PF305 ANALYSES OF CYTOKINE RELEASE SYNDROME AND NEUROTOXICITY BY AGE AND LYMPHODEPLETING CHEMOTHERAPY USE IN ADULTS WITH RELAPSED OR REFRACTORY DIFFUSE LARGE B‐CELL LYMPHOMA TREATED WITH TISAGENLECLEUCEL

Abstract

Background:Autologous anti‐CD19 chimeric antigen receptor (CAR) T‐cell therapy tisagenlecleucel has demonstrated efficacy with a favorable risk/benefit profile in adults with relapsed or refractory diffuse large B‐cell lymphoma (r/r DLBCL) in the phase 2 JULIET trial. CAR T‐cell therapy is associated with unique toxicities, e.g., cytokine release syndrome (CRS) and neurotoxicity (NT), that require specialized management. It is important to understand how patient (pt) characteristics (e.g., age) and prior lymphodepleting (LD) chemotherapy regimens could inform risk and management of CRS and NT.Aims:This JULIET post‐hoc analysis aims to understand unique safety profiles associated with CAR T‐cell therapy in adults with r/r DLBCL treated with tisagenlecleucel, stratified by age or LD chemotherapy use.Methods:CRS and NT events observed in JULIET were graded with approaches common to CAR T‐cell therapy (CRS: Penn, Lee; NT: Common Terminology Criteria for Adverse Events v4.03 [CTCAE], modified CAR‐T Related Encephalopathy Syndrome [mCRES]). Penn grades were reported per trial protocol; other grades were obtained via expert consensus grading (ASH2018 Schuster #4190; Maziarz #4183). Four medical experts independently re‐graded CRS using the Lee scale while blinded to original Penn grades, and graded NT using CTCAE and mCRES. Experts discussed grading assessments and reconciled disagreements at live meetings. Per the investigational charter, in cases that could not be reconciled, the most conservative final assessment of any reviewer determined final grading for any individual case.This analysis stratified JULIET pts who received tisagenlecleucel by age at screening (<65 or ≥65 years old [yo]) or LD chemotherapy use before infusion and type (fludarabine+cyclophosphamide [FC], bendamustine [B], or no LD therapy [permitted if white blood cell count ≤1,000 cells/μL ≥1wk pre‐infusion]). Demographics, disease characteristics, medical history, and CRS and NT grades were summarized and compared among subgroups.Results:Of 111 JULIET pts infused with tisagenlecleucel as of Dec 8, 2017, 86 were <65yo and 25 were ≥65yo. Overall, pts were similar in demographics and medical history, except that more pts aged <65yo received prior HSCT (<65 vs ≥65yo: 55.8% vs 24.0%, p < 0.01). The rates of any CRS grade (Lee: 59.3% vs 48.0%; Penn: 60.5% vs 48.0%), and grade ≥3 (Lee: 19.8% vs 8.0%; Penn: 22.1% vs 20.0%) were similar between age subgroups (Table 1). Among 64 pts with CRS per Penn (<65yo n = 52; ≥65yo n = 12), there were no differences in time to CRS onset, CRS duration, or proportions of pts with tocilizumab use. Overall, NT grades were similar between two subgroups with numerically higher rates in older pts: any grade (CTCAE: 43.0% vs 52.0%; mCRES: 14.0% vs 28.0%), grade ≥3 (CTCAE: 12.8% vs 20.0%; mCRES: 9.3% vs 24.0%). No grade 5 CRS or NT was observed in any JULIET pts. In stratification based on LD therapy (FC n = 81; B n = 22; no LD therapy n = 8), CRS and NT grades were similar across three subgroups (Table 1). Again, among CRS patients in LD therapy subgroups, no differences were observed regarding CRS characteristics and tocilizumab use.Summary/Conclusion:This JULIET post‐hoc analysis observed similar rates of CRS and NT grades using various grading approaches among pts aged <65 vs ≥65yo. In the only CAR T‐cell study for an approved product allowing different LD chemotherapies, no differences in safety were observed across LD regimens used in JULIET. These findings will inform clinical decision making in CAR T‐cell therapy use in r/r DLBCL.image