Murine Colon Adenocarcinoma: Syngeneic Orthotopic Transplantation and Subsequent Hepatic Metastases 1. 2 M. H. Tan, 3. 4 E. D. Holyoke, 3. 4 and M. H. Goldrosen 3.4.5 ABSTRACT-Syngeneic murine colon adenocarcinoma (MCA· 38) cells were transplanted In the submucosa of distal colon, proximal colon, cecum, ileum, jejunum, and duodenum of male C57BL/6 mice, with local lymphoid follicles used as points of entry. The tumor grew best at the cecum and led to liver and mesenteric lymph node metastases in 8 and 9 weeks, respective· Iy, after transplantation. Histologically, a local Inflammatory reaction Involving polymorphonuclear leukocytes was observed within 48-72 hours following transplantation; after this time, the microscopic tumor foci began to grow progressively. Mononu· clear lymphoid cells of the gut-associated lymphoid tissue did not infiltrate the progressively growing tumor; however, polymor· phonuclear leukocytes were constantly observed at the tumor periphery in the lamina propria. The studies indicated that or· thotoplc transplantation as a model system can provide a means of examining the role of the local Immune response as a focus of host resistance and as a factor in metastatic tumor spread. The findings also suggested the usefulness of this model in immunotherapeutic and chemotherapeutic studies of secondary hepatic disease.-J Natl Cancer Inst 59: 1537-1544,1977.
Several investigators have used murine colon cancer induced by DMH as a model for the study of colorectal cancer (1-4). This model has been useful in defining the etiologic factors involved in the induction of colon carcinoma (5-10), potential chemotherapeutic agents useful in its treatment (3), and cell-surface antigens expressed and the nature of the immune responses evoked through these antigens (11-15). Whereas most of these studies were performed on rodents with primary colon adenocarcinomas, some studies (16, 17) were done with a model in which the colon tumor was implanted into an sc site that was not representative of the natural environment of this cancer (18). Although the DMH-induced cancer in the rodent model has certain characteristics in common with human colon cancer, important differences also exist. DMH does induce tumors in organs other than the colon in certain murine species, and colon tumors in mice generally arise as multiple foci; although in humans multiple tumor foci are also observed, most human tumors appear as a single focus. Furthermore, the liver is a primary site of distant metastases in humans, whereas in the murine model, DMH-induced colon tumors have not been reported to metastasize to the liver to an extent where meaningful experiments could be performed. This report primarily describes basic observations on an orthotopic ally transplanted murine colon tumor. The primary feature of this model is the reproducible development ofliver metastases.
MATERIALS AND METHODS Am·mal5.-Adult 10- to 12-week-old male C57BL/6 mice (West Seneca Laboratories, West Seneca, N.Y.) were maintained on a standard mouse diet of pellets and water ad libitum. Tumor cells. - Murine colon adenocarcinoma (MCA -38) VOL. 59. NO.5. NOVEMBER 1977
1537
was established in tissue culture by differential enzymatic techniques in this laboratory (19). The original tumor line was obtained by chemical induction with DMH in C57BL/6 mice (3) and was a gift from Dr. T. Corbett of the Southern Research Institute, Birmingham, Alabama. MCA-38 cells from tissue culture passage generations 13-17 were used. After trypsinization, tumor cells were washed and' resuspended in medium consisting of RPMI-1640 with 9% heatinactivated fetal calf serum and 0.05% penicillin-streptomycin-neomycin solution (GIBCO #564; Grand Island Biological Co., Grand Island, N. Y.). Cell viability was assessed by trypan blue dye exclusion (GIBCO #525), and cell number was assessed on a hemacytometer. Preparations of 108 cells/ml with greater than 98% viability were used. Orthotopz'c transplantatz'on. -Mice were systemically anesthetized by peritoneal injection of pentobarbital sodium (0.05 mg/g body wt; Abbott Laboratories, North Chicago, Ill.). After the abdomen was completely shaved, the mouse was placed supine in a 10-em petri dish and its limbs were held down by adhesive tape; the skin was prepared with 70% ethanol and a 1.0- to 1.5-em cutaneous incision was made along the abdominal midline followed by an incision through the parietal peritoneum. Then 0.01 ml cell inoculum (10 6 tumor cells) was prepared in a I-emS tuberculin syringe with a 30-gauge I-inch stainless steel hypodermic needle (Jeffrey-Fell Co., Buffalo, N.Y.). The cecum was carefully exteriorized and the tumor cell suspension injected into the apical lymphoid follicle. A successful inoculation turned the follicle initially white and resulted in the formation of a "bleb." The hypodermic needle was withdrawn 15-20 seconds after injection of the cell suspension to allow for the equilibration of the extracellular fluid pressure. Inoculum seepage or bleeding was stopped with a cotton-tipped applicator (Chesebrough-Pond's Inc., New York, N.Y.). The lack of "bleb" formation generally resulted in intraluminal seepage. After the cecum was replaced, the abdominal wall was closed with hemoclips (Edward Week &
ABBREVIATIONS USED: DMH=I.2-dimethylhydrazine; H & E=hematoxylin and eosin.
Received April 7 • 1977; accepted June 27. 1977. Supported in part by Public Health Service grant CA15263 from the National Cancer Institute. S Tumor Immunology Group. Department of Surgical Oncology. New York State Department of Health. Roswell Park Memorial Institute. 666 Elm St .. Buffalo. N. Y. 14263. • Department of Experimental Pathology. Roswell Park Memorial Institute Division. State University of New York at Buffalo. Roswell Park Memorial Institute. 5 We thank Dr. Stephen H. Leveson for his valuable suggestions throughout this study and Ms. Joann Vandrei for her assistance in the preparation of this manuscript. I
2
J NATL CANCER INST
15SS
TAN, HOLYOKE, AND GOLDROSEN
Co., Inc., Long Island City, N.Y.) and the skin closed with wound clips (Clay Adams Inc., New York, N.Y.). Following this operation, the mouse was wrapped in a hand towel, returned to its cage, and allowed to recover under the warmth of incandescent light. The wound clips were removed a week later. The same procedure was used at the other sites along the intestinal tract, and whenever applicable, the thickness of the local lymphoid follicles were used as a site of transplantation. Experiment I.-Mice in groups of 10 were given transplants of 10 6 viable MCA-SS cells at the following sites: distal colon, proximal colon, cecum, ileum, jejunum, and duodenum; they were killed 14 weeks later. Control mice in groups of 5 were given corresponding injections of 0.01 ml suspension medium at the same sites along the intestinal tract. Experiment 2.-Mice in groups of 10 (1 group/wk for 10 wk) were given transplants of 10 6 MCA-S8 cells in the cecal submucosae and were killed 1 week after the last injection. In addition, 1 group at 5 days and groups at 72, 48, and 24 hours before being killed were inoculated in the cecal submucosae. Then 0.01 ml of suspension medium was correspondingly injected into the cecal submucosae of a group of 5 control mice. Observations and necropsy. - The general health of these mice was observed daily and their weights were noted weekly. When the animals were killed, the tumor foci and other internal organs were grossly assessed under a stereomicroscope. The proportion of mice with tumor at injection sites and other sites was recorded. The presence of tumor within the peritoneal cavity but not at the injection site was presumed due to extraluminal seepage, whereas the absence of tumor at the injection site or any other site was presumed due to intraluminal seepage. Tumor and lymphoid tissues were preserved in 10% formalin-phosphate-buffered solution for histologic examination. The same procedure was used in control and experimental mice that died before the end of the experiment.
RESULTS Orthotopic transplantation of syngeneic MCA-S8 cells was well tolerated at the local lymphoid follicles of the intestinal tract. Intra-abdominal seepage of cell inoculum was less than 5% at all sites. Intraluminal seepage was less than 10% at the cecum; however, it exceeded 50% at the distal colon. This represented the lowest and highest intraluminal seepage incidence, respectively, among all tested intestinal sites. In experiment I, more than 40% of the mice died of intraluminal obstruction by the growing tumor 5-8 weeks after transplantation at the distal colon, proximal colon, ile~m, jejunum, and duodenum. At 14 weeks, the remaining mIce appeared healthy and were killed. The primary transplanted tumor measured 1.0-1.5 cm at the smallest diameter, a~d seco~dary tumor foci (0.2-0.8 cm) were commonly found III the hver. Furthermore, the spleen and mesenteric lymph nodes were enlarged. Table 1 presents the incidence of tumors at different sites along the submucosae of the bowel walls in syngeneic C57BL/6 mice. The tumor-take rate was lowest in the proximal and distal colon, intermedi-
J
NA TL CANCER INST
1.-MCA-38 orthotopic transplantation 14 weeks post
TABLE
injection G Site Distal colon Proximal colon Cecum Ileum Jejunum Duodenum
Primary tumor incidence b
Liver metastases·
Mesenteric node metastases d
3/10
2/3
112
4/10 9/10 6/10 6/10 6/10
2/4 5/9 4/6 3/6 2/6
0/2 2/5 114 113 0/2
Dose: 106 MCA-38 cells per transplant. No. of mice with tumorslN o. in group. e No. of mice with liver metastaseslN o. with tumors. d No. of mice with mesenteric node metastases/No. with liver metastases. G
b
ate in the three areas of the small intestine, and highest in the cecum (90%); 50% ofthe mice that developed tumors at t~e primary site of injection also developed macroscopic hver metastases. Macroscopic metastases in organs other than the liver were not observed. The number of mice with regional lymph node involvement was less than that with liver metastases; this suggested hematogenous spread via the portal vein rather than lymphatic spread as the major route for dissemination of disease to the liver. Microscopically, the transplanted tumor retained its prototype histology (fig. 1A). Local mononuclear lymphoid cells were nonreactive adjacent to the growing tumor (figs. 1B, 1C). Marked infiltration of blood polymorphonuclear ce~ls was observed at the tumor periphery in the lamina proprIa (figs. 2A-2F). Figures SA-SC document the micrometastases observed in a cecal venous radicle, the liver, and a mesenteric lymph node. Within 48-72 hours, an inflammatory reaction characterized by the infiltration of polymorphonuclear cells was noted (fig. 4A). At 5 days, this local inflammatory reaction could be observed at the tumor periphery (fig. 4B). At 3 weeks, the transplanted tumor became macroscopically evident and measured 0.2 cm at the smallest diameter. Macroscopic hepatic metastases were 2.4 2.0 E
~ &..I N VI
-~ ~
::;) ~
1.6
IUoCllOSCXlPtC LIVER METASTASES
-
-
-
-20% 60%
65%
1.2 0.8 0.4 0 0 WEEKS POST INJECTION
TEXTFIGU~
I.-Tumor gro~th comparison of MCA·38, sc (0) versus orthotopIC (0) transplantatIOn. Tumor cells (IXI06) inoculated. Or· thotopic transplantation was made at the cecum. Each point represents mean value from at least 8 tumors; vertical bar= the range of tumor diameters. VOL. 59, NO.5, NOVEMBER 1977
1539
ORTHOTOPIC TRANSPLANTATION OF MURINE COLON ADENOCARCINOMA
first noted at 8 weeks and microscopic metastases to the mesenteric lymph nodes were observed at 9 weeks, which again suggested that hepatic metastases were not due to lymphatic spread. The growth curve of the orthotopic transplant was compared to an sc transplant of the same number of MCA-38 cells (text-fig. 1). The results indicate that the growth rate in the submucosa of the bowel wall is one-fifth that of the sc inoculation. Furthermore, 65% of the animals with a primary orthotopic transplant had macroscopic hepatic metastases 10 weeks post injection, whereas no hepatic metastases were noted with the animals given sc transplants. In contrast, less than 20% of the mice with an sc transplant developed pulmonary metastases, whereas no mice with the orthotopic transplant developed pulmonary metastases.
DISCUSSION The DMH-treated rodent model for the study of colorectal cancer has provided valuable insight into the nature of this disease from the viewpoint of several disciplines; however, this animal model differs from human colon cancer in several major respects of their comparative natural history. In many patients, colon carcinoma occurring above the rectosigmoid junction will metastasize to the liver, whereas carcinoma occurring below the peritoneal reflection may spread directly to adjacent organs; in the murine model, hepatic metastases have not been reported. This may be due to the fact that DMH at a dose sufficient to induce colon tumors causes nonspecific cell injury, particularly in the liver (20). Alternatively, these animals may die of intestinal obstruction before the primary tumor has had an opportunity to metastasize. Thus the complementary approach of orthotopic transplantation was developed. A comparative-study was done to determine -the best site of transplantation along the small and large intestines. Tumor take was lowest in the proximal and distal colon, intermediate in the small intestine, and highest in the cecum (90%). The submucosa of the cecum offered several additional advantages as a site of transplantation. Tumor cells can be implanted in the submucosa of the cecum with relative ease and minimal manipulation in comparison to other sites. Death due to intestinal obstruction occurs in a large percentage of animals (40%) at all sites, with the exception of the cecum. Progressively growing tumors in the submucosa of the cecum can be surgically removed, whereas surgical removal of colon tumors at any other site along the murine bowel wall is technically difficult. A companson of the growth rate of an equal inoculum of MCA-38 cells in the submucosa of the cecum and as an sc implant revealed major differences in behavior of this colon tumor. The sc transplant became palpable within 1 week, whereas the orthotopic transplant required 3 weeks to achieve similar size. Once palpable, the colon tumor grew five times slower in the submucosa of the cecum (0.1 cm/wk) than in the subcutaneous area under the skin (0.5 cm/wk). Histologically, an infiltration of polymorphonuclear leukocytes around the periphery of the tumor cells in the orthotopic transplant was observed between 48 and 72 hours. Mononuclear cells of the adjacent lymphoid follicles appeared static and did not infiltrate the progressively growing VOL. 59, NO.5, NOVEMBER 1977
orthotopic tumor. The presence or absence of a mononuclear cell infiltrate appears to be an important factor in the control of tumor growth. Progressively growing, lethal murine tumors lack mononuclear cell infiltrates (unpublished data). The survival of patients with breast cancer correlates directly with the extent of mononuclear cell infiltration in the primary breast carcinoma (21). In the DMHinduced model, Martin et al. (13) observed a prominent infiltration of polymorphonuclear cells in the connective tissue of the growing tumor. More direct evidence was presented by Russell et al. (22) in the Moloney sarcoma system where both tumor progression and tumor regression occur. Polymorphonuclear leukocytes could be isolated from the tumors that grew progressively and subsequently regressed, whereas mononuclear cells were isolated primarily from those tumors that regressed. These observations are consistent with the histologic findings on the orthotopic transplants because no tumor regression was observed. Furthermore, these observations do not rule out a potential immunologic function of the gut-associated lymphoid tissue of the local immune system in the observed delay of tumor growth in the submucosa of the large intestine. A key observation of this study is the finding of hepatic metastases. Hepatic metastases in this model appear to be hematogenous in origin since the number of mesenteric lymph node metastases is lower than the number of liver metastases, regardless of the site of implantation; the microscopic liver foci appear earlier than the mesenteric lymph node metastases. This contrasts with the finding of pulmonary metastases after sc implantation and excision [(3) and unpublished data]. Thus the site of implantation of the initial cell inoculum influenced the dominant site of metastases. However, the sites of implantation of primary inoculum are not the only factor that influences formation of metastases. Direct intraportal injection of syngeneic murine colon tumor cells failed to produce hepatic metastases (unpublished data), which suggested that the presence of the primary inoculum and/or the environment of the primary inoculum promoted formation of hepatic metastases. This model system will open up several new areas of research that currently cannot be explored. The immunobiology of this model can be analyzed from the perspective of the local immune system as a focus of host resistance to progressively growing tumors. Changes in immunologic parameters in relation to the timing of hepatic metastases will give us a better idea of the function of these factors in resisting or facilitating metastatic tumor spread. Immunotherapy and chemotherapy studies aimed at secondary hep_atic neoplastic disease can now be performed in a natural model, which may help us gain a better understanding of colorectal cancer and help us design and test new modalities of therapy in an environment that parallels the natural history of human colon carcinoma more closely than do the models currently in use.
REFERENCES (1) DESCHNER EE: Experimentally induced cancer of the colon. Cancer
S4:824-828, 1974 (2) MARTIN MS, MARTIN F, MICHIELS R, et al: An experimental model for
J NA TL
CANCER INST
1540
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
TAN, HOLYOKE, AND GOLDROSEN
cancer of the colon and rectum. Intestinal carcinoma induced in the rat by 1,2-dimethylhydrazine. Digestion 8:22-34, 1973 CORBETT TH, GRISWOLD DP, ROBERTS BJ, et al: Tumor induction relationships in development of transplantable cancers of the colon in mice for chemotherapy assays, with a note on carcinogen structure. Cancer Res 35:2434-2439, 1975 THURNHERR N, DESCHNER EE, STONEHILL EH, et al: Induction of adenocarcinomas of the colon in mice by weekly injections of 1,2dimethylhydrazine. Cancer Res 33:940-945, 1973 EVANS JT, HAUSCHKA TS, MITTELMAN A: Differential susceptibility of four mouse strains to induction of multiple large-bowel neoplasms by 1,2-dimethylhydrazine. J Nat! Cancer Inst 52:999-1000,1974 EVANS JT, SHOWS TB, SPROUL EE, et al: Genetics of colon carcinogenesis in mice treated with 1,2-dimethylhydrazine. Cancer Res 37:134-136,1977 FILIPE MI: Mucous secretion in rat colonic mucosa during carcinogenesis induced by dimethylhydrazine. A morphological and histological study. Br TCancer 32:60-77, 1975 POZHARISSKI KM: Morphology and morphogenesis of experimental epithelial tumors of the intestine. J Nat! Cancer Inst 54: 1115-1135, 1975 REDDY BS, NARISAWA T, WRIGHT P, et al: Colon carcinogenesis with azoxymethane and dimethylhydrazine in germ-free rats. Cancer Res 35:287-290,1975 ROGERS AE, NEWBERNE PM: Dietary effects on chemical carcinogenesis in animal models for colon and liver tumors. Cancer Res 35:3427-3431,1975 ABEYOUNIS CJ, MILGROM F: A thermostable antigen characteristic for carcinogen-induced rat intestinal tumors. J Immunol 116:30-34, 1976 GARMAISE AB-K, ROGERS AE. NEWBERNE PM. et al: Immunological
J NATL CANCER INST
(13)
(14) (15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
detection of antigen(s) associated with rat colon carcinoma. Nature 248:706-707.1974 MARTIN F, MARTIN MS. BORDES M, et al: Antigens associated with chemically induced intestinal carcinomas in rats. Int J Cancer 15:144-151,1975 SJOGREN HO: Immunobiology of colon carcinomas. Dig Dis 19: 1033-1035.1974 STEELE G, SJOGREN HO, PRICE MR: Tumor-associated and embryonic antigens in soluble fractions of a chemically induced rat colon carcinoma. Int JCancer 16:33-51.1975 MCCALL DC. COLE JW: Transplantation of chemically induced adenocarcinomas of the colon in an inbred strain of rats. Cancer 33:1021-1026, 1974 WARD JM, YAMAMOTO RS, WEISBURGER JH, et al: Transplantation of chemically induced metastatic mucinous adenocarcinomas of the jejunum and colon in rats. J Nat! Cancer Inst :) I: 1997 -1999. 1973 BARTLETT GL. KREIDER JW. PURNELL DM: Immunotherapy of cancer in animals: Models or muddles? J NatI Cancer Inst 56:207210, 1976 TAN MH, HOLYOKE ED. GOLDROSEN MH: Murine colon adenocarcinomas: Methods for selective culture in vitro. J NatI Cancer Inst 56:871-873. 1976 HAASE P, COWEN DM, KNOWLES JC, et al: Evaluation of dimethylhydrazine induced tumours in mice as a model system for colorectal cancer. Br J Cancer 28:530-543, 1973 BLACK MM, OPLER SR, SPEER FD: Survival in breast cancer cases in relation to the primary tumor and regional lymph nodes. Surg Gynecol Obstet 100:543-551,1955 RUSSELL SW. GILLESPIE GY, HANSEN B. et al: Inflammatory cells in solid murine neoplasms. II. Cell types found throughout the course of Moloney sarcoma regression or progression. Int J Cancer 18:331-338,1976
VOL. 59, NO.5, NOVEMBER 1977
I.-A) MCA-38 orthotopic transplant showing grade III adenocarcinoma; B) nonreactive gut-associated lymphoid follicle with adjacent tumor; and C) nonreactive mesenteric lymph node with adjacent tumor _H & E. x 400
FIGURE
TAN. HOLYOKE. AND GOLDROSEN
1541
FIGURE 2. - Infiltration of polymorphonuclear leukocytes at the tumor in the lamina propria. A) distal colon; B) proximal colon; C) cecum; D) ileum; E) jejunum; and F) duodenum. H & E. X 400
1542
TAN, HOLYOKE, AND GOLDROSEN
3. - Dissemination of MCA·38 cells from an orthotopic transplant (H & E) to A) a cecal venous radicle (x 450); B) liver (x 400); and C) mesenteric lymph node (x 350).
FIGURE
TAN, HOLYOKE, AND GOLD ROSEN
1543
FIGURE 4.-A) Inflammatory reaction characterized by infiltration of polymorphonuclear cells. H & E. x 350. B) Local in· flammatory reaction at the tumor periphery. H & E. x 400
1544
TAN, HOLYOKE, AND GOLDROSEN
Several investigators have used murine colon cancer induced by DMH as a model for the study of colorectal cancer (1-4). This model has been useful in defining the etiologic factors involved in the induction of colon carcinoma (5-10), potential chemotherapeutic agents useful in its treatment (3), and cell-surface antigens expressed and the nature of the immune responses evoked through these antigens (11-15). Whereas most of these studies were performed on rodents with primary colon adenocarcinomas, some studies (16, 17) were done with a model in which the colon tumor was implanted into an sc site that was not representative of the natural environment of this cancer (18). Although the DMH-induced cancer in the rodent model has certain characteristics in common with human colon cancer, important differences also exist. DMH does induce tumors in organs other than the colon in certain murine species, and colon tumors in mice generally arise as multiple foci; although in humans multiple tumor foci are also observed, most human tumors appear as a single focus. Furthermore, the liver is a primary site of distant metastases in humans, whereas in the murine model, DMH-induced colon tumors have not been reported to metastasize to the liver to an extent where meaningful experiments could be performed. This report primarily describes basic observations on an orthotopic ally transplanted murine colon tumor. The primary feature of this model is the reproducible development ofliver metastases.
MATERIALS AND METHODS Am·mal5.-Adult 10- to 12-week-old male C57BL/6 mice (West Seneca Laboratories, West Seneca, N.Y.) were maintained on a standard mouse diet of pellets and water ad libitum. Tumor cells. - Murine colon adenocarcinoma (MCA -38) VOL. 59. NO.5. NOVEMBER 1977
1537
was established in tissue culture by differential enzymatic techniques in this laboratory (19). The original tumor line was obtained by chemical induction with DMH in C57BL/6 mice (3) and was a gift from Dr. T. Corbett of the Southern Research Institute, Birmingham, Alabama. MCA-38 cells from tissue culture passage generations 13-17 were used. After trypsinization, tumor cells were washed and' resuspended in medium consisting of RPMI-1640 with 9% heatinactivated fetal calf serum and 0.05% penicillin-streptomycin-neomycin solution (GIBCO #564; Grand Island Biological Co., Grand Island, N. Y.). Cell viability was assessed by trypan blue dye exclusion (GIBCO #525), and cell number was assessed on a hemacytometer. Preparations of 108 cells/ml with greater than 98% viability were used. Orthotopz'c transplantatz'on. -Mice were systemically anesthetized by peritoneal injection of pentobarbital sodium (0.05 mg/g body wt; Abbott Laboratories, North Chicago, Ill.). After the abdomen was completely shaved, the mouse was placed supine in a 10-em petri dish and its limbs were held down by adhesive tape; the skin was prepared with 70% ethanol and a 1.0- to 1.5-em cutaneous incision was made along the abdominal midline followed by an incision through the parietal peritoneum. Then 0.01 ml cell inoculum (10 6 tumor cells) was prepared in a I-emS tuberculin syringe with a 30-gauge I-inch stainless steel hypodermic needle (Jeffrey-Fell Co., Buffalo, N.Y.). The cecum was carefully exteriorized and the tumor cell suspension injected into the apical lymphoid follicle. A successful inoculation turned the follicle initially white and resulted in the formation of a "bleb." The hypodermic needle was withdrawn 15-20 seconds after injection of the cell suspension to allow for the equilibration of the extracellular fluid pressure. Inoculum seepage or bleeding was stopped with a cotton-tipped applicator (Chesebrough-Pond's Inc., New York, N.Y.). The lack of "bleb" formation generally resulted in intraluminal seepage. After the cecum was replaced, the abdominal wall was closed with hemoclips (Edward Week &
ABBREVIATIONS USED: DMH=I.2-dimethylhydrazine; H & E=hematoxylin and eosin.
