Preliminary Safety Profile of the CD19-Directed Defined Composition CAR T Cell Product JCAR017 in Relapsed/Refractory Aggressive B-NHL Patients: Potential for Outpatient Administration

Introduction

JCAR017 is a CD19-directed 4-1BB CAR T cell product administered in a defined composition at a precise dose of CD8 and CD4 CAR T cells. TRANSCEND NHL 001 is the first multicenter Ph1 trial of JCAR017 in relapsed/refractory (R/R) B-cell NHL (NCT02631044; interim results previously reported (Abramson, 14-ICML 2017)) that will enroll a planned pivotal cohort later this year. Preliminary safety data show lower rates of all grade cytokine release syndrome (CRS) and neurotoxicity (NT) than reported for other CD19-directed CAR T cell products with broad heterogeneity of CAR T cell dosing and cell composition characteristics, though all grade events for all of those products have not been reported. Based on these data, we have begun to administer JCAR017 in the outpatient setting and plan to have a proportion of patients in the pivotal cohort treated as outpatients. Herein we update the safety and tolerability of JCAR017 and adverse event (AE) management in R/R B-NHL patients on the ongoing TRANSCEND NHL 001 trial (NCT02631044).

Methods

Patients with R/R DLBCL NOS (de novo or transformed from indolent lymphoma), PMBCL, FL grade 3B, and adequate organ function are eligible for the DLBCL cohort. No minimum ALC is required for apheresis and poor-prognosis patients with previous failed allo-SCT, secondary CNS involvement, and ECOG 2 are eligible. Patients receive lymphodepletion (LD), with fludarabine and cyclophosphamide, followed by a single flat dose of JCAR017 at one of two dose levels (DL1, 5 × 10⁷ cells; DL2, 1 × 10⁸ cells); a few received a 2-dose schedule at DL1.

All AE data were reported and collected from LD through 90-days post JCAR017. Severity of cytokine release syndrome (CRS) is graded per Lee et al (Blood 2014) and NT is graded per CTCAE 4.03 for a given neurologic event preferred term. The safety profile of all patients, patients treated inpatient and those treated in the outpatient setting will be compared among the TRANSCEND NHL 001 study population.

Results

As of July 7, 2017, 69 DLBCL NOS (de novo or transformed from indolent lymphoma, PMBCL, FL grade 3B) patients were treated (51 male, 23 female). The most common treatment-emergent AEs (TEAEs) other than CRS or NT included neutropenia (41%, 28/69), fatigue (30%, 21/69), thrombocytopenia (30%, 21/69), and anemia (26%, 18/69). One Gr 5 related TEAE was diffuse alveolar damage.

No acute infusional toxicity occurred, and the majority of patients, 64% (44/69), had no CRS or NT, suggesting outpatient delivery of JCAR017 may be feasible. Rates of CAR T cell-associated toxicities, including CRS and NT, did not differ between dose levels. Among the 25 patients (36%) who experienced any grade CRS or NT, 21 (30%) had CRS and 14 (20%) had NT. No patients had Gr 3 CRS and only one (1%, 1/69) had Gr 4 CRS and required ICU care; the other 29% (20/69) had Gr 1-2 CRS. Of the 20% of patients with NT, 6% (4/69) had Gr 1-2 and 14% (10/69) had Gr 3-4; 2 (3%) had seizure. No Gr 5 CRS or NT occurred. All CRS and NT events resolved except one case of Gr 1 tremor, which was ongoing at the time of data cutoff. Median time to onset of first CRS and NT was 5 days (range 2, 12) and 10 days (range 5, 23), respectively. In the first 72 hours post infusion, no pts had NT, and only 10% (7/69) had CRS (all Gr 1); NT was preceded by CRS in >70% of patients. Overall, <20% required anti-cytokine therapy (tocilizumab alone 1 (1%), dexamethasone alone 6 (9%), and both 6 (9%)) and only one required any vasopressor support. Median doses of tocilizumab and dexamethasone were 1 and 6, respectively. Median CRS and NT duration was 5 days and 11 days, respectively.

Given the low incidence and late onset of CRS or NT, a strategy of hospital admission at the first sign of fever made outpatient infusion a feasible approach; 4 out of a planned ~20 patients have been treated in the outpatient setting to date. Additional data will be presented on rates of hospitalization, days in ICU, updated enrollment, toxicity rates, rates of intervention with tocilizumab and dexamethasone, and resolution in this safety cohort overall, and those treated in the outpatient setting.

Conclusion

The preliminary safety profile of JCAR017, including all grade CRS and NT, supports outpatient administration in a majority of R/R aggressive B-NHL patients. Additional safety data on JCAR017, which is currently being evaluated in both the inpatient and outpatient setting, will be reported at the meeting.