| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics, Molecular Targets, and Chemical Biology |
1 Micromet AG, Munich, Germany; 2 Experimental Pharmacology and Oncology, Berlin-Buch, Berlin, Germany; and 3 Department of Gynecology and Obstetrics, University of Essen, Essen, Germany
Requests for reprints: Patrick A. Baeuerle, Micromet AG, Staffelseestrasse 2, 81477 Munich, Germany. Phone: 49-89-895277601; Fax 49-89-895277205; E-mail: patrick.baeuerle{at}micromet.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Key Words: Ep-CAM • bispecific antibody • xenograft model • T lymphocyte • tumor immunity
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bispecific T-cell–recruiting antibody constructs offer a completely different route to reactivate T cells against tumor cells that has the promise to be independent from MHC class I expression and peptide antigen processing by tumor cells and T-cell receptor specificity of T cells. Bispecific antibodies are designed to recruit cytotoxic T cells by a common signaling molecule, such as CD3, against tumor cells bearing a frequently and differentially expressed tumor-associated cell surface antigen. The latter can also be recognized by recombinant monoclonal antibody therapies, but is not directly recognized by T cells. In contrast to vaccination strategies, bispecific antibodies will not mount a tumor-specific oligoclonal but a polyclonal T-cell response. As a consequence, essentially all differentiated T cells of the CD8 as well as CD4 phenotype will become activated and participate in a reaction against target antigen-positive tumor cells. Many different formats of bispecific antibodies have been developed during the past 20 years (5–8). The necessary clinical specifications for such antibodies are, however, difficult to meet. T-cell–recruiting bispecific antibodies ought to be active at low effector-to-target ratios as present in tumors. This means they need to support serial killing by activated T cells. For clinical use, bispecific antibodies have to be simple and robust molecules that can be easily manufactured for large patient populations. Many attempts have been made but few formats fulfilled the above specifications and finally progressed to clinical trials (9, 10).
The class of bispecific T-cell engager (BiTE) has been well studied by a CD19–/CD3–bispecific antibody construct called bscCD19xCD3 (11–14). It is composed of two tandemly arranged single-chain antibodies forming a single 55,000 daltons polypeptide chain. The CD19-specific BiTE molecule was developed for the treatment of human B-cell malignancies and is currently in phase I trials. With its T cell–dependent cytotoxic activity at subpicomolar concentrations, effectiveness at very low effector-to-target ratios, ease of production, and high-stability bscCD19xCD3 fulfills most requirements for the development of a bispecific antibody in malignant diseases. The efficacy of bscCD19xCD3 was investigated in vitro with cell lines (11, 13), ex vivo with B lymphoma patient samples (14), in vivo with a nonobese diabetes/severe combined immunodeficiency (NOD/SCID) mouse model (12), and in chimpanzee, the only relevant species apart from man.4 All experiments showed a high biological activity of the BiTE molecule.
Here we have investigated another member of this class of bispecific antibodies, called bscEp-CAMxCD3, for its in vivo activity with respect to inhibition of tumor take and activity against established adenocarcinoma-derived tumors in immunodeficient NOD/SCID mouse models. BscEp-CAMxCD3 is specific for epithelial cell adhesion molecule (Ep-CAM), a target antigen that has been extensively used by various antibody-based therapies (6, 15–19). Like most other tumor-associated antigens, Ep-CAM is expressed on normal tissue. However, Ep-CAM expression on a wide variety of carcinomas does exceed expression and accessibility of the antigen on normal cells, thereby establishing a useful therapeutic window for a targeted antibody approach (20, 21). BscEp-CAMxCD3 is here shown to be highly efficacious in inhibition of tumor growth and in eradication of established tumors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Anti–Ep-CAM antibody BerP-4 was purchased from DAKO (Hamburg, Germany), anti-CD3 antibody UCHT-1 from Sigma-Aldrich (Taufkirchen, Germany), anti-CD4 antibody RPA-T4, and anti-CD8a antibody RPA-T8 from BD Biosciences (Heidelberg, Germany). Simultest IMK-Lymphocyte subpopulation differentiation kit was purchased from BD Biosciences and LSAB immunohistochemistry detection kit from DAKO. SW480 human colon carcinoma cell line was purchased from American Type Culture Collection (Manassas, VA). CHO cells stably expressing human Ep-CAM were generated at Micromet. Primary ovarian tumor specimens were kindly provided by Prof. Dr. R. Kimmig (University of Essen, Essen, Germany). All patients had an advanced ovarian cancer stage (Federation Internationale de Gynecologie et d'Obstetrique stage IIIC or IV). Histology mainly was serous papillary. Informed consent was obtained from each subject.