Received April 7 • 1977; accepted June 27. 1977. Supported in part by Public Health Service grant CA15263 from the National Cancer Institute. S Tumor Immunology Group. Department of Surgical Oncology. New York State Department of Health. Roswell Park Memorial Institute. 666 Elm St .. Buffalo. N. Y. 14263. • Department of Experimental Pathology. Roswell Park Memorial Institute Division. State University of New York at Buffalo. Roswell Park Memorial Institute. 5 We thank Dr. Stephen H. Leveson for his valuable suggestions throughout this study and Ms. Joann Vandrei for her assistance in the preparation of this manuscript. I
2
J NATL CANCER INST
15SS
TAN, HOLYOKE, AND GOLDROSEN
Co., Inc., Long Island City, N.Y.) and the skin closed with wound clips (Clay Adams Inc., New York, N.Y.). Following this operation, the mouse was wrapped in a hand towel, returned to its cage, and allowed to recover under the warmth of incandescent light. The wound clips were removed a week later. The same procedure was used at the other sites along the intestinal tract, and whenever applicable, the thickness of the local lymphoid follicles were used as a site of transplantation. Experiment I.-Mice in groups of 10 were given transplants of 10 6 viable MCA-SS cells at the following sites: distal colon, proximal colon, cecum, ileum, jejunum, and duodenum; they were killed 14 weeks later. Control mice in groups of 5 were given corresponding injections of 0.01 ml suspension medium at the same sites along the intestinal tract. Experiment 2.-Mice in groups of 10 (1 group/wk for 10 wk) were given transplants of 10 6 MCA-S8 cells in the cecal submucosae and were killed 1 week after the last injection. In addition, 1 group at 5 days and groups at 72, 48, and 24 hours before being killed were inoculated in the cecal submucosae. Then 0.01 ml of suspension medium was correspondingly injected into the cecal submucosae of a group of 5 control mice. Observations and necropsy. - The general health of these mice was observed daily and their weights were noted weekly. When the animals were killed, the tumor foci and other internal organs were grossly assessed under a stereomicroscope. The proportion of mice with tumor at injection sites and other sites was recorded. The presence of tumor within the peritoneal cavity but not at the injection site was presumed due to extraluminal seepage, whereas the absence of tumor at the injection site or any other site was presumed due to intraluminal seepage. Tumor and lymphoid tissues were preserved in 10% formalin-phosphate-buffered solution for histologic examination. The same procedure was used in control and experimental mice that died before the end of the experiment.
RESULTS Orthotopic transplantation of syngeneic MCA-S8 cells was well tolerated at the local lymphoid follicles of the intestinal tract. Intra-abdominal seepage of cell inoculum was less than 5% at all sites. Intraluminal seepage was less than 10% at the cecum; however, it exceeded 50% at the distal colon. This represented the lowest and highest intraluminal seepage incidence, respectively, among all tested intestinal sites. In experiment I, more than 40% of the mice died of intraluminal obstruction by the growing tumor 5-8 weeks after transplantation at the distal colon, proximal colon, ile~m, jejunum, and duodenum. At 14 weeks, the remaining mIce appeared healthy and were killed. The primary transplanted tumor measured 1.0-1.5 cm at the smallest diameter, a~d seco~dary tumor foci (0.2-0.8 cm) were commonly found III the hver. Furthermore, the spleen and mesenteric lymph nodes were enlarged. Table 1 presents the incidence of tumors at different sites along the submucosae of the bowel walls in syngeneic C57BL/6 mice. The tumor-take rate was lowest in the proximal and distal colon, intermedi-
J
NA TL CANCER INST
1.-MCA-38 orthotopic transplantation 14 weeks post
TABLE
injection G Site Distal colon Proximal colon Cecum Ileum Jejunum Duodenum
Primary tumor incidence b
Liver metastases·
Mesenteric node metastases d
3/10
2/3
112
4/10 9/10 6/10 6/10 6/10
2/4 5/9 4/6 3/6 2/6
0/2 2/5 114 113 0/2
Dose: 106 MCA-38 cells per transplant. No. of mice with tumorslN o. in group. e No. of mice with liver metastaseslN o. with tumors. d No. of mice with mesenteric node metastases/No. with liver metastases. G
b
ate in the three areas of the small intestine, and highest in the cecum (90%); 50% ofthe mice that developed tumors at t~e primary site of injection also developed macroscopic hver metastases. Macroscopic metastases in organs other than the liver were not observed. The number of mice with regional lymph node involvement was less than that with liver metastases; this suggested hematogenous spread via the portal vein rather than lymphatic spread as the major route for dissemination of disease to the liver. Microscopically, the transplanted tumor retained its prototype histology (fig. 1A). Local mononuclear lymphoid cells were nonreactive adjacent to the growing tumor (figs. 1B, 1C). Marked infiltration of blood polymorphonuclear ce~ls was observed at the tumor periphery in the lamina proprIa (figs. 2A-2F). Figures SA-SC document the micrometastases observed in a cecal venous radicle, the liver, and a mesenteric lymph node. Within 48-72 hours, an inflammatory reaction characterized by the infiltration of polymorphonuclear cells was noted (fig. 4A). At 5 days, this local inflammatory reaction could be observed at the tumor periphery (fig. 4B). At 3 weeks, the transplanted tumor became macroscopically evident and measured 0.2 cm at the smallest diameter. Macroscopic hepatic metastases were 2.4 2.0 E
~ &..I N VI
-~ ~
::;) ~
1.6
IUoCllOSCXlPtC LIVER METASTASES
-
-
-
-20% 60%
65%
1.2 0.8 0.4 0 0 WEEKS POST INJECTION
TEXTFIGU~
I.-Tumor gro~th comparison of MCA·38, sc (0) versus orthotopIC (0) transplantatIOn. Tumor cells (IXI06) inoculated. Or· thotopic transplantation was made at the cecum. Each point represents mean value from at least 8 tumors; vertical bar= the range of tumor diameters. VOL. 59, NO.5, NOVEMBER 1977
1539
ORTHOTOPIC TRANSPLANTATION OF MURINE COLON ADENOCARCINOMA
first noted at 8 weeks and microscopic metastases to the mesenteric lymph nodes were observed at 9 weeks, which again suggested that hepatic metastases were not due to lymphatic spread. The growth curve of the orthotopic transplant was compared to an sc transplant of the same number of MCA-38 cells (text-fig. 1). The results indicate that the growth rate in the submucosa of the bowel wall is one-fifth that of the sc inoculation. Furthermore, 65% of the animals with a primary orthotopic transplant had macroscopic hepatic metastases 10 weeks post injection, whereas no hepatic metastases were noted with the animals given sc transplants. In contrast, less than 20% of the mice with an sc transplant developed pulmonary metastases, whereas no mice with the orthotopic transplant developed pulmonary metastases.
DISCUSSION The DMH-treated rodent model for the study of colorectal cancer has provided valuable insight into the nature of this disease from the viewpoint of several disciplines; however, this animal model differs from human colon cancer in several major respects of their comparative natural history. In many patients, colon carcinoma occurring above the rectosigmoid junction will metastasize to the liver, whereas carcinoma occurring below the peritoneal reflection may spread directly to adjacent organs; in the murine model, hepatic metastases have not been reported. This may be due to the fact that DMH at a dose sufficient to induce colon tumors causes nonspecific cell injury, particularly in the liver (20). Alternatively, these animals may die of intestinal obstruction before the primary tumor has had an opportunity to metastasize. Thus the complementary approach of orthotopic transplantation was developed. A comparative-study was done to determine -the best site of transplantation along the small and large intestines. Tumor take was lowest in the proximal and distal colon, intermediate in the small intestine, and highest in the cecum (90%). The submucosa of the cecum offered several additional advantages as a site of transplantation. Tumor cells can be implanted in the submucosa of the cecum with relative ease and minimal manipulation in comparison to other sites. Death due to intestinal obstruction occurs in a large percentage of animals (40%) at all sites, with the exception of the cecum. Progressively growing tumors in the submucosa of the cecum can be surgically removed, whereas surgical removal of colon tumors at any other site along the murine bowel wall is technically difficult. A companson of the growth rate of an equal inoculum of MCA-38 cells in the submucosa of the cecum and as an sc implant revealed major differences in behavior of this colon tumor. The sc transplant became palpable within 1 week, whereas the orthotopic transplant required 3 weeks to achieve similar size. Once palpable, the colon tumor grew five times slower in the submucosa of the cecum (0.1 cm/wk) than in the subcutaneous area under the skin (0.5 cm/wk). Histologically, an infiltration of polymorphonuclear leukocytes around the periphery of the tumor cells in the orthotopic transplant was observed between 48 and 72 hours. Mononuclear cells of the adjacent lymphoid follicles appeared static and did not infiltrate the progressively growing VOL. 59, NO.5, NOVEMBER 1977
orthotopic tumor. The presence or absence of a mononuclear cell infiltrate appears to be an important factor in the control of tumor growth. Progressively growing, lethal murine tumors lack mononuclear cell infiltrates (unpublished data). The survival of patients with breast cancer correlates directly with the extent of mononuclear cell infiltration in the primary breast carcinoma (21). In the DMHinduced model, Martin et al. (13) observed a prominent infiltration of polymorphonuclear cells in the connective tissue of the growing tumor. More direct evidence was presented by Russell et al. (22) in the Moloney sarcoma system where both tumor progression and tumor regression occur. Polymorphonuclear leukocytes could be isolated from the tumors that grew progressively and subsequently regressed, whereas mononuclear cells were isolated primarily from those tumors that regressed. These observations are consistent with the histologic findings on the orthotopic transplants because no tumor regression was observed. Furthermore, these observations do not rule out a potential immunologic function of the gut-associated lymphoid tissue of the local immune system in the observed delay of tumor growth in the submucosa of the large intestine. A key observation of this study is the finding of hepatic metastases. Hepatic metastases in this model appear to be hematogenous in origin since the number of mesenteric lymph node metastases is lower than the number of liver metastases, regardless of the site of implantation; the microscopic liver foci appear earlier than the mesenteric lymph node metastases. This contrasts with the finding of pulmonary metastases after sc implantation and excision [(3) and unpublished data]. Thus the site of implantation of the initial cell inoculum influenced the dominant site of metastases. However, the sites of implantation of primary inoculum are not the only factor that influences formation of metastases. Direct intraportal injection of syngeneic murine colon tumor cells failed to produce hepatic metastases (unpublished data), which suggested that the presence of the primary inoculum and/or the environment of the primary inoculum promoted formation of hepatic metastases. This model system will open up several new areas of research that currently cannot be explored. The immunobiology of this model can be analyzed from the perspective of the local immune system as a focus of host resistance to progressively growing tumors. Changes in immunologic parameters in relation to the timing of hepatic metastases will give us a better idea of the function of these factors in resisting or facilitating metastatic tumor spread. Immunotherapy and chemotherapy studies aimed at secondary hep_atic neoplastic disease can now be performed in a natural model, which may help us gain a better understanding of colorectal cancer and help us design and test new modalities of therapy in an environment that parallels the natural history of human colon carcinoma more closely than do the models currently in use.
REFERENCES (1) DESCHNER EE: Experimentally induced cancer of the colon. Cancer
S4:824-828, 1974 (2) MARTIN MS, MARTIN F, MICHIELS R, et al: An experimental model for
J NA TL
CANCER INST
1540
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
TAN, HOLYOKE, AND GOLDROSEN
cancer of the colon and rectum. Intestinal carcinoma induced in the rat by 1,2-dimethylhydrazine. Digestion 8:22-34, 1973 CORBETT TH, GRISWOLD DP, ROBERTS BJ, et al: Tumor induction relationships in development of transplantable cancers of the colon in mice for chemotherapy assays, with a note on carcinogen structure. Cancer Res 35:2434-2439, 1975 THURNHERR N, DESCHNER EE, STONEHILL EH, et al: Induction of adenocarcinomas of the colon in mice by weekly injections of 1,2dimethylhydrazine. Cancer Res 33:940-945, 1973 EVANS JT, HAUSCHKA TS, MITTELMAN A: Differential susceptibility of four mouse strains to induction of multiple large-bowel neoplasms by 1,2-dimethylhydrazine. J Nat! Cancer Inst 52:999-1000,1974 EVANS JT, SHOWS TB, SPROUL EE, et al: Genetics of colon carcinogenesis in mice treated with 1,2-dimethylhydrazine. Cancer Res 37:134-136,1977 FILIPE MI: Mucous secretion in rat colonic mucosa during carcinogenesis induced by dimethylhydrazine. A morphological and histological study. Br TCancer 32:60-77, 1975 POZHARISSKI KM: Morphology and morphogenesis of experimental epithelial tumors of the intestine. J Nat! Cancer Inst 54: 1115-1135, 1975 REDDY BS, NARISAWA T, WRIGHT P, et al: Colon carcinogenesis with azoxymethane and dimethylhydrazine in germ-free rats. Cancer Res 35:287-290,1975 ROGERS AE, NEWBERNE PM: Dietary effects on chemical carcinogenesis in animal models for colon and liver tumors. Cancer Res 35:3427-3431,1975 ABEYOUNIS CJ, MILGROM F: A thermostable antigen characteristic for carcinogen-induced rat intestinal tumors. J Immunol 116:30-34, 1976 GARMAISE AB-K, ROGERS AE. NEWBERNE PM. et al: Immunological
J NATL CANCER INST
(13)
(14) (15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
detection of antigen(s) associated with rat colon carcinoma. Nature 248:706-707.1974 MARTIN F, MARTIN MS. BORDES M, et al: Antigens associated with chemically induced intestinal carcinomas in rats. Int J Cancer 15:144-151,1975 SJOGREN HO: Immunobiology of colon carcinomas. Dig Dis 19: 1033-1035.1974 STEELE G, SJOGREN HO, PRICE MR: Tumor-associated and embryonic antigens in soluble fractions of a chemically induced rat colon carcinoma. Int JCancer 16:33-51.1975 MCCALL DC. COLE JW: Transplantation of chemically induced adenocarcinomas of the colon in an inbred strain of rats. Cancer 33:1021-1026, 1974 WARD JM, YAMAMOTO RS, WEISBURGER JH, et al: Transplantation of chemically induced metastatic mucinous adenocarcinomas of the jejunum and colon in rats. J Nat! Cancer Inst :) I: 1997 -1999. 1973 BARTLETT GL. KREIDER JW. PURNELL DM: Immunotherapy of cancer in animals: Models or muddles? J NatI Cancer Inst 56:207210, 1976 TAN MH, HOLYOKE ED. GOLDROSEN MH: Murine colon adenocarcinomas: Methods for selective culture in vitro. J NatI Cancer Inst 56:871-873. 1976 HAASE P, COWEN DM, KNOWLES JC, et al: Evaluation of dimethylhydrazine induced tumours in mice as a model system for colorectal cancer. Br J Cancer 28:530-543, 1973 BLACK MM, OPLER SR, SPEER FD: Survival in breast cancer cases in relation to the primary tumor and regional lymph nodes. Surg Gynecol Obstet 100:543-551,1955 RUSSELL SW. GILLESPIE GY, HANSEN B. et al: Inflammatory cells in solid murine neoplasms. II. Cell types found throughout the course of Moloney sarcoma regression or progression. Int J Cancer 18:331-338,1976
VOL. 59, NO.5, NOVEMBER 1977
I.-A) MCA-38 orthotopic transplant showing grade III adenocarcinoma; B) nonreactive gut-associated lymphoid follicle with adjacent tumor; and C) nonreactive mesenteric lymph node with adjacent tumor _H & E. x 400
FIGURE
TAN. HOLYOKE. AND GOLDROSEN
1541
FIGURE 2. - Infiltration of polymorphonuclear leukocytes at the tumor in the lamina propria. A) distal colon; B) proximal colon; C) cecum; D) ileum; E) jejunum; and F) duodenum. H & E. X 400
1542
TAN, HOLYOKE, AND GOLDROSEN
3. - Dissemination of MCA·38 cells from an orthotopic transplant (H & E) to A) a cecal venous radicle (x 450); B) liver (x 400); and C) mesenteric lymph node (x 350).
FIGURE
TAN, HOLYOKE, AND GOLD ROSEN
1543
FIGURE 4.-A) Inflammatory reaction characterized by infiltration of polymorphonuclear cells. H & E. x 350. B) Local in· flammatory reaction at the tumor periphery. H & E. x 400
1544
TAN, HOLYOKE, AND GOLDROSEN
