Supplementary Figure SF1. Schematic of sipuleucel-T treatment doses (infusions) and serum collection time points in IMPACT and ProACT.

Supplementary Figure 2. Schematic showing how sipuleucel-T is designed to activate a patient's immune system by ex vivo cellular activation during manufacture and subsequent in vivo immune response.

Supplementary Figure 3. CONSORT diagram with patient disposition.

Supplementary Figure S2 - Assessments of unique TCR sequences in the blood for healthy donors and for sipuleucel-T-treated subjects were performed on PBMCs taken before the preparation of each sipuleucel-T-treatment. A, TCR sequence profiles in healthy donors and sipuleucel-T-treated subjects with prostate cancer over time. Sipuleucel-T-treated subjects had a significantly higher mean number of productive TCR sequences at baseline (Week 0) compared with Week 2 and Week 4 (P < 0.01, paired t-test on the log scale). No differences in mean productive sequences in healthy donors over time were detected (NS (not significant), paired t-test on the log scale). The number of unique sequences was higher in sipuleucel-T-treated subjects at Week 0 and Week 2 compared with healthy donors (Wilcoxon one-sided test, P = 0.025 and 0.053, respectively). B, Shannon diversity index for TCR repertoire in blood from healthy donors and sipuleucel-T-treated subjects with prostate cancer. The median Shannon diversity index of healthy donors and a sipuleucel-T-treated subject at baseline (Week 0) and Weeks 2 and 4 is shown. TCR sequence diversity was higher in sipuleucel-T-treated subjects compared with healthy donors. Shannon diversity index decreased at Weeks 2 and 4 compared with baseline in sipuleucel-T-treated subjects (P < 0.001), whereas Shannon diversity index remained stable over time in healthy donors.

Supplementary Figure 1. Schematic of the STAMP trial design.

Supplementary data - Table S1. Incidence of all adverse events occurring in {greater than or equal to}15% of patients as well as grade 3 to 5 adverse events* Figure S1. Study schematic. Figure S2. Time-to-PSA progression* from testosterone recovery to {greater than or equal to}175 ng/dL. Figure S3. Time-to-testosterone recovery. Figure S4. Time-to-next anti-cancer intervention in both arms. Figure S5. Comparison of humoral antigen spread at week 2 between STAND versus IMPACT and STAMP. Figure S6. Sipuleucel-T parameters: (A) TNC count, (B) APC count, and (C) APC activation in autologous peripheral blood mononuclear cells after ex vivo PA2024 activation. Figure S7. Comparison of APC activation (CD54+ upregulation) between STAND and IMPACT.

Supplementary Tables 1-3, Figure Legends. ST1. Median cumulative product parameters, all randomized patients; ST2. Summary of maximal percent decrease from baseline in serum PSA; ST3: Most common AEs occurring within 1 day of infusion in >5% of patients.

Supplementary Figure SF2. Shown in this figure are four Kaplan-Meier (KM) plots comparing the OS of patients in the sipuleucel-T arm of IMPACT with no IgG response at week 10 (IgG responses = 0) to patients with {greater than or equal to}1 IgG responses (A), {greater than or equal to}2 IgG responses (B), {greater than or equal to}3 IgG responses (C), or {greater than or equal to}4 IgG responses (D) at week 10.

Supplementary materials and Methods, Tables ST1-13. Table ST1: Baseline clinical characteristics of patients in IMPACT Table ST2: Baseline clinical characteristics of patients in ProACT. Table ST3: Increase in levels of IgG against candidate antigens at weeks 2 and 22 in IMPACT as measured with ProtoArray Table ST4: Protein reagents used in Luminex xMAP assays. Table ST5: Overlap of the number of patients who were IgG responders to different antigens at week 10 after treatment in the sipuleucel-T arm of IMPACT. Table ST6: Evaluation of IgG responses to candidate antigens at week 12 in ProACT using Luminex xMAP. Table ST7: Association of post-treatment changes in serum levels of IgG at week 10 with OS in the sipuleucel-T arm of IMPACT. Table ST8: Comparison of OS in sipuleucel-T-treated IgG responders and IgG non-responders at week 10 with that in control patients in IMPACT. Table ST9: Evaluation of IgG responses against candidate antigens at weeks 2 and 22 in IMPACT using Luminex xMAP. Table ST10: Assessment of IgG responses against candidate antigens at weeks 4 and 20 in ProACT using Luminex xMAP. Table ST11: Overlap of the number of sipuleucel-T-treated patients who were IgG responders to antigens across the post-treatment time points in IMPACT Table ST12: Association of changes in serum IgG levels with OS at week 2 or 22 in the sipuleucel-T arm of IMPACT. Table ST13: Comparison of OS in sipuleucel-T-treated IgG responders and IgG non-responders with that in control patients at weeks 2 and 22 in IMPACT.

