Background: Autologous T cells modified to express a chimeric antigen receptor (CAR-T) has demonstrated exciting efficacy in treating leukemia and there has been some reports about the toxicities recently. However, the spectrum of capillary-leak syndrome (CLS) associated with CAR-T cell therapy has not been systematically evaluated, which can be a life threatening complication as results of the cytokine release syndrome (CRS). Therefore, as the use of CAR-T therapy continues to expand to broader applications, it is prudent to characterize the profile of CLS to help providers guide safe management.
Method: We reviewed all acute lymphoblastic leukemia (ALL) patients who had participated in the clinical trial from our center to receive CAR-T therapy between 2016-2018. Patients analyzed in the study received either CD19 CAR-T cells or CD19 plus CD22 CAR-T cells. The diagnosis of CLS includes edema, acute hypotension and hemoconcentration with hypoproteinemia or hypoalbuminemia. CRS grading was evaluated with Lee's criteria for CRS.
Result: 42 ALL patients were included in this study with the mean age of 27(8-52) years old. 11(11/42, 26.2%) patients were diagnosed as CLS and 31 were not. It was observed that CLS was more common in patients who developed severe CRS. Patients with CLS was found to have high rate of hypotension and use of gamma globulin.(Table 1) Top level concentration of serum IL-6 in CLS patients was much higher than that in non-CLS patients (16438.7 vs 3292.7 pg/mL, p=0.0016), which is consistent with the well recognized concept of IL-6 as an indicator of CRS.(Figure 1) It is important to notice that CLS patients had lower levels of serum total protein (TP, 43.7 vs 52.8 g/L, p=0.0005) and serum albumin (ALB, 27.4 vs 33.8 g/L, p=0.0011), while the hemoglobin (HGB) concentration showed no difference, suggesting that TP and ALB might be better indicators for CLS than HGB, although hemoconcentration, hypoproteinemia and hypoalbuminemia are both important in diagnosis.(Figure 2) Moreover, there was no significant difference in age, gender, Ph type of ALL, type of CAR-T cells infused and death ratio.(Table 1) Although CRS has been reported to be related with disease burden before the therapy, our data showed no difference of it between the patients with and without CLS.
Conclusion: In conclusion, we have evaluated a basic profile of CLS among CAR-T patients in our center and the study indicates that CLS warrants extra attention for patients who receive CAR-T therapy. Further investigations are required to elucidate best practices for prevention and management of CLS in CAR-T therapy.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.