This Article Abstract Full Text (PDF) Supplementary Data Publisher's Note Erratum (v25,p3184) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Birrer, M. J. Articles by Mok, S. C. Search for Related Content PubMed PubMed Citation Articles by Birrer, M. J. Articles by Mok, S. C. Social Bookmarking                            What's this?

Whole Genome Oligonucleotide-Based Array Comparative Genomic Hybridization Analysis Identified Fibroblast Growth Factor 1 As a Prognostic Marker for Advanced-Stage Serous Ovarian Adenocarcinomas

Michael J. Birrer, Michael E. Johnson, Ke Hao, Kwong-Kwok Wong, Dong-Choon Park, Aaron Bell, William R. Welch, Ross S. Berkowitz, Samuel C. Mok

From the Cell and Cancer Biology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD; Department of Obstetrics and Gynecology, Gynecologic Oncology; and the Department of Pathology, Brigham and Women's Hospital, Harvard Medical School; Department of Biostatistics, Harvard School of Public Health; Gillette Center for Women's Cancer, Dana-Farber Harvard Cancer Center, Boston, MA; Department of Gynecologic Oncology, University of Texas M.D. Anderson Cancer Center, Houston, TX; and the Department of Obstetrics and Gynecology, Saint Vincent Hospital, The Catholic University of Korea, Suwon, Korea

Address reprint requests to Samuel C. Mok, PhD, Laboratory of Gynecologic Oncology Brigham and Women's Hospital, BLI-447, 221 Longwood Ave, Boston, MA 02115; e-mail: scmok{at}rics.bwh.harvard.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
Purpose To identify markers that can predict overall survival in patients with high-grade advanced stage serous adenocarcinomas.

Patients and Methods Oligonucleotide array comparative genomic hybridization (aCGH) was performed on 42 microdissected high-grade serous ovarian tumor samples. aCGH segments were obtained and a prediction Cox model was built and validated by the standard leave one out analysis. Both DNA and mRNA copy numbers of selected genes located on the candidate aCGH segments were determined by quantitative polymerase chain reaction (qPCR) and quantitative reverse transcriptase PCR (qRT-PCR) analyses. The gene that showed the highest correlation was further validated on an independent set of specimens and was selected for further functional studies.

Results Two chromosomal regions, 4p16.3 and 5q31-5q35.3, exhibited the strongest correlation with overall survival (P < .01). From the 5q31 region, fibroblast growth factor 1 (FGF-1) was selected for further validation study. FGF-1 mRNA copy number was significantly correlated with DNA copy number and protein expression levels (P = .021 and < .001), and both FGF-1 mRNA and protein levels were significantly associated with overall survival (P = .018 and .042). This association was validated for protein expression on an independent set of 81 samples, significant to P = .006. Further studies showed significant correlation between FGF-1 protein expression and CD31+ staining in the tumor stroma (P = .024). Finally, both cancer cells and endothelial cells treated with exogenous FGF-1 showed a significant increase in cell motility and survival.

Conclusion Amplification of FGF-1 at 5q31 in ovarian cancer tissues leads to increased angiogenesis, and autocrine stimulation of cancer cells, which may result in poorer overall survival in patents with high-grade advanced stage serous ovarian cancer.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
Advanced ovarian cancer (stages III and IV) accounted for the majority (> 75%) of the approximately 20,180 new cases of epithelial ovarian cancer in 2006 in the United States.¹ More than 16,000 deaths per year result from ovarian cancer, making this cancer the most lethal gynecologic malignancy. Minimal improvements have been made in overall survival over the past three decades. Ovarian cancer is notable for initial chemotherapy sensitivity using combination platinum and taxane chemotherapy after debulking surgery. However, the vast majority of these women will have their cancer recur within 12 to 24 months after diagnosis and will die of progressively chemotherapy-resistant tumor.²

Prognostic factors for ovarian cancer include stage, grade of the tumor, extent of debulking surgery, and platinum and taxane sensitivity. Debulking surgery has been called an independent predictor for prognosis but many argue that the ability to optimally debulk ovarian cancer to smaller than 1 cm of tumor may reflect an intrinsically biologically more favorable and indolent cancer. Prediction for platinum and taxane sensitivity has not yet been accomplished but is reflected clinically by decrease in serum CA125 during initial chemotherapy, platinum-free interval, and long-term survival. Despite these findings, prognostic biomarkers for ovarian cancer are still lacking.

Multiple comparative genome hybridization (CGH) platforms have been developed, and have been used to identify prognostic markers for ovarian cancer.^3,4 Although BAC array CGH (aCGH) allows for the identification of DNA copy number variation at an increased mapping resolution on a locus-by-locus basis, the identification of specific genes that are involved remains tedious and challenging. cDNA array analysis is an alternative platform which can be used to assess DNA copy number variation on a gene-by-gene basis. Such an approach has been used successfully to identify DNA copy number changes in specific genes in breast cancer,^5,6 ovarian cancer cell lines,⁷ and microdissected ovarian tumor tissue samples.⁴ Taken together, CGH analyses have identified multiple prognostic markers in ovarian cancer. However, most of these markers have limited discriminatory power due to the fact that the number of patients used in the studies is small, and clinical outcomes are influenced by multiple factors, which have not been taken into consideration. This is particularly true in ovarian cancer, which is a heterogeneous type of disease.

In this study, we report the use of a new 60-mer 22K oligonucleotide-based aCGH platform combined with DNA isolated from microdissected tumor tissue samples to identify DNA copy number abnormalities that may be used as prognostic markers for patients with advanced stage serous adenocarcinomas.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
Oligonucleotide Array Preparations
The Human Release 1.0 oligonucleotide library, containing 60-mer oligonucleotides representing 22,000 unique genes were obtained from Sigma-Genosys (Sigma, St Louis, MO). The oligonucleotides were dissolved at 10 µmol/L concentration in 50 mmol/L sodium phosphate buffer pH 8.5 and single spotted onto CodeLink slides, using an OmniGrid 100 microarrayer (Genomic Solutions, Ann Arbor, MI) at the Advanced Technology Center (Rockville, MD).

CGH
Tumor cells from 42 late-stage, high-grade serous adenocarcinomas (including 14 cases used in a previous study⁴) were procured by laser-based microdissection, as previously described.⁴ DNA was extracted, amplified, labeled, and hybridized onto a 60-mer 22K oligonucleotide array platform overnight at 42°C. Scanning and signal quantification were performed as previously described.⁴ Normalized sample to control ratios were used for analysis. Dye swap comparisons were performed to assure no bias in the hybridization process. In addition, female versus female pairs were performed to assure normal distributions of fluorescence for each channel.

Identification of Chromosome Segments Associated With Cancer Survival
Copy number abnormalities (CNA), as well as the boundaries within the genome where these changes occurred, were determined using a changing point algorithm,⁸ implemented in Bioconductor package DNAcopy.

The test for association of the aCGH profile with survival was conducted in two steps. In the first step, the prediction model was trained, where one of 42 samples was reserved for validation and the remaining 41 samples were used in active training. For each aCGH segment, a Cox proportional hazard model (adjusted for debulking status) was fitted, and the two segments with the largest Cox scores were selected as signature associated with survival time. The prediction model was then built. In the second step, the hazard of the one reserved sample based on its aCGH profile and debulking was predicted. The standard leave one out procedure was then conducted by repeating the aforementioned two steps 42 times, where each sample was in turn reserved for validation. To generate the predicted hazard for each patient, and to evaluate the accuracy of our prediction, the samples were equally divided into low- or high-risk group using median predicted hazard as cutoff. The nonparametric log-rank test was used to determine significance.

Segment Validation by Real-Time Quantitative PCR
After statistical analysis of the CGH data, the two chromosomal segments with the highest Cox score were chosen for further analysis. Genes were selected at random from these chromosome segments for analysis by quantitative polymerase chain reaction (qPCR) to assess relative copy numbers changes on a gene-by-gene basis. Sixty-three DNA samples (including the 42 samples used for CGH analysis) were extracted as previously described for CGH analysis, and amplified by GenomiPhi whole genome amplification system (Amersham, Piscataway, NJ) according to the manufacturer's instructions. qPCR was then performed using primers specific for each gene (Table A1, online only) using SYBR green as described.⁹ DNA content was normalized to that of Line-1, a repetitive element for which copy numbers per haploid genome are similar among all normal and neoplastic cells.¹⁰ Relative DNA copy number was determined by normalizing to that of a normal HOSE brushing. Correlations between DNA copy number fold change and survival data were performed using Cox regression and Kaplan-Meier analysis.

Validation of Candidate Gene Expression in Selected Regions by Real-Time Quantitative Reverse-Transcriptase PCR
Genes with strong correlation values between relative DNA copy number and survival were subsequently validated for RNA expression levels by real-time quantitative reverse transcriptase (RT)-PCR analysis in 50 cases (including the 42 samples used for CGH analysis) with RNA extracted from microdissected tumor tissue available. Amplification and cDNA synthesis were performed using the Ovation Aminoallyl RNA Amplification and Labeling System (NuGen Technologies Inc, San Carlos, CA). qRT-PCR was performed using Taqman primers and normalized to cyclophilin (Applied Biosystems, Foster City, CA) as described.⁹ Relative mRNA copy number was determined by normalizing to that of a normal HOSE brushing. Correlations between DNA and mRNA expression levels were determined by Spearman's rho, with a P value of less than .05 considered as significant.

Immunohistochemistry. Immunolocalization of fibroblast growth factor 1 (FGF-1) protein and CD31+ endothelial cells was performed on 50 formalin-fixed paraffin-embedded samples (including the 42 samples used for CGH analysis) using a commercially available anti-FGF-1 polyclonal antibody (R&D Systems, Minneapolis, MN), and an anti-CD31 monoclonal antibody (Dako Cytomation, Carpinteria, CA), respectively. FGF-1 protein expression grading was conducted by determining color intensity by the Image-Pro Plus 5.1 software (Media Cybermetics, Silver Spring, MD) in three different areas of both tumor and stroma at 10x magnification. Intensity scores were calculated as the average difference in staining intensity between the tumor and stroma areas. CD31 staining was localized to blood vessels in and around tumor areas. Grading was performed by manual count of stained blood vessels in the three highest expressing areas of each sample, observed at 20x magnification. Scoring was performed by reviewers blinded to any clinical data associated with the investigation.

Validation Studies
FGF-1 protein expression and DNA copy number were performed on a tissue microarray containing an independent set of 81 samples and on 17 independent DNA samples of high-grade, advanced-stage serous ovarian cancer tissues, respectively.

Effects of Exogenous FGF-1 on Cell Proliferation, Motility, and Survival
Human umbilical vein endothelial cells (HUVEC), as well as ovarian cancer cell lines (UCI101, SKOV3, OVCAR3, 222, CAOV3, A2780), were treated with FGF-1 (10 ng/mL) or control buffer in media containing 10% or 0.1% fetal calf serum for 3 to 5 days. Cell growth or survival was determined by tetrazolium colorimetric assays (MTT assays). Experiments were performed in quadruplicates at each data point. Cell motility assay was performed using the Boyden chamber in the presence of FGF-1 (0, 2.5, 5, and 10 ng/mL).

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
Identification of CNA on Two Chromosome Segments That Are Associated With Survival
A total of 176 unique chromosomal segments showing CNA were identified (Fig 1). The nonparametric log-rank test showed the high- and low-risk groups were significantly different (P < .001) in survival (Figs 2 and 3). Amplification in section 5q31-35 was found to be significantly associated with poor survival (P = .003; Figs 2 and 3). Deletion within the 4p16.3 section was conversely shown to offer a strong protective effect on overall patient survival (P = .003; Figs 2 and 3).

View larger version (14K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 1. Score for normalized chromosome segments. Bar graph showing Cox regression score of individual chromosomal segments against overall patient survival, adjusted for debulking status. Yellow bars denote statistically significant correlation (P < .05).

View larger version (8K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 2. Aberrant chromosomes were investigated using array comparative genome hybridization (aCGH) with printed oligo microarray. Genomic regions on chromosomes 4 and 5 with recurrent gain or loss among cancer tumors were shown. x-axis indicates chromosomal positions of oligo markers; and y-axis indicates the number of patients carrying aberrant chromosomal changes (mean log intensity ratio of all probes on the segment > 0.1 or < –0.1). Segments that are circled indicate areas of interest. For chromosome 4, the circled area relates to a region, which correlated with longer overall survival (> 30 months) in patients with deletion. In chromosome 5, the circled area corresponds to shorter survival (< 30 months) in patients with amplification of the region.

View larger version (11K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 3. Kaplan-Meier plot showing the association between predicted risk and survival. Predicted hazard for patients was based on Cox regression score. Two array comparative genome hybridization (aCGH) segments correspond to 5q31-35 and 4p16.3 with highest Cox score were selected. Samples were equally divided into low- or high-risk group using median predicted hazard as cutoff.

Correlation Between FGF-1 mRNA and Protein Expression Levels and Overall Survival
FGF-1 mRNA and protein expression levels were examined by Cox regression and Kaplan-Meier survival analysis. The results showed that FGF-1 mRNA overexpression was significantly correlated with poor survