`
`Cancer Therapy: Preclinical
`
`Genetically Targeted T Cells Eradicate Systemic Acute
`Lymphoblastic Leukemia Xenografts
`RenierJ. Brentjens,1Elmer Santos,2 Yan Nikhamin,1Raymond Yeh,1Maiko Matsushita,1Krista La Perle,3
`Alfonso Quinta¤ s-Cardama,1Steven M. Larson,2 and Michel Sadelain1,4,5
`
`Abstract Purpose: HumanTcells targeted to the B cell ^ specific CD19 antigen through retroviral-mediated
`transfer of a chimeric antigen receptor (CAR), termed 19z1, have shown significant but partial
`in vivo antitumor efficacy in a severe combined immunodeficient (SCID)-Beige systemic human
`acute lymphoblastic leukemia (NALM-6) tumor model. Here, we investigate the etiologies of
`treatment failure in this model and design approaches to enhance the efficacy of this adoptive
`strategy.
`Experimental Design: A panel of modified CD19-targeted CARs designed to deliver combined
`activating and costimulatory signals to theT cell was generated and tested in vitro to identify an
`optimal second-generation CAR. Antitumor efficacy of Tcells expressing this optimal costimula-
`tory CAR,19-28z, was analyzed in mice bearing systemic costimulatory ligand-deficient NALM-6
`tumors.
`Results: Expression of the 19-28z CAR, containing the signaling domain of the CD28 receptor,
`enhanced systemic T-cell antitumor activity when compared with 19z1 in treated mice. A treat-
`ment schedule of 4 weekly T-cell injections, designed to prolong in vivo T-cell function, further
`improved long-term survival. Bioluminescent imaging of tumor in treated mice failed to identify a
`conserved site of tumor relapse, consistent with successful homing by tumor-specific T cells to
`systemic sites of tumor involvement.
`Conclusions: Both in vivo costimulation and repeated administration enhance eradication of
`systemic tumor by genetically targeted T cells. The finding that modifications in CAR design as
`well asT-cell dosing allowed for the complete eradication of systemic disease affects the design
`of clinical trials using this treatment strategy.
`
`The majority of adult B-cell malignancies, including acute
`lymphoblastic leukemia (ALL), chronic lymphocytic leukemia,
`and non – Hodgkin’s lymphoma, are incurable despite currently
`available therapies. For this reason, novel therapeutic strategies
`are needed to treat
`these diseases. Adoptive therapy with
`genetically engineered autologous T cells is one such approach.
`T cells may be modified to target tumor-associated antigens
`
`Authors’ Affiliations: Departments of 1Medicine and 2Radiology; 3Research
`Animal Resource Center; 4Immunology Program; and 5Gene Transfer and Somatic
`Cell Engineering Laboratory, Memorial Sloan Kettering Cancer Center, New York,
`New York
`Received 3/23/07; revised 5/17/07; accepted 5/23/07.
`Grant support: CA95152, CA59350, CA08748, CA86438, and CA96945;
`The Alliance for Cancer Gene Therapy (M. Sadelain); The Annual Terry Fox Run for
`Cancer Research (New York, NY) organized by the Canada Club of New York,
`William H. Goodwin and Alice Goodwin, and the Commonwealth Cancer
`Foundation for Research and the Experimental Therapeutics Center of Memorial
`Sloan Kettering Cancer Center (R.J. Brentjens and M. Sadelain); Amgen Career
`Development Award (R.J. Brentjens); and the Bocina Cancer Research Fund.
`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.
`Requests for reprints: Renier J. Brentjens, Department of Medicine, Memorial
`Sloan Kettering Cancer Center, Box 242, 1275 York Avenue, New York, NY 10021.
`Phone: 212-639-7053; E-mail: brentjer@mskcc.org.
`F 2007 American Association for Cancer Research.
`doi:10.1158/1078-0432.CCR-07-0674
`
`through the introduction of genes encoding artificial T-cell
`receptors, termed chimeric antigen receptors (CAR), specific to
`such antigens (1 – 5).
`CD19 is an attractive target for immune-mediated therapies
`as it is expressed on most B-cell malignancies and normal B
`cells, but not on bone marrow stem cells. We previously
`constructed a ‘‘first-generation’’ CAR, termed 19z1, specific to
`the CD19 antigen. The 19z1 CAR contains a CD19-specific
`murine single-chain fragment length antibody (scFv) fused to
`the extracellular and transmembrane regions of CD8, which, in
`turn, is fused to the intracellular signaling domain of the CD3 ~
`chain. Human T cells retrovirally transduced to express the
`19z1 receptor specifically lyse heterologous and autologous
`CD19+ human tumor cells in vitro (6). A single i.v. injection of
`19z1+ T cells into SCID-Beige mice bearing established systemic
`Raji tumor, a human Burkitt lymphoma tumor cell line that
`expresses the costimulatory ligands CD80 and CD86, success-
`fully eradicates disease in 50% of mice (6). Unfortunately, a
`similar therapy fails to fully eradicate systemic NALM-6 tumor,
`a human pre-B cell ALL cell line that lacks expression of both
`CD80 and CD86. However, 19z1+ T-cell therapy of SCID-Beige
`mice bearing systemic NALM-6 tumors genetically engineered
`to express CD80 (NALM-6/CD80) enhanced long-term survival
`and resulted in complete NALM-6/CD80 tumor eradication in
`40% of mice (6). Although these data show a role for in vivo
`costimulation in tumor eradication by genetically modified
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`5426
`
`www.aacrjournals.org
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Eradication of Systemic ALL with CD19-Targeted T Cells
`
`elsewhere (8). For ex vivo expansion studies and cytokine release assays,
`transduced T cells were cocultured for 7 days after
`retroviral
`transduction in 24-well tissue culture plates (Falcon, Becton Dickinson)
`with confluent NIH 3T3 AAPCs in RPMI medium supplemented with
`10% FCS, L-glutamine, streptomycin, and penicillin, with no added
`cytokines. For in vivo experiments, transduced T cells were injected after
`for the Pz1+ T cell control,
`expansion on 3T3(CD19/CD80) or,
`3T3[prostate-specific membrane antigen (PSMA)/CD80] AAPCs
`in RPMI medium as above, supplemented with 20 IU IL-2/mL and
`10 ng/mL IL-15. For mice treated with multiple injections of modified
`T cells for 4 weeks, CAR+ T cells were generated by weekly restimulation
`on AAPCs as described above.
`Western blot analysis. Western blot analysis of T-cell lysates under
`reducing conditions with 0.1 mol/L DTT (Sigma) was done as
`previously described (10). Briefly, transduced T cells were washed in
`PBS and resuspended in radioimmunoprecipitation assay buffer
`(Boston Bioproducts) with mini complete protease inhibitor as per
`the manufacturer’s instructions (Roche Diagnostics). Resulting proteins
`were separated on 12% SDS-PAGE mini gels (Bio-Rad) after the
`
`addition of 6 reducing loading buffer (Boston Bioproducts) and
`heating at 100jC for 10 min. Separated proteins were subsequently
`transferred to Immobilon membranes and probed using an anti-human
`CD3 ~ chain monoclonal antibody (BD Biosciences). Antibody binding
`was detected by probing the blot with goat anti-mouse horse radish
`peroxidase – conjugated antibody followed by luminescent detection
`using Western Lighting Chemiluminescence Reagent Plus (Perkin-Elmer
`Life Sciences) as per the manufacturer’s instructions.
`Cytotoxicity assays. We determined the cytotoxic activity of trans-
`duced T cells by standard 51Cr release assays as described elsewhere (8).
`Briefly, transduced T cells were assessed by fluorescence-activated cell
`sorting analysis for CAR expression as well as CD4:CD8 ratio on day 4
`after transduction. NALM-6 tumor cells were labeled with 51Cr for 1 h at
`37jC, washed with RPMI medium supplemented with 10% FCS, and
`resuspended in the same medium at a concentration of 1 105 tumor
`
`cells/mL. Transduced T cells were added to tumor cells at varying
`effector to target cell ratios in 96-well tissue culture plates in a final
`volume of 200 AL, and incubated for 4 h at 37jC. Thereafter, 30 AL of
`supernatant from each well was analyzed using Lumaplate-96 micro-
`plates (Packard Bioscience) by a Top Count NXT microplate scintilla-
`tion counter (Packard Bioscience). Effector cell number in all assays was
`calculated based on the total number of CD8+ CAR+ T cells.
`Cytokine detection assays. Cytokine assays were done per manufac-
`turer’s specifications using the multiplex Human Cytokine Detection
`System (Upstate, Inc.). Luminescence was assessed using the Luminex
`IS100 system and analyzed for cytokine concentration using IS 2.2
`software (Luminex Corp.).
`Flow cytometry. We did flow cytometry using a FACScan cytometer
`with Cellquest software (BD Biosciences). Cells were labeled with either
`phycoerythrin-conjugated, CAR-specific polyclonal goat antibody (Cal-
`tag Laboratories) or phycoerythrin-labeled anti-human CD8 and FITC-
`labeled anti-human CD4 monoclonal antibodies (Caltag Laboratories).
`Retroviral transduction of NALM-6 tumor cells with GFP-FFLuc. The
`GFP-FFLuc gene (Clontech Laboratories) was subcloned into the SFG
`retroviral vector. VSV-G pseudotyped retroviral supernatants derived
`from gpg29 fibroblasts transduced with the resulting SFG (GFP-FFLuc)
`plasmid were used to transduce NALM-6 tumor cells as described
`elsewhere (10). Resulting tumor cells were sorted by fluorescence-
`activated cell sorting for GFP expression.
`In vivo SCID-Beige mouse tumor models. We inoculated 8- to 12-
`week-old FOX CHASE C.B-17 (SCID-Beige) mice (Taconic) with tumor
`cells by tail vein injection. We subsequently treated mice by tail vein
`injection with transduced T cells. In the Raji tumor model, mice were
`
`injected by tail vein with 5 105 Raji tumor cells on day 1, and on day
`6 were treated with a single i.v. dose of 1 107 CAR+ T cells. In the
`vein on day 1 with 1 106 NALM-6 tumor cells, and on days 2 to
`4 were injected i.v. with 1 107 CAR+ T cells daily. In the weekly
`
`NALM-6 upfront treatment tumor model, mice were injected by tail
`
`tumor-targeted T cells, they neither fully explain the etiologies
`of treatment failure in a majority of treated mice nor provide a
`strategy to overcome these limitations.
`To further investigate the in vivo limitations of this adoptive
`T-cell strategy, we chose to pursue the treatment of NALM-6
`tumors in SCID-Beige mice due to the fact that the tumor in
`this model has several features that mimic B-cell ALL disease in
`human subjects: First, the disease is systemic; second, similar to
`the clinical setting, NALM-6 tumor displays an anatomic
`disease pattern that includes involvement of the bone marrow
`and central nervous system (CNS; ref. 7); and third, the NALM-
`6 tumor cell line fails to express costimulatory ligands as do
`most B-cell leukemias, including B-cell ALL. In this report, we
`use this NALM-6 tumor model to address the limitation of
`failed in vivo T-cell costimulation, and further investigate T-cell
`persistence and homing as potential etiologies of treatment
`failure. We found that both in vivo costimulation as well as
`repeated T-cell administration were critical to the complete
`eradication of NALM-6 tumor in SCID-Beige mice. Subsequent
`modifications in our treatment strategy based on these findings
`resulted in a markedly improved rate of complete tumor
`eradication in treated mice.
`To our knowledge, this is the first report demonstrating
`complete eradication of a systemic human tumor lacking
`costimulatory ligands using genetically targeted T cells. Further-
`more, complete eradication is achieved in the absence of further
`in vivo therapy, including prior chemotherapy or subsequent
`cytokine support. These optimized treatment strategies are likely
`to be applicable to future human trials enrolling patients with
`B-cell malignancies, including B-cell ALL.
`
`Materials and Methods
`
`Cell lines and T cells. Raji and NALM-6 tumor cell lines were
`cultured in RPMI 1640 (Life Technologies) supplemented with 10%
`heat-inactivated FCS, nonessential amino acids, HEPES buffer, pyru-
`vate, and BME (Life Technologies). PG-13 and gpg29 retroviral
`producer cell
`lines were cultured in DMEM (Life Technologies)
`supplemented with 10% FCS, and NIH-3T3 artificial antigen-presenting
`cells (AAPC), described previously (6), were cultured in DMEM
`supplemented with 10% heat-inactivated donor calf serum. All media
`were supplemented with 2 mmol/L L-glutamine (Life Technologies),
`100 units/mL penicillin, and 100 Ag/mL streptomycin (Life Technol-
`ogies). Where indicated, medium was supplemented with 10 ng/mL
`interleukin 15 (IL-15; R&D Systems).
