- Email: [email protected]
Serum levels of macrophage colony-stimulating factor in patients with ovarian cancer undergoing second-look laparotomy
Serum levels of macrophage colony-stimulating factor in patients with ovarian cancer undergoing second-look laparotomy Steven A. Elg, MD,.,b Yin Yu, MD: Linda F. Carson, MD: Leon L. Adcock, MD: Leo B. Twiggs, MD: Konald A. Prem, MD,. and Sundaram Ramakrishnan, PhD b Minneapolis, Minnesota OBJECTIVE: The purpose of this study was to evaluate the prognostic significance of macrophage colony-stimulating factor serum levels in patients with ovarian cancer undergoing second-look laparotomy. STUDY DESIGN: The presurgical serum levels of macrophage colony-stimulating factor from 33 consecutive patients with ovarian cancer undergoing second-look laparotomy were determined and compared with those of 50 healthy control subjects. Mean differences in groups were evaluated with the Student t test. RESULTS: Sixteen of 33 patients had a positive result at second look and a mean serum macrophage colony-stimulating factor level of 2.31 ± 1.45 ng/ml. Seventeen of 33 patients had a negative result at second look and a mean macrophage colony-stimulating factor level of 1.90 ± 0.86 ng / ml (p > 0.05). The mean macrophage colony-stimulating factor level in the control group was 1.20 ± 0 5. 1 ng/ml. This was statistically lower than the mean level found in patients with a negative second-look result (p < 0.05). CONCLUSION: Regardless of tumor status, serum macrophage colony-stimulating factor levels tend to be elevated at the time of second-look laparotomy. (AM J OBSTET GVNECOL 1992;166:134-7.)
Key words: Ovarian cancer, tumor marker, macrophage colony-stimulating factor, second laparotomy Tumor cells produce growth factors, some of which may regulate the proliferation of cancer cells by autocrine and paracrine pathways.'·7 The aberrant production of trophic factors by cancer cells is due to either gene amplification or genetic translocation, which may bring the growth factor genes under the control of different regulatory elements. Interestingly, among the different growth factors studied thus far, colony-stimulating activity has been detected in many nonhematologic tumors. 6 • 7 Concomitant with the presence of colony-stimulating factor activity, higher levels of circulating leukocytes have been seen in some of these patients." 5 These results, along with studies using murine experimental model systems wherein human tumors were transplanted, indicate constitutive production of cytokines by malignant cells. 7 Although the importance of these factors remains uncertain at the present time, their levels in serum may prove useful as From the Women's Cancer Center' and the Department of Pharmacology,' Umverslty of Minnesota Health Science Center. Supported in part by a grant from the National Cancer Institute (CA48608) and a grant from the Minnesota MedIcal Foundatwn (SMF-
637-89).
Received for publication October 17, 1990; revised April 23, 1991; accepted May 28, 1991. Reprint requests: Steven A. Elg, MD, UMHC Box 395, 420 Delaware St. SE , Minneapolis. MN 55455. 6/1131296 134
Table I. Patient age Group
No.
Control Second-look laparotomy Positive Negative
50
Age (yr, mean ± SD)
33
16 17
41 52 55 49
± ± ± ±
11 14 12 16
Table II. Initial diagnosis of patients undergoing second-look laparotomy Initial diagnosis
No.
Adenocarcinoma (serous or mucinous) Malignant endometrioid tumor Clear cell carcinoma
28 1 4
TOTAL
33
markers to monitor primary tumors or to detect metastases. In a recent study it was reported that each of six ovarian cancer cell lines secrete macrophage colonystimulating factor into the culture medium in the absence of any external stimuli. 8 Whereas lymphoid and myeloid cell lines failed to show any detectable amount
Macrophage CSF serum levels at second look in ovanan cancer
Volume 166 !';umber 1. Part I
135
7
-
6
•
E
m c:
5
•
:E
:::l
a: w
•
4
C/)
•
z C/)
...J W
• •• •••
3
...
>
W ...J
u..
