Abstract
Currently, the lineage-specific cell-surface molecules CD19 and CD20 present on many B-cell malignancies are targets for both antibody- and cell-based therapies. Coupling these two treatment modalities is predicted to improve the antitumor effect, particularly for tumors resistant to single-agent biotherapies. This can be shown using an immunocytokine, composed of a CD20-specific monoclonal antibody fused to biologically active interleukin 2 (IL-2), combined with ex vivo expanded human umbilical cord blood–derived CD8+ T cells, that have been genetically modified to be CD19 specific, for adoptive transfer after allogeneic hematopoietic stem-cell transplantation. We show that a benefit of targeted delivery of recombinant IL-2 by the immunocytokine to the CD19+CD20+ tumor microenvironment is improved in vivo persistence of the CD19-specific T cells, and this results in an augmented cell-mediated antitumor effect. Phase I trials are under way using anti-CD20-IL-2 immunocytokine and CD19-specific T cells as monotherapies, and our results warrant clinical trials using combination of these two immunotherapies. [Cancer Res 2007;67(6):2872–80]
Introduction
Malignant B cells express a pattern of cell surface molecules that define their lineage commitment (1–3), and these are the targets of monoclonal antibody (mAb)–based (4–6) and T-cell–based treatment approaches (7–11). However, these immunotherapies may fail to eradicate tumor as a result of an inability of tumor-specific mAb to fully activate the effector functions of the recipient (12–15) and curtailed T-cell persistence after adoptive immunotherapy (16–18). Therefore, strategies that augment mAb function and T-cell survival are predicted to improve the therapeutic effect.
An approach to improve the clinical potential of mAb is to fuse interleukin 2 (IL-2) to a tumor-specific recombinant mAb (e.g., CD20-specific mAb) to deliver this immunostimulatory cytokine to the tumor microenvironment, which leads to recruitment and activation of immune cells that express the cytokine receptor (19–21). Ex vivo propagated genetically modified T cells that have been rendered tumor specific are a population of effector cells whose survival is predicted to benefit from this locoregional deposition of IL-2.
To obtain large numbers of clinical-grade, tumor-specific T cells that target B-lineage lymphoma and leukemia, we and others have enforced expression of a CD19-specific chimeric immunoreceptor (designated CD19R), which combines antibody recognition with T-cell effector functions (7, 10, 22). In particular, CD19-specific T cells can be manufactured from umbilical cord blood to augment the graft-versus-tumor effect after allogeneic hematopoietic stem-cell transplantation (23). However, factors that may limit the successful therapeutic use of these ex vivo expanded CD8+ T cells include a dependence on exogenous IL-2 to achieve and sustain their proliferative potential after adoptive transfer (24).
With the generation of an anti-CD20-IL-2 immunocytokine (DI-Leu16-IL-2; ref. 25), we now ask if targeted delivery of IL-2 to sites of CD20 binding on malignant B cells could improve the survival and antitumor effect of CD19-specific T cells. In the present study, we show that the anti-CD20-IL-2 immunocytokine binds specifically to CD20+ tumors as well as IL-2R+ (IL-2 receptor positive) T cells and that infusing a combination of anti-CD20-IL-2 immunocytokine with CD19R+ T cells improves in vivo T-cell persistence, which leads to an augmented clearance of CD20+CD19+ tumor beyond that achieved by delivery of the immunocytokine or T cells alone.
Materials and Methods
Plasmid expression vectors. The plasmid vector CD19R/ffLucHyTK-pMG, described previously, coexpresses the CD19R chimeric immunoreceptor gene and the tripartite fusion gene ffLucHyTK (22). Truncated CD19, lacking the cytoplasmic domain (26), was expressed in ffLucHyTK-pMG to generate the plasmid tCD19/ffLucHyTK-pMG to coexpress the CD19 and ffLucHyTK transgenes. The bifunctional hRLucZeo fusion gene that coexpresses the Renilla koellikeri (sea pansy) luciferase hRLuc and the zeomycin-resistance gene (Zeo) was cloned from the plasmid pMOD-LucSh (InvivoGen, San Diego, CA) into pcDNA3.1+ (Invitrogen, Carlsbad, CA) to create the plasmid hRLuc:Zeocin-pcDNA3.1.
Propagation of cell lines and primary human T cells. Daudi, ARH-77, Raji, SUP-B15, and K562 cells were obtained from American Type Culture Collection (Manassas, VA). Granta-519 cells were obtained from DSMZ (Braunschweig, Germany). An EBV-transformed lymphoblastoid cell line was kindly provided by Drs. Phillip Greenberg and Stanley Riddell (Fred Hutchinson Cancer Research Center, Seattle, WA). These cells were maintained in tissue culture as described (23). IL-2Rβ+ TF-1β cells were kindly provided by Dr. Paul M Sondel (University of Wisconsin, Madison, WI; ref. 27). Human T-cell lines were derived from umbilical cord blood mononuclear cells after informed consent and were cultured as previously described (22, 28).
Immunocytokines. The anti-CD20-IL-2 (DI-Leu16-IL-2) immunocytokine was derived from a deimmunized anti-CD20 murine mAb (Leu16). Anti-GD2-IL-2 (14.18-IL-2), which recognized GD2 disialoganglioside, served as a control immunocytokine with irrelevant specificity for B-lineage tumor line used in this study (EMD Lexigen Research Center, Billerica, MA; ref. 29).
Nonviral gene transfer of DNA plasmid vectors. OKT3-activated umbilical cord blood–derived T cells were genetically modified by electroporation with CD19R/ffLucHyTK-pMG (23). ARH-77 was electroporated with hRLuc:Zeocin-pcDNA3.1 using the Multiporator device (250V/40 μs, Eppendorf, Hamburg, Germany) and propagated in cytocidal concentration (0.2 mg/mL) of zeocin (InvivoGen).
Flow cytometry. FITC- or phycoerythrin-conjugated reagents were obtained from BD Biosciences (San Jose, CA): anti-TCRαβ, anti-CD3, anti-CD4, anti-CD8, anti-CD25, and anti-CD122. F(ab′)2 fragment of FITC-conjugated goat anti-human Fcγ (Jackson Immunoresearch, West Grove, PA) was used at 1/20 dilution to detect cell surface expression of CD19R transgene. Leu16 and anti-CD20-IL-2 immunocytokine (100 μg each) were conjugated to Alexa Fluor 647 (Molecular Probes, Eugene OR). Data acquisition was on a FACSCalibur (BD Biosciences) using CellQuest version 3.3 (BD Biosciences), and analysis was undertaken using FCS Express version 3.00.007 (Thornhill, Ontario, Canada).
Chromium release assay. The cytolytic activity of T cells was determined by 4-h chromium release assay (CRA; ref. 22). CD19-specific T cells were incubated with 5×103 chromium-labeled target cells in a V-bottomed 96-well plate (Costar, Cambridge, MA). The percentage of specific cytolysis was calculated from the release of 51Cr, as described earlier, using a TopCount NXT (Perkin-Elmer Life and Analytical Sciences, Inc., Boston, MA). Data are reported as mean ± SD.
Immunofluorescence microscopy. CD19R+ T cells (106) and CD19+CD20+tumor cells (106) were centrifuged at 200 × g for 1 min and incubated at 37°C for 30 min. After gentle resuspension, the cells were sedimented, the supernatant was removed, and the pellet was fixed for 20 min with 3% parafomaldehyde in PBS on ice. After washing, the fixed T cell–tumor cell conjugates were incubated for 30 min at 4°C with anti-CD3-FITC or Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine. Nuclei were counterstained with Hoechst 33342 (Molecular Probes; 0.1 μg/mL). Cells were examined on a Zeiss LSM 510 META NLO Axiovert 200M inverted microscope. Hoechst 33342 was excited at 750 nm using Coherent Ti:Sapphire multiphoton laser, Alexa Fluor 647 at 633 nm using helium-neon laser, and FITC at 488 nm using argon ion laser. Images were acquired with a Zeiss plan-neofluar 20×/0.5 air lens or plan neofluar 40×/1.3 numerical aperture oil immersion lens, and fields of view were then examined using Zeiss LSM Image Browser version 3.5.0.223.
Persistence of adoptively transferred T cells. Before the initiation of the experiment, 6- to 10-week-old female NOD/scid (NOD/LtSz-Prkdcscid/J) mice (The Jackson Laboratory, Bar Harbor, ME) were γ-irradiated to 2.5 Gy using an external 137Cs source (JL Shepherd Mark I Irradiator, San Fernando, CA) and maintained under pathogen-free conditions at City of Hope National Medical Center (COH) Animal Resources Center. On day −7, the mice were injected in the peritoneum with 2×106 hRLuc+ CD19+CD20+ARH-77 cells. Tumor engraftment was evaluated by biophotonic imaging (see “Biophotonic imaging”) and mice with progressively growing tumors were segregated into four treatment groups to receive 107 CD19-specific T cells (day 0) either alone or in combination with 75,000 units/injection (equivalent to ∼25 μg immunocytokine; ref. 25) of IL-2 (Chiron, Emeryville, CA), 5 μg/injection of anti-CD20-IL-2 immunocytokine (DI-Leu16-IL-2), or 5 μg/injection of anti-GD2-IL-2 immunocytokine, given by additional separate i.p. injections. Animal experiments were approved by COH institutional committees.
In vivo efficacy of combination immunotherapies. Six- to 10-week-old γ-irradiated NOD/scid mice were injected with 2×106 hRLuc+ CD19+CD20+ARH-77 cells in the peritoneum. Sustained tumor engraftment was documented within 7 days of injection by biophotonic imaging. Mice in the four treatment groups received combinations of CD19-specific T cells (107 cells in the peritoneum on day 0), anti-CD20-IL-2 immunocytokine, or anti-GD2-IL-2 immunocytokine (5 μg/injection in the peritoneum).
Biophotonic imaging. Anesthetized mice were imaged using a Xenogen IVIS 100 series system as previously described (30). Briefly, each animal was serially imaged in an anterior-posterior orientation at the same relative time point after 100 μL (0.068 mg/mouse) of freshly diluted Enduren Live Cell Substrate (Promega, Madison, WI), or 150 μL (4.29 mg/mouse) of freshly thawed d-luciferin potassium salt (Xenogen, Alameda, CA) solution injection. Photons were quantified using the software program Living Image (Xenogen). Statistical analysis of the photon flux at the end of the experiment was accomplished by comparing area under the curve using two-sided Wilcoxon rank sum test. Biological T-cell half-life was calculated as A = I × (1/2)(t/h), where A is flux at time t, I is day 0 flux, and h is rate of decay.
Results
Redirecting T cells specificity for CD19. The genetic modification of umbilical cord blood–derived T cells to render them specific for CD19 was accomplished by nonviral electrotransfer of a DNA expression plasmid designated CD19R/ffLucHyTK-pMG, which codes for the CD19R transgene (22) and a recombinant multifunction fusion gene that combines firefly luciferase (ffLuc), hygromycin phosphotransferase, and herpes virus thymidine kinase (HyTK; ref. 31), permitting in vitro selection of CD19R+ T cells with cytocidal concentration of hygromycin B and in vivo imaging after infusion of d-luciferin. Genetically modified ex vivo expanded T cells were CD8+, expressed components of the high-affinity IL-2R and CD19R transgene, as detected by using a Fc-specific antibody (Fig. 1A). CD19R+ T cells could specifically lyse leukemia and lymphoma targets expressing CD19 with ∼50% to 70% of CD19+ tumor cells killed at an effector to target ratio of 50:1 in a 4 h CRA (Fig. 1B). The variability of lysis of the various B-cell lines could be attributed to the expression of various cell surface markers, particularly the adhesion molecules (22). Specific lysis of CD19+ K562 compared with CD19− K562 cells showed that the killing of CD19+ tumor targets occurred through the chimeric immunoreceptor.
Phenotype and function of genetically modified T cells. A, multivariable flow cytometry showing that the genetically modified T cells are predominantly CD3+TCR+CD8+CD25+CD122+CD132+. Isotype-matched fluorescent mouse mAb or nonspecific goat control antibody was used to establish the negative gates. The percentage of gated+ cells is shown. B, lysis of tumor targets by 4-h CRA. CD19+ B-cell tumor lines are Daudi, ARH-77, SUP-B15, Granta-519, Raji, and genetically modified K562 (CD19+ K562; ref. 30). Background lysis of CD19− (parental) K562 cells is shown as a control for specificity and endogenous NK-T activity. Spontaneous release of each target was ≤9%. Points, mean for triplicate wells at effector to target cell ratios between 50:1 and 1:1; bars,± 1 SD.
Phenotype and function of genetically modified T cells. A, multivariable flow cytometry showing that the genetically modified T cells are predominantly CD3+TCR+CD8+CD25+CD122+CD132+. Isotype-matched fluorescent mouse mAb or nonspecific goat control antibody was used to establish the negative gates. The percentage of gated+ cells is shown. B, lysis of tumor targets by 4-h CRA. CD19+ B-cell tumor lines are Daudi, ARH-77, SUP-B15, Granta-519, Raji, and genetically modified K562 (CD19+ K562; ref. 30). Background lysis of CD19− (parental) K562 cells is shown as a control for specificity and endogenous NK-T activity. Spontaneous release of each target was ≤9%. Points, mean for triplicate wells at effector to target cell ratios between 50:1 and 1:1; bars,± 1 SD.
Binding of anti-CD20-IL-2 immunocytokine. The ability of the anti-CD20-IL-2 immunocytokine to bind to both B-lineage tumors and T cells was examined using flow cytometry and confocal microscopy. This immunocytokine bound to CD20+ ARH-77 but not to CD20− SUP-B15 (data not shown) and K562 cells, consistent with recognition of parental Leu16 mAb for CD20 (Fig. 2A; ref. 32). The anti-CD20-IL-2 immunocytokine, but not parental Leu16 mAb, bound to CD25+ genetically modified T cells and TF-1β, a tumor cell line genetically modified to express CD122 (IL-2Rβ; ref. 27), which is consistent with binding of chimeric IL-2 via the IL-2R (Fig. 2A and data not shown). The greater median fluorescent intensity (MFI) on T cells, compared with TF-1β, is consistent with binding of the immunocytokine to the high-affinity IL-2R. Immunofluorescence confocal microscopy was done to evaluate the localization of immunocytokine on conjugates of CD19-specific T cells and CD20+ tumors. The confocal micrographs showed cell surface labeling of conjugates of tumor and T cells with Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine (red) and T cells labeled with FITC-conjugated anti-CD3 (green). Areas of overlapping binding between deposition of immunocytokine and anti-CD3 is depicted by a yellow color (Fig. 2B). We hypothesize that T cells show colocalization of CD3 and immunocytokine on their surface initially; however, as they form a synapse with the tumor cell, there seems to be a rearrangement of IL-2R on the T cells toward the synapse leading to the presence of yellow signal extending well outside the synapse and leaving a green pocket opposite the synapse. The Alexa Fluor 647–conjugated parental anti-CD20 Leu16 mAb, lacking the chimeric IL-2 domain, binds CD20+ tumors, but not the genetically modified T cells (data not shown). In aggregate, these data show that anti-CD20-IL-2 immunocytokine can bind to CD20 molecules on B-lineage tumors and IL-2R on T cells and that this immunocytokine can be deposited at the interface between tumor and T cells.
Binding of anti-CD20-IL-2 immunocytokine to B cells and T cells. A, flow cytometry analysis of Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine (ICK) binding through CD20 to CD20+CD25− ARH-77 and CD20−CD25− K562 cell lines; and IL-2 receptors to genetically modified CD20−CD25+ T cells and CD20−CD25−CD122+ TF-1β (filled histograms). Unfilled histograms, fluorescence of unstained cells. The percentage of gated+ cells and MFI (in brackets) are indicated. Alexa Fluor 647 emission (668 nm) was revealed in the APC/Cy5 channel (FL-4). B, confocal micrographs of tumor and T-cell conjugates stained with Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine (red) and FITC-conjugated anti-CD3 (green); cell nuclei were counterstained with Hoechst (blue). Yellow, areas that show overlapping binding of immunocytokine and anti-CD3 mAb.
Binding of anti-CD20-IL-2 immunocytokine to B cells and T cells. A, flow cytometry analysis of Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine (ICK) binding through CD20 to CD20+CD25− ARH-77 and CD20−CD25− K562 cell lines; and IL-2 receptors to genetically modified CD20−CD25+ T cells and CD20−CD25−CD122+ TF-1β (filled histograms). Unfilled histograms, fluorescence of unstained cells. The percentage of gated+ cells and MFI (in brackets) are indicated. Alexa Fluor 647 emission (668 nm) was revealed in the APC/Cy5 channel (FL-4). B, confocal micrographs of tumor and T-cell conjugates stained with Alexa Fluor 647–conjugated anti-CD20-IL-2 immunocytokine (red) and FITC-conjugated anti-CD3 (green); cell nuclei were counterstained with Hoechst (blue). Yellow, areas that show overlapping binding of immunocytokine and anti-CD3 mAb.
