CD137 stimulation enhances the antilymphoma activity of anti-CD20 antibodies

Abstract

Antibody-dependent cell-mediated cytotoxicity (ADCC), which is largely mediated by natural killer (NK) cells, is thought to play an important role in the efficacy of rituximab, an anti-CD20 monoclonal antibody (mAb) used to treat patients with B-cell lymphomas. CD137 is a costimulatory molecule expressed on a variety of immune cells after activation, including NK cells. In the present study, we show that an anti-CD137 agonistic mAb enhances the antilymphoma activity of rituximab by enhancing ADCC. Human NK cells up-regulate CD137 after encountering rituximab-coated tumor B cells, and subsequent stimulation of these NK cells with anti-CD137 mAb enhances rituximab-dependent cytotoxicity against the lymphoma cells. In a syngeneic murine lymphoma model and in a xenotransplanted human lymphoma model, sequential administration of anti-CD20 mAb followed by anti-CD137 mAb had potent antilymphoma activity in vivo. These results support a novel, sequential antibody approach against B-cell malignancies by targeting first the tumor and then the host immune system.

Figure 1

Rituximab induces CD137 up-regulation on human NK cells after incubation with CD20⁺ tumor B cells. Peripheral blood from 3 healthy donors was analyzed for CD137 expression on CD3⁻CD56⁺ NK cells after 24-hour culture with lymphoma cell lines and trastuzumab or rituximab. (A) Percentage of CD137⁺ cells among CD3⁻CD56⁺ NK cells from 3 healthy donors cultured with a CD20⁻ lymphoma cell line (OCI-Ly19) or CD20⁺ lymphoma cell lines (Ramos, DHL-4, Raji). (B) CD20 surface expression on lymphoma cell lines (OCI-Ly19, Ramos, DHL-4, Raji). Histograms were colored according to the log10-fold increase in MFI of lymphoma cell lines relative to isotype. (C) CD137 expression on the NK-cell subsets CD3⁻CD56^bright and CD3⁻CD56^dim from a representative healthy donor after 24-hour culture with the CD20⁺ lymphoma cell line Ramos and rituximab.

Figure 2

Anti-CD137 agonistic mAb increases rituximab-mediated NK-cell cytotoxicity on tumor cells. NK cells isolated and purified from the peripheral blood of healthy donors were analyzed for degranulation by CD107a mobilization after 24-hour culture in 5 conditions: medium alone; CD20⁺ lymphoma cell line (Raji, Ramos, DHL-4); tumor and rituximab; tumor and anti-CD137 antibody; or tumor, rituximab, and anti-CD137 agonistic antibody. (A) Raji, *P = .01. (B) Ramos, *P = .003. (C) DHL-4, *P = .002. A representative flow cytometric plot of CD107a expression with Ramos target cells is shown in supplemental Figure 3. NK-cell cytotoxicity on Raji, Ramos, and DHL-4 tumor cells was analyzed in a chromium-release assay (D-F). Preactivated NK cells (as described in “Methods”) were purified before being incubated with chromium-labeled Raji, Ramos, and DHL-4 cells for 4 hours. (D-F) Percent lysis of target cells by chromium release at varying effector (activated NK cells)-to-target (Raji) cell ratios cultured with medium alone (♦), anti-CD137 (▴), rituximab (●), or rituximab and anti-CD137(■) antibodies. (D) Raji, *P = .01. (E) Ramos, *P = .01. (F) DHL-4, *P = .009.

Figure 3

Anti-CD137 agonistic mAb enhances antilymphoma activity of murine anti-CD20 mAb in vivo. C57BL/6 mice were inoculated with 5 × 10⁶ BL3750 lymphoma tumor cells subcutaneously on the abdomen. (A-B) After tumor inoculation, mice received rat IgG control on day 3 (●), anti-CD20 antibody on day 3 (■), anti-CD137 antibody on day 4 (♦), or anti-CD20 antibody on day 3 and anti-CD137 antibody on day 4 (▴). Mice (10 per group) were then monitored for tumor growth (A; *P < .001) and overall survival (B; *P = .048). (C-D) Tumor growth and survival with identical treatment sequence but with treatment delayed until day 8 after tumor inoculation. Mice received rat IgG control on day 8 (●), anti-CD20 antibody on day 8 (■), anti-CD137 antibody on day 9 (♦), or anti-CD20 antibody on day 8 and anti-CD137 antibody on day 9 (▴). Mice (10 per group) were then monitored for tumor growth (C; *P < .001) and overall survival (D; *P < .001).