2
C/)
~ ••
880
....•
8
•
08~~~800
()
::E
06 g 0
Fig. 1. Macrophage CSF levels in sera of normal control subjects (open mc/es) and patients with ovarian cancer before second-look laparotomy (solzd czrcles) , Sera samples were analyzed III specific radioimmunoassay. Each value is mean of triplicate independent estimations from individual samples.
of macrophage CSF, a limited number of breast canCer cell lines produced the growth factor in significant quantities. Presence of the colony-stimulating factor was confirmed not only by immunologic (radioimmunoassay) and biologic assay (murine bone marrow cultures) but also by analysis of total cellular ribonucleic acid for macrophage CSF -related transcripts." Kacinski et al. 9 have detected expression of macrophage CSF transcripts in ovarian carcinoma cells by in situ hybridization. Macrophage CSF is a dimeric glycoprotein of about 70 kd and belongs to a group of cytokines that regulate the ontogeny of blood cells. Macrophage CSF selectively stimulates proliferation and differentiation of the mononuclear phagocytic lineage and mature macrophages. 1O II Recently, macrophage CSF serum levels were reported to be elevated in patients with epithelial ovarian cancer." It has been suggested that macrophage CSF may be useful as a potential tumor marker. The purpose of this study is to evaluate the prognostic significance of macrophage CSF levels in ovarian cancer patients undergoing second-look laparotomy. Material and methods
Serum samples from 33 consecutive patients undergoing second-look laparotomy for epithelial ovarian cancer were retrieved from the serum bank of the U niversity of Minnesota Women's Cancer Center. The sera, which had been drawn just before surgery, were ob-
tained between Dec. I, 1985, and June 1, 1988, and stored at - 20° C until analysis. Macrophage CSF levels in these samples were determined in the following way. A polyclonal antiserum against human recombinant macrophage CSF (provided by the Genetics Institute, Cambridge, Mass.) was generated in rabbits by hyperimmunization. The antiserum used in these studies has been well characterized and does not cross react with other growth factors or cytokines. Details of the assay conditions are published elsewhere." This assay system can detect concentrations as low as 500 pg/ml. To minimize interassay variations, aliquots of individual sera samples from five normal subjects were used as internal controls. Macrophage CSF serum levels were determined in a similar manner in 50 healthy female controls. All samples were also assayed for the presence of tumor marker CA 125 with an immunoradiometric method. 12 CA 125 levels were expressed in relative units per milliliter. Mean differences in groups were evaluated with the Student t teSt. A p value <0.05 was considered significant. Results
No clinical evidence of infection was noted at the time of obtaining the serum. Mean ages of the study groups are given in Table I. A summary of initial pathologic diagnoses is shown in Table II. Macrophage CSF levels as determined by radioimmunoassay showed a wide
136
Eig et al.
January 1992
Am
J Obstet G~necol
7
6
E
e.,
5
r
c::
~
:E ::>
c: w en
z en
4
6 [J
3
:~
...J
w > w
2
...J
u..
en
'-?
:E
1
r
I:b dJ
0
Fig. 2. Macrophage CSF levels in sera of patients with ovarian cancer. Open squares, Abnormal pathologic findings at second-look laparotomy; open triangles, normal pathologic findings at secondlook laparotomy. range between nondetectable levels and 5.3 ng/ml in one patient. The distribution of individual samples is shown in scatter plots (Figs. 1 and 2). Sixteen patients with positive pathologic studies at the time of second look had a mean serum macrophage CSF level of 2.31 ± 1.45 ngl ml. The mean level in the 17 patients with negative findings was 1.90 ± 0.86 ng/ml. The difference in the~e levels is not statistically significant
(P>
0.05).
The mean macrophage CSF level in the control group was 1.20 ± 0.51 ng I ml. This is statistically lower than that found in patients with a negative second look (P < 0.005). Subsequently. macrophge CSF levels were compared with those of another tumor marker. CA 125. In all but five of the 33 patients, CA 125 levels were negative «35 U/ml). Two of these five had a macrophage CSF level <3.62 ng/ml (mean ± SD 1.90 + 2) and a negative second look, one had a positive second look and a macrophage CSF level >3.62 ng/ml, and two had a positive second look and a macrophage CSF level <3.62 ng/ml (0.98 and 3.50 ng/ml). The variations seen in the levels of macrophage CSF might be attributed to tumor load. However, in the current study the sample size was too small to evaluate a possible correlation at second-look laparotomy.
Comment Under normal conditions, macrophage CSF is produced constitutively by a variety of cells, including fibroblasts, endothelial and epithelial cells, and placental
tissue." 13. 14 Certain functionally differentiated cells such as monocytes do not secrete detectable quantities of macrophage CSF unless stimulated with tumor-promoting reagents (e.g., phorbol myristate acetate) or with bacterial lipopolysaccharides. IJ . 17 Normal serum will contain some basal amount of colony-stimulating factor. A recent report by Shaddle et al.li< indicated that normal serum contained a mean level of 1.5 ng/ml of macrophage CSF, which is comparable to the levels observed in our study. The steady-state level of macrophage CSF in serum is governed by the rates of production and clearance. In normal individuals these two parameters are at equilibrium. Although the clearance rate and compartments involved in the clearance process have not been completely resolved for macrophage CSF, alveolar macrophages and Kupffer cells in the liver contribute to the removal of macrophage CSF from the circulation. I" Macrophage CSF has a very short half-life, and injection of radioiodinated macrophage CSF is quickly cleared from circulation within 15 minutes. It has been noted that elevated macrophage CSF levels are associated with renal dysfunction. In an earlier study Kacinski et al. 9 have shown that macrophage CSF levels in patients with ovarian cancer were significantly elevated. The increase had a positive correlation with CA 125 levels and showed a marked decrease after tumor debulking. These observations suggested that macrophage CSF levels could be used as an adjunct diagnostic parameter in ovarian cancer.
Volume 166 Number I, Part I
Macrophage CSF serum levels at second look in ovarian cancer
In this study we investigated the possible predictive value of macrophage CSF levels in patients undergoing second-look operation who had low CA 125 levels. All but one of the 33 patients studied received cisplatin before second look. One patient had a renal transplant before ovarian cancer developed. Although significant nephrotoxicity was documented in only two patients, subtle renal toxicity may also account for the elevated macrophage CSF levels found in patients with no evidence of disease at the time of second look. Renal function was not evaluated in the control group. Macrophage CSF might be elevated simply because the patients in the study had been on a regimen of chronic leukocyte-suppressing chemotherapy. All patients, however, had normal white blood cell counts at the time of second-look surgery. Further study is warranted to see whether all patients undergoing chemotherapy have an elevated macrophage CSF level. We conclude that, regardless of tumor status, macrophage CSF levels tend to be elevated at the time of second-look laparotomy. Further work is currently under way to determine if macrophage CSF levels may predict eventual recurrences in those patients with negative second-look laparotomy. REFERENCES l. Marguardt H, Todaro GJ. Human transforming growth
2.
3.
4.
5.
factor: production from a melanoma cell line, purification, and initial characterization. J BioI Chern 1982;257:5220-5. Huang JS, Huang SS, Duel TF. Transforming protein of simian sarcoma virus stimulates autocrme cell growth of SSV-transformed cells through PDGF cell-surface receptors. Cell 1984;39:79-87. Cuttitta F, Carney DN, Mulshine J, et at. Bombesin-like peptides can function as autocrine growth factors in human small cell lung cancer. Nature (Lond) 1985;316: 823-6. Obara T, Ito Y, Kodama T, et at. A case of gastric carcinoma associated with excessive granulocytosis. Production of a colony-stimulating activity by the tumor. Cancer 1985;56:782-8. Suda T, Miura Y, Mizoguchi H, Kubota K, Takaku F. A case of lung cancer associated with granulocytosis and production of colony-stimulating activity by the tumor. Br .I Cancer 1980;41 :980-4.