In vivo T-cell persistence given in combination with immunocytokine. Having determined that the anti-CD20-IL-2 immunocytokine could bind to tumor and T cells, we evaluated whether infusions of anti-CD20-IL-2 immunocytokine could improve the in vivo persistence of adoptively transferred genetically modified CD8+ T cells. To achieve sustained locoregional depositions of the anti-CD20-IL-2 immunocytokine, we chose the tumor line ARH-77 as a target for immunotherapy because this is relatively resistant to killing by anti-CD20–specific mAb (33), and these results were confirmed in vivo in NOD/scid mice using rituximab (data not shown). Initially, a dose of immunocytokine was established that could both improve the in vivo survival of CD8+CD19R+ffLuc+ T cells, compared with adoptive immunotherapy in the absence of immunocytokine, and not statistically alter tumor growth as monotherapy (Fig. 4). We showed that an immunocytokine dose of both 5 and 25 μg could improve the persistence of infused T cells, resulting in a T-cell ffLuc-derived signal detectable above background luminescence measurements (≤106 p/s/cm2/sr) 14 days after adoptive immunotherapy (Fig. 3A). Biological half-life of the infused T cells was determined by calculating the rate of T-cell decay (ffLuc activity) at the end of the experiment and expressed as the number of days required by the cells to achieve half the initial (day 0) flux. Indeed, the biological half-life of the infused T cells was twice as long in mice that received immunocytokine (1.09 days) compared with T cells given alone (0.43 days). As a further indication that infusion of the immunocytokine may enhance the survival of adoptively transferred T cells, we observed an ∼300% (3-fold) increase in the ffLuc-derived signal (day 12) compared with day 11 when the immunocytokine was injected in both the groups. As the relative in vivo T-cell persistence was similar for both of the immunocytokine doses (P = 0.86), we used 5 μg per immunocytokine injection for subsequent experiments, a dose equivalent to ∼15,000 units of human recombinant IL-2 (25).
Effect of immunocytokine on persistence of adoptively transferred T cells. NOD/scid mice (four mice per group) bearing ARH-77 tumors were treated with 107 CD19R+ffLuc+ umbilical cord blood T-cell clone (day 0, open arrow) along with (A) anti-CD20-IL-2 immunocytokine (5 and 25 μg; solid arrows; on days 0, 4, 7, and 11) or no immunocytokine, or (B) anti-CD20-IL-2 immunocytokine/GD2-IL-2 immunocytokine (5 μg/injection) or rhIL-2 (75,000 units/injection) on days 0, 2, 5, 10, 15, 21, and 45 (closed arrows). The persistence of T cells was measured as ffLuc-derived flux from mice and graphed over time (mean flux ± SD is shown in A and B, and flux for individual mice is shown in C). One mouse (red line) was selected from each group for the display of sequential bioluminescence images of T cells in vivo. Comparison (day 83) between groups receiving combination of T cells and anti-CD20-IL-2 immunocytokine and no treatment (*, P = 0.01); T cells and control immunocytokine (anti-GD2-IL-2 immunocytokine; **, P = 0.05); or T cells and IL-2 (***, P = 0.02). Background luminescence (gray area) was defined from mice that were imaged after receiving d-luciferin along with treatment mice, but which did not receive ffLuc+ T cells. In vitro ffLuc activity of the genetically modified T cells was 0.35 ± 0.02 cpm/cell (mean ± SD) compared with 0 ± 0 cpm/cell (mean ± SD) for parental unmodified cells. Supplementary Data contain a movie of the relative in vivo T-cell persistence in the four treatment groups.
Effect of immunocytokine on persistence of adoptively transferred T cells. NOD/scid mice (four mice per group) bearing ARH-77 tumors were treated with 107 CD19R+ffLuc+ umbilical cord blood T-cell clone (day 0, open arrow) along with (A) anti-CD20-IL-2 immunocytokine (5 and 25 μg; solid arrows; on days 0, 4, 7, and 11) or no immunocytokine, or (B) anti-CD20-IL-2 immunocytokine/GD2-IL-2 immunocytokine (5 μg/injection) or rhIL-2 (75,000 units/injection) on days 0, 2, 5, 10, 15, 21, and 45 (closed arrows). The persistence of T cells was measured as ffLuc-derived flux from mice and graphed over time (mean flux ± SD is shown in A and B, and flux for individual mice is shown in C). One mouse (red line) was selected from each group for the display of sequential bioluminescence images of T cells in vivo. Comparison (day 83) between groups receiving combination of T cells and anti-CD20-IL-2 immunocytokine and no treatment (*, P = 0.01); T cells and control immunocytokine (anti-GD2-IL-2 immunocytokine; **, P = 0.05); or T cells and IL-2 (***, P = 0.02). Background luminescence (gray area) was defined from mice that were imaged after receiving d-luciferin along with treatment mice, but which did not receive ffLuc+ T cells. In vitro ffLuc activity of the genetically modified T cells was 0.35 ± 0.02 cpm/cell (mean ± SD) compared with 0 ± 0 cpm/cell (mean ± SD) for parental unmodified cells. Supplementary Data contain a movie of the relative in vivo T-cell persistence in the four treatment groups.
To determine if the improved T-cell persistence was due to the binding of the immunocytokine in the ARH-77 tumor microenvironment, we used a control immunocytokine (anti-GD2-IL-2 immunocytokine) that does not bind to GD2− ARH-77 (data not shown). Furthermore, we compared the ability of the anti-CD20-IL-2 immunocytokine to potentiate T-cell survival compared with administration of exogenous recombinant human IL-2. Longitudinal measurement of ffLuc-derived flux revealed that the infused T cells persisted longer in mice that received anti-CD20-IL-2 immunocytokine, compared with the untreated (P = 0.01), IL-2–treated (P = 0.02), and control immunocytokine-treated (P = 0.05) groups (Fig. 3B and C); the biological half-lives of T cells in the groups are 1.7, 0.5, 1.0, and 0.7 days, respectively. There was a difference (P < 0.05) in the in vivo persistence of T cells accompanied by IL-2, compared with T cells given without this cytokine, which is consistent with the dependence of these T cells to receive T-cell help in the form of exogenous IL-2 to survive in vivo. No apparent difference was observed in the persistence (P = 0.5) or biological half-life (P = 0.2) of adoptively transferred T cells between the mice receiving exogenous IL-2 or control immunocytokine. These data support the hypothesis that the locoregional deposition of the anti-CD20-IL-2 immunocytokine at the CD19+CD20+ tumor site significantly augments in vivo persistence of CD8+ CD19-specific T cells.
In vivo efficacy of immunocytokine in combination with CD19-specific T cell to treat established B-lineage tumor. We investigated in vivo whether the immunocytokine-mediated improved persistence of genetically modified CD19-specific T cells could lead to augmented clearance of established CD19+CD20+ tumor. A dose of T cells (107 cells) was selected because this dose by itself does not control long-term tumor growth (Fig. 4; data not shown). CD19-specific CD8+ T cells were adoptively transferred into groups of mice bearing established CD19+CD20+hRLuc+ ARH-77 tumor along with anti-CD20-IL-2 immunocytokine or control anti-GD2-IL-2 immunocytokine. Tumor growth was serially monitored by in vivo bioluminescence imaging of ARH-77 tumor-derived hRLuc enzyme activity. Mice that received both CD19-specific T cells and anti-CD20-IL-2 immunocytokine experienced a reduction in tumor growth, with 75% of mice obtaining complete remission, as measured by bioluminescence imaging, at the end of the experiment (50 days after adoptive immunotherapy; Fig. 4). We found that the combination therapy of CD19R+ T cells and anti-CD20-IL-2 immunocytokine was effective in reducing tumor growth compared with no immunotherapy (P = 0.01) and T cells given with an equivalent dosing of the control immunocytokine (P = 0.03). Although the tumor burden seems to be increasing in the treated group, no visible tumor as seen by hRLuc signal was observed at the end of the experiment, as the flux remained below background level, consistent with a complete antitumor response. Mouse groups receiving T cells alone or T cells with control immunocytokine showed a similar pattern of tumor growth, with an initial reduction around day 8 followed by relapse. All mice in the control group, which received no immunotherapy, experienced sustained tumor growth. We saw similar tumor growth kinetics in mice that did or did not receive anti-CD20-IL-2 immunocytokine in the absence of T cells (P > 0.05 through day 50), and this is presumably a reflection of the dose regimen chosen for the immunocytokine in this experiment. Increased doses of T cells or anti-CD20-IL-2 immunocytokine delivered as monotherapies results in a sustained antitumor effect; however, using these doses would preclude our ability to measure the ability of the immunocytokine to potentiate T-cell persistence and improve tumor killing.
Combined antitumor efficacy of immunocytokine and CD19-specific T cells. A, the tumor burden was monitored longitudinally by quantification of ARH-77 tumor-derived hRLuc activity in five groups of NOD/scid mice receiving combinations of T cells (107 on day 0, open arrow); anti-CD20-IL-2 immunocytokine (5 μg/injection); anti-GD2-IL-2 immunocytokine (5 μg/injection) given on days 0, 5, 8, 12, 15, 19, and 22 (closed arrows); and graphed over time as mean flux ± SD. B, serial pseudocolor images representing light intensity from hRLuc ARH-77 cells in selected mice before and after immunotherapy. Comparison (day 50) between groups receiving combination of T cells and anti-CD20-IL-2 immunocytokine and no treatment (*, P = 0.01), and T cells and control immunocytokine (anti-GD2-IL-2 immunocytokine; **, P = 0.03). Genetically modified ARH-77 (transfected with hRLuc:Zeocin-pcDNA3.1) hRLuc activity in vitro was 0.32 ± 0.04 cpm/cell (mean ± SD) compared with 0.004 ± 0.0008 cpm/cell (mean ± SD) for parental unmodified cells. Supplementary Data contain a movie of the relative in vivo antitumor effects of immunotherapy in the five mouse groups.
Combined antitumor efficacy of immunocytokine and CD19-specific T cells. A, the tumor burden was monitored longitudinally by quantification of ARH-77 tumor-derived hRLuc activity in five groups of NOD/scid mice receiving combinations of T cells (107 on day 0, open arrow); anti-CD20-IL-2 immunocytokine (5 μg/injection); anti-GD2-IL-2 immunocytokine (5 μg/injection) given on days 0, 5, 8, 12, 15, 19, and 22 (closed arrows); and graphed over time as mean flux ± SD. B, serial pseudocolor images representing light intensity from hRLuc ARH-77 cells in selected mice before and after immunotherapy. Comparison (day 50) between groups receiving combination of T cells and anti-CD20-IL-2 immunocytokine and no treatment (*, P = 0.01), and T cells and control immunocytokine (anti-GD2-IL-2 immunocytokine; **, P = 0.03). Genetically modified ARH-77 (transfected with hRLuc:Zeocin-pcDNA3.1) hRLuc activity in vitro was 0.32 ± 0.04 cpm/cell (mean ± SD) compared with 0.004 ± 0.0008 cpm/cell (mean ± SD) for parental unmodified cells. Supplementary Data contain a movie of the relative in vivo antitumor effects of immunotherapy in the five mouse groups.
The ability to measure both ffLuc and hRLuc enzyme activities in the same mice allowed us to determine whether the persistence of adoptively transferred T cells directly correlated with tumor size for individual mice. This was accomplished by plotting ffLuc-derived T-cell flux versus hRLuc-derived tumor-cell flux from Fig. 3. Both group of mice, which received CD19-specific T cells along with anti-CD20-IL-2 immunocytokine/anti-GD2-IL-2 immunocytokine, showed a drop in tumor burden at day 8, which is due to the T cells infused. However, the highest numbers of T cells (ffLuc activity; mean flux 4.7×106 versus 1.5×106 p/s/cm2/sr) and lowest tumor burden (hRLuc activity; mean flux 1.4×107 versus 4×107 p/s/cm2/sr) by day 83 (Fig. 5) was observed in the group receiving anti-CD20-IL-2 immunocytokine when compared with the control immunocytokine-treated group. This analysis shows that half of the mice achieved an antitumor response (absence of detectable hRLuc activity) after combination immunotherapy with CD19R+ T cells and anti-CD20-IL-2 immunocytokine. We note that there was continued T-cell persistence (ffLuc activity) in the anti-CD20-IL-2 immunocytokine–treated group compared with the control immunocytokine–treated group (P < 0.05) at day 83. Although tumor burden (hRLuc activity) was reduced in the CD20 immunocytokine– compared with the control immunocytokine–treated group at day 83, no statistical significance was observed. Thus, we note a trend toward continued T-cell persistence and desired antitumor effect in the anti-CD20-IL-2 immunocytokine–treated group.
Measurement of both T-cell persistence and antitumor effect of immunotherapies in individual mice. Mice were treated as in Fig. 3B, and T-cell persistence (ffLuc signal, Y axis) along with tumor burden (hRLuc signal, X axis) was measured in the same mouse at the days mentioned. Improved T-cell persistence (ffLuc signal) and reduced tumor burden (antitumor effect, hRLuc signal) in mice that received a combination of CD19-specific T cells and anti-CD20-IL-2 immunocytokine is observed at day 83. Shaded gray areas, background fluorescence.
Measurement of both T-cell persistence and antitumor effect of immunotherapies in individual mice. Mice were treated as in Fig. 3B, and T-cell persistence (ffLuc signal, Y axis) along with tumor burden (hRLuc signal, X axis) was measured in the same mouse at the days mentioned. Improved T-cell persistence (ffLuc signal) and reduced tumor burden (antitumor effect, hRLuc signal) in mice that received a combination of CD19-specific T cells and anti-CD20-IL-2 immunocytokine is observed at day 83. Shaded gray areas, background fluorescence.
We believe that this is the first time that bioluminescence imaging has been used to connect the persistence of genetically modified T cells to an antitumor effect. These data further reveal that the mice that received the tumor-specific immunocytokine control their tumor burden to a greater extent than the mice that received the control immunocytokine (which does not bind the tumor). As a treatment for minimal residual disease in patients undergoing hematopoietic stem-cell transplantation, this combination therapy shows the ability to keep the disease relapse in check for almost 3 months in this mouse model.