`Construction of second-generation CAR fusion genes. Construction of
`the 19z1 and Pz1 scFv-~ chain fusion proteins have been previously
`published (6, 8). The resulting fusion genes were cloned into the SFG
`retroviral vector (9). VSV-G pseudotyped retroviral supernatants derived
`from transduced gpg29 fibroblasts were used to construct stable PG-13
`gibbon ape leukemia virus (GaLV) envelope-pseudotyped retroviral
`producing cell lines using polybrene (Sigma) as described previously
`(8). All second-generation fusion receptors contain the scFv derived
`from 19z1. Human CD28, DAP10, 4-1BB, and OX40 coding regions
`were PCR amplified from a human activated T-cell cDNA library, and
`subcloned into the TopoTA PCR 2.1 cloning vector (Invitrogen). All
`receptor constructs were generated using overlapping PCR. The
`resulting cassettes were designed to facilitate the exchange of the
`transmembrane and signaling domains of the 19z1 construct by NotI/
`BamHI restriction sites encoded in flanking primers.
`Retroviral
`transduction and expansion of human T lymphocytes.
`Retroviral
`transduction of healthy donor T cells, obtained under
`institutional review board – approved protocol 90-095,
`is described
`
`www.aacrjournals.org
`
`5427
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Cancer Therapy: Preclinical
`
`treatment model, mice were inoculated with NALM-6 tumor cells on
`
`day 1, and subsequently treated with i.v. injections of 1 107 CAR+
`
`T cells on days 2, 8, 15, and 22. In all experiments, mice that developed
`hind limb paralysis or decreased response to stimuli were sacrificed by
`CO2 asphyxiation. All murine studies were done in the context of
`an Institutional Animal Care and Use Committee – approved protocol
`(no. 00-05-065).
`In vivo bioluminescence of NALM-6 tumors. Bioluminescence
`imaging was done using Xenogen IVIS Imaging System (Xenogen) with
`Living Image software (Xenogen) for acquisition of imaging data sets.
`injection with 150 mg/kg D-luciferin
`Mice were infused by i.p.
`(Xenogen) suspended in 200 AL PBS. Ten minutes later, mice were
`imaged while under 2% isoflurane anesthesia. Image acquisition was
`done on a 15- or 25-cm field of view at medium binning level for
`0.5- to 3-min exposure time. Both dorsal and ventral views were
`obtained on all animals. Tumor bulk, as determined by IVIS imaging,
`was assessed as described previously (11).
`Histologic analysis of mouse tissue sections. Mouse tissues were fixed
`in 10% buffered formalin phosphate (Fisher Scientific). Osseous
`samples (head with brain, vertebral column with spinal cord, and
`hind limbs) were fixed and decalcified in SurgiPath Decalcifier I
`(SurgiPath Medical Industries) as per the manufacturer’s specifications.
`All tissues were processed by routine methods and embedded in
`paraffin wax. Five-micrometer sections were stained with H&E (Poly
`Scientific).
`Statistics. Statistical analysis of survival data by log-rank analysis
`was obtained using GB-STAT software (Dynamic Microsystems).
`
`Results
`
`Construction of second-generation costimulatory CARs. We
`have previously shown that T cells which express the first-
`generation 19z1 CAR successfully eradicate systemic CD80/
`CD86+ Raji tumor in SCID-Beige mice. However, in the same
`report, we further show the inability of 19z1+ T cells to fully
`eradicate systemic NALM-6 tumors, which fail to express the
`costimulatory CD80 and CD86 ligands. The genetic modifica-
`tion of NALM-6 tumors to express CD80 allowed for the
`complete eradication of tumor in a significant number of
`treated mice (6) consistent with the notion that in vivo T-cell
`costimulation enhances the antitumor efficacy of
`tumor-
`specific T cells.
`Because most B-cell tumors fail to express costimulatory
`ligands, we addressed this limitation of our treatment strategy
`by constructing a series of second-generation CARs designed to
`deliver an additional costimulatory signal in the absence of
`exogenous costimulatory ligand by inserting the transmem-
`brane and cytoplasmic signaling domains of the CD28, DAP10,
`and 4-1BB costimulatory receptors into the 19z1 CAR (Fig. 1A).
`Alternative DAP-10 – and 4-1BB – containing receptors, as
`well as the OX-40 – containing CAR, were designed to contain
`the CD8 transmembrane domain from the original 19z1
`
`Fig. 1. T-cell expression of second-generation CARs. A, pictorial representations of 19z1with second-generation CAR genes demonstrating the genetic fragments used to
`generate the costimulatory CARs. B, Western blot analysis of Tcells retrovirally transduced with CAR genes under reducing conditions. Membranes were probed with a
`monoclonal antibody specific to the cytoplasmic domain of the human ~ chain and show the expression of CARs at the expected molecular weights. The nativeT-cell ~ chain
`is indicated. Lane 1, 19z1; lane 2, 19-28z; lane 3, 19(DAPic)z; lane 4, 19z(DAPic); lane 5, 19(DAP)z; lane 6, 19(4-1BBic)z; lane 7, 19(4-1BB)z; and lane 8, 19(OX-40ic)z.
`C, NALM-6 tumor lysis by CAR+ Tcells was assessed by standard 4 h 51Cr release assays. All Tcells expressing CD19-targeted second-generation CARs lysed target tumor
`cells equally well when compared withTcells expressing the first-generation 19z1CAR. Control Tcells expressing the irrelevant Pz1CAR did not significantly lyse NALM-6
`tumors. Effector to target ratios (E:T Ratio) representTcells normalized to the CD8+ CAR+ T-cell fraction. Data represent one of three different experiments using three different
`healthy donorT-cell populations with similar results.
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`5428
`
`www.aacrjournals.org
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Eradication of Systemic ALL with CD19-Targeted T Cells
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Fig. 2. In vitro T-cell costimulation of CAR+ Tcells as assessed by proliferation and cytokine secretion after coculture on NIH 3T3 fibroblast AAPCs. A, CAR-transduced
`Tcells, normalized to the CAR+ T-cell fraction, were cocultured on AAPC monolayers either without exogenous costimulation [3T3(CD19); left], or with exogenous
`costimulation [3T3(CD19/CD80); right]. On day 1, 5 105 CAR+ Tcells were cocultured on AAPC monolayers in 24-well tissue culture plates in cytokine-free medium.
`On days 4 and 7, total viableT-cell counts were obtained by trypan blue exclusion assays, and FACS analysis was done on these samples to calculate the total number of
`CAR+ Tcells.Whereas 19-28z, 19(4-1BB)z, and 19(4-1BBic)z transduced Tcells expanded by day 4, only 19-28z+ Tcells continued to expand in the absence of exogenous
`CD80 costimulation (left). Control Pz1+ Tcells failed to expand in either setting. B and C, equal numbers of CAR+ Tcells were incubated in the absence of additional cytokines
`AAPC monolayers. Cell-free tissue culture supernatants were analyzed at 24 h after coculture for the presence of cytokines: lane 1, 19z1; lane 2, 19-28z; lane 3, Pz1;
`lane 4, 19(DAP)z; lane 5, 19(DAPic)z; lane 6, 19z(DAPic); lane 7, 19(OX-40ic)z; lane 8, 19(4-1BB)z; and lane 9, 19(4-1BBic)z. OnlyTcells that expressed the 19-28z CAR
`secreted significant levels of IL-2 (B) and IFNg (C) in the absence of exogenous CD80.These data are representative of one of three different experiments using three different
`healthy donorT-cell populations, all of which showed the same proliferation and cytokine secretion patterns.
`
`construct to assess whether this CAR design could enhance
`costimulatory signaling. Finally, we also assessed whether
`placing the DAP-10 signaling domain either proximal or distal
`to the ~ chain signaling domain improved CAR function.
`T cells from healthy human donors were retrovirally
`transduced to express these CAR constructs. T-cell transduction
`efficiency, as assessed by flow cytometric analysis, ranged from
`50% to 80% (data not shown). Western blot analysis of
`transduced T cells, normalized to the CAR+ fraction, showed
`comparable levels of CAR proteins present at the predicted
`molecular weights (Fig. 1B).
`To verify that second-generation CARs when expressed in
`human T cells retained the ability to lyse CD19+ tumor cells
`
`in vitro, we conducted standard 4 h 51Cr release assays using
`transduced healthy donor T cells targeting 51Cr-labeled CD19+
`NALM-6 tumor cells (Fig. 1C). Effector to target ratio between
`different CAR constructs was normalized to the CD8+ CAR+
`T-cell population. All tested CARs, with the exception of the
`Pz1+ T-cell control, specific to the PSMA, were able to mediate
`tumor cell lysis equally well when compared with the first-
`generation 19z1 CAR.
`Characterization of second-generation CAR costimulatory
`function. We have previously generated a series of AAPCs
`derived from NIH-3T3 murine fibroblasts genetically engi-
`neered to express either CD19 alone [3T3(CD19)] or both
`CD19 and CD80 [3T3(CD19/CD80); ref. 6]. When cocultured
`
`www.aacrjournals.org
`
`5429
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Cancer Therapy: Preclinical
`
`with CD19-specific CAR+ T cells, the former [3T3(CD19)] pro-
`vides T cells with a CAR-mediated activation signal (signal 1)
`alone, whereas the latter [3T3(CD19/CD80)] provides T cells
`with signal 1 as well as an exogenous costimulatory signal
`(signal 2) mediated through CD80 on the AAPC binding to
`CD28 on the T cell.
`transduced
`To assess CAR-mediated T-cell proliferation,
`T cells were cocultured on both 3T3(CD19) and 3T3(CD19/
`CD80) AAPCs in the absence of exogenous cytokine. CAR+
`T-cell number was determined after coculture on 3T3(CD19)
`and 3T3(CD19/CD80) AAPCs (Fig. 2A) on days 4 and 7.
`Following coculture on 3T3(CD19) AAPCs, only T cells
`transduced with the 19-28z, 19(4-1BB)z, and 19(4-1BBic)z
`CARs expanded after 4 days of coculture, whereas at day 7, only
`T cells transduced with the 19-28z CAR had undergone
`significant proliferation (10-fold expansion) consistent with
`CAR-mediated T-cell costimulation. All other constructs failed
`to promote T-cell proliferation in the absence of exogenous
`costimulation. Significantly, 19-28z CAR+ T cells proliferated
`equally well on either 3T3(CD19) or 3T3(CD19/CD80) AAPCs,
`and proliferated equally well when compared with 19z1+ T cells
`in the setting of 3T3(CD19/CD80) AAPC stimulation.
`We next assessed CAR+ T-cell secretion of IL-2 and IFN-g as
`surrogate makers of costimulation. Equal numbers of CAR+
`T cells with similar CD4:CD8 ratios were cocultured on
`3T3(CD19) and 3T3(CD19/CD80) AAPCs in cytokine-free
`medium after initial T-cell transduction. At 24 h after T-cell
`coculture on AAPCs, tissue culture supernatants were analyzed
`for the presence of IL-2 and IFN-g. Coculture of 19-28z+ T cells
`on 3T3(CD19) AAPCs resulted in significantly elevated levels of
`IL-2 (Fig. 2B) and IFN-g (Fig. 2C) when compared with the
`first-generation 19z1 CAR as well as the other six tested second-
`generation CARs. These data further confirm the ability of the
`19-28z CAR to elicit a costimulatory signal independent of
`exogenous costimulatory ligand. Significantly, 4-1BB and DAP-
`10 constructs containing either the native or CD8 transmem-
`brane domains showed no significant differences in function,
`
`and in the setting of the DAP-10 constructs, no differences in
`function were observed when the signaling domain was placed
`either proximal or distal to the ~ chain.
