Individual Patient Data Meta-Analysis from 16 Trials for Safety Factors in Cytokine Release Syndrome After CAR-T Therapy in Patients with Non-Hodgkin Lymphoma (NHL) and Acute Lymphoblastic Leukemia

Introduction

Chimeric antigen receptor T cells (CAR-T) with anti-CD19 have shown great promise in the treatment of relapsed and refractory non-Hodgkin lymphoma (NHL) and acute lymphoblastic leukemia (ALL). Cytokine release syndrome (CRS) is the most significant and life-threatening side effect. This individual patient data (IPD) meta-analysis is to investigate the association of severe CRS with CAR-T dose and baseline factors.

Methods

We collected the individual patient-level data of 237 patients with NHL or ALL from 16 published papers. A logistic model was used to analyze the association of severe CRS incidence with CAR-T dose and baseline factors including age and baseline tumor burden. A generalized additive model (GAM) with logit link function was used to estimate the nonlinear response for severe CRS incidence with CAR-T dose and baseline factors.

Results

Severe CRS incidence was positively associated with current proposed CAR-T treatment infusion dose at a range of 0.2 × 10⁶–5.0 × 10⁶ T cells per kg of body weight in patients less than or equal to 25 years old. For patients over 25 years old the association was not significant. Significant association between severe CRS incidence and baseline tumor burden was also shown in this study.

Conclusions

Our results provide novel insights that association between CAR-T treatment dose and severe CRS incidence only exists in patients less than or equal to 25 years old. Severe CRS incidence is associated with baseline tumor burden which indicates that tumor burden needs to be controlled with induced chemotherapy before CAR-T treatment.