Cancer Therapy: Preclinical

Augmented Antitumor Activity against B-Cell Lymphoma by a Combination of Monoclonal Antibodies Targeting TRAIL-R1 and CD20 Sreeram Maddipatla,3 FranciscoJ. Hernandez-Ilizaliturri,1, 2 , 3 Joy Knight,1, 2 and Myron S. Czuczman1, 2 , 3

Abstract Purpose: Mapatumumab and are fully humanized, high-affinity immunoglobulin G1Emonoclonalantibodies (mAb) that target/activate the tumornecrosis factor-related apoptosis- inducingligand (TRAIL) receptor1 (TRAIL-R1)and receptor 2 (TRAIL-R2),respectively,triggering the extrinsic apoptoticpathway.Theoretically,synergistic antitumoractivity shouldbe observedby combining TRAIL-R mAbswithagents(e.g.,)that activatetheintrinsicapoptoticpathway. Experimental Design: To this end, targeted antigen expression in a NHL-cell panel was evaluated by flow cytometry. NHL cells were exposed to mapatumumab or lexatumumab followed by rituximab, isotype, or RPMI. DNA synthesis was quantified by [3H]-thymidine incorporation assays. Induction of apoptosis was detected by flow-cytometric analysis. For antibody-dependent cellular cytotoxicity (ADCC) and complement-mediated cytotoxicity (CMC) studies, standardized 51Cr-release assays were done. We inoculated severe combined immunodeficiency (SCID) mouse with Raji cells i.v. The animals then were treated with various combinations of rituximab, mapatumumab, lexatumumab, and isotype alone or in combination. Results: In vitro exposure to mapatumumab resulted in significant apoptosis (30-50%) and decreased DNA synthesis in sensitive lymphoma cells. Mapatumumab/rituximab combination resulted in a significant inhibition of cell proliferation (90% reduction) when compared with mapatumumab (60% reduction) or rituximab (5% reduction). In vivo, the median survival time of animals treated with mapatumumab and rituximab was longer (not reached) than those treated with rituximab monotherapy [33 days (95% confidence interval, 29-37), log-rank test, P =0.05]. Conclusions: Mapatumumab induces apoptosis, cell growth arrest, ADCC, and CMC. The combination of mapatumumab plus rituximab is more effective in controlling lymphoma growth in vivo than either antibody. Rituximab and mapatumumab warrant further evaluation against B-cell lymphoma.

Background cancer cells. In contrast to most traditional chemotherapies, conjugated or unconjugated monoclonal antibodies (mAb) can Targeted therapy was largely developed as a consequence induce significant antitumor activity with improved therapeutic of advances in our understanding of the molecular biology of indices (1). B-cell lymphomas, which identified tumor-associated antigens Rituximab, a chimeric mAb targeting the CD20 antigen and/or key regulatory pathways that govern the development present on B-cell lymphocytes and most of B-cell lymphomas, and maintenance of a given neoplastic process. Targeted is already a universal biological agent used in the treatment of therapies were also developed in response to the challenge B-cell neoplasms. Although combinations of rituximab plus of optimizing antitumor activity while decreasing nonspecific chemotherapeutic agents lead to augmented antitumor activity, toxicity. Monoclonal antibodies against tumor-associated anti- a significant number of patients show limited responses, gens are an attractive form of targeted biological therapy against difficulty tolerating rituximab-chemotherapy combination regimens, and/or relapse post-therapy. There is a dire need to further develop innovative, less toxic combination therapies Authors’ Affiliations: 1Departments of Medicine and 2Immunology, Roswell Park exploring the use of two or more biological agents against B-cell Cancer Institute, and 3School of Medicine, State University of NewYork at Buffalo, lymphomas. Buffalo, New York More than a decade ago, it was shown that the deregulation Received 3/23/07; revised 5/24/07; accepted 6/5/07. of apoptosis is a key step necessary for the development, Grant support: NIHPO1CA103985-1. The costs of publication of this article were defrayed in part by the payment of page maintenance, and progression of various subtypes of B-cell charges. This article must therefore be hereby marked advertisement in accordance lymphomas. In normal B cells, programmed cell death or with 18 U.S.C. Section 1734 solely to indicate this fact. apoptosis is a vital process that is tightly regulated in situ by Requests for reprints: Myron S. Czuczman, Lymphoma/Myeloma Service, complex intracellular and extracellular signals (2). Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263. Phone: 716-845-3221;Fax: 716-845-3894; E-mail: [email protected]. Cytokines such as the (TNF) family F 2007 American Association for Cancer Research. members are an example of extracellular molecules capable of doi:10.1158/1078-0432.CCR-07-0680 regulating apoptosis because they bind to specific cell surface

Clin Cancer Res 2007;13(15) August1,2007 4556 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Mapatumumab and Rituximab Improved Antitumor Activity receptors on the target cells and regulate cellular proliferation TRAIL-R2, respectively; both of these antibodies are currently and differentiation (3). TNF is the classic member of the family undergoing preclinical and clinical evaluation (13, 19). Agonist of cytokines that interact with the corresponding set of mAbs can induce apoptosis in human cancer cell lines in vitro receptors from the TNF receptor (TNF-R) family and function and in vivo in murine xenograft cancer models (12, 20, 21). The as important mediators of immunoregulation and the inflam- theoretical advantages of agonist mAb against DR over TRAIL matory response by regulating programmed cell death via the ligand are both (a) the specificity for only functional TRAIL-Rs extrinsic and intrinsic apoptotic pathways (4, 5). The activation (and not decoy receptors) and (b) their longer half-life as of apoptosis via death receptors (DR) seems to be preserved in compared with soluble TRAIL. various subtypes of B-cell lymphomas, making the targeting Similarly to TRAIL, agonist mAbs mapatumumab and of tumor necrosis factor-related apoptosis-inducing ligand lexatumumab are capable of inducing effective antilymphoma (TRAIL) receptors (TRAIL-R) an appealing strategy in cancer effects in vitro and in vivo (20). Various groups of investigators therapeutics. TRAIL-Rs are part of the TNF-R gene superfamily had shown that the activation of the TRAIL-Rs by either ligands and have a broad range of biological functions like regulation or mAbs sensitize cancer cells to the effects of various chemo- of cell death and survival (6). TRAIL-DRs have a homologous therapeutic and/or biological agents. Combining two biologi- cytoplasmic sequence termed as the death domain, in addition cally active and well-tolerated agents with different mechanisms to a cysteine-rich extracellular domain, which regulates whether of action, such as rituximab and agonist mAbs against DR, is a these receptors activate or prevent apoptosis (7). To date, six potentially attractive treatment strategy for patients with B-cell members of the TNF family receptors with death domains lymphoma. In this report, we study the effects of targeting B-cell have been identified and include TNF-R1 (DR1), Fas (DR2), lymphomas in vitro and in vivo with agonist DR humanized Apo-3 (DR3), TRAIL-1 (DR4), TRAIL-2 (DR5), and DR6. antibodies (mapatumumab/lexatumumab) alone and in com- In addition to TNF, other TNF-R family ligands have been bination with rituximab. identified such as Fas/Apo1L or CD95L (binds to DR CD95), and -a (binds to TNF-R1), or Apo3 ligand or Apo3L or TWEAK (binds to DR3), each capable of inducing Materials and Methods apoptosis in transformed cells and activated lymphocytes. Their therapeutic use has been limited due to their toxic effects on Culture medium and reagents. Cells were maintained in RPMI 1640 (Sigma Chemical) supplemented with 10% heat-inactivated (60jC, normal tissues (i.e., systemic inflammatory response syndrome, in vivo 45 min) fetal bovine serum (FBS, Atlanta Biologicals), 5 mmol/L or hepatotoxicity) (5, 7, 8). HEPES, 100 units/mL penicillin, and 100 Ag/mL of streptomycin A novel member in the TNF family was recently identified (Invitrogen Corp.). Ficoll-Hypaque used to isolate peripheral blood and designated TRAIL or Apo-2L (for Apo-2 ligand; refs. 4, 9). mononuclear cells (PBMC) was purchased from Sigma Chemical. TRAIL is a type II membrane protein resembling Fas/Apo1L in Sodium chromate51 (51Cr; Perkin-Elmer Life Inc.) and [3H]-thymidine amino acid sequence expressed in a variety of human tissues radioisotopes (Perkin-Elmer Life Inc.) were used in functional assays that is capable of inducing apoptosis by binding to its assessing antibody-associated cytotoxicity and cell proliferation, respec- appropriate DR (9). tively. Triton X-100 was purchased from Sigma Chemical. Cell lines. The Raji and Ramos cell lines are well-characterized TRAIL, like several other TNF family members, forms an + + + oligomeric structure that helps to cross-link and trimerize its B-cell lymphoblastic cell lines (phenotype: CD20 , CD19 , CD22 ) derived from a patient with Burkitt’s lymphoma (obtained from the receptors and transduce a signal to the target cell, resulting in American Type Culture Collection). The SU-DHL-4 and SU-DHL-10 cell apoptosis (4, 10). lines were derived from patients with diffuse large-cell B-cell lympho- On the other hand, there are other TRAIL-Rs that lack mas and immortalized by concomitant EBV infection and were a kind functional death domains (TRAIL-R3/DcR1, TRAIL-R4/DcR2, gift from Dr. Steven Treon (Dana-Farber Cancer Institute, Boston, MA). osteoprotegrin/OPG); these are known as decoy receptors, and Experiments were conducted in four rituximab-sensitive cell lines Raji, their function has not been clearly established. Triggering these SU-DHL4, SU-DHL-10, and Ramos cells. decoy receptors may protect cells from signaling via the death- All cell lines were cultured in RPMI 1640 supplemented with 10% domain–containing TRAIL-Rs (5, 7, 11–13). heat-inactivated FBS (Life Technologies BRL), 1% L-glutamine, and j TRAIL induces apoptosis in a wider range of cancer cells penicillin/streptomycin in a humid environment of 37 C and 5% CO2. (1, 13). Upon binding of TRAIL, there is aggregation of Antibodies and reagents. Mapatumumab (TRAIL-R1 mAb) and lexatumumab (TRAIL-R2 mAb) are fully human immunoglobulin TRAIL-Rs on the cell surface, triggering the proximal signaling G1 (IgG1) mAbs and were provided by Human Genome Sciences. of cell death (13). TRAIL-induced apoptosis in cancer cells is Rituximab (IDEC/Genentech Inc.) was obtained from the Roswell Park carried out by the activation of both extrinsic and intrinsic Cancer Institute (RPCI) Pharmacy Department at a stock concentration intracellular death signaling pathways (14). TRAIL promotes of 10 mg/mL. Unless otherwise specified, mAbs were used at a final caspase-8 and caspase-10 activation after binding to appropriate concentration of 10 Ag/mL for in vitro studies. Antibodies were dosed at cognate receptors (TRAIL-R1 and TRAIL-R2), which, in turn, 10 mg/kg and diluted in sterile PBS for tail vein injection into the severe results in the downstream activation of caspase-3, caspase-6, and combined immunodeficiency (SCID) mice for in vivo studies. Trastu- caspase-7 (13). In some of the cancer cell lines, Apo-2L/TRAIL– zumab (Herceptin, Genentech Inc.) was used as an isotype control at the same dose. induced activation of caspase-3 is further augmented through 51 51 the activation of the intrinsic apoptotic pathway (1, 11). In vivo Chromium ( Cr) was used for antibody-dependent cellular cytoto- xicity and complement-mediated cytotoxicity assays. [3H]-Thymidine antitumor activity has been published using various xenograft was used for the assessment of DNA synthesis and cell proliferation. cancer models with TRAIL ligands (12–18). Alternatively, PBMCs were obtained from healthy donors on an Institutional Review TRAIL-R–mediated apoptosis can be triggered by mAbs against Board–approved protocol and isolated by Phycoll centrifugation and TRAIL-R1 or TRAIL-R2. Mapatumumab and lexatumumab are used at an effector/target ratio of 40:1. Pooled human serum was used as two fully human mAbs that target and activate TRAIL-R1 and a source of complement for complement-mediated cytotoxicity (CMC)

