Trem-Cel, a CRISPR/Cas9 Gene-Edited Allograft Lacking CD33, Shows Rapid Primary Engraftment with CD33-Negative Hematopoiesis in Patients with High-Risk Acute Myeloid Leukemia (AML) and Avoids Hematopoietic Toxicity during Gemtuzumab Ozogamicin (GO) Maintenance Post-Hematopoietic Cell Transplant (HCT).

Background Maintenance therapy intended to control post-alloHCT relapse is often limited by hematopoietic toxicity towards normal engrafted cells. GO (MylotargTM) is an antibody-drug conjugate targeting CD33+ AML, but its use is limited by on-target, off-tumor hematopoietic toxicity. Trem-cel (formerly VOR33) is manufactured from CD34+ cells isolated from matched-donor apheresis and modified by CRISPR/Cas9 gene-editing to remove CD33, with the goal of protecting normal hematopoietic cells from post-HCT CD33-directed therapies. Methods VBP101(NCT04849910) is a first-in-human Phase 1/2 open-label multicenter trial to establish the safety of trem-cel for CD33+ AML patients who are at high risk of relapse. Eligible patients (18-70 y) must have CD33+ AML with high-risk features for relapse and a 10/10 HLA-matched related or unrelated donor. Donors undergo mobilization with G-CSF and plerixafor prior to apheresis and manufacturing of trem-cel. Prior to HCT with trem-cel, patients receive myeloablative conditioning. After 60 days post-HCT, patients may begin maintenance GO in a 3+3 dose escalation strategy starting at 0.5 mg/m2 every 28d for 4-8 cycles. Results We report on 6 initial patients who received trem-cel at a median dose of 5.2 × 106 CD34+ cells/kg (2.6 - 7.6) and CD33 editing efficiency of 88% (80 – 91). Primary neutrophil engraftment occurred in all patients at a median of 10 days (9 – 11), and platelet recovery occurred at a median of 16 days (15-22), excluding one patient with documented anti-platelet antibody (Fig 1). At the day 28 assessment, full peripheral blood myeloid chimerism was achieved in all patients, and CD33 expression was absent in a median of 94% of neutrophils (86–99) and 92% of monocytes (82 - 94). GO dosing is ongoing at 0.5 mg/m2 dose, and neutrophil and platelet counts were minimally affected, suggesting CD33-edited donor cells were protected from GO-mediated hematotoxicity. Following the 0.5 mg/ m2 dose, exposures of GO were predictively higher in the context of CD33-negative hematopoiesis and corresponded to doses used in relapsed/refractory AML patients. After GO dosing, the percent of CD33-negative cells increased, consistent with elimination of unedited residual CD33+ donor cells. Conclusions Initial results demonstrate trem-cel has rapid primary neutrophil engraftment similar to non-edited CD34-selected grafts (Luznik et al 2022, JCO 40:356-368) and a high CD33 editing efficiency leading to a majority of myeloid cells lacking CD33 expression and supporting the biologic dispensability of CD33. The exposure at 0.5mg/m2 GO dose corresponded to higher doses of GO in AML patients, possibly due to reduction of the CD33 hematopoietic antigen sink. Minimal cytopenias have been observed thus far after treatment with GO, supporting the hypothesis that CD33 deletion can protect donor cells from CD33-targeted therapies.