UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
Novel applications of growth factors in solid tumors
Westermann, A.
Publication date
1999
Document Version
Final published version
Link to publication
Citation for published version (APA):
Westermann, A. (1999). Novel applications of growth factors in solid tumors.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
Novel applications of growth factors
in solid tumors
:•.... ' ' '. :'•••
• • • • • • . •
' '
• ' " '. ' '.
• .
•• ' • ' .'•••. • • •
Novel applications of growth factors
in solid tumors
:::;•-. • ••.-.••••
•
,..•••. . '
. • ' . ' •
. . .' ' • • • '•
. ' : . . . ' : • "
•
• . . . • • • • '• . •:••
NOVEL APPLICATIONS OF GROWTH FACTORS IN SOLID TUMORS
1. Lysophosphatidylzuur speelt een belangrijke rol bij het biologisch gedrag van het
ovarium carcinoom.
2. Het remmen van lysophosphatidylzuur is een veelbelovende strategie bij de
behandeling van het ovarium carcinoom.
3. Intraperitoneale toediening van suramine leidt tot een belangrijk farmacokinetisch
voordeel ten opzichte van intraveneuze toediening.
4. De verbeterde haalbaarheid van hooggedoseerde chemotherapie met
stamceltransplantatie is vrijwel volledig te danken aan de ontwikkeling van
groeifactoren, moderne anti-emetica en moderne antibiotica.
5. De symptomen in de herstelperiode na hooggedoseerde chemotherapie met
stamceltransplantatie zijn dusdanig verbeterd, dat de zorg in deze periode vrijwel
geheel thuis kan plaatsvinden.
6. Hoogbegaafdheid is niet zelden een compensatie voor een gebrek, hetzij bij het
kind, hetzij bij de ouders.
7. Het is een misvatting te denken dat stervenden over meer informatie over de
dood beschikken dan anderen.
8. De oriëntatie van het standbeeld van Willem van Oranje in Den Haag, gericht
naar McDonalds en Novotel, en van het parlementsgebouw afgewend, toont aan
dat de Nederlandse democratie functioneert zonder enig historisch besef.
9. In de Nederlandse media worden formuleringen vaak op zo paternalistische wijze
versimpeld, dat juist de elitaire wereldbeschouwing van de journalist wordt bloot
gegeven.
10. Wie niet groot is, moet op een stoel gaan staan.
11. Het Nederlandse arbeidsrecht verlaagt niet het globale stressniveau, maar
verplaatst de stress naar anderen (Wet van behoud van stress).
12. Politieambtenaren boezemen onschuldigen angst in, terwijl boosdoeners hen als
lastposten zien. Daarom is de kans klein dat ze als groep ooit een beste vriend
worden.
13. De verovering van het heelal is even buiten de dampkring tot stilstand gekomen.
Anne Marie Westermann Thesis University of Amsterdam
Cover design: Elma van Imhoff
Lay-out: Chris Bor Medische Fotografie en Illustratie, AMC, Amsterdam Printed by: Thela Thesis, Amsterdam
Publication of this thesis was financially supported by The Netherlands Cancer Institute/Antoni van Leeuwenhoekhuis, Ruitinga van Swieten Stichting, Rhône-Poulenc Rorer, Bayer, Roche, Hoechst Marion Roussel, Bristol-Myers Squibb, Glaxo Wellcome and Amgen.
Novel applications of growth factors in
solid tumors
ACADEMISCH PROEFSCHRIFT
Ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof.dr. J.J.M. Franse
ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit
op woensdag 1 december 1999, te 15.00 uur
door
Anne Marie Westermann
Prof dr S. Rodenhuis Prof dr J.H. Beijnen
Universiteit van Amsterdam Universiteit Utrecht Promotiecommissie Prof dr C.H.N. Veenhof Prof dr W.H. Moolenaar Prof dr D.W. Erkelens Prof dr L. Smets Prof dr M.P.M. Burger Prof dr G.H. Blijham
Universiteit van Amsterdam Universiteit Leiden
Universiteit Utrecht Universiteit van Amsterdam Universiteit van Amsterdam Universiteit Utrecht
1. Background and outline of the thesis [summary] 7 2. Growth factors in human ovarian cancer 17
Cancer Treat Rev 1997;23:113-31
3. Malignant effusions contain lysophosphatidic acid (LPA)-like activity 41
Ann Oncol 1998;9:437-42
4. Feasibility and pharmacokinetics of intraperitoneal suramin in advanced 55 malignancy Submitted
5. Successful intraperitoneal suramin treatment of peritoneal mesothelioma 67
Ann Oncol 1997;8:801-2
6. Feasibility of multiple courses of high-dose cyclophosphamide, thiotepa, and 73 carboplatin for breast cancer or germ cell cancer
J Clin Oncol 1996;14:1473-83
7. At home management of aplastic phase following high-dose chemotherapy 95
with stem-cell rescue for hematological and non-hematological malignancies
Ann Oncol 1999;10:511-7
8. Nederlandse samenvatting voor niet-medici 109
9. Curriculum vitae 115 10. Dankwoord 117
n
• o
Background and outline of the thesis
[summary]
Background
In this century, chemotherapy has attained increasing importance in the treatment of advanced cancer, fostered by the rapidly expanding insights into the origins and behavior of malignant disease. Since the clinical observation that the sulfur mustard gas used in World War I caused lymphoid aplasia, and its subsequent evaluation as an antitumor agent in the 1930s and 1940s ' , a range of natural and synthetic compounds have made their way into the clinic. In the 1960s and 1970s, the dramatically improved outcome for patients with previously incurable tumors such as lymphomas, leukemias and advanced germ cell tumors through the use of newly discovered anticancer agents further enhanced the interest in chemotherapy. However, many of the more common tumors either did not respond to chemotherapy, or rapidly relapsed after the induction of remission.
Skipper et al2 observed that in certain laboratory tumors the growth fraction and cell loss
fraction were highly stable, independent of the size of the cell population. Further study showed that when treated with cytotoxic drugs, a certain dose killed a certain constant fraction of these tumor cells, regardless of the number of cells present, an observation often referred to as the
log kill phenomenon 3. In the absence of regrowth, repeated doses of the same drug will,
therefore, lead to an exponential decrease in the number of surviving tumor cells. This is the theoretical basis for the common practice of chemotherapy administration in repeated cycles of fixed duration.
Since the log kill fraction in the murine tumors used by Skipper often increases as the dose increases, and each drug in combination treatment contributes its own log kill, combination chemotherapy, with enough cycles in high enough doses of each individual agent, should eradicate more, or even all tumor cells. This hypothesis led to the evaluation of increased doses of chemotherapy in the 1980s, a concept called dose escalation. This evolution was greatly aided by improvements in supportive care to overcome toxicity in normal tissues, such as anti-emetics for nausea and vomiting, growth factors for bone marrow suppression, and novel antibiotics for
both prophylaxis and treatment of infectious complications.
The application of dose escalation to cancer treatment culminated in the use of very high, myelo-ablative doses of chemotherapy with alkylating agents, supported by autologous bone marrow transplants at first, and stem cell transplants from the 1990s onwards. While a very hazardous venture in the 1980s, with high mortality and morbidity rates, the risk of high dose chemotherapy with autologous stem cell transplant in the 1990s has decreased formidably through the aforementioned supportive measures. In the 1990s, multiple high dose chemotherapy
courses with stem cell support became feasible 4. The reduction in morbidity and mortality
eased the way for transfer of most of the treatment to the home setting rather than the prolonged 4 to 6-week hospital stays common in the 1980s 5. High dose chemotherapy has improved the
outcome for hematologic malignancies, childhood tumors and germ cell tumors. However, although the likelihood of tumor response and complete remission has increased for most cancers, no survival benefit has been proven for a range of more common malignancies.
In the 1980s, the recognition that many of the more common tumors in the advanced stages could not be cured by even very high dose chemotherapy, led to a change in the direction of anticancer research. The study of tumor cell kinetics and the way it is affected by anticancer agents further focused research on drug scheduling rather than dose alone. Clinical experience, such as the observation that larger tumors are not always more sensitive to treatment than smaller tumors, did not seem to support the murine tumor model-derived concept of Skipper, with assumed exponential proliferation, or constant doubling times, as the only kinetic paradigm in human tumor cell growth. The sigmoid-shaped curve based on the 1825 Gompertz' equation (the 'law of mortality') as a model for tumor cell kinetics does not presuppose a constant doubling time, but shows increasing doubling time with increasing tumor size, with a maximum growth fraction at about 37% of the maximum tumor size. Although the likelihood of a response of a tumor to cytotoxic drugs is independent of the size of that tumor, the relative magnitude of that response will depend on where the tumor is in its Gompertzian growth curve, because smaller tumors will have a larger growth fraction than larger tumors. It then follows that the more tumor cells are killed by chemotherapy, the faster the regrowth of the surviving cells, thus severely restricting the ability to cure tumors6. Norton showed that Gompertzian sigmoid-curve
kinetics are consistent with the shape of survival curves of untreated breast cancer patients, and with the curves of freedom of progression after mastectomy 7.
Further mathematical computations by Norton and Day 8 suggested a model, that predicted
that a schedule of sequential chemotherapy would lead to higher cell kill rates than the alternating schedules that had been the cornerstone of especially hematological and childhood malignancies for some time. Although the first results of such a sequential regimen in breast cancer showed an advantage for this approach over alternating schedules 9, and promising results in small cell lung
cancer were obtained10, these results have not been confirmed by other studies11_13. Thus, sequential
regimens have not found a place in common practice so far.
High-dose chemotherapy in itself is not very likely to overcome the obstacle of regrowth in the Gompertzian model, since even a very small remaining tumor burden at the end of treatment
will rapidly proliferate. However, reducing the dose interval for standard dose combination chemotherapy (a method of dose intensification known as increasing dose-density), could minimize regrowth between chemotherapy cycles, and thus increase cell kill u. Theoretically,
this implies that if dose-dense combination chemotherapy at standard doses cannot be achieved, single-agent therapy at standard doses might be preferred over low-dose combination chemotherapy. This concept of dose-dense, frequent administration of single-agent chemotherapy in standard doses has gained recognition in the dose-dense weekly cisplatin regimens in the treatment of ovarian cancer15, head and neck cancer16, melanoma 17, and non-small cell lung
Although the efficacy of chemotherapy may be gradually increasing, partly due to the implementation of dose-intensification and improved supportive care, long-term results in the advanced stages of the most common tumors such as breast cancer, lung cancer and colorectal cancer have remained modest. One explanation could lie in the survival of a small but consistent, significant number of tumor cells irrespective of the amount of cytoreductive therapy, as was suggested by the results of a phase III study of high-dose chemotherapy in patients with stage IV breast cancer achieving complete remission on standard chemotherapy. Patients randomized to immediate high-dose treatment had longer relapse-free survival, but shorter overall survival than the patients randomized to undergo high dose chemotherapy only at relapse 20. After
high-dose chemotherapy, relapse-free and overall survival were identical, suggesting that there is a limit to the amount of cytoreduction that can be achieved through high-dose chemotherapy. If this is the case, no amount of chemotherapy can be expected to further minimize tumor burden, after a certain maximum cytoreduction has been reached. This does not make sense, however, since some patients are cured by chemotherapy. More likely, regrowth based on Gompertzian kinetics may counteract chemotherapy as its efficacy increases, and a method to inhibit regrowth by non-myelosuppressive means between high-dose courses could eliminate this mechanism. Therefore a strategy directed at inhibition of regrowth of residual disease could be very important.
In recent years, an array of novel targets for restriction of tumor cell proliferation without impairment of chemotherapy effect has been recognized and has led to clinical studies. Examples are ras-farnesylation inhibition, thought to suppress the growth of ras-transformed cancer cells21,22, vaccines
against cancer antigens 2324 and inhibition of angiogenesis, essential for tumor proliferation and
metastasis 25. The recent identification of growth factors and their receptors, in addition to the
elucidation of signal transduction pathways, suggests yet another treatment modality that can be integrated into systemic therapy. Peptide growth factors and their receptors, such as the epidermal
growth factor receptor (EGFR)-superfamily, the platelet-derived growth factor (PDGF) family, basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF), transforming growth factor ß (TGFß), the insulin-like growth factors (IGF), and the interleukins are often overexpressed in human tumors. In addition to these bioactive peptides, the recent recognition of the unique phospholipid, lysophosphatidic acid (LPA), as a growth factor for ovarian cancer and possibly other cancers has further expanded our knowledge and possibilities for treatment
26'27. Growth factors induce cell proliferation through stimulation of specific tyrosine kinase
receptors. Specific patterns of growth factor activity and growth factor receptor expression may explain the behavior of particular tumor types, e.g. a role of LPA that has been suggested in ovarian cancer2S.
Inhibition of growth factor receptor loops is one of the novel strategies for therapy that has spawned a large research interest, and the number of growth factor receptor-inhibitors in development has boomed. Clinical trials of such, generally non-toxic agents include the use of anti-p185HER2 monoclonal antibody (of the EGFR-family) in phase III studies in breast cancer, after promising preclinical and phase II results29'30. For instance, an anti-EGFR monoclonal antibody
is currently being tested in phase II trials in combination with topotecan, with encouraging preclinical results 31'32. We have chosen to exploit the blocking of growth factor-receptor
interactions by suramine, specifically aiming at the inhibition of the bioactive LPA-like lipid growth factors33. In addition, myriad inhibitors of growth factors and their receptors are expected to
enter clinical trials soon, after completion of preclinical testing.
It may well be that the future of systemic anti-cancer therapy in advanced tumors lies in the combination of dose-dense chemotherapy with the non-toxic biologic agents mentioned above. The novel targeted interventions are expected to have the greatest impact in situations of minimal residual disease, such as after intensive chemotherapy, between courses of chemotherapy, and in the adjuvant setting.
Outline of the thesis
The research described in this thesis focuses on the global strategies outlined above. We have developed intensive, repetitive high-dose chemotherapy with peripheral blood stem-cell transplantation (Chapter 6) and we have adapted its practical application to the requirements of cancer patients by removing part of the treatment from the hospital setting (Chapter 7). Although this approach leads to a high proportion of complete remissions in solid tumors, eradication of
all tumor cells is achieved in only a minority of patients, and either inhibition of regrowth between chemotherapy cycles, or effective treatment of minimal residual disease (or both) are required to realize cure.
Regrowth inhibition between chemotherapy cycles might be reached through the use of appropriate growth factor receptor-inhibitors, that are not myelosuppressive. We have studied a novel lipid growth factor, LPA, and its role in peritoneal malignancies such as ovarian cancer and mesothelioma (Chapter 2). LPA may be important as a growth-promoting factor in ovarian cancer (Chapter 3), and we have attempted to inactivate it using intraperitoneal suramin (Chapter 4). Intraperitoneal administration is feasible, and has a distinct pharmacokinetic advantage over systemic administration. Its effect as an antitumor agent remains to be established, but we have observed an encouraging response in a single patient with peritoneal meothelioma (Chapter 5). The integration of LPA-inhibition as a regrowth-inhibitor between cycles of high-dose chemotherapy is envisaged, but is, at present, no more than a challenging prospect.
