- - - - - - - - - - - - - - - SYMPOSIUM ARTICLE - -
Radio-immunotargeting in experimental animal models of intraperitoneal cancer J. G. FJELD" I. VERGOTE',2, L.
DE Vos' AND
K.
NUSTAD'
From the 'Central Laboratory, and the 2Department of Gynecologic Oncology, the Norwegian Radium Hospital, Oslo, Norway
Acta Obstet Gynecol Scand 1992: 71 Suppl 155: 105-111 The field of immunotargeting, and the challenges met when this technique is applied in experimental animals or in patients, are reviewed. Even with highly specific monoclonal antibodies, non-specific uptake in normal tissues and high background level of unbound radioactivity in blood and extravascular body fluids remain significant problems. Further experimental work in animal model systems is needed to bring this technique from the state of being an experimental method, with limited clinical application, to a routine diagnostic or therapeutic method. Different animal models are available, and their potential for elucidation of the various methodological problems in radio-immunotargeting are discussed in the present paper. In our laboratory. two intraperitoneal models were devised. having relevance for gynecologic and other forms of intraperitoneal malignancies. These models were elaborated with special emphasis on the possibility for exact measurement of important parameters in immunotargeting reactions. In the first model, hybridoma cells are inoculated intraperitoneally to mimic intraperitoneal carcinomatosis, and the monoclonal antibody produced by the hybridoma is used as serum tumor marker. In the second model the tumor cells are contained within intraperitoneally implanted micropore chambers, resembling a localized tumor. An artificial tumor like this allows control with the antigen load in the target, and measurement of the concentration of the injected antibody in the fluid within the target.
Radio-immunotargeting The basic idea of immunotargeting is to exploit the exquisite specificity of antigen-antibody reactions in the diagnosis or treatment of cancer. The lack of specificity for the pathologic cells is a general problem with systemic therapy, resulting in detrimental side effects on normal tissues. The idea to conjugate radio-isotopes with a tumor specific antibody was pioneered almost 40 years ago with polyclonal antibodies (1). With the polyclonal antisera affinity purification procedures were necessary to obtain antibody preparations of high purity. However, because
the fraction of specific immunoglobulin in antisera is low, it was difficult to collect a sufficient quantity of tumor-specific immunoglobulins. Like many other fields in immunology, the immunotargeting technique did benefit when homogeneous immunoglobulin preparations of specified antigen specificity, i.e. monoclonal antibodies (MoAb), were made available in gram quantities by the hybridoma technology (2). Many of the problems with in vivo targeting with polyclonal antisera were presumed to be due to factors specific for the polyclonicity. Consequently, the hybridoma technique was suggested to revolutionize both cancer diagnosis (3-5), and therapy (6, 7).
Abbreviations used: MoAb, monoclonal antibody(-ies); HAMA, human anti mouse antibody(-ies). Acta Obstet Gynecol Scand Suppl ISs
106
J. G. Fjeld et al.
Are labeled MoAb 'magic bullets'?
While an overall improvement was obtained with MoAb, it was soon discovered that also MoAb gave a series of problems when applied for immunotargeting. Therefore, considerable effort was still needed to explore and understand the fundamental mechanisms of the shortcomings of MoAb in vivo, with the purpose of finding techniques to overcome the problems. Heralding the labeled MoAb as 'magic bullets' or 'missiles' may mislead the reader to think of the targeting reaction as an active, target-seeking process. This is far from correct, and may explain some of the disappointment expressed when the low tumor uptake of antibodies relative to injected dose is presented. The physiochemical laws for antigen-antibody reactions seem frequently to be forgotten in studies on antibody-antigen reactions in vivo. When a tumorspecific labeled MoAb is injected i.v., it reaches the tumor cells via (i) the blood circulation, (ii) extravasation across the blood vessel wall to the interstitium, (iii) transport with the interstitial fluid to the specific antigen in the tumor (8). The transport mechanisms are convection and diffusion. The immunological binding of antibody to the antigen maintains the antibody concentration gradient towards the tumor antigen. The efficacy of the transport of a certain injected dose from blood to antigen is hampered by (i) antibody excretion, (ii) immunoglobulin catabolism by the reticuloendothelial system (9), (iii) non-specific binding to normal tissues or to soluble molecules, and (iv) specific binding to free tumor antigen in the body fluids. Moreover, the tumor tissues have physiologic barriers responsible for the poor antibody uptake. The tumor tissue blood supply is heterogeneous, and may be poor in some areas, the interstitial pressure in tumors is frequently elevated, and the antibody transport distances in the interstitium are often large (10). Several methods such as radiation, heat and vasoactive drugs, have been tried out to improve the tumor blood flow and thereby increase the antibody uptake (11-15). However, the effect was always of short duration. This low access to the tumor antigen is in contrast to an in vitro antibody-antigen reaction, where all the antibody administered is exposed to the antigen, except for a minor fraction of nonspecifically bound antibodies. Thus, with high affinity antibodies and a concentration close to antigen excess, nearly 100% of the added antibody dose will be bound in vitro, while only a minor fraction of the same antibodies are bound in vivo. The vast majority of animal experiments have been carried out with nude mice carrying relatively large subcutaneous xenografts, and the results in Acla Obstet Gynecol Scand Suppl/55
these experiments are regularly far better than in patients. Considering the above-mentioned basic physiological and physiochemical mechanisms that affect the targeting result, it is easy to explain the low tumor uptake in large animals or patients (16). When the antibody distribution volume is large compared with the tumor volume, the fraction of injected dose reaching the tumor will be significantly lower than in a small mouse carrying a relatively large xenograft. Human Anti-Mouse Antibodies (HAMA)
The xenogeneic transfer of murine MoAb is a stimulus for anti-MoAb production by the patients (6, 17). This HAMA production has three aspects. First, HAMA may initiate dangerous allergic reactions. However, this has not been a significant problem so far, partly due to the precautions taken against this potential danger in clinical studies. The second aspect is that HAMA may interfere with the tumor uptake of MoAb. Thirdly, since HAMA may have anti-idiotypic activity (18, 19), these antibodies may result in the production of anti-anti-idiotypic antibodies, i.e. anti-tumor antibodies (18, 20-23). Thus, one might speculate that HAMA in some cases could be beneficial for the patient. Trying to avoid the HAMA problem, mouse/human chimeric antibodies have been produced by transfection of a non-producing mouse myeloma line with a chimeric gene construct composed of murine variable regions and constant regions from Man (23). Moreover, human MoAb were produced by fusing lymphocytes from glioma or bronchial carcinoma patients with a human myeloma cell line (24). Determination of the optimal dose of labeled MoAb
As discussed by others (25), conflicting results have been obtained in studies on the optimal MoAb dose to be administered (26-29). According to the mass action law, an increase in antibody dose implies an increase in uptake. Moreover, this increase in antibody uptake should vary with the relative concentration levels of antigen and antibody. The optimal dose is presumably the dose resulting in a tumor uptake corresponding to the upper part of the sigmoid binding curve obtained when plotting the uptake of different amounts of antibody to a constant amount of antigen. When approaching antibody excess the net effect of further increase in antibody dose will be minimal, and contribute mainly to the background of unbound antibodies.
