Background:

JCAR017 is a CD19-directed 4-1BB CAR T cell product with a highly controlled manufacturing process that enables administration in a defined composition at a precise dose of CD8 and CD4 CAR T cells. Preliminary safety data on JCAR017 in relapsed/refractory (R/R) B-cell NHL (TRANSCEND NHL 001, NCT02631044; interim results previously reported [Abramson, 14-ICML 2017]) showed lower rates of 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. The manufacturing process control strategy for JCAR017 was designed to minimize between-lot variability in drug product quality attributes.

Methods:

JCAR017 was manufactured from non-Hodgkin lymphoma (NHL) patient-derived autologous T cells in TRANSCEND NHL 001. JCAR017 drug product attributes were analyzed for phenotypic, functional, and cell health related attributes. The memory and cell health phenotypes of CAR T cells were examined using flow cytometry and T cell functionality was assessed using in vitro antigen-specific bioassays.

Results:

JCAR017 process and product control begins with an automated wash and T cell purification that results in pure CD8+ (median 99%, Inter Quartile Range [IQR] 98-100%) and CD4+ (median 99%, IQR 99-100%) populations (data presented for patient lots manufactured through 01AUG2017). Separate CD4+ and CD8+ T cell populations are transduced with lentivirus to express a CD19-directed CAR with a 4-1BB/CD3ζ endodomain. The CD4+ and CD8+ T cell fractions are expanded separately to a target CAR T cell dose and formulated / cryopreserved separately. Cultures are optimized for the unique biology of each phenotype. To minimize between-lot variance in cell health, the drug product is held at constant volume with a tight range of viable cell concentrations (CD8+: median 31x106 cells/mL, IQR 28-40 x106 cells/mL, N=38; CD4+: median 35x106 cells/mL, IQR 31-40x106, N=36). Additionally, JCAR017 cell health is assessed by measuring post-thaw cell surface Annexin V and active intracellular Caspase 3. The median % Annexin V+ is 11%, IQR 9-18% (N=33) of CD8+CAR+ T cells and 10%, IQR 8-17% (N=31) of CD4+CAR+ T cells. Caspase 3 expression is similar to Annexin V. In vitro antigen-specific cytokine accumulation and intracellular cytokine bioassays show similar low variance between lots for multiple cytokines. For example, after CD19 stimulation TNFα production has a tight range with relative standard deviation (RSD) of 37% for CD4+CAR+ T cells (N=59) and 51% for CD8+CAR+ T cells (N=61). Details will be presented and compared to a similar CAR T cell product without a defined composition.

JCAR017 is administered in precise quantities of CD4+CAR+ and CD8+CAR+ T cells to patients. In an ongoing, multicenter, Phase 1 trial (TRANSCEND NHL 001), JCAR017 is administered as a flat dose at dose level 1 (DL1) of 5×107 CAR+ T cells (2.5x107 CD4+CAR+ T cells and 2.5x107 CD8+CAR+ T cells) or at dose level 2 (DL2) 10×107 CAR+ T cells (5x107 CD4+CAR+ T cells and 5x107 CD8+CAR+ T cells). At the site, JCAR017 is thawed and a patient-specific volume is withdrawn and administered. The range of administered dose was found to have low variability: 2.4-2.7x107 (target ±8%) CD4+CAR+ or CD8+CAR+ T cells at DL1 (N=48) and 4.6-5.1x107 (target ±8%) CD4+CAR+ or CD8+CAR+ T cells at DL2 (N=20). After administration to the patient, JCAR017 has a slow expansion profile with peak expansion at approximately 15 days with low rates of severe CRS and NT and an emerging dose:response relationship. Full PK, clinical efficacy, and safety of JCAR017 will be presented separately.

Conclusions:

The process and product control of JCAR017 results in high T cell purity and low between-lot variance. Three aspects of JCAR017 manufacturing contribute to the low variability between lots and encouraging clinical profile: 1) a precise, consistent flat dose of administered CD4+ and CD8+ CAR T cells; 2) control and optimization of CD4 and CD8 T cell culture conditions that results in low variability of cytokine production (e.g. IL-2, TNFα and IFNγ, etc.); and 3) constant formulation and volume of drug product that contributes to consistent cell health. The process control strategy and defined cellular composition of JCAR017 could contribute to the lower rates of CRS and NT than reported for other CD19-targeted CAR T cell products in clinical development for NHL.

Disclosures

Ramsborg: Juno Therapeutics: Employment, Equity Ownership, Patents & Royalties. Guptill: Juno Therapeutics: Employment, Equity Ownership. Weber: Juno Therapeutics: Employment, Equity Ownership. Christin: Juno Therapeutics: Employment, Equity Ownership. Larson: Juno Therapeutics: Employment, Equity Ownership, Patents & Royalties. Lewis: Juno Therapeutics: Employment, Equity Ownership. Mallaney: Juno Therapeutics: Employment, Equity Ownership. Bowen: Juno Therapeutics: Employment, Equity Ownership. Higham: Juno Therapeutics: Equity Ownership. Albertson: Juno Therapeutics: Employment, Equity Ownership, Patents & Royalties.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution