Background

Acute myeloid leukaemia (AML) is a malignancy characterized by abnormal clonal expansion of myeloid blasts in the bone marrow that impairs normal haematopoiesis, causing infection, bleeding, and anaemia [1]. Despite standard treatment with chemotherapeutic agents, including anthracycline and cytarabine, to achieve complete remission (CR) for patients with acute myeloid leukaemia, many patients still relapse within a short period [2, 3]. Only 35–45% of patients under the age of 60 years experience long-term remission with conventional treatment, and the proportion drops to 10–15% for those older than 60 years [2, 4]. Relapse and leukaemia-related complications are the most common causes of death, and only 10% of patients with a first relapse survive long term [4, 5].

In recent years, cellular immunotherapy of chimeric antigen receptor T cells (CAR-T) for relapsed or refractory acute myeloid leukaemia (R/R-AML) targeting myeloid-lineage antigens, such as CD123, CLL-1, and CD33, has shown promising prospects in many preclinical studies [6,7,8,9,48]. Similarly, we demonstrated this finding in a subset of 16 R/R AML patients (5/16, 31.25%) with minimal residual disease (MRD). Similar to previous studies, both R/R AML patients and HDs had high expression of CD7 in their normal cell populations, such as NK cells and T cells. In contrast, B cells, monocytes, and neutrophils did not express CD7 [15, 49]. The lack of CD7 expression in myeloid cells prevents the killing of myeloid cells.

CD7-positive T-cells cultured with CD7 CAR-T cells lead to diminished proliferation and increased cell death, as demonstrated in our study and previous studies [15]. Despite the diminished proliferation, our previous study demonstrated that the required dose of naturally selected CD7 CAR-T cells for transfusion back to patients could be achieved [34].

In addition, similar to our previous report, CD7 CAR-T cells were dominated by CD7-negative subpopulations on Day 12 post transduction possibly due to antigenic masking/intracellular sequestration by CD7 CAR [34]. This finding suggests that the naturally selected CD7 CAR-T cells remained CD7 negative before they were administered back to the patients.

Finally, CD7 was positively expressed on T cells and NK cells, which are important components of the human immune system [49, 50]. Previous studies have shown that CD7 CAR-T cells cause defects in CD7+ T cells in recipients, but CD7− T-cell subsets appear to replace their function to some extent, thereby alleviating treatment-related T-cell immunodeficiency. In addition, CD7+ T-cell populations and NK cell populations were restored after bridging allogeneic haematopoietic stem cell transplantation [30, 34]. Compared with other myeloid-targeted CAR-T cells, CD7 CAR-T cells can avoid damage to normal myeloid cells and reduce haematopoietic toxicity to a certain extent because CD7 is not expressed on other myeloid cells.

Persistence, a key factor affecting CAR-T-cell efficacy, has been widely investigated in previous studies, and it has been shown that the T-cell memory phenotype subpopulation is an important factor in maintaining CAR-T-cell persistence [41, 42]. In the present study, we demonstrated that both CD4+ and CD8+ T-cell subpopulations exhibited a higher percentage of cells with central memory phenotypes in the naturally selected CD7 CAR-T cells compared with the NTR group, which facilitated a more durable effect of CD7 CAR-T cells. In addition, there was no evidence of accelerated terminal differentiation, even in the CD8+ T-cell subpopulation, and the proportion of TEff with naturally selected CD7 CAR-T cells decreased. The above study demonstrated the persistence of naturally selected CD7 CAR T cells in the in vitro phase of the study. However, further validation of its durability in comparison with other reported CD7 CAR-T cells and in clinical trials of R/R-AML needs to be performed [15, 51,52,53,54].

Naturally selected CD7 CAR-T exhaustion marker assays were performed, and the results suggested a trend or statistically significant increase in both PD-1 and TIM-3 expression. In a previous clinical report of T-ALL/LBL, TIM-3 and PD-1 expression levels were not significantly different in patients who achieved CR compared to those who achieved less than CR [34]. Therefore, the impact of increased exhaustion markers on naturally selected CD7 CAR-T cells for R/R-AML needs to be validated by further clinical trials and long-term observations.

In addition, CD7+ cell lines and primary AML blasts from R/R-AML patients in vitro and CD7+AML xenograft models were used to assess the antileukaemic ability of naturally selected CD7 CAR-T cells, demonstrating powerful cytotoxic effects on CD7+ AML cells.

Although our proposed naturally selected CD7 CAR-Ts exhibit diminished proliferation due to cell death at the in vitro culture stage, the CAR-T dose for patient transfusion can still be achieved. In addition, CD7 CAR-Ts may be less costly and exhibit lower risks associated with gene editing. In addition, in recent years, the construction of CAR-T cells by isolating CD7-negative cell populations or the use of ibrutinib and dasatinib to inhibit fratricide has demonstrated good potential value [52, 54]. Hai-** Dai et al. demonstrated the safety and efficacy of CD7 CAR-T cells using protein expression blockers to block CD7 expression at the CAR-T-cell membrane for the treatment of R/R early T-cell precursor lymphoblastic leukaemia/lymphoma (ETP-ALL/LBL) in patients with TP53 mutations [51]. Yongxian Hu et al. investigated allogeneic CD7 CAR-T cells for R/R CD7-positive haematologic malignancies, including one patient with CD7-positive acute myeloid leukaemia and 11 patients with T-cell leukaemia/lymphoma. This study demonstrated the safety and efficacy of allogeneic CD7 CAR-T cells for CD7-positive haematologic malignancies [53]. Both the abovementioned studies and our proposed naturally selected CD7 CAR-T cells have explored the use of CD7 CAR-T cells in the treatment of haematologic malignancies. However, it seems that CAR-T cells targeting CD7 show different characteristics in clinical trials compared with CAR-T cells targeting CD19. Thus, the actual performance of CD7 CAR-T cells in the treatment of R/R-AML patients should be assessed in further clinical trials and long-term clinical observations [55].

In conclusion, naturally selected CD7 CAR-T cells, as a CAR-T treatment strategy without additional treatments, such as CD7 knockdown, can reduce the cost and additional unknown risks associated with gene knockdown to some extent. Patients can benefit from avoiding the risk of additional gene knockouts and reduced production costs of naturally selected CD7 CAR-T cells as a bridging allogeneic HSCT pretreatment for R/R-AML. This study demonstrated the feasibility of naturally selected CD7 CAR-T cells in the treatment of patients with CD7+ R/R-AML in the preclinical study phase. However, R/R-AML is a complex disease type, and the effectiveness of naturally selected CD7 CAR-T therapy needs to be further tested in clinical trials.