Experimental radio-immunotargeting Table I. Intraperitoneal xenograft models Ovarian cancer
NIH-OVCAR-3 HEY and SKOV-3 SW626 HTB77IP3 Fresh tumor cells
Hamilton, 1984 (39) Baumal, 1986 (40) Ripamonti, 1987 (41) Wahl, 1989 (42) Ward, 1987 (46)
Colon cancer
LS174T HT29 RPMI4788
Esteban, 1990 (43) Ripamonti, 1987 (41) Naomoto, 1987 (44)
Mesothelioma
H-MESO-1
Reale, 1987 (45)
Radiolabels, and the effect of labeling on the pharmacokinetics The radiolabel and the labeling technique may change the antibody biodistribution and the specific target uptake. Various radio-isotopes are utilized, selected on the basis of their radiation characteristics. 211Astatine, lJOYttrium, 67Copper, I88Rhenium, 32phosphorous and 212Bismuth are radiometals with theoretical potential for radiotherapy, while IIIIn_ dium and 99mTechnetium are imaging labels. Radioiodines such as 1251 and 131 1 were used in numerous experimental biodistribution and targeting studies, and 131 1 was used both for clinical imaging (30), and for therapy (31). A third radio-iodide, the shortlived isotope 1231, has good imaging characteristics (5, 32). Labeling with iodine is easy and convenient, and applicable to all proteins with accessible tyrosine residues. However, loss of the iodine label due to dehalogenation has been reported (33, 34). The free radio-iodine is then incorporated in the thyroid, or excreted, mainly through the kidneys. Novel and improved labeling methods have been developed during recent years, improving the iodination of antibodies (35), or making it possible to utilize a series of new radio-isotopes. Recently, the method for protein coupling with technetium has been improved significantly (36). A main problem with the lllIn-labeled antibodies is that they are sequestered in the liver. Modification of the oligosaccharide moieties of MoAb was demonstrated to be preferable to modification of aminoacids, when conjugating MoAb with 1I1In and 1251 (37). Labels or labeling procedures may also affect the immunological binding characteristics of the antibody. Therefore, preclinical immunoreactivity and biodistribution experiments should be carried out with each new radiolabel and coupling method.
107
Experimental animal model systems The nude mouse xenograft system The nude mouse xenograft tumor model is the most frequently utilized animal model. In this model, different human tumors are transferred to immunodeficient mice, either by s.c. implantation of fresh tissue specimens, or by s.c. injection of a small volume of suspended cells from a human cell line. After a variable period of time, the grafted cells have grown to a tumor of about 1 gram. This model mimics the real tumor situation fairly well, but has several limitations. The athymic mice must be kept under carefully controlled and rather expensive housing conditions in order to protect them from infections, and effects of the immune system on tumor growth cannot be studied. Moreover, the grafted tumor cells must grow to a tumor of appropriate size, and the growth rate and tumor size will vary from animal to animal. Lastly, the antibody binding capacity is unknown, and the value of this parameter may not be proportional to the tumor size (38). Transperitoneal spread and intra-abdominal carcinomatosis are common in patients with ovarian and colorectal cancer. Hence, intraperitoneal tumor models mimic more closely the spread pattern of ovarian and colorectal cancer than do subcutaneous models. A variety of experimental i.p. tumor models have been applied to elucidate the biology of intraabdominal carcinomatosis, and to evaluate the potential benefit of radio-immunotherapy. Recently, i.p. diffuse tumor spread has been described in nude mice, using human ovarian (39-42), colorectal (41, 43,44) and malignant mesothelioma cell lines (45), while Ward et al. (46) established i.p. xenografts using fresh primary human tumor material (Table I). Table II. Intraperitoneal xenograft versus hybridoma as tumor model Xenograft Carcinoma Human derived Uniform intra-abdominal spread Median survival Blood tumor marker Cost price Growth from culture cells Tumor take Immunocompetent host a
6
Hybridoma
++ ++
+
++ 10 days
25-43 days (+)' High ++/_6
Low ++
+
++
++
++
Only for the NIH-OVCAR-3 model (see main text). Some models use in vitro cultured cells, while in other models, tumor transplants are used. ACla Obstet Gynecot Scand SIIppl155
108
J. G. Fjeld et al.
Like the subcutaneously xenografted mice, all these intraperitoneal models have several limitations (Table II). Cell lines derived from human tumors may differ from the original tumor cells due to selection in the establishment of the cell line or modulation of the biological features of the cell line by the abnormal condition associated with its propagation. Therefore Ward et at. (46) proposed the use of intraperitoneal xenografts from primary tumor material, but passage time varied in this model from 1 to 7 months, and only three out of ten tumors were easily passageable. Intraperitoneal tumor growth has also been reported using murine ovarian teratoma and ovarian reticulum cell sarcoma cell lines, or tumor induction by carcinogens (47-51). These models use murine or rat tumors and the tumor type does not represent a common human cancer. In both these and the intraperitoneal xenograft models it is difficult to measure the delay in tumor growth, because they often provide multiple metastases at sites where they are difficult to measure. The NIH-OYCAR-3 xenograft is the only i.p. tumor model with a serum marker, i.e CA 125 (52). However, in this study as much as 50!l1 blood had to be sampled from the orbital plexus, and the serum CA 125 level was elevated in all animals after at least 12 days. Artificial targets in immunocompetent animals For exploration of all the problems that are not directly related to the tumor tissue, artificial targets are well suited. Such targets give better control with the antigen burden, so that systematic comparative studies are easier to perform. In one such model system the target consisted of antigen-coated agarose beads that were injected i.v. into rats, being sequestered in the lung capillary bed (53). In a second model, antigen-coated Sepharose beads were implanted s.c. in rats (54). In our laboratory, we have elaborated a tumor model based on Micropore diffusion chambers implanted intraperitoneally in animals (16, 55). The chambers consist of two Millipore GSWP filters (Millipore'P Corp., USA) with a mean pore diameter of 0.22 urn, heat-sealed to both sides of an acrylic plastic ring that is 2 mm thick and has an outer diameter of 13 mm (56). A suspension of antigencoated polymer particles, or fixed or viable tumor cells, are filled into the chamber through a hole in the ring. The chamber is sealed with a plastic plug, implanted i.p., and the labeled MoAb with specificity for the chamber content can be injected i.v. The chamber walls allow free diffusion of molecules, including antibodies, while the tumor cells or polymer particles have diameters larger than the membrane Acta Obstet Gynecol Scand Suppl J55
pores. A model like this can in principle be utilized in any experimental animal species. We have applied this model for a comparative study of the targeting potential of an IgG and its F(ab')2 and Fab' fragments (57). The model has also been used to obtain, for the first time, experimental data on the avidity of an antibody in vivo (58). The antibody association constant estimated from the in vivo binding results indicated that the antibody avidity is'not changed in vivo, when compared with the in vitro estimate. Thus, reduction in avidity seems not to account for the often insufficient tumor uptake in immunotargeting. A main advantage with this diffusion chamber model is the possibility for measurement of the concentration of free antibody within the target. A second advantage is that the target can be filled with target cells or antigen-coated particles that have been preincubated with the labeled antibody in vitro, such that the in vivo stability of labeling methods, the stability of the antibody-antigen reaction, or antigen shedding can be studied under controlled conditions. A theoretical mathematical model, trying to take all present knowledge about the in vivo antibody biodistribution and tumor uptake into account, has been published (25). Artificial targeting models should also be preferred for experimental determination of many of the variables and parameters in such mathematical models. Hybridoma as tumor model for intraperitoneal carcinomatosis An interesting tumor model of intraperitoneal carcinomatosis has been developed using murine B-cell hybridoma. This model has not yet been applied for targeting purposes. However, its potential as a general model of all kinds of therapy, including immunotargeted therapy with labeled MoAb, warrants its discussion in this setting. The hybridoma grows rapidly in immunocompetent mice and produces a highly specific and sensitive tumor marker, i.e. the specific monoclonal antibody which can be detected in the blood circulation (59). In selecting the hybridoma cells, the following properties were sought: rapid growth of the clone in the abdominal cavity without any pretreatment of the host animal, the possibility of using cells cultured in vitro for direct inoculation in the animal, a short and uniform median survival time, uniform intraabdominal tumor spread, early detection of the monoclonal antibody in the murine blood, and the monoclonal antibody had to be directed against an easily available antigen. Based on these criteria, the clone K13, producing a monoclonal antibody specific for the kappa light
Experimental radio-immunotargeting chains of human immunoglobulin, was chosen. An assay was developed for the measurement of this MoAb in the host blood. This was performed by coupling antigen (human IgG, kappa) to polymer particles, so that the MoAb K13 could be extracted from the blood samples. In this model only 5 III blood were sampled from the tail vein, and K13 reached a measurable level in the blood of all mice as early as 24 h after the injection of 106 hybridoma cells. Kinetic studies suggested that only 105 secreting cells were needed to give a detectable K13 blood level. Intraperitoneal injection of 106 K13 hybridoma cells resulted in a standard tumor growth pattern with diffuse superficial i.p. tumor spread and ascites, and a short mean survival (Table II). These findings made the hybridoma model attractive for investigating the treatment of intraperitoneal superficial tumor implants. Moreover, because immunocompetent mice were used, the immunological response of the host can be evaluated.
