1 [CANCER RESEARCH 47, June ] Characterization of a Human Malignant Mesothelioma Cell Line (H-MESO-1): A Biphasic Solid and Ascitic Tumor Model1 Frank R. Reale,2 Thomas W. Griffin, Joyce M. Compton, Suzanne Graham, Philip L. Townes, and Arthur Bogden Departments of Pathology [F. R. R., J. M. C.J, Medicine [T. W. G.Â,and Pediatrics [P. L. T.J, University of Massachusetts Medical School, Worcester, Massachusetts 01605; H. J. D. Orthopedic Institute, New York, New York fs. G.J; and Bogden Laboratories [A. B.J, E G & G Mason Research Institute, Worcester, Massachusetts 0/605 ABSTRACT Human malignant mesothelioma of the pleura was successfully trans planted s. c. into athymic nude mice and grew as a solid neoplastic mass. Tumor growth resulted in death of the animals between 98 and 161 days after implantation. Minced samples of the growing tumor were propa gated as a malignant peritoneal effusion. Animals with malignant ascites died predictably at 32 to 33 days. Light and electron microscopy, and immunocytochemistry demonstrated a similarity of the transplanted solid and fluid malignancies with the human primary mesothelioma. Cytogenetic analysis demonstrated a predominance of cells with a triploid number of identifiable but abnormal human chromosomes. This model, which mimics the clinical behavior of malignant mesothelioma in the human, may be of value in animal trials of chemotherapy and immunotherapy. INTRODUCTION Malignant mesothelioma is a neoplastic process arising from serosal surfaces, most frequently in the pleura and peritoneum. The incidence of malignant mesothelioma has increased in North America and industrial parts of Europe as reported in the early 1960s (1, 2). This greater incidence correlates with increased occupational exposure to asbestos. Fifty five % of pleural mesotheliomas and 53% of peritoneal mesotheliomas in men have been associated with an occupational exposure to asbestos fibers (3). In an unexposed population the incidence of mesothelioma is about one per million. However in the asbestos exposed populations the incidence increases to ap proximately 0.2% of consecutive autopsies (4). Not only are workers who are directly exposed to asbestos fibers at risk but a substantial "bystander" risk in families of asbestos workers has been documented (5). The clinical behavior of malignant mesothelioma is characterized by growth on serosal surfaces with progressive encasement of heart and viscera with the concomitant development of malignant effusions (6). In con trast, blood born mã tastasesto other organs occur late (1, 5). This distinctive pattern of growth suggests that mesothelioma may be a candidate for regional therapy. Human tumors serially carried in nude mice have proven to be a useful model for testing therapeutic modalities. Their value, however is dependent on preservation of biological behavior and morphological features of the initial tumor (7). The serial transplantation of human malignant mesothelioma in nude mice as a s. c. xenograft has been previously reported as has a transplantable xenograft from a fresh human tissue and from an in vitro cultivated cell line (8, 9). A transplantable malignant mesothelioma that mimics the clinical behavior of this tumor cell type in the human with tumor mass formation and malig- nant effusions has not been reported. We believe that the development of a cell line that can grow predominantly either as a solid form and as an effusion with a predictable course can lead to broadened investigation of therapeutic methods in this disease. MATERIALS AND METHODS Initiation and Propagation of Cell Line. A 35-year-old previously healthy male with a known history of asbestos exposures 12 to 20 years previously underwent exploratory thoracotomy for shortness of breath and opacification of the right hemithorax. Malignant mesothelioma was diagnosed by pathology and a portion of the resected malignant pleura was immediately placed in medium (TC 199; Grand Island Biological Co., Grand Island, NY) with gentamicin. This specimen was transported by common carrier and received within 24 h by Bogden Laboratories [Mason Research Laboratory in Worcester, MA (EG&G)]. Nine 1-mm3 portions of the tumor were implanted s. c. into nu/nu mice of BALB/c lineage, of either sex, 4-6 weeks of age. Initial measurable growth was seen from 21 to 29 days after implantation. A portion of the above transplanted tumor from the 12th passage was thoroughly minced by sterile techniques and injected i. p. into similar nu/nu mice. After several weeks cell growth was detected by the development of obvious ascites. 0.5 ml of thick rose-colored ascitic fluid was mixed with 1.5 ml of TC 199 medium and injected i.p. into 4-6-wk-old similar nu/nu mice. Serial transplantation was carried out every days. Both the ascitic and solid cell lines were tested for 12 possible viral contaminants by Microbiological Associates, Inc. (Log numbers B3273 and B4594) and were found to be negative for organisms. A portion of the original tumor from the left pleural surface was fixed in 10% buffered formalin and processed in the routine histological fashion. Samples of the solid transplanted tumor (passage 51) were likewise processed in a similar manner. Sections (5 //in) were stained with hematoxylin and eosin as well as PAS3 with and without diastase, alcian blue at ph 2.5 and 0.4 with and without hyaluronidase, and MInek-arm Â in.- according to the standard techniques (10). Sections (5 pm) of the tumor to be analyzed for the presence of keratin and cytokeratin specific antigens were digested on 0.1% trypsin (type III; Sigma Chemical Co., St. Louis, MO) and 0.1% CaCl2 (ph 7.8) for 15 min at 37Â C.Tissues were then treated with anticytokeratin (kit preparation; Beckton Dickinson, Mountain View, CA) or antikeratin, 1:100 and 1:500 (Enzo Biochemical, Inc., NY, NY). Undigested tissues were studied with anticarcinoembryonic antigen, 1:100 and 1:500, (DAKO Corp., Santa Barbara, CA) or Vimentin, 1:10 and 1:20 (Enzo). Tissue sections were exposed sequentially to the above primary antiserum for 18 h followed by 1:100 biotinylated horse ant imouse IgG of 1:100 biotinylated sheep anti-rabbit IgG for l h at room temperature (Vector Laboratories Inc., Burlingame, CA). Sections were then ex posed to avidin-conjugated horseradish peroxidase (Vector) 1:100 for 30 min at room temperature. Identification of antigen-labeled cells was accomplished after incubation in 3.3' diaminobenzydine (30 mg/100 Received 5/29/86; revised 2/11/87; accepted 3/9/87. The costs of publication of this article were defrayed in part by the payment ml) and 0.05% H2O2 in 0.05 M tris buffer (ph 7.6). of page charges. This article must therefore be hereby marked advertisement in A sample of the solid form of the H-MESO-1 tumor was minced accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1Supported in part by Grant R01-CA from the National Cancer Insti under sterile conditions and plated into RPMI1640 medium containing 1% penicillin, 1% streptomycin, 1% kanamycin, and 10% inactivated fetal calf serum (all from GIBCO). Cultures were maintained at 37'C tute. In whom requests for reprints should be addressed, at Department of Pathology, University of Massachusetts Medical School, 55 Lake Avenue North. Worcester, MA ' The abbreviation used is: PAS, periodic acid Schiff. 3199
2 in a humidified 5% COj atmosphere for 3 weeks, until a monolayer of tumor cells formed. Medium was changed twice a week. Six milliliters of ascitic fluid were retrieved and centrifugea and the cell sediment was washed with a balanced salt solution. Six smears were prepared from the cell button; all residual cells were resuspended in the salt solution and filtered through a Gelman Metrice! Membrane Filter (5-Ã Â mpore). Both preparations were wet fixed in 95% ethyl alcohol and stained by the Papanicolaou method. Ultrastructural analysis was accomplished by transmission electron microscopy on tissue from the solid form and on cell blocks from the ascitic form. These were fixed with 2.5% v/v glutaraldehyde in 0.1 M phosphate buffer, ph 7.4, postfixed with osmium tetroxide, and em bedded. No electron microscopy was done on the original biopsy. Cytogenetic Analysis. Cytogenetic studies were performed on the ascitic form of the H-MESO-1 from passage 42 and on cells cultured in monolayer from the solid form. Giemsa-trypsin banding was accom plished using the method of Seabright (11). RESULTS Morphology. On gross examination of the mice two distinct growth patterns could be appreciated. Those animals that re ceived s. c. implants of H-MESO-1 died between 98 and 161 days after transplantation with an average survival time of days. At the time of their death, large solid tumor was present at the site of implantation (average 29 g) with smaller mã tas tases seen throughout the peritoneal cavity. Moderate ascites F^ffiÂ &f!ãª&~&'m X Â«"~%- Ã **t' -*' t "***â *â * *-*^*Â«2t :Ã "' ' ' / *â ^r- ^?# Â»â ž â T* P. ' T - "â Â»,Â». *3Trf*-. ^ Â«U^. '. *'JWB ' â '5Â 5Â Â 1-^^ Fig. I. Similar morphological appearance is seen in the histology of the original tumor (A) and the transplanted tumor (B and C). Both tumors consist of epithelioid cells growing in nests divided by fibrous stroma. Multinucleated tumor cells and abnormal mitosis are frequent. Immunohistochemical staining for cytokeratin was strongly positive in the transplanted tumor (D). Prominent nucleoli are uniformly seen. In the ascitic form (Ej, Ft and Fj) of H-MESO-l cell clusters with "windows," nucleoli and a single cell in mitosis are present. Cell-in-cell arrangement is compared in the solid form and ascitic form (Â /).Hematoxylin and eosin (A-E,) and Papanicolaou (E* FI, and Fj): bar, 50 tâ m. 3200
3 formed in this group terminally. In those animals that received implantation by i. p. injection of IO6 H-MESO-1 cells, death occurred predictably of massive malignant ascites at 32 to 33 days after injection with an average survival of 32.3 days. At the time of their death encasement of most intraabdominal organs and diffuse thickening of the peritoneum by tumor was seen. Intraabdominal mã tastasesvaried in size from 0.2 to 1.4 mm. Thus these methods of implantation seem to correlate well with the mode of tumor spread and with survival time of the animals. The original tumor demonstrated a moderately pleomorphic population of epithelioid cells with little attempt at organiza tion. Nests of epithelioid cells were entrapped in variable bands of fibrous stroma. There were a few papillary fronds and tumor nests in focal areas. Blood vessels were also present within the tumor. The individual tumor cells varied in shape from colum nar and cuboidal to polygonal. The cytoplasm stained with varying density with eosin. The nuclei had a coarse or stippled chromatin pattern with a few nuclei demonstrating a vesicular appearance. A nucleolus was frequently present in a nucleus. Rare cells contained bizarre or multiple nuclei. A few foci of necrosis and pyknotic nuclei were also present. Inflammatory cells consisting of both polymorphonuclear leukocytes and lym phocytes were identifiable near areas of necrosis (Fig. 1/4). Few of the tumor cells stained faintly positive with PAS and this reaction was not removed with diastase. Rare tumor cells stained positive for alcian blue at ph 2.5 and 0.4 and this intracellular and extracellular staining was completely removed by hyaluronidase. Mucicarmine stains of the tumor were nega tive. The epithelioid cell to stremai cell ratio was 95:5. All of the above findings were consistent with a malignant mesothe- 1ionia of the epithelioid type. Though not performed on the original tumor, immunohistochemical staining for keratin, cytokeratin, carcinoembryonic antigen, and vimentin were studied on the solid form of H- MESO-1. Tumor cells were interpreted as strongly positive if all cells stained for the antigen, moderately positive if many cells stained, weakly positive if only an occasional cell stained, and negative if no cells stained. Appropriate positive and neg ative controls were used with all antibodies. The tumor stained strongly positive for cytokeratin (see Fig. ID), moderately pos itive for keratin, weakly positive for vimentin (1:20) and nega tive for carcinoembryonic antigen. This immunohistochemical profile is well known for human malignant mesothelioma (12). The histology of the transplanted solid tumor at passage 51 (Fig. l, B and Q retained identical morphological and cytochemical staining characteristics with hematoxylin and eosin, PAS, alcian blue, and mucicarmine. The overall architecture was unchanged. Cytology. The Papanicolaou-stained smears from the H- MESO-1 ascitic form contained abundant basophilic epithe lioid neoplastic cells and rare blood elements. The majority of the malignant cells were isolated with cell diameters which measured between 12 and 20 urn and a mean diameter of 17.5 urn. Many doublets and triplets were present which exhibited pincer-like prongs and cell in cell arrangements (13, 14) (see Fig. l ET). Occasional aggregates of between five and 40 cells were present in each xloo field and "windows" (clear zones thought to represent the brush borders) were present between the individual cells as described in mesothelial cells of human origin and human mesotheliomas (13, 15) (see Fig. 1, F, and F2). A similar cellular orientation was also seen in the histology (Fig. \EÂ ).The nuclear size measured between 8 and 16 /Â m with a mean of 15 urn and an increased nuclearxytoplasmic ratio. The nuclei in the abnormal cells were usually located centrally with the cytoplasm barely visible in many cells. The chromatin was clumped and irregularly distributed with variable nuclear outlines. Nucleoli were present in the majority of the cells. Multinucleation and mitoses could be seen in every three high power fields. Two cell populations were apparent in the smears, one of well differentiated neoplastic mesothelial cells and a second of more poorly differentiated cells. This finding though not unique is often helpful in making the diagnosis of malignant mesothe lioma. This is especially true when the morphology of the neoplastic cells closely resemble normal mesothelium (14). Ultrastructure. The ultrastructural characteristics of both the solid form and the malignant ascitic fluid of H-MESO-1 were characteristic of epithelioid mesothelioma, although identifia ble differences were noted between the two (16-21). The solid tumor showed characteristic epithelial-like cells with abundant cytoplasm and a single nucleus. The cells showed increasing clumping of chromatin and prominence of nucleoli (Fig. 2) as their cytoplasm and surface specialization became more com-, Â ' ^: â Fig. 2. Electron micrograph of a cell from the solid tumor phase showing a characteristic nucleus with prominent nucleoli (N) and abun dant cytoplasm. Active unpolarized cytoplasmic organdã es with abundant mitochon-..; dria are seen, x ri^ ' ' " â *: ' -VvV. *.*Â,. â i.'*- 1".\ *?Ã Sr k-.-ã fci-*.. -' â - "â "' 3201
4 Fig. 3. Electron micrograph demonstrating the complex long irregular and branched microvilli between tumor cells suggestive of mesothelioma. Abundant cytoplasmic organdã es can be seen, x Fig. 4. Electron micrograph of a cell from solid phase showing a less differentiated cell than Fig. 2 with a tight junction (T) and less complex cytoplasm. The nucleus is convoluted with a single nucleolus (AO.x Fig. 5. Electron micrograph of a binucleated tumor cell from the ascitic phase with prominent macronucleoli (W) and coarse chromatin clumping (C). Cytoplasmic organelles are polarized around the nucleus and associ ated with 10-nm filaments (arrows), x 5000.
5 Fig. 6. Electron micrograph of two tumor cells from the ascitic phase in the "cell in cell" arrangement similar to those seen in Fig. I/. Decreased microvilli are present in the ascitic form of this tumor as compared to the solid form (Fig. 4). x Fig. 7. Giemsa-trypsin banded karyotype from monolayer form of H-MESO-1. ComplÃ ment is pseudotriploid with 68 chromosomes including multiple abnormalities (monosomy, translocations, deletion, trisomy, and tetrasomy) and 14 marker chromosomes represent ing rearrangements of undetermined identity. Arrows, identified rearrangements c "E ->- -11 ; - i G X Y plex. Characteristic cells often showed complex interdigitations of cellular membranes with macula adherens and tight junctions at luminal surface borders. Rare intracellular lumina were noted. Abundant long irregular and often branching microvilli were seen along the cell surfaces (Fig. 3). An active if unpolarized cytoplasmic organelle complement was seen with abundant mitochondria. Prominent rough endoplasmic reticulum con taining moderately osmophilic granular material was noted; similar material being seen in the extracellular spaces. Fila 3203 ments (10 pm) were relatively inconspicuous. Less differen tiated cells showed simplification of surface and junctional specialization and a relatively "empty" cytoplasm (Fig. 4). The ultrastructure of the cells from ascitic fluid shared all of the aforementioned characteristics. Nonspecific lo-^m fila ments, however, were abundant and, along with the bulk of the cytoplasmic organdã es, were circumferentially oriented about the nucleus (Fig. 5). The peripheral cytoplasm contained pri marily rough endoplasmic reticulum and free ribosomes formed
6 bulbus pseudopodia. Surface microvilli were present as were intracellular lumina, but were notably less prominent than in the solid tissue. The "cell in cell" arrangement was also seen in electron microscopy (Fig. 6). Cytogenetic Analysis. Forty six Giemsa-trypsin banded metaphase plates from ascitic and monolayer cells were examined. The modal number of chromosomes in the ascitic and monolayer cells is with a range of 56 to 135 chromosomes. The former is considered to represent broken, incomplete cells and the latter represents adjacency of two cells. The chromo some complement is a markedly abnormal form of pseudotriploidy (Fig. 7) with many chromosomes having additional translocated segments of unknown origin, and with 14 marker chromosomes of uncertain identity. DISCUSSION BIPHASIC HUMAN MALIGNANT MESOTHELIOMA CELL LINE The morphological, ultrastructural, and cytogenetic features of H-MESO-1 were consistent with derivation from malignant mesothelioma. H-MESO-1 cells both in the solid and effusion form possessed very similar histological morphology and ultrastructural appearance to the human malignancy. Morphologi cally, the tumor expiants demonstrated epithelioid malignant cells with histological and immunological profile of malignant mesothelioma. By electron microscopy and cytology the cells grew in a few small tight clusters and occasionally as individual single and multinucleated cells supporting the diagnosis of mesothelioma. Classic intragroup windows are present and thought to be strongly suggestive for mesothelial origin. Occa sional individual cells are seen with multiple nucleoli as are commonly seen in mesothelium. The nuclei of the mouse model mesothelioma growing in the ascitic form were identical to those seen in the histological section of the human primary malignant mesothelioma. Intermediate filaments being more prominent in the tumor cells reflect a more dominant role in the cytoskeletal integrity, although their intracellular role is far from clear. Chromosome abnormalities are commonly seen in malignant tumors including malignant mesothelioma (22, 23). In contrast to the pseudohypodiploidy reported by Mark (23), H-MESO-1 is pseudotriploid. in mesothelioma The monosomy 14 and 22 previously reported (23) is not observed in H-MESO-1. However, monosomy 9, 13, 16, and 17 are found. Comparison of the chromosome complements fails to reveal fundamental similar ity between the mesotheliomas previously reported and the mesothelioma reported here. Suitable animal models of human mesothelioma would pro vide an important opportunity for experimental research in the understanding of mesothelioma. The majority of previous ani mal models have involved induction of mesothelioma in rodents by asbestos fibers or other chemicals (24-28). This technique required long and variable periods of observation. Furthermore, the relevance of these animal tumors to the drug sensitivity of a human malignant mesothelioma may be questioned. The availability of the ascitic form with peritoneal seeding and diffuse thickening and with its relatively short and reproducible survival time may be highly advantageous to study variable modalities in the treatment of this disease. Chahinian et al. reported successful transplantation of human malignant mesothelioma into nude mice in 1981 (8). Human malignant mesotheliomas were successfully transplanted into genetically athymic mice from two of three patients. The tumor expiants of the first generation grew in six of 20 mice. The mean delay between transplantation and tumor growth was days for the range of 18 to 104 days. Pathological examination by light and electron miscroscopy confirmed that the nature of the growing tumors in nude mice was malignant mesothelioma. In these studies, transplantation of the human mesothelioma tissue was performed s. c. with growth limited to the formation of solid masses. It is therefore of note that H-MESO-1 is capable of growth both as a s. c. nodule and as an ascitic fluid with peritoneal seeding and diffuse peritoneal thickening. The latter form strongly mimics the growth pattern of this tumor in humans (6). Since the clinical behavior of human mesothelioma involves a production of serosal tumor implants with dense adhesions and ascites, its retention of biological behavior of the human malignant mesothelioma by this tumor model is notable. Fur thermore, the ability to grow this tumor in ascitic form may allow the evaluation of innovative forms of treatment such as intrapleural or i. p. chemotherapy or immunotherapy (29-31). ACKNOWLEDGMENTS The authors wish to express their gratitude to Angela Farkas and Patricia Crowley for technical assistance in electron miscroscopy and cytogenetics, respectively. REFERENCES 1. Taryle, D. A., Lakshlyinarayan, S., and Sahn, S. A. Pleural mesotheliomaâ an analysis of 18 cases and a review of the literature. Medicine, 55: , Wagner, J. C., Sleggs, C. A., and Marchand, J. A. Diffuse pleural mesothe lioma and asbestos exposure in the Northwestern Industr. Med., 17: , Cape Province. Br. J. 3. MacDonald, A. D., and McDonald, 3. C. Malignant America. Cancer (Phila.), 46: , mesothelioma in North 4. Otto, H. Das berufsbedingte Mesotheliom in der BRD. Pathologe, 2: 8-18, Animan, K. H. Malignant mesothelioma. N. Engl. J. Med., 303: , Antman, K. H. Clinical Presentation and Natural History of Benign and Malignant Mesothelioma. Semin. Oncol., 8: , Shimosato, Y., Kameya, T., Najai, K., Hirohashi, S., Koide, T., Hayashi, H., and Nomura, T. Transplantation of human tumors Cancer Inst., 56: , in nude mice. J. Nati. 8. Chahinian. A. P., Beranek, J. T., Suzuki, Y., Bekesi, J. G., Wisniewski, L., Selikoff, I. J., and Holland, J. F. Transplantation of human malignant mesothelioma into nude mice. Cancer Res., 40: , Arnold, W., Naundorf, H., Wildner, G. R., Nissen, E., and Tanneberger, S. T. Biological characterization of a mesothelioma line in nude mice. I. Transplantation of in vitro cultivated cells of a human ascitic tumor effusion. Arch. Geschwulstforsch., 49: , Luna, L. G. (Ed.) Manual of Histologie Staining Methods of the Armed Forces Institute of Pathology, Ed. 3. New York: McGraw-Hill Book Com pany, Seabright, M. A rapid banding technique for human chromosomes. 2: , Lancet, 12. Kahn, H. J., Thorner, P. S., Yeger, H., Bailey, D., and Baumal, R. Distinct keratin patterns demonstrated by immunoperoxidase staining of adenocarcinomas, carcinoids, and mesotheliomas using polyclonal anti-keratin antibodies. Am. J. Clin. Pathol., 86: , and monoclonal Naylor, B. The exfoliative cytology of diffuse malignant Pathol. Bacteriol., 86: , mesothelioma. J. 14. Klempman, S. Exfoliative cytology of diffuse pleural mesothelioma. (Phila.), 15: , Cancer 15. Koss, L. G. Diagnostic Cytology and its Histopathologic Bases, Ed. 3, p Philadelphia: J. B. Lippincott Co., Davis, J. M. G. Ultrastructure of human mesothelioma. J. Nati. Cancer Inst., 52: , Kannerstein, M., Churg, J., and McCaughey, W. T. E. Asbestos and meso thelioma: a review. Pathol. Annu., 13: , Legrand, M., and Pariente, R. Ultrastructure thelioma. Thorax, 29: , study of pleural fluid in meso 19. Virtanen, I., and Lehto, V. P. Intermediate (lonm) filaments in human malignant mesothelioma. Virch. Arch. B Cell. Pathol., 28: , Uys, C. J. Observations on the pathology and ultrastructure of mesothelioma. In: F. Muggia and M. Rosencqing (eds.), Lung Cancer: Progress in Thera peutic Research, pp New York: Raven Press, Wang, N. S. Electron microscopy in the diagnosis of pleural mesotheliomas. Cancer (Phila.), 31: , 1973.
7 22. Sondberg, A. A. The Chromosomes in Human Cancer and Leukemia. New 27. Kannerstein, M., and Chung, J. Mesothelioma in man and experimental York: Elsevier, p. 249, animals. Environ. Health Perspect., 34: 31-36, Mark J Monosomy 'l4, Monosomy 22 and 13qâ three chromosomal ab- 28. Berman, J. J., and Rice, J. M. Mesotheliomas and proliferative lesions of normalities observed in cells of two malignant mesotheliomas studied by 'he testicular mesothelium produced in Fischer, Sprague-Dawley, and Buffalo banding techniques. Acta Cytol.. 22: , Â«Jfby methyl (Acetoxy methyl) nitrosamiril (DMN-OAC). Vet. Pathol., 16: 24. Smith, W. E., Hubert, D. D., Holiat, S. M., Sobel, H. J., and Davis, S. An,â ž ;2Ã Â X. Â»i,,, â Â»â *, A-K., ir,. r, r- IDÃ.-I\JI.M 29. Chahmian, A. P., and Holland, J. F. Treatment of diffuse malignant mesoexperimental model for treatment of mesothel.oma. Cancer (Ph.la.), 47:658- the ioma. g revjew M( sinaâ } Med Â J;54_?6,9?g Markman, M., Howell, S. B., Lucas, W. E., Pfeifle, C. E., and Green, M. R. 25. Shin, M. L., and Firmmger, H. I. Acute and chronic effects of mtraperitoneal Combination intraperitoneal chemotherapy with cisplatin, cytarabine, and injection of two types of asbestos in rats with a study of the histopathogenesis doxorubicin for refractory ovarian carcinoma and other malignancies prinand ultrastructure of resulting mesothelioma. Am. J. Pathol., 70: , cipally confined to the peritoneal cavity. J. Clin. Oncol., 2: , Wahl, R. L., Geatti, O., and Fischer, S. Intraperitoneal delivery of monoclo- 26. Davis, J. M. G. The use of animal models for studies on asbestos bioeffects. nal antibodies: influence of class and fragmentation on kinetics and intraper- Ann. N. Y. Acad. Sci., 79: , itoneal dosing advantage. J. NucÃ.Med., 26: 114,