Effector and target cell preparation. SW480 cells were grown in RPMI (Life Technologies, Karlsruhe, Germany) supplemented with 10% FCS (Life Technologies) at 37°C in a 5% CO2 humidified incubator. Subconfluent cultures were detached by incubation with 0.25% trypsin and checked for vitality by trypan blue exclusion. Only cultures with viability >95% were used for in vitro cytotoxicity assays and in vivo experiments.
Peripheral blood mononuclear cells (PBMC) of healthy donors were used as effector cells for in vitro and in vivo efficacy analysis of the bscEp-CAMxCD3 antibody. PBMCs were prepared by Ficoll density centrifugation from enriched lymphocyte preparations (buffy coats) obtained from local blood banks. Erythrocytes were depleted by incubation for 15 minutes with erythrocyte lysis buffer (155 mmol/L NH4Cl, 10 mmol/L KHCO3, and 100 µmol/L EDTA) and thrombocytes removed by supernatant aspiration following centrifugation at 100 x g for 10 minutes. Purified PBMCs were cultured in RPMI/10% FCS and 25 mmol/L HEPES (Sigma) until use. The relative numbers of different lymphocyte subpopulations (CD3-, CD4-, and CD8-positive lymphocytes) were determined by fluorescence-activated cell sorting (FACS) analysis using the Simultest IMK-Lymphocyte kit (BD Biosciences) according to the manufacturer's instruction.
In vitro cytotoxicity assay. Bioactivity of bscEp-CAMxCD3 was analyzed by a FACS-based in vitro cytotoxicity assay using the Ep-CAM–positive colon carcinoma cell line SW480 and a stably Ep-CAM–expressing CHO cell line as target cells and human PBMCs (CD3-enriched or not enriched) as effector cells. CD3 enrichment was done by using the Human T-Cell Enrichment Column kit (R&D Systems, Wiesbaden-Nordenstadt, Germany) according to the manufacturer's instructions.
Target cells were washed twice with PBS and labeled with PKH26 dye (Sigma-Aldrich) according to the manufacturer's instructions. Labeled target cells were washed twice with RPMI/10% FCS and mixed with freshly isolated effector cells at an effector-to-target ratio of 10:1 or 5:1. Target (2 x 104) and effector cells (1 x 105 to 2 x 105) in a volume of 50 µL RPMI/10% FCS were added per well in a 96-well round-bottomed plate. Serial dilutions of bscEp-CAMxCD3 (5- or 4-fold) were prepared in RPMI/10% FCS starting with a concentration of 200 ng/mL. Fifty microliters of the different bscEp-CAMxCD3 solutions were added in triplicates to the corresponding wells. Individual cytotoxicity mixtures were incubated for 0, 2, 8, 16, and 24 hours at 37°C/5% CO2. Propidium iodide was added to a final concentration of 1 µg/mL per well and plates were incubated for 10 minutes at room temperature. The number of PKH and propidium iodide–positive target cells was determined by FACS. Cytotoxicity was measured as the ratio of propidium iodide–positive (dead cells) over total number of target cells (PKH-positive) according to the formula: cytotoxicity (%) = [(propidium iodide–positive cells / PKH-positive cells) x 100]. Sigmoidal dose response killing curves were analyzed by Prism Software (GraphPad Software, Inc., San Diego, CA) and the bscEp-CAMxCD3 concentration calculated that induced half maximal killing (EC50 value).
Mice and experimental conditions. In vivo experiments were done in female 6- to 10-week-old C57BL/6 and immunodeficient NOD/SCID mice characterized by T cell, B cell, and natural killer cell deficiency and lack of macrophage function (The Jackson Laboratory, Bar Harbor, ME). The mice were maintained under sterile and standardized environmental conditions (20 ± 1°C room temperature, 50 ± 10% relative humidity, 12 hours light-dark rhythm). Mice received autoclaved food and bedding (Ssniff, Soest, Germany) and acidified (pH 4.0) drinking water ad libitum. NOD/SCID mice were tested for leakiness of immunodeficiency and only mice with IgG levels below 100 ng/mL were used. All experiments were done according to the German Animal Protection Law with permission from the responsible local authorities. In compliance with the Animal Protection Law, mice had to be euthanized when tumor volumes exceeded 10% of their body weight. Statistical analysis of the mean tumor volume of the corresponding treatment groups versus the vehicle control group was done by the nonparametric Mann-Whitney U test.
SW480 colon carcinoma model. For each animal, 2.5 x 106 or 5 x 106 human PBMC were mixed with 5 x 106 SW480 colon carcinoma cells in a final volume of 0.2 mL PBS. The PBMC effector-to-target cell mixture (E:T, 1:2 or 1:1) was s.c. injected into the right flank of each NOD/SCID mouse. Two different variants of the tumor model (early treatment and established tumor model) were used. For the early treatment model, four to six animals per group were i.v. treated with bscEp-CAMxCD3, PBS control vehicle, or bispecific antibody control starting 1 hour after SW480/PBMC inoculation at the indicated doses and treatment was repeated for 4 consecutive days. In the established tumor model, initiation of treatment was delayed until s.c. growing SW480 tumors had developed (50-200 mm3). Six animals per group were treated starting at days 4, 8, and 12, respectively, and treatment was repeated for 4 consecutive days. Tumors were measured on the indicated days with a caliper in two perpendicular dimensions and tumor volumes were calculated according to the following formula: tumor volume = [(width2 x length) / 2].
Primary ovarian human xenograft model. Following surgical resection of peritoneal metastasis of histologically proven ovarian cancer patients, primary tumor specimens were cut into 50 to 100 mm3 cubes and s.c. implanted into the right flank of six female 6- to 10-week-old NOD/SCID mice per group. Animals were i.v. treated with 5 µg bscEp-CAMxCD3/injection or PBS control vehicle for 5 consecutive days at the indicated time points post tumor implantation. Tumor sizes were measured on the indicated days with a caliper in two perpendicular dimensions and tumor volumes calculated according to the following formula: tumor volume = [(width2 x length) / 2].
In parallel, at least four representative pieces of the individual ovarian tumor specimens were subjected to immunohistochemistry analysis to determine the level of Ep-CAM expression and presence of T cells. In brief, cryo-slides of the primary ovarian carcinomas were fixed with glutaraldehyde. Endogenous peroxidase activity was quenched with freshly prepared 3% hydrogen peroxide for 5 minutes at room temperature. Nonspecific antibody binding was blocked with bovine serum albumin (BSA) for 30 minutes at room temperature. Sections were then incubated with the primary anti-human antibodies: Ep-CAM (BerEP4/FITC, DAKO), CD3 (clone UCHT-1, Sigma), CD4 (clone RPA-T4, BioScience), and CD8a (clone RPA-T8, BioScience) for 30 minutes at room temperature. Secondary anti-mouse or anti-rabbit antibodies and the streptavidin biotin method with peroxidase staining (DAKO-LSAB kit) were used for staining. Between 200 and 400 cells in at least four microscopic views were semiquantitatively evaluated. The percentage of positively stained cells was assessed.
Pharmacokinetic analysis of bscEp-CAMxCD3. Twenty C57BL/6 mice were i.v. injected with 100 µg dimeric bscEp-CAMxCD3 and allocated to four different groups of five mice each. Different groups were alternatingly bled at different time points after bscEp-CAMxCD3 injection (pre-dose, 2, 15, and 30 minutes and 2, 4, 6, and 24 hours) with a maximum of two bleeds per animal. BscEp-CAMxCD3 serum concentrations were quantified by a bscEp-CAMxCD3–specific ELISA.
ELISA plates (Nunc, Wiesbaden, Germany) were coated with 100 µL anti-His antibody (Penta-His; Qiagen) diluted in PBS to a final concentration of 2 µg/mL and incubated overnight at 4°C. Plates were blocked by PBS/1% BSA for 60 minutes at 25°C and washed thrice. Test samples were diluted in PBS/10% mouse serum pool, 100 µL added per well, and incubated for 60 minutes at 25°C. Plates were washed thrice and incubated with 100 µL soluble Ep-CAM protein conjugated with biotin (Micromet, Munich, Germany) at a final concentration of 2 µg/mL for 60 minutes at 25°C. Plates were washed thrice followed by incubation for 60 minutes at 25°C with 100 µL streptavidin conjugated with alkaline phosphatase (DAKO) at a final concentration of 0.5 µg/mL. Plates were washed thrice, incubated for 60 minutes with 180 µL substrate (1 mg/mL p-NPP dissolved in 0.2 mol/L Tris buffer; Sigma) at 25 °C and the absorbance (405 nm) was read on Power WaveX select (Bio-Tek Instruments, Bad Friedrichshall, Germany). The bscEp-CAMxCD3 concentration in different test samples was calculated referring to a bscEp-CAMxCD3 standard calibration curve.
Pharmacokinetic calculations of bscEp-CAMxCD3 were done by the pharmacokinetic software package WinNonlin Professional 4.1 (Pharsight Corporation, Mountain View, CA; 2003). Parameters were determined by noncompartmental analysis. The noncompartmental analysis was based on model 201 (i.v. bolus injection). The terminal elimination half-life t1/2 was calculated using a log-linear regression of the concentration data including the last four sampling time points with a measurable concentration of bscEpCAMxCD3.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
) of 0.59 ± 0.17 hours and a terminal elimination half-life (t1/2 ß) of 5.3 ± 0.4 hours. Bolus injection of 100 µg bscEp-CAMxCD3 resulted in a mean maximum serum concentrations (Cmax) of 77.6 ± 8.3 µg/mL and an associated exposure (AUClast) of 206.3 ± 3.6 hour x µg/mL. The clearance was determined with 0.47 ± 0.01 mL/h and the volume of distribution with 3.54 ± 0.25 mL. Establishment of a human colon carcinoma xenograft model in nonobese diabetes/severe combined immunodeficiency mice. To evaluate the in vivo efficacy of bscEp-CAMxCD3, the human SW480 cell line was used to establish a xenograft model in NOD/SCID mice. With a s.c. inoculum of 1 x 107 tumor cells, all injected NOD/SCID mice developed rapidly growing tumors of similar growth kinetics (data not shown). Inoculation of 1 x 106 SW480 cells also resulted in 100% tumor take but growth kinetics were slower and interanimal variability was higher. Injection of 1 x 105 SW480 cells did not result in the formation of palpable tumors. An inoculum of 5 x 106 SW480 cells was, therefore, selected for in vivo experiments to assure 100% tumor take, to keep tumor variability low, and to induce an intermediate kinetic of SW480 tumor growth.
Because bscEp-CAMxCD3 is specific for both human CD3 and human Ep-CAM, human-derived effector cells had to be coinjected with the tumor cells expressing human Ep-CAM. To this end, unstimulated human PBMC from healthy donors were used and premixed with Ep-CAM–positive SW480 tumor cells followed by s.c. injection of cell mixes into NOD/SCID mice. To mimic the situation within human tumors more closely, a low effector-to-target ratio of 1:1 or 1:2 (5 x 106 or 2.5 x 106 PBMC plus 5 x 106 SW480 cells) was used. PBMC donor samples were analyzed by flow cytometry to determine the effector-to-target ratio with respect to CD3-, CD4-, and CD8-positive T effector cells. All PBMC samples used had similar percentages of T-cell subpopulations (i.e., 46.5 ± 6.8% CD3-positive cells and 31.7% ± 4.9% CD3/CD4 and 14.1% ± 2.6% CD3/CD8 double-positive cells). The effector-to-target ratios, as are relevant for bscEp-CAMxCD3 activity, were therefore smaller than 1:1; it was 
1:2 for total CD3-, 1:3 for CD4-, and 1:7 for CD8-positive T cells.
BscEp-CAMxCD3 dose finding in the SW480 human xenograft model. Based on the in vitro efficacy and the pharmacokinetic analysis of bscEp-CAMxCD3, a daily i.v. dosing scheme was selected for the dose finding study. Four animals per treatment group were i.v. treated with 0.01, 0.1, 1, and 10 µg bscEp-CAMxCD3 per injection starting 1 hour after inoculation of the PBMC/SW480 cell mixture. BscEp-CAMxCD3 treatment was repeated once daily for 4 consecutive days and outgrowth of solid s.c. SW480 tumors determined by caliper measurements (Fig. 2). A clear dose response for the antitumor activity of bscEp-CAMxCD3 was observed. Following treatment with the lowest dose level of 5 x 0.01 µg bscEp-CAMxCD3, four of four animals developed SW480 tumors and there was no reduction in tumor growth observed compared with the SW480 only and the SW480 + PBMC control group. Five consecutive doses of 0.1, 1, and 10 µg bscEp-CAMxCD3 per injection induced complete growth inhibition of SW480 tumors in four of four mice until day 60 after SW480 tumor cell inoculation. At the end of the study on day 82, all mice of the 10 and 1 µg dose group were still free of detectable tumor and only one animal in the 0.1 µg dose group had recurrence of a SW480 tumor.
|
|
Treatment of established SW480 tumors by bscEp-CAMxCD3 at different time points. We next explored the therapeutic effect of bscEp-CAMxCD3 on palpable s.c. growing SW480 tumors. Six animals per group were either treated with PBS control buffer or bscEp-CAMxCD3 1 hour after inoculation of the PBMC/SW480 mixture. In the other groups, start of treatment was delayed to days 4, 8, and 12 after inoculation, respectively, when s.c. growing tumors had reached volumes between 50 and 200 mm3. A dose of 1 µg bscEp-CAMxCD3 on 5 consecutive days was selected because this treatment schedule had been shown to strongly inhibit SW480 tumor outgrowth (see Figs. 2 and 3).
In the early treatment group, bscEp-CAMxCD3 induced complete inhibition of SW480 tumor growth in six of six animals (Fig. 4A). BscEp-CAMxCD3 treatment at days 4 to 8 reduced the size of established SW480 tumors to the limit of detection in six of six animals until day 18 (Fig. 4B). At day 25, two of six mice had recurring SW480 tumors. At the end of the study on day 42, three of six animals had their tumors eradicated. Statistical analysis of the mean tumor volume of the control group versus the bscEp-CAMxCD3 treatment group showed highly significant differences until the end of the study (P < 0.0005). Treatment of established tumors between days 8 to 12 strongly inhibited tumor growth in six of six animals (Fig. 4C). At day 25, tumors were still absent in two of six animals. By the end of the study on day 42, one of six animals was still tumor free and mean tumor volume of the bscEp-CAMxCD3 treatment group was significantly lower between day 12 and the end of the study than that of the control group (P < 0.03). BscEp-CAMxCD3 treatment between days 12 to 16 showed a growth inhibitory trend in four of six animals (Fig. 4D). At the end of the study, six of six animals developed SW480 tumors and mean tumor volume was not significantly lower than in the control group.
|
|
Primary ovarian tumor samples from patients were cut into equal pieces of 100 mm3 and s.c. implanted into NOD/SCID mice. From a total of six patient samples, tumor pieces were partially accepted in immunodeficient mice in four cases and showed either stable size or growth during the observation period of 70 days. The primary ovarian tumor samples from these four patients (patients 1-4) were each implanted as 12 equally sized pieces into two groups of six NOD/SCID mice each. One group of six NOD/SCID mice was i.v. treated with PBS control buffer at the indicated time points, whereas the other group was treated with 5 µg bscEp-CAMxCD3 per i.v. injection. Four additional pieces of each tumor were characterized for intersample and intrasample heterogeneity with respect to the percentage of tumor cells by morphologic criteria, and for the relative number of Ep-CAM-, CD3-, CD4-, and CD8-positive cells by immunohistochemical staining with respective antibodies. H&E staining of tumor specimen revealed a tumor cell percentage of 70% to 95%, 50% to 90%, 10% to 50%, and 10% for samples 1 to 4, respectively. The number of Ep-CAM–expressing cells was closely correlated with the number of tumor cells. Analysis of the effector cell populations resulted in 0.83 ± 1.2% CD3-positive and 0.38 ± 0.42% CD8-positive cells for sample 1. Sample 2 showed 0.83 ± 0.45% CD3-positive and 0.2 ± 0.23% CD8-positive cells; sample 3 had 6.8 ± 5% CD3-positive, 1.6 ± 1.8% CD4-positive and 4.1 ±2.8% CD8-positive cells. Sample 4 contained 5.8 ± 7.4% CD3-positive and 0.8% CD8-positive cells. In tumor samples 1, 2, and 4, no CD4-positive cells could be detected.
Tumor pieces from the four patient samples grew slowly in NOD/SCID mice. With samples 3 and 4, tumor take was poorest and even in PBS control group only one of six and two of six animals, respectively, had measurable xenografts at the end of the study (i.e., 70 days after implantation). In the bscEp-CAMxCD3–treated cohort with samples 3 and 4, mean tumor volume during the study period was lower compared with the PBS control group, and at the end of the study all bscEp-CAMxCD3–treated animals were free of detectable tumor (data not shown). With patient samples 1 and 2, the rate of tumor take was higher and tumor growth was more consistent (Fig. 6). In both experiments, five of six animals in the PBS control group were tumor positive at the end of the study whereas only three of six mice had detectable tumors in the bscEp-CAMxCD3–treated group. With patient sample 1, mean tumor volume in bscEp-CAMxCD3 treated mice was significantly lower between days 24 and 31 compared with the PBS-treated animals (P < 0.05; Fig. 6A). With patient sample 2, a therapeutic effect was most pronounced resulting in a significant difference in mean tumor volume of the control group versus the bscEp-CAMxCD3–treated group (P < 0.05) between day 29 until end of the study on day 70 (Fig. 6B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The high in vivo potency of bscEp-CAMxCD3 is comparable with that of bscCD19xCD3, a single-chain bispecific antibody present in a monomeric form, which has shown high potency in preventing outgrowth of s.c. CD19-positive human B lymphoma in NOD/SCID mice (12). Previously, bscCD19xCD3 could not be tested for its activity against solid tumors because the long delay in s.c. B lymphoma outgrowth surpassed the limited life span of human T effector cells in NOD/SCID mice.
To evaluate the effect of bscEp-CAMxCD3 on established tumors, we had to establish a tumor model with fast outgrowth resulting in palpable tumors within 1 week after cell inoculation. The reduced efficacy seen with delayed treatments of bscEp-CAMxCD3 beyond 8 days post inoculation may have several reasons. One is the limited half-life of human T cells within NOD/SCID mice. Another one is the higher tumor volume that will increase the workload for cytotoxic T cells as well as decrease tumor penetration for the bispecific antibody construct.
BscEp-CAMxCD3 and other BiTE molecules harboring a human-specific anti-CD3 single-chain antibody are strictly dependent in their activity on human T cells. No antitumor activity is possible in the absence of human T cells (11) and no cross-reactivity with T cells from other species with the exception of chimpanzee was observed.4 In the case of the human colon carcinoma xenograft model, T cells were supplied in the form of freshly isolated, unstimulated human PBMC. The activation status of such T cells was documented earlier (12). In the present experiments, the relevant fraction of cytotoxic T cells with the CD8 phenotype were present at a low effector-to-target ratio of only 1:7 in the mixtures with tumor cells. This effector-to-target ratio may still be overestimated given that a significant proportion of CD8-positive T cells are still na
uml;ve and have no cytotoxic granules yet.
In the case of metastatic ovarian cancer tissue, human T cells could only have originated from the tumor itself. As shown in a vast body of literature (22–25) and has been shown in the present study by immunohistochemical analysis, only a few percent of cells in tumors are CD3/CD8 double-positive cytotoxic T cells. These could have either rested next to tumor cells as TILs or were present in blood of tumor vasculature. Our efficacy data suggest that these few T cells can be activated by i.v. administered bscEpCAMxCD3 against Ep-CAM–positive tumor cells, resulting in a significant reduction of tumor size. TILs were frequently described as being anergic and occasionally specific for tumor-associated peptide antigens (26, 27). Our data suggest that such tumor-specific anergic T cells may become reactivated by BiTEs and mount a tumor-specific T-cell response in addition to the BiTE-activated response. Moreover, BiTEs were shown to fully activate T cells leading to proliferation, cytokine production, and expression of cell adhesion molecules (12). In particular, secretion of tumor necrosis factor, IFN
, and chemokines by activated T cells may add to a therapeutic effect through concerted immune cell attraction and stimulation.
Cancer cells in an established tumor show high heterogeneity at various levels. They can differ with respect to target antigen expression, acquired immune escape mechanisms, growth behavior, and accessibility to drug and T cells. Likewise, the local frequency and state of activity of cytotoxic T cells can be very different in tumors. These variables of primary and metastatic human tumor tissue may reduce response rates to T cell–recruiting bispecific antibodies at all conceivable levels. By studying metastatic ovarian cancer tissue in immunodeficient mice, we attempted to take the heterogeneity of human tumor tissue into consideration. With all four patient xenografts that were accepted, we have seen that bscEp-CAMxCD3 caused a reduction in tumor size or even loss of the xenograft, whereas vehicle controls had significantly less rejections. These effects suggest that i.v. administered bispecific antibody reached tumor-resident T cells and that reactivated T cells were present in sufficient number to diminish tumor cell numbers. The current data are limited for a number of reasons. For future experiments, larger patient sample numbers are required to estimate an in vivo response rate and even longer observation periods are desirable to study the durability of responses. Likewise, larger sample numbers from one tumor are also desirable to allow inclusion of more control conditions, such as parental antibodies and BiTEs of different target specificity. Lastly, authentic tumor material from humans presents with a high degree of heterogeneity with respect to growth characteristics in immunodeficient mice. Our in vivo data are corroborated by an ex vivo study with metastatic peritoneal ovarian cancer samples (18). In this study, resected tumor samples were dissociated and resulting cell mixtures incubated with bscEp-CAMxCD3. After 24 hours, the number of Ep-CAM/CA125–positive tumor cells was determined by cell sorting and the percentage of cell tumor lysis was determined. The response rate to bscEp-CAMxCD3 was 80% (17 of 21 ovarian cancer patient samples). Again, the only source of human T cells was the tumor tissue itself.
In metastatic ovarian cancer, a high proportion of tumor cells is Ep-CAM positive (18, 19) and considerable overexpression of Ep-CAM on ovarian cancer tissue relative to normal ovarian epithelium has been reported (21). Late-stage ovarian cancer is, therefore, an interesting target indication for treatment with Ep-CAM–specific immunotherapeutics. An Ep-CAM–specific antibody of the trioma class is currently being developed as a palliative treatment of metastatic ovarian cancer. It has shown promising clinical results in the reduction of malignant ascites (28). Triomas are different from BiTEs in that this class of recombinant antibodies activates antigen-presenting cells in addition to T cells leading to a broader, but perhaps less controllable, immune response with the potential of a vaccination effect (29). BiTEs are considerably smaller than triomas, which might be beneficial with respect to tumor penetration and activation of TILs, and exclusively activate T cells in the presence of tumor target cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received 7/23/04. Revised 1/28/05. Accepted 1/28/05.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Bargou, E. Leo, G. Zugmaier, M. Klinger, M. Goebeler, S. Knop, R. Noppeney, A. Viardot, G. Hess, M. Schuler, et al. Tumor Regression in Cancer Patients by Very Low Doses of a T Cell-Engaging Antibody Science, August 15, 2008; 321(5891): 974 - 977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Asano, Y. Sone, K. Ikoma, H. Hayashi, T. Nakanishi, M. Umetsu, Y. Katayose, M. Unno, T. Kudo, and I. Kumagai Preferential heterodimerization of a bispecific diabody based on a humanized anti-EGFR antibody 528 Protein Eng. Des. Sel., July 11, 2008; (2008) gzn037v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Stork, K. A. Zettlitz, D. Muller, M. Rether, F.-G. Hanisch, and R. E. Kontermann N-Glycosylation as Novel Strategy to Improve Pharmacokinetic Properties of Bispecific Single-chain Diabodies J. Biol. Chem., March 21, 2008; 283(12): 7804 - 7812. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Amann, K. Brischwein, P. Lutterbuese, L. Parr, L. Petersen, G. Lorenczewski, E. Krinner, S. Bruckmeier, S. Lippold, R. Kischel, et al. Therapeutic Window of MuS110, a Single-Chain Antibody Construct Bispecific for Murine EpCAM and Murine CD3 Cancer Res., January 1, 2008; 68(1): 143 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Stork, D. Muller, and R. E. Kontermann A novel tri-functional antibody fusion protein with improved pharmacokinetic properties generated by fusing a bispecific single-chain diabody with an albumin-binding domain from streptococcal protein G Protein Eng. Des. Sel., November 3, 2007; (2007) gzm061v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Muller, A. Karle, B. Meissburger, I. Hofig, R. Stork, and R. E. Kontermann Improved Pharmacokinetics of Recombinant Bispecific Antibody Molecules by Fusion to Human Serum Albumin J. Biol. Chem., April 27, 2007; 282(17): 12650 - 12660. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Hammond, R. Lutterbuese, S. Roff, P. Lutterbuese, B. Schlereth, E. Bruckheimer, M. S. Kinch, S. Coats, P. A. Baeuerle, P. Kufer, et al. Selective Targeting and Potent Control of Tumor Growth Using an EphA2/CD3-Bispecific Single-Chain Antibody Construct Cancer Res., April 15, 2007; 67(8): 3927 - 3935. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research |
). For bscEp-CAMxCD3–treated animals, individual tumor growth curves are shown (
). Points, means; bars, SD. Arrows, bscEp-CAMxCD3 treatment time points.
) via tail vein injection on days 6 to 10, 20 to 24, and 34 to 38 (A); or on days 3 to 7 and 17 to 21 (B). Points, mean values of tumor growth curves (n = 6); bars, SD. Arrows, bscEp-CAMxCD3 treatment time points. Asterisks, readings that were statistically significant (P < 0.05) from readings of the PBS group.