To the Editor,

Acute myeloid leukemia (AML) patients who experience relapse after allogeneic hematopoietic stem cell transplantation (allo-HSCT) have a dismal prognosis, with a one-year overall survival (OS) rate of only 20% [1, 2]. Recently, chimeric antigen receptor T (CAR-T) cell therapy has emerged as one of the most promising therapeutic approaches for those with relapsed/refractory B-cell acute lymphoblastic leukemia (r/r B-ALL) [3, 4]. Although numerous tumor antigens, such as CD33, CD123 and CLL1, have been explored as potential target antigens for AML treatment in the past few decades [5,6,7], CAR-T cell therapy in AML remains challenging due to the lack of an ideal specific antigen target and the risk of fatal “off-tumor, on-target” side effects [8].

CD38 is known to be expressed on most AML blast cells or plasma cells in multiple myeloma but not on healthy hematopoietic stem cells (HSCs) [9]. Studies of daratumumab monotherapy targeting CD38 report encouraging clinical efficacy, favorable safety profiles and mild hematologic toxicities in multiple myeloma patients [10]. To investigate whether a similar immunotherapeutic approach can be applied for the treatment of AML, we selected the same CD38 epitope (CD38 scFv) as daratumumab for our CD38-CAR-T cells (CAR-T-38) (Fig. 1a). Theoretically, CAR-T-38 cells should exhibit anticancer effects and safety profiles similar to those of daratumumab; moreover, CAR-T-38 cells possess innovative “live-drug” functionality. In this prospective study (NCT04351022), we evaluated the clinical therapeutic efficacy and safety of CAR-T-38 cell therapy in six patients with relapsed AML post-allo-HSCT. None of the cases responded to multiple lines of salvage treatments, and all had CD38-positive blasts in their bone marrow at the time of relapse. The patients’ clinical characteristics are summarized in Table 1. For the initial tumor-reduction chemotherapy (Fig. 1b), only patient 1 achieved blast reduction from 15.5% to 5%, whereas the remaining five showed no response. All patients were pretreated with fludarabine and cyclophosphamide (FC) regimens prior to CAR-T-38 immunotherapy. CAR-T-38 cells (4 from autologous and 2 from donors) at a median dose of 8.05 (6.1–10) × 10 [6] cells per kilogram of body weight were administered after the FC regimens. The median marrow donor chimerism recovered from 46.3% (26.6–66.4%) to 92.5% (58.9–98.1%) at one week and to 97.5% (13.5–98.4%) at two weeks post-CAR-T-38 infusion (Fig. 1c). Moreover, at four weeks after CAR-T-38 infusion, 66.7% of patients (4/6; patients 1, 2, 4 and 5) achieved complete remission (CR) (including 1 with CR and 3 with CR with incomplete count recovery (CRi)) and full donor chimerism. In contrast, patient 6 exhibited a 44% reduction in bone marrow blasts; patient 3 showed no response (NR). Interestingly, patient 4 experienced relapse 117 days after the first CAR-T-38 infusion but did achieve remission after a second CAR-T-38 treatment (Fig. 1d). OS was calculated from the date of CAR-T-38 infusion to the date of death. Leukemia-free survival (LFS) was calculated from the date of CR or CRi post-CAR-T-38 infusion to the date of relapse or death or the last follow-up. The 6-month OS and LFS rates were both 50%, and the median OS and LFS were 7.9 and 6.4 months, respectively (Fig. 1d). Multiparameter flow cytometry (FCM) revealed that CD38-positive blasts were effectively eliminated after CAR-T-38 infusion (Fig. 1e). However, CD38-positive granulocytes and monocytes increased gradually after CAR-T-38 infusion (Fig. 1f, g).

Fig. 1
figure 1

CAR-T-38 therapy regimen and treatment response among 6 relapsed AML patients post-allo-HSCT. a Schematic of the CAR-T-38 structure. b Schematic of the CAR-T-38 therapy regimen (* and ** supplementary file). c Variation in donor chimerism in the bone marrow (BM) and peripheral blood (PB) was measured by short tandem repeat-polymerase chain reaction (STR-PCR). d Patients 1, 2, 4, and 5 achieved complete remission (CR) or CR with incomplete count recovery (CRi) at 4 weeks after CAR-T-38 infusion. Patients 1 and 4 relapsed within six months. Patient 4 experienced relapse 117 days after the first CAR-T-38 infusion but exhibited remission after the second CAR-T-38 treatment. The 6-month OS and LFS rates were both 50%. CD38-positive blasts were reduced, and CD38-positive granulocytes and monocytes gradually increased at 1 and 4 weeks after CAR-T-38 infusion. e The CD38-positive blast population is shown in the upper right (red) of the CD34/CD38 plot. f, g CD38-positive granulocytes and monocytes, based on forward versus side scatter and CD38 expression, are shown in the upper left (crimson) and the lower left (green), respectively

Full size image
Table 1 Clinical characteristics of patients
Full size table

Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are well-known major adverse events that limit the clinical application of CAR-T cell therapy. In our study, five patients presented mild CRS (Grade I–II), and only one experienced grade III hepatotoxicity with elevated serum transaminase and bilirubin levels according to the Common Terminology Criteria for Adverse Events (CTCAE) v4.0. All of these events were transient and clinically manageable. All patients had pancytopenia before CAR-T-38 infusion, and neutropenia appeared to persist during CAR-T-38 therapy. The median duration of neutropenia (absolute neutrophil count (ANC) of < 500/μl) was 22 (7–35) days. The median duration of platelet recovery was 25 (18–123) days. No patients manifested neurological toxicities or experienced severe infections.

Although the sample size was small and the follow-up time was limited, our preliminary prospective study highlights that the clinical utility and safety of CAR-T-38 therapy for treating AML patients with relapse post-allo-HSCT.