Predicted Fludarabine Exposure Is Independently Associated with CRS and ICANS after CD19 CAR T-Cell Therapy

Abstract

CD19-targeted chimeric antigen receptor-engineered (CD19 CAR) T-cell therapy has transformed the treatment landscape for relapsed or refractory B-cell malignancies but remains limited by non-durable responses and severe toxicities, such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). Fludarabine (Flu)-based lymphodepletion has been shown to have critical impact on the in vivo expansion and persistence of CD19 CAR T cells. In addition, it has been suggested that the predicted Flu exposure could be associated with improved outcomes in children with B-cell acute lymphoblastic leukemia (B-ALL) undergoing CD19 CAR T-cell therapy (Fabrizio et al., 2022) but such association has not been evaluated in adults. we applied multivariable regression to model the relationship between Flu exposure and clinical outcomes in patients with B-cell malignancies undergoing CD19 CAR T-cell therapy. We retrospectively analyzed data from 131 patients with B-cell malignancies undergoing Flu and cyclophosphamide (Cy) lymphodepletion (LD) and CD19 CAR T-cell therapy (1:1 CD8+:CD4+ composition) on a phase I/II clinical trial (NCT01865617) at our institution. Patients were excluded if they received LD other than Flu/Cy, had a delay in CAR T-cell infusion more than 3 days after LD, or had missing data needed for Flu exposure calculation. Cumulative Flu exposure was defined by the predicted (calculated rather than measured) area under the curve (Flu p[AUC]) using a population pharmacokinetic model, which takes into account the weight, estimated creatinine clearance, and the administered Flu dose (Langenhorst et al., 2019). Event-free survival (EFS) was defined as time from CAR T-cell infusion to disease progression or death and was estimated using Cox model analysis. Clinical responses were defined according to the 2019 NCCN guidelines for B-ALL (multiparameter flow-negative marrow complete response [CR] or CR/partial response [PR] by PET or CT in the absence of marrow disease), 2018 iwCLL for CLL (CR or PR), and the 2014 Lugano criteria for non-Hodgkin lymphoma (CR or PR). CRS and ICANS grade were determined using 2014 Lee et al. and CTCAE4.03 criteria, respectively. Univariate and multivariate Cox and proportional odds logistic regression were used to estimate the association between Flu pAUC and EFS, and between peak CRS/ICANS grade, respectively. Reported odds ratios (OR) reflect an increase from the 1st to 3rd quartile of Flu pAUC. Linear regression was used to estimate associations between Flu pAUC and serum cytokine concentrations. A total of 131 patients (ALL: 47, CLL: 41, NHL: 43) were included. Eligible patients had a median age of 55 years (IQR, 42-64), 4 prior lines of therapy (IQR, 3-6), and Flu pAUC of 19.95 mg x h/L (IQR, 17-23.5). We observed CR to CD19 CAR T-cell therapy in 39 of 47 ALL (83%), 6 of 41 (14.6%) CLL, and 22 of 43 NHL (51%). Following CAR T-cell infusion, the median EFS was 203 days. In a Cox model including disease type, CAR T-cell dose level, pre-LD LDH, and prior lines of therapy, there was no statistically significant association between Flu pAUC and the EFS (HR, 1.08; 95%CI, 0.83-1.42, p=0.56). In a multivariable model including disease type, CAR T-cell dose level, pre-LD LDH, and prior lines of therapy, higher Flu pAUC was independently associated with increased odds of CRS (OR, 1.7; 95%CI, 1.13-2.57; p=0.01) and ICANS (OR, 1.87; 95%CI,1.14-3.07; p=0.01). Probability curves of grade ≥3 CRS and ICANS are shown in the figure. While we could not confirm a correlation between Flu pAUC and peak in vivo CAR T-cell expansion (p=0.6), we observed a positive correlation between Flu pAUC and day-0 pre-infusion IL-18 (p=0.01), day-0 TNFRp75 (p=0.007), and peak IL-2 (p=0.036). In contrast, we measured a negative correlation between Flu pAUC and day-0 pre-infusion IL-22 (p=0.04). Flu pAUC was independently associated with CRS and ICANS after CD19 CAR T-cell therapy. Higher Flu pAUC was associated with higher day-0 pre-infusion serum concentrations of CRS-related cytokines, suggesting an impact on the cytokine milieu independent of CAR T-cell in vivo activation. Our findings support the optimization of Flu exposure (e.g. dose adjustments based on PK models and/or PK measurements) to mitigate severe toxicities after CD19 CAR T-cell therapy and select patient populations (e.g. baseline renal dysfunction or obesity) may require closer toxicity monitoring. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal