Comparison of the biologic effects of MA5 and B-B4 monoclonal antibody labeled with iodine-131 and bismuth-213 on multiple myeloma
Version of Record online: 12 FEB 2002
Copyright © 2002 American Cancer Society
Supplement: Eighth Conference on Radioimmunodetection and Radioimmunotherapy of Cancer
Volume 94, Issue Supplement 4, pages 1202–1209, 15 February 2002
How to Cite
Supiot, S., Faivre-Chauvet, A., Couturier, O., Heymann, M. F., Robillard, N., Kraeber-Bodéré, F., Morandeau, L., Mahé, M. A. and Chérel, M. (2002), Comparison of the biologic effects of MA5 and B-B4 monoclonal antibody labeled with iodine-131 and bismuth-213 on multiple myeloma. Cancer, 94: 1202–1209. doi: 10.1002/cncr.10286
- Issue online: 12 FEB 2002
- Version of Record online: 12 FEB 2002
- Manuscript Accepted: 14 NOV 2001
- Manuscript Received: 31 OCT 2001
- multiple myeloma;
- cell cycle
Using a specific monoclonal antibody (MAb), B-B4, coupled to bismuth-213 (213Bi) by a chelating agent (CITC-DTPA), the feasibility of alpha-radioimmunotherapy (RIT) for multiple myeloma (MM) has been demonstrated previously.
In this study, the two MAbs tested, MA5 and B-B4, target the epithelial antigens Muc-1 and syndecan-1, respectively, which are both expressed by MM cell lines. Antibody characterization was evaluated by flow cytometric analysis of normal and tumoral hematopoeitic cells of MM patients as well as immunohistochemical tests of normal, nonhematopoetic tissues. Radiobiologic effects were evaluated for 213Bi- and iodine-131 (131I)–labeled antibodies. We assessed in vitro mortality (thymidine incorporation, MTT, and clonogenic assays) and cell cycle modifications with propidium iodide staining. These tests were performed on MM cell lines until 120 hours postirradiation at several time points, using radiolabeled antibody concentrations ranging from 0.5 to 20 nM and specific activities ranging from 240 to 1200 MBq/mg of MAb.
MA5 stained all MM cells in only 50% of patients, whereas B-B4 recognized all MM cells in all patients. B-B4 principally showed hepatic, pulmonary, and duodenal staining, whereas MA5 marked renal and pulmonary tissues. RIT with 213Bi-B-B4 induced specific mortality and G2/M phase cell cycle arrest, which depended on the concentrations and specific activity. For 213Bi-MA5, this arrest appeared at concentrations above 10 nM, an amount fivefold higher than that required with B-B4. This difference was also found in thymidine incorporation assays. Furthermore, with 213Bi-B-B4, the arrest at the G2/M phase appeared quickly, within 24 hours after irradiation, and affected up to 60% of the cells (for 20 nM of 213Bi-B-B4 at 1,200 MBq/mg). Conversly, 131I-B-B4 had a very limited effect on cell mortality and did not induce any cell cycle arrest.
The results of this study show that B-B4 might be the more effective therapeutic antibody and suggest that alpha-RIT might be more suitable than beta-RIT for treating single-cell tumor models. Thus, these findings set the stage for the beginning of clinical trials using alpha-emitter–radiolabeled B-B4, with special attention paid to hepatic, pulmonary, and intestinal side effects. Cancer 2002;94:1202–9. © 2002 American Cancer Society.
Multiple myeloma is responsible for 1% of cancer-related deaths and represents 10% of malignant blood diseases. For the vast majority of patients, multiple myeloma is an incurable disease with a very poor prognosis and a median survival of about 30 months.1
Radioimmunotherapy (RIT) has proved efficient in the treatment of radiosensitive tumors such as non-Hodgkin lymphoma, but it is less suitable for more radioresistant solid tumors.2 Several methodologic approaches intended to improve RIT efficacy are under evaluation. One involves optimizing the use of radionuclides—for instance, by replacing iodine 131 (131I), which has been widely used until now, with a more suitable radioisotope, such as yttrium 90. This isotope is particularly useful for lymphomas, which are often treated only when the tumor mass has become large. In the case of multiple myeloma, cells are found either isolated in bone marrow or in small clusters. Among the different beta-emitters available, it would be logical to choose 131I for the treatment of multiple myeloma because the energy of the beta particles emitted is distributed within a range of about a millimeter, whereas more energetic beta-emitters (yttrium-90, rhenium-186 and -188) deliver their energy over a greater range. However, the notion of energy deposition is not the only parameter to consider; high gamma emission, physical half-life, or the coupling technique used can make the choice of a suitable radionuclide more complicated. For instance, in the case of isolated tumor cells, alpha particles have a theoretic advantage over beta particles because of their high linear energy transfer and shorter range of action. Presumably, cell destruction would be more selective and irradiation less harmful to adjacent tissues. Various studies have already demonstrated the efficacy of alpha-emitters for cell mortality,3 particularly in a multiple myeloma model.4 However, a review of the literature clearly showed that beta particles are most often used. Thus, it appeared that an in vitro study would be useful to determine whether alpha particles actually have a theoretic advantage over beta particles.
Improvement could also be made in the vector used, particularly within the parameters governing antigen-antibody interaction in the multiple myeloma model. Among the different antigens concerned, syndecan-1 or CD 138 is present in 100% of myeloma cells.5–8 This antigen corresponds to a linking protein in the extracellular matrix. B-B4 antibody of murine origin recognizes the external part of this antigen, showing a good affinity constant of around 1 nM.4 Another antigen, Muc1, is rather well represented on the surface of myeloma cells.9–11 This antigen is a mucin (a type I transmembrane protein) composed of a single extracellular domain made up of repeated tandem domains.12 The MA5 antibody, which recognizes this antigen, is used for the detection of breast tumor cells.13 Thus, we decided that it would be interesting to compare the efficacy of these two antibodies as radioimmunotherapy vectors.
The current study analyzed the effects of bismuth-213 (213Bi)– or 131I–labeled MA5 and B-B4 antibodies on cell survival and the cell cycle, to compare not only the two radionuclides but also the two antibodies likely to be used in RIT protocols for multiple myeloma. This study also considered the possible combined use of two vectors in order to reduce the nonspecific effects of uptake by nontarget cells.
MATERIALS AND METHODS
Multiple Myeloma Model
Human myeloma cell line U266 was obtained from the American Type Culture Collection (Rockville, MD) and cultured in RPMI-1640 supplemented with 10% heat-inactivated fetal calf serum (FCS) and 2 mM L-glutamine in humidified air at 37 °C and 5% CO2. Forty-eight hours before each assay, cells were transferred into fresh culture medium, and density was adjusted to 2.5 × 105 cells/mL.
Three monoclonal antibodies (Mabs) were used in the study: B-B4 (Diaclone Research, Besançon, France);6 MA5, kindly provided by Dr. D. Goldenberg (Garden State Cancer Center, Belleville, NJ); and immunoglobulin G (IgG) 134, kindly provided by Immunotech Coulter, Marseille, France. The IgG did not react with myeloma cells and was used as a nonspecific antibody.
Labeling of Monoclonal Antibodies
The 213Bi generator was kindly provided by the European Institute for Transuranium Elements.
B-B4, MA5, and nonspecific IgG 134 MAbs were conjugated to the bifunctional chelating agent CITC-DTPA (pentaacetic triamine diethylene p-aminobenzyl acid), which was synthesized in our laboratory using the procedure described by Brechbiel et al.14
Labeling Monoclonal Antibody with Bismuth-213
Radiolabeling was obtained by incubating, for 15 minutes at room temperature, 10–100 μg of MAbs-CITC-DTPA with a buffered, calibrated solution of freshly eluted 213Bi, according to the method described by Kaspersen et al.15
B-B4 was labeled with 131I (Cis Biointernational, Gif sur Yvette, France) by the iodogen method.16 Radiochemical purity, checked by ITLC-SG using 10% TCA as solvent, was usually > 90%.
Histology and Immunoperoxidase Staining
First, the best dilution of B-B4 was determined (1:400, with microwave pretreatment 3 × for 5 minutes each) on frozen and formalin-fixed breast tumors, specifically, well-differentiated infiltrating duct breast tumors and mastosis. For MA5, the suitable dilution was 1:500 (without microwave pretreatment). Next, we studied the reactions of B-B4 antibody with different normal human tissues: liver, kidney, lung, intestinal tract, and heart. All immunohistochemical reactions were performed on deparaffinized sections (4 μm) of formalin-fixed tissue samples. Sections were rehydrated through a series of staged ethanol solutions and incubated in distilled water containing 0.3% hydrogen peroxide to inhibit endogenous peroxidase. After washing, sections were incubated for 30 minutes at room temperature with primary antibody (B-B4 and MA5). Antibody-antigen complexes were revealed by avidin-biotin technique (Dako kit 0675, Trappes, France). The reaction product was visualized by incubation in AEC (3-amino-9-ethylcarbazole) (Dako K 0697) for 10 minutes and then washed with water before counterstaining with Mayer hematoxylin. The slides were washed before dehydration and coverslip mounting. B-B4 and MA5 antibodies were suppressed in immunostaining for negative controls. A case of breast tumor was used as a positive control. The number of positive cells for B-B4 or MA5 (determined by systematic scanning) as well as the intensity and the location of the immunostaining (membrane cells or cytoplasm) were noted.
For each assay, the experiment was done in quadriplicate and repeated at least twice.
3H-Thymidine proliferation assays, limited dilution analysis
These were described in our previously published report.4
The cells were washed, counted, resuspended in RPMI-1640 medium containing 10% FCS at 104 cells/mL, and plated in 96-well microtiter plates at 100 μL/well. MTT reagent (5 mg/mL in phosphate-buffered saline [PBS]) was added, and cells were incubated at 37 °C for 1 hour. Then 100 μL isopropanol was added to each well, and mixed by whirling to solubilize the formazen product. Finally, the plates were read for absorbance in a spectrophotometer at 550 nm.
Measurement of cell cycle distribution
To perform this measurement, 1 × 106 cells were pelleted, resuspended in 0.2 mL PBS, and fixed by the addition of 2 mL of ice-cold 70% ethanol/30% PBS. Fixed cells were pelleted, vigorously resuspended in PBS, and incubated for 30 minutes at 37 °C with 100 μg/mL RNAse and 40 μg propidium iodide. The fluorescence of the stained cells was analyzed using a FACScan (Becton Dickinson, San Jose, CA). Data were analyzed with ModFit LT 2 (Becton Dickinson).
Analysis of the Characteristics of MA5 and B-B4 Antibodies
Flow cytometry was used to analyze the uptake of labeled antibodies on normal plasma cells and myeloma cells of patients and on myeloma cell lines (Table 1A). MA5 antibody recognized 43% (3 of 7) of normal plasma cells; 56% (17 of 30) of myeloma cells from patients with multiple myeloma or monoclonal gammapathy of undetermined significance (MGUS); and 100% of the cell lines studied, including U266. Scatchard tests showed that the affinity of MA5 on U266 cells was 9 nM (moderate affinity). The characteristics of B-B4 antibody had been determined previously. No cellular internalization of this antibody was found by the method described by Preijers et al.17
|A||Normal plasma cells||3/7 (43%)||100%|
|MGUS and MM patient cells||17/30 (56%)||100%|
|MM cell lines||3/3 (100%)||100%|
|Smooth muscle of hilus vessels||0||0|
|Bronchial epithelium||M||M weak|
Tissue Uptake of B-B4 and MA5
Immunohistochemical analysis of fresh tissue sections revealed membrane uptake of B-B4 on renal urothelial cells, hepatocytes, biliary epithelium, lungs (alveoli, bronchi, and bronchial glands), and duodenal glands (Table 1B). Cytoplasmic labeling was observed on renal tubules and myocardial fibers. MA5 immunolabeling showed renal membrane uptake on tubes and urothelial cells. MA5 labeled the biliary epithelium, and uptake was also noted in the lung (alveoli, bronchi, and bronchial glands). Weak cytoplasmic labeling was observed in the smooth muscles of renal vessels and in myocardial fibers.
B-B4 Antibody Radiolabeled with Alpha-Emitters Caused Blockage of Myeloma Cells in G2/M Phase
The effects of B-B4 antibody labeled with 213Bi at a specific activity of 400 MBq/mg were observed on the U266 line (Fig. 1). Cells were blocked in G2/M phase within 24 hours after the antibody was introduced into the medium. This blockage remained stable or increased slightly after 48 hours until the end of analysis (96 hours). The degree of blockage depended on the antibody concentration (67% of cells for 10 nM vs. 54% for 1 nM). For a concentration of 0.5 nM, the initial rise at 24 hours seemed to disappear after 48 hours. This increase in the proportion of cells blocked in G2/M phase was also observed for uptake of a nonspecific antibody labeled with the same radionuclide (data not shown).
G2/M Blockage Was Related to Myeloma Cell Mortality
The number of cells blocked in G2/M phase was considered in terms of the different parameters of cell survival. For each antibody concentration studied, the percentage of cells blocked in G2/M phase at 72 hours was related to cell survival, as measured by tritiated thymidine incorporation, MTT metabolization, and the limit dilution test (cell mortality was expressed by a semilogarithmic coordinate). A positive correlation was found between the percentage of cells blocked in G2/M phase and cell survival as measured by thymidine incorporation (R2 = 0.92), MTT metabolization (R2 = 0.94), and clonogenic survival (R2 = 0.98).
Comparison of B-B4 and MA-5 at the Same Specific Activity
Our group previously determined that the optimal conditions for the use of B-B4 antibody are a concentration of 1–5 nM at a specific activity of less than 400 MBq/mg or at added volumic activities of 4–20 MBq/L to ensure that the effect is relatively specific. For a specific activity of 240 MBq/mg, cell survival was determined by tritiated thymidine incorporation 48 hours after incubation with B-B4-213Bi, MA5-213Bi, and a 213Bi-labeled nonspecific antibody (Fig. 2). Marked inhibition of cell growth (about 25.7% of surviving cells) was observed for an added volumic activity of 40 MBq/L, and inhibition remained relatively constant (17.4–34.2%) for higher activities up to 400 MBq/L. For the same added activities, cell survival was significantly lower (3.7–8.2%) with B-B4 than with MA5. Cell proliferation was also reduced after nonspecific irradiation, showing a reduction of 46.2–74.8% for low and up to 27.8% for high added activities. Blockage in G2/M phase was then determined for the same antibodies at a specific activity of 240 MBq/mg (Fig. 3). For B-B4, blockage occurred early (after 24 hours) and remained stable at a high level (29.1–44.2%). Blockage occurred later for MA5 (after 72 hours) and at a lower level (18–33.2%). Blockage related to nonspecific irradiation remained stable (7.7–16.2%).
Effects of Beta Radioimmunotherapy
MTT metabolization studies were performed on U266 cells to compare the effects of 131I and 213Bi labeling of B-B4 antibody (Fig. 4). Reduction of cell viability was very slight (less than 6% mortality) for low specific activities of 131I and about 16% for high activities. However, 213Bi-B-B4 showed a more marked effect on cell viability for both low and high added activities (28–43% and 47–54%, respectively). Cell cycle analyses were then conducted to determine the percentage of cells blocked in G2/M phase with 131I and 213Bi. After incubation with 131I-labeled antibody, blockage after 72 hours was moderate (17–23%) for low added activities (Fig. 5) and more marked (20–27%) for 800 MBq/L. The same antibody, when labeled with 213Bi, produced earlier and higher blockage (48–62%) both for low and high added activities. Only 9–13% of nonirradiated cells were in G2/M phase. Finally, cell survival curves were determined by a limit dilution technique (Fig. 6). For 131I-labeled B-B4 antibody, mortality was 43.1–55.5% for high added activities. For the same antibody labeled with 213Bi, more than 99.99% of irradiated cells were destroyed when comparable added activities were used.
Comparison of B-B4 and MA5 Antibodies
B-B4 antibody showed better affinity (0.9 nM) than MA5 antibody (9 nM) on multiple myeloma cell lines. B-B4 recognized 100% of cells of patients with multiple myeloma, whereas MA5 recognized only 56%. Thymidine analysis indicated that cell mortality was greater with 213Bi-B-B4 than with 213Bi-MA5 (3.7–8.2% of survival vs. 17.4–34.2%). Moreover, a large part of the effect observed for MA5 was apparently related to nonspecific irradiation, since cell survival at high added activities was comparable to that for nonspecific antibody. Moreover, there was a greater accumulation of cells in G2/M phase for B-B4 (29.1–44.2%) than MA5 (18–33.2%). Thus, MA5 seems less capable than B-B4 of targeting and destroying multiple myeloma cells. B-B4 does not bind to soluble syndecan 1, whereas a part of MA5 may bind to soluble MUC1,10 which could distort results for affinity and cell mortality. However, MA5 expression may be induced by dexamethasone,11 which would presumably provide better efficacy for an association with corticosteroids.
Immunohistochemical studies showed B-B4 uptake on lung parenchyma, intestinal epithelium, kidney, and heart. These potentially toxic effects need to be carefully watched for during the clinical development of RIT of myeloma. Various technical parameters related to the denaturation-renaturation conditions of fixed cells can cause false-positive labeling due to cryptic antigens. As B-B4 is not internalized, the cytosolic uptake observed intensely in the heart and weakly in renal epithelium has no clinical implications. However, potential toxicity cannot be studied in the mouse, as there is no multiple myeloma tumor model for this species, but only subcutaneous tumor models that are not representative of the clinical occurrence of multiple myeloma. Only imaging data can indicate whether the biodistribution of radiolabeled antibody confirms the experimental results of immunohistochemistry. If toxicity (notably in the liver) proves to be a limiting factor for B-B4 efficacy, MA5 may be an interesting alternative; it appears to be less toxic to the liver and duodenum. Moreover, an association of these two radiolabeled antibodies could allow the tissue toxicity of B-B4 to be reduced. Many other malignancies express Muc-1; thus, it may be better suited to some other histologic types of cancer.
Comparison of Iodine 131 and Bismuth 213
213Bi-B-B4 showed markedly greater efficacy than 131I-B-B4 for cell viability (MTT test, Fig. 5A), blockage in G2/M phase (Fig. 5B), and clonogenic survival (Fig. 5C). Thus, 213Bi, an alpha-emitter, seems more suitable than 131I for antibody labeling in RIT of myeloma. The modest effect of 131I RIT could not have been due to deiodination of the antibody, since no internalization of B-B4 was detected. Ninety-five percent of the disintegration of the 131I bound to the antibody occurs within a radius of 1000 μm.18 If myeloma cells are at a distance of 100 μm and in a concentration of 5.105 cells per mL, this energy is delivered far from the tumor target. Thus, RIT is most suitable for tumors with 105 to 107 cells.19 As the maximal range for 213Bi is 80 μm, energy is delivered close to the cell nucleus. Moreover, the linear energy transfer of alpha particles is around 100 keV/μm, but only 0.2 keV/μm for yttrium-90, the most energetic beta-emitter.20 These results are consistent with those for RIT cell models in which a similar rate of survival was found for lymphoma lines treated with yttrium-90.21 A microdosimetric study comparing alpha and beta particles would be necessary to calculate the dose delivered to the nucleus of each cell and relate it to cell survival. In vivo application of this cellular model should determine whether alpha-emitters are more effective than beta-emitters. In fact, the toxic effects of alpha particles on normal bone marrow cells may prevent the delivery of a tumoricidal dose. Moreover, beta particles are likely to be more destructive in the context of microcell clusters within bone marrow than to isolated cells in vitro.
The choice of radionuclide will be an important consideration in the clinical development of RIT for multiple myeloma. The fact that B-B4 antibody is not internalized excludes the use of Auger electron-emitters, such as 125I. Among beta-emitters, yttrium-90 particles are more energetic than those of other radioisotopes, such as 131I. Thus, nonspecific effects may be increased by the delivery of energy at some distance from the target cell, in which case yttrium-90 might be less effective than 131I. The proximity of a cyclotron is not necessary for the alpha-emitter 213Bi, which can be produced by chemical generators. However, short physical half-life makes it less useful for clinical applications than other alpha-emitters, such as astatine-211 or actinium-225.
G2/M Phase Blockage
Alpha and beta RIT of multiple myeloma causes an accumulation (up to 67%) of U266 cells in G2/M phase. After radiation-induced DNA damage, many cells are blocked at the G1/S control point of the cell cycle.22 This point seems to be closely controlled by p53 protein. In multiple myeloma, p53 protein seems to be mutated rather rarely in the early stages of the disease, showing mutation rates of less than 20%. In advanced stages, mutations of p53 protein alter its capacity to bind to DNA and control the cell cycle.23 Inactivating point mutations of this protein occur in the U266 line24 and prevent it from playing an inhibitory role at the G1/S control point.This absence of G1/S control explains why cells are blocked in G2/M phase, where p53 protein appears to have little influence. Thus, in vivo RIT would show different efficacy for early and advanced forms of multiple myeloma.
A relation exists between the dose delivered to cells and the interval before blockage occurs, as well as blockage amplitude. Higher added activities produce earlier and more marked blockage. Blockage is closely correlated with cell survival, as measured by the incorporation of tritiated thymidine, MTT metabolization, and clonogenic survival. The G2/M phase, which follows the S phase of DNA synthesis, is used by the cell to repair any replication errors that occurred during synthesis. The material required for mitosis is then developed before mitosis finally takes place. The period of prolonged blockage in G2/M phase could be used by the cell for more thorough repair of radiation damage experienced during S phase. However, the blockage period does not appear to be correlated with variations in radiosensitivity.25 In fact, cells blocked in G2/M phase show markedly greater sensitivity than those synchronized in the other phases of the cycle.26 Radioimmunotherapy of multiple myeloma causes intense, prolonged blockage, depending on the dose delivered to cells. Blockage maintains the cells in a more radiosensitive phase, which accounts for the greater cell destruction. Thus, Ning has shown that caffeine-induced inhibition of G2/M blockage reduces cell death after irradiation and that preincubation with nocodazole (which prolongs G2/M blockage) causes even greater cell death.27
Clinical Use of Radioimmunotherapy
Various studies have demonstrated the efficacy of beta RIT in non-Hodgkin lymphoma.28 Alpha RIT is currently being studied for application to isolated tumor cells, such as acute myeloid leukemia cells.29 Our results indicate that B-B4 antibody has better recognition and destructive efficacy for tumor cells than MA5 antibody. Moreover, B-B4 was more destructive when labeled with 213Bi than with 131I. This would seem to confirm our initial hypothesis that alpha particles are more capable than beta particles of destroying isolated tumor cells.
It would be useful to study associations of alpha RIT with agents such as melphalan or prednisone (still the reference treatment for patients older than 65 years)30 or with other substances such as cyclophosphamide, anthracyclines, vincristine, or interféron-α-2b (which has provided better response rates but no definite improvement in survival).1 The blockage observed in G2/M phase suggests that alpha RIT could be associated with chemotherapeutic agents that prolong the blockage of tumor cells, such as doxorubicin or paclitaxel (which also seems to have radiosensitizing properties).21 In cases of marked hematologic toxicity with RIT, it would be advisable to study an association with effective but nonmyelotoxic agents, such as thalidomide.31 Our ongoing studies will try to specify the possible roles of these drugs in RIT. Otherwise, the use of B-B4 antibody radiolabeled with myeloablative doses might be comparable to that of whole-body irradiation, but with much more selective destruction of myeloma cells and milder extramedullary toxicity for patients younger than 65 years. It has been demonstrated in these patients that more intense therapy involving whole-body irradiation, chemotherapy, and autograft of stem cells improves results.32
The results reported in this article are promising for alpha RIT with B-B4 antibody. The continued overexpression of syndecan-1 in progressive forms of multiple myeloma suggests that B-B4 labeled with alpha-emitters would be effective even in the treatment of refractory forms of the disease. On the basis of these encouraging results, a Phase I/II trial will soon be carried out to develop this therapeutic approach for clinical application.
The authors thank Catherine Saï-Maurel, Jean Le Boterff, and James Gray for their technical assistance and M. Bardiès for helpful discussion.
- 5B-B2 and B-B4, two new Moabs against secreting plasma cells. In: Schlossman S, editor. Proceedings of the fifth international workshop and conference on human leukocyte differentiation antigens, Boston; 1995:174., , , .
- 29Phase I trial of targeted alpha-particle therapy for myeloid leukemias with bismuth-213-HuM195 (anti-CD33). American Society of Clinical Oncology, 1999:Abstract 22., , , , , , , , , , .