Potent ex vivo armed T cells using recombinant bispecific antibodies for adoptive immunotherapy with reduced cytokine release

Abstract

Background: T cell-based immunotherapies using chimeric antigen receptors (CAR) or bispecific antibodies (BsAb) have produced impressive responses in hematological malignancies. However, major hurdles remained, including cytokine release syndrome, neurotoxicity, on-target off-tumor effects, reliance on autologous T cells, and failure in most solid tumors. BsAb armed T cells offer a safe alternative.

Methods: We generated ex vivo armed T cells (EATs) using IgG-[L]-scFv-platformed BsAb, where the anti-CD3 (huOKT3) scFv was attached to the light chain of a tumor-binding IgG. BsAb density on EAT, in vitro cytotoxicity, cytokine release, in vivo trafficking into tumors, and their antitumor activities were evaluated in multiple cancer cell lines and patient-derived xenograft mouse models. The efficacy of EATs after cryopreservation was studied, and gamma delta (γδ) T cells were investigated as unrelated alternative effector T cells.

Results: The antitumor potency of BsAb armed T cells was substantially improved using the IgG-[L]-scFv BsAb platform. When compared with separate BsAb and T cell injection, EATs released less TNF-α, and infiltrated tumors faster, while achieving robust antitumor responses. The in vivo potency of EAT therapy depended on BsAb dose for arming, EAT cell number per injection, total number of EAT doses, and treatment schedule intensity. The antitumor efficacy of EATs was preserved following cryopreservation, and EATs using γδ T cells were safe and as effective as αβ T cell-EATs.

Conclusions: EATs exerted potent antitumor activities against a broad spectrum of human cancer targets with remarkable safety. The antitumor potency of EATs depended on BsAb dose, cell number and total dose, and schedule. EATs were equally effective after cryopreservation, and the feasibility of third-party γδ-EATs offered an alternative for autologous T cell sources.

Trial registration: ClinicalTrials.gov NCT02173093 NCT00027807 NCT04137536.

Keywords: adoptive; immunotherapy; investigational; neuroblastoma; sarcoma; therapies; translational medical research.

Conflict of interest statement

Competing interests: Both MSK and N-KVC have financial interest in Y-mAbs, Abpro-Labs and Eureka Therapeutics. N-KVC reports receiving commercial research grants from Y-mAbs Therapeutics and Abpro-Labs. N-KVC was named as inventor on multiple patents filed by MSK, including those licensed to Y-mAbs Therapeutics, Biotec Pharmacon, and Abpro-Labs. N-KVC is a SAB member for Abpro-Labs and Eureka Therapeutics. LL is cofounder of Transtarget and is a SAB member for Rapa Therapeutics. JAP has no disclosures to report.

© Author(s) (or their employer(s)) 2021. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.

Figures

Figure 1

Bispecific antibody platform has profound effects on the antitumor activity of bispecific antibody armed T cell immunotherapy. (A) BsAb structural platform. (B) BsAb armed T cells were stained with anti-idiotype antibody, and surface BsAb densities were measured as MFI by flow cytometry. (C) Antibody-dependent T cell-mediated cytotoxicity assay of GD2-BsAb armed T cells at increasing BsAb arming doses. The effector to target cell ratio (E:T ratio) was 10:1. (D) 2×107 of T cells were armed with 2 µg of each BsAb and administered intravenously two doses per week for 2–3 weeks in GD2(+) osteosarcoma PDX and GD2(+) neuroblastoma PDX mouse models. In vivo antitumor responses were compared among groups. (E) Immunohistochemical staining of CD3(+) tumor infiltrating lymphocytes in neuroblastoma PDX tumors treated with a variety of GD2-BsAb armed T cells (on day 10 after the beginning of treatment). (F) Surface BsAb density (MFI) of HER2-BsAb armed T cells. (G) HER2-BsAb armed T cell induced cytotoxicity against osteosarcoma cell line U-2 OS at an E:T ratio of 10:1. (H) HER2-BsAb armed T cells (2 µg of HER2-BsAb/2×107 of T cells) were administered intravenously twice per week for 3 weeks in HER2(+) osteosarcoma PDX-bearing mice. In vivo antitumor effect was compared. **P<0.01, ***P<0.001, ****P<0.0001. BiTE, Bi-specific T-cell engagers; BsAb, bispecific antibodies; GD2, disialoganglioside; HER2, human epidermal growth factor receptor 2; MFI, mean fluorescence intensity; PDX, patient-derived tumor xenograft; scFv, single-chain variable fragment.

