home | clinical cancer research - vascular endothelial growth factor levels in ovarian ... · vegf...

Vascular Endothelial Growth Factor Levels in Ovarian Cyst FluidCorrelate with Malignancy1

Darcie Hazelton, Roberto F. Nicosia,2 andSanto V. NicosiaDepartment of Pathology and Laboratory Medicine, MCP HahnemannUniversity, Philadelphia, Pennsylvania 19102 [D. H., R. F. N.],Department of Pathology and Laboratory Medicine, University ofSouth Florida, and H. Lee Moffitt Cancer Center, Tampa, Florida33612 [S. V. N.]

ABSTRACTOvarian cancer is a richly vascularized neoplasm with

solid and cystic components. The purpose of this study was todetermine whether cyst fluid could be used to quantitativelyevaluate production of angiogenic factors in ovarian lesions.ELISA was used to measure vascular endothelial growth factor(VEGF) and basic fibroblast growth factor (bFGF) in the cystfluid of patients with ovarian cancer (n 5 13), benign cysts andcystadenomas (n 5 23), borderline tumors (n 5 5), and func-tional cysts (n 5 8). VEGF levels were markedly elevated in thefluid of malignant cysts (38.56 8.2 ng/ml) as compared withbenign (1.6 6 0.4 ng/ml; P < 0.001), borderline (5.76 1.5ng/ml; P < 0.001), or functional cysts (3.86 2.0 ng/ml; P <0.001). The presence of VEGF in cancer cells was confirmed byimmunohistochemistry. Follow-up of patients with malignantand borderline lesions demonstrated a correlation betweenVEGF levels in cyst fluid and tumor recurrence (P 5 0.03).bFGF in malignant cysts was either undetectable or very low(0.36 0.2 ng/ml), and no significant differences were found inbFGF levels among malignant, benign, borderline, and func-tional cysts. This study demonstrates that ovarian malignancyis associated with dramatic elevation of VEGF levels in ovariancyst fluid. Conversely, there is no correlation between cyst fluidbFGF levels and malignant transformation. The high levels ofVEGF in malignant cysts are consistent with the hypothesisthat this growth factor plays an important role in ovariancancer related-angiogenesis and tumor progression and repre-sents a potentially important target of antiangiogenic therapy.

INTRODUCTIONOvarian cancer is one of the most common gynecological

neoplasms. Each year, 26,000 women develop ovarian cancer,but only 25% of them are diagnosed when the tumor is confined

to the ovary (1). As a result, fifty percent of these patients diewithin five years. There is, therefore, a need for new approachesin the early diagnosis, prognostic evaluation, and treatment ofthis insidious and lethal disease.

During the early stages of tumor progression, cancer cellsacquire the capacity to stimulate angiogenesis and secrete growthfactors that promote endothelial migration, proliferation, capillarytube formation, and extracellular matrix degradation (2). The newlyformed vessels carry oxygen and nutrients to the growing tumor,dispose of its metabolic waste products and secrete factors thatfurther promote the proliferation and invasive behavior of tumorcells through paracrine mechanisms (3, 4). They also function as aport of entry for the spread of metastatic cells to distant sites. Thehypothesis that tumor growth is angiogenesis dependent is partic-ularly relevant for ovarian cancer, which becomes very large inspite of the small size of the organ in which it develops. Histolog-ical studies have confirmed that ovarian cancer is richly vascular-ized and have shown a correlation between microvascular densityand tumor aggressiveness (5).

Because of the importance of angiogenesis in tumor growth,the identification of growth factors responsible for ovarian cancer-related angiogenesis may have important implications for the di-agnostic and prognostic evaluation of this disease. These factorsand their signaling mechanisms may also be potential targets ofadjuvant therapy (6). Among the known endothelial growth factors,VEGF3 has emerged as a critical regulator of the angiogenicprocess (7). VEGF is anMr 32,000–45,000 homodimer producedby a variety of cell types including cancer cells. VEGF has endo-thelial target specificity and is secreted as four alternatively splicedforms, VEGF121, VEGF165, VEGF189, and VEGF210. VEGF pro-motes endothelial migration, proliferation, protease activity, andcapillary tube formation (7). VEGF is also known as vascularpermeability factor because of its potent effect on the permeabilityof postcapillary venules (8). The deletion of even a singleVEGFallele in mice results in defective vasculogenesis and intrauterinedeath of the embryos (9, 10). Similarly, the deletion of genes for theVEGF receptorsflk-1 andflt-1 causes severe vascular defects thatare incompatible with the survival of the mouse embryo (11, 12).VEGF contributes also to angiogenic responses in the adult and hasbeen shown to mediate tumor angiogenesis in experimental ani-mals (13). The safety and efficacy of a humanized anti-VEGFantibody is presently being tested in cancer patients (14).

Because VEGF has been implicated as a regulator of an-giogenesis in ovarian cancer (15), production of VEGF mayreflect the angiogenic activity of this neoplasm. Ovarian cancersgenerate fluid-filled cysts that contain secretory products ofcancer cells. On this basis, we hypothesized that cyst fluid couldbe used to quantitatively evaluate VEGF production in ovarian

Received 10/2/98; revised 12/28/98; accepted 1/4/99.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 This project was supported by NIH Grant HL52585.2 To whom requests for reprints should be addressed, at Department ofPathology and Laboratory Medicine, MCP Hahnemann University, MailStop # 435, Broad and Vine Streets, Philadelphia, PA 19102. Phone:(215) 762-8971; Fax: (215) 762-1741; E-mail: [email protected].

3 The abbreviations used are: VEGF, vascular endothelial growth factor;bFGF, basic fibroblast growth factor.

823Vol. 5, 823–829, April 1999 Clinical Cancer Research

Research. on February 18, 2021. © 1999 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

lesions. In this study, we measured VEGF in ovarian cyst fluidusing a highly sensitive ELISA. We also measured bFGF which,like VEGF, has been proposed as a regulator of tumor angio-genesis (16, 17). Our data demonstrate that malignant ovariancysts have markedly elevated levels of VEGF. Benign ovariancysts have either undetectable or low levels of VEGF, whereasborderline tumors secrete low to intermediate amounts ofVEGF. bFGF levels in malignant cysts are either undetectable orvery low, and no significant differences are found in bFGFlevels among malignant, benign, borderline, and functionalcysts. These findings indicate that VEGF levels in ovarian cystfluid may represent a useful biomarker of angiogenesis andtumor progression. They also support the idea that VEGF playsan important role in ovarian cancer-related angiogenesis andmay represent an important target of antiangiogenic therapy.

MATERIALS AND METHODSFluid Collection. Fluid was collected from ovarian cysts in

the surgical pathology laboratories of the Medical College of Penn-sylvania Hospital (Philadelphia PA) and of the H. L. Moffitt Can-cer Center and Research Institute (Tampa, FL). Each operativespecimen was obtained with the patient’s informed consent. Fluidwas obtained by puncturing the cyst wall with an 18-gauge needlemounted on a 10-ml syringe. If a tumor contained more than onecyst, fluid was collected from the dominant cyst. After collection,the syringe was labeled with the patient name and surgical pathol-ogy number and stored at270°C. Before each assay, the fluidswere thawed and centrifuged at 2000 rpm for 10 min. The super-natants were used for the assay. Excess supernatant fluid wasaliquoted and stored at270°C. Ovarian fluids were evaluated forVEGF, bFGF, and total protein.

ELISA. VEGF and bFGF were measured by captureELISA (R & D Systems, Minneapolis, MN) using a Biotekmicroplate reader. The sensitivity of the assays was,9 pg/mlfor bFGF and,1 pg/ml for VEGF. Fluids were diluted 1:10 to1:100 using dilution buffer and measured in microtiter platesaccording to the manufacturer’s recommendations. Standardcurves were prepared with human recombinant VEGF andbFGF (R & D Systems). Endothelial Cell Growth Supplement(Sigma Chemical Co., St. Louis, MO) was used as a positivecontrol for the bFGF ELISA. Each mg of Endothelial CellGrowth Supplement contained 2.3 ng of bFGF. Total proteinswere measured by the Bradford method using the Bio-RadProtein Assay (Bio-Rad, Burlingame, CA).

All samples were measured without prior knowledge of thediagnosis. Differences among experimental groups were evalu-ated by ANOVA, followed by Student-Newman-Keuls test. Forevaluation of tumor recurrence in patients with malignant cysts,differences were assessed by Student’st test. Statistical signif-icance was set atP , 0.05.

Immunohistochemical Studies. Representative tumorsections mounted on Fisher Plus slides (Fisher) were baked at60°C overnight, deparaffinized in xylene, and rehydrated ingraded ethanols. Endogenous peroxidase activity was quenchedwith 3% hydrogen peroxide in methanol. Antigen unmaskingwas performed in a microwave oven using sodium citrate buffer(pH 6.0). The sections were blocked in a 20% solution of goatserum (Sigma) in PBS and reacted with either primary rabbit

anti-human VEGF antibody (Santa Cruz Biotechnology, Inc.,Santa Cruz, CA) or nonimmune rabbit IgG (Sigma). The anti-bodies were diluted to 0.1mg/ml with 1% goat serum in PBS.To demonstrate the specificity of the VEGF antibody reaction,the primary antibody was incubated overnight at 4°C with thecorresponding peptide (Santa Cruz Biotechnology, Inc.), ac-cording to the manufacturer’s protocol. The neutralized anti-body was then used as a negative control. The sections werethen incubated in biotinylated goat anti-rabbit IgG (Dako,Carpinteria, CA; 1:100 in 1% goat serum/PBS), reacted withavidin-biotin complex (Vector ABC Elite), developed with dia-minobenzidine, and counterstained with Meyer’s hematoxylin.

RESULTSELISA of Ovarian Cyst Fluid. VEGF and bFGF levels

were evaluated in 23 patients with benign cysts, 8 patients withfunctional cysts, 5 patients with borderline tumors, and 13 patientswith malignant tumors. Data on the various patients are listed inTables 1–4. Benign lesions consisted of 9 simple cysts, 11 serouscystadenomas, 1 mucinous cystadenoma, and 1 paratubal cyst. Thisgroup included also one hydrosalpynx, which grossly appeared asan ovarian cyst at the time of fluid collection. Five corpus luteumcysts, two follicular cysts, and one endometriotic cyst were groupedtogether as functional cysts. The term “functional cyst” was used ina broad sense to include cysts composed of tissues that physiolog-ically produce a variety of growth factors including VEGF duringthe female menstrual cycle (7). Borderline lesions included twoserous and three mucinous tumors of low malignant potential.Malignant tumors consisted of nine serous cystadenocarcinomas,

Table 1 Summary of VEGF, bFGF, and total protein levels in thefluid of benign ovarian cysts

Age DiagnosisVEGF(ng/ml)

bFGF(ng/ml)

Totalprotein(mg/ml)

22 Simple serous cyst 0.9 0.001 2.634 Simple serous cyst 8.0 7.0 1.138 Simple serous cyst 1.2 0.7 1.440 Simple serous cyst 0.02 0.0 2.042 Simple serous cyst 0.5 0.0 1.743 Simple serous cyst 0.0 0.001 0.668 Simple serous cyst 0.1 0.0 1.070 Simple serous cyst 5.9 0.0 1.074 Simple serous cyst 0.02 N/A N/A31 Serous cystadenoma 1.0 0.0 1.245 Serous cystadenoma 3.1 0.0 1.056 Serous cystadenoma 0.1 0.1 1.159 Serous cystadenoma 2.1 0.0 2.762 Serous cystadenoma 0.0 0.0 1.169 Serous cystadenoma 0.2 0.0 0.871 Serous cystadenoma 4.2 0.9 2.572 Serous cystadenoma 1.4 0.0 2.673 Serous cystadenoma 1.6 0.0 3.575 Serous cystadenoma 2.1 0.2 2.880 Serous cystadenoma 1.0 0.0 0.955 Mucinous cystadenoma 0.4 0.0 1.250 Hydrosalpynxa 1.2 0.1 3.461 Paratubal cyst 1.2 0.0 1.6

a Lesion grossly appearing as ovarian cyst at the time of fluidcollection.

824 VEGF in Ovarian Cancer



two endometrioid carcinomas, one mucinous cystadenocarcino-mas, and one granulosa cell tumor.

There was a marked difference in VEGF levels betweenmalignant cysts (38.56 8.2 ng/ml) and benign (1.66 0.4ng/ml), borderline (5.76 1.5 ng/ml), or functional (3.86 2.0ng/ml) cysts (Fig. 1). Malignant neoplasms had an average24-fold increase in VEGF over benign lesions and a 6-foldincrease over borderline tumors. VEGF levels in malignanttumors ranged from 7.3 to 97.5 ng/ml. Benign cysts had eitherundetectable or relatively low levels of VEGF. The highestconcentration of VEGF among the nonmalignant lesions wasdetected in a case of cystic endometriosis. Borderline tumorsexhibited an average 3-fold increase in VEGF concentrationover benign lesions. VEGF levels in these tumors, however,were low to intermediate and overlapped with those of theVEGF-producing benign and functional cysts. There was a

slight difference in total proteins between benign and malignanttumors, which, however, could not account for the dramaticallyelevated levels of VEGF in the latter (Fig. 2).

Unlike VEGF, bFGF was nonmeasurable or very low inmalignant cysts except in a serous cystadenocarcinoma (2.5ng/ml). Although small amounts of bFGF were found in themajority of these cysts, bFGF levels did not correlate withmalignancy (Fig. 3). In fact, the highest concentration of bFGF(7 ng/ml) was found in a simple serous cyst.

