Abstract number: 4838 Preclinical Evaluation of TC-110: CD19-Specific T Cell Receptor (TCR) Fusion Construct (TRuC™) T Cells for the Treatment of Hematologic Malignancy Holly Horton1, Jian Ding1, Seema Shah1,Jessica Geirut1, Nikolaus Thorausch2, Rashmi Choudhary1, Olga Kiner1, Irene Scarfo3, Bonnie Le1, Anna Morath2, Wolfgang Schamel2, Marcela Maus3, Daniel R. Getts1, Patrick A. Baeuerle1, Robert Hofmeister1 1 TCR2 Therapeutics, Cambridge, MA, USA, 2 Department of Immunology, Faculty of Biology, BIOSS Centre for Biological Signaling Studies, University of Freiburg, Freiburg, Germany, 3 Massachusetts General Hospital Background Results

Several CD19 targeting T cell therapies have been approved for treatment of Figure 1. In Vitro Characterization of TC-110 Figure 4. TC-110 Clears Tumor Faster in IV and SC NALM-6 and TC-110 T Cells Expand hematologic malignancies. These second generation CAR T cell therapies consist of at Early Timepoints NALM-6 and Raji cells express CD19 TC-110 Transduction Efficiency donor T cells engineered to express an extracellular CD19-targeting domain attached to IV NALM-6 SC NALM-6 1600 z 1011 an intracellular signaling domain such as 41BB (Kymriah) or CD28 (Yescarta). In the Vehicle 1400 TM present research, we evaluate a novel T Cell Receptor Fusion Construct, TRuC , in 1010 NT 1200

TC-110 1000 which the CD19 targeting domain is fused to the T cell receptor (TCR) at the CD3-e 9 10 CD19 41BBz CAR subunit. While second generation CARs act in isolation from the TCR through an 800 8 CD19 CD28 CAR independent signaling domain, TRuCs employ the full signaling machinery of the TCR 10 600 400 107 complex. Fusion of the tumor antigen binder to the TCR may yield more efficient 200

6 Mean Tumor Volume mm3 signaling upon binding of the T cell to the tumor antigen. In the present study, we TC-110 Tumor Cell Lysis 10 0 evaluate TC110 TRuC in which a CD19-targeting domain is fused to the CD3-e subunit Luminescence (photons/sec) 105 0 10 20 30 40 50 0 10 20 30 40 50 Study Day of the TCR. In mouse models of human leukemia and lymphoma, TC-110 treatment led Study Day to faster tumor regression compared to CD19 CAR with minimal cytokine release. TC110 Expansion in NALM-6 Non-Transduced 5

L blood L 10 TC-110 µ CD19 41BB CAR Material and Methods 104 CD19 CD28 CAR 3

T cells/100 10 + T cell generation: CD19 TRuC, TC-110, was generated by recombinantly fusing FMC63, Surface expression of CD19 was demonstrated on NALM-6 and Raji tumor cells. The transduction efficiency 102 CD19-specific single chain Fv (ScFv), to the CD3ɛ subunit. CD19-specific CARs (CD19- of the TRuC and CARs on T cells was detected using an anti-GFP antibody in flow cytometry. Tumor cell lysis 0 10 20 30

CAR) were generated by fusing FMC63 to the 2nd generation CAR with 4-1BB or CD28 of luciferase-expressing NALM-6 (24 hrs) and Raji tumor cells (72 hrs) was performed at an effector to target # of hCD3 Study Day intracellular domain. The constructs were introduced into primary human T cells via ratio of 10:1. NSG mice were implanted IV or SC with NALM-6 (5x105 cells per mouse). Six days later, mice were injected IV with 1x106 TRuC+ or CD19 BBz CAR or CD19 CD28 CAR T cells (1x106 CAR+) on Study Day 0. lentiviral transduction. After standard stimulation and expansion, TRuC™ surface Figure 2. TC-110 Dose Response in NALM-6 and Raji expression and T cell activation were analyzed by flow cytometry. Blood samples were collected at different time points and the number of T cells in the blood was In Vivo Experimental Design determined by FACs. During active tumor clearance (days 1-8 for IV model and 1-14 for SC model), TC- In vivo efficacy: The anti-tumor efficacy of TC-110 and CD19-CAR T cells was tested in 110 T cell numbers were highest compared to day 20 (post tumor regression). CD19 41BBz CAR had peak NSG mouse models of acute lymphoblastic leukemia (ALL, NALM-6) and Burkitt’s T cell expansion on day 14. CD19 CD28 CAR had minimal T cell expansion. T cell expansion on day 29 lymphoma (Raji). Blood samples were collected at different time points and the was most likely due to GVHD which sometimes occurs in xenogeneic tumor models treated with human presence of TC-110 and CD19-CAR T cells was detected by flow cytometry analysis. T cells. Luminex assay: The levels of human cytokines in mouse serum was measured using Figure 5. Serum Cytokines after Treatment with TC-110 in IV NALM-6 the Luminex-based Cytokine Magnetic Bead Panel Multiplex Assay (Millipore-Sigma). The mouse serum was collected on days 1, 2, 4, 7, and 10 days post T cell injection and stored at -80C until sample analysis. TC-110 Dose Response in Systemic NALM-6 and Raji