Figure 2

Ex vivo arming of T cells with IgG-[L]-scFv-platformed BsAb. (A) Antibody-dependent T cell-mediated cytotoxicity assay of GD2-EATs and HER2-EATs at increasing effector to target ratios (E:T ratios) and at increasing BsAb doses. (B) Surface BsAb density on EAT was analyzed using anti-human IgG Fc-specific antibody and anti-mouse quantum beads. Geometric MFIs of GD2-EATs and HER2-EATs were measured with increasing arming dose of either GD2-BsAb or HER2-BsAb, and BsAb density (MFI) on EAT was referenced to antibody binding capacity (ABC). (C) Comparison of in vitro cytotoxicity between EATs and soluble BsAb with identical E:T ratios, for both anti-GD2 and anti-HER2 systems. BsAb, bispecific antibodies; EATs, ex vivo armed T cells; GD2, disialoganglioside; HER2, human epidermal growth factor receptor 2; MFI, mean fluorescence intensities.

Figure 3

Ex vivo arming of T cells could significantly reduce cytokine release. (A) Th1 cell cytokines (IL-2, IL-6, IL-10, IFN-γ, and TNF-α) released by T cells were measured in each step of arming: (a) in the supernatants after 20 min of incubation with GD2-BsAb (prewash), (b) after the second washing step (postwash), and (c and d) after incubation with target cells (GD2(+) M14 melanoma cell line) at 37℃ for 4 hours. (B) Cytokine release was compared between prewash supernatants (a) and postwash supernatants (b). (C) Cytokine release was compared between GD2-BsAb plus T cells (c), GD2-EATs (d) and unarmed T cells after exposure to target antigen. (D) After intravenous injection of GD2-EATs (armed with 10 µg of GD2-BsAb/2×107 cells) or GD2-BsAb (10 µg) plus unarmed T cells (2×107 cells) into osteosarcoma PDX-bearing mice, serum Th1 cytokine levels were measured at different time points. *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. BsAb, bispecific antibody; EATs, ex vivo armed T cells; GD2, disialoganglioside; IFN, interferon; IL, interleukin; PBS, phosphate-buffered saline; PDX, patient-derived tumor xenograft; Th1, T helper 1; TNF, tumor necrosis factor.

Figure 4

EATs showed faster tumor homing kinetics than BsAb-directed unarmed T cells, rapidly bypassing lung sequestration. Luciferase transduced T cells were expanded and armed with BsAb (Luc(+) GD2-EATs). Luc(+) GD2-EATs (10 µg of GD2-BsAb/2×107 T cells) or Luc(+) unarmed T cells (2×107 cells) with or without GD2-BsAb (10 µg) were administered intravenously into GD2(+) neuroblastoma PDX mice when the average tumor volume reached 100 mm3. (A) Bioluminescence images of GD2-EATs trafficking into tumors over days. (B) Bioluminescence images of GD2-BsAb directed unarmed T cell trafficking into tumors over days. (C) Quantitation of T cell infiltration into tumors over time measured by bioluminescence (n=5 mice/group) expressed as total flux or radiance (photons/s) per pixel integrated over the tumor contour (ROI). (D) Tumor growth curves of individual mice treated with GD2-EATs, GD2-BsAb plus unarmed T cells, or unarmed T cells. To test the in vivo persistence of target antigen-specific EATs, Luc(+) GD2-EATs (armed with 10 µg of GD2-BsAb/2×107 T cells) or Luc(+) HER2-EATs (10 µg of HER2-BsAb/2×107 T cells) were intravenously administered into osteosarcoma TEOS1C PDX-bearing mice. Additional two doses of non-luciferase transduced GD2-EATs or HER2-EATs were administered on day 7 and day 14. (E) Quantitation of bioluminescence of Luc(+) EATs in tumors post-treatment. (F) Tumor growth curves of individual mice treated with GD2-EATs, HER2-EATs, or unarmed T cells, or no treatment. (G) Bioluminescence images of Luc(+) GD2-EATs (upper) or Luc(+) HER2-EATs (lower) in tumors over time. Bioluminescence of Luc(+) EATs was detected over 28 days postinjection. BsAb, bispecific antibodies; EATs, ex vivo armed T cells; GD2, disialoganglioside; HER2, human epidermal growth factor receptor 2; PDX, patient-derived tumor xenograft; ROI, region of interest.