www.aacrjournals.org 4557 Clin Cancer Res 2007;13(15) August 1, 2007 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Cancer Therapy: Preclinical assays. Annexin V detection kit (Oncogene Inc.), FITC-conjugated The antiproliferative effects of mapatumumab or lexatumumab alone or mouse anti-human TRAIL-R1 and TRAIL-R2 mAb were obtained from in combination with rituximab on NHL cell lines. Lymphoma cell lines BD Biosciences (clones numbers) and used for flow-cytometric studies. were placed in 96-well plates (1 Â 105 cells per well) and exposed to Animals. For the in vivo experiments, 6- to 8-week-old SCID mice 10 Ag/mL concentrations of mapatumumab (TRAIL-R1 mAb) or were bred and maintained at the Department of Laboratory Animal lexatumumab (TRAIL-R2 mAb), alone or in the presence of rituximab Resources facility at RPCI. The experimental design was approved by the (10 Ag/mL) or isotype (10 Ag/mL). The final concentration was adjusted Institutional Animal Care and Use Committee at RPCI under protocol to 200 AL with 10% RPMI 1640. Cell lines were then incubated at 37jC

P966M. All animals were housed and maintained in laminar flow and 5% CO2 for 24 and 48 h. Following 24 or 48 h, 1 ACi per well of cabinets or microisolator units and provided with sterilized food and [3H]-thymidine (New England Nuclear Products) was added, and cells water. Our laboratory facility has been certified by the American were incubated for 16 h more. Cells were then harvested using the Association for Accreditation of Laboratory Animal Care and in Harvest system (Tomac Cell Harvester, Tomtec) into the 96-well glass accordance with current regulation and standards of the U.S. filters (Wallac Inc.), and [3H]-thymidine uptake was measured using an Department of Agriculture and the U.S. Department of Health and automated scintillation counter (Wallace). Each experiment was done Human Services. in triplicate at three different times; results are presented as the mean F Evaluation of apoptosis following in vitro exposure of NHL cell lines to SD of counts per minute (cpm) at 24 and 48 h. HGS ETR1 (TRAIL-R1 mAb) or lexatumumab (TRAIL-R2 mAb) with/ 51Cr release assays to study the effects of rituximab in combination with without rituximab. In vitro effects of either mapatumumab (TRAIL-R1 anti–TRAIL-R antibodies mapatumumab and lexatumumab to induce mAb) or lexatumumab (TRAIL-R2 mAb) as single agent or in antibody-dependent cellular cytotoxicity or CMC in vitro. NHL cells combination with rituximab were evaluated by flow-cytometric (5 Â 106 cells) were labeled with 51Cr for up to 2 h as previously analysis. Lymphoma cell lines (1 Â 106 cells) were exposed to either described (22). Subsequently, 51Cr-labeled cell suspensions [2 Â 106 mapatumumab (TRAIL-R1 mAb), lexatumumab (TRAIL-R2 mAb), or cells/mL for CMC assays and 1 Â 105 cells/mL for antibody-dependent isotype control () alone or in combination with rituximab. cellular cytotoxicity (ADCC)] were plated in 96-well plates. Labeled cells Each antibody was used at a final concentration of 10 Ag/mL. were then exposed to mapatumumab or lexatumumab mAb at 10 Ag/mL

Subsequently, cells were incubated at 37jC, 5% CO2 for a period of for 2 h and subsequently incubated in the presence of rituximab 24 or 48 h, and apoptosis was detected by staining-treated cells with (10 Ag/mL) or isotype control mAb (10 Ag/mL) and human serum at a FITC-labeled Annexin-V and propidium iodine (Oncogene). All 1:4 dilution (CMC) or PBMCs at an effector/target ratio of 40:1 (ADCC). samples were analyzed by multicolor flow-cytometric analysis using a After a 6-h period of incubation at 37jC and 5% CO2, the supernatant fluorescence-activated cell sorter/FACStar Plus (Becton Dickinson) flow was harvested, and 51Cr release was measured by standard gamma cytometer. Cells were scored as apoptotic if they were Annexin-V counter. The percentage of lysis was calculated as follows: % Lysis = [test positive and propidium iodine negative/positive (early and late cpm - background cpm]/[maximum cpm - background cpm]. apoptosis, respectively). The experiments were done in triplicate, and Correlation between in vitro effects of mapatumumab and lexatumu- the results were given as mean values F SE. mab with surface expression of TRAIL-R1 and TRAIL-R2. To correlate

Fig. 1. Changes in DNA synthesis and cell proliferation following 24 and 48 h exposure to anti ^ TRAIL-R1and anti ^ TRAIL-R2 humanized mAbs as determined by [3H]-thymidine incorporation. In vitro exposure of Ramos cell lines to HGS-ETR1 (mapatumumab) resulted in a significant decrease in DNA synthesis at 24 and 48h(P < 0.05) when compared with isotype or untreated controls (A). Similar finding were observed in Raji cell lines exposed to HGS-ETR1 (B). No antitumor activity was observed in HGS-ETR-2 (lexatumumab)^ treated cells.

Clin Cancer Res 2007;13(15) August1,2007 4558 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Mapatumumab and Rituximab Improved Antitumor Activity

Fig. 2. Effects of anti ^ TRAIL-R1and anti ^ TRAIL-R2 humanized mAbs in combination with rituximab in the DNA synthesis of Burkitt’s lymphoma cell lines. The in vitro exposure of Ramos cell lines to rituximab (10 Ag/mL) and HGS-ETR1 (mapatumumab; 10Ag/mL) resulted in an additive antitumor activity by decreasing DNA synthesis at 48 h (P < 0.05; A). Although significant antitumor activity by HGS-ETR1 (mapatumumab) was observed in Raji cells, no significant changes were observed in cell growth and DNA synthesis exposed to rituximab and HGS-ETR1for up to48h(B). No antitumor activity was observed in cells exposed to HGS-ETR2 (lexatumumab) mAb.

the in vitro activity of mapatumumab or lexatumumab, with or without Statistical analysis. In vitro experiments were repeated in triplicates, rituximab, we determined baseline expression of CD20, TRAIL-R1, and and the results are reported as the mean with standard error (ste). TRAIL-R2 in all cell lines tested. Surface CD20, TRAIL-R1, and TRAIL-R2 Statistical differences between treatment groups and controls were expression were determined by flow cytometry using FITC-labeling determined by m2 testing. In addition, differences in survival between using antigen-specific mAbs in each cell line. treatment groups were calculated using the Kaplan-Meier curves (SPSS In vivo effects of mapatumumab or lexatumumab alone or in 11.0 for Windows 2000 software). P values were calculated by log-rank combination with rituximab against human B-cell lymphoma-bearing test. Differences were considered significant when the P value xenografts. These studies were carried out using a disseminated was <0.05. human lymphoma-bearing SCID mouse xenograft model. Raji cells were harvested from confluent cultures, and only suspensions with Results >90% viable cells were used for animal inoculation as previously described (22). On day 0, 6- to 8-week-old SCID mice were inoculated In vitro exposure of NHL cells to mapatumumab results in a  6 with 1 10 Raji cells via tail vein injection. After 72 h (to allow tumor significant decrease in DNA synthesis. Antiproliferation activity engraftment), the animals were divided into seven cohorts. The first was observed in Raji and Ramos cells exposed to mapatumu- cohort (group A) served as control and received no treatment. Groups B mab for 24 and 48 h. In standard proliferation assays, and C consisted on animals treated with either HGS ETR1 (TRAIL-R1 mAb) at 10 mg/kg or HGS ETR2 (TRAIL-R2 mAb) at 10 mg/kg given via mapatumumab was able to inhibit the growth of Ramos cells tail vein injection on days +3, +5, +7, +9, +11, +13, +15, and +17. by 81% and 73% at 24 and 48 h, respectively. When compared Groups D and E were treated with rituximab and trastuzumab (isotype with untreated cells, in vitro exposure to mapatumumab control) monotherapy given via tail vein injection at 10 mg/kg on days resulted in a significant decrease in DNA synthesis. Mean +3, +5, +7, +9, +11, +13, +15, and +17. Groups G and H were treated cpm decreased from 124,085 cpm F 6,733 ste and 315,800 with either of mapatumumab (TRAIL-R1 mAb) at 10 mg/kg or cpm F 4,458 standard error (ste) in untreated cells at 24 and lexatumumab (TRAIL-R2 mAb) at 10 mg/kg on days +5, +9, +13, and 48 h, respectively, to 23,650 cpm F 2,931 ste and 85,932 cpm +17, alternating with rituximab at 10 mg/kg given via tail vein injection F 897 ste in mapatumumab-exposed Ramos cells (P = 0.007 on days +3, +7, +11, and +15. After completion of therapy, animals and P < 0.001), respectively (Fig. 1A). were observed for up to a 90-day time period. The end point of the Similar antitumor activity was observed in Raji cells where study was survival defined as the time to development of limb paralysis. in vitro Animals that reached the end point or survived after 3 months of exposure to mapatumumab resulted in growth observation were sacrificed by cervical dislocation. Pathologic exami- inhibition by 44% and 56% at 24 and 48 h, respectively. The nation of all organs (e.g., liver, lung and brain) was done to detect any mean cpm of HGS-ETR-1–exposed Raji cells were 2,008 cpm F residual disease. The experiments were repeated on three separate 147 ste (24 h) and 3,739 F 384 ste (48 h) and were lower occasions. than control exposed cells, at 3,582 cpm F 189 ste (24 h) and

www.aacrjournals.org 4559 Clin Cancer Res 2007;13(15) August 1, 2007 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Cancer Therapy: Preclinical

8,477 F 739 ste (48 h), P = 0.014 and P = 0.025, respectively exposed to rituximab or lexatumumab (15.7% F 3.67 ste versus (Fig. 1B). No significant antiproliferative effects were observed 4.13 F 3.06% versus 0%, P = 0.056; Fig. 3). with lexatumumab in the cell lines tested (Fig. 1). Mapatumumab activates the complement cascade and is In vitro exposure of B-cell lymphoma cell lines to rituximab capable of inducing effective CMC in various lymphoma cell did not result in significant changes in DNA synthesis at the lines. Depending on their Fc region, certain subclasses of time intervals studied. The combination of rituximab and mAbs can effectively activate the classic complement pathway mapatumumab or lexatumumab did not result in additive or and induce cell cytotoxicity. Mapatumumab and lexatumumab synergistic antiproliferative effects when compared with cells have an IgG1 backbone and can potentially activate CMC. In exposed to mapatumumab or lexatumumab alone in the the cell lines tested by our group, we found that mapatumumab majority of cell lines tested (Fig. 2A). Of interest, the in vitro is capable of inducing a variable degree of cell death by exposure of Ramos cells to rituximab plus mapatumumab for complement activation as determined by 51Cr release assays. 48 h resulted in the near-complete inhibition of DNA synthesis In vitro exposure of NHL cells to mapatumumab alone in the as compared with Ramos cells exposed to mapatumumab presence of human serum resulted in significant cell lysis in (P = 0.001) or rituximab (P < 0.001) alone (Fig. 2B). Ramos (mean F SE, 71.47 F 1.21%) and SU-DHL4 (27.65 F Exposure of NHL cell lines to mapatumumab or lexatumumab 0.92%) and cell lines as compared with untreated cells. induces a variable degree of apoptosis. We found a variable Additive effects where seen when mapatumumab was degree of apoptosis following the in vitro exposure of NHL cell combined with rituximab in some cell lines. In vitro exposure lines to mapatumumab, but not HGS-ETR-2, which is in of Raji cells to rituximab resulted in 27% cell lysis. The in vitro accordance to previous reports demonstrating that binding of addition of mapatumumab to rituximab increased the CMC DR receptors triggers apoptosis (6, 7, 9–11) in cancer cells. lysis to 40.3 F 1.48% (P = 0.047; Fig. 4). Rituximab, as a single Direct apoptosis was detected only in Ramos (56.2% F 4.43 agent, was very effective in inducing CMC in Ramos (94 F ste) after in vitro exposure to mapatumumab for 48 h as 2.1%) and SU-DHL-4 (75 F 2.2%) and masked any potential compared with untreated cells. On the other hand, in vitro augmented effect induced by combining it with mapatumumab exposure to rituximab did not result in the induction of mAb (Fig. 4). apoptosis at the same time interval (4.13 F 3.06%). Although Mapatumumab can induce ADCC in NHL cells and results in no significant improvement in apoptosis was noted by additive effects when combined with rituximab in vitro. As combining mapatumumab or lexatumumab with rituximab, shown by other investigators, ADCC is believed to be one of the the exposure of Ramos cells to both rituximab and lexatumu- most important in vivo mechanisms of action of rituximab and mab resulted in improved activity as compared with Ramos other mAbs (23–25). We further investigated the capacity of mapatumumab and lexatumumab to induce ADCC in vitro, with or without rituximab, in a panel of lymphoma cell lines. In vitro exposure to single agent mapatumumab, but not lexatumumab, resulted in a variable degree of ADCC (primarily in Ramos and Raji cell lines). Following the exposure of Ramos cells to mapatumumab, ADCC was observed in 55 F 9.1% of the cells. Although the addition of rituximab to mapatumumab increased the ADCC to 69 F 20.4%, this was not statistically significant when compared with single agent mapatumumab (P = 0.68; Fig. 5). In vitro exposure of Raji cells to mapatumumab resulted in minimal ADCC (6 F 3.1%). However, when combined with rituximab, the combination resulted in higher cell lysis (20.7 F 2.4%) than either rituximab (6.1 F 2%) or mapatumumab monotherapy alone (P = 0.046; Fig. 5). The in vitro antitumor activity of ETR-1 does not correlate with TRAIL-R1 antigen density expression on NHL cells. It is a commonly accepted observation that there exists a direct correlation between the amount of antigen expression on a cancer cell and the degree of biological activity of its associated mAb. In the present work, we aimed to correlate the in vitro responses to mapatumumab or lexatumumab (with or without rituximab) to the surface antigen expression of TRAIL-R1, Fig. 3. Induction of apoptosis following in vitro exposure to HGS-ETR1with or TRAIL-R2, and CD20 antigen, respectively. The panel of cell without rituximab or trastuzumab (isotype control). Ramos cells were exposed to lines studied expressed a similar degree of CD20, TRAIL-R1, HGS-ETR1 (mapatumumab) or HGS-ETR2 (lexatumumab; 10 Ag/mL) followed by either rituximab, isotype (10 Ag/mL), or media. Induction of apoptosis was detected and TRAIL-R2 antigens. No correlation was observed between by multiparameter flow-cytometric analysis (FITC-AnnexinV/PI staining) after sensitive cells and insensitive cells in vitro to a given antibody 48 h of incubation. HGS-ETR1 (mapatumumab) induced significant apoptosis in Ramos cells (56.2 % F 4.43 ste) when compared with rituximab (4.13% F 3 ste) given either alone or in combination versus antigen expression alone (P = 0.006).The addition of rituximab to mapatumumab did not increase the (data not shown). Our data suggest that the biological percent of apoptotic cells (58.47% F 7.36 ste, P = 0.552).The exposure of Ramos responsiveness of sensitive cells in vitro to mapatumumab is cellstobothrituximabandHGS-ETR2(15.7%F 3.67 ste, apoptosis) resulted in additive antitumor activity when compared with Ramos cells exposed to independent to the TRAIL-R1 antigen density. Alternate HGS-ETR2 alone (P =0.056). mechanisms of action may explain differences in tumor

Clin Cancer Res 2007;13(15) August1,2007 4560 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Mapatumumab and Rituximab Improved Antitumor Activity

responses between the cell lines tested; one possibility could be differences in heterodimerization with other TNF-R family members upon TRAIL-R1 binding by mapatumumab. In vivo effects of anti–TRAIL-receptor antibodies in combina- tion with rituximab: concurrent administration of mapatumumab with rituximab is more effective in controlling lymphoma growth and in prolonging survival than rituximab or mapatumumab monotherapy alone. Treatment of human lymphoma-bearing SCID mice with rituximab resulted in prolongation in survival as compared with placebo-treated controls. The median survival time for rituximab-treated animals was 33 days (95% CI, 29-37) as compared with a median survival of 21 days (95% CI, 20-22) for those animals receiving placebo (log-rank test, P = 0.002). No significant antitumor activity was observed in animals treated with lexatumumab alone, and the median survival (95% CI, 20-22) was similar to control mice for lexa- tumumab treated (Fig. 6A and B). On the other hand, mapa- tumumab seems to be able to better control lymphoma growth as a single agent as compared with placebo controls. The median survival for mapatumumab treated animals was longer (26 days; 95% CI, 22-30) than placebo-treated animals, but did not reach statistical significance (P = 0.23). The administration of alternating doses of rituximab in combination with mapatumumab for a total of eight doses resulted in the most effective antitumor activity and prolonga- tion of survival of human lymphoma-bearing SCID mice. Statistically, significant differences were observed between animals treated with rituximab versus mapatumumab plus rituximab (Fig. 6A and B). The median survival time of animals treated with mapatumumab and rituximab was longer (not reached after 90 days) than those treated with rituximab monotherapy alone [median survival of 33 days (95% CI, 29- 37); log-rank test, P = 0.05 (Fig. 6A and B)]. Contrary to what was observed in animals treated with rituximab plus mapatumumab, lexatumumab when combined with rituximab did not improve the survival of human lymphoma-bearing SCID mice. Animals receiving rituximab plus lexatumumab had a median survival that was not different from those animals treated with rituximab monotherapy [49 days (95% CI, 17-81) versus 33 days (95% CI, 29-37)]. Despite a trend toward an improved survival, differences between rituximab-treated animals and those receiving the concurrent administration of rituximab plus lexatumumab did not reach statistical significance (log-rank P = 0.887; Fig. 6A and B). After a follow-up period of 3 months, survival rates were the highest for animals treated with rituximab plus mapatu- mumab (60%) when compared with animals treated with rituximab alone (40%) or combination of rituximab plus lexatumumab (26.7%). Pathologic examination of surviving animals sacrificed at the end of the study failed to show any residual disease.

Discussion

Novel combination treatment strategies that attempt to Fig. 4. 51Cr-labeled NHL cells were exposed to HGS-ETR1 (mapatumumab), maximize rituximab activity while minimizing nonspecific HGS-ETR2 (lexatumumab), or isotype (5 Ag/mL) F rituximab (10 Ag/mL) and human serum (dilution, 1:4). 51Cr release was measured after a 6-h period of toxicity is a major focus of current preclinical and clinical incubation, and the percentage of lysis was calculated. In vitro exposure of Raji cells lymphoma research (26–30). Integrating novel and active to both rituximab and HGS-ETR1resulted in additive antiactivity (P =0.047). biological agents, such as mapatumumab, may potentially Complement-mediated cell lysis was more effectively induced in Raji cells exposed to both rituximab and HGS, when compared with single antibody therapy or increase the treatment options for many lymphoma patients. In HGS-ETR2. our current work, we evaluated the biological activity of two

www.aacrjournals.org 4561 Clin Cancer Res 2007;13(15) August 1, 2007 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Cancer Therapy: Preclinical

mapatumumab can activate the innate immune system and is capable of activating ADCC and CMC in sensitive B-cell lymphoma cells. Of interest, the degree of antitumor activity of mapatumumab in vitro did not correlate with TRAIL-R1 antigen density. A plausible explanation for our results could be differences in the heterodimerization between various B-cell lymphoma cells upon TRAIL-R1 binding to mapatumumab. It has been published that following in vitro exposure to TRAIL ligand or mAbs targeting TRAIL-R1, there is recruitment by trimerization of other DRs, which results in signaling transduction and apoptosis (13). It is feasible that responses to mapatumumab are dependent on the type of DRs undergoing trimerization upon TRAIL-R1 binding rather than the degree of antigen density. The same principle could explain the absence of

Fig. 5. HGS-ETR1in sensitive cells is capable of inducing effective ADCC (A). In addition, when combined with rituximab in vitro against 51Cr-labeled NHL cells, this resulted in better antitumor activity than rituximab monotherapy (P =0.046). B, cell were labeled with 51Cr and exposed in vitro to HGS-ETR1 (mapatumumab), HGS-ETR2 (lexatumumab), or isotype (5 Ag/mL) F rituximab (5 Ag/mL) and PBMCs (effector/target ratio, 40:1). 51Cr release was measured, and the percentage of lysis was calculated after 6 h.

fully human mAbs targeting TRAIL-Rs alone or in combination with rituximab against a panel of various B-cell lymphoma cell Fig. 6. Kaplan-Meier analysis of cumulative survival of SCID mice bearing Raji lymphoma xenografts treated with rituximab and/or mapatumumab. SCID mice lines. were inoculated with 1 Â10 6 Raji cells via tail vein i.v. injection (day 0) and treated As has been shown by other investigators, the DR pathway is either with placebo, rituximab, isotype (10 mg/kg/dose on days +3, +5, +7, +11, intact and functional in various types of cancers, including +13, +15, and +17), mapatumumab, and lexatumumab (10 mg/kg/dose on days +3, +5, +7, +11, +13, +15, and +17), or either mapatumumab (anti ^ TRAIL-R1mAb) at B-cell lymphomas (19, 31). Moreover, binding and subsequent 10 mg/kg or lexatumumab (anti ^ TRAIL-R2 mAb) at10 mg/kg on days +5, +9, +13, activation of the TNF death family receptors by either ligands or and +17, alternating with rituximab at 10 mg/kg administered via tail vein injection on days +3, +7, +11, and +15 and analyzed for survival as a function of time. antibodies result in significant apoptosis and antitumor activity A, rituximab in combination with mapatumumab results in an improved antitumor in a variety of solid tumor malignancies (5, 9, 12–15, 17–20). activity when compared with mAb monotherapy.The median survival of lymphoma As shown by our data, mapatumumab, but not lexatumumab, bearing SCID mice treated with rituximab in combination with mapatumumab has not been reached as compared with a shorter median survival observed in animals is active and capable of inducing cell growth inhibition and treated with rituximab- or mapatumumab-treated animals (B). n = 15 mice per apoptosis in NHL cell lines and xenografts. In addition, group.

Clin Cancer Res 2007;13(15) August1,2007 4562 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Mapatumumab and Rituximab Improved Antitumor Activity

antitumor activity of lexatumumab despite ample surface that the activation of intrinsic caspase pathway may be required expression of TRAIL-R2 antigen in the cell lines tested. Addi- for the amplification of the apoptotic effect. Based on these tional studies are needed to further address such hypothesis. findings, we decided to study potential biological interactions A phase II clinical study with mapatumumab was recently between mapatumumab or lexatumumab and rituximab. reported in patients with relapsed/refractory B-cell lymphomas. We believe that we are the first group to report on the Patients included on the study had relapsed/refractory NHL of augmented antilymphoma activity from the combination of any histologic subtype and were heavily pretreated. Mapatumu- rituximab plus mapatumumab in vivo and, to a lesser degree, mab was administered at two dose levels: 3 mg/kg (8 subjects) in vitro. In our animal studies, the survival of lymphoma- or 10 mg/kg (32 subjects) every 21 days in the absence of bearing SCID mice was longest in animals receiving rituximab disease progression or prohibitive toxicity for up to six cycles. plus mapatumumab, as compared with animals receiving Preliminary reports on the efficacy and toxicity showed that rituximab, mapatumumab, or isotype monotherapy alone. mapatumumab was well tolerated with patients experiencing Because all the animals received a total of eight doses of primarily grade 1 or 2 toxicities. In addition, objective antitumor antibody, either single agent or in combination, the prolonga- responses were shown in a subset of patients, all with follicular tion in survival most likely is the result of biological interactions histologies. Moreover, 30% of all patients had disease stabiliza- between two mAbs rather than excess activity from a single tion following mapatumumab therapy (32). biologically active agent. Because mapatumumab has activity in NHL and can To a lesser degree, we found that mapatumumab, when elucidate antitumor activity against B-cell lymphomas by combined with rituximab, results in additive effects with regard different mechanisms of action than rituximab, we further to cell growth inhibition and activation of effective ADCC and studied its interactions with rituximab in the preclinical setting. CMC in sensitive B-cell lymphoma cell lines. Because the Rituximab has been shown to be effective in activating the mechanisms of action of each mAb tested (rituximab and complement cascade via the classic pathway and in recruiting mapatumumab) is diverse, in vitro testing using standardized effector cells (e.g., neutrophils and natural killer cells) into the immunologic (i.e., ADCC or CMC assays), apoptosis, and/or tumor bed, inducing ADCC (22–25). In addition, in vitro proliferation assays may largely underestimate the antitumor exposure of lymphoma cells to rituximab induces programmed interactions between rituximab and mapatumumab. As a result cell death via the intrinsic apoptotic pathway. On the other of such limitations, in vivo animal studies may more accurately hand, mapatumumab can trigger apoptosis by caspase-8 activa- reflect and predict what may occur in clinical studies. Our tion via the extrinsic apoptotic pathway. An earlier study (19) findings support future evaluation of this unique combination that showed resistance to mapatumumab and lexatumumab in (i.e., mapatumumab and rituximab) in clinical trails of patients lymphoma cell lines is due to lack of Bid expression, suggesting with CD-20–positive, B-cell lymphomas.