Immunohistochemical Localization of VEGF. Immu-nostaining of histological sections confirmed the presence ofVEGF at the tissue level. VEGF localized to the cell cytoplasmof tumor cells in both solid and papillary areas (Fig. 4A). VEGFstaining correlated with malignant transformation because be-nign cysts, which were occasionally found next to malignantcells, were lined by a VEGF-negative epithelium. Stromal ele-ments such as fibroblasts, endothelial cells, and vascular smoothmuscle cells exhibited variable immunoreactivity. Plasma cellsand leukocytes were positive. Borderline serous tumors stainedfor VEGF particularly in regions with papillary projections andepithelial tufting (Fig. 4C). Borderline mucinous tumors exhib-ited a distinctively focal pattern of immunoreactivity. In thesetumors, cystic glandular formations lined by stratified atypicalepithelium stained for VEGF, whereas the well-differentiated,mucin-secreting epithelium of adjacent cysts was negative (Fig.4D). The epithelium of benign serous cysts and serous cystade-nomas was either negative or weakly to moderately positive(Fig. 4E). Among the functional cysts, the corpus luteum cystsand an endometriotic cyst showed the strongest staining.

Fig. 1 ELISA of VEGF in ovarian cyst fluid. Malignant cysts containmarkedly elevated levels of VEGF as compared with benign cysts (P,0.001), functional cysts (P, 0.001), and borderline tumors (P, 0.01).Bar, SE.

Table 2 Summary of VEGF, bFGF, and total protein levels in thefluid of functional ovarian cysts


bFGF(ng/ml)

Total protein(mg/ml)

49 Follicular cyst 0.0 0.1 1.672 Follicular cyst 2.1 0.0 2.224 Corpus luteum cyst 6.1 0.1 2.041 Corpus luteum cyst 0.4 0.0 1.627 Corpus luteum cyst 5.1 0.1 2.649 Corpus luteum cyst 0.0 0.1 1.550 Corpus luteum cyst 0.4 0.0 4.436 Cystic endometriosis 16.3 0.0 2.7

Table 3 Summary of VEGF, bFGF, and total protein levels in thefluid of borderline ovarian cysts


bFGF(ng/ml)


76 Borderline serous tumor 4.4 0.1 1.040 Borderline serous tumor 9.8 0.2 6.924 Borderline mucinous tumor 8.5 0.01 1.652 Borderline mucinous tumor 2.7 0.0 2.759 Borderline mucinous tumor 3.0 0.0 2.8

Table 4 Summary of VEGF, bFGF, and total protein levels in thefluid of malignant ovarian cysts


bFGF(ng/ml)


62 Serous cystadenocarcinoma 11.5 2.6 2.648 Serous cystadenocarcinoma 15.1 0.0 2.367 Serous cystadenocarcinoma 51.6 0.1 2.367 Serous cystadenocarcinoma 55.0 0.1 2.671 Serous cystadenocarcinoma 97.5 0.0 2.971 Serous cystadenocarcinoma 70.4 0.1 4.660 Serous cystadenocarcinoma 11.4 0.02 5.463 Serous cystadenocarcinoma 7.3 0.1 4.566 Serous cystadenocarcinoma 12.6 0.01 1.459 Mucinous cystadenocarcinoma 11.8 0.0 0.642 Endometrioid carcinoma 71.5 N/A 3.656 Endometrioid carcinoma 38.4 0.01 2.164 Granulosa cell tumor 46.2 0.05 1.5

825Clinical Cancer Research



Relationship between VEGF Levels and Tumor Recur-rence. Twelve patients with malignant cysts and 5 patientswith borderline cysts were followed up to 3 years after initialsurgery. Patients with malignant cysts were diagnosed at stagesI (n 5 2), III (n 5 7), or IV (n 5 3), whereas patients withborderline cysts were diagnosed at stages I (n5 4) or II (n 5 1).Six patients with malignant cysts had evidence of disease atfollow-up. They included all stage IV patients, two stage IIIpatients, and one patient initially diagnosed at stage I. There wasno evidence of disease at follow-up in patients with borderlinetumors. VEGF levels in the primary tumor of patients withevidence of disease were significantly higher than those ofpatients with no evidence of tumor recurrence (Fig. 5).

DISCUSSIONThe results of this study demonstrate that malignant ovar-

ian cysts contain markedly elevated levels of VEGF as com-pared with benign, borderline, or functional cysts. This findingsupports the idea that VEGF plays an important role in ovarian-related tumor angiogenesis and indicates that high VEGF levelsin ovarian cyst fluid may represent an indicator of malignancyand tumor progression in the ovary. Usingin situ hybridization,Abu-Jawdehet al. (18) found strong expression of VEGF in 29of 29 ovarian cancers and 2 of 8 borderline tumors. They alsomeasured by an immunofluorometric method VEGF in the cystfluid of two cancer patients and found it to be elevated inrelation to benign patients. Paleyet al. (19) observed overex-pression of VEGF byin situ hybridization in 29 of 68 ovariancancers and reported that median disease-free survival for pa-tients with positive VEGF was 22 months, compared with.120months for the VEGF-negative group. Our findings confirm andexpand these observations and establish ELISA as a rapid and

sensitive method for measuring VEGF in the fluid of ovariancystic lesions. The range of VEGF levels found in our seriesmay reflect differences in the angiogenic capacity and, possibly,biological behavior of different ovarian neoplasms.

Dvorak et al. (8) have shown that VEGF is a potentvasopermeability factor. In ovarian cancer, VEGF may causeaccumulation of cyst fluid by increasing the permeability of thetumor microvasculature (20). A similar mechanism may explainthe massive accumulation of ascitic fluid, which follows the i.p.spread of this tumor. This idea is supported by experimentalstudies in guinea pigs that demonstrate a link between VEGFproduction, angiogenesis, increased vascular permeability, andi.p. growth of tumors (21).

VEGF may not only function as an angiogenic and vasoper-meability factor but also as an autocrine factor for the tumor cellsthemselves. Although it is known as an endothelium-specific mi-togen, VEGF may potentially influence the behavior of ovariancancer cells because these cells may express KDR, which is a highaffinity VEGF receptor (22). It is, however, not clear what effects,if any, VEGF may have on ovarian cancer cells. Boococket al.(22)tested the responsiveness of ovarian cancer cell lines to VEGF butwere unable to demonstrate a mitogenic effect. The observationthat VEGF promotesin vitro papillogenesis by rabbit ovariansurface epithelial cells has raised the possibility that this growthfactor may contribute to the morphogenesis of papillae in ovariancancer (23). In addition, a close correlation has been demonstratedbetween VEGF expression and tumor cell proliferation in humanovarian cancer (24).

An interesting result of our study was the demonstration of anaverage 3-fold increase of VEGF levels in borderline tumors, ascompared with benign lesions. Borderline tumors are neoplasms oflow malignant potential. There is, however, a subset of borderline

Fig. 2 Protein assay in ovarian cyst fluid. There is a minimal increasein total protein concentration in malignant cysts.Bars,SE.

Fig. 3 ELISA of bFGF in ovarian cyst fluid. There is no difference inbFGF levels among malignant, benign, borderline, and functional cysts.Bars,SE.




Fig. 4 Photomicrographs of endometrioid carcinoma (A), borderline serous tumor (C), borderline mucinous tumor (D), and simple serous cyst (E)immunostained for VEGF by the ABC immunoperoxidase method.B, endometrioid carcinoma reacted with nonimmune IgG. Note: the carcinomastains strongly for VEGF. Immunoreactivity is observed also in the borderline tumors. In the borderline mucinous tumor, there is a VEGF-positiveatypical gland (arrow) next to a negative benign gland (arrowhead). The epithelium of the simple serous cyst shows no significant immunoreactivity.3350.




tumors that behaves aggressively and progresses to carcinoma (25).It is possible that this process is preceded by a switch of theborderline tumors to an angiogenic phenotype, as described duringtumor progression in experimental animals (26). Our immunohis-tochemical findings that VEGF expression in borderline mucinouslesions is focal and correlates with cellular atypia are compatiblewith this possibility. If the hypothesis of the angiogenic switch isvalidated in ovarian cancer, borderline tumors prone to malignanttransformation may be identified by demonstrating high levels ofVEGF in the cyst fluid. Interestingly, in our series, the VEGF levelsof two borderline tumors were within the range of VEGF levels inmalignant tumors.

Previous studies have demonstrated a correlation betweenangiogenesis, measured by counting microvessels on histologicalsections, and the tendency of malignant tumors to metastasize (27).This observation, although reproduced in many laboratories, hasbeen questioned by some who have not been able to confirm it (28).This discordance may be due to sampling errors, the subjectiveevaluation of histological sections by different observers, or thebiological variability among different tumor types. To that end,VEGF ELISA of ovarian cyst fluid represents an alternative ap-proach for measuring tumor angiogenic activity, which is moreobjective than evaluation of microvascular density.