References
1 Adair CPJ, Bagg HJ. Experimental and clinical studies on the treatment of cancer by dichloroethyl-sulphide (mustard gas). Ann Surg 1931;93:190
2 Skipper HE. Kinetics of mammary tumor cell growth and implications for therapy Cancer 1971;28:1479-99
3 Skipper HE, Schabel FM Jr, Wilcox WS. Experimental evaluation of potential anticancer agents xiii: On the criteria and kinetics associated with 'curability' of experimental leukemia. Cancer Chemother Rep 1964;35:1
4 Rodenhuis S, Westermann AM, Holtkamp MJ, Nooijen WJ, Baars JW, van der Wall E et al. Feasibility of multiple courses of high-dose cyclophosphamide, thiotepa, and carboplatin for breast cancer or germ cell cancer. J Clin Oncol 1996;14:1473-83
5 Westermann AM, Holtkamp MMJ, Linthorst GAM, Van Leeuwen L, Willemse EJM, Van Dijk WC et al. At home management of aplastic phase following high-dose chemotherapy with stem-cell rescue for hematological and non-hematological malignancies. Ann Oncol 1999;10:511-17 6 Norton L. Adjuvant breast cancer therapy: Current status and future strategies - growth kinetics
and the improved drug therapy of breast cancer. Semin Oncol 1999;26(suppl 3): 1-4 7 Norton L. A Gompertzian model of human breast cancer. Cancer Res 1988;48:7067-71
8 Norton L, Day R. Potential innovations in scheduling of cancer chemotherapy. Important Adv Oncol 1991:57-72
9 Bonadonna G, Zambette M, Valagusa P. Sequential or alternating doxorubicin and CMF regimens in breast cancer with more than three positive nodes. JAMA 1995;273:542-7
10 Twelves CJ, Goldman J, Ash CM, Souhami RL, Harper PG, Spiro SG et al. Sequential chemotherapy in good-prognosis patients with small-cell lung cancer. Cancer Chemother Pharmacol 1991;28:139-41 11 Laurie JA, Hahn RG, Therneau TM, Patel SR, Mailliard JA, Windschitl HE et al. Chemotherapy for
hormonally refractory advanced prostate carcinoma. A comparison of combined versus sequential treatment with mitomycin C, doxorubicin, and 5-fluorouracil. Cancer 1992;69:1440-4
12 Haioun C, Lepage E, Gisselbrecht C, Bastion Y, Coiffier B, Brice P et al. Benefit of autologous bone marrow transplantation over sequential chemotherapy in poor-risk aggressive non-Hodgkin's lymphoma: updated results of the prospective study LNH87-2. Groupe d'Etude des Lymphomes de l'Adulte. J Clin Oncol 1997;151131-7
13 Ueoka H, Kiura K, Tabata M, Kamei H, Gemba K, Sakae K et al. A randomized trial of hybrid administration of cyclophosphamide, doxorubicin, and vincristine (CAV)/cisplatin and etoposide (PVP) versus sequential administration of CAV-PVP for the treatment of patients with small cell lung carcinoma: results of long term follow-up. Cancer 1998;83(2):283-90
14 Levein L, Hryniuk WM. Dose intensity analysis of chemotherapy regimens in ovarian carcinoma. J Clin Oncol 1987;756-67
1 5 Bolis G, Favalli G, Danese S, Zanaboni F, Mangili G, Scarabelli C et al. Weekly cisplatin given for 2 months versus cisplatin plus cyclophosphamide given for 5 months after cytoreductive surgery for advanced ovarian cancer. J Clin Oncol 1997;15:1938-44
16 Planting AS, de Mulder PH, de Graeff A, Verweij J. Phase II study of weekly high-dose cisplatin for six cycles in patients with locally advanced squamous cell carcinoma of the head and neck. Eur J Cancer 1997;33:61-5
17 Planting AS, van der Burg ME, Goey SH, Schellens JH, Vecht C, de Boer-Dennert M et al. Phase II study of a short course of weekly high-dose cisplatin combined with long-term oral etoposide in metastatic malignant melanoma. Eur J Cancer 1996;32A:2026-8
18 Planting A, Kho S, van der Burg M, Goey H, Schellens J, van den Bent M et al. A phase II study of weekly high-dose cisplatin combined with oral etoposide in advanced non-small-cell lung cancer. Cancer Chemother Pharmacol 1997;40:347-52
19 Planting AS, de Wit R, van der Burg ME, Stoter G, Verweij J. Phase II study of a closely spaced ifosfamide-cisplatin schedule with the addition of G-CSF in advanced non-small-cell lung cancer and malignant melanoma. Ann Oncol 1996;7:1080-2
20 Peters WP, Jones RB, Vredenberg J et al. A large, prospective , randomized trial of high-dose combination alkylating agents with autologous bone marrowsupport as consolidation for patients -with metastastic breast cancer achieving complete remission after intensive doxorubicin-based
induction therapy. Proc Am Soc Clin Oncol 1996; 1 5:121 [abstr]
21 Gibbs JB, Oliff A, Kohl NE. Farnesyltransferase inhibitors: Ras research yields a potential cancer therapeutic. Cell 1994;77:175-8
22 Zujewski J, Horak ID, Bol CJJG, Woestenborghs R, End D, Chiao J et al. A phase I and pharmacokinetic study of farnesyltransferase inhibitor, R11 5777, in advanced cancer. Proc Am Soc Clin Oncol 1999; 18:739 [abs]
23 Hsueh EC, Nathanson L, Foshag LJ, Essner R, Nizze JA, Stern SL et al. Active specific immunotherapy with polyvalent melanoma cell vaccine for patients with in-transit melanoma metastases Cancer 1999;85:2160-9
24 Vermorken JB, Claessen AM, van Tinteren H, Gall HE, Ezinga R, Meijer S et al. Active specific immunotherapy for stage II and stage III human colon cancer: a randomised trial Lancet 1999;353:345-50
25 Folkman J.Seminars in Medicine of the Beth Israel Hospital, Boston. Clinical applications of research on angiogenesis [Review] N Engl J Med 1995;333:1757-63
26 Xu Y, Mills GB. Activation of human ovarian cancer cells: role of lipid factors in ascitic fluid. In: Sharp F, Mason P, Blackett T, Berek J [eds.] Ovarian cancer 3, 1 st edition. London, Chapman and hall Medical, 1995;121-35
27 Westermann AM, Havik E, Postma FR, Beijnen JH, Dalesio O, Moolenaar WH, Rodenhuis S. Malignant effusions contain LPA-like activity. Annals Oncol 1998;437-42
28 Westermann AM, Beijnen JH, Moolenaar WH, Rodenhuis S. Growth factors in human ovarian cancer. Cancer Treatm Rev 1997;23:113-31
29 Baselga D, Tripathy D, Mendelsohn J et al. Phase II study of weekly intravenous humanized recombinant anti-p185 HER2 monoclonal antibody in patients with HER2/neu-overexpressing metastaic breast cancer. J Clin Oncol 1996; 14:737-44 Proc Am Assoc Cancer Res 1999;40:155 [abs] 30 Norton L, Slamon D, Leyland-Jones B, Wolter J, Fleming T, Eirmann W et al. Overall survival advantage
to simultaneous chemotherapy (CRx) plus the humanized anti-HER2 monoclonal antibody Herceptin in HER2-overexpressing metastatic breast cancer. Proc Am Soc Clin Oncol 1999;18:483 [abstr] 31 Baselga J, Norton L, Masui H et al. Anti-tumor effects of doxorubicin in combination with
anti-epidermal growth factor receptor monoclonal antibody. J Natl Cancer Inst 1993;85:1327-33 32 Ciardiello F, Bianco R, Damiano V, De Lorenzo S, Pepe S, De Placido S et al. Antitumor activity of
sequential treatment with topotecan and anti-epidermal growth factor receptor monoclonal antibody C225. Proc Am Ass Cancer Res 1999;40:1 55 [abstr]
33 Westermann AM, Dubbelman R, Moolenaar WH, Beijnen JH, Rodenhuis S. Successful intraperitoneal suramin tretament of peritoneal mesothelioma. Ann Oncol 1997;8:801-2
Anneke M. Westermann
1, Jos H. Beijnen
1, Wouter H. Moolenaar
2,
Sjoerd Rodenhuis
1From
1the Department of Medical Oncology and
2the Division of
Cellular Biochemistry, The Netherlands Cancer Institute,
Amsterdam, The Netherlands.
n
• o
CD
Growth factors in human ovarian NJ
cancer
Abstract
Background Ovarian cancer remains confined to the peritoneal cavity, even in the late stages of the disease. This particular growth pattern might be the result of the activity of growth factors present in malignant ascites, a common feature of advanced disease. However, presently identified growth factors cannot fully account for the mechanics of ovarian cancer.
Methods Papers on growth of ovarian cancer cell lines in vitro and the clonogenic assay of human ovarian cancers, and the factors influencing them are summarized. Presently known peptide growth factors were reviewed for their association with human ovarian cancer, as were cytokines. The information on the growth-promoting properties of ascites was surveyed, and the available evidence for possible novel growth factors was collected.
Results Growth of ovarian cancer cells, both in cell lines and in the clonogenic assay, is stimulated by peptide growth factors, although no combination of peptide growth factors can replace serum or cell free ascites. The growth-promoting characteristics of ascites seem to be completely explained by the presence of lysophosphatidic acid (LPA), a bioactive phospholipid with mitogenic features.
Conclusion The bioactive lipid LPA seems to be an important growth promoter present in ascites of ovarian cancer patients. LPA inhibition may be a novel target for therapy.
Introduction
Human ovarian cancer is the leading cause of death from gynecologic malignancy. In most cases, the initial diagnosis is established at an advanced stage, when current therapy can only cure a small proportion of patients. Even in the presence of advanced disease, there is rarely evidence of tumor cells outside the peritoneal cavity.1 Local growth factors may play a part in
promoting intraperitoneal spread and growth, and it is conceivable that these could represent unique targets for therapy.
Although many studies suggest that malignant ascites may contain one or more specific growth factors required for the proliferation of ovarian cancer cells, so far no specific ascites-borne growth factors have been isolated and identified that are candidates for such a role. A group of bioactive lipids, the lysophosphatidic acids, are present in both serum and in malignant effusions.2 It is possible that the elusive ascitic growth factor may belong to this novel group of
established mitogens. The present review summarizes current insights into the function of well-characterized growth factors in ovarian cancer growth, and discusses the potential influence of the abovementioned ascites-borne lipid growth promoters.
Growth of ovarian carcinoma cell lines in vitro
A large number of ovarian cancer cell lines have been described since the early seventies.323
Media used to maintain cell lines were routinely supplemented with antibiotics, vitamins and fetal calf serum (FCS) or horse serum. Although insulin, transferrin, gonadotropins, estradiol, dexamethasone, L-thyroxin or epidermal growth factor (EGF) were added in some instances, human ovarian cancer cell lines almost universally required the addition of serum. In Table 1, a selection of human ovarian carcinoma cell lines with their requirements for growth in vitro is listed.
Establishment of ovarian cancer cell lines in serum-free medium is apparently difficult. In the early 1980s, however, investigators succeeded in growing rat ovarian cancer cells in serum-free medium by adding EGF and transferrin.24 The serum requirement to sustain human ovarian
cancer cell lines could be reduced from 20 down to 0.5% when suitable growth factors were employed (table 1). In the complete absence of serum, serum components such as fetuin or bovine serum albumin were usually necessary to maintain the cell lines.10'2022'23 Rare exceptions
to this rule include certain subcultures of established ovarian cancer cell lines, as reported by Jozan ef a/.22 and Golombick etal.25 The latter described a cell line grown in a solution of amino
acids, vitamins and simple salts. Berchuck etal. could sustain growth of ovarian cancer cell lines in serum-free medium for 48 hours, but not in permanent culture.26
It has been suggested that cell lines requiring little or no serum for in vitro growth are less well differentated than those that do not grow in chemically defined media,10 but this has not
Table 1. Reports of permanent human ovarian cancer cell lines in vitro.
Cell line Supplements References
154, 163 20% fetal calf serum (FCS), antibiotics DiSaia 3
#2774 20% FCS, antibiotics Freedman 4
163 20% FCS, antibiotics Kanabus 5
HeW 20% FCS Pattillo 6
A7, A10 10% FCS, antibiotics Abu Sinna 7
COLO 110, 316, 319, 330 10-20% FCS, antibiotics Woods 8 NIH:OVCAR-3 2% FCS, insulin 10 lU/ml, antibiotics Hamilton 9 EFO-3, EFO-21, EFO-27, 10% FCS, fetuin 5 |ig/ml, transferrin 2.5 ng/ml, insulin Simon 10 EFO-47, EFM-63 0.08 lU/ml, human menopausal gonadotropin
0.04 lU/ml, L-thyroxine 10 nM, antibiotics
EFO-21, EFO-47 fetuin 5 u.g/ml, 0 . 1 % bovine serum albumin, transferrin 2.5 ng/ml, insulin 0.08 lU/ml, L-thyroxine 10 nM, antibiotics
OAW42 10% FCS Wilson 11
CAOV-3, SKOV-3, HOC-1, 10-15 % FCS Buick 12
HOC-7, HEY
la 288 10% horse serum, 10% FCS, 1 \i/\ transferrin, T3 6.5 ng/ml, dexamethasone 14 ng/ml, 108M estradiol, insulin 20 |ig/ml, antibiotics
Clamon 13
IGROV1 10% FCS Benard 14
JA-1.TR175, TR170 10% FCS Hill 15
HTOA 15% FCS Ishiwata 16
PE01, PE04, PE06, PE014, 10% FCS Langdon 17
PE016, PE023, PEA1, PEA2, T014
DO-s 2-15% FCS Briers 18
A1, A69, A90, A121 20% FCS Crickard 19
UWOV1, UWOV2 2-5% FCS, antibiotics Golombick 20
UWOV2 (Sf) insulin 1 mg/l, transferrin 1 mg/l, bovine serum albumin 500 mg/l
OMC-3 10% FCS Yamada 21
OVCCR1/sf insulin, transferrin, epidermal growth factor (EGF) in varying concentrations3
Jozan 22 SCHM-1, RIC-2, MAC-2, 3% FCS, antibiotics, EGF 5 ng/ml, insulin 5 ng/ml,
SIB-1 transferrin 10 u,g/ml
MAC-2, SIB-1 Bovine serum albumin 3 mg/ml, antibiotics, EGF 5 ng/ml, insulin 5 ng/ml, transferrin 10 ng/ml
Hirte 23
UWOV2 (Pf) vitamins Golombick 25
a0.5% serum was required for each passage to allow cell adhesion to the plastic. been confirmed in later publications.23
The scarcity of serum-free cell lines strongly suggests the presence of as yet unknown growth promoting factors in serum. A large number of potential candidates for this serum-derived growth factor have been proposed and will be discussed below.
Clonogenic assay of human ovarian carcinoma cells
In the late 1970s, interest arose in the isolation of human tumor cells directly from patients for drug sensitivity testing and other studies. Successful direct cultures of human cancer cells were rarely described27 until Hamburger and Salmon developed a clonogenic assay providing
conditions for cell growth of most tumor types.28 In their original bioassay, tumor cells were
cultured in a two-layer system, with the upper layer containing tumor cells suspended in agar supplemented with horse serum, antibiotics, glutamin, CaCI2, insulin, asparagine, dextran and
2-mercaptoethanol. The lower agar layer contained medium conditioned by adherent spleen cells of mineral oil-primed BALB/c mice. The majority (85 to 100%) of cells obtained from either biopsies or effusions of ovarian cancer patients could form colonies using this method.29 32 Selected reports on conditions for the clonogenic assay of human ovarian cancer cells are
summarized in Table 2.
It later turned out that feeder layers were not the only method to allow clonogenic in vitro growth, although it proved impossible to clone ovarian cancer cells in serum-free conditions with the addition of wellknown growth factors alone.33 For in vitro growth, ovarian cancer cells
may require a specific and individual set of growth factors, as has been suggested for other tumors grown in serum-free conditions, such as small cell lung cancer.34
Peptide growth factors
Peptide growth factors play a pivotal role in the control of proliferation of normal and tumor cells.35 They exert their action through binding of a specific cell surface receptor after secretion
into the extracellular environment. The capacity of tumor cells to produce and react to their own growth factors via functional external receptors is a central principle that has been termed 'autocrine secretion' or 'autocrine loop'. This paradigm includes the notion that growth factors can cause malignant transformation through both overproduction of positive autocrine growth factors and the loss of negative growth control present in normal cell growth.36
The role of peptide growth factors in human ovarian cancer cell lines and in direct cultures is summarized in Table 3.