Supplementary Table S1. Baseline demographics and disease characteristics in patients with mCRPC from the phase III sipuleucel-T trials included in the eosinophil analysis and the overall pooled population. Supplementary Figure S1. Overview of the study designs of three phase III clinical trials from which data were pooled for this retrospective analysis. Supplementary Figure S2. Changes in eosinophil counts for sipuleucel-T-treated patients with or without an elevated eosinophil count in randomized phase III trials, versus the control group. Supplementary Figure S3. Association of immune response with change in eosinophil counts for sipuleucel-T-treated patients in randomized phase III trials.

Supplementary Figure S1 - Study schema

Statistical Modeling Approaches for Survival Subsequent to Disease Progression for Control-Arm Patients Including Parameter Estimates for all Model Terms.

2537 Background: Orva-cel is a B-cell maturation antigen–targeted chimeric antigen receptor (CAR) T cell therapy being evaluated in the phase 1/2 EVOLVE study (NCT03430011) in pts with RRMM who had at least 3 prior lines of therapy (Tx). We previously reported safety and efficacy in the phase 1 study and established the recommended dose (RD) of orva-cel as 600 × 106 CAR+ T cells (Mailankody et al, ASCO 2020). Cytokine release syndrome (CRS), a dominant toxicity of CAR T cell therapy, is mediated in part by IL-1. We explore the role of ppx with AKR, an IL-1 signaling inhibitor, on reducing the incidence of grade (G) ≥2 CRS after orva-cel treatment at the RD. Methods: Fourteen pts were enrolled sequentially for AKR ppx and treated with orva-cel at the RD. The non-AKR ppx control group comprised the remainder of the phase 1 pts receiving orva-cel at the RD (n = 19). The median follow-up (range) was 3.0 mo (1.8–6.2) for the AKR ppx group and 8.8 mo (5.3–12.2) for the non-AKR ppx group. AKR was administered as 100 mg SC the night before orva-cel infusion, 3 h before the infusion (Day 1), and q24 h on Days 2–5. Dosing was increased to q12 h if CRS developed. CRS was graded by Lee (2014) criteria. Tocilizumab (T) and steroids (S) were used per protocol-specified treatment management guidelines. Results: Disease characteristics and outcomes are shown in the table. In AKR ppx and non-AKR ppx groups, median number of prior regimens was 6 and 5, and bridging Tx was used in 57% and 68% of pts, respectively. The total frequency of CRS was similar in the 2 groups, but with less G 2 in the AKR ppx pts; relative risk (95% CI) = 0.54 (0.21, 1.38). No G ≥3 CRS was seen in either group. The incidence of neurological events (NE), G ≥3 infection, and macrophage activation syndrome/hemophagocytic lymphohistiocytosis (MAS/HLH) was similar. T and S use was numerically lower with AKR ppx. Orva-cel expansion kinetics were similar in the 2 groups. All pts had a 2-month efficacy assessment, with ORR in 100% of AKR ppx and 95% of non–AKR ppx pts. Conclusions: In this nonrandomized evaluation of AKR ppx with orva-cel treatment, the incidence of G ≥2 CRS was lower in pts receiving AKR ppx. The use of AKR ppx produced no adverse effect on the incidence of NE, infection, or MAS/HLH, nor on orva-cel expansion or disease response. These results warrant further study of AKR ppx in CAR T cell therapy. Clinical trial information: NCT03430011. [Table: see text]

Background: Orva-cel is an investigational B-cell maturation antigen (BCMA)-targeted chimeric antigen receptor (CAR) T cell product genetically modified with a lentiviral vector to express a CAR construct with a unique fully human single-chain variable fragment, optimized spacer, and 4-1BB costimulatory and CD3ζ activation domains. Orva-cel is currently being evaluated for efficacy and safety in the ongoing phase 1/2 EVOLVE study (NCT03430011) in heavily pretreated patients with relapsed/refractory multiple myeloma. We characterized orva-cel drug products, manufactured using the process in place for the phase 2 portion of the study and intended for commercial manufacturing, for CAR+ T cell purity, phenotype, and function. Methods: Immunophenotyping was performed by flow cytometry of both surface and intracellular markers, including CD3, CD4, CD8, CD45, CCR7, CD45RA, CD28, CD27, and active caspase 3. Cytokine production after challenge with BCMA+ target cells was assessed by intracellular cytokine staining and Luminex multiplex assay of secreted cytokines, including interferon (IFN)-γ, tumor necrosis factor (TNF)-α, interleukin (IL)-2, and granzyme B (GrB). CAR-mediated in vitro proliferative capacity was measured after anti-idiotypic antibody stimulation using the IncuCyte Live-Cell Analysis System (Sartorius, Göttingen, Germany). In vivo CAR+ T cell proliferation and persistence were assessed by quantitative polymerase chain reaction (qPCR). Results: The orva-cel manufacturing process was designed to enable consistent production of highly pure CD3+ cell products (median frequency of CD3+ T cells, 99.96%; quartiles 1-3 interquartile range, 99.9%-100.0%; n = 81). Orva-cel drug products were characterized by high frequencies of less-differentiated CAR+ T cells, leading to a dominant central memory-like population (CCR7+CD45RA- CAR+ T cells) and substantial frequencies of naïve-like cells (CCR7+CD45RA+ CAR+ T cells) (Figure). When assayed for in vitro functional activity, orva-cel drug products showed robust antigen-specific cytokine and effector molecule production (IFN-γ, TNF-α, IL-2, and GrB) upon challenge with BCMA+ tumor cells, as well as vigorous proliferation in response to CAR stimulation. Preliminary correlative analysis suggested that the early memory phenotype may be linked to increased CAR+ T cell proliferative capacity, as determined by in vitro experiments and in vivo PK parameters (ie, maximum CAR+ T cell concentration observed in the blood [Cmax], time to Cmax, area under the curve from Day 0 to 28 [AUC0-28], and CAR+ T cell persistence at Month 3 and Month 6). Consistent with the early CAR T cell memory phenotype, qPCR analysis showed robust in vivo proliferation of CAR+ T cells after infusion, with a median Cmax of 1.54 × 105 transgene copies/µg DNA and median AUC0-28 of 1.61 × 106 transgene copies/µg DNA*day, as well as long-term in vivo persistence, with CAR+ T cells detected in 69% of patients at 6 months postinfusion. Conclusions: The orva-cel manufacturing process results in drug products characterized by highly pure T cells, with high frequencies of early memory and polyfunctional CAR+ T cells. Orva-cel drug products showed robust antigen-specific degranulation, production of multiple cytokines, sustained in vitro and in vivo proliferation, and in vivo persistence. Disclosures Colonna: Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Navarro:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Devries:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Beckett:Bristol-Myers Squibb Company: Current equity holder in publicly-traded company; Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment. Amsberry:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Radhakrishnan:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Piasecki:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Heipel:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Li:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Kavita:Bristol Myers Squibb Company: Current Employment, Current equity holder in publicly-traded company. Works:Bristol-Myers Squibb Company: Current equity holder in publicly-traded company; Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment. Mujacic:Bristol-Myers Squibb Company: Current equity holder in publicly-traded company; Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment.

Background: Orva-cel, a B-cell maturation antigen (BCMA)-targeted chimeric antigen receptor (CAR) T cell therapy, has shown promising preliminary efficacy and a favorable safety profile in patients with relapsed/refractory multiple myeloma in the ongoing phase 1/2 EVOLVE study (NCT03430011). Patients enrolled in the trial were heavily pretreated (median of 6 prior lines of therapy) and refractory to their last regimen per International Myeloma Working Group criteria. Five orva-cel dose levels (50, 150, 300, 450, and 600 × 106 CAR+ T cells) have been evaluated in 115 patients as of May 1, 2020. Here, we present the serum soluble BCMA (sBCMA) results and levels of immune-related serum factors associated with study safety and efficacy endpoints. Methods: Baseline and postinfusion levels of sBCMA (preinfusion through disease progression) and 39 immune-related serum factors (preinfusion through Day 29) were quantitated in serum by immunoassay (MesoScale Discovery, Rockville, MD). Linear regression models were used to evaluate the effect of orva-cel dose on biomarkers. Logistic regression models were used to evaluate the association between biomarkers and binary clinical response and safety endpoints. All patients with ≥1-month follow-up were included in the safety analyses (n = 115); all patients with ≥3-months potential follow-up were included in the efficacy analyses (n = 102). For assessment of sBCMA levels at nadir, only patients with sampling at Day 75 or later were considered (n = 91). Results & Conclusions: Baseline sBCMA levels (median [range], 480 [3.4-5126.5] ng/mL) correlated with established clinical measures of tumor burden, including percentage of malignant plasma cells in the bone marrow, paraprotein levels, and serum free light chain concentrations. Baseline sBCMA levels also correlated with laboratory measures predictive of prognosis, including β2-microglobulin, lactate dehydrogenase, and ferritin levels. As high tumor burden and high levels of these laboratory measures are both associated with poorer outcomes and increased frequency of adverse events (AEs), baseline sBCMA may provide a useful composite marker of disease burden and prognostic indicators. Patients with higher baseline sBCMA levels were more likely to experience AEs, including cytokine release syndrome (CRS) and neurological events (NEs) (P < 0.05). Overall response rate (ORR) was not influenced by baseline sBCMA levels, indicating that circulating sBCMA did not interfere with the ability of orva-cel to bind to and kill tumor cells expressing BCMA such that patients achieved at least a short-term response. Despite a similar ORR, patients with high baseline sBCMA levels were less likely to have an ongoing response at Month 6 (P < 0.05). These associations between baseline sBCMA levels and safety and efficacy endpoints appeared to hold within each dose level cohort, despite lacking power to detect a statistical difference. sBCMA levels decreased after orva-cel infusion in all responding patients (n = 86), with nadir reached within 2-3 months (median [range] time to minimum concentration, 2.8 [0.4-17.9] months). A dose effect was observed, with a 34% decrease in sBCMA nadir, on average, for every 150 × 106 CAR+ T cell dose increase (P < 0.05). sBCMA levels at nadir were associated with ORR, complete response rate, and Month 6 response rate (P < 0.05). sBCMA concentrations within the first month postinfusion allowed for stratification of Month 6 responders vs nonresponders. The significance of this stratification increased from Days 4-5 through Month 2, indicating that sBCMA may serve as an important early biomarker of durable response. Baseline levels of immune-related serum factors did not correlate with safety or efficacy endpoints in the current data set. A subset of serum factors characteristic of CAR T cell activation increased after orva-cel infusion; the peak levels of some biomarkers were associated with the orva-cel dose, including interleukin (IL)-2 and IL-6. Peak postinfusion levels of several inflammatory factors were associated with CRS (eg, IL-6 and IL-8) and NE (eg, tumor necrosis factor-α) grades (P < 0.05). Correlative analysis is ongoing, and updated results will be presented. Disclosures Piasecki: Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Devries:Bristol-Myers Squibb Company: Current equity holder in publicly-traded company; Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment. Radhakrishnan:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Li:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Heipel:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Fox:Bristol-Myers Squibb Company: Current Employment, Current equity holder in publicly-traded company. Beckett:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Cota Stirner:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Conte:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Doerr:Juno Therapeutics, a Bristol-Myers Squibb Company: Current Employment; Bristol-Myers Squibb Company: Current equity holder in publicly-traded company. Mailankody:PleXus Communications: Honoraria; Takeda Oncology: Research Funding; Janssen Oncology: Research Funding; Allogene Therapeutics: Research Funding; Juno Therapeutics, a Bristol-Myers Squibb Company: Research Funding; Physician Education Resource: Honoraria. Wong:Fortis: Research Funding; Amgen: Consultancy; Bristol Myers Squibb: Research Funding; Janssen: Research Funding; Roche: Research Funding; GSK: Research Funding; Sanofi: Membership on an entity's Board of Directors or advisory committees. McCarthy:Janssen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board; Magenta: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board; Karyopharm: Consultancy, Honoraria; Juno Therapeutics, a Bristol-Myers Squibb Company: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board , Research Funding is to Roswell Park, Research Funding; Starton: Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board; Genentech: Consultancy, Honoraria; AbbVie: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board; Takeda: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other: Advisory Board.

8504 Background: Orva-cel is an investigational, BCMA-directed CAR T cell product with a fully human binder. Over 100 pts have been treated in the EVOLVE phase 1 study. Pts treated at 50 and 150 × 10 6 CAR+ T cells were previously reported (Mailankody ASH 2018 #957). We now report results of the higher dose levels (DLs) in 51 pts who received orva-cel manufactured using the process intended to support commercial use. Methods: Pts with RRMM who had ≥3 prior regimens, a proteasome inhibitor (PI), an immunomodulatory drug (IMiD), and an anti-CD38 monoclonal antibody (mAb), received orva-cel at 300, 450, and 600 × 10 6 CAR+ T cells after lymphodepletion with fludarabine/cyclophosphamide. Results: Median pt age was 61 (range, 33–77) y; median time from diagnosis was 7.0 (range, 1.7–23.6) y, with a median of 6 (range, 3–18) prior regimens. Overall, 92% of pts were penta-exposed (2 IMiDs, 2 PIs, and an mAb); 61% of pts received bridging therapy (77% were refractory to bridging therapy). Two pts had dose-limiting toxicities: grade 3 neurological event (NE) for >7 d at 300 × 10 6 CAR+ T cells and grade 4 neutropenia for >28 d at 450 × 10 6 CAR+ T cells. Key efficacy and safety outcomes are shown in the Table. Cytokine release syndrome (CRS)/NEs were managed with tocilizumab and/or steroids (78%), anakinra (14%), and/or vasopressors (6%). Grade ≥3 anemia, neutropenia, and thrombocytopenia at 29 d occurred in 21%, 55%, and 44% of pts (median time to resolution to grade ≤2 of any cytopenia, ≤2.1 mo). Grade ≥3 infections occurred in 14%. After a median follow-up (F/U) of 5.9 mo, median progression-free survival was not reached. Conclusions: Orva-cel at 300, 450, and 600 × 10 6 CAR+ T cells demonstrated manageable safety (CRS grade ≥3: 2%; NE grade ≥3: 4%) and compelling efficacy in heavily pretreated pts with RRMM, with a 91% objective response rate (ORR) and 39% complete response (CR)/stringent CR (sCR) rate. Updated results will be presented, including minimal residual disease, durability of response, and recommended phase 2 dose. Clinical trial information: NCT03430011 . [Table: see text]

122 Background: JCAR017 is a defined composition, CD19-directed 4-1BB CAR T cell product administered at a precise dose of CD8 and CD4 CAR T cells in a seamless design Ph1 pivotal trial of R/R B-cell NHL (TRANSCEND NHL 001; NCT02631044). Methods: Blood samples were collected for biomarker analyses at protocol-defined time points. PK (CAR T cell expansion and persistence) was measured using flow cytometry. Cytokines were measured on a Luminex platform. Additional analytes will be presented. All reported p-values are 2-sided without multiplicity adjustment. Results: Safety (n = 59) and efficacy (n = 54) outcomes were analyzed for correlations with patient (pt) characteristics and biomarkers. Dose level did not correlate with cytokine release syndrome (CRS) or neurotoxicity (NT) despite higher median Cmax and median AUC0-28 at DL2. In pts with NT or ≥Gr 2 CRS, CD4 and CD8 CAR T cell levels were 5-10 fold and 3-5 fold higher, respectively, than median DL2 levels. Pt factors that correlated with any grade CRS and NT were ECOG 2 (p = 0.03) and high disease burden (p < 0.05). Higher levels of IL-8, IL-10, and CXCL10 before CART cell infusion were associated with Gr 3-4 NT (each p< 0.05), suggesting that inherent pt factors may result in higher CAR T expansion and associated CRS and NT. Lower pre-CAR T cell ferritin, LDH, CXCL10, G-CSF, and IL-10 were associated with CR/PR, and lower pre-CAR T cell ferritin, CRP, LDH, CXCL10, IL-8, IL-10, IL-15, MCP-1, MIP-1β, TNF-α were associated with 3-month durable response (each p< 0.05). Median Cmax and AUC0-28 of CD8 CAR T cells were higher in responding patients and with durable response at Month 3 (CD8 Cmax median = 20.8 vs 5.5; CD8 AUC median = 235 vs 55 in CR/PR vs PD at Month 3). Of pts evaluable for persistence at 3 months (n = 29), 90% and 93% had detectable CD8+ and CD4+ CAR+ T cells; of those with available PK results at time of relapse (n = 11), 82% had persistence at time of relapse. Conclusions: JCAR017 demonstrated increased CAR T cell expansion and persistence and higher durability of response at higher dose levels, with manageable toxicities. CAR T cells were also detected at time of relapse, suggesting potential opportunities for future combination clinical trials. Clinical trial information: NCT02631044.

Introduction: B-cell maturation antigen (BCMA) is expressed on malignant plasma cells and is an attractive therapeutic target for multiple myeloma. BCMA CAR T-cells, antibody drug conjugates and bispecific T-cell engagers have demonstrated substantial preclinical and clinical activity to date. JCARH125 is a BCMA-targeting CAR T product containing a lentiviral CAR construct with a fully human scFv, optimized spacer, 4-1BB co-stimulatory and CD3z activation domains. The construct has shown minimal tonic signaling and lack of inhibition by soluble BCMA. JCARH125 is generated using a manufacturing process developed to optimize various aspects, including increased consistency of cell health, in the drug product.

Abstract JCAR017 is a CD19-directed 41BB chimeric antigen receptor (CAR) T cell product administered in a defined composition at a precise dose of CD8 and CD4 CAR T cells. JCAR017 manufacturing employs process controls that enable precise independent control of the infused dose of CD8 and CD4 cells. Preliminary safety data from JCAR017 administered in relapsed/refractory B cell non-Hodgkin lymphoma (NHL) demonstrated lower rates of cytokine release syndrome (CRS) and neurotoxicity (NT) compared to those reported for other CD19-directed CAR T cell therapies with heterogeneity in total infused dose and CD8 and CD4 composition. Multiple patient factors and blood biomarkers have been identified that correlate with JCAR017 CAR T cell in vivo expansion, antitumor activity, and toxicity (Heipel M et al. and Siddiqi T et al. ASH 2017). However, little is known regarding the contribution of CAR T cell product differentiation state to pharmacokinetics (PK), clinical outcome, or toxicity. Product characterization is executed on JCAR017 to define the CAR T cell memory phenotype composition and antigen-specific function. Product attributes were assessed for relationships with clinical response, safety, and PK through univariate, multivariate, and machine learning-based analyses. JCAR017 memory T cell composition demonstrated strong relationships with cytokine production profile observed following in vitro CD19 stimulation of the CAR T cell drug product. Specifically, products with elevated frequencies of CCR7+ central memory T cells exhibited increased production of IL-2 (Spearman ρ=0.55, P<0.0001), whereas patient drug product with increased frequencies of effector differentiated T cells demonstrated increased production of IFNγ (ρ=0.51, P<0.0001) and IL-13 (ρ=0.45, P<0.0001). The correlations between CAR T cell drug product memory phenotype and function translated to positive correlations between central memory subset composition and peak in vivo CAR T cell expansion (ρ=0.42, P=0.002) and progression-free survival (Kaplan-Meier survival estimate, P=0.0164). In addition, an increased frequency of a central memory subpopulation in drug product demonstrated relationships with CRS (P=0.0069) and severe NT (P=0.0014) events. The phenotype to functional links described above offer insights as to how specific memory subpopulations contribute to the complex CAR T cell mechanism of action. CAR T cell products with increased CCR7+ central memory composition demonstrated increased peak CAR T cell expansion and persistence, suggesting less differentiated CAR T cells contribute to PK and progression-free survival. These findings can be used to further define the next generation of gene-engineered T cell products (e.g., a fixed dose of CCR7+CAR+) that will decrease the dose-to-dose variability. Citation Format: Ryan P. Larson, Rachel Lower, Todd DeVries, Yue Jiang, Ronald J. Hause, Rich Getto, Brian Christin, Nathan K. Yee, Michael A. Bowen, Clinton Weber, Daniel Li, Tina Albertson, Claire Sutherland, Christopher G. Ramsborg. Defined cell composition and precise control over JCAR017 dose enables identification of relationships between chimeric antigen receptor T cell product attributes, pharmacokinetics, and clinical endpoints in NHL [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 960.