APMIS 100: 772-778, 1992
Growth patterns of pulmonary metastases and primary tumours from five murine fibrosarcoma cell clones KNUT H. FALKVOLL' and IVAR AMUND GRIMSTAD* 'Institute for Surgical Research and 'Institute of Clinical Biochemistry and Institute of Pathology, University of Oslo, Rikshospitalet, Oslo, Norway
Falkvoll, K. H. & Grimstad, I. A. Growth patterns of pulmonary metastases and primary tumours from five murine fibrosarcoma cell clones. APMIS 100: 772-778, 1992. The growth patterns, including the size, shape and regional preferences, of lung metastases from five rnurine fibrosarcoma cell clones were studied. Spontaneous metastases developed from tumours formed by subcutaneous inoculation of the cell clones. Lung colonies (experimental metastases) were established by i.v. injection of cells. The numbers of both spontaneously and experimentally formed subpleural lung metastases were counted through a stereomicroscope. The fraction of colonies that was located subpleurally was determined in histological sections of lungs. The growth kinetics of clonally derived primary tumours, and the number of spontaneous and experimental lung metastases, differed greatly between certain cell clones. The number of spontaneous lung metastases was correlated with the maximum size of primary tumours. No close correlation was observed between the size of the primary tumours and the size of experimental metastases. There were differences between the cell clones in the shape and regional preferences of their lung deposits. The subpleural colonies were generally larger than the intraphlmonary ones. Thus, both the regional distribution and the growth pattern of lung deposits differed between the clones. Key words: Growth pattern; primary tumours; pulmonary metastases; murine fibrosarcomas.
K. H. Falkvoll, Institute for Surgical Research, Rikshospitalet, Langesgate 1, N-0027 Oslo 1, Norway.
Cancer metastasis is a non-random process. The sites of distant dissemination of several cancers cannot be explained by the distribution of blood flow, but may be determined by specific interactions between tumour cells and the target organs ( 17). Such specificity may reside in differential interactions of tumour cells with organ-derived microvessel endothelial cells and subendothelial matrix ( I 5). Differential responses of various tumour cell populations to tissue-mediated growth stimulators and inhibitors have also been demonstrated to be involved in the determination of organ preferences of metastases
Received December 27, 1991. Accepted April 4, 1992.
772
from some tumours (14, 15). Metastasis formation may, in addition, show regional preferences within organs (2, 5, 18, 19). Identification of tumour cell and host properties accounting for preferential location of metastases in particular organs and in certain regions within organs may shed light on what enables some neoplastic cells, but not others, to form metastases (1, 2). The aim of the present investigation was to study the growth pattern, including the size, shape and regional distribution in the lungs, of metastases from five different cell clones derived from a murine fibrosarcoma. In order to relate these data to the metastatic propensity of the clones, the formation of lung metastases from transplanted footpad tumours, as well as the formation of lung colonies following i.v. injec-
METASTASES
-
CIBNAL DIFFERENCES
tion of cells, was studied. Factors are discussed which may be involved in preferential metastasis formation in particular organs and certain regions within these organs.
MATERIALS AND METHODS Tumour cell clones The tumour cell clones originate from a fibrosarcoma induced subcutaneously by methylcholanthrene in C57 BL/6J mice. (24) The cloning by microscopically controlled limiting dilution, and initial characterization have been described elsewhere (9, 22). The cells were subcultured at least twice before use, but were not kept in culture for more than five weeks in order to prevent clonal diversification. The culture medium was RPMI 1640 supplemented with L-glutamine (292 mg/l), 10% foetal bovine serum, penicillin (50 units/ ml), streptomycin (50 pg/ml) and amphotericin B (125 ng/ml). The cells were kept at 37'C in a humidified atmosphere of air supplemented with 5% CO,. The five cell clones used in this study were labelled C,, Cz. C1,C4and C,, which correspond to the respective designations l.l/clone 3, l.l/clone 38, 1.1/L1 clone 2, l .l / L 2 clone 3 and l.l/anti-B' clone 3 used in previous reports (9, 10). Primary tumour growth Subconfluent rapidly growing cultured cells were harvested by brief trypsinization, washed, and 2 x lo5 cells, dispersed in 50 p1 serum-free RPMI 1640 medium, were inoculated into the right hind footpads of female syngeneic (C57 BL/6J) mice (Bommice, Ry, Denmark) weighing 18-20 g. The tumour growth rate was determined at regular intervals by measuring the dorsoplantar diameter of the hind limb using accurate calipers, with the opposite limb serving as control. Spontaneous metastasis Twenty-eight days after cell inoculation into the right hind footpad of the mice, 15-16 animals per clone, the legs were amputated proximal to the tumour (9). The limbs were amputated because further growth of some tumours would give ulceration of the overlying skin. When the mice were killed 14 days later, the lungs were removed, inflated with a suspension of India ink and placed in a bleaching solution as described by Wexler (28). This preparation greatly increased the visibility of surface lung metastases counted through a stereomicroscope. Randomly selected lungs were studied histologically. The presence (or absence) of metastases in other organs and the presence of primary tumours were verified histologically. Experimental metastasis - lung colony formation Lung colonies were established by i.v. injection of cells in single cell suspensions. Single cell suspensions
were obtained by allowing sedimentation of cell clumps among the trypsinized washed cells in tubes on ice for about 10 min (3). A cooled 1.0 ml syringe supplied with a 27-gauge needle was used for i.v. injection of 2 x lo5 single cells into the lateral tail vein. The trypan blue exclusion test confirmed that the viability of injected cells was always greater than 90%. After 21 days the lungs from nine to 15 mice (for each clone) were removed and prepared as described above, and the surface metastases were counted. The distribution of metastases was studied in histological sections of the lungs (five to eight mice in each group), and the numbers of subpleural and intrapulmonary metastases were scored. Two perpendicular diameters, for subpleural metastases the diameter parallel to the surface and the diameter at right angles to the pleural surface, of metastatic structures were measured on video screen projections of the sections. The cross-sectional area of metastatic structures was calculated as being elliptical. This adequately determined the cross-sectional area of nodular metastases but might have underestimated the cross-sectional area of metastases forming thin subpleural discs.
Histological preparation Lungs, hind limb tumours and other organs were fixed in 4% buffered formaldehyde. Lungs prepared for histology were inflated with 4% buffered formaldehyde immediately after removal to prevent collapse of lung tissue. Five pm-thick paraffin histological sections were cut and stained with haematoxylin and eosin. Serial sections were cut parallel to the sternum. Statistical analysis Multiple regression analysis of repeated measurements was used to compare the growth of footpad tumours. The Spearman rank correlation test was used to study the relation between data on primary tumours, spontaneous metastasis and experimental metastasis. The Mann-Whitney test was used to compare similar metastasis data from different cell clones. The Bonferroni correction was applied to control the experiment-wise error rate, and the significance level was set at 0.05.
RESULTS In vivo growth rate The growth curves of hind limb tumours of the cloned cell lines are shown in Fig. 1. Except for tumours of the C , and C, clones, the tumours differed markedly with respect t o both onset of growth and growth rate. Tumours of the C4cell clone showed sluggish growth the first week, whereafter the growth rate was similar to that
773
FALKVOLL & GRIMSTAD
"1
GROWTH OF HIND LIMB TUMOURS
,
I
0
5
10
15
20
I
25
30
DAYS AFTER INOCULATION
Fig. 1. Growth of transplanted footpad tumours. The increase in footpad diameter is plotted versus time after cell inoculation. The data are presented as median values.
of the C, and C3tumours. The C2 tumours had lower growth rates. The C5 tumours, whose presence was confirmed histologically, showed spontaneous regression beyond the second week after cell inoculation. Metastasis formation Metastases, originating either from hind limb tumours or after i.v. injection of cells, occurred almost exclusively in the lungs. Spontaneous metastasis formation was occasionally observed in regional draining lymph nodes in the tumour-
bearing hind limb and in lymph nodes in the lung hilus. After i.v. injection of the C, cell clone, colony formation in the suprarenal gland was observed in three of twelve mice. Histological examination showed that the spontaneous metastases varied more in size than did the lung colonies formed after i.v. injection of cells. The shape and distribution of pulmonary spontaneous metastases and colonies were nevertheless similar for individual cell clones. The total number of lung deposits, originating spontaneously from hind limb tumours or as colonies following i.v. injection, was estimated from the number of subpleural colonies and the fraction of subpleural colonies. Table 1 shows the number of spontaneous lung metastases and lung colonies, the fraction of subpleural colonies, and the cross-sectional area of subpleural and intrapulmonary colonies for the five cell clones. The cell clones differed greatly with respect to fonnation of both spontaneous and experimental lung metastases. The fraction of colonies located subpleurally differed significantly between the clones C, and C,, C2 and C3, as well as C, and C5. Although the cross-sectional area of some subpleural colonies was probably underestimated, the subpleural colonies were significantly larger than the intrapulmonary ones for the C,, C3, and C , clones. There was no close correlation between the size of intrapulmonary colonies 21 days after i.v. injection and the maximum
TABLE 1. Lung metastases formation by cell clones from a murine fibrosarcoma Cell Clone Number of Number of Fraction of CrossCrossspontaneous subpleural subpleural sectional area sectional area subpleural colonies colonies of subpleural of intrapulmetastases colonies monary colonies (mm2x (mm2x lo-? 23 0.5 31.1 19.1 c, 7 (4-9) (17-39) (0.4-0.6) (24.6-40.5) (1 1.9-30.6) 53 0.3 2.0 0.8 C* 0 (0-1 )" (45-66) (0.1-0.4) (0.5-5.9) (0.4-1.2) 105 0.7 9.5 2.9 c, 35.5 (1 3-53) (79-123) (0.6-0.8) (7.6-1 1.3) (1G4.3) 149 0.4 17.1 8.5 c 4 8 (6.8-12.3) ( 16.2-3 1.4) (1 15-190) (0.3-0.6) (3-26) C5 0 6.5 0.3 2.4 1.o (0.42.6) ( 1.6-1 4.1) (0.1-0.5) (69) (0- 1)2) Spontaneous lung metastases were formed from transplanted footpad tumours, and lung colonies (experimental metastases) were formed after i.v. injection of cells. The data are presented as median value with the lower and upper quartile in brackets. The minimum and maximum number of spontaneous metastases from the C2 and C5 clones were (0-2) and (0-5), respectively.
774
METASTASES
~
CLONAL DIFFERENCES
increase in the size of hind limb tumours at 21 days after inoculation. The cell clones forming the most slowly growing tumours, Cz, and the spontaneously regressive C5 tumours, however, had significantly smaller colonies than the other cell clones. The cross-sectional area of intrapulmonary colonies differed significantly between the CI and Cz, C, and C,, CI and C5,and C4 and C5clones. The size and shape of lung deposits differed between the cell clones. The subpleural metastatic tumours were mostly nodular for the CI, Cz and C5cell clones, but generally formed thin subpleural disc-shaped layers for the C3 and C4 cell clones. Interestingly, these two clones formed the largest number of lung metastases both from footpad tumours and following i.v. injection. The relationship between the colony diameter parallel to the pleural surface and the perpendicular diameter (or thickness of colonies growing as subpleural discs) is shown in Fig. 2. The intrapulmonary colonies were mostly nodular. Diffuse infiltration of alveolar septa, however, was occasionally observed in intrapulmonary colonies from the C2 and C3cell clones. The intrapulmonary colonies from the C, and C5 cell clones formed demarcated nodules and were
thus similar to the respective subpleural ones. The colonies from the CI, C3 and C4 cell clones were often located at the periphery of larger blood vessels and bronchial structures, and many colonies from the most metastatic cell clones, the C, and C4 clones, surrounded larger blood vessels. Several mitotic cells, apoptotic cells and cells apparently dying in mitosis were observed within the metastatic foci. Capillarylike vessels were observed within larger colonies. Apart from compression of surrounding lung tissue, destruction of normal tissue was not observed. In Fig. 3 the calculated total number of spontaneous lung metastases 42 days after cell inoculation into the footpad is plotted versus the maximum size of hind limb tumours (at 28 days after inoculation for the CI, C,, C, and C4 clones, and 14 days after inoculation for the C5 clone). When correlating the above data on total number of spontaneous lung metastases with maximum size of hind limb tumours for all the five clones seen together, the Spearman rank coefficient was 0.77. This indicated a strong correlation between the maximum hind limb tumour diameter and the number of spontaneous lung metastases. When the relationship between the primary data on the number of spontaneous subpleural metastases and the
SUBPLEURAL METASTASES (mm
Y
10.')
T
I
0
'
I
'
1
'
1
'
1
2 4 6 8 1 0 1 DIAMETER PARALLEL TO PLEURAL SURFACE
'
1
2
Fig. 2. This figure gives information on the shape
of subpleural experimental metastases. The diameter parallel to the pleural surface is plotted versus the perpendicular diameter. For metastases forming a subpleural disc, the diameter of the tumour disc parallel to the pleural surface and the maximum thickness of the tumour cell layers at right angles to the pleural surface were measured. Median +upper and lower quartile.
I
0
~
I
~
I
~
I
I
4 6 8 10 MAXIMUM INCREASE IN HIND LIMB DIAMETER (mm) 2
Fig. 3. The estimated total number of spontaneous lung metastases is plotted versus the maximum increase in the hind limb diameter (at 28 days after inoculation for the C,, C,, C , and C4 clones, and 14 days after inoculation for the Csclone). The data are presented as median k upper and lower quartile.
775
J
f
,
FALKVOLL & GRIMSTAD
maximum hind limb tumour diameter was studied for the individual cell clones, a positive correlation was found for all clones except the C2 cell clone (data not shown). DISCUSSION The growth pattern, including the size, shape and distribution of colonies in the lung, differed between the cell clones. The size of the primary tumours and the size of the lung colonies derived from the same cell clones were not closely correlated. Subcutaneously inoculated cells may depart from the injection site already within 24 h after inoculation (7, 11) probably because of migration into lymphatic or blood vessels. In the present study, variation in size of the spontaneously formed lung metastases, as compared to the colonies resulting from i.v. cell injection, suggested that cell release from footpad tumours occurred at different times. The rate of cell release from tumours into the circulation, however, is not easily determined. When cancer cells are injected i.v. to produce disseminated tumour colonies, this factor is defined, though there are some potential problems associated with i.v. cell injection as compared to spontaneous metastasis formation from primary tumours (27). Cells injected i.v. bypass any interactions with host cells and extracellular tissue components at the primary tumour site, and may invoke a different immunological response than cells spontaneously released from a tumour (6, 16). It has been reported for some cell lines that colony formation following i.v. injection is minimally influenced by previous exposure of the host to tumour cell antigens (26, 29); for other cell lines there has been reported good agreement between spontaneous metastasis formation from equally sized transplanted tumours and lung colony formation following i.v. injection of the same cells (4, 27). In the present study the correlation between the size of primary tumours and the number of metastases they form agrees with several previous studies involving animal and human tumours (21, 25). It is commonly assumed that the number of cells released into the circulation depends on tumour size and to a certain degree determines the number of metastases (13, 20, 776
30). The correlation between tumour size and metastasis formation, however, may not merely be determined by the number of cells released from the tumours (3, 8). The data in Table 1 on the formation of lung colonies indicate that other properties of the cell clones are also involved. Previous studies of the presently used clones and other murine fibrosarcoma cells have demonstrated considerable differences between strongly and weakly metastatic cells in the expression of cell surface adhesion molecules (9, 23). Grimstad & Prydz (10) showed that the release of thromboplastin into in vitro culture medium of the presently used cells correlated with spontaneous metastasis formation. The expression of surface adhesion molecules and release of thromboplastin from the murine fibrosarcoma cells may enhance their retention and growth in the lung microcirculation. Although the cell clones forming the smaller hind limb tumours also formed the smaller lung colonies, the size of the hind limb tumours was not closely correlated with the size of the lung colonies. It is interesting to note that the hind limb tumours which regressed spontaneously (from C5 cells) were capable of forming metastases spontaneously. Regressive growth of hind limb tumours occured even when 1 x lo6 C5cells were inoculated. After injection of more than 5 x lo6 C5cells, however, the footpad tumours grew progressively, but the number of lung metastases did not increase (LA. Grimstad, unpublished data). The lung colonies from the C5 cells decreased in size during the third week after cell injection (unpublished data). Thus, regressive growth of the C5cells occurred both in the hind limb when a limited number of cells were used for inoculation, and in the lungs following i.v. injection. This may be due to inherent properties of the cells, or the presence of growth inhibitory factors. The difference in fraction of subpleural lung deposits of these murine fibrosarcoma clones shows that the whole organ should be investigated when comparing the metastatic propensity of cell clones from a tumour. The distribution of metastases in the lungs, including the lung surface, may differ between different tumour systems (5, 12), but there is a tendency for preferential localization at the surface of metastases in the lungs and the liver (2, 18, 28).
METASTASES - CLONAL DIFFERENCES
The cell clones with the larger fraction of subpleural experimental metastases formed the largest metastases, and their intrapulmonary metastases tended to be associated with larger blood vessel and bronchial structures. The observation that subpleural metastases were larger than intrapulmonary metastases agrees with the finding of Orr et al. (18). Autoradiographic studies of fibrosarcoma cells forming unevenly distributed lung colonies have shown a uniform distribution of trapped cells in the lungs 24 h after i.v. injection (18). The differences in the distribution and size of metastases within the lungs may involve regional variations in the lungs with respect to sustained retention of arrested cells and/or conditions for tumour cell growth. Regional differences in blood flow, lymphatic drainage and mechanical properties of the lungs as well as in the composition of the extracellular matrix have all been suggested to contribute to this (2, 18). Differential tumour cell adhesion to microvessel walls and differential effects of tissue-derived growth stimulators or inhibitors may be involved in organ preferences of metastases (14, 15), and these factors may also be involved in regional metastatic preferences within organs. We wish to express our thanks to Professor Ole I? Clausen, Institute of Pathology, University of Oslo, for providing help with preparation of histological sections. We are grateful to Dr J. Eng, Department of Bacteriology, National Institute for Public Health, Oslo, for performing tests for mycoplasma and bacterial contamination of the cell cultures. Financial support was given by the Norwegian Cancer Society.
REFERENCES 1. Cavanaugh, l? G. & Nicolson. G. L.: Purification and some properties of a lung-derived growth factor that differentially stimulates the growth of tumor cells metastatic to the lung. Cancer Res. 49: 3928-33, 1989. 2. Dingemans, K. I?, Van Spronsen, R. & Thunnissen, E.: B16 metastases in mouse liver and lung. I. Localization. Invasion Metastasis 5: 50-60, 1985. 3. Fidler, I. J.: The relationship of embolic homogeneity, number, size and viability to the incidence of experimental metastasis. Eur. J. Cancer 9: 223-227, 1973. 4. Fidler, I. J.: Biological behavior of malignant melanoma cells correlated to their survival in vivo. Cancer Res 35: 218-224. 1975.
5. Fidler, I. J.: General considerations for studies of experimental cancer metastasis. In: Busch, H. (Ed.): Methods in Cancer Research. 15. Academic Press, New York 1978, pp. 399439. 6. Fidler, I. J., Gersten, D. M. & Riggs, C. W: Quantitative analysis of tumor:host interaction and the outcome of experimental metastasis. In: Day, S. B., Myers, W I? L., Stansly, l?, Garattini, S. & Lewis, M. G. (Eds.): Cancer invasion and metastasis: Biologic mechanisms and therapy. Raven Press, New York 1977, pp. 277-303. 7. Fisher, B. & Fisher, E.: Studies of metastatic mechanisms employing labeled tumor cells. The proliferation and spread of neoplastic cells. The Williams and Wilkins Company, Baltimore 1968, pp. 555-581. 8. Gorelik, E., Segal, S., Shapiro, J., Katzav, S., Ron, I:& Feldman, M.: Interactions between the local tumor and its metastases. Cancer Metastasis Rev. I: 83-94, 1982. 9. Grimstad, I. A. & Bosnes, l!: Cell-surface lamininlike molecules and alpha-D-galactopyranosyl end-groups of cloned strongly and weakly metastatic murine fibrosarcoma cells. Int. J. Cancer 40: 505-510, 1987. 10. Grimstad, I. A. & Prydz. H.: Thromboplastin release, but not content, correlates with spontaneous metastasis of cancer cells. Int. J. Cancer 41: 427431, 1988. 1 1. Hofer, K. G., Prensky, W & Hughes, W L.: Death and metastasis of tumor cells in mice monitored with 125-I-iododeoxyuridine. J. Natl. Cancer Inst. 43: 763-773, 1969. 12. Ketcham, A. S., Wexler, H. & Mantel, N.: The effect of removal of a “primary” tumor on the development of spontaneous metastases. I. Development of a standardized experimental technic. Cancer Res. 19: 940-944, 1959. 13. Liotta, L. A., Kleinerman, J. & Saidel, G. M.: Stochastic model of metastases formation. Biometrics 32: 535-550, 1976. 14. Naito, S., Giavazzi, R. & Fidler, I. J.: Correlation between the in vitro interaction of tumor cells with an organ environment and metastatic behaviour in vivo. Invasion Metastasis 7: 16-29, 1987. 15. Nicolson, G. L.: Organ specificity of tumor metastasis: role of preferential adhesion, invasion and growth of malignant cells at specific secondary sites. Cancer Metastasis Rev. 7: 143-88, 1988. 16. Nicolson, G. L. & Custead, S. E.: Tumor metastasis is not due to adaption of cells to a new organ environment. Science 215: 176-178, 1982. 17. Nicolson, G. L. & Poste, D. l! M.: Tumor cell diversity and host responses in cancer metastasis - part I. Curr. Probl. Cancer VII: 1-83, 1982. 18. Orr, F W, Young, L., King, G. M. & Adamson, I. Y : Preferential growth of metastatic tumors at the pleural surface of mouse lung. Clin. Exp. Metastasis 6: 221-232, 1988.
777
FALKVOLL & GRIMSTAD
19. Schackert, G. & Fidler, I. J.: Site-specific metastasis of mouse melanomas and a fibrosarcoma in the brain or meninges of syngeneic animals. Cancer Res. 48: 3478-84, 1988. 20. Sugarbaker, E. V: Cancer metastasis: A product of tumor-host interactions. Curr. Probl. Cancer 3: 1-59, 1979. 21. Sugarbaker, E. Y & Cohen, A. M.: Altered antigenicity in spontaneous pulmonary metastases from an antigenic murine sarcoma. Surgery 72; 155-161, 1972. 22. Varani, J., Fligiel, S. E. G. & Perone, I!: Directional motility in strongly malignant murine tumor cells. Int. J. Cancer 35: 559-564, 1985. 23. Virani, J., Grimstad, I. A., Knibbs, R. N., Hovig, ?: & McCoy, J. I!: Attachment, spreading and growth in vitro of highly malignant and low malignant murine fibrosarcoma cells. Clin. Exp. Metastases 3: 45-59, 1985. 24. Varani, J., Lovett, E. J. III, Wicha, M . , Malinoff; H. & McCoy, J. I! Jr: Cell surface a1pha-Dgalactopyranosyl end groups: use as markers in the isolation of murine tumor cell lines with different cancer-causing potentials. J. Natl. Cancer Inst. 71: 1281-1287. 1983.
778
25. Ware, J. L. & DeLong, E. R.: Influence of tumor size on human prostate tumor metastasis in athymic nude mice. Br. J. Cancer 51: 419423, 1985. 26. Weiss, L. & Glaves, D.: The immunospecificity of altered initial arrest patterns of circulating cancer cells in tumour-bearing mice. Int. J. Cancer 18: 774777, 1976. 27. Welch, D.R., Neri, A. & Nicolson, G. L.: Comparison of ‘spontaneous’ and ‘experimental’ metastasis using rat 13762 mammary adenocarcinoma metastatic cell clones. Invasion Metastasis 3: 65-80, 1983. 28. Wexler, H.: Accurate identification of experimental pulmonary metastases. J. Natl. Cancer Inst. 36: 641-645, 1966. 29. Wexler, H., Chretien, I! B. & Ketcham, A . S.: Effect of tumor immunity on production of lung tumors after intravenous inoculation of antigenically identical tumor cells. J. Natl. Cancer Inst. 48: 657-663, 1972. 30. Wexler, H., Ryan, J. J. & Ketcham, A. S.: The study of circulating tumor cells by the formation of pulmonary embolic tumor growths in a secondary host. Cancer 23: 946-51, 1969.
Growth patterns of pulmonary metastases and primary tumours from five murine fibrosarcoma cell clones KNUT H. FALKVOLL' and IVAR AMUND GRIMSTAD* 'Institute for Surgical Research and 'Institute of Clinical Biochemistry and Institute of Pathology, University of Oslo, Rikshospitalet, Oslo, Norway
Falkvoll, K. H. & Grimstad, I. A. Growth patterns of pulmonary metastases and primary tumours from five murine fibrosarcoma cell clones. APMIS 100: 772-778, 1992. The growth patterns, including the size, shape and regional preferences, of lung metastases from five rnurine fibrosarcoma cell clones were studied. Spontaneous metastases developed from tumours formed by subcutaneous inoculation of the cell clones. Lung colonies (experimental metastases) were established by i.v. injection of cells. The numbers of both spontaneously and experimentally formed subpleural lung metastases were counted through a stereomicroscope. The fraction of colonies that was located subpleurally was determined in histological sections of lungs. The growth kinetics of clonally derived primary tumours, and the number of spontaneous and experimental lung metastases, differed greatly between certain cell clones. The number of spontaneous lung metastases was correlated with the maximum size of primary tumours. No close correlation was observed between the size of the primary tumours and the size of experimental metastases. There were differences between the cell clones in the shape and regional preferences of their lung deposits. The subpleural colonies were generally larger than the intraphlmonary ones. Thus, both the regional distribution and the growth pattern of lung deposits differed between the clones. Key words: Growth pattern; primary tumours; pulmonary metastases; murine fibrosarcomas.
K. H. Falkvoll, Institute for Surgical Research, Rikshospitalet, Langesgate 1, N-0027 Oslo 1, Norway.
Cancer metastasis is a non-random process. The sites of distant dissemination of several cancers cannot be explained by the distribution of blood flow, but may be determined by specific interactions between tumour cells and the target organs ( 17). Such specificity may reside in differential interactions of tumour cells with organ-derived microvessel endothelial cells and subendothelial matrix ( I 5). Differential responses of various tumour cell populations to tissue-mediated growth stimulators and inhibitors have also been demonstrated to be involved in the determination of organ preferences of metastases
Received December 27, 1991. Accepted April 4, 1992.
772
from some tumours (14, 15). Metastasis formation may, in addition, show regional preferences within organs (2, 5, 18, 19). Identification of tumour cell and host properties accounting for preferential location of metastases in particular organs and in certain regions within organs may shed light on what enables some neoplastic cells, but not others, to form metastases (1, 2). The aim of the present investigation was to study the growth pattern, including the size, shape and regional distribution in the lungs, of metastases from five different cell clones derived from a murine fibrosarcoma. In order to relate these data to the metastatic propensity of the clones, the formation of lung metastases from transplanted footpad tumours, as well as the formation of lung colonies following i.v. injec-
METASTASES
-
CIBNAL DIFFERENCES
tion of cells, was studied. Factors are discussed which may be involved in preferential metastasis formation in particular organs and certain regions within these organs.
MATERIALS AND METHODS Tumour cell clones The tumour cell clones originate from a fibrosarcoma induced subcutaneously by methylcholanthrene in C57 BL/6J mice. (24) The cloning by microscopically controlled limiting dilution, and initial characterization have been described elsewhere (9, 22). The cells were subcultured at least twice before use, but were not kept in culture for more than five weeks in order to prevent clonal diversification. The culture medium was RPMI 1640 supplemented with L-glutamine (292 mg/l), 10% foetal bovine serum, penicillin (50 units/ ml), streptomycin (50 pg/ml) and amphotericin B (125 ng/ml). The cells were kept at 37'C in a humidified atmosphere of air supplemented with 5% CO,. The five cell clones used in this study were labelled C,, Cz. C1,C4and C,, which correspond to the respective designations l.l/clone 3, l.l/clone 38, 1.1/L1 clone 2, l .l / L 2 clone 3 and l.l/anti-B' clone 3 used in previous reports (9, 10). Primary tumour growth Subconfluent rapidly growing cultured cells were harvested by brief trypsinization, washed, and 2 x lo5 cells, dispersed in 50 p1 serum-free RPMI 1640 medium, were inoculated into the right hind footpads of female syngeneic (C57 BL/6J) mice (Bommice, Ry, Denmark) weighing 18-20 g. The tumour growth rate was determined at regular intervals by measuring the dorsoplantar diameter of the hind limb using accurate calipers, with the opposite limb serving as control. Spontaneous metastasis Twenty-eight days after cell inoculation into the right hind footpad of the mice, 15-16 animals per clone, the legs were amputated proximal to the tumour (9). The limbs were amputated because further growth of some tumours would give ulceration of the overlying skin. When the mice were killed 14 days later, the lungs were removed, inflated with a suspension of India ink and placed in a bleaching solution as described by Wexler (28). This preparation greatly increased the visibility of surface lung metastases counted through a stereomicroscope. Randomly selected lungs were studied histologically. The presence (or absence) of metastases in other organs and the presence of primary tumours were verified histologically. Experimental metastasis - lung colony formation Lung colonies were established by i.v. injection of cells in single cell suspensions. Single cell suspensions
were obtained by allowing sedimentation of cell clumps among the trypsinized washed cells in tubes on ice for about 10 min (3). A cooled 1.0 ml syringe supplied with a 27-gauge needle was used for i.v. injection of 2 x lo5 single cells into the lateral tail vein. The trypan blue exclusion test confirmed that the viability of injected cells was always greater than 90%. After 21 days the lungs from nine to 15 mice (for each clone) were removed and prepared as described above, and the surface metastases were counted. The distribution of metastases was studied in histological sections of the lungs (five to eight mice in each group), and the numbers of subpleural and intrapulmonary metastases were scored. Two perpendicular diameters, for subpleural metastases the diameter parallel to the surface and the diameter at right angles to the pleural surface, of metastatic structures were measured on video screen projections of the sections. The cross-sectional area of metastatic structures was calculated as being elliptical. This adequately determined the cross-sectional area of nodular metastases but might have underestimated the cross-sectional area of metastases forming thin subpleural discs.
