Int. J. Cancer: 16, 559-570 (1975)
HUMAN ENDOMETRIAL CARCINOMAS SERIALLY TRANSPLANTED IN NUDE MICE AND ESTABLISHED IN CONTINUOUS CELL LINES
C. MERENDA B. SORDAT 2, J. P. MACH and S. CARREL Department of Biochemistry, University of Lausanne; Department of Immunology, Swiss Institute jor Experimental Cancer Research, Lausanne; and Unit of Human Cancer immunology, Lausanne Branch Ludwig Institute for Cancer Research, in conjunction with the Policlinique Midicale Universitaire, Lausanne, Switzerland.
Three out of five human endometrial carcinomas were successfully grafted into nude mice (BALBlclnulnu). Two of these tumors could be maintained by serial transplantation. The morphological characteristics displayed by the grafted tumors were comparable to those of the original carcinomas. Permanent cell lines were established from these two tumors. Reinjection of cells grown in vitro into nude mice produced nodules of identical histology as compared to original solid transplants. The influence of medroxyprogesterone acetate on tumor growth in vivo and cell proliferation in vitro was studied. This hormonal treatment did not produce any significant effect on tumor cells, either in vitro or in vivo, for the two endometrial carcinomas. After medroxyprogesterone administration, a slight but non-significant growth inhibition of the tumor cells in vitro was observed and the tumor transplants in vivo did not appear to be influenced. The experiments illustrate the possible use of this model for testing potential anti-cancer agents.
Endometrial cancer is one of the most common cancers in women, occurring after the age of 50 in 75% of cases (Nordqvist, 1974). Research into the biology of this tumor, its treatment and the mechanism of action of hormones is hampered by the absence of adequate experimental models. In an attempt to overcome this problem, we have tried to develop a suitable model by implanting human endometrial carcinomas subcutaneously into nude mice. These mice have been shown to exhibit deficiencies in all T-cellmediated immunological responses and therefore to accept heterologous grafts of human tumors (Rygaard and Povlsen, 1969; Povlsen and Rygaard, 1971). Attempts were made to establish long-term tissue culture lines from these human tumors grown in nude mice. Since progestins appear to be of therapeutic value
in cases of advanced endometrial carcinoma (Carbone et al., 1974), we attempted to study the effects of medroxyprogesterone acetate both on tumor explants growing in nude mice and on the same tumor cells cultivated in vitro. MATERIAL AND METHODS
Surgical specimens
Five different endometrial adenocarcinomas obtained from patients who had undergone hysterectomy or curetting were supplied by Dr. V. Barrelet, Department of Gynecology and Obstetrics, University of Lausanne and Dr. P. Vassilakos, Laboratory of Cytology, University of Geneva. Two of the tumors (END-1 and END-2) were of a n undifferentiated
Received: July 2, 1975.
559
MEKENDA ET AL.
type as seen by light microscopy (Figs. 1 and 2) while the other three (END-3, END-4 a n d END-5) were well-differentiated adenocarcinomas, an example of which is given in Figure 3. Heterotransplantation
Tissue fragments were minced under sterile conditions and inoculated subcutaneously with
a trocar into the dorso-lateral region of the mice within 20 to 40 min following surgical excision. I n all experiments, 8-week-old female nude mice, back-cross mated with BALB/c, were used. Originally they were supplied by G I . Bomholtgard Ltd., Ry, Denmark, a n d were later bred in our own colony under conventional conditions which allowed an average life-span
FIGURE IA
An area from the 01 iginal END-I tumor. The poorly-differentiated carcinoma is composed of epithelial cells forming cords separated by vascularized stroma. The cells have large ovoid nuclei with distinct nucleoli, Giemsa ~ 2 6 0 0 .
560
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
of 3 months. The animals were observed twice a week and tumor growth was followed by external measurements of the major and minor diameters with slide calipers up to the day of killing. The following formula was applied for evaluating the mean tumor diameter:
2
(a
= major
diameter, b
=
minor
diameter), and standard deviations were calculated. For serial passage of the transplants, the tumor-bearing animal was killed and the tumor dissected, freed from necrotic tissue, minced and reinoculated in other animals, as described above. Samples from each tumor were taken for examination by light and electron microscopy.
FIGURE l~ END-I tumor after 11 passages in nude mice. The tumor cells also form masses separated by fine septa of murine stroma. Cytological features similar to A, such as cytoplasmic basophilia, nuclear shape and nucleoli, can be recognized. Giemsa x 2080.
56 1
M E R E N D A ET AL.
FIGURE 2A An example of the original END-2 tumor material. The carcinoma shows areas of papillary type histology. Giemsa x 835.
Cell crtltures Cell cultures from these transplanted tumors were prepared by mincing solid specimens into 1- to 2-mm fragments which were dispersed into the medium and subsequently treated with 0.25% trypsin. One million tumor cells obtained by this procedure were explanted in tissue culture flasks (Falcon No. 3013) directly on plastic
562
without any interposed biological substrate. The nutrient solution consisted of Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. Chromosomal analysis was carried out on END-1 cells after treatment with colchicine according to the technique of Visfeldt et a/. (1972).
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
FIGURE 20 END-2 tumor after three passages in nude mice. Septa of murine strorna are more developed than for END-1, maintaining the papillary type histology. Numerous mitoses. Nuclei show dispersed chromatin with numerous nucleoli. Giemsa x 835. Hormone treatment
Forty-six female nude mice, divided into two groups, were grafted with 1 3 0 f 1 0 m g of solid tumor material. One group was kept as controls while the other was injected twice with 5 mg (0.1 ml) medroxyprogesterone acetate (MPA) IP (Depo-Provera 150, Upjohn Co., Kalamazoo, Mich., USA) on days 3 and 10 after tumor grafting. This represents a n average
of 0.4mg/g body weight. The tumors were measured twice a week. After 17 days, the animals were killed and the mean tumor diameters and corresponding weights were obtained. The effect of M P A in vitro was tested o n the two cell lines growing in suspension and established from END-1 a n d END-2. Tumor cells (5x104) were grown in 30-ml tissue culture flasks a n d fed with 5 ml of medium
563
MERENDA ET AL.
FIGURE 3A An example of the original E N D - 3 tumor which is a differentiated carci-
noma. Glandular formations are maintained. Note the presence of cells with clear nuclei and distinct nucleoli. Giemsa x 835.
to which MPA was added in concentrations of 0.5 ,ug/ml, I ,ug/ml, 4 ,ug/ml, 10 ,ug/ml. MPA stock solutions were prepared in 0.25 % dimethylsulfoxide (DMSO). DMSO alone was added to cultures of endometrial cells as a control. Two other cell lines, a malignant melanoma (SK Me-I) and a lymphoblastoid cell line (LIK) were handled in the same manner. All experiments were performed in duplicate or triplicate. The
564
cultures were sampled each day, the cells counted in a hemocytometer, and results expressed as cell numbers per ml of culture fluid. RESULTS
Heterotransp'antntion Successful tumor grafts were observed in three of the five endometrial carcinomas trans-
TRANSPLANTED H U M A N ENDOMETRIAL CARCINOMA
FIGURE 3B
END-3 tumor after the 1st passage in nude mice. The murine stroma is more abundant and includes cords of tumor cells comparable to those on A. Note the cell infiltration between the tumor cords. Giemsa x 835.
planted. They were characterized by a palpable nodule appearing 3 t o 5 weeks after inoculation. The growth rate of END-1 increased progressively during the first three passages and then remained stable with an average increment of 4 to 5 g of tumor mass in 4 weeks. The growth rate of END-2 was slower, yielding approximately 1.5 g of tumor mass in 4 weeks. One hundred percent takes were observed with END-1 and END-2 when transferred from solid explants.
END-3 grew very slowly whereas no actual growth was observed with END-4 and END-5. Serial transplantation from animal to animal was achieved for END-1 and END-2 which are presently at their 30th passage. END-3 did not grow after the third passage. To date, no regional or distant metastases have been observed in the transplanted mice. Repeated chromosomal analysis of END-1 constantly demonstrated human female karyotype.
565
MERENDA ET AL.
566
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
Cell cultures
~
4
Attempts were made to establish cell cultures from END-1 and END-2 which were serially transplantable. They were performed on material of both tumors originating from different passages. Permanent cell lines of epithelioid morphology could be obtained (Fig. 4~ and 8 ) . Both of them grew chiefly in suspension. Less than 1 % of the cells adhered to the surface of the plastic, where they formed epithelial-like colonies. At first, fibroblast-like cells, presumably of murine origin, were admixed with the tumor cells but disappeared in subsequent passages in vitro. This was facilitated by the fact that only cells in suspension were utilized for the transfers. Under these conditions, the five attempts which were made to establish cell cultures from END-I and END-2 solid transplants were successful. The doubling time was calculated for both END-I and END-2 in vitro cultures and found to be respectively 14 and 16 h; these values increased after the two first divisions under the culture conditions. Cells of END-1 established in vitro (Fig. 4A) were retransplantable into nude mice. Inoculation of 5x10g tumor cells SC gave rise to lobulated solid masses averaging 0.5 g in weight after 2 weeks in approximately 50% of the experiments. These nodules proved to be easily retransplantable and are, at the present time, in the 28th passage. Inoculation of END-2 cells grown in vitro (Fig. 48) also gave rise to retransplantable nodules but with a slower growth rate than END-1.
composed of neoplastic epithelial cells which formed irregular cords and masses and were separated by septa of vascularized mouse connective tissue (Figs. 18 and 28). For the three endometrial carcinomas investigated, the degree of differentiation of the original tumor was preserved during transfers. As illustrated in Figures 1, 2 and 3, the histological features for each type of carcinoma correlated well with those of the tumors from which they were derived. Likewise, the morphology of the tumor tissue did not change when transplants were initiated from cultured cells (Figs. 18 and 4c). In both instances, the tumor cells showed comparable nuclear/cytoplasmic ratios, basophilia and high mitotic activity. As the tumors became larger (>1 g for END-I), necrosis set in and increased progressively. Viable cells with mitotic activity formed perivascular sheaths and were supported by a stroma of murine origin consisting of capillaries, collagen fibers, fibroblastic and monocytic cells. At the ultrastructural level, as illustrated by the undifferentiated tumor END-1 in Figure 4c, the cells exhibited a large nucleus with numerous perichromatin granules, a cytoplasm rich in polysomes with occasional ergastoplasmic profiles. Cell-to-cell junctions were rare. Prolonged survivals of END-4 and END-5 carcinoma cells could be demonstrated histologically in spite of the lack of active proliferation and of grossly visible nodules, Furthermore, in these two tumors, apparently non-neoplastic glands present within the original tissue fragment appeared to survive somewhat better than the carcinoma cells when transplanted in nude mice.
Morphological examination
Hormone treatment in vitro
Microscopic investigation showed that all the tumors obtained in nude mice, either from solid material or from suspension cultures, were
The effect of MPA on END-1 and END-2 cell proliferation was studied in vitro. MPA/ DMSO at the highest concentration tested
FIGURE 4 A : An epithelial-type colony established in vifro from END-1 solid transplants. Cells are of large size with a basophilic cytoplasm, a large nucleus and numerous nucleoli. Giemsa x 2,700. B : Cell colony established in vitro from END-2 transplant. Cytological features are quite comparable to those of END-I in A. Giemsa ~2,700. c: END-I cells established in vitro and reinjected into the nude mice. This illustrates the 10th passage after retransplantation. Morphological characteristics are similar to those represented in Figure 18. Giemsa x 2,200. D : Cells at the ultrastructural level from END-I transplant (2nd passage). The cytoplasm is rich in polysomes with occasional profiles of ergastoplasm. No cell-to-celljunction can be seen in this field. Uranyl-lead x 12,500.
567
MERENDA ET AL.
(10 ,ug/ml) produced a 25 % inhibition as compared to the DMSO controls when followed for 8 days of continuous culture. This corresponded to a prolongation of the doubling time from approximately 14 h to 21 h in the END-1 tumor cells and from 16 h to 21 h in the END-2 cells, in the presence of MPA. Similar results, however, were obtained with two unrelated human cell lines, the SK-Me-I melanoma and the lymphoblastoid LIK cells. The doubling time of LIK cells which, under the culture conditions used, was estimated to be 18 h and thus found t o be comparable t o the values obtained for END-1 and END-2 cells, was prolonged in the presence of 10 ,ug/ml MPA to 23 h, resulting in a decrease in cell numbers of about 20% over 7 days of culture. A similar reduction of approximately 1 5 %
20
-
could also be observed in SK-Me4 cells when cultured for 8 days. No significant difference in sensitivity between the two endometrial and the two unrelated human cell lines could therefore be observed, suggesting absence of specific inhibition of END-1 and END-2 by the MPA concentration tested. Hormone treatment in vivo
The effects of treating nude mice bearing END-1 solid transplants (from the 24th passage) with MPA are represented in Figure 5 . Tumor size is expressed in mean diameters plotted versus days following inoculation. No significant difference was detected, either in the size of the tumors or in their weights at the day of killing, between the MPA-treated and the control animals. Mean tumor weights were respectively
0--.
NON TREATED CONTROLS
(19 MICE)
0-0
TREATED WITH DEPO-PROVERA
(16 MICE)
7
?
-
2 rt
E
15
-
2 a
-
-
K
0
-
3t
g5 10 -
I'
i
5 1
4
568
4
I T
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
4.6812.49 g for the treated and 3.9*1.16 g for the untreated mice. The correlation coefficient (r) between mean diameters and weights was found to be satisfactory (r = 0.877). Similar results were obtained with END-2 solid transplants following MPA treatment. The END-2 had a slower growth rate and the tumor weights, 15 days after grafting, were respectively 0.24*0.18 g for control and 0.28*0.23 g for treated mice. Moreover, histological examination of tumor material taken at the day of killing from treated and untreated groups did not demonstrate any detectable morphological difference in the course of these experiments. DISCUSSION
The results demonstrate the possibility of grafting and serially maintaining endometrial carcinomas of human origin in nude mice. Two poorly-differentiated tumors grew well and could be transferred whereas three well-differentiated tumors did not grow continuously in this system. Whether the degree of differentiation or of hormonal dependence, or local factors such as the extent of vascularization, play a role in the successful transplantation of these carcinomas is open to question, In our experience, breast carcinomas could not be transplanted o r serially maintained when obtained from patients as surgical specimens; however, breast carcinoma cells from a permanent line were observed to form nodules when transplanted in nude mice (Ozzello et al., 1974). In contrast, a high percentage of colorectal carcinomas representing various degrees of differentiation were successfully transplanted and maintained in nude mice (Povlsen and Rygaard, 1971; Sordat et al., 1974). The results reported here for endometrial carcinomas END-1 and END-2 fulfill certain criteria for an experimental tumor model: (1) persistence of the original morphological characteristics of the tumor tissues; (2) a 100% tumor take when retransplanted; (3) local subcutaneous spreading, allowing quantitative measurements of tumor growth; and (4) the possibility of establishing continuous cell lines in vitro from the transplanted tumors. The limiting factor in this system appears to be the relatively short survival of nude mice under
conventional conditions. It is of interest t o note that apparently normal glands present within the original well-differentiated tumors survived somewhat better than carcinoma cells. Our results of MPA treatment for two poorlydifferentiated endometrial carcinomas transplanted in nude mice and maintained as established cell lines show no significant effect of the hormone either in vivo or in vitro. Sekiya et al. (1974~)described a reduction in vitro of the survival rate of a chemically-induced rat endometrial carcinoma in the presence of 8 ,ug/ml of progesterone, as compared to untreated controls. In our results, the inhibitory effect observed in vitro is considered t o be not specific when compared to the effect of MPA on two unrelated human cell lines. On the other hand, Nordqvist (1974), using suspensions of human endometrial carcinomas in short-term cultures and treated with 10 ,ug/ml of progesterone, obtained a significant inhibition of DNA synthesis in two out of 14 specimens. The same author suggested a possible correlation between these in vitro results and the clinical therapeutic response in patients. In contrast, Sekiya et al. (19746) reported an enhancement in the growth of a rat endometrial cell line in female animals treated with a total dose of 5 mg of MPA. In humans, it is known that a temporary, and in some cases a complete remission of uterine tumor, occurs in roughly 1/3 of the patients treated with progesterone (Reifenstein, 1974). It has also been reported that histologically well-differentiated tumors are more likely to respond to progesterone therapy than poorlydifferentiated types (Carbone, 1974). There have been few reports on non-hormone chemotherapy of human endometrial carcinomas and the experimental system presented here may prove to be useful for the investigation of other potential anti-cancer agents.
ACKNOWLEDGEMENTS
We thank Mrs. L. Kolly, Mrs. J. Bamat and Miss U. Atzli for expert technical assistance. We thank also Profs. L. Ozzello and K. T. Brunner for advice and criticism, and Dr. H. Engers for reviewing the manuscript. This work was partially supported by grants from the Swiss National Foundation for Scientific Research.
569
MERENDA ET AL.
CARCINOMES ENDOMETRIAUX HU MA1 NS GREFFES DE FACON SERIEE CHEZ DES SOURIS “ N U D E ” ET ETABLIS EN LIGNEES CELLULAIRES CONTINUES Sur cinq carcinomes endomttriaux humains, trois ont pu Ptre greffis avec succi..~ chez des souris ‘‘ nude ’’ (BALBlclnulnu). Deux de ces trois tumeurs ont t t k maintenues de fagon strike. Les caracttristiques morphologiques presenttes par les tumeurs greffkes ttaient identiques a celles des tumeurs d’origine. Des ligntes cellulaires continues ont pu Ptre Ptablies ir partir de transplants de ces deux tumeurs. La reinjection a des souris ‘‘ nude” de cellules provenant de ces deux ligne‘es a induit la croissance de nodules histologiquement identiques aux transplants d’origine solide. L’effet de I’administration d’acttate de mPdroxyprogestProne sur la croissance tumorale in vivo a PtP comparP a celiri exercP par la mime substance sur la prolifiration cellulaire in vitro. Le traitement hormonal n’a provoque aucun effet significatif sur les cellules tumorales in vitro et in vivo pour les deux carcinomes endometriaux examinis. Apris administration de mtdroxyprogestkrone, une inhibition non sptcifique de la prolifPuation cellulaire in vitro a Ptt constatPe et, de mPrne, la croissance des greffes i n vivo n’a pas paru influencee de fagon signijicative. Ces experiences mettent en Pvidence I’utilisation possible de ce modkle pour tester d’autres agents anti-canctreux potentiels.
REFERENCES
CARBONE, P. P., and CARTER, S. K., Endometrial cancer: approach to development of effective chemotherapy. Gynec. Oncol., 2, 348-353 (1974). NORDQVIST, S. R. B., In vitro effects of progestins on D N A synthesis in metastatic endometrial carcinoma. Gynec. Oncol., 2, 41 5-428 (1974). C., CARREL, S., OZZELLO,L., SORDAT,B., MERENDA, HURLIMANN, J., and MACH,J. P., Transplantation of a human mammary carcinoma cell line (BT20) into nude mice. J. nut. Cancer Ins?., 52, 1669-1672 (1 974). POVLSEN,C. O., and RYGAARD,J., Heterotransplantation of human adenocarcinomas of the colon and rectum to the mouse mutant nude. A study of nine consecutive transplantations. Acta path. microhiol. scand. Sect. A, 79,153-169 (1971). REIFENSTEIN, E. C., The treatment of advanced endometrial canc-,r with hydroxyprogesterone capronate. Gynec. Oncol., 2, 377-414 (1974). RYGAARD,J., and POVLSEN,C. O., Heterotransplantation of a human malignant tumor to nude mice. Acta path. microbiol. scand., 77, 758-760 (1969).
570
SEKIYA, S., TAKAMIZAWA, H., KIKUCHI, Y., YANO,A., and KAMIYAMA, M., Effect of progesterone on uterine adenocarcinoma of the rat in vitro and in vivo. Abstract Xlth Int. Cancer Congress, Florence, Italy, p. 408 (1974a). SEKIYA,S., YAND, A., and TAKAMIZAWA, H., Enhancement of tumor growth and metastases by medroxyprogesterone acetate in transplanted uterine adenocarcinoma cells of rat. 1.nut. Cancer Inst., 52, 297-298 (19746). SORDAT,B., FRITSCHE, R., MACH,J. P., CARREL, S., OZZELLO, L., and CEROTTINI, J.-C., Morphological and functional evaluation of human solid tumors serially transplanted into nude mice. In: Proceedings Ist International Workshop on Nude Mice, Scanticon, Aarhus, Denmark 1973, p. 269278, Gustav Fischer Verlag, Stuttgart, Germany (1974). VISFELDT,J., POVLSEN,C. O., and RYGAARD,J., Chromosome analyses of human tumours following heterotransplantation to the mouse mutant nude. Acta path. microbiol. scand., 80, 169-176 (1972).
HUMAN ENDOMETRIAL CARCINOMAS SERIALLY TRANSPLANTED IN NUDE MICE AND ESTABLISHED IN CONTINUOUS CELL LINES
C. MERENDA B. SORDAT 2, J. P. MACH and S. CARREL Department of Biochemistry, University of Lausanne; Department of Immunology, Swiss Institute jor Experimental Cancer Research, Lausanne; and Unit of Human Cancer immunology, Lausanne Branch Ludwig Institute for Cancer Research, in conjunction with the Policlinique Midicale Universitaire, Lausanne, Switzerland.
Three out of five human endometrial carcinomas were successfully grafted into nude mice (BALBlclnulnu). Two of these tumors could be maintained by serial transplantation. The morphological characteristics displayed by the grafted tumors were comparable to those of the original carcinomas. Permanent cell lines were established from these two tumors. Reinjection of cells grown in vitro into nude mice produced nodules of identical histology as compared to original solid transplants. The influence of medroxyprogesterone acetate on tumor growth in vivo and cell proliferation in vitro was studied. This hormonal treatment did not produce any significant effect on tumor cells, either in vitro or in vivo, for the two endometrial carcinomas. After medroxyprogesterone administration, a slight but non-significant growth inhibition of the tumor cells in vitro was observed and the tumor transplants in vivo did not appear to be influenced. The experiments illustrate the possible use of this model for testing potential anti-cancer agents.
Endometrial cancer is one of the most common cancers in women, occurring after the age of 50 in 75% of cases (Nordqvist, 1974). Research into the biology of this tumor, its treatment and the mechanism of action of hormones is hampered by the absence of adequate experimental models. In an attempt to overcome this problem, we have tried to develop a suitable model by implanting human endometrial carcinomas subcutaneously into nude mice. These mice have been shown to exhibit deficiencies in all T-cellmediated immunological responses and therefore to accept heterologous grafts of human tumors (Rygaard and Povlsen, 1969; Povlsen and Rygaard, 1971). Attempts were made to establish long-term tissue culture lines from these human tumors grown in nude mice. Since progestins appear to be of therapeutic value
in cases of advanced endometrial carcinoma (Carbone et al., 1974), we attempted to study the effects of medroxyprogesterone acetate both on tumor explants growing in nude mice and on the same tumor cells cultivated in vitro. MATERIAL AND METHODS
Surgical specimens
Five different endometrial adenocarcinomas obtained from patients who had undergone hysterectomy or curetting were supplied by Dr. V. Barrelet, Department of Gynecology and Obstetrics, University of Lausanne and Dr. P. Vassilakos, Laboratory of Cytology, University of Geneva. Two of the tumors (END-1 and END-2) were of a n undifferentiated
Received: July 2, 1975.
559
MEKENDA ET AL.
type as seen by light microscopy (Figs. 1 and 2) while the other three (END-3, END-4 a n d END-5) were well-differentiated adenocarcinomas, an example of which is given in Figure 3. Heterotransplantation
Tissue fragments were minced under sterile conditions and inoculated subcutaneously with
a trocar into the dorso-lateral region of the mice within 20 to 40 min following surgical excision. I n all experiments, 8-week-old female nude mice, back-cross mated with BALB/c, were used. Originally they were supplied by G I . Bomholtgard Ltd., Ry, Denmark, a n d were later bred in our own colony under conventional conditions which allowed an average life-span
FIGURE IA
An area from the 01 iginal END-I tumor. The poorly-differentiated carcinoma is composed of epithelial cells forming cords separated by vascularized stroma. The cells have large ovoid nuclei with distinct nucleoli, Giemsa ~ 2 6 0 0 .
560
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
of 3 months. The animals were observed twice a week and tumor growth was followed by external measurements of the major and minor diameters with slide calipers up to the day of killing. The following formula was applied for evaluating the mean tumor diameter:
2
(a
= major
diameter, b
=
minor
diameter), and standard deviations were calculated. For serial passage of the transplants, the tumor-bearing animal was killed and the tumor dissected, freed from necrotic tissue, minced and reinoculated in other animals, as described above. Samples from each tumor were taken for examination by light and electron microscopy.
FIGURE l~ END-I tumor after 11 passages in nude mice. The tumor cells also form masses separated by fine septa of murine stroma. Cytological features similar to A, such as cytoplasmic basophilia, nuclear shape and nucleoli, can be recognized. Giemsa x 2080.
56 1
M E R E N D A ET AL.
FIGURE 2A An example of the original END-2 tumor material. The carcinoma shows areas of papillary type histology. Giemsa x 835.
Cell crtltures Cell cultures from these transplanted tumors were prepared by mincing solid specimens into 1- to 2-mm fragments which were dispersed into the medium and subsequently treated with 0.25% trypsin. One million tumor cells obtained by this procedure were explanted in tissue culture flasks (Falcon No. 3013) directly on plastic
562
without any interposed biological substrate. The nutrient solution consisted of Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. Chromosomal analysis was carried out on END-1 cells after treatment with colchicine according to the technique of Visfeldt et a/. (1972).
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
FIGURE 20 END-2 tumor after three passages in nude mice. Septa of murine strorna are more developed than for END-1, maintaining the papillary type histology. Numerous mitoses. Nuclei show dispersed chromatin with numerous nucleoli. Giemsa x 835. Hormone treatment
Forty-six female nude mice, divided into two groups, were grafted with 1 3 0 f 1 0 m g of solid tumor material. One group was kept as controls while the other was injected twice with 5 mg (0.1 ml) medroxyprogesterone acetate (MPA) IP (Depo-Provera 150, Upjohn Co., Kalamazoo, Mich., USA) on days 3 and 10 after tumor grafting. This represents a n average
of 0.4mg/g body weight. The tumors were measured twice a week. After 17 days, the animals were killed and the mean tumor diameters and corresponding weights were obtained. The effect of M P A in vitro was tested o n the two cell lines growing in suspension and established from END-1 a n d END-2. Tumor cells (5x104) were grown in 30-ml tissue culture flasks a n d fed with 5 ml of medium
563
MERENDA ET AL.
FIGURE 3A An example of the original E N D - 3 tumor which is a differentiated carci-
noma. Glandular formations are maintained. Note the presence of cells with clear nuclei and distinct nucleoli. Giemsa x 835.
to which MPA was added in concentrations of 0.5 ,ug/ml, I ,ug/ml, 4 ,ug/ml, 10 ,ug/ml. MPA stock solutions were prepared in 0.25 % dimethylsulfoxide (DMSO). DMSO alone was added to cultures of endometrial cells as a control. Two other cell lines, a malignant melanoma (SK Me-I) and a lymphoblastoid cell line (LIK) were handled in the same manner. All experiments were performed in duplicate or triplicate. The
564
cultures were sampled each day, the cells counted in a hemocytometer, and results expressed as cell numbers per ml of culture fluid. RESULTS
Heterotransp'antntion Successful tumor grafts were observed in three of the five endometrial carcinomas trans-
TRANSPLANTED H U M A N ENDOMETRIAL CARCINOMA
FIGURE 3B
END-3 tumor after the 1st passage in nude mice. The murine stroma is more abundant and includes cords of tumor cells comparable to those on A. Note the cell infiltration between the tumor cords. Giemsa x 835.
planted. They were characterized by a palpable nodule appearing 3 t o 5 weeks after inoculation. The growth rate of END-1 increased progressively during the first three passages and then remained stable with an average increment of 4 to 5 g of tumor mass in 4 weeks. The growth rate of END-2 was slower, yielding approximately 1.5 g of tumor mass in 4 weeks. One hundred percent takes were observed with END-1 and END-2 when transferred from solid explants.
END-3 grew very slowly whereas no actual growth was observed with END-4 and END-5. Serial transplantation from animal to animal was achieved for END-1 and END-2 which are presently at their 30th passage. END-3 did not grow after the third passage. To date, no regional or distant metastases have been observed in the transplanted mice. Repeated chromosomal analysis of END-1 constantly demonstrated human female karyotype.
565
MERENDA ET AL.
566
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
Cell cultures
~
4
Attempts were made to establish cell cultures from END-1 and END-2 which were serially transplantable. They were performed on material of both tumors originating from different passages. Permanent cell lines of epithelioid morphology could be obtained (Fig. 4~ and 8 ) . Both of them grew chiefly in suspension. Less than 1 % of the cells adhered to the surface of the plastic, where they formed epithelial-like colonies. At first, fibroblast-like cells, presumably of murine origin, were admixed with the tumor cells but disappeared in subsequent passages in vitro. This was facilitated by the fact that only cells in suspension were utilized for the transfers. Under these conditions, the five attempts which were made to establish cell cultures from END-I and END-2 solid transplants were successful. The doubling time was calculated for both END-I and END-2 in vitro cultures and found to be respectively 14 and 16 h; these values increased after the two first divisions under the culture conditions. Cells of END-1 established in vitro (Fig. 4A) were retransplantable into nude mice. Inoculation of 5x10g tumor cells SC gave rise to lobulated solid masses averaging 0.5 g in weight after 2 weeks in approximately 50% of the experiments. These nodules proved to be easily retransplantable and are, at the present time, in the 28th passage. Inoculation of END-2 cells grown in vitro (Fig. 48) also gave rise to retransplantable nodules but with a slower growth rate than END-1.
composed of neoplastic epithelial cells which formed irregular cords and masses and were separated by septa of vascularized mouse connective tissue (Figs. 18 and 28). For the three endometrial carcinomas investigated, the degree of differentiation of the original tumor was preserved during transfers. As illustrated in Figures 1, 2 and 3, the histological features for each type of carcinoma correlated well with those of the tumors from which they were derived. Likewise, the morphology of the tumor tissue did not change when transplants were initiated from cultured cells (Figs. 18 and 4c). In both instances, the tumor cells showed comparable nuclear/cytoplasmic ratios, basophilia and high mitotic activity. As the tumors became larger (>1 g for END-I), necrosis set in and increased progressively. Viable cells with mitotic activity formed perivascular sheaths and were supported by a stroma of murine origin consisting of capillaries, collagen fibers, fibroblastic and monocytic cells. At the ultrastructural level, as illustrated by the undifferentiated tumor END-1 in Figure 4c, the cells exhibited a large nucleus with numerous perichromatin granules, a cytoplasm rich in polysomes with occasional ergastoplasmic profiles. Cell-to-cell junctions were rare. Prolonged survivals of END-4 and END-5 carcinoma cells could be demonstrated histologically in spite of the lack of active proliferation and of grossly visible nodules, Furthermore, in these two tumors, apparently non-neoplastic glands present within the original tissue fragment appeared to survive somewhat better than the carcinoma cells when transplanted in nude mice.
Morphological examination
Hormone treatment in vitro
Microscopic investigation showed that all the tumors obtained in nude mice, either from solid material or from suspension cultures, were
The effect of MPA on END-1 and END-2 cell proliferation was studied in vitro. MPA/ DMSO at the highest concentration tested
FIGURE 4 A : An epithelial-type colony established in vifro from END-1 solid transplants. Cells are of large size with a basophilic cytoplasm, a large nucleus and numerous nucleoli. Giemsa x 2,700. B : Cell colony established in vitro from END-2 transplant. Cytological features are quite comparable to those of END-I in A. Giemsa ~2,700. c: END-I cells established in vitro and reinjected into the nude mice. This illustrates the 10th passage after retransplantation. Morphological characteristics are similar to those represented in Figure 18. Giemsa x 2,200. D : Cells at the ultrastructural level from END-I transplant (2nd passage). The cytoplasm is rich in polysomes with occasional profiles of ergastoplasm. No cell-to-celljunction can be seen in this field. Uranyl-lead x 12,500.
567
MERENDA ET AL.
(10 ,ug/ml) produced a 25 % inhibition as compared to the DMSO controls when followed for 8 days of continuous culture. This corresponded to a prolongation of the doubling time from approximately 14 h to 21 h in the END-1 tumor cells and from 16 h to 21 h in the END-2 cells, in the presence of MPA. Similar results, however, were obtained with two unrelated human cell lines, the SK-Me-I melanoma and the lymphoblastoid LIK cells. The doubling time of LIK cells which, under the culture conditions used, was estimated to be 18 h and thus found t o be comparable t o the values obtained for END-1 and END-2 cells, was prolonged in the presence of 10 ,ug/ml MPA to 23 h, resulting in a decrease in cell numbers of about 20% over 7 days of culture. A similar reduction of approximately 1 5 %
20
-
could also be observed in SK-Me4 cells when cultured for 8 days. No significant difference in sensitivity between the two endometrial and the two unrelated human cell lines could therefore be observed, suggesting absence of specific inhibition of END-1 and END-2 by the MPA concentration tested. Hormone treatment in vivo
The effects of treating nude mice bearing END-1 solid transplants (from the 24th passage) with MPA are represented in Figure 5 . Tumor size is expressed in mean diameters plotted versus days following inoculation. No significant difference was detected, either in the size of the tumors or in their weights at the day of killing, between the MPA-treated and the control animals. Mean tumor weights were respectively
0--.
NON TREATED CONTROLS
(19 MICE)
0-0
TREATED WITH DEPO-PROVERA
(16 MICE)
7
?
-
2 rt
E
15
-
2 a
-
-
K
0
-
3t
g5 10 -
I'
i
5 1
4
568
4
I T
TRANSPLANTED HUMAN ENDOMETRIAL CARCINOMA
4.6812.49 g for the treated and 3.9*1.16 g for the untreated mice. The correlation coefficient (r) between mean diameters and weights was found to be satisfactory (r = 0.877). Similar results were obtained with END-2 solid transplants following MPA treatment. The END-2 had a slower growth rate and the tumor weights, 15 days after grafting, were respectively 0.24*0.18 g for control and 0.28*0.23 g for treated mice. Moreover, histological examination of tumor material taken at the day of killing from treated and untreated groups did not demonstrate any detectable morphological difference in the course of these experiments. DISCUSSION
The results demonstrate the possibility of grafting and serially maintaining endometrial carcinomas of human origin in nude mice. Two poorly-differentiated tumors grew well and could be transferred whereas three well-differentiated tumors did not grow continuously in this system. Whether the degree of differentiation or of hormonal dependence, or local factors such as the extent of vascularization, play a role in the successful transplantation of these carcinomas is open to question, In our experience, breast carcinomas could not be transplanted o r serially maintained when obtained from patients as surgical specimens; however, breast carcinoma cells from a permanent line were observed to form nodules when transplanted in nude mice (Ozzello et al., 1974). In contrast, a high percentage of colorectal carcinomas representing various degrees of differentiation were successfully transplanted and maintained in nude mice (Povlsen and Rygaard, 1971; Sordat et al., 1974). The results reported here for endometrial carcinomas END-1 and END-2 fulfill certain criteria for an experimental tumor model: (1) persistence of the original morphological characteristics of the tumor tissues; (2) a 100% tumor take when retransplanted; (3) local subcutaneous spreading, allowing quantitative measurements of tumor growth; and (4) the possibility of establishing continuous cell lines in vitro from the transplanted tumors. The limiting factor in this system appears to be the relatively short survival of nude mice under
conventional conditions. It is of interest t o note that apparently normal glands present within the original well-differentiated tumors survived somewhat better than carcinoma cells. Our results of MPA treatment for two poorlydifferentiated endometrial carcinomas transplanted in nude mice and maintained as established cell lines show no significant effect of the hormone either in vivo or in vitro. Sekiya et al. (1974~)described a reduction in vitro of the survival rate of a chemically-induced rat endometrial carcinoma in the presence of 8 ,ug/ml of progesterone, as compared to untreated controls. In our results, the inhibitory effect observed in vitro is considered t o be not specific when compared to the effect of MPA on two unrelated human cell lines. On the other hand, Nordqvist (1974), using suspensions of human endometrial carcinomas in short-term cultures and treated with 10 ,ug/ml of progesterone, obtained a significant inhibition of DNA synthesis in two out of 14 specimens. The same author suggested a possible correlation between these in vitro results and the clinical therapeutic response in patients. In contrast, Sekiya et al. (19746) reported an enhancement in the growth of a rat endometrial cell line in female animals treated with a total dose of 5 mg of MPA. In humans, it is known that a temporary, and in some cases a complete remission of uterine tumor, occurs in roughly 1/3 of the patients treated with progesterone (Reifenstein, 1974). It has also been reported that histologically well-differentiated tumors are more likely to respond to progesterone therapy than poorlydifferentiated types (Carbone, 1974). There have been few reports on non-hormone chemotherapy of human endometrial carcinomas and the experimental system presented here may prove to be useful for the investigation of other potential anti-cancer agents.
ACKNOWLEDGEMENTS
We thank Mrs. L. Kolly, Mrs. J. Bamat and Miss U. Atzli for expert technical assistance. We thank also Profs. L. Ozzello and K. T. Brunner for advice and criticism, and Dr. H. Engers for reviewing the manuscript. This work was partially supported by grants from the Swiss National Foundation for Scientific Research.
569
MERENDA ET AL.
CARCINOMES ENDOMETRIAUX HU MA1 NS GREFFES DE FACON SERIEE CHEZ DES SOURIS “ N U D E ” ET ETABLIS EN LIGNEES CELLULAIRES CONTINUES Sur cinq carcinomes endomttriaux humains, trois ont pu Ptre greffis avec succi..~ chez des souris ‘‘ nude ’’ (BALBlclnulnu). Deux de ces trois tumeurs ont t t k maintenues de fagon strike. Les caracttristiques morphologiques presenttes par les tumeurs greffkes ttaient identiques a celles des tumeurs d’origine. Des ligntes cellulaires continues ont pu Ptre Ptablies ir partir de transplants de ces deux tumeurs. La reinjection a des souris ‘‘ nude” de cellules provenant de ces deux ligne‘es a induit la croissance de nodules histologiquement identiques aux transplants d’origine solide. L’effet de I’administration d’acttate de mPdroxyprogestProne sur la croissance tumorale in vivo a PtP comparP a celiri exercP par la mime substance sur la prolifiration cellulaire in vitro. Le traitement hormonal n’a provoque aucun effet significatif sur les cellules tumorales in vitro et in vivo pour les deux carcinomes endometriaux examinis. Apris administration de mtdroxyprogestkrone, une inhibition non sptcifique de la prolifPuation cellulaire in vitro a Ptt constatPe et, de mPrne, la croissance des greffes i n vivo n’a pas paru influencee de fagon signijicative. Ces experiences mettent en Pvidence I’utilisation possible de ce modkle pour tester d’autres agents anti-canctreux potentiels.
REFERENCES
CARBONE, P. P., and CARTER, S. K., Endometrial cancer: approach to development of effective chemotherapy. Gynec. Oncol., 2, 348-353 (1974). NORDQVIST, S. R. B., In vitro effects of progestins on D N A synthesis in metastatic endometrial carcinoma. Gynec. Oncol., 2, 41 5-428 (1974). C., CARREL, S., OZZELLO,L., SORDAT,B., MERENDA, HURLIMANN, J., and MACH,J. P., Transplantation of a human mammary carcinoma cell line (BT20) into nude mice. J. nut. Cancer Ins?., 52, 1669-1672 (1 974). POVLSEN,C. O., and RYGAARD,J., Heterotransplantation of human adenocarcinomas of the colon and rectum to the mouse mutant nude. A study of nine consecutive transplantations. Acta path. microhiol. scand. Sect. A, 79,153-169 (1971). REIFENSTEIN, E. C., The treatment of advanced endometrial canc-,r with hydroxyprogesterone capronate. Gynec. Oncol., 2, 377-414 (1974). RYGAARD,J., and POVLSEN,C. O., Heterotransplantation of a human malignant tumor to nude mice. Acta path. microbiol. scand., 77, 758-760 (1969).
570
SEKIYA, S., TAKAMIZAWA, H., KIKUCHI, Y., YANO,A., and KAMIYAMA, M., Effect of progesterone on uterine adenocarcinoma of the rat in vitro and in vivo. Abstract Xlth Int. Cancer Congress, Florence, Italy, p. 408 (1974a). SEKIYA,S., YAND, A., and TAKAMIZAWA, H., Enhancement of tumor growth and metastases by medroxyprogesterone acetate in transplanted uterine adenocarcinoma cells of rat. 1.nut. Cancer Inst., 52, 297-298 (19746). SORDAT,B., FRITSCHE, R., MACH,J. P., CARREL, S., OZZELLO, L., and CEROTTINI, J.-C., Morphological and functional evaluation of human solid tumors serially transplanted into nude mice. In: Proceedings Ist International Workshop on Nude Mice, Scanticon, Aarhus, Denmark 1973, p. 269278, Gustav Fischer Verlag, Stuttgart, Germany (1974). VISFELDT,J., POVLSEN,C. O., and RYGAARD,J., Chromosome analyses of human tumours following heterotransplantation to the mouse mutant nude. Acta path. microbiol. scand., 80, 169-176 (1972).
