causes consisting of renal failure, sinusoidal obstruction syndrome/veno-occlusive disease related to GO, and sepsis, all after day 100.

Acute GVHD of grade 2 to 4 occurred in 20% of patients and grade 3 to 4 in 7%, which the investigators stated to be consistent with those reported for standard unedited CD34-selected grafts in patients who did not receive prophylaxis. All cases of disease relapse retained CD33 expression by flow cytometry, indicating that antigen escape was not the primary resistance mechanism.

Study Termination and Next Steps

The trial was terminated early for fiscal reasons, making the published report, primarily covering the completed phase 1 portion, the final analysis from the trial.

The early termination of the study limits conclusions about efficacy and precludes definitive statements about the contribution of GO maintenance to the observed survival outcomes. Nonetheless, the trial establishes proof of principle that CRISPR-mediated CD33 deletion in donor hematopoietic cells can transform a shared myeloid antigen into an AML-specific target amenable to post-transplant immunotherapy without compromising engraftment kinetics, immune reconstitution, or baseline hematologic recovery.

The authors noted that GO, currently the only approved anti-CD33 agent, may ultimately be superseded by CD33-directed chimeric antigen receptor (CAR) T cells or bispecific T-cell engagers with greater antileukemic potency and fewer non-hematologic toxicities such as hepatotoxicity. Patients with disease relapse during the study were eligible for VBP301, a CD33 CAR T trial (NCT05984199) using donor-derived T cells.

In a case study published in JCO Precision Oncology in October 2025, a 65-year-old male patient treated in the VBP101 trial proceeded to the VBP301 trial and received the donor-derived CAR T product VCAR33 after AML progression was determined on day 104 from trem-cel treatment.³ After treatment with VCAR33 and 2 incidences of grade 1 cytokine release syndrome, investigators observed pancytopenia and aplasia with no detectable AML on day 16 after receiving CAR T. With continued follow-up up to day 147 after CAR T, he achieved complete response with incomplete count recovery, needing continued transfusions and growth factor support for thrombocytopenia. His peripheral blood neutrophils did not express CD33, indicating maintained trem-cel hematopoiesis.

The investigators discussed the unexpected challenge of the aplasia but proposed that the combination of trem-cel with VCAR33 can keep CAR T cells from removing normal myeloid cells “and may increase efficacy by expanding the therapeutic window of VCAR33 and by preserving myeloid cells that could support CAR T activity.”

Longer follow-up and adequately powered randomized studies will be required to determine whether trem-cel combined with optimized CD33-directed post-HCT therapies can meaningfully reduce relapse rates in this historically difficult-to-treat population.

“In the future, we are hopeful we will be able to combine this with CD33-targeted immunotherapies, such as CAR T cells, and improve treatment options for patients with these very aggressive blood cancers,” DiPersio stated in the news release.²

REFERENCES

1. DiPersio JF, Koehne G, Shah NN, et al. CRISPR–Cas9 CD33-deleted allogeneic hematopoietic cell transplantation with gemtuzumab ozogamicin maintenance in AML: a phase 1/2 trial. Nat Med. Published online May 12, 2026. doi:10.1038/s41591-026-04362-1

2. Strait, JE. Gene-edited stem cell transplant shows promise for aggressive blood cancers. News release. Siteman Cancer Center. May 12, 2026. Accessed May 12, 2026. https://tinyurl.com/3jn2fndp

3. Koehne G, Mushtaq MU, Muffly LS, et al. Remission of TP53-mutant AML after transplantation with trem-cel, a CRISPR/Cas9 gene-edited allograft lacking CD33, followed by a donor-derived anti-CD33 chimeric antigen receptor T (VCAR33). JCO Precis Oncol. 2025;9:e25000556. doi:10.1200/PO-25-00556