Histological preparation Lungs, hind limb tumours and other organs were fixed in 4% buffered formaldehyde. Lungs prepared for histology were inflated with 4% buffered formaldehyde immediately after removal to prevent collapse of lung tissue. Five pm-thick paraffin histological sections were cut and stained with haematoxylin and eosin. Serial sections were cut parallel to the sternum. Statistical analysis Multiple regression analysis of repeated measurements was used to compare the growth of footpad tumours. The Spearman rank correlation test was used to study the relation between data on primary tumours, spontaneous metastasis and experimental metastasis. The Mann-Whitney test was used to compare similar metastasis data from different cell clones. The Bonferroni correction was applied to control the experiment-wise error rate, and the significance level was set at 0.05.
RESULTS In vivo growth rate The growth curves of hind limb tumours of the cloned cell lines are shown in Fig. 1. Except for tumours of the C , and C, clones, the tumours differed markedly with respect t o both onset of growth and growth rate. Tumours of the C4cell clone showed sluggish growth the first week, whereafter the growth rate was similar to that
773
FALKVOLL & GRIMSTAD
"1
GROWTH OF HIND LIMB TUMOURS
,
I
0
5
10
15
20
I
25
30
DAYS AFTER INOCULATION
Fig. 1. Growth of transplanted footpad tumours. The increase in footpad diameter is plotted versus time after cell inoculation. The data are presented as median values.
of the C, and C3tumours. The C2 tumours had lower growth rates. The C5 tumours, whose presence was confirmed histologically, showed spontaneous regression beyond the second week after cell inoculation. Metastasis formation Metastases, originating either from hind limb tumours or after i.v. injection of cells, occurred almost exclusively in the lungs. Spontaneous metastasis formation was occasionally observed in regional draining lymph nodes in the tumour-
bearing hind limb and in lymph nodes in the lung hilus. After i.v. injection of the C, cell clone, colony formation in the suprarenal gland was observed in three of twelve mice. Histological examination showed that the spontaneous metastases varied more in size than did the lung colonies formed after i.v. injection of cells. The shape and distribution of pulmonary spontaneous metastases and colonies were nevertheless similar for individual cell clones. The total number of lung deposits, originating spontaneously from hind limb tumours or as colonies following i.v. injection, was estimated from the number of subpleural colonies and the fraction of subpleural colonies. Table 1 shows the number of spontaneous lung metastases and lung colonies, the fraction of subpleural colonies, and the cross-sectional area of subpleural and intrapulmonary colonies for the five cell clones. The cell clones differed greatly with respect to fonnation of both spontaneous and experimental lung metastases. The fraction of colonies located subpleurally differed significantly between the clones C, and C,, C2 and C3, as well as C, and C5. Although the cross-sectional area of some subpleural colonies was probably underestimated, the subpleural colonies were significantly larger than the intrapulmonary ones for the C,, C3, and C , clones. There was no close correlation between the size of intrapulmonary colonies 21 days after i.v. injection and the maximum
TABLE 1. Lung metastases formation by cell clones from a murine fibrosarcoma Cell Clone Number of Number of Fraction of CrossCrossspontaneous subpleural subpleural sectional area sectional area subpleural colonies colonies of subpleural of intrapulmetastases colonies monary colonies (mm2x (mm2x lo-? 23 0.5 31.1 19.1 c, 7 (4-9) (17-39) (0.4-0.6) (24.6-40.5) (1 1.9-30.6) 53 0.3 2.0 0.8 C* 0 (0-1 )" (45-66) (0.1-0.4) (0.5-5.9) (0.4-1.2) 105 0.7 9.5 2.9 c, 35.5 (1 3-53) (79-123) (0.6-0.8) (7.6-1 1.3) (1G4.3) 149 0.4 17.1 8.5 c 4 8 (6.8-12.3) ( 16.2-3 1.4) (1 15-190) (0.3-0.6) (3-26) C5 0 6.5 0.3 2.4 1.o (0.42.6) ( 1.6-1 4.1) (0.1-0.5) (69) (0- 1)2) Spontaneous lung metastases were formed from transplanted footpad tumours, and lung colonies (experimental metastases) were formed after i.v. injection of cells. The data are presented as median value with the lower and upper quartile in brackets. The minimum and maximum number of spontaneous metastases from the C2 and C5 clones were (0-2) and (0-5), respectively.
774
METASTASES
~
CLONAL DIFFERENCES
increase in the size of hind limb tumours at 21 days after inoculation. The cell clones forming the most slowly growing tumours, Cz, and the spontaneously regressive C5 tumours, however, had significantly smaller colonies than the other cell clones. The cross-sectional area of intrapulmonary colonies differed significantly between the CI and Cz, C, and C,, CI and C5,and C4 and C5clones. The size and shape of lung deposits differed between the cell clones. The subpleural metastatic tumours were mostly nodular for the CI, Cz and C5cell clones, but generally formed thin subpleural disc-shaped layers for the C3 and C4 cell clones. Interestingly, these two clones formed the largest number of lung metastases both from footpad tumours and following i.v. injection. The relationship between the colony diameter parallel to the pleural surface and the perpendicular diameter (or thickness of colonies growing as subpleural discs) is shown in Fig. 2. The intrapulmonary colonies were mostly nodular. Diffuse infiltration of alveolar septa, however, was occasionally observed in intrapulmonary colonies from the C2 and C3cell clones. The intrapulmonary colonies from the C, and C5 cell clones formed demarcated nodules and were
thus similar to the respective subpleural ones. The colonies from the CI, C3 and C4 cell clones were often located at the periphery of larger blood vessels and bronchial structures, and many colonies from the most metastatic cell clones, the C, and C4 clones, surrounded larger blood vessels. Several mitotic cells, apoptotic cells and cells apparently dying in mitosis were observed within the metastatic foci. Capillarylike vessels were observed within larger colonies. Apart from compression of surrounding lung tissue, destruction of normal tissue was not observed. In Fig. 3 the calculated total number of spontaneous lung metastases 42 days after cell inoculation into the footpad is plotted versus the maximum size of hind limb tumours (at 28 days after inoculation for the CI, C,, C, and C4 clones, and 14 days after inoculation for the C5 clone). When correlating the above data on total number of spontaneous lung metastases with maximum size of hind limb tumours for all the five clones seen together, the Spearman rank coefficient was 0.77. This indicated a strong correlation between the maximum hind limb tumour diameter and the number of spontaneous lung metastases. When the relationship between the primary data on the number of spontaneous subpleural metastases and the
SUBPLEURAL METASTASES (mm
Y
10.')
T
I
0
'
I
'
1
'
1
'
1
2 4 6 8 1 0 1 DIAMETER PARALLEL TO PLEURAL SURFACE
'
1
2
Fig. 2. This figure gives information on the shape
of subpleural experimental metastases. The diameter parallel to the pleural surface is plotted versus the perpendicular diameter. For metastases forming a subpleural disc, the diameter of the tumour disc parallel to the pleural surface and the maximum thickness of the tumour cell layers at right angles to the pleural surface were measured. Median +upper and lower quartile.
I
0
~
I
~
I
~
I
I
4 6 8 10 MAXIMUM INCREASE IN HIND LIMB DIAMETER (mm) 2
Fig. 3. The estimated total number of spontaneous lung metastases is plotted versus the maximum increase in the hind limb diameter (at 28 days after inoculation for the C,, C,, C , and C4 clones, and 14 days after inoculation for the Csclone). The data are presented as median k upper and lower quartile.
775
J
f
,
FALKVOLL & GRIMSTAD
maximum hind limb tumour diameter was studied for the individual cell clones, a positive correlation was found for all clones except the C2 cell clone (data not shown). DISCUSSION The growth pattern, including the size, shape and distribution of colonies in the lung, differed between the cell clones. The size of the primary tumours and the size of the lung colonies derived from the same cell clones were not closely correlated. Subcutaneously inoculated cells may depart from the injection site already within 24 h after inoculation (7, 11) probably because of migration into lymphatic or blood vessels. In the present study, variation in size of the spontaneously formed lung metastases, as compared to the colonies resulting from i.v. cell injection, suggested that cell release from footpad tumours occurred at different times. The rate of cell release from tumours into the circulation, however, is not easily determined. When cancer cells are injected i.v. to produce disseminated tumour colonies, this factor is defined, though there are some potential problems associated with i.v. cell injection as compared to spontaneous metastasis formation from primary tumours (27). Cells injected i.v. bypass any interactions with host cells and extracellular tissue components at the primary tumour site, and may invoke a different immunological response than cells spontaneously released from a tumour (6, 16). It has been reported for some cell lines that colony formation following i.v. injection is minimally influenced by previous exposure of the host to tumour cell antigens (26, 29); for other cell lines there has been reported good agreement between spontaneous metastasis formation from equally sized transplanted tumours and lung colony formation following i.v. injection of the same cells (4, 27). In the present study the correlation between the size of primary tumours and the number of metastases they form agrees with several previous studies involving animal and human tumours (21, 25). It is commonly assumed that the number of cells released into the circulation depends on tumour size and to a certain degree determines the number of metastases (13, 20, 776
30). The correlation between tumour size and metastasis formation, however, may not merely be determined by the number of cells released from the tumours (3, 8). The data in Table 1 on the formation of lung colonies indicate that other properties of the cell clones are also involved. Previous studies of the presently used clones and other murine fibrosarcoma cells have demonstrated considerable differences between strongly and weakly metastatic cells in the expression of cell surface adhesion molecules (9, 23). Grimstad & Prydz (10) showed that the release of thromboplastin into in vitro culture medium of the presently used cells correlated with spontaneous metastasis formation. The expression of surface adhesion molecules and release of thromboplastin from the murine fibrosarcoma cells may enhance their retention and growth in the lung microcirculation. Although the cell clones forming the smaller hind limb tumours also formed the smaller lung colonies, the size of the hind limb tumours was not closely correlated with the size of the lung colonies. It is interesting to note that the hind limb tumours which regressed spontaneously (from C5 cells) were capable of forming metastases spontaneously. Regressive growth of hind limb tumours occured even when 1 x lo6 C5cells were inoculated. After injection of more than 5 x lo6 C5cells, however, the footpad tumours grew progressively, but the number of lung metastases did not increase (LA. Grimstad, unpublished data). The lung colonies from the C5 cells decreased in size during the third week after cell injection (unpublished data). Thus, regressive growth of the C5cells occurred both in the hind limb when a limited number of cells were used for inoculation, and in the lungs following i.v. injection. This may be due to inherent properties of the cells, or the presence of growth inhibitory factors. The difference in fraction of subpleural lung deposits of these murine fibrosarcoma clones shows that the whole organ should be investigated when comparing the metastatic propensity of cell clones from a tumour. The distribution of metastases in the lungs, including the lung surface, may differ between different tumour systems (5, 12), but there is a tendency for preferential localization at the surface of metastases in the lungs and the liver (2, 18, 28).
METASTASES - CLONAL DIFFERENCES
The cell clones with the larger fraction of subpleural experimental metastases formed the largest metastases, and their intrapulmonary metastases tended to be associated with larger blood vessel and bronchial structures. The observation that subpleural metastases were larger than intrapulmonary metastases agrees with the finding of Orr et al. (18). Autoradiographic studies of fibrosarcoma cells forming unevenly distributed lung colonies have shown a uniform distribution of trapped cells in the lungs 24 h after i.v. injection (18). The differences in the distribution and size of metastases within the lungs may involve regional variations in the lungs with respect to sustained retention of arrested cells and/or conditions for tumour cell growth. Regional differences in blood flow, lymphatic drainage and mechanical properties of the lungs as well as in the composition of the extracellular matrix have all been suggested to contribute to this (2, 18). Differential tumour cell adhesion to microvessel walls and differential effects of tissue-derived growth stimulators or inhibitors may be involved in organ preferences of metastases (14, 15), and these factors may also be involved in regional metastatic preferences within organs. We wish to express our thanks to Professor Ole I? Clausen, Institute of Pathology, University of Oslo, for providing help with preparation of histological sections. We are grateful to Dr J. Eng, Department of Bacteriology, National Institute for Public Health, Oslo, for performing tests for mycoplasma and bacterial contamination of the cell cultures. Financial support was given by the Norwegian Cancer Society.