137
6. Takeda A, Suzumori K, Sugimoto Y, et at. Clear cell carcinoma of the ovary with colony-stimulating factor production. Cancer 1984;54: 1019-23. 7. Ikeda K, Motoyoshi K, Ishizaka Y, et at. Human colonystimulating activity-producing tumor: production of very low mouse-active colony-stimulating activity and induction of marked granulocytosis in mice. Cancer Res 1985;45:4144-9. 8. Ramakrishnan S, Xu FJ, Brandt SJ, Niedel JE, Bast RC Jr, Brown EL. Constitutive production of macrophage colony-stimulating factor by human ovarian and breast cancer cell lines. J Clin Invest 1989;83:921-6. 9. Kacinski BM, Bloodgood RS, Schwartz PE, Carter DC, Stanley ER. The macrophage colony-stimulating factor CSF-l is produced by human ovarian and endometrial adenocarcinoma-derived cell lines and is present at abnormally high levels in the plasma of ovarian carcinoma patients with active disease. Cancer Cells 1989;7:333-7. 10. Stanley ER. Guilbert LJ. Methods for the purification, assay, characterization and target cell binding of a colonystimulating factor (CSF-l). J Immunol Methods 1981 ;42:253-84. II. Metcalf D. The hemopoietic colony-stimulating factors. Amsterdam: Elsevier, 1984. 12. Klug TL, Bast RC Jr, Niloff JM, Knapp RC, Zurawski VR Jr. Monoclonal antibody immunoradiometric assay for an antigenic determinant (CA-125) associated with human epithelial ovarian carcinomas. Cancer Res 1984;44: 104853. 13. Stanley ER, Heard PM. Factors regulating macrophage production and growth. Purification and some properties of the colony-stimulating factor from medium conditioned by mouse L cells. J Bioi Chern 1977;252:4305-12. 14. Sakai N, Kubota M, Shikita M, Yokota M, Ando K. Intraclonal diversity of fibrosarcoma cells for the production of macrophage colony-stimulating factor and granulocyte colony-stimulating factor. J Cell Physiol 1987;133:400-4. 15. Rambaldi A, Young DC, GriffinJD. Expression of the MCSF (CSF-I) gene by human monocytes. Blood 1987;69:1409-14. 16. HoriguchiJ, Warren MK, Kufe D. Expression of the macrophage specific colony-stimulating factor (CSF-l) during human monocytic differentiation. Biochem Biophys Res Commun 1987;141:924-92. 17. Ralph P, Warren MK, Lee MT, et al. Inducible production of human macrophage growth factor, CSF-1. Blood 1986;68:633-63. 18. Shadd Ie PJ , Allen JI , Geier MD, Koths K. Detection of endogenous macrophage colony, stimulating factor (MCSF) in human blood. Exp HematoI1989;17:154-9. 19. Bartocci A, Mastrogiannis DS, Migliorati G, Stockert RJ, Wolkoff AW, Stanley ER. Macrophages specifically regulate the concentration of their own growth factor in the circulation . Cell Bioi 1987;84:6179-83 .
Key words: Ovarian cancer, tumor marker, macrophage colony-stimulating factor, second laparotomy Tumor cells produce growth factors, some of which may regulate the proliferation of cancer cells by autocrine and paracrine pathways.'·7 The aberrant production of trophic factors by cancer cells is due to either gene amplification or genetic translocation, which may bring the growth factor genes under the control of different regulatory elements. Interestingly, among the different growth factors studied thus far, colony-stimulating activity has been detected in many nonhematologic tumors. 6 • 7 Concomitant with the presence of colony-stimulating factor activity, higher levels of circulating leukocytes have been seen in some of these patients." 5 These results, along with studies using murine experimental model systems wherein human tumors were transplanted, indicate constitutive production of cytokines by malignant cells. 7 Although the importance of these factors remains uncertain at the present time, their levels in serum may prove useful as From the Women's Cancer Center' and the Department of Pharmacology,' Umverslty of Minnesota Health Science Center. Supported in part by a grant from the National Cancer Institute (CA48608) and a grant from the Minnesota MedIcal Foundatwn (SMF-
637-89).
Received for publication October 17, 1990; revised April 23, 1991; accepted May 28, 1991. Reprint requests: Steven A. Elg, MD, UMHC Box 395, 420 Delaware St. SE , Minneapolis. MN 55455. 6/1131296 134
Table I. Patient age Group
No.
Control Second-look laparotomy Positive Negative
50
Age (yr, mean ± SD)
33
16 17
41 52 55 49
± ± ± ±
11 14 12 16
Table II. Initial diagnosis of patients undergoing second-look laparotomy Initial diagnosis
No.
Adenocarcinoma (serous or mucinous) Malignant endometrioid tumor Clear cell carcinoma
28 1 4
TOTAL
33
markers to monitor primary tumors or to detect metastases. In a recent study it was reported that each of six ovarian cancer cell lines secrete macrophage colonystimulating factor into the culture medium in the absence of any external stimuli. 8 Whereas lymphoid and myeloid cell lines failed to show any detectable amount
Macrophage CSF serum levels at second look in ovanan cancer
Volume 166 !';umber 1. Part I
135
7
-
6
•
E
m c:
5
•
:E
:::l
a: w
•
4
C/)
•
z C/)
...J W
• •• •••
3
...
>
W ...J
u..
2
C/)
~ ••
880
....•
8
•
08~~~800
()
::E
06 g 0
Fig. 1. Macrophage CSF levels in sera of normal control subjects (open mc/es) and patients with ovarian cancer before second-look laparotomy (solzd czrcles) , Sera samples were analyzed III specific radioimmunoassay. Each value is mean of triplicate independent estimations from individual samples.
of macrophage CSF, a limited number of breast canCer cell lines produced the growth factor in significant quantities. Presence of the colony-stimulating factor was confirmed not only by immunologic (radioimmunoassay) and biologic assay (murine bone marrow cultures) but also by analysis of total cellular ribonucleic acid for macrophage CSF -related transcripts." Kacinski et al. 9 have detected expression of macrophage CSF transcripts in ovarian carcinoma cells by in situ hybridization. Macrophage CSF is a dimeric glycoprotein of about 70 kd and belongs to a group of cytokines that regulate the ontogeny of blood cells. Macrophage CSF selectively stimulates proliferation and differentiation of the mononuclear phagocytic lineage and mature macrophages. 1O II Recently, macrophage CSF serum levels were reported to be elevated in patients with epithelial ovarian cancer." It has been suggested that macrophage CSF may be useful as a potential tumor marker. The purpose of this study is to evaluate the prognostic significance of macrophage CSF levels in ovarian cancer patients undergoing second-look laparotomy. Material and methods
Serum samples from 33 consecutive patients undergoing second-look laparotomy for epithelial ovarian cancer were retrieved from the serum bank of the U niversity of Minnesota Women's Cancer Center. The sera, which had been drawn just before surgery, were ob-
tained between Dec. I, 1985, and June 1, 1988, and stored at - 20° C until analysis. Macrophage CSF levels in these samples were determined in the following way. A polyclonal antiserum against human recombinant macrophage CSF (provided by the Genetics Institute, Cambridge, Mass.) was generated in rabbits by hyperimmunization. The antiserum used in these studies has been well characterized and does not cross react with other growth factors or cytokines. Details of the assay conditions are published elsewhere." This assay system can detect concentrations as low as 500 pg/ml. To minimize interassay variations, aliquots of individual sera samples from five normal subjects were used as internal controls. Macrophage CSF serum levels were determined in a similar manner in 50 healthy female controls. All samples were also assayed for the presence of tumor marker CA 125 with an immunoradiometric method. 12 CA 125 levels were expressed in relative units per milliliter. Mean differences in groups were evaluated with the Student t teSt. A p value <0.05 was considered significant. Results
No clinical evidence of infection was noted at the time of obtaining the serum. Mean ages of the study groups are given in Table I. A summary of initial pathologic diagnoses is shown in Table II. Macrophage CSF levels as determined by radioimmunoassay showed a wide
136
Eig et al.
January 1992
Am
J Obstet G~necol
7
6
E
e.,
5
r
c::
~
:E ::>
c: w en
z en
4
6 [J
3
:~
...J
w > w
2
...J
u..
en
'-?