Epidermal growth factor family
Human epidermal growth factor (EGF) was among the first polypeptide growth factors isolated.37
It stimulates growth of normal human epithelial cells in culture as well as clonogenic growth of human tumor cell lines and freshly isolated tumor material, although inhibition of some cell lines
EGF-Table 2. , Reports on short term culture of human ovarian carcinoma from bi opsies or effusions. Mat.a Medium Supplements Growth col/5x10 5 cells Ref. T/E 78 Eagle's modified essential
medium (MEM) E 8 two-layer system (TLS) of
0.3% agar in CMRL 1066, on medium conditioned by adherent spleen cells of mineral oil-primed BALB/c mice T/E 31 -TLS with medium conditioned
by adherent spleen cells of mineral oil-primed BALB/c mice -TLS with type-0 human
red blood cells
-TLS with medium conditioned by CD-1 or DBA/2 or non-oil primed BALB/c mice -TLS with medium conditioned
by MA-184, Wi-38 cell lines -TLS with cell-free autologous effusion 1:4
E 27 standard TLS T 51 enriched Eagle's MEM
E 13
T6 enriched Eagle's MEM in 0.3% agar
15%FCS, 10% human cord serum
15% PCS, 20% horse serum, antibiotics, 2-ME
15% horse serum, antibiotics, +/- 2-ME
10%FCS, 10% human cord serum, antibiotics, irradiated xenogeneic feeder layer of adherent peritoneal cells
10%FCS, 10% human cord serum, antbiotics
10%FCS, 10% human cord serum, antibiotics, irradiated xenogeneic feeder layer of adherent peritoneal cells 10% PCS, 10% human cord serum, antbiotics
69% 100% 10% PCS, 15% horse serum 81 % 4 1 % 3 1 % 100-800 85% 99-2549 320-833 109-320 0 29 20-620 100% 299(26-1464) 100% 507(50-2528) 27 28 29,30 32 31 T/E 16 T35 standard TLS 15% FCS, 20% horse 53% standard TLS, with 0-75% enriched CMRL 1066 in top layer serum 15% FCS, 20% horse serum, substitution of CMRL with 25-100% cell free ascites 71%
standard TLS 2.5-15% horse serumb EGF 50 ng/ml, insulin 5 p.g/ml, transferrin 5 u,g/ml, selenium 5 ng/ml, 17B-estradiol W M , hydrocortisone lO^M 66% 0 113 33
Table 3. Overview of growth factors in ovarian cancer Substance Effect on ovarian
cancer cell growth
References
Epidermal growth factor (EGF)
Amphiregulin (AR)
Transforming growth factor ß (TGF-ß) Platelet derived growth factor (PDGF) Basic fibroblast growth factor (bFGF) Insulin-like growth factors (IGF) Interleukin 1 (IL-1)
Interleukin 6 (IL-6)
Tumor necrosis factor a (TNF-oc)
Macrophage-colony stimulating factor (M-CSF) Cell free ascites (CFA)
Lysophosphatidic acid (LPA)
dual 26,39
stimulating 22,38,40,41 (cell lines),33,43-46 (direct cultures)
dual 65
inhibiting 39,42,70-72 (cell lines),73 (direct cultures) no effect 26 stimulating 19 stimulating 85 stimulating 93 no effect 95 stimulating 93 no effect 97 stimulating 11,38,58,116,117 (celllines),29,30 105, 113,118 (direct cultures) stimulating 128, Jalink 1995 (unpublished
observations)
receptor commonly present on many different cell types, including ovarian cancer cells. This general g r o w t h stimulating effect has been reported in most ovarian cancer cell lines26-3842 and also in direct clonogenic assays of ovarian carcinoma.33'43-46 Some cell lines, however, showed no response to EGF,26 or were inhibited by EGF. Varying responses of cell lines to peptide g r o w t h factors are common. They are not usually dependent on concentration, but they are possibly due to different post-receptor events in the various cell lines.39
EGF was later shown to be part of a family of structurally homologous peptides also including transforming growth factor a (TGF-a)47 and amphiregulin (AR).48 The members of this family bind to the EGF-receptor (EGFR), causing receptors to dimerize and initiate intracellular signalling. The EGFR has marked homology with the human c-erbB-2 gene (which is identical to HER-2 or the rat
neu oncogene) product, a 185 kDa protein (p185e*B2) with tyrosine kinase activity.49 Although TGF-a and EGF do not bind to p185e*B2, the not yet fully characterized 30 kDa glycoprotein ligand of the erbB-2 oncogene product has been shown to interact directly with the EGFR.50 Other members of this receptor family, c-enbB-3 and c-erbB-4, have recently been identified, and the dose similarity in structure of the four receptors allows extensive heterodimerization of these different receptor
a T denotes solid tumor, E effusion, and T/E both as a source of malignant cells, followed by the number of patients tested.
b The serum requirement could be reduced by the growth factors and hormones specified in the serum-free experiments.
subtypes.51-52 Immunohistochemically, increased expression of these receptor proteins has been
demonstrated in malignant ovarian tumors compared to benign ones.53
TGF-a differs from EGF in that it can induce anchorage-independent growth not only in cancer cells but also in untransformed, non-neoplastic, indicator cells.54 TGF-a promotes ovarian cancer
cell line growth in vitro and enhances direct clonogenic growth of ovarian tumor cells, thereby mimicking EGF,3M1 while the amphiregulin effects are similar but less pronounced.48
Evidence to support the autocrine loop concept in ovarian cancer was presented by Bauknecht
et al., who documented elevated levels of EGF and of EGF-like factors in crude cellular extracts
of ovarian tumors compared to non-malignant tissue.55 In contrast, neither Berchuck etal.26 nor
Ottensmeier et a I.56 found EGF activity in ovarian cancer cell-conditioned media. Production of
EGF and TGF-a by ovarian cancer cell lines was, however, demonstrated by Kurachi etal.40 and
Stromberg et al.57 Hanauske et al. showed a higher frequency and level of TGF-a in malignant
effusions in a variety of tumors including ovarian cancer, as compared to benign effusions.58
Further support for a TGF-a/EGFR autocrine loop in ovarian cancer was provided by experiments with primary tumor cultures in which TGF-a but not EGF was detected in EGFR-positive ovarian tumors.59 Gordon etal. reported mRNA expression of EGF, TGF-a, AR, EGFR and erbB-2 in three
ovarian cancer cell lines, both in serum-supplemented and serum-reduced conditions.60 Other
authors found that in the presence of exogenous EGF the amount of serum needed to sustain in
vitro proliferation of cell lines42 and direct clonogenic assay33 could be lowered, though it could
not compensate completely for the serum requirement.
EGFR-positive ovarian cancer cells appear indeed to require the presence of a growth stimulating EGFR ligand for growth in vivo, as shown by experiments in nude mice who were EGF depleted by sialoadenectomy. Subcutaneous injection of EGFR-positive ovarian cancer cells led to tumors in animals receiving exogenous EGF, but not in those not receiving the growth factor.61
In ovarian cancer cell lines overexpression of erbB-2 is frequently encountered.64 erbB-2
expression can be detected by immunohistochemistry in 30 to 70% of human ovarian cancers,62'63
but in only 5 to 10% of normal ovarian cells.64
Both normal and ovarian cancer cells express and respond to amphiregulin. At low concentrations (picomolar range), amphiregulin inhibits growth of certain normal and malignant ovarian cells. At higher concentrations (nanomolar range) growth stimulation can be observed. These effects may vary between cell lines,65 as described above for EGF effects.
Transforming growth factor ß
The hallmark of transforming growth factors (TGF) is their ability to confer a transformed phenotype on untransformed indicator cells. In the early 1980s the TGFs were found to include two groups: TGF-a mentioned previously, which is structurally and functionally homologous
with EGF, and TGF-ß.66 TGF-ß does not bind to the EGFR, but a synergistic effect was observed
with EGF orTGF-a, leading to a potent transforming action on untransformed indicator cells.67
Later research indicated that the predominant effect of TGF-ß on normal cells in vitro is inhibition of proliferation. This dual effect makes TGF-ß a central mediator of cell growth regulation. TGF-ß is considered a principal negative control element in cell growth, and it is present in most body fluids and tissues.68 Loss of responsiveness to the growth inhibitory effects of TGF-ß may be a
general characteristic of malignant transformation. The TGF-ß family can be divided into at least three different subtypes in humans with similar in vitro functions and with multiple receptor subtypes.69
Studies on the effects of TGF-ß on ovarian tissue showed inhibition of growth of normal epithelial ovarian cells,70 ovarian carcinoma cell lines,39'427072 and directly cultured ovarian cancer cells,73
However, some cell lines in these studies showed no response to exogenous TGF-ß at all.26'70 or
only to one or two particular subtypes of TGF-ß.72
Production of TGF-ß was demonstrated in normal ovarian tissues,7074 ovarian cancer cell
lines,4270-7275 and directly cultured ovarian cancer cells.73-74 Though some tumors seemed to
have lost the capacity to produce TGF-ß,73 others were found to have increased expression of
certain TGF-ß isoforms.74 TGF-ß is secreted by most cells in a biologically inactive form and
regulation of bioactivation may be of physiologic importance. As a result, immunohistochemical demonstration of the peptide in tumors or cells may not directly reflect its biological significance.
Platelet derived growth factor
Platelet derived growth factor (PDGF) is a polypeptide growth factor first described in the 1980s as a megakaryocyte-derived mitogen for mesenchymal cells, but later it was also shown to be synthesized by a plethora of other normal and transformed cells. It stimulates both epithelial and endothelial cells in an autocrine fashion.76 The growth factor has three isoforms and binds
to two defined receptors (termed PDGFRa and PDGFRß).77
Normal ovarian tissue or benign ovarian tumors do not show PDGF expression, although PDGF-expression is common (70-100%) in ovarian cancer tissue and in cell lines,78 as detected
by immunohistochemistry. PDGFRa is less frequently expressed than PDGF in human tumors (0-35%).79 PDGF was found to have an autocrine role in a cell line growing in a defined
serum-free medium.75 Berchuck et al. found no stimulation of ovarian cancer cell lines by exogenous
PDGF,26 consistent with the lack of expression of PDGF receptors in other studies. It has therefore
been suggested that PDGF acts in a paracrine fashion in ovarian tumors and mediates supportive connective tissue stroma formation.71
Basic fibroblast growth factor
Basic fibroblast growth factor (bFGF) is part of a family of heparin-binding growth factors for many mesodermal and ectodermal cells, though originally described as a mitogen for cultured fibroblasts. The family consists of at least seven distinct proteins and four receptors. (For review see 80 ,81)
bFGF as well as the FGFR1 receptor subtype have been detected in ovarian cancer cell lines, while exogenous bFGF stimulated ovarian cancer cell line proliferation.82 An earlier study indicated
growth stimulation by FGF of only one of four ovarian carcinoma cell lines.26
Insulin-like growth factors
The insulin-like growth factors (IGF) IGF-1 and IGF-2 are polypeptides with endocrine, paracrine and autocrine effects on cell proliferation. Their mitogenic effects in vitro are suggestive of a role in growth regulation and even in the oncogenesis of malignant tumors. The IGFs are present in body fluids, bound to specific binding proteins that not only carry IGF, but also regulate access to the IGF receptors, thereby mitigating IGF's mitogenic action, (reviewed in 83) Both IGF-1, its binding proteins and its receptor are expressed in ovarian cancer cell lines and ovarian tumors.84 In a small study, Karasik et al. found that levels of IGF-1 and one of its binding
proteins could discriminate between malignant and benign ovarian tumors.85 IGF-1 stimulated
the growth of two ovarian cancer cell lines, whereas specific IGF-1 receptor RNA anti-sense oligodeoxynucleotides markedly stilted the growth of those cell lines. This hints at a role in the autocrine growth of ovarian cancer cell lines for IGF and its receptor,86 though more extensive
data are required to corroborate these findings.
Cytokines
Cytokines are peptide molecules initially thought to modulate the proliferation and bioactivity of cells of the immune system. It is presently clear that cytokines are peptide factors that mediate a host of different reactions in different cell populations and in the extracellular environment, and the distinction between cytokines and other peptide growth regulatory factors has lost most of its significance. The cytokine family was originally divided into interleukins (thought to be produced by leukocytes), lymphokines (originating from lymphocytes), monokines (released by monocytes), interferons (IFN) and colony-stimulating factors (CSF). These names stem from historical concepts about function and origin of these proteins, that are largely structurally unrelated, (reviewed in 87)
type, dose and circumstances.88 A large body of evidence has been accumulated that cytokines
are produced by ovarian tumors, but growth modulatory effects of cytokines have not been decisively documented, (for a review of the role of cytokines in ovarian cancer see ref 89).
Tumor necrosis factor
TNF was one of the first cytokines to be identified.90 Although the subtypes TNF-a and TNF-ß
have a similar scale of biological activities, TNF-ß effects are generally weaker, and TNF has come to mean TNF-a unless specified otherwise in most publications. Both TNF and its encoding mRNA have been detected in uncultured ovarian cancer specimens,9193 although Wu eta/.found
no expression in ovarian cancer cell lines in vitro.93 Exogenous TNF as well as interleukin 1 (IL-1)
could reinduce TNF expression in a clonogenic assay of ovarian cancer cells that had lost this ability after 7 days in culture.93
Interleukin 1
IL-1 is a cytokine produced primarily by monocytes and macrophages, with both inhibitory and stimulatory effects on human tumor cells. Two subtypes, IL-1 a and IL-1 ß, can be discerned, that bind to the same receptors. Promotion of tumor metastases by IL-1 has been described for some tumors,94 and in a proportion of ovarian tumors IL-1 can be detected.89 IL-1 can induce
TNF expression in a primary culture of ovarian cancer cells, as described above.93
Interleukin 6
II-6 is a pleiotropic cytokine with both growth- and differentiation-inducing activity, that can be produced by T lymphocytes, monocyte/macrophages, and a variety of tumor cells. It is also secreted by ovarian cancer cell lines and by ovarian tumor cells in the clonogenic assay, and it can be detected in malignant ascites.95 IL-6 mRNA expression has furthermore been demonstrated
in normal ovarian epithelial cells.96 However, exogenous IL-6 did not stimulate proliferation of
six different ovarian cancer cell lines.95
Culture supernatants from human monocytes contain significant amounts of IL-6, TNF-a and IL-1, and stimulate clonogenic growth of ovarian cancer cell lines in vitro.97 These growth factors
could, however, not completely account for the growth enhancement observed when ovarian cancer cells were grown in peripheral blood monocyte-conditioned medium,97 suggesting that
as yet unidentified monocyte-derived factors also play a part.
Macrophage colony stimulating factor
Macrophage-colony stimulating factor (M-CSF, also called CSF-1) enhances monocytopoiesis, attracts macrophages and can stimulate production of TNF-a. In ovarian cancer cell lines and in
ovarian cancer in vivo, expression of M-CSF,98100 and its receptor101J02 could be demonstrated,
while normal ovarian epithelial cells do not express M-CSF. Addition of either one of the colony stimulating factors M-CSF, G-CSF or GM-CSF did not influence growth of ovarian tumor cell lines,97 although this has been reported for other CSF-receptor positive cancer cell lines.103
Cytokines in ascites
Increased levels of CSF,104'105 TNF-a,100'106<107 |L-1 ß95-100-106 and |L-6 95-100-l0S-108 have been detected
in effusions of ovarian cancer patients,100 while benign ovarian follicular fluid contained much
lower levels, especially of IL-6 and ll_-1ß.106 It has been suggested that high CSF levels in ascites
correlate with a poor prognosis.105 The cellular source of these cytokines is uncertain. Malignant
effusions are known to contain large numbers of macrophages. An increased number of cytokine-producing macrophages found at ovarian cancer sites may account for some of the biological characteristics of the tumor.
In conclusion, there is strong evidence that the peptide growth factor/cytokine network is inappropriately functioning in human ovarian cancer, and that this imbalance contributes to ovarian cancer cell growth. The growth promoting characteristics of serum can not, however, be fully substituted for or explained by the activities of the known peptide growth factors and cytokines.
Growth promoting properties of malignant effusions
As in many other solid tumors, no single group of growth factors or cytokines has been shown to be unequivocally responsible for ovarian tumor growth. Research into possibly specific growth substances has enhanced interest in novel growth substances present in ascites. In addition, cell lines grew more rapidly during establishment in the presence of cell-free ascites (CFA) compared to serum,17 suggesting the presence of adequate levels of one or more critical
growth promoters in ascites.
Macrophages
For clonogenic growth of ovarian cancer cells obtained from effusions, feeder layers were not required in the clonogenic assay. When tumor cells freshly obtained from patients were kept suspended in their native ascites, colony growth could be observed until 24 hours after collection.32
This was ascribed to the abundant macrophages present in malignant effusions, providing a feeder layer effect. This hypothesis was supported by the fact that removal of macrophages from cell populations in malignant effusions lowered the average number of colonies in the original clonogenic assay by Hamburger ef a/.2930 Buick ef a/, showed that this decrease in
plating efficiency was reversible in reconstitution experiments that used the removed autologous adherent cells as an underlayer in a two-layer agar system.109 Even xenogeneic peritoneal
macrophages (obtained from mice) were able to improve the cloning efficiency of primary ovarian cancer cells in culture.31
Mantovani etal. found an array of effects of macrophages isolated from human ascitic ovarian tumors on in vitro tumor growth of established murine and human cell lines, with both cytotoxic, cytostatic and stimulatory effects observed in different experiments.110
These data suggest that adherent, phagocytic cells - possibly macrophages - may release or produce one or more factors vital to clonogenic growth of human tumors.111
Cell-free ascites
Vaage and Agarwal have reported growth stimulation of several murine neoplastic cells not only by serum but even more so by cell-free ascites (CFA).112 Uitendaal etal. also demonstrated
the growth promoting properties of cell-free ascites on ovarian tumors in the clonogenic assay.113
Replacement of the medium supplements (including FCS) by CFA from ovarian cancer patients resulted in increased plating efficiency in the clonogenic assay of ovarian tumor specimens, and this effect was maintained if ascites of other ovarian cancer patients was used. In further experiments by Broxterman et al.,38 CFA consistently enhanced clonogenic growth of ovarian
cancer cell lines more efficiently than a range of sera.
When Hamburger and Salmon first added CFA to the culture medium in the clonogenic assay, they found decreased plating efficiency as compared to BALB/c spleen cell-conditioned media.30 In
addition to growth-stimulating factors, ascites may thus also contain growth inhibiting molecules. The presence of growth inhibiting factors in ascites that co-purified with TGF-a was shown in experiments with rat ascites fluids that suppressed the growth of human colon carcinoma cell lines CBS and MOSER.114 The growth of other cell lines, HCT 116 and HCT C was not inhibited.
These observations are suggestive of a balance between growth promoting and inhibitory factors present in ascites, with a net effect of growth promotion in malignant ascites consistent with the familiar autocrine concept of the role of growth factors in tumorigenesis. The ascitic growth factors have not yet been fully characterized, and their cellular source remains undefined, with both peritoneal macrophages and tumor cells being possible candidates.
In a study of Wilson etal., a comparison was made between the growth promoting properties of ascitic fluids, cyst fluids and peritoneal washings from patients with benign or malignant gynecological conditions. Effects on a variety of tumor- and normal cell lines were studied. All fluids stimulated colony formation in soft agar, and this activity was not tumor or cancer-related. The 'colony stimulating activity' could be inhibited by the addition of heparin and thrombin.115
Evidence for a non-peptide growth factor
Mills et al. provided evidence for the presence of a novel growth factor in ascites. They reported that ascitic fluids from ovarian cancer patients but not from patients with other cancers or benign diseases induced proliferation of fresh ovarian cancer cells and of the ovarian cancer cell line HEY. The proliferative response was associated with rapid increases in phospholipid hydrolysis and free cytosolic calcium. None of a number of well characterized growth regulators such as EGF, PDGF, FGF, thrombin, vasopressin, angiotensin, IL-1, IL-2, interferons, TGF-ß, or TNF-oc, increased intracellular calcium concentrations. Pretreatment of the tested cell lines with these factors did not in any way alter the response to CFA, suggesting that CFA contains hitherto unknown growth promoters that stimulate cells through phospholipid hydrolysis. CFA could completely substitute for human plasma or FCS in these experiments. "6 Subsequent experiments
showed that the human ovarian cancer cell line HEY could grow intraperitoneal^ in immunodeficient nude mice in the presence, but not in the absence of ascitic fluid from ovarian cancer patients. Ascitic fluid from patients with benign disease could not support this growth. After the intraperitoneal tumors progressed, CFA was no longer needed for tumor growth. Eventually the xenografted tumors developed ascites with potent growth factor activity, suggesting an autocrine mechanism.117
These data suggest that CFA contains a growth promoting substance that is different from known polypeptide mediators. Xu and Mills termed the putative growth factor present in ascites from ovarian cancer patients OCAF (ovarian cancer ascites factor) and postulated that it might be a lipid, based on the observations that the growth promoting fraction in ascites is resistant to all known proteases, and stable under such drastic conditions as boiling, pH extremes and detergent treatment that should destroy proteins. It was furthermore soluble in 80% acetone and 100% methanol, though it could be purified by techniques used in protein chemistry such as reverse phase chromatography and isoelectric focusing, probably due to a tight association with albumin or other serum proteins.118 In their analysis of extracted purified OCAF, OCAF
appeared to be a phospholipid, with biochemical characteristics identical to lysophosphatidic acid (LPA), an established bioactive phospholipid with growth promoting properties.
Phospholipid growth factors - lysophosphatidic acid (LPA)
Since the 1980s, it has become increasingly clear that phospholipids do not only have a structural function in biological systems, but are also active in intracellular signalling and can act as growth factors. The best known of these is platelet-activating factor (PAF), a phospholipid that induces a rapid increase in intracellular calcium levels after binding to a high affinity
OH receptor.119120 It has a role in hypersensitivity and inflammatory
Q _ p Q u responses such as asthma and septic shock, but has so far not been ! implicated in cancer growth nor in the control of normal cell y proliferation.121
Q Q Q Lysophosphatidic acid (LPA; monoacylglycerol-3-phosphate) (See
I | fig. 1 for structure) is the simplest naturally occurring glycero-ls phospholipid, that has long been known to be a critical intermediate 0 = C in de novo phospholipid biosynthesis.122
Recently, the broad spectrum of biological activities of LPA has p
been found to include platelet activation, induction of neuronal shape LPA
changes, smooth muscle contraction and stimulation of cell
Figure 1. Structure of proliferation when added to fibroblasts and epithelial cells in
lysophosphatidic acid (LPA).
culture.123 LPA acts on its cognate G-protein coupled cell surface
receptor, which is apparently present in many cell types.124
LPA is a normal constituent of serum, where it is present in albumin-bound form, but it is absent in platelet-poor plasma. Serum contains a factor that co-purifies with albumin and that causes neurite retraction in PC12 cells, inhibits the proliferation of certain tumor cells in vitro, and activates the phophatidylinositol/Ca2+ second messenger system in a variety of cells. There
is evidence that albumin-bound LPA is responsible for at least part of these activities. Synthetically prepared lysophosphatidates reproduce the biological activities of the natural serum factor.125
Kreps etal. demonstrated that serum effects on cyclic AMP accumulation in nine different cell lines were completely mimicked by LPA.126 Jalink etal. reported identical neuronal shape changes
in neuroblastoma cell lines after exposure to LPA or serum.127
LPA is formed and released by activated platelets, but this is probably not the only source of LPA, as there is evidence that LPA is also produced by growth factor stimulated fibroblasts and injured cells.128 It is of interest that earlier studies have demonstrated that platelet lysates can
enhance the growth of some, but not all epithelial and mesodermally-derived tumors in soft agar,129'130 and that a platelet-derived TGF-like compound present in serum, distinct from PDGF,
EGF or insulin can stimulate fibroblast growth in vitro.'13'
The mitogenic action of LPA on fibroblasts in culture is well-documented,132 and occurs in
the absence of synergizing peptide growth factors. The concentration of LPA required for the full mitogenic effect is micromolar as opposed to the nanomolar concentration necessary for neurite retraction and for the other responses mentioned previously. The mitogenic activity critically depends on the length of the fatty acid chain, with 1-oleoyl LPA being most potent and 1-decanoyl LPA showing hardly any mitogenic activity.133
LPA promotes invasion of hepatoma and carcinoma cells into monolayers of mesothelial cells. Serum produced a similar effect, but peptide growth factors could not substitute for serum or LPA in this study.134 Xu et al. found that LPA could stimulate proliferation of the human Jurkat T
cell line in the absence of serum.135 LPA activates breast and ovarian cancer cell lines SKOV-3
and OCC1 in completely serum-free medium,136 as well as OVCAR-3 and Lucy (Jalink K et al.
Unpublished observations.), suggesting that LPA might be the serum-derived growth factor that makes it so difficult to culture cells in serum-free media.
Since Xu ef a/, suggested that LPA might be the potent growth enhancing factor present in ascites,118 both their group and ours have shown the presence of an LPA-like mitogenic activity
in ascites of ovarian cancer patients, and found its levels to be substantially higher than those in
Conclusion
Known peptide growth factors and cytokines have a role in ovarian tumor growth, but they cannot fully account for the growth promoting properties of both serum and malignant effusions of ovarian cancer patients. We postulate that the so far elusive growth promoter might be the bioactive phospholipid LPA, based on the known properties of LPA and on the substitution experiments described above. More data are required to confirm the connection between LPA and ovarian cancer. Further studies should include quantitation of LPA-levels in malignant effusions of ovarian cancer patients compared to other patients, as well as identification of the cellular source of ascitic LPA.
Ovarian cancer remains confined to the peritoneal cavity until late in its clinical course. If LPA in ascites is indeed an important growth factor in this disorder, therapy might be targeted towards inhibiting its activity. The multifunctional growth factor inhibiting agent suramin currently is the only drug available to block LPA activity at the receptor level, but the development of more potent and more specific LPA antagonists may be a viable enterprise.
References
1. Ozols, R.F., Rubin, S.C, Dembo, A.J. & Robboy, S.J. (1992) Epithelial ovarian cancer. In: Hoskins, W.J., Perez, C.A. & Young, R.C. eds. Principles and practice of gynecologic oncology, 1 st edition. Philadelphia, PA: Lippincott, pp.732-781.
2. Moolenaar, W.H., Van der Bend, R.L., Van Corven, E.J., Jalink, K., Eichholtz, T. & Van Blitterswijk, W.J. (1992) Lysophosphatidic acid: a novel phospholipid with hormone- and growth factor-like activities. Cold Spring Harb. Symp. Quant. Biol. 17:163-167.
3. DiSaia, P.J., Morrow, M., Kanabus, J., Piechal, W. STownsend D.E. (1975) Two new tissue cultures lines from ovarian cancer. Gyn. Oncol. 3:215-219.
4. Freedman, R.S., Pihl, E., Kusyk, C, Gallager, H.S. & Rutledge, F. (1978) Characterization of an ovarian carcinoma cell line. Cancer 42:2352-2359.
5. Kanabus, J., Braunstein, G.D., Emry, P.K., DiSaia, P.J. & Wade, M.E. (1978) Kinetics of growth and ectopic production of human chorionic gonadotropin by an ovarian cystadenocarcinoma cell line maintained in vitro. Cancer Res. 38:765-770.
6. Pattillo, R.A , Story, M.T. & Ruckert, A.C.F. (1979) Expression of cell-mediated immunity and blocking factor using a new line of ovarian cancer cells in vitro. Cancer Res. 39:1185-1191.
7. Abu Sinna, G., Beekman, G., Lundgren, E., Nordenson, I. & Roos, G. (1979) Characterization of two new human ovarian carcinoma cell lines. Gyn. Oncol. 7:267-280.
8. Woods, L.K., Morgan, R.T., Quinn, L.A., Moore, G.E., Semple, T.U. & Stedman, K.E. (1979) Comparison of four new cell lines from patients with adenocarcinoma of the ovary. Cancer Res. 39:4449-4459.
9. Hamilton, T.C., Young, R.C, McKoy, W.M. et al. (1983) Characterization of a human ovarian carcinoma cell line (NIH:OVCAR-3) with androgen and estrogen receptors. Cancer Res. 43:5379-5389.
10. Simon, W.E., Albrecht, M., Haensel, M., Dietel, M. & Hoelzel, F. (1983) Cell lines derived from human ovarian carcinomas: growth stimulation by gonadotrophic and steroid hormones. J. Natl. Cancer Inst. 70:839-845.
11. Wilson, A.P. (1984) Characterization of a cell line derived from the ascites of a patient with papillary serous cystadenocarcinoma of the ovary. J. Natl. Cancer Inst. 72:513-521.
12. Buick, R.N., Pullano, R. & Trent, J.M. (1985) Comparative properties of five human ovarian adenocarcinoma cell lines. Cancer Res. 45:3668-3676.
13. Clamon, G., Loh, P-M., White, M. & Robinson, R. (1985) Development and characterization of a human ovarian cancer cell line which clones without agar in vitro. Gynecol. Oncol. 20:92-99. 14. Benard, J., Da Silva, J., de Blois, M-C. et al. (1985) Characterization of a human ovarian
adenocarcinoma line, IGROV1, in tissue culture and in nude mice. Cancer Res. 45:4970-4979. 15. Hill, B.T., Whelan, R.D.H., Gibby, E.M. et al. (1987) Establishment and characterization of three
new human ovarian carcinoma cell lines and initial evaluation of their potential in experimental chemotherapy studies. Int. J. Cancer 39:219-225.
16. Ishiwata, I., Ishiwata, C, Soma, M., Nozawa, S. & Ishikawa H. (1987) Characterization of newly established human ovarian carcinoma cell line: special reference of the effects of cis-platinum on cellular proliferation and release of CA 125. Gynecol. Oncol. 26:340-354.
17. Langdon, S.P., Lawrie, S.S., Hay, F.G. et al. (1988) Characterization and properties of nine human ovarian adenocarcinoma cell lines. Cancer Res. 48:6166-6172.
18. Briers, T.W., Stroobants, P., Vandeputte, T.M. et al. (1989) Establishment and characterization of a human ovarian neoplastic cell line DO-s. Cancer Res. 49:5153-5161.
19. Crickard, K„ Niedbala, M.J., Crickard, U. et al. (1989) Characterization of human ovarian and endometrial carcinoma cell lines established on extracellular matrix. Gynecol. Oncol. 32:163-173. 20. Golombick, T., Dansey, R., Bezwoda, W.R. & Rosendorff, J. (1990) Establishment and
characterization of two new human ovarian cancer cell lines UWOV1 and UWOV2 and a subline UWOV2 (Sf) growing in serum-free conditions: growth characteristics, biochemical and cytogenetic studies. In Vitro Cell. Dev. Biol. 26:447-454.
