A Comparison Study of Anti-CLL1 CART Cells Equipped with Different Co-Stimulatory Domains in the Treatment of Children with Refractory/Relapsed Acute Myeloid Leukemia

Background: It's well established that conventional chemotherapy doesn't benefit children with refractory/relapsed acute myeloid leukemia (R/R-AML). Though multiple attempts, i.e., novel target agents, and intensive chemotherapy, have been translated into clinics, few have been proved to be successful. Chimeric antigen receptor (CAR) T cell therapy is increasingly becoming a curative candidate for refractory hematological malignancies. Recently, anti-CLL1 CAR T-cells have been reported to effectively treat children with R/R-AML by several groups even though with a small sample size. However, the impact of co-stimulatory domain (CD28/CD27 versus 4-1BB) on the efficacy of CAR T-cells could not be neglected, which is not investigated so far. Thus, we here compare the impact of different costimulatory molecules on the safety and efficacy of anti-CLL1 CAR T-cells in treating children with R/R-AML.

Methods: Seven R/R-AML children treated with anti-CLL1-based CAR T-cells were enrolled into this preliminary comparison study. The baseline characteristics, time of CAR T-cells generation, in vivo persistence of CAR T-cells, cytokine release profiling, cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), secondary hemophagocytic lymphohistiocytosis (sHLH) or macrophage activation syndrome (MAS), and other adverse events, the length of hospitalization, remission status, and the follow-up data were collected for analysis.

Results: Among these seven patients, four received CD28/CD27 based CAR T-cells, and three received 4-1BB based CAR T-cells therapy. The CD28/CD27 group consisted of two boys and two girls, with the median age of 7.5 (range; 5 to 11), while the 4-1BB group consisted of two boys and one girl, with the median age of 12 (range; 8 to 12). Generation of CD28/CD27 equipped CAR T-cells (median, 14 days) were faster than that of 4-1BB ones (median, 21 days). All patients experienced grade 1 to 3 CRS and were well managed with one patient experiencing grade 1-2 ICANS but without any sHLH or MAS. The maximum durations were 156 days for CD28/CD27, ranging from 0 to 156 days and 274 days for 4-1BB, ranging from 16 to 274 days) respect. For hospitalization, CD28/CD27 group experienced subtle longer hospital stays (median, 23 days; range, 14 to 30 days) than 4-1BB group (median, 21 days; range, 18 to 54 days). The overall response rates were 67 % and 75 % in 4-1BB group and CD28/CD27 group respectively. Three of four patients in CD28/CD27 group achieved morphological remission with MRD negativity, while two of three patients in 4-1BB group achieved morphological CR, with one as being MRD negativity. Of all the seven cases, only one patient proceeded to hematopoietic stem cell transplant (HSCT) and died of GVHD seven months after HSCT, however, another four patients who achieved CR could not received HSCT due to poor economic status. Of the remaining six patients, two non-respondent died of disease progression, one patient relapsed and subsequently died, and the other three remained alive and being followed up.

Conclusion: Though the enrollment was too small for us to draw a conclusion, we have noticed a similar efficacy/safety index between 4-1BB and CD28/CD27 equipped anti-CLL1-based CAR T-cells in treating children with R/R-AML. In sum, the relatively high efficacy and low grade and manageable side effects were also validated in this study.