In aggregate, these data show that the combination of anti-CD20-IL-2 immunocytokine and CD19R+ T cells results in augmented control of tumor growth, as predicted from the in vivo T-cell persistence data.
Discussion
We show that anti-CD20-IL-2 immunocytokine specifically binds to CD20+ tumor, that infusions of the anti-CD20-IL-2 immunocytokine can augment persistence of adoptively transferred CD19-specific T cells in vivo, and that this leads to improved control of an established CD19+CD20+ tumor. We believe that these observations are due to the deposition of IL-2 at sites of CD20 binding, which provides a positive survival stimulus to infused CD19R+IL-2R+ effector T cells residing in the tumor microenvironment.
The development of an anti-CD20-IL-2 immunocytokine has implications for future immunotherapy of B-lineage malignancies. Although rituximab has been extensively used to treat CD20+ malignancies (34–36), some patients become unresponsive to this mAb therapy, leading to disease progression (37). The development of an anti-CD20-IL-2 immunocytokine with its ability to activate immune effector cells may rescue these patients, and a clinical trial at COH is under way to determine the safety and feasibility of infusing this immunocytokine. Modifications other than the addition of cytokines (38, 39), such as radionucleotides (40) and cytotoxic agents (41, 42), may also improve the therapeutic potential of unconjugated clinical-grade mAbs. Indeed, combining mAb therapy with therapeutic modalities that exhibit nonoverlapping toxicity profiles is an attractive strategy to improve the antitumor effect without compromising patient safety.
One novel combination therapy for treating B-lineage tumors, described in this report, is to combine immunocytokine with T-cell therapy. The two immunotherapies used, anti-CD20-IL-2 immunocytokine and CD19-specific T cells, have the potential to improve the eradication of tumor because (a) the targeting of different cell surface molecules reduces the possibility emergence of antigen-escape variants, (b) the mAb conjugated to IL-2 can recruit and activate effector cells (such as CD19-specific T cells) expressing the cytokine receptor in the tumor microenvironment, and (c) T cells can kill independent of host factors, which may limit the effectiveness of mAb-mediated complement dependent cytotoxicity and antibody-dependent cell cytotoxicity (12–15). These immunotherapies will target both malignant and normal B cells. However, as loss of normal B-cell function has not been an impediment to rituximab therapy and as clinical conditions associated with hypogammaglobulinemia could be corrected with infusions of exogenous immunoglobulin, a loss of B-cell function may be an acceptable side effect in patients with advanced B-cell leukemias and lymphomas receiving CD19- and/or CD20-directed therapies.
Another potential advantage of immunocytokine therapy is that the locoregional delivery of T-cell help in the form of IL-2 may avoid the systemic toxicities observed with i.v. infusion of the IL-2 cytokine (43–45), and this may be particularly beneficial in the context of allogeneic hematopoietic stem-cell transplantation. We have recently described that umbilical cord blood–derived CD8+ T cells can be rendered specific for CD19 to augment the graft-versus-tumor effect after hematopoietic stem-cell transplantation. Moreover, because the immunocytokine improves the in vivo immunobiology of umbilical cord blood–derived CD19-specific T cells, this study provides the groundwork for combining these two immunotherapies after umbilical cord blood transplantation.
Alternative immunocytokines and T cells with shared specificities for tumor types other than B-lineage malignancies could also be considered for combination immunotherapy. For example, immunocytokines might be combined with T cells that have been rendered specific by the introduction of chimeric immunoreceptors for breast (46, 47), ovarian (48), colon (49), and brain (50) malignancies. Furthermore, immunocytokines bearing other cytokines might be infused with T cells to deliver IL-7, IL-15, or IL-21 to further augment T-cell function in the tumor microenvironment.
In summary, the clinical testing of anti-CD20-IL-2 immunocytokine and CD19R+ T cells as monotherapy will provide Phase I safety and feasibility data. It is anticipated that the data in this report will be used to justify next-generation clinical trials to evaluate combinations of the immunocytokine and T cells.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Acknowledgments
Grant support: CA30206, CA003572, and CA107399; Alliance for Cancer Gene Therapy; Amy Phillips Charitable Foundation; The Leukemia and Lymphoma Society; Lymphoma Research Foundation; National Foundation for Cancer Research; Pediatric Cancer Research Foundation; National Marrow Donor Program; and Marcus Foundation.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Adrian Castro and Vanessa Reeves for assistance with animal experiments, COH Animal Resource Center (under the direction of Dr. Richard Ermel), and the Light Microscopy and Flow Cytometry Cores.