References 1. Kelley SK, Ashkenazi A. Targeting death receptors in 11. LeBlanc HN, Ashkenazi A. Apo2L/TRAIL and its primary and cultured lymphoma cells: induction of cancer with Apo2L/TRAIL. Curr Opin Pharmacol death and decoy receptors. Cell Death Differ 2003; apoptosis and enhancement of doxorubicin- and 2004;4:333 ^ 9. 10 :66 ^ 75. -induced cell death. Br J Haematol 2005; 2. Vaux DL, Cory S, Adams JM. Bcl-2 gene promotes 12. Pukac L, Kanakaraj P, Humphreys R, et al. Mapa- 130:501^10. haematopoetic cell survival and cooperates with tumumab, a fully human TRAIL-receptor 1 monoclo- 20. Motoki K, Mori E, Matsumoto A, et al. Enhanced c-myc to immortalize pre-B cells. Nature 1998;335: nal antibody, induces cell death in multiple tumour apoptosis and tumor regression induced by a direct 440^2. types in vitro and in vivo. Br J Cancer 2005; agonist antibody to tumor necrosis factor-related 3. Nagata S. Apoptosis by death factor. Cell 1997;88: 92:1430^41. apoptosis-inducing ligand receptor 2. Clin Cancer Res 355^65. 13 . Cretney E, Shanker A, Yagita H, Smyth MJ, Sayers 2005;11:3126^35. 4. WileySR,SchooleyK,SmolakPJ,etal.Identification TJ. TNF-related apoptosis-inducing ligand as a thera- 21. Carnahan J, Stein R, Qu Z, et al. Epratuzumab, a and characterization of a new member of theTNF fam- peutic agent in autoimmunity and cancer. Immunol CD22-targeting recombinant ily that induces apoptosis. Immunity 1995;3:673 ^ 82. CellBiol2006;84:87^98. with a different mode of action from rituximab. Mol 5. Walczak H, Miller RE, Ariail K, et al. Tumoricidal 14. Keane MM, Ettenberg SA, Nau MM, Russell EK, Immunol 2007;44:1331 ^ 41. activity of tumor necrosis factor-related apoptosis- Lipkowitz S. Chemotherapy augmentsTRAIL-induced 22. Hernandez-Ilizaliturri FJ, Jupudy V, Reising S, inducing ligand in vivo. Nat Med 1999;5:157 ^ 63. apoptosis in breast cell lines. Cancer Res 1999;59: Repasky EA, Czuczman MS. Concurrent administra- 6. Debatin KM, Krammer PH. Death receptors in 734 ^ 41. tion of granulocyte colony-stimulating factor or granu- chemotherapy and cancer. Oncogene 2004;23: 15. GriffithTS, Lynch DH. TRAIL: a molecule with multi- locyte-monocyte colony-stimulating factor enhances 2950 ^ 66. ple receptors and control mechanisms. Curr Opin the biological activity of rituximab in a severe com- 7. Tartaglia LA, Ayres TM, Wong GH, Goeddel DV. A Immunol 1998;10:559 ^ 63. bined immunodeficiency mouse lymphoma model. novel domain within the 55 kd TNF receptor signals 16. Van der Sloot AM,Tur V, Szegezdi E, et al. Designed Leuk Lymphoma 2005;46:1775 ^ 84. cell death. Cell 1993;10:845 ^ 53. tumor necrosis factor-related apoptosis-inducing 23. Eisenberg R, Looney RJ. The therapeutic potential 8. NakaT, Sugamura K, Hylander B, Widmer M, Rustum ligand variants initiating apoptosis exclusively via the of anti-CD20 ‘‘what do B-cells do?’’ Clin Immunol Y, Repasky E. Effects of tumor necrosis factor-related DR5 receptor. Proc Natl Acad Sci U S A 2006;103: 2005;117:207 ^ 13. apoptosis-inducing ligand alone and in combination 8634 ^9. 24. Fischer L, Penack O, Gentilini C, et al. The anti- with chemotherapeutic agents on patients’ colon 17. Almasan A, Ashkenazi A. Apo2L/TRAIL: apo- lymphoma effect of antibody-mediated immunother- tumors grown in SCID mice. Cancer Res 2002;62: ptosis signaling, biology, and potential for cancer apy is based on an increased degranulation of 5800^6. therapy. Cytokine Growth Factor Rev 2003;14: peripheral blood natural killer (NK) cells. Exp Hematol 9. Pitti RM, Marsters SA, Ruppert S, Donahue CJ, 337^48. 2006;34:753^9. Moore A, Ashkenazi A. Induction of apoptosis by 18. Mitsiades CS, Treon SP, Mitsiades N, et al. TRAIL/ 25. Jazirehi AR. Bonavida B. Cellular and molecular Apo-2 ligand, a new member of the tumor necrosis Apo2L ligand selectively induces apoptosis and over- signal transduction pathways modulated by rituxi- factor cytokine family. J Biol Chem 1996;271: comes drug resistance in : therapeu- mab (rituxan, anti-CD20 mAb) in non-Hodgkin’s 12 6 87 ^ 9 0. tic applications. Blood 2001;98:795 ^ 804. lymphoma: implications in chemosensitization and 10. Gruss HJ, Dower SK. Tumor necrosis factor ligand 19. Georgakis GV, Li Y, Humphreys R, et al. Activity of therapeutic intervention. Oncogene 2005;24: superfamily: involvement in the pathology of malig- selective fully human agonistic antibodies to the 2121^ 43. nant lymphomas. Blood 1995;85:3378 ^ 404. TRAIL TRAIL receptors TRAIL-R1 and TRAIL-R2 in 26. Davis TA, Maloney DG, Grillo-Lopez AJ, et al.

www.aacrjournals.org 4563 Clin Cancer Res 2007;13(15) August 1, 2007 Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Cancer Therapy: Preclinical

Combination immunotherapy of relapsed or refracto- tion immunotherapy with a CpG oligonucleotide (1018 of interleukin-12 in combination with rituximab in ry low-grade or follicular non-Hodgkin’s lymphoma ISS) and rituximab in patients with non-Hodgkin’s patients with B-cell non-Hodgkin lymphoma. Blood with rituximab and interferon-a-2a. Clin Cancer Res lymphoma: increased interferon-a/h ^ inducible gene 2002;99:67 ^ 74. 2000;6:2644 ^ 52. expression without significant toxicity. Blood 2005; 31. SnellV, Clodi K, Zhao S, et al. Activity ofTNF-related 27. van der Kolk LE, Grillo-Lopez AJ, Baars JW, van 105:489^95. apoptosis-inducing ligand (TRAIL) in haematological Oers MH. Treatment of relapsed B-cell non-Hodgkin’s 29. Gluck WL, Hurst D, Yuen A, et al. Phase I studies malignancies. Br J Haematol 1997;99:618^ 24. lymphoma with a combination of chimeric anti-CD20 of interleukin (IL)-2 and rituximab in B-cell non- 32. Younes A, Vose JM, Zelenetz AD, et al. Results of monoclonal antibodies(rituximab) and G-CSF: final Hodgkin’s lymphoma: IL-2 mediated natural killer cell a phase 2 trial of HGSETR1 (agonistic human mono- report on safety and efficacy. Leukemia 2003;17: expansion correlations with clinical response. clonal antibody to TRAIL receptor 1) in subjects with 1658 ^ 64. Clin Cancer Res 2004;10:2253 ^ 64. relapsed/refractory non-Hodgkin’s lymphoma (NHL) 28. Friedberg JW, Kim H, McCauley M, et al. Combina- 30. Ansell SM,WitzigTE, Kurtin PJ, et al. Phase 1study [abstract 489]. Blood 2005;106:146a.

Clin Cancer Res 2007;13(15) August1,2007 4564 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research. Augmented Antitumor Activity against B-Cell Lymphoma by a Combination of Monoclonal Antibodies Targeting TRAIL-R1 and CD20

Sreeram Maddipatla, Francisco J. Hernandez-Ilizaliturri, Joy Knight, et al.

Clin Cancer Res 2007;13:4556-4564.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/13/15/4556

Cited articles This article cites 32 articles, 11 of which you can access for free at: http://clincancerres.aacrjournals.org/content/13/15/4556.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/13/15/4556.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/13/15/4556. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2007 American Association for Cancer Research.