CAR-T and Other Immunotherapies in Myeloma Ivan Borrello, M.D. - PowerPoint PPT Presentation

CAR-T and Other Immunotherapies in Myeloma Ivan Borrello, M.D.

bb2121: BCMA CAR T Cell Design bb2121 CAR Design CD8 MND SP Anti-BCMA scFv 4-1BB CD3z Promoter Linker Signaling Domains Tumor binding domain • Autologous T cells transduced with a lentiviral vector encoding a CAR specific for human BCMA • Optimal 4-1BB costimulatory signaling domain: associated with less acute toxicity and more durable CAR T cell persistence than CD28 costimulatory domain 1 1. Ali SI, et al. Blood . 2016;128(13):1688-700.

CAR-T Toxicities 3

4

Treatment History Escalation Expansion Parameter (N=21) (N=22) Median (min, max) prior regimens 7 (3, 14) 8 (3, 23) Prior autologous SCT, n (%) 21 (100) 19 (86) 0 0 3 (14) 1 15 (71) 14 (64) >1 6 (29) 5 (23) Escalation (N=21) Expansion (N=22) Parameter Exposed Refractory Exposed Refractory Prior therapies, n (%) Bortezomib 21 (100) 14 (67) 22 (100) 16 (73) Carfilzomib 19 (91) 12 (57) 21 (96) 14 (64) Lenalidomide 21 (100) 19 (91) 22 (100) 18 (82) Pomalidomide 19 (91) 15 (71) 22 (100) 21 (96) Daratumumab 15 (71) 10 (48) 22 (100) 19 (86) Cumulative exposure, n (%) Bort/Len 21 (100) 14 (67) 22 (100) 14 (64) Bort/Len/Car/Pom/Dara 15 (71) 6 (29) 21 (96) 7 (32) Data cutoff: March 29, 2018. SCT, stem cell transplant.

Cytokine Release Syndrome Cytokine Release Syndrome Parameters Cytokine Release Syndrome By Dose Level Dosed Patients 100 Parameter (N=43) Maximum Toxicity Grade a 82% Patients with a CRS event, n (%) 27 (63) 80 3 2 1 9,1 Maximum CRS grade a Patients, % 22,7 60 None 16 (37) 1 16 (37) 39% 40 2 9 (21) 3 2 (5) 22,2 50,0 4 0 20 16,7 Median (min, max) time to onset, d 2 (1, 25) 0 150 × 10 6 >150 × 10 6 Median (min, max) duration, d 6 (1, 32) 150 x 106 >150 x 106 (n=18) (n=22) Tocilizumab use, n (%) 9 (21) Dose Level b Corticosteroid use, n (%) 4 (9) Data cutoff: March 29, 2018. a CRS uniformly graded according to Lee DW, et al. Blood. 2014;124(2):188-195. b 3 patients were treated at the 50 x 10 6 dose level for a total of 43 patients.

bb2121 CAR+ T Cell Expansion Median (Q1, Q3) Vector Copies in CD3-Enriched Peripheral Peak bb2121 Vector Copies in Responders vs Blood by Dose Cohorts Nonresponders 1 0 7 Median Vector Copies/µg of P =0.005 1 0 6 Genomic DNA 1 0 5 1 0 4 1 0 3 Patients with a post-baseline vector copy value were included. One patient was dosed at 205 × Patients with ≥2 months of response data and 1 month of vector copy data (N=36). 10 6 CAR+ T cells instead of the planned 450 × 10 6 and was included in the 450 × 10 6 dose group. P value based on a 2-sided Wilcoxon rank sum test. • Comparable C max in active dose cohorts (≥ 150 × 10 6 Month 1 Month 3 Month 6 Month 12 CAR+ T cells) At risk, n 32 26 16 10 • Durable bb2121 persistence (≥6 months) in 44% With detectable vector, n (%) 31 (97) 22 (85) 7 (44) 2 (20) • Higher peak expansion in patients with response Data cutoff: March 29, 2018. C max , maximum serum concentration; LLOQ, lower limit of quantitation. 7

Tumor Response: Deep MRD- negative responses observed 50 × 10 6 150 × 10 6 450 × 10 6 800 × 10 6 Response Total MRD- 0 4 11 1 16 evaluable responders MRD-neg a 0 4 (100) 11 (100) 1 (100) 16 (100) Data cutoff: March 29, 2018. a Of 16 MRD-negative responses: 4 at 10 -6 , 11 at 10 -5 , 1 at 10 -4 sensitivity by Adaptive next-generation sequencing assay. • All responding patients evaluated for MRD were MRD negative at 1 or more time points • 2 nonresponders evaluated for MRD were MRD positive at month 1

Progression-Free Survival • mPFS of 11.8 months at active doses (≥ 150 × 10 6 CAR+ T cells) in 18 subjects in dose escalation phase • mPFS of 17.7 months in 16 responding subjects who are MRD-negative PFS at Inactive (50 × 10 6 ) and Active (150 – 800 × 10 6 ) Dose PFS in MRD-Negative Patients a Levels a 150 – 800 × 10 6 50 × 10 6 150 – 800 × 10 6 (n=3) (n=18) (n=16) Events 3 10 mPFS (95% CI), 17.7 mPFS (95% CI), 2.7 11.8 (5.8 – NE) mo (1.0 – 2.9) (8.8 – NE) mo mPFS = 11.8 mo mPFS = 17.7 mo mPFS = 2.7 mo Data cutoff: March 29, 2018. Median and 95% CI from Kaplan-Meier estimate. NE, not estimable. a PFS in dose escalation cohort.

Marrow Infiltrating Lymphocytes MILs Exhibit Significant Anti-Myeloma Specificity aMILs Effectively Kill Myeloma Cells 160000 Nothing 140000 CD33 120000 CD138 CPM/ 10^5 CD3 100000 80000 60000 40000 20000 0 PBL aPBL MILs aMILs MILs eradicate pre-established disease 700 HBSS aMIL 600 aPBL Human Kappa (ng/ml) 500 400 300 200 100 0 0 30 50 70 90 110 130 150 170 190 240 Noonan et al Ca Res 2005; 65(5) Days post tumor challenge

MILs Persist in the Bone Marrow and Eradicate Myeloma aMILs aPBLs day 79 day 110 Human CD3 + 0 10 1 10 2 10 3 10 4 10 0 10 1 10 2 10 3 10 4 APC 10 Human CD138 + Control APC No Stain M1 M1 100 101 102 103 104 100 101 102 103 104 PE PE

First MILs Clinical Trial

Tumor Specificity of aMILs Product 40 p=0.07 35 %CD3 + /CFSE low /IFNg + 30 25 20 15 10 5 0 CR PD 13 (Noonan et al. STM 2015; 7:288)

Tumor-specific Response in the BM Correlates with Clinical Outcomes 40 * CR PR 35 *p<0.001 * SD PD 30 %CD3+/CFSElow/IFN γ + 25 20 15 10 5 0 D60 D180 D360 (Noonan et al. STM 2015; 7:288)

41BB Expression with Expansion PBL Normoxia Hypoxia Pre-Activation CD4 + /41BB + CD4 + /41BB + CD4 + /41BB + =0% =8.14% =2.8% CD4 + MIL CD4 + /41BB + CD4 + /41BB + CD4 + /41BB + CD4 + =43.4% =10.67% =18.21%

Hypoxia Enhances Function in 4- 1BB+ T cell Subset 10 9 8 7 6 5 2139 untouched 4 3 2 2139 4- 1BBneg 1 0 2139 4- 1BBpos 16

In vivo MILs Expansion 4000 J0770 J0997 J1343 MILs J1343 No MILs N=21 N=30 N=5 N=2 3500 3000 Average ABS Lymph Count Hypoxia MILs Normoxia MILs 2500 2000 1500 1000 No MILS 500 0 d+3 7 14 21 28 60 180

Superior Killing by MIL-CARs Compared to PBL-CARs Day 9 Day 5 Day 3 PBLs CAR 57.8% 5.7% 36.1% CD3 MILs CAR 0% 1.6% 7.1% CD138 N.B: 8226 cells was added on days 3 or 7 days after the primary 8226 challenge

MIL CARs: More Data Showing Superior Killing via the CAR in MIL CARs vs. PBL CARs Rechallenge Primary Challenge (48hr) 1 0 0 1 0 0 P = 0 . 0 2 1 * P = 0 . 0 0 7 7 * * % 8 2 2 6 C e l l K i l l i n g % 8 2 2 6 C e l l K i l l i n g 5 0 5 0 0 0 C A R M I L s C A R P B L s C A R M I L s C A R P B L s CART:Target ratio = 1:10 19

MIL CARs: Preserve the Endogenous TCR-mediated Killing Tumor Specificity Assay Testing ability of Native TCR to Recognize Tumor Ag: CD38 Stimulated NT MILs CD38 MIL CARs CD38 MIL CARs 4.8% 14.9% 21.0% Native TCR in MIL CARS works even after the CAR has fired

Conclusions • Tumor specificity of MILs correlates with clinical outcomes • Memory phenotype, broad antigenic specificity are properties unique to MILs and not found on PBLs • T cell persistence correlates with responses • Hypoxia augments T cell function of MILs through – upregulation of 4-1BB – increase in anti-apoptotic proteins and survival cytokines – Enhance ex vivo and in vivo expansion • The absence of a PFS plateau with BCMA CARs limits the long- term efficacy of this approach in MM • MILs appear to show better anti-tumor activity as a source of CAR-modified T cells than PBLs 21

Acknowledgements Borrello Lab Myeloma Group Funding Abbas Ali Megan Heiman NIH BMT PO1 Valentina Hoyos Carol Ann Huff Bill Matsui Luca Biavati Danielle Dillard Amy Sidorski Ervin Griffin Satish Shanbhag Jenn Hanle Amy Thomas Commonwealth Clinical Research WindMIL Foundation Kim Noonan Laura Cucci Eric Lutz Leo Luznik Phil Imus Lakshmi Rudraraju Maria Yankouski Amanda Stevens Cell Therapy Lab Janice Davis Vic Lemas 22 Sue Fiorino