`In vivo antitumor activity of alternative costimulatory
`CARs. To define the in vivo costimulatory activity of the
`19-28z CAR, we initially compared the antitumor activity of
`19z1+ and 19-28z+ T cells in a previously established systemic
`Raji tumor model. Because both 19z1+ T cells and 19-28z+
`T cells are costimulated in vivo by CD80/CD86+ Raji tumor
`cells, we predicted and observed equal long-term survival (50%)
`in both treatment groups (Fig. 3A). We next compared the same
`treatment groups (19z1 versus 19-28z) in mice bearing systemic
`NALM-6 tumor, which fails to express the CD80 and CD86
`costimulatory ligands. We observed improved antitumor
`efficacy in the 19-28z treatment group (Fig. 3B). However, the
`improved long-term overall survival was modest (0 of 15 19z1+
`T cell – treated mice versus 3 of 17 or 18% of 19-28z+ T cell –
`treated mice; P < 0.03). Treatment with 19(4-1BB)z+ and
`19(4-1BBic)z+ T cells enhanced survival when compared with
`mice treated with the control Pz1+ T cells, but overall survival
`was similar to 19z1+ T cell – treated mice, with no long-term
`surviving mice (data not shown). Fluorescence-activated cell
`sorting analysis of single-cell suspensions of tissues derived
`from mice that failed 19-28z+ T cell treatment showed persistent
`expression of CD19 on the tumor cells, ruling out the possibility
`that down-regulation of the CD19 target antigen was a source of
`treatment failure (data not shown).
`Repeated administration of CD19-targeted T cells enhances
`complete NALM-6 tumor eradication. Using bioluminescent
`imaging of treated SCID-Beige mice bearing NALM-6(GFP-
`FFLuc) tumors, we found a 10- to 14-day delay of tumor
`progression after 19z1+ T-cell therapy when compared with
`Pz1+ T-cell
`therapy (Fig. 4A-B). A similar delay of tumor
`progression was seen after 19-28z+ T-cell therapy in mice with
`relapsed disease (data not shown). Fluorescence-activated cell
`sorting analysis of single-cell suspensions derived from tissues
`of 19-28z+ T cell – treated tumor-bearing mice confirmed a
`
`Fig. 3. In vivo analysis of CD19-specific CAR+ Tcells in SCID-Beige mice bearing systemic human CD19+ tumors. A, treatment with 19z1+ or 19-28z+ Tcells eradicated
`systemic Raji tumors in SCID-Beige mice equally well. Briefly, mice were injected by tail vein with 5 105 Raji tumor cells on day 1. Mice were subsequently treated with a
`single dose of 1 107 CAR+ Tcells by tail vein injection on day 6. Mice treated with either 19z1+ or 19-28z+ Tcells eradicated tumor equally well with an overall 50% long-term
`survival in both treatment groups. B, treatment of NALM-6 bearing SCID-Beige mice with 19-28z+ Tcells (1 107 CAR+ Tcells injected daily 3 d) statistically enhanced
`survival when compared with similar treatment with 19z1+ Tcells (18% versus 0% long-term survival; P < 0.03).
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`5430
`
`www.aacrjournals.org
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Eradication of Systemic ALL with CD19-Targeted T Cells
`
`Fig. 4. Enhanced antitumor activity through weekly
`CD19-targeted T-cell infusions. A, treatment of
`NALM-6 tumor-bearing mice with daily targeted
`T-cell injections on days 1to 3 after tumor cell
`injection delays tumor growth for 10 to 14 d as
`assessed by analysis of IVIS images of mice treated
`with either 19z1+ Tcells or the control Pz1+ Tcells.
`NALM-6 (GFP-Luc) tumor burden was assessed
`based on ventral mouse images in photons per
`second emitted after luciferin injection. Similar
`data were obtained in mice treated with 19-28z+
`Tcell ^ treated mice (not shown). B, representative
`dorsal and ventral images of NALM-6 (GFP-Luc) ^
`bearing mice obtained weekly afterT-cell therapy
`shows delayed tumor growth in 19z1+ Tcell ^ treated
`mice when compared with the Pz1+ T-cell control
`mice. C, weekly 19-28z+ Tcell tail vein injection for
`4 wks on days 2, 9, 16, and 22, after NALM-6 tumor
`injection on day 1, enhanced long-term survival of
`treated mice >4-fold (44% versus 10%) when
`compared with similar treatment with either 19z1+
`Tcells (P < 0.002) or the control Pz1+ Tcell ^ treated
`mice. Results represent pooled data from two
`separate experiments.
`
`short in vivo T-cell survival (data not shown), consistent with
`previously published data demonstrating limited in vivo
`survival of human T cells in an immunocompromised murine
`host (11 – 14).
`To assess whether enhanced in vivo T-cell persistence
`promoted long-term survival, we generated a sustained in vivo
`population of tumor-targeted T cells in NALM-6 tumor bearing
`
`SCID-Beige mice by injecting CAR+ T cells weekly for 4 weeks.