Aknowledgements The i.p, diffusion chamber immunotargeting model was developed in cooperation with Dr Haakon B. Benestad, Institute of Basic Sciencies, Dept. of Physiology, University of Oslo. Norway. J. G. Fjeld, I. Vergote and L. Vergote-De Vos are research fellowsof the Norwegian Cancer Society, and K. Nustad is supported by the same society.
References I. Pressman D, Korngold L. The ill vivo localization of anti-Wagner osteogenic sarcoma antibodies. Cancer 1953;6: 619-23. 2. Kohler G, Milstein C. Continuous culture of fused cells secreting antibody of predefined specificity. Nature 1975; 256: 495-97. 3. Mach JP. Buchegger F, Forni M, et al. Use of radiolabelled anti-CEA antibodies for the detection of human carcinomas by external photoscanning and tomoscintigraphy. Immunol Today 1981; 2: 239-49. 4. Farrands PA, Perkins AC, Pimm MV, Baldwin RW, Hardcastle JD. Radio-immunodetection of human colorectal cancers using an anti-tumour monoclonal antibody. Lancet 1982; i: 397-400. 5. Epenetos AA, Britton KE, Mather S, et al. Targeting of 12-'1-labelled tumour associated monoclonal antibodies to ovarian, breast and gastrointestinal tumours. Lancet 1982; ii: 999-1006. 6. Sears HF, Atkinson B, Mattis J, Ernst C, Herlynn D, Koprowski H. Phase I clinical trial of monoclonal antibody in the treatment of gastrointestinal tumours. Lancet 1982; i: 762-65. 7. Miller RA, Oseroff AR, Stratte PT, Levy R. Treatment of B-cell lymphoma with monoclonal antiidiotype antibody. N Engl J Med 1982; 306: 517-22.
109
8. Jain RK. Transport of molecules in the tumour interstitium: A review. Cancer Res 1987;47: 3039- 51. 9. Sands H, Jones PL, Brown BA, Nason T. Uptake of radiolabel in rat liver cells after administration of radiolabelled B72.3 and its F(ab'h fragments. Br J Cancer 1989;59: 306 (Abstract). 10. Jain RK. Delivery of novel therapeutic agents in tumors: Physiological barriers and strategies. J Natl Cancer Inst. 1989; 81: 570-76. II. Jain RK and Ward-Hartley K. Tumor blood flow characterization, modifications, and role in hyperthermia. IEEE Trans. Sonics Ultrasonics 1984; SU-31: 504-26. 12. Stickney DR, Gridley DS, Kirk GA, Slater JM. Enhancement of monoclonal antibody binding to melanoma with single dose radiation or hyperthermia. Natn Cancer Inst Monogr 1985; 3: 47-52. 13. Msirikale JS, Klein JL, Schroeder J, Order SE. Radiation enhancement of radiolabelled antibody deposition in tumors. Int J Radiat Oncol BioI Phys 1987; 13: 1839-44. 14. Jain RK. Tumor blood flow response to heat and pharmacological agents. In: Fielden, Fowler, Hendry, Scott, cds. Radiation research (Proc. 8th ICRR), vol. 2, pp 813-18. London: Taylor & Francis, 1987. 15. Jain RK. Determinants of tumor blood flow. A review. Cancer Res. 48: 2641-58, 1988. 16. Fjeld JG, Bruland OS, Benestad HB, Schjerven T, Stigbrand T, Nustad, K. Radio-imrnunotargeting of human tumour cells in immunocompetent animals. Br J Cancer 1990; 62: 573-78. 17. Sears HF, Herlyn D. Steplewski Z, et al. Phase II clinical trial of a murine monoclonal antibody cytotoxic for gastrointestinal carcinoma. Cancer Res. 1985; 45: 5910-13. 18. Koprowski H, Herlyn D, Lubeck M, et al. Human anti-idiotype antibodies in cancer patients. Is the monulation of the immune response beneficial for the patient? Proc Natl Acad Sci 1984; 81: 216-19. 19. Herlyn D, Lubeck M, Sears HF, et al. Specific detection of anti-idiotypic immune response in cancer patients treated with murine monoclonal antibody. J Immunol Methods 1985; 85: 27-28. 20. Wistar Symposium on Immunodiagnosis and Immunotherapy with C017-IA MAb in gastrointestinal cancer. Hybridoma, suppl. I: SI-SI87, 1986. 21. Lee VK, Harriot TG, Kuchroo VK, et al. Monoclonal anti-idiotypic antibodies related to a murine oncofetal bladder tumor antigen induce specific cell-mediated tumor immunity. Proc Nall Acad Sci 1985; 82: 6286-90. 22. Herlyn D, Ross A, Koprowski H. Anti-idiotypic antibodies bear the internal image of a human tumor antigen. Science 1986; 232: 100-02. 23. Shaw DR, Khazaeli MB, Sun LK. Ghrayeb J, Daddona PE, McKinney S, LoBuglio AF. Characterization of a mouse/human chimeric monoclonal antibody (17-1a) to a colon cancer tumor-associated antigen. J Immunol 1987; 138: 4534-38. 24. Sikora K, Alderson T, Nethersell A, Smedley H. Tumor localisation by human monoclonal antibodies. Med Oncol Tumor Pharmacother 1985; 2: 77-86. Acta Obstet Gynecol Scand Suppl155
110
J. G. Fjeld et at.
25. Thomas GD, Chappell MJ, Dykes PW, Ramsden DB, Godfrey KR, Ellis JR, Bradwell AR. Effect of dose, molecular size, affinity, and protein binding on tumor uptake of antibody or ligand: A biomathematical model. Cancer Res 1989; 49: 3290-96. 26. Carasquillo JA, Abrams PG, Schroff RW, et al. Effect of antibody dose on the imaging and biodistribution of indium-111 9.2.27 anti-melanoma monoclonal antibody. J Nucl Med 1988; 29: 39-47. 27. Carasquillo JA, Sugarbaker P, Colcher D, et al. Radioscintigraphy of colon cancer with iodine-131-labeled B72.3 monoclonal antibodies. J Nucl Med 1988; 29: 1022-30. 28. Koizumi K, DeNardo GL, DeNardo SJ, et al. Multicompartmental analysis of the kinetics of radiolabelled monoclonal antibody in patients with cancer. J Nucl Med 1986; 27: 1243-54. 29. Rogers GT, Harwood PJ, Pedley RB, Boden J, Bagshawe KD. Dose-dependent localisation and potential for therapy of F(ab')2 fragments against CEA studied in human tumour xenograft model. Br J Cancer 1986; 54: 341-44. 30. Larson SM, Brown JP, Wright PW, et al. Imaging of melanoma with 1311-labeled monoclonal antibodies. J Nucl Med 1983; 24: 123-29. 31. Epenetos AA, Munro AJ, Stewart S, et al. Antibody guided irradiation of advanced ovarian cancer with intraperitoneally administered radiolabeled monoclonal antibodies. J Clin Oncol 1987; 5: 1890-99. 32. Leroy M, Teillac P, Rain JD, Saccavini JC, Le Due A, Najean Y. Radio-immunodetection of lymph node invasion in prostatic cancer. The use of iodine-123 labeled monoclonal anti-prostatic acid phosphatase (PAP) 227 and F(ab')2 antibody fragments in vivo. Cancer 1989; 64: 1-5. 33. Halpern SE, Stern PH, Hagan PL, et al. Radiolabeling of monoclonal antitumor antibodies: comparison of 1-125 and In-111 CEA with Ga-67 in a nude mouse! human colon tumor model. Clin Nucl Med 1981; 6: 453. 34. Keenan AM, Harbert JC, Larson SM. Monoclonal antibodies in nuclear medicine. J Nucl Med 1985; 26: 531-37. 35. Zalutsky MR, Narula AS. Radio-halogenation of a monoclonal antibody using an N-succinimidyl 3-(tri-nbutylstannyl) benzoate intermediate. Cancer Res 1988; 48: 1446-50. 36. Hnatowich DJ. Recent developments in the radiolabeling of antibodies with Iodine, Indium and Technetium. Sem Nucl Med 1990; 20: 80. 37. Rodwell JD, Alvarez VL, Lee C, et al. Site-specific covalent modification of monoclonal antibodies: In vitro and in vivo evaluations. Proc Nat! Acad Sci 1986; 83: 2632-36. 38. Philben VJ, Jacowatz, JG, Beatty BG, et al. The effect of tumor CEA content and tumor size on tissue uptake of Indium l l l-labeled anti-CEA monoclonal antibody. Cancer 1986; 57: 571-76. 39. Hamilton TC, Young RC, Louie KG, et al. Characterization of a xenograft model of human ovarian carcinoma which produces ascites and intra-abdominal carcinomatosis in mice. Cancer Res 1984; 44: 5286-90. Acta Obstet Gyneco/ Scand SlIpp//55
40. Baumal R, Law J, Buick RN, et al. Monoclonal antibodies to an epithelial ovarian adenocarcinoma: Distinctive reactivity with xenografts of the original tumor and a cultured cell line. Cancer Res 1986; 46: 3994-4000. 41. Ripamonti M, Canevari S, Menard S, et al. Human carcinoma cell lines xenografted in athymic mice: biological and antigenic characteristics of an intra-abdominal model. Cancer Immunol Immunother 1987; 24: 13-18. . 42. Wahl R, and Liebert M. Improved radiolabeled monoclonal antibody uptake by lavage of intraperitoneal carcinomatosis in mice. J Nucl Med 1989; 30: 60-65. 43. Esteban J, Hyams DM, Beatty BG, Merchant B, Beatty JD. Radio-immunotherapy of human colon carcinomatosis Xenograft with 9OY-ZCE025 monoclonal antibody: Toxicity and tumor phenotype studies. Cancer Res 1990; (Suppl.) 50: 989s-992s. 44. Naomoto Y, Kondo H, Tannaka N, Orita K. Novel experimental models of human cancer metastasis in nude mice: lung metastasis, intra-abdominal carcinomatosis with ascites, and liver metastasis. J Cancer Res Clin Oncol 1987; 113: 544-49. 45. Reale FR, Griffin TW, Compton JM, Graham S, Townes PL, Bogden A. Characterization of a human malignant mesothelioma cell line (H-MESO-l): A biphasic solid and ascitic tumor model. Cancer Res 1987; 47: 3199-205. 46. Ward BG, Wallace K, Shepherd JH, Balkwill FR. Intraperitoneal xenografts of human epithelial ovarian cancer in nude mice. Cancer Res 1987; 47: 2662-67. 47. Fekete E, Ferrigno MA. Studies on a transplantable teratoma of the mouse. Cancer Res 1952; 12: 438-40. 48. Mannel RR, Stratton JA, Rettenmaier MA, Liao SY, Disaia PJ. Use of a murine model for comparison of intravenous and intraperitoneal cisplatin in the treatment of microscopic ovarian cancer. Gynecol Oncol 1988; 31: 50-55. 49. Pour PM. Transplacental induction of gonadal tumors in rats by a nitrosamine. Cancer Res 1986;46: 4135-38. 50. Taguchi 0, Michael SD, Nishizuka Y. Rapid induction of ovarian granulosa cell tumors by 17, 12-Dimethylbenz(a)anthracene in neonatally estrogenized mice. Cancer Res 1988; 48: 425-29. 51. Los G, Mutsaers PHA, van der Vijgh WJF, Baldew GS, de Graaf PW, McVie JG. Direct diffusion of cisdiaminedichloroplatinum (II) in intraperitoneal rat tumors after intraperitoneal chemotherapy: a comparison with systemic chemotherapy. Cancer Res 1989;49: 3380-84. 52. Holt JA, Waggoner SE, Lee EY, Hubby MM, Hamilton TC. Serum CA 125 and survival of mice inoculated with ovarian carcinoma and treated with antiestrogen, estrogen, or progestin. Gynecol Oncol 1987; 27: 282-93. 53. Otsuka FL, Welch MJ, McElvany KD, Nicolotti RA, Fleischman JB. Development of a model system to evaluate methods for radiolabelling monoclonal antibodies. J Nucl Med 1984; 25: 1343-49. 54. Walker KZ, Seymour-Munn K, Keech FK, et al. A rat model system for radio-imrnunodetection of kappa
Experimental radio-immunotargeting
55.
56.
57.
58.
myeloma antigen on malignant B cells. Eur J Nucl Med 1986; 12: 461-67. Fjeld JG, Benestad HB, Stigbrand T, Nustad K. In vivo evaluation of radiolabelled antibodies with antigen-coated polymer particles in diffusion chambers. J Immunol Methods 1988; 109: 1-7. Benestad HB, Reikvam A. Diffusion chamber culturing of haematopoietic cells: Methodological investigations and improvement of the technique. Exp Hematol 1975; 3: 249-60. Fjeld JG, Michaelsen TE, Benestad HB, Nustad K. The effect of the biodistribution differences between IgG, F(ab')2 and Fab' on their immunotargeting potential for human tumor cells in immunocompetent mice. Antib Immunoconjug Radiopharm 1991; 4: 443-51. Fjeld JG, Benestad HB, Stigbrand T, Nustad K. In
111
vivo measurement of the association constant of a radiolabeled monoclonal antibody in experimental irnmunotargeting. Br J Cancer [in press, 1992]. 59. Vergote I, De Vos L, Fjeld JG, et al. B-cell hybridoma as intraperitoneal tumor model. Correlation between tumor growth and monoclonal antibody production. Hybridome [in press, 1992].
Address for correspondence:
Kjell Nustad, Ph.D., M.D. Central Laboratory The Norwegian Radium Hospital N-031O Oslo 3 Norway
Acta Obstet Gynecol Scand Suppl 155
Radio-immunotargeting in experimental animal models of intraperitoneal cancer J. G. FJELD" I. VERGOTE',2, L.
DE Vos' AND
K.
NUSTAD'
From the 'Central Laboratory, and the 2Department of Gynecologic Oncology, the Norwegian Radium Hospital, Oslo, Norway
Acta Obstet Gynecol Scand 1992: 71 Suppl 155: 105-111 The field of immunotargeting, and the challenges met when this technique is applied in experimental animals or in patients, are reviewed. Even with highly specific monoclonal antibodies, non-specific uptake in normal tissues and high background level of unbound radioactivity in blood and extravascular body fluids remain significant problems. Further experimental work in animal model systems is needed to bring this technique from the state of being an experimental method, with limited clinical application, to a routine diagnostic or therapeutic method. Different animal models are available, and their potential for elucidation of the various methodological problems in radio-immunotargeting are discussed in the present paper. In our laboratory. two intraperitoneal models were devised. having relevance for gynecologic and other forms of intraperitoneal malignancies. These models were elaborated with special emphasis on the possibility for exact measurement of important parameters in immunotargeting reactions. In the first model, hybridoma cells are inoculated intraperitoneally to mimic intraperitoneal carcinomatosis, and the monoclonal antibody produced by the hybridoma is used as serum tumor marker. In the second model the tumor cells are contained within intraperitoneally implanted micropore chambers, resembling a localized tumor. An artificial tumor like this allows control with the antigen load in the target, and measurement of the concentration of the injected antibody in the fluid within the target.
Radio-immunotargeting The basic idea of immunotargeting is to exploit the exquisite specificity of antigen-antibody reactions in the diagnosis or treatment of cancer. The lack of specificity for the pathologic cells is a general problem with systemic therapy, resulting in detrimental side effects on normal tissues. The idea to conjugate radio-isotopes with a tumor specific antibody was pioneered almost 40 years ago with polyclonal antibodies (1). With the polyclonal antisera affinity purification procedures were necessary to obtain antibody preparations of high purity. However, because
the fraction of specific immunoglobulin in antisera is low, it was difficult to collect a sufficient quantity of tumor-specific immunoglobulins. Like many other fields in immunology, the immunotargeting technique did benefit when homogeneous immunoglobulin preparations of specified antigen specificity, i.e. monoclonal antibodies (MoAb), were made available in gram quantities by the hybridoma technology (2). Many of the problems with in vivo targeting with polyclonal antisera were presumed to be due to factors specific for the polyclonicity. Consequently, the hybridoma technique was suggested to revolutionize both cancer diagnosis (3-5), and therapy (6, 7).
Abbreviations used: MoAb, monoclonal antibody(-ies); HAMA, human anti mouse antibody(-ies). Acta Obstet Gynecol Scand Suppl ISs
106
J. G. Fjeld et al.
Are labeled MoAb 'magic bullets'?
While an overall improvement was obtained with MoAb, it was soon discovered that also MoAb gave a series of problems when applied for immunotargeting. Therefore, considerable effort was still needed to explore and understand the fundamental mechanisms of the shortcomings of MoAb in vivo, with the purpose of finding techniques to overcome the problems. Heralding the labeled MoAb as 'magic bullets' or 'missiles' may mislead the reader to think of the targeting reaction as an active, target-seeking process. This is far from correct, and may explain some of the disappointment expressed when the low tumor uptake of antibodies relative to injected dose is presented. The physiochemical laws for antigen-antibody reactions seem frequently to be forgotten in studies on antibody-antigen reactions in vivo. When a tumorspecific labeled MoAb is injected i.v., it reaches the tumor cells via (i) the blood circulation, (ii) extravasation across the blood vessel wall to the interstitium, (iii) transport with the interstitial fluid to the specific antigen in the tumor (8). The transport mechanisms are convection and diffusion. The immunological binding of antibody to the antigen maintains the antibody concentration gradient towards the tumor antigen. The efficacy of the transport of a certain injected dose from blood to antigen is hampered by (i) antibody excretion, (ii) immunoglobulin catabolism by the reticuloendothelial system (9), (iii) non-specific binding to normal tissues or to soluble molecules, and (iv) specific binding to free tumor antigen in the body fluids. Moreover, the tumor tissues have physiologic barriers responsible for the poor antibody uptake. The tumor tissue blood supply is heterogeneous, and may be poor in some areas, the interstitial pressure in tumors is frequently elevated, and the antibody transport distances in the interstitium are often large (10). Several methods such as radiation, heat and vasoactive drugs, have been tried out to improve the tumor blood flow and thereby increase the antibody uptake (11-15). However, the effect was always of short duration. This low access to the tumor antigen is in contrast to an in vitro antibody-antigen reaction, where all the antibody administered is exposed to the antigen, except for a minor fraction of nonspecifically bound antibodies. Thus, with high affinity antibodies and a concentration close to antigen excess, nearly 100% of the added antibody dose will be bound in vitro, while only a minor fraction of the same antibodies are bound in vivo. The vast majority of animal experiments have been carried out with nude mice carrying relatively large subcutaneous xenografts, and the results in Acla Obstet Gynecol Scand Suppl/55
these experiments are regularly far better than in patients. Considering the above-mentioned basic physiological and physiochemical mechanisms that affect the targeting result, it is easy to explain the low tumor uptake in large animals or patients (16). When the antibody distribution volume is large compared with the tumor volume, the fraction of injected dose reaching the tumor will be significantly lower than in a small mouse carrying a relatively large xenograft. Human Anti-Mouse Antibodies (HAMA)
The xenogeneic transfer of murine MoAb is a stimulus for anti-MoAb production by the patients (6, 17). This HAMA production has three aspects. First, HAMA may initiate dangerous allergic reactions. However, this has not been a significant problem so far, partly due to the precautions taken against this potential danger in clinical studies. The second aspect is that HAMA may interfere with the tumor uptake of MoAb. Thirdly, since HAMA may have anti-idiotypic activity (18, 19), these antibodies may result in the production of anti-anti-idiotypic antibodies, i.e. anti-tumor antibodies (18, 20-23). Thus, one might speculate that HAMA in some cases could be beneficial for the patient. Trying to avoid the HAMA problem, mouse/human chimeric antibodies have been produced by transfection of a non-producing mouse myeloma line with a chimeric gene construct composed of murine variable regions and constant regions from Man (23). Moreover, human MoAb were produced by fusing lymphocytes from glioma or bronchial carcinoma patients with a human myeloma cell line (24). Determination of the optimal dose of labeled MoAb
As discussed by others (25), conflicting results have been obtained in studies on the optimal MoAb dose to be administered (26-29). According to the mass action law, an increase in antibody dose implies an increase in uptake. Moreover, this increase in antibody uptake should vary with the relative concentration levels of antigen and antibody. The optimal dose is presumably the dose resulting in a tumor uptake corresponding to the upper part of the sigmoid binding curve obtained when plotting the uptake of different amounts of antibody to a constant amount of antigen. When approaching antibody excess the net effect of further increase in antibody dose will be minimal, and contribute mainly to the background of unbound antibodies.
Experimental radio-immunotargeting Table I. Intraperitoneal xenograft models Ovarian cancer
NIH-OVCAR-3 HEY and SKOV-3 SW626 HTB77IP3 Fresh tumor cells
Hamilton, 1984 (39) Baumal, 1986 (40) Ripamonti, 1987 (41) Wahl, 1989 (42) Ward, 1987 (46)
Colon cancer
LS174T HT29 RPMI4788
Esteban, 1990 (43) Ripamonti, 1987 (41) Naomoto, 1987 (44)
Mesothelioma
H-MESO-1
Reale, 1987 (45)
Radiolabels, and the effect of labeling on the pharmacokinetics The radiolabel and the labeling technique may change the antibody biodistribution and the specific target uptake. Various radio-isotopes are utilized, selected on the basis of their radiation characteristics. 211Astatine, lJOYttrium, 67Copper, I88Rhenium, 32phosphorous and 212Bismuth are radiometals with theoretical potential for radiotherapy, while IIIIn_ dium and 99mTechnetium are imaging labels. Radioiodines such as 1251 and 131 1 were used in numerous experimental biodistribution and targeting studies, and 131 1 was used both for clinical imaging (30), and for therapy (31). A third radio-iodide, the shortlived isotope 1231, has good imaging characteristics (5, 32). Labeling with iodine is easy and convenient, and applicable to all proteins with accessible tyrosine residues. However, loss of the iodine label due to dehalogenation has been reported (33, 34). The free radio-iodine is then incorporated in the thyroid, or excreted, mainly through the kidneys. Novel and improved labeling methods have been developed during recent years, improving the iodination of antibodies (35), or making it possible to utilize a series of new radio-isotopes. Recently, the method for protein coupling with technetium has been improved significantly (36). A main problem with the lllIn-labeled antibodies is that they are sequestered in the liver. Modification of the oligosaccharide moieties of MoAb was demonstrated to be preferable to modification of aminoacids, when conjugating MoAb with 1I1In and 1251 (37). Labels or labeling procedures may also affect the immunological binding characteristics of the antibody. Therefore, preclinical immunoreactivity and biodistribution experiments should be carried out with each new radiolabel and coupling method.
107
Experimental animal model systems The nude mouse xenograft system The nude mouse xenograft tumor model is the most frequently utilized animal model. In this model, different human tumors are transferred to immunodeficient mice, either by s.c. implantation of fresh tissue specimens, or by s.c. injection of a small volume of suspended cells from a human cell line. After a variable period of time, the grafted cells have grown to a tumor of about 1 gram. This model mimics the real tumor situation fairly well, but has several limitations. The athymic mice must be kept under carefully controlled and rather expensive housing conditions in order to protect them from infections, and effects of the immune system on tumor growth cannot be studied. Moreover, the grafted tumor cells must grow to a tumor of appropriate size, and the growth rate and tumor size will vary from animal to animal. Lastly, the antibody binding capacity is unknown, and the value of this parameter may not be proportional to the tumor size (38). Transperitoneal spread and intra-abdominal carcinomatosis are common in patients with ovarian and colorectal cancer. Hence, intraperitoneal tumor models mimic more closely the spread pattern of ovarian and colorectal cancer than do subcutaneous models. A variety of experimental i.p. tumor models have been applied to elucidate the biology of intraabdominal carcinomatosis, and to evaluate the potential benefit of radio-immunotherapy. Recently, i.p. diffuse tumor spread has been described in nude mice, using human ovarian (39-42), colorectal (41, 43,44) and malignant mesothelioma cell lines (45), while Ward et al. (46) established i.p. xenografts using fresh primary human tumor material (Table I). Table II. Intraperitoneal xenograft versus hybridoma as tumor model Xenograft Carcinoma Human derived Uniform intra-abdominal spread Median survival Blood tumor marker Cost price Growth from culture cells Tumor take Immunocompetent host a
6
Hybridoma
++ ++
+
++ 10 days
25-43 days (+)' High ++/_6
Low ++
+
++
++
++
Only for the NIH-OVCAR-3 model (see main text). Some models use in vitro cultured cells, while in other models, tumor transplants are used. ACla Obstet Gynecot Scand SIIppl155
108
J. G. Fjeld et al.
Like the subcutaneously xenografted mice, all these intraperitoneal models have several limitations (Table II). Cell lines derived from human tumors may differ from the original tumor cells due to selection in the establishment of the cell line or modulation of the biological features of the cell line by the abnormal condition associated with its propagation. Therefore Ward et at. (46) proposed the use of intraperitoneal xenografts from primary tumor material, but passage time varied in this model from 1 to 7 months, and only three out of ten tumors were easily passageable. Intraperitoneal tumor growth has also been reported using murine ovarian teratoma and ovarian reticulum cell sarcoma cell lines, or tumor induction by carcinogens (47-51). These models use murine or rat tumors and the tumor type does not represent a common human cancer. In both these and the intraperitoneal xenograft models it is difficult to measure the delay in tumor growth, because they often provide multiple metastases at sites where they are difficult to measure. The NIH-OYCAR-3 xenograft is the only i.p. tumor model with a serum marker, i.e CA 125 (52). However, in this study as much as 50!l1 blood had to be sampled from the orbital plexus, and the serum CA 125 level was elevated in all animals after at least 12 days. Artificial targets in immunocompetent animals For exploration of all the problems that are not directly related to the tumor tissue, artificial targets are well suited. Such targets give better control with the antigen burden, so that systematic comparative studies are easier to perform. In one such model system the target consisted of antigen-coated agarose beads that were injected i.v. into rats, being sequestered in the lung capillary bed (53). In a second model, antigen-coated Sepharose beads were implanted s.c. in rats (54). In our laboratory, we have elaborated a tumor model based on Micropore diffusion chambers implanted intraperitoneally in animals (16, 55). The chambers consist of two Millipore GSWP filters (Millipore'P Corp., USA) with a mean pore diameter of 0.22 urn, heat-sealed to both sides of an acrylic plastic ring that is 2 mm thick and has an outer diameter of 13 mm (56). A suspension of antigencoated polymer particles, or fixed or viable tumor cells, are filled into the chamber through a hole in the ring. The chamber is sealed with a plastic plug, implanted i.p., and the labeled MoAb with specificity for the chamber content can be injected i.v. The chamber walls allow free diffusion of molecules, including antibodies, while the tumor cells or polymer particles have diameters larger than the membrane Acta Obstet Gynecol Scand Suppl J55
pores. A model like this can in principle be utilized in any experimental animal species. We have applied this model for a comparative study of the targeting potential of an IgG and its F(ab')2 and Fab' fragments (57). The model has also been used to obtain, for the first time, experimental data on the avidity of an antibody in vivo (58). The antibody association constant estimated from the in vivo binding results indicated that the antibody avidity is'not changed in vivo, when compared with the in vitro estimate. Thus, reduction in avidity seems not to account for the often insufficient tumor uptake in immunotargeting. A main advantage with this diffusion chamber model is the possibility for measurement of the concentration of free antibody within the target. A second advantage is that the target can be filled with target cells or antigen-coated particles that have been preincubated with the labeled antibody in vitro, such that the in vivo stability of labeling methods, the stability of the antibody-antigen reaction, or antigen shedding can be studied under controlled conditions. A theoretical mathematical model, trying to take all present knowledge about the in vivo antibody biodistribution and tumor uptake into account, has been published (25). Artificial targeting models should also be preferred for experimental determination of many of the variables and parameters in such mathematical models. Hybridoma as tumor model for intraperitoneal carcinomatosis An interesting tumor model of intraperitoneal carcinomatosis has been developed using murine B-cell hybridoma. This model has not yet been applied for targeting purposes. However, its potential as a general model of all kinds of therapy, including immunotargeted therapy with labeled MoAb, warrants its discussion in this setting. The hybridoma grows rapidly in immunocompetent mice and produces a highly specific and sensitive tumor marker, i.e. the specific monoclonal antibody which can be detected in the blood circulation (59). In selecting the hybridoma cells, the following properties were sought: rapid growth of the clone in the abdominal cavity without any pretreatment of the host animal, the possibility of using cells cultured in vitro for direct inoculation in the animal, a short and uniform median survival time, uniform intraabdominal tumor spread, early detection of the monoclonal antibody in the murine blood, and the monoclonal antibody had to be directed against an easily available antigen. Based on these criteria, the clone K13, producing a monoclonal antibody specific for the kappa light
Experimental radio-immunotargeting chains of human immunoglobulin, was chosen. An assay was developed for the measurement of this MoAb in the host blood. This was performed by coupling antigen (human IgG, kappa) to polymer particles, so that the MoAb K13 could be extracted from the blood samples. In this model only 5 III blood were sampled from the tail vein, and K13 reached a measurable level in the blood of all mice as early as 24 h after the injection of 106 hybridoma cells. Kinetic studies suggested that only 105 secreting cells were needed to give a detectable K13 blood level. Intraperitoneal injection of 106 K13 hybridoma cells resulted in a standard tumor growth pattern with diffuse superficial i.p. tumor spread and ascites, and a short mean survival (Table II). These findings made the hybridoma model attractive for investigating the treatment of intraperitoneal superficial tumor implants. Moreover, because immunocompetent mice were used, the immunological response of the host can be evaluated.
Aknowledgements The i.p, diffusion chamber immunotargeting model was developed in cooperation with Dr Haakon B. Benestad, Institute of Basic Sciencies, Dept. of Physiology, University of Oslo. Norway. J. G. Fjeld, I. Vergote and L. Vergote-De Vos are research fellowsof the Norwegian Cancer Society, and K. Nustad is supported by the same society.
References I. Pressman D, Korngold L. The ill vivo localization of anti-Wagner osteogenic sarcoma antibodies. Cancer 1953;6: 619-23. 2. Kohler G, Milstein C. Continuous culture of fused cells secreting antibody of predefined specificity. Nature 1975; 256: 495-97. 3. Mach JP. Buchegger F, Forni M, et al. Use of radiolabelled anti-CEA antibodies for the detection of human carcinomas by external photoscanning and tomoscintigraphy. Immunol Today 1981; 2: 239-49. 4. Farrands PA, Perkins AC, Pimm MV, Baldwin RW, Hardcastle JD. Radio-immunodetection of human colorectal cancers using an anti-tumour monoclonal antibody. Lancet 1982; i: 397-400. 5. Epenetos AA, Britton KE, Mather S, et al. Targeting of 12-'1-labelled tumour associated monoclonal antibodies to ovarian, breast and gastrointestinal tumours. Lancet 1982; ii: 999-1006. 6. Sears HF, Atkinson B, Mattis J, Ernst C, Herlynn D, Koprowski H. Phase I clinical trial of monoclonal antibody in the treatment of gastrointestinal tumours. Lancet 1982; i: 762-65. 7. Miller RA, Oseroff AR, Stratte PT, Levy R. Treatment of B-cell lymphoma with monoclonal antiidiotype antibody. N Engl J Med 1982; 306: 517-22.
109
8. Jain RK. Transport of molecules in the tumour interstitium: A review. Cancer Res 1987;47: 3039- 51. 9. Sands H, Jones PL, Brown BA, Nason T. Uptake of radiolabel in rat liver cells after administration of radiolabelled B72.3 and its F(ab'h fragments. Br J Cancer 1989;59: 306 (Abstract). 10. Jain RK. Delivery of novel therapeutic agents in tumors: Physiological barriers and strategies. J Natl Cancer Inst. 1989; 81: 570-76. II. Jain RK and Ward-Hartley K. Tumor blood flow characterization, modifications, and role in hyperthermia. IEEE Trans. Sonics Ultrasonics 1984; SU-31: 504-26. 12. Stickney DR, Gridley DS, Kirk GA, Slater JM. Enhancement of monoclonal antibody binding to melanoma with single dose radiation or hyperthermia. Natn Cancer Inst Monogr 1985; 3: 47-52. 13. Msirikale JS, Klein JL, Schroeder J, Order SE. Radiation enhancement of radiolabelled antibody deposition in tumors. Int J Radiat Oncol BioI Phys 1987; 13: 1839-44. 14. Jain RK. Tumor blood flow response to heat and pharmacological agents. In: Fielden, Fowler, Hendry, Scott, cds. Radiation research (Proc. 8th ICRR), vol. 2, pp 813-18. London: Taylor & Francis, 1987. 15. Jain RK. Determinants of tumor blood flow. A review. Cancer Res. 48: 2641-58, 1988. 16. Fjeld JG, Bruland OS, Benestad HB, Schjerven T, Stigbrand T, Nustad, K. Radio-imrnunotargeting of human tumour cells in immunocompetent animals. Br J Cancer 1990; 62: 573-78. 17. Sears HF, Herlyn D. Steplewski Z, et al. Phase II clinical trial of a murine monoclonal antibody cytotoxic for gastrointestinal carcinoma. Cancer Res. 1985; 45: 5910-13. 18. Koprowski H, Herlyn D, Lubeck M, et al. Human anti-idiotype antibodies in cancer patients. Is the monulation of the immune response beneficial for the patient? Proc Natl Acad Sci 1984; 81: 216-19. 19. Herlyn D, Lubeck M, Sears HF, et al. Specific detection of anti-idiotypic immune response in cancer patients treated with murine monoclonal antibody. J Immunol Methods 1985; 85: 27-28. 20. Wistar Symposium on Immunodiagnosis and Immunotherapy with C017-IA MAb in gastrointestinal cancer. Hybridoma, suppl. I: SI-SI87, 1986. 21. Lee VK, Harriot TG, Kuchroo VK, et al. Monoclonal anti-idiotypic antibodies related to a murine oncofetal bladder tumor antigen induce specific cell-mediated tumor immunity. Proc Nall Acad Sci 1985; 82: 6286-90. 22. Herlyn D, Ross A, Koprowski H. Anti-idiotypic antibodies bear the internal image of a human tumor antigen. Science 1986; 232: 100-02. 23. Shaw DR, Khazaeli MB, Sun LK. Ghrayeb J, Daddona PE, McKinney S, LoBuglio AF. Characterization of a mouse/human chimeric monoclonal antibody (17-1a) to a colon cancer tumor-associated antigen. J Immunol 1987; 138: 4534-38. 24. Sikora K, Alderson T, Nethersell A, Smedley H. Tumor localisation by human monoclonal antibodies. Med Oncol Tumor Pharmacother 1985; 2: 77-86. Acta Obstet Gynecol Scand Suppl155
110
J. G. Fjeld et at.
25. Thomas GD, Chappell MJ, Dykes PW, Ramsden DB, Godfrey KR, Ellis JR, Bradwell AR. Effect of dose, molecular size, affinity, and protein binding on tumor uptake of antibody or ligand: A biomathematical model. Cancer Res 1989; 49: 3290-96. 26. Carasquillo JA, Abrams PG, Schroff RW, et al. Effect of antibody dose on the imaging and biodistribution of indium-111 9.2.27 anti-melanoma monoclonal antibody. J Nucl Med 1988; 29: 39-47. 27. Carasquillo JA, Sugarbaker P, Colcher D, et al. Radioscintigraphy of colon cancer with iodine-131-labeled B72.3 monoclonal antibodies. J Nucl Med 1988; 29: 1022-30. 28. Koizumi K, DeNardo GL, DeNardo SJ, et al. Multicompartmental analysis of the kinetics of radiolabelled monoclonal antibody in patients with cancer. J Nucl Med 1986; 27: 1243-54. 29. Rogers GT, Harwood PJ, Pedley RB, Boden J, Bagshawe KD. Dose-dependent localisation and potential for therapy of F(ab')2 fragments against CEA studied in human tumour xenograft model. Br J Cancer 1986; 54: 341-44. 30. Larson SM, Brown JP, Wright PW, et al. Imaging of melanoma with 1311-labeled monoclonal antibodies. J Nucl Med 1983; 24: 123-29. 31. Epenetos AA, Munro AJ, Stewart S, et al. Antibody guided irradiation of advanced ovarian cancer with intraperitoneally administered radiolabeled monoclonal antibodies. J Clin Oncol 1987; 5: 1890-99. 32. Leroy M, Teillac P, Rain JD, Saccavini JC, Le Due A, Najean Y. Radio-immunodetection of lymph node invasion in prostatic cancer. The use of iodine-123 labeled monoclonal anti-prostatic acid phosphatase (PAP) 227 and F(ab')2 antibody fragments in vivo. Cancer 1989; 64: 1-5. 33. Halpern SE, Stern PH, Hagan PL, et al. Radiolabeling of monoclonal antitumor antibodies: comparison of 1-125 and In-111 CEA with Ga-67 in a nude mouse! human colon tumor model. Clin Nucl Med 1981; 6: 453. 34. Keenan AM, Harbert JC, Larson SM. Monoclonal antibodies in nuclear medicine. J Nucl Med 1985; 26: 531-37. 35. Zalutsky MR, Narula AS. Radio-halogenation of a monoclonal antibody using an N-succinimidyl 3-(tri-nbutylstannyl) benzoate intermediate. Cancer Res 1988; 48: 1446-50. 36. Hnatowich DJ. Recent developments in the radiolabeling of antibodies with Iodine, Indium and Technetium. Sem Nucl Med 1990; 20: 80. 37. Rodwell JD, Alvarez VL, Lee C, et al. Site-specific covalent modification of monoclonal antibodies: In vitro and in vivo evaluations. Proc Nat! Acad Sci 1986; 83: 2632-36. 38. Philben VJ, Jacowatz, JG, Beatty BG, et al. The effect of tumor CEA content and tumor size on tissue uptake of Indium l l l-labeled anti-CEA monoclonal antibody. Cancer 1986; 57: 571-76. 39. Hamilton TC, Young RC, Louie KG, et al. Characterization of a xenograft model of human ovarian carcinoma which produces ascites and intra-abdominal carcinomatosis in mice. Cancer Res 1984; 44: 5286-90. Acta Obstet Gyneco/ Scand SlIpp//55
40. Baumal R, Law J, Buick RN, et al. Monoclonal antibodies to an epithelial ovarian adenocarcinoma: Distinctive reactivity with xenografts of the original tumor and a cultured cell line. Cancer Res 1986; 46: 3994-4000. 41. Ripamonti M, Canevari S, Menard S, et al. Human carcinoma cell lines xenografted in athymic mice: biological and antigenic characteristics of an intra-abdominal model. Cancer Immunol Immunother 1987; 24: 13-18. . 42. Wahl R, and Liebert M. Improved radiolabeled monoclonal antibody uptake by lavage of intraperitoneal carcinomatosis in mice. J Nucl Med 1989; 30: 60-65. 43. Esteban J, Hyams DM, Beatty BG, Merchant B, Beatty JD. Radio-immunotherapy of human colon carcinomatosis Xenograft with 9OY-ZCE025 monoclonal antibody: Toxicity and tumor phenotype studies. Cancer Res 1990; (Suppl.) 50: 989s-992s. 44. Naomoto Y, Kondo H, Tannaka N, Orita K. Novel experimental models of human cancer metastasis in nude mice: lung metastasis, intra-abdominal carcinomatosis with ascites, and liver metastasis. J Cancer Res Clin Oncol 1987; 113: 544-49. 45. Reale FR, Griffin TW, Compton JM, Graham S, Townes PL, Bogden A. Characterization of a human malignant mesothelioma cell line (H-MESO-l): A biphasic solid and ascitic tumor model. Cancer Res 1987; 47: 3199-205. 46. Ward BG, Wallace K, Shepherd JH, Balkwill FR. Intraperitoneal xenografts of human epithelial ovarian cancer in nude mice. Cancer Res 1987; 47: 2662-67. 47. Fekete E, Ferrigno MA. Studies on a transplantable teratoma of the mouse. Cancer Res 1952; 12: 438-40. 48. Mannel RR, Stratton JA, Rettenmaier MA, Liao SY, Disaia PJ. Use of a murine model for comparison of intravenous and intraperitoneal cisplatin in the treatment of microscopic ovarian cancer. Gynecol Oncol 1988; 31: 50-55. 49. Pour PM. Transplacental induction of gonadal tumors in rats by a nitrosamine. Cancer Res 1986;46: 4135-38. 50. Taguchi 0, Michael SD, Nishizuka Y. Rapid induction of ovarian granulosa cell tumors by 17, 12-Dimethylbenz(a)anthracene in neonatally estrogenized mice. Cancer Res 1988; 48: 425-29. 51. Los G, Mutsaers PHA, van der Vijgh WJF, Baldew GS, de Graaf PW, McVie JG. Direct diffusion of cisdiaminedichloroplatinum (II) in intraperitoneal rat tumors after intraperitoneal chemotherapy: a comparison with systemic chemotherapy. Cancer Res 1989;49: 3380-84. 52. Holt JA, Waggoner SE, Lee EY, Hubby MM, Hamilton TC. Serum CA 125 and survival of mice inoculated with ovarian carcinoma and treated with antiestrogen, estrogen, or progestin. Gynecol Oncol 1987; 27: 282-93. 53. Otsuka FL, Welch MJ, McElvany KD, Nicolotti RA, Fleischman JB. Development of a model system to evaluate methods for radiolabelling monoclonal antibodies. J Nucl Med 1984; 25: 1343-49. 54. Walker KZ, Seymour-Munn K, Keech FK, et al. A rat model system for radio-imrnunodetection of kappa
Experimental radio-immunotargeting
55.
56.
57.
58.
myeloma antigen on malignant B cells. Eur J Nucl Med 1986; 12: 461-67. Fjeld JG, Benestad HB, Stigbrand T, Nustad K. In vivo evaluation of radiolabelled antibodies with antigen-coated polymer particles in diffusion chambers. J Immunol Methods 1988; 109: 1-7. Benestad HB, Reikvam A. Diffusion chamber culturing of haematopoietic cells: Methodological investigations and improvement of the technique. Exp Hematol 1975; 3: 249-60. Fjeld JG, Michaelsen TE, Benestad HB, Nustad K. The effect of the biodistribution differences between IgG, F(ab')2 and Fab' on their immunotargeting potential for human tumor cells in immunocompetent mice. Antib Immunoconjug Radiopharm 1991; 4: 443-51. Fjeld JG, Benestad HB, Stigbrand T, Nustad K. In
111
vivo measurement of the association constant of a radiolabeled monoclonal antibody in experimental irnmunotargeting. Br J Cancer [in press, 1992]. 59. Vergote I, De Vos L, Fjeld JG, et al. B-cell hybridoma as intraperitoneal tumor model. Correlation between tumor growth and monoclonal antibody production. Hybridome [in press, 1992].
Address for correspondence:
Kjell Nustad, Ph.D., M.D. Central Laboratory The Norwegian Radium Hospital N-031O Oslo 3 Norway
Acta Obstet Gynecol Scand Suppl 155