Figure 5

Factors determining the in vivo efficacy of EATs. (A) The effect of BsAb dose on in vivo efficacy of EATs. 2×107 of T cells were armed with increasing dose of GD2-BsAb or HER2-BsAb and their antitumor effects were compared. (B) The effect of EAT cell number on the in vivo efficacy of EATs. At a fixed arming dose of 0.5 µg of BsAb/1×106 cells, increasing numbers of GD2-EAT or HER2-EAT (5×106 cells, 1×107 cells, and 2×107 cells) were administered. (C) The effect of EAT treatment schedule on the in vivo efficacy of EATs. GD2-EATs (2 µg of GD2-BsAb/2×107 cells) were administered as low intensity (1 dose/week in arm 1), standard (2 doses/week in arm 2), or dose-dense (3 doses/week in arm 3) and their in vivo antitumor response and survival were compared. (D) The effect of total number of doses of EATs on long-term tumor control and survival. Arm 1, GD2-EATs (10 µg of GD2-BsAb/2×107 cells) were administered twice weekly for 1 week, and GD2-BsAb (10 µg/dose) were followed twice weekly for 3 weeks; arm 2, four doses of GD2-EATs followed by four doses of GD2-BsAb; arm 3, eight doses of GD2-EATs. *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. BsAb, bispecific antibody; EATs, ex vivo armed T cells; GD2, disialoganglioside; HER2, human epidermal growth factor receptor 2; PDX, patient-derived tumor xenograft; sc, subcutaneous.

Figure 6

Cryopreserved EATs retained target antigen-specific cytotoxicity and exerted a comparable antitumor activity. (A) Geometric mean fluorescence intensities of BsAb on GD2-EATs (0.5 µg of GD2-BsAb/106 cells) and HER2-EATs (0.5 µg of HER2-BsAb/106 cells) before and after cryopreservation. (B) Antibody-dependent T cell-mediated cytotoxicity assay of GD2-EATs and HER2-EATs against GD2(+) and HER2(+) osteosarcoma cell lines before and after cryopreservation (n=6). (C) In vivo antitumor response of thawed GD2-EATs. T cells were first armed with GD2-BsAb (10 µg of BsAb/2×107 cells) before cryopreservation. These cells were then thawed and administered into mice bearing GD2(+) osteosarcoma PDX to compare their in vivo antitumor potency with freshly armed GD2-EATs derived from the same donor. Four to six doses of cryopreserved (thawed) GD2-EATs or freshly armed (fresh) GD2-EATs were administered intravenously twice per week. (D) Relative body weights over time after treatment with fresh or thawed GD2-EATs. (E) Flow cytometry analyses of peripheral blood (PB) T cells in the mice treated with fresh or thawed GD2-EATs on day 35 after treatment. (F) In vivo antitumor response of thawed GD2-EATs or HER2-EATs against telangiectatic osteosarcoma PDXs. T cells were first armed with GD2-BsAb or HER2-BsAb (10 µg of BsAb/2×107 cells) before cryopreservation. Six doses of GD2-EATs or HER2-EATs were given twice a week for 3 weeks with SC interleukin-2. Both thawed EATs significantly suppressed tumor growth without weight loss, improving survival (p<0.0001). ***P<0.001, ****P<0.0001. ATC, activated T cells; EATs, ex vivo armed T cells; E:T, effector to target cell ratio; GD2, disialoganglioside; HER2, human epidermal growth factor receptor 2; PDX, patient-derived tumor xenograft.

Figure 7

EAT as a versatile platform to harness γδ T cells. (A) Flow cytometry of GD2-BsAb or HER2-BsAb on γδ T cells (γδTs) or unselected polyclonal T cells (mostly αβ T cells (αβTs)) after arming with GD2-BsAb or HER2-BsAb. (B) ADTC assays of γδ-GD2-EATs and γδ-HER2-EATs compared with αβ-GD2-EATs and αβ-HER2-EATs. Non-specific tumor cell killing by unarmed γδ T cells and unarmed αβ T cells (background) were subtracted. (C) ADTC assay of γδ-GD2-EATs and γδ-HER2-EATs at increasing effector to target ratios (E:T ratios) and at increasing BsAb arming doses. (D) In vitro cytotoxicity against breast cancer cell line HCC1954 was compared between fresh γδ-HER2-EATs and thawed γδ-HER2-EATs at increasing E:T ratios and at increasing BsAb arming doses. (E) Flow cytometry analyses of peripheral blood T cells after second dose of γδ-GD2-EATs plus zoledronate and supplementary IL-2 or IL-15. (F) Three doses of γδ-GD2-EATs (10 µg of GD2-BsAb/2×107 cells) were administered with supplementary IL-2 or IL-15 to treat osteosarcoma PDXs and the antitumor response was compared among groups. *P<0.05, ***P<0.001. ADTC, antibody-dependent T cell-mediated cytotoxicity; BsAb, bispecific antibodies; EATs, ex vivo armed T cells; γδ T cells, gamma delta T cells; IL, interleukin; MFI, mean fluorescence intensity; PDX, patient-derived tumor xenograft.