CD7-targeted universal CAR T-cell therapy in relapsed or refractory (R/R) acute myeloid leukemia (AML): Clinical results from CTD401/402 studies

Abstract

Abstract Introduction: CAR-T cell therapy has transformed lymphoid malignancy treatment but remains challenging in AML due to lack of safe targets and antigen heterogeneity, and targeting pan-myeloid antigens (e.g., CD123, CD33, CLL-1) has shown early dose-limiting toxicities (DLTs) and modest efficacy. CD7 is aberrantly expressed on blasts/progenitors in ~30% of AML patients but absent on normal myeloid/erythroid lineages, representing a potential target. We report clinical outcomes from ten CD7⁺ R/R AML patients treated with two allogeneic CD7-targeted CAR T-cell products: CTD401 (n=2) and CTD402 (n=8), the latter representing an optimized construct derived from CTD401. Both products were originally designed for T-cell malignancies and later explored in AML. (NCT04538599, NCT05902845, NCT05895994, NCT05454241) Methods: In these investigator-initiated trials, eligible CD7+ R/R AML patients received lymphodepletion with fludarabine (25-30mg/m2) and cyclophosphamide (400-850mg/m2) for 3-5 days, with/without etoposide (100mg/m2) for 3 or 5 days. CD7 expression was assessed by a methodologically validated flow cytometry assay. Patients received a single infusion of CAR+ T cells at doses of 2×10⁸ (n=3), 4×10⁸ (n=3) or 0.9-1×10⁹ (n=4) CAR+ T cells. Results: As of July 1, 2025, ten R/R AML patients (three under compassionate use) received CTD401/402. Median age was 34.5 years (range, 3.5-67); 20% (2/10) were female. Median prior therapy lines: 3 (range, 1-7); two had prior hematopoietic stem cell transplantation (HSCT). Median blast was 39.25% (range, 0.35%-83.15%). All patients experienced mild CRS (G1, n=9; G2, n=1). One developed Grade 2 ICANS after CTD402 redosing due to absent expansion following initial infusion. Immune-related events including CRS and ICANS were manageable. The most common Grade ≥3 treatment-emergent adverse events were cytopenias: thrombocytopenia (40%), neutropenia (30%), and leukopenia (30%). No DLTs or treatment-related deaths occurred. Among 10 evaluable patients, 8 exhibited high CD7 expression (≥80%) on baseline blasts, while 2 had partial (<80%). Best overall response (CR/CRi) rate was 50.0% (5/10), increasing to 62.5% (5/8) among high CD7 expression patients. All responders achieved MRD negativity. In addition, among all responders, 2 patients received HSCT, and 3 patients received hematopoietic stem cell boost (HCB) after achieving CR/CRi, with no preconditioning as we have previously published (Hu et al., NEJM 2024), with a median interval of 31 days (range, 14–93). Median duration of response was 10.1 months (range, 0.7-53.6). Of 5 non-responders, 4 cases with baseline CD7 expression ranging from 38% to ≥90% showed only CD7-negative blasts remaining post-infusion, suggesting selective expansion of CD7-negative clones or potential downregulation/loss of CD7 expression in resistant cases. CAR-T expansion was robust across doses, with median Cmax of 2,835,796 copies/μg DNA (range, 14,845-7,359,551) and median Tmax of 8.5 days (range, 4-21). Median persistence is 31 days (range, 7-640+) with the longest exceeding 21 months and durable CR maintained. Conclusion: CTD401/402 were well tolerated up to 1×10⁹ CAR⁺ T cells with no DLTs. Among patients with high CD7 expression, best overall response rate reached 62.5%, supporting CD7 as a viable target in a molecularly defined AML subset. Subsequent HSCT/HCB may be crucial for sustaining remission after CAR T-cell therapy, and in selected AML patients, CD7-directed CAR T cells may allow foregoing/reduction of myeloablation, enabling potentially curative transplantation with reduced toxicity as we have previously published (NEJM 2024).