The lack of correlation between cyst fluid bFGF and ma-lignancy does not preclude the possibility that this growth factoris involved in ovarian cancer-related angiogenesis. Indeed, otherinvestigators have reported the presence of bFGF in this type oftumor (16, 17). These studies, however, were based on immu-nochemical analysis of tissue extracts, and no attempts weremade at measuring bFGF in the fluid phase. The absence or

relatively low levels of bFGF in ovarian cyst fluid may be dueto the fact that bFGF is not secreted through conventionalpathways because it lacks a signal peptide sequence (29). Inaddition, bFGF may become sequestered in the extracellularmatrix due to its heparan sulfate-binding properties (29). Thus,although bFGF is probably involved in ovarian cancer-relatedangiogenesis, its molecular properties may hamper its accumu-lation in the cyst fluid. Our findings that the majority of ovarianmalignancies have low but detectable levels of bFGF are con-sistent with previous reports demonstrating bFGF in cerebrospi-nal fluid (30), serum (31), and urine (32) of cancer patients. Inthese studies, however, bFGF levels of cancer patients werecompared with those of normal individuals. In our study, bFGFlevels in malignant ovarian cysts were compared with those ofborderline, benign, and functional cysts. Interestingly, bFGFwas found in three corpus luteum cysts, one follicular cyst, fourserous cysts, three serous cystadenomas, and one hydrosalpynx.

In future studies VEGF ELISA of ovarian cyst fluid may beused for diagnostic and prognostic applications. Early cysticlesions of the ovary detected by intravaginal ultrasounds or bymagnetic resonance imaging (33) may be tested for VEGFproduction. Measuring VEGF in these cysts could differentiatebenign from malignant lesions. This approach, however, ispresently limited by the concern that puncturing a cyst for fluidcollection may spread a possible malignancy. Alternatively, themalignant nature of these cysts might be assessed by measuringVEGF in the circulation. VEGF ELISA of cyst fluid performedafter surgery may provide information concerning the angio-genic activity and aggressiveness of the tumor. Our observationthat VEGF levels are higher in ovarian cancers that tend to recursupports this possibility and warrants further investigation in alarger prospective series. Recurrence of disease after surgerycould be monitored by measuring serum levels of VEGF, whichare elevated and correlate with tumor burden in ovarian cancerpatients (34). VEGF levels in ovarian cyst fluid may be used toidentify borderline tumors that are prone to malignant behavior.VEGF levels should be interpreted with caution in cases ofendometriosis because they may be significantly elevated inthese patients, as reported previously (35), and as confirmed inour study. In addition, the age of the patient should be taken intoaccount because VEGF regulates the vascularization of thecorpus luteum during the reproductive years (36). However,VEGF levels in the follicular and corpus luteum cysts includedin our study were all below those of the malignant cysts.

More studies are needed to better understand the prognosticsignificance of the increased VEGF levels in ovarian cancer.Particularly important will be to increase the number of obser-vations and determine the range of VEGF concentration in thedifferent types of ovarian cysts. Evaluation of VEGF in thecirculation will establish whether VEGF measurement can beused for early detection of ovarian cancer and for the monitoringof patients after surgery. Finally, it will be important to inves-tigate the biological significance of VEGF production in ovariancancer progression and determine the therapeutic efficacy ofdrugs that specifically target VEGF and its receptor system.

ACKNOWLEDGMENTSWe gratefully acknowledge Dr. Enrique Hernandez, Department of

Obstetrics and Gynecology, Allegheny University of the Health Sciences,

Fig. 5 Cyst fluid VEGF levels and tumor recurrence in patients withmalignant (n5 12) and borderline (n5 5) ovarian lesions. The con-centration of VEGF is significantly higher in patients with evidence ofrecurrent disease (P 5 0.03).Horizontal bars, mean.




Philadelphia, PA, and Dr. James Fiorica, Department of Obstetrics andGynecology, University of South Florida, Tampa, FL, for providing clinicaldata and follow-up on the patients included in this study.

REFERENCES1. Ovarian cancer. Screening, treatment and followup. NIH ConsensusStatement,12: 1–30, 1994.2. Folkman, J. The role of angiogenesis in tumor growth. Semin.Cancer Biol.,3: 65–71, 1992.3. Rak, J., Filmus, J., and Kerbel, R. S. Reciprocal paracrine interac-tions between tumor cells and endothelial cells: “the angiogenesis pro-gression” hypothesis. Eur. J. Cancer,32A: 2438–2450, 1996.4. Nicosia, R. F., Tchao, R., and Leighton, J. Angiogenesis-dependenttumor spread in reinforced fibrin clot culture. Cancer Res.,43: 2159–2166, 1983.5. Hollingsworth, H. C., Kohn, E. C., Steinberg, S. M., Rothenberg,M. L., and Merino, M. J. Tumor angiogenesis in advanced stage ovariancarcinoma. Am. J. Pathol.,147: 33–41, 1995.6. Folkman, J. Fighting cancer by attacking its blood supply. Sci. Am.,275: 150–154, 1996.7. Ferrara, N., and Davis-Smyth, T. The biology of vascular endothelialgrowth factor. Endocr. Rev.,18: 4–25, 1997.8. Dvorak, H. F., Brown, L. F., Detmar, M., and Dvorak, A. M.Vascular permeability factor/vascular endothelial growth factor, micro-vascular permeability and angiogenesis. Am. J. Pathol.,146: 1029–1039, 1995.9. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L.,Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt,C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., andNagy, A. Abnormal blood vessel development and lethality in embryoslacking a single VEGF allele. Nature (Lond.),380: 435–439, 1996.10. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L.,O’Shea, K. S., Powell-Braxton, L., Hillan, K. J., and Moore, M. W.Heterozygous embryonic lethality induced by targeted inactivation ofthe VEGF gene. Nature (Lond.),380: 439–442, 1996.11. Fong, G., Rossant, J., Gertsenstein, M., and Breitman, M. L. Role ofthe flt-1 receptor tyrosine kinase in regulating the assembly of vascularendothelium. Nature (Lond.),376: 66–70, 1995.12. Shalaby, F., Rossant, J., Yamaguchi, T. P., Gertsenstein, M., Wu,X. F., Breitman, M. L., and Schuh, A. C. Failure of blood-islandformation and vasculogenesis in flk-1 deficient mice. Nature (Lond.),376: 62–66, 1995.13. Borgstrom, P., Hillan, K. J., Sriramarao, P., and Ferrara, N. Com-plete inhibition of angiogenesis and growth of microtumors by anti-vascular endothelial growth factor neutralizing antibody: novel conceptsof angiostatic therapy from intravital videomicroscopy. Cancer Res.,56:4032–4039, 1996.14. Presta, L. G., Chen, H., O’Connor, S. J., Chisholm, V., Meng, Y. G.,Krummen, L., Winkler, M., and Ferrara, N. Humanization of an anti-vascular endothelial growth factor monoclonal antibody for the therapy ofsolid tumors and other disorders. Cancer Res.,57: 4593–4599, 1997.15. Nakanishi, Y., Kodama, J., Yoshinouchi, M., Tokumo, K., Kamimura,S., Okuda, H., and Kudo, T. The expression of vascular endothelial growthfactor and transforming growth factor-b associates with angiogenesis inovarian cancer. Int. J. Gynecol. Pathol.,16: 256–262, 1997.16. Crickard, K., Gross, J. L., Crickard, U., Yoonessi, M., Lele, S.,Herblin, W. F., and Eidsvoog, K. Basic fibroblast growth factor andreceptor expression in human ovarian cancer. Gynecol. Oncol.,56:277–284, 1994.17. Di Blasio, A. M., Cremonesi, L., Vigano, P., Ferrari, M., Gospo-darowicz, D., Vignali, M., and Jaffe R. B. Basic fibroblast growth factorand its receptor messenger ribonucleic acids are expressed in humanovarian epithelial neoplasms. Am. J. Obstet. Gynecol.,169:1517–1523,1993.18. Abu-Jawdeh, G. M., Faix, J. D., Niloff, J., Tognazzi, K., Manseau,E., Dvorak, H. F., and Brown, L. Strong expression of vascular perme-ability factor (vascular endothelial growth factor) and its receptors inovarian borderline and malignant neoplasms. Lab. Invest.,74: 1105–1115, 1996.

19. Paley, P. J., Staskus, K. A., Gebhard, K., Mohanraj, D., Twiggs,L. B., Carson, L. F., Ramakrishnan, S. Vascular endothelial growthfactor expression in early stage ovarian carcinoma. Cancer (Phila.),80:98–106, 1997.20. Gordon, J. D., Mesiano, S., Zaloudek, C. J., and Jaffe, R. B.Vascular endothelial growth factor localization in human ovary andfallopian tubes: possible role in reproductive function and ovarian cystformation. J. Clin. Endocrinol. Metabol.,81: 353–359, 1996.21. Yeo, K. T., Wang, H. H., Nagy, J. A., Sioussat, T. M., Ledbetter,S. R., Hoogewerf, A. J., Zhou, Y., Masse, E. M., Senger, D. R., Dvorak,H. F., and Yeo, T. K. Vascular permeability factor (vascular endothelialgrowth factor) in guinea pig and human tumor and inflammatory effu-sions. Cancer Res.,53: 2912–2918, 1993.22. Boocock, C. A., Charnock-Jones, D. S., Sharkey, A. M., McLaren,J., Barker, P. J., Wright, K. A., Twentyman, P. R., and Smith, S. K.Expression of vascular endothelial growth factor and its receptors flt andKDR in ovarian carcinoma. J. Natl. Cancer Inst.,87: 506–516, 1995.23. Pierro, E., Nicosia, S. V., Saunders, B., Fultz, C. B., Nicosia, R. F.,and Mancuso, S. Influence of growth factors on proliferation and mor-phogenesis of rabbit ovarian mesothelial cellsin vitro. Biol. Rep.,54:660–669, 1996.24. Mattern, J., Stammler, G., Koomagi, R., Wallwiener, D., Kauf-mann, M., and Volm, M. Association of vascular endothelial growthfactor expression with tumor cell proliferation in ovarian cancer. Anti-cancer Res.,17: 621–624, 1997.25. Silva, E. G., Kurman, R. J., Russell, P., and Scully, R. E. Sympo-sium: ovarian tumors of borderline pathology. Int. J. Gynecol. Pathol.,15: 281–302, 1996.26. Hanahan, D., Christofori, G., Naik, P., and Arbeit, J. Transgenicmouse models of tumor angiogenesis: the angiogenic switch, its molec-ular controls, and prospects for preclinical therapeutic models. Eur. J.Cancer,32A: 2386–2393, 1996.27. Weidner, N. Intratumor microvessel density as a prognostic factorin cancer. Am. J. Pathol.,147: 9–19, 1995.28. Hall, N. R., Fish, D. E., Hunt, N., Goldin, R. D., Guillou, P. J., andMonson J. R. T. Is the relationship between angiogenesis and metastasisin breast cancer real? Surg. Oncol.,1: 223–229, 1992.29. Mignatti, P., and Rifkin, D. B. Release of basic fibroblast growthfactor, an angiogenic factor devoid of secretory signal sequence: a trivialphenomenon or a novel secretion mechanism. J. Cell. Biochem.,47:201–207, 1991.30. Li, V. W., Folkerth, R. D., Watanabe, H., Yu, C., Rupnick, M.,Barnes, P., Scott, R. M., Black, P., Sallan, S. E., and Folkman, J.Microvessel count and cerebrospinal fluid basic fibroblast growth factorin children with brain tumours. Lancet,344: 82–86, 1994.31. Fujimoto, K., Ichimori, Y., Kakizoe T., Okajima, E., Sakamoto, H.,Sugimura, T., and Terada, M. Increased serum levels of basic fibroblastgrowth factor in patients with renal cell carcinoma. Biochem. Biophys.Res. Commun.,180: 386–392, 1991.32. Nguyen, M., Watanabe, H., Budson, A. E., Richie, J. P., Hayes,D. F., and Folkman, J. Elevated levels of an angiogenic peptide, basicfibroblast growth factor, in the urine of patients with a wide spectrum ofcancers. J. Natl. Cancer Inst.,86: 356–361, 1994.33. Outwater, E. K., Huang, A. B., Dunton, C. J., Talerman, A., andCapuzzi, D. M. Papillary projections in ovarian neoplasms: appearanceon MRI. J. Magn. Resonance Imaging,7: 689–695, 1997.34. Yamamoto, S., Konishi, I., Mandai, M., Kuroda, J., Komatsu, T.,Nanbu, K., Sakahara, H., and Mori, T. Expression of vascular endothe-lial growth factor (VEGF) in epithelial ovarian neoplasms: correlationwith clinicopathology and patient survival, and analysis of serum VEGFlevels. Br. J. Cancer,76: 1221–1227, 1997.35. McLaren, J., Prentice A., Charnock-Jones, D. S., and Smith S. K.Vascular endothelial growth factor (VEGF) concentrations are elevatedin peritoneal fluid of women with endometriosis. Hum. Reprod.,11:220–223, 1996.36. Kamat, B. R., Brown, L. F., Manseau, E. J., Senger, D. R., andDvorak, H. F. Expression of vascular permeability factor/vascular en-dothelial growth factor by human granulosa and theca lutein cells, rolein corpus luteum development. Am. J. Pathol.,146: 157–165, 1995.




1999;5:823-829. Clin Cancer Res Darcie Hazelton, Roberto F. Nicosia and Santo V. Nicosia Fluid Correlate with MalignancyVascular Endothelial Growth Factor Levels in Ovarian Cyst

Updated version

http://clincancerres.aacrjournals.org/content/5/4/823

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/5/4/823.full#ref-list-1

This article cites 32 articles, 4 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/5/4/823.full#related-urls

This article has been cited by 5 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/5/4/823To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/5/4/823.full#ref-list-1

http://clincancerres.aacrjournals.org/content/5/4/823.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



home | clinical cancer research - vascular endothelial growth factor levels in ovarian ... · vegf...

Documents