IV NALM-6 11 IV Raji Vehicle 1×10 1×1011 NT 1e7 cells TC-110 1e5 cells 1×1010 1×1010 TC-110 3e5 cells Platform Overview TC-110 1e6 cells TC-110 3e6 cells 1×109 1×109 T Cell Receptor Fusion Construct (TRuC™) TC-110 1e7 cells 8 1×108 1×10 Natural TCR e TRuC g TRuC b TRuC 1×107 1×107 Luminescence (photons/sec)

1×106 Luminescence (photons/sec) α β α β α β α β 1×106 V V V V V V V V

1×105 α β β 5 ε δ γ ε ε δ α β γ ε ε δ α β γ ε ε δ α γ ε -5 0 5 10 15 20 25 30 1×10 c -5 0 5 10 15 20 25 30 c c c c c c c Study Day Study Day 5 ζ ζ ζ ζ ζ ζ ζ ζ Tumor models were established by i.v. or s.c implantation of either NALM-6-luc or Raji-luc (5x10 cells per mouse) cells in 7-week-old female NSG mice. T cells were injected 6 days after tumor implantation at Features: doses ranging from 1x105 to 1x107 (TRuC+ T cells) . Tumor growth was monitored by IVIS imaging. - Integration of the complete TCR - Targeting of tumor surface antigens Figure 3. Faster Tumor Regression for TC-110 versus CD19 CAR in SC Raji - No HLA restriction SC Raji 3000 - No need for additional signaling domains Vehicle 2500 NT CD19 TRuC TC-110 2000 CD19 41BBz CAR CD19 CD28 CAR 1500

1000

500 Mean tumor volume (mm3) 0 Blood was collected on Days 1, 2, 4, 7 and 10 post treatment with TC-110 or CD19 BBz CAR T cells in mice bearing IV NALM-6. Serum was assayed for cytokine levels using Luminex (top panel). For TC-110- -500 CD19 CAR Constructs for Comparative Studies 0 10 20 30 40 treated mice, IFNg, TNFa, IL-13 and GM-CSF levels were highest on day 1, followed by rapid decline to Study Day CD19 41BBz CAR baseline. IL-4 was high in CD19 BBz CAR-treated mice on day 4. For both TC-110 and CD19 CAR, highest Tumor Size on Days 0, 6, 13, 16 and 27 post TC-110 or CD19 CAR T Cell Treatment levels of cytokines correlated with times of active tumor regression (days 1-4) and cytokine levels were lowest at time of complete tumor regression (days 8-10). Day 0 Day 6 Day 13 Data 16 Day 27 ) CD19 CD28 CAR 3 1500 1500 1500 1500 1500 Conclusions 1000 1000 1000 1000 1000 • TC-110, a CD19-targeting TRuC T cell therapy, demonstrated potent in vivo efficacy 500 500 500 500 500

Tumor Volume (mm Volume Tumor in xenogeneic models of acute lymphoblastic leukemia (NALM-6) and Burkitt’s 0 0 0 0 0 lymphoma (Raji). NT NT NT NT NT TC110 TC110 TC110 TC110 TC110 Collaborators Vehicle Vehicle Vehicle Vehicle Vehicle • Tumor-bearing mice had faster tumor regression after treatment with TC-110

CD19 CD28 CAR CD19 CD28 CAR CD19 CD28 CAR CD19 CD28 CAR CD19 CD28 CAR CD19 41BBz CAR CD19 41BBz CAR CD19 41BBz CAR CD19 41BBz CAR CD19 41BBz CAR compared to treatment with CD19 BBz CAR and CD19 CD28 CAR in both NALM-6 NSG mice were implanted SC with Raji (5x105 cells per mouse). Six days post tumor implant, mice were and Raji tumor models. injected IV with TC-110 (1x106 TRuC+ cells per mouse) or CD19 BBz CAR or CD28 CAR (1x106 CAR+ cells • Cytokine release declined rapidly after TC-110 treatment and correlated with time per mouse). Tumor growth was monitored by caliper measurement of SC tumors. Tumors regressed of active tumor regression. more quickly after TC-110 treatment (top panel). Tumors completely regressed by day 13 for TC-110– treated mice (lower panel). In contrast, complete tumor regression occurred by day 27 for CD19 BBz CAR • These results support clinical development of TC-110 for treatment of hematologic and CD28 CAR-treated mice. malignancies.