REFERENCES 1. Cavanaugh, l? G. & Nicolson. G. L.: Purification and some properties of a lung-derived growth factor that differentially stimulates the growth of tumor cells metastatic to the lung. Cancer Res. 49: 3928-33, 1989. 2. Dingemans, K. I?, Van Spronsen, R. & Thunnissen, E.: B16 metastases in mouse liver and lung. I. Localization. Invasion Metastasis 5: 50-60, 1985. 3. Fidler, I. J.: The relationship of embolic homogeneity, number, size and viability to the incidence of experimental metastasis. Eur. J. Cancer 9: 223-227, 1973. 4. Fidler, I. J.: Biological behavior of malignant melanoma cells correlated to their survival in vivo. Cancer Res 35: 218-224. 1975.
5. Fidler, I. J.: General considerations for studies of experimental cancer metastasis. In: Busch, H. (Ed.): Methods in Cancer Research. 15. Academic Press, New York 1978, pp. 399439. 6. Fidler, I. J., Gersten, D. M. & Riggs, C. W: Quantitative analysis of tumor:host interaction and the outcome of experimental metastasis. In: Day, S. B., Myers, W I? L., Stansly, l?, Garattini, S. & Lewis, M. G. (Eds.): Cancer invasion and metastasis: Biologic mechanisms and therapy. Raven Press, New York 1977, pp. 277-303. 7. Fisher, B. & Fisher, E.: Studies of metastatic mechanisms employing labeled tumor cells. The proliferation and spread of neoplastic cells. The Williams and Wilkins Company, Baltimore 1968, pp. 555-581. 8. Gorelik, E., Segal, S., Shapiro, J., Katzav, S., Ron, I:& Feldman, M.: Interactions between the local tumor and its metastases. Cancer Metastasis Rev. I: 83-94, 1982. 9. Grimstad, I. A. & Bosnes, l!: Cell-surface lamininlike molecules and alpha-D-galactopyranosyl end-groups of cloned strongly and weakly metastatic murine fibrosarcoma cells. Int. J. Cancer 40: 505-510, 1987. 10. Grimstad, I. A. & Prydz. H.: Thromboplastin release, but not content, correlates with spontaneous metastasis of cancer cells. Int. J. Cancer 41: 427431, 1988. 1 1. Hofer, K. G., Prensky, W & Hughes, W L.: Death and metastasis of tumor cells in mice monitored with 125-I-iododeoxyuridine. J. Natl. Cancer Inst. 43: 763-773, 1969. 12. Ketcham, A. S., Wexler, H. & Mantel, N.: The effect of removal of a “primary” tumor on the development of spontaneous metastases. I. Development of a standardized experimental technic. Cancer Res. 19: 940-944, 1959. 13. Liotta, L. A., Kleinerman, J. & Saidel, G. M.: Stochastic model of metastases formation. Biometrics 32: 535-550, 1976. 14. Naito, S., Giavazzi, R. & Fidler, I. J.: Correlation between the in vitro interaction of tumor cells with an organ environment and metastatic behaviour in vivo. Invasion Metastasis 7: 16-29, 1987. 15. Nicolson, G. L.: Organ specificity of tumor metastasis: role of preferential adhesion, invasion and growth of malignant cells at specific secondary sites. Cancer Metastasis Rev. 7: 143-88, 1988. 16. Nicolson, G. L. & Custead, S. E.: Tumor metastasis is not due to adaption of cells to a new organ environment. Science 215: 176-178, 1982. 17. Nicolson, G. L. & Poste, D. l! M.: Tumor cell diversity and host responses in cancer metastasis - part I. Curr. Probl. Cancer VII: 1-83, 1982. 18. Orr, F W, Young, L., King, G. M. & Adamson, I. Y : Preferential growth of metastatic tumors at the pleural surface of mouse lung. Clin. Exp. Metastasis 6: 221-232, 1988.
777
FALKVOLL & GRIMSTAD
19. Schackert, G. & Fidler, I. J.: Site-specific metastasis of mouse melanomas and a fibrosarcoma in the brain or meninges of syngeneic animals. Cancer Res. 48: 3478-84, 1988. 20. Sugarbaker, E. V: Cancer metastasis: A product of tumor-host interactions. Curr. Probl. Cancer 3: 1-59, 1979. 21. Sugarbaker, E. Y & Cohen, A. M.: Altered antigenicity in spontaneous pulmonary metastases from an antigenic murine sarcoma. Surgery 72; 155-161, 1972. 22. Varani, J., Fligiel, S. E. G. & Perone, I!: Directional motility in strongly malignant murine tumor cells. Int. J. Cancer 35: 559-564, 1985. 23. Virani, J., Grimstad, I. A., Knibbs, R. N., Hovig, ?: & McCoy, J. I!: Attachment, spreading and growth in vitro of highly malignant and low malignant murine fibrosarcoma cells. Clin. Exp. Metastases 3: 45-59, 1985. 24. Varani, J., Lovett, E. J. III, Wicha, M . , Malinoff; H. & McCoy, J. I! Jr: Cell surface a1pha-Dgalactopyranosyl end groups: use as markers in the isolation of murine tumor cell lines with different cancer-causing potentials. J. Natl. Cancer Inst. 71: 1281-1287. 1983.
778
25. Ware, J. L. & DeLong, E. R.: Influence of tumor size on human prostate tumor metastasis in athymic nude mice. Br. J. Cancer 51: 419423, 1985. 26. Weiss, L. & Glaves, D.: The immunospecificity of altered initial arrest patterns of circulating cancer cells in tumour-bearing mice. Int. J. Cancer 18: 774777, 1976. 27. Welch, D.R., Neri, A. & Nicolson, G. L.: Comparison of ‘spontaneous’ and ‘experimental’ metastasis using rat 13762 mammary adenocarcinoma metastatic cell clones. Invasion Metastasis 3: 65-80, 1983. 28. Wexler, H.: Accurate identification of experimental pulmonary metastases. J. Natl. Cancer Inst. 36: 641-645, 1966. 29. Wexler, H., Chretien, I! B. & Ketcham, A . S.: Effect of tumor immunity on production of lung tumors after intravenous inoculation of antigenically identical tumor cells. J. Natl. Cancer Inst. 48: 657-663, 1972. 30. Wexler, H., Ryan, J. J. & Ketcham, A. S.: The study of circulating tumor cells by the formation of pulmonary embolic tumor growths in a secondary host. Cancer 23: 946-51, 1969.