`The incidence of long-term survival with 19-28z+ T-cell therapy
`increased from 18% as seen in the upfront daily 3 treatment
`
`schedule to 44% in the weekly treatment schedule (Figs. 3B
`and 4C, respectively). Consistent with an additive effect of
`in vivo T-cell persistence and costimulation, we found superior
`antitumor activity of 19-28z+ T cells, when compared with
`
`www.aacrjournals.org
`
`5431
`
`Clin Cancer Res 2007;13(18) September 15, 2007
`
`UPenn Ex. 2041
`Miltenyi v. UPenn
`IPR2022-00855
`
`
`
`Downloaded from http://aacrjournals.org/clincancerres/article-pdf/13/18/5426/1972030/5426.pdf by guest on 17 July 2022
`
`Cancer Therapy: Preclinical
`
`19z1+ T cells, as shown by 44% versus 10% long-term survival
`(P < 0.002), using this weekly T-cell dosing schedule (Fig. 4C).
`CD19-targeted T cells eradicate NALM-6 tumor cells at
`disparate anatomic sites in SCID-Beige mice after adoptive
`T-cell therapy. Despite improved survival with the weekly
`T-cell dosing treatment schedule, 56% of 19-28+ T cell – treated
`mice ultimately failed therapy. To assess whether these mice
`developed persistent or recurrent tumor in consistent anatomic
`sites potentially restricted from T-cell penetration or detection,
`we monitored in vivo NALM-6(GFP-FFLuc) tumor progression
`by weekly bioluminescent (IVIS) imaging in the setting of the
`weekly T-cell treatment regimen. In these studies, all control
`mice treated weekly with Pz1+ T cells showed diffuse tumor
`progression at multiple sites over the treatment period (Fig. 5A).
`By weekly IVIS monitoring of the 19z1+ T cell – treated mice,
`40% (4 of 10) had no detectable disease upon completion of
`therapy (day 28). In the nine mice that developed progressive
`disease in this treatment group, initial disease persistence or
`relapse was noted at distinct anatomic sites in different mice,
`including the bone marrow (n = 2; Fig. 5B), the periodontal
`region (n = 3), the spleen/calvarium (n = 1), the abdomen
`(n = 1), and the CNS/calvarium (n = 1). Overall, 69% (11 of 16)
`of mice treated with weekly 19-28z+ T-cell infusions showed no
`evidence of disease by weekly IVIS imaging upon completion
`
`of therapy. The 56% (9 of 16) of 19-28z+ T cell – treated mice,
`which ultimately failed therapy, likewise showed detectable
`disease at disparate anatomic sites, including the spleen (n = 1;
`Fig. 5C), the CNS/calvarium (n = 2; Fig. 5D), the periodontal
`region (n = 4; Fig. 5E), and the bone marrow (n = 2; Fig. 5F).
`Significantly, persistent disease in the periodontal region,
`despite ongoing T-cell therapy, was noted in 30% of 19z1-
`treated mice, and 25% of 19-28z – treated mice, accounting for
`a large majority of the mice with evidence of disease upon
`completion of treatment. The presence of tumor as assessed
`by bioluminescent imaging in 19-28z+ T cell – treated mice was
`subsequently confirmed by histologic analysis, demonstrating
`NALM-6 tumor infiltration in the bone marrow (Fig. 6A-B), the
`spleen (Fig. 6C-D), and the periodontal region (Fig. 6E-F).
`
`Discussion
`
`The successful clinical application of adoptive therapy of
`cancer with genetically targeted T cells requires both a better
`understanding of the in vivo biology of genetically modified
`T cells, as well as innovative means of enhancing the in vivo
`potency of these T cells. In this report, we use the clinically
`relevant NALM-6 systemic murine model of human B-cell ALL,
`which involves the bone marrow and CNS, and lacks
`
`Fig. 5. Bioluminescent imaging of NALM-6(GFP-
`FFLuc) tumor cells after weekly CAR+ T-cell
`treatment for 4 wks.Weekly treatment with Pz1+
`Tcells in mice bearing systemic NALM-6
`(GFP-FFLuc) resulted in a typical diffuse pattern of
`tumor cell expansion involving the bone marrow,
`spleen, liver, lymph nodes, and periodontal