Figure 4

Anti-CD20 and anti-CD137 mAbs combination activity is enhanced by sequential mAb administration. C57BL/6 mice were inoculated with 5 × 10⁶ BL3750 lymphoma tumor cells subcutaneously on the abdomen. After tumor inoculation, mice received rat IgG control on day 3 (●), anti-CD20 mAb on day 3 and anti-CD137 mAb on day 4 (▴), or anti-CD20 mAb on day 3 and anti-CD137 mAb on day 3 (■). Mice (10 per group) were then monitored for tumor growth (A; *P < .001) and overall survival (B; * P = .001).

Figure 5

Enhancement of the antilymphoma activity of anti-CD20 mAb by anti-CD137 agonistic mAb is dependent on NK cells and macrophages. (A) Peripheral blood cell subsets from lymphoma-bearing C57BL/6 mice 4 days after tumor inoculation treated on day 3 with either IgG control or anti-CD20 antibody were analyzed for CD137 expression on CD3⁻NK1.1⁺ NK cells (NK), F4/80⁺ macrophages (Mϕ), CD3⁺CD8⁺ T cells (CD8), and CD3⁺CD4⁺ T cells (CD4) (n = 3 mice per group, *P = .001). (B) Tumor-infiltrating lymphocytes from lymphoma-bearing C57BL/6 mice 7 days after tumor inoculation treated on day 3 with either IgG control or anti-CD20 antibody were analyzed for CD137 expression on CD3⁻NK1.1⁺ NK cells (NK), F4/80⁺ macrophages (Mϕ), CD3⁺CD8⁺ T cells (CD8), and CD3⁺CD4⁺ T cells (CD4) (n = 3 mice per group, *P = .012; NS, not significant). (C-D) C57BL/6 mice were inoculated with 5 × 10⁶ BL3750 lymphoma tumor cells. After tumor inoculation, mice received rat IgG control on day 3 (●), anti-Asialo-GM1 on days −1, 0, 5, 10, 15, 20, and 25 with anti-CD20 antibody on day 3 and anti-CD137 antibody on day 4 (■), liposomal clodronate on days −2, 0, 4, 8, 12, 16, 20, and 24 with anti-CD20 antibody on day 3 and anti-CD137 antibody on day 4 (▾), anti-CD8 mAb on days −1, 0, 5, 10,15, 20, and 25 with anti-CD20 antibody on day 3 and anti-CD137 antibody on day 4 (♦), or anti-CD20 antibody on day 3 and anti-CD137 antibody on day 4 (▴). Mice (10 per group) were then monitored for tumor growth (C; *P = .002) and overall survival (D; *P < .001).

Figure 6

Anti-CD137 agonistic mAb enhances antilymphoma activity of rituximab in vivo in a disseminated human lymphoma xenotransplant model. SCID mice were inoculated with 3 × 10⁶ luciferase-labeled Raji lymphoma tumor cells intravenously through the retro-orbital sinus. After tumor inoculation, mice received rat IgG control on day 3 (●), rituximab on day 3 (■), anti-CD137 antibody on day 4 (♦), or rituximab on day 3 and anti-CD137 antibody on day 4 (▴). Treatment was continued weekly for a total of 4 weeks. (A) Luciferase imaging of representative mice 10, 20, and 30 days after treatment are shown. Mice (5 per group) were then monitored for quantified bioluminescence (B; *P = .001) and overall survival (C; *P = .013).

Figure 7

Rituximab-coated, autologous lymphoma cells induce CD137 up-regulation on NK cells from human patients with B-cell malignancies. (A-B) Peripheral blood from patients with B-cell malignancies and CTCs were analyzed for CD137 expression on CD3⁻CD56⁺ NK cells after 24-hour culture with medium alone, trastuzumab, or rituximab. (A) CD16 and CD137 expression on CD3⁻CD56⁺ NK cells for a patient with marginal zone lymphoma (MZL) with 70% CTCs. (B) Percentage of CD137⁺ cells among CD3⁻CD56⁺ NK cells in a cohort of 25 patients with follicular lymphoma (FL), chronic lymphocytic leukemia (CLL), MZL, mantle cell lymphoma (MCL), diffuse large B-cell lymphoma (DLBCL), or CD20⁺ acute lymphoblastic leukemia (ALL). (C) Correlation (r² = 0.87, P < .001) between the percentage of peripheral blood CTCs and CD137 surface expression on CD3⁻CD56⁺ NK cells after 24-hour culture with rituximab from patient samples with FL (◐ ), CLL (●), MZL (◒), DLBCL (◑), or CD20⁺ ALL (◓).