:E
1
r
I:b dJ
0
Fig. 2. Macrophage CSF levels in sera of patients with ovarian cancer. Open squares, Abnormal pathologic findings at second-look laparotomy; open triangles, normal pathologic findings at secondlook laparotomy. range between nondetectable levels and 5.3 ng/ml in one patient. The distribution of individual samples is shown in scatter plots (Figs. 1 and 2). Sixteen patients with positive pathologic studies at the time of second look had a mean serum macrophage CSF level of 2.31 ± 1.45 ngl ml. The mean level in the 17 patients with negative findings was 1.90 ± 0.86 ng/ml. The difference in the~e levels is not statistically significant
(P>
0.05).
The mean macrophage CSF level in the control group was 1.20 ± 0.51 ng I ml. This is statistically lower than that found in patients with a negative second look (P < 0.005). Subsequently. macrophge CSF levels were compared with those of another tumor marker. CA 125. In all but five of the 33 patients, CA 125 levels were negative «35 U/ml). Two of these five had a macrophage CSF level <3.62 ng/ml (mean ± SD 1.90 + 2) and a negative second look, one had a positive second look and a macrophage CSF level >3.62 ng/ml, and two had a positive second look and a macrophage CSF level <3.62 ng/ml (0.98 and 3.50 ng/ml). The variations seen in the levels of macrophage CSF might be attributed to tumor load. However, in the current study the sample size was too small to evaluate a possible correlation at second-look laparotomy.
Comment Under normal conditions, macrophage CSF is produced constitutively by a variety of cells, including fibroblasts, endothelial and epithelial cells, and placental
tissue." 13. 14 Certain functionally differentiated cells such as monocytes do not secrete detectable quantities of macrophage CSF unless stimulated with tumor-promoting reagents (e.g., phorbol myristate acetate) or with bacterial lipopolysaccharides. IJ . 17 Normal serum will contain some basal amount of colony-stimulating factor. A recent report by Shaddle et al.li< indicated that normal serum contained a mean level of 1.5 ng/ml of macrophage CSF, which is comparable to the levels observed in our study. The steady-state level of macrophage CSF in serum is governed by the rates of production and clearance. In normal individuals these two parameters are at equilibrium. Although the clearance rate and compartments involved in the clearance process have not been completely resolved for macrophage CSF, alveolar macrophages and Kupffer cells in the liver contribute to the removal of macrophage CSF from the circulation. I" Macrophage CSF has a very short half-life, and injection of radioiodinated macrophage CSF is quickly cleared from circulation within 15 minutes. It has been noted that elevated macrophage CSF levels are associated with renal dysfunction. In an earlier study Kacinski et al. 9 have shown that macrophage CSF levels in patients with ovarian cancer were significantly elevated. The increase had a positive correlation with CA 125 levels and showed a marked decrease after tumor debulking. These observations suggested that macrophage CSF levels could be used as an adjunct diagnostic parameter in ovarian cancer.
Volume 166 Number I, Part I
Macrophage CSF serum levels at second look in ovarian cancer
In this study we investigated the possible predictive value of macrophage CSF levels in patients undergoing second-look operation who had low CA 125 levels. All but one of the 33 patients studied received cisplatin before second look. One patient had a renal transplant before ovarian cancer developed. Although significant nephrotoxicity was documented in only two patients, subtle renal toxicity may also account for the elevated macrophage CSF levels found in patients with no evidence of disease at the time of second look. Renal function was not evaluated in the control group. Macrophage CSF might be elevated simply because the patients in the study had been on a regimen of chronic leukocyte-suppressing chemotherapy. All patients, however, had normal white blood cell counts at the time of second-look surgery. Further study is warranted to see whether all patients undergoing chemotherapy have an elevated macrophage CSF level. We conclude that, regardless of tumor status, macrophage CSF levels tend to be elevated at the time of second-look laparotomy. Further work is currently under way to determine if macrophage CSF levels may predict eventual recurrences in those patients with negative second-look laparotomy. REFERENCES l. Marguardt H, Todaro GJ. Human transforming growth
2.
3.
4.
5.
factor: production from a melanoma cell line, purification, and initial characterization. J BioI Chern 1982;257:5220-5. Huang JS, Huang SS, Duel TF. Transforming protein of simian sarcoma virus stimulates autocrme cell growth of SSV-transformed cells through PDGF cell-surface receptors. Cell 1984;39:79-87. Cuttitta F, Carney DN, Mulshine J, et at. Bombesin-like peptides can function as autocrine growth factors in human small cell lung cancer. Nature (Lond) 1985;316: 823-6. Obara T, Ito Y, Kodama T, et at. A case of gastric carcinoma associated with excessive granulocytosis. Production of a colony-stimulating activity by the tumor. Cancer 1985;56:782-8. Suda T, Miura Y, Mizoguchi H, Kubota K, Takaku F. A case of lung cancer associated with granulocytosis and production of colony-stimulating activity by the tumor. Br .I Cancer 1980;41 :980-4.
137
6. Takeda A, Suzumori K, Sugimoto Y, et at. Clear cell carcinoma of the ovary with colony-stimulating factor production. Cancer 1984;54: 1019-23. 7. Ikeda K, Motoyoshi K, Ishizaka Y, et at. Human colonystimulating activity-producing tumor: production of very low mouse-active colony-stimulating activity and induction of marked granulocytosis in mice. Cancer Res 1985;45:4144-9. 8. Ramakrishnan S, Xu FJ, Brandt SJ, Niedel JE, Bast RC Jr, Brown EL. Constitutive production of macrophage colony-stimulating factor by human ovarian and breast cancer cell lines. J Clin Invest 1989;83:921-6. 9. Kacinski BM, Bloodgood RS, Schwartz PE, Carter DC, Stanley ER. The macrophage colony-stimulating factor CSF-l is produced by human ovarian and endometrial adenocarcinoma-derived cell lines and is present at abnormally high levels in the plasma of ovarian carcinoma patients with active disease. Cancer Cells 1989;7:333-7. 10. Stanley ER. Guilbert LJ. Methods for the purification, assay, characterization and target cell binding of a colonystimulating factor (CSF-l). J Immunol Methods 1981 ;42:253-84. II. Metcalf D. The hemopoietic colony-stimulating factors. Amsterdam: Elsevier, 1984. 12. Klug TL, Bast RC Jr, Niloff JM, Knapp RC, Zurawski VR Jr. Monoclonal antibody immunoradiometric assay for an antigenic determinant (CA-125) associated with human epithelial ovarian carcinomas. Cancer Res 1984;44: 104853. 13. Stanley ER, Heard PM. Factors regulating macrophage production and growth. Purification and some properties of the colony-stimulating factor from medium conditioned by mouse L cells. J Bioi Chern 1977;252:4305-12. 14. Sakai N, Kubota M, Shikita M, Yokota M, Ando K. Intraclonal diversity of fibrosarcoma cells for the production of macrophage colony-stimulating factor and granulocyte colony-stimulating factor. J Cell Physiol 1987;133:400-4. 15. Rambaldi A, Young DC, GriffinJD. Expression of the MCSF (CSF-I) gene by human monocytes. Blood 1987;69:1409-14. 16. HoriguchiJ, Warren MK, Kufe D. Expression of the macrophage specific colony-stimulating factor (CSF-l) during human monocytic differentiation. Biochem Biophys Res Commun 1987;141:924-92. 17. Ralph P, Warren MK, Lee MT, et al. Inducible production of human macrophage growth factor, CSF-1. Blood 1986;68:633-63. 18. Shadd Ie PJ , Allen JI , Geier MD, Koths K. Detection of endogenous macrophage colony, stimulating factor (MCSF) in human blood. Exp HematoI1989;17:154-9. 19. Bartocci A, Mastrogiannis DS, Migliorati G, Stockert RJ, Wolkoff AW, Stanley ER. Macrophages specifically regulate the concentration of their own growth factor in the circulation . Cell Bioi 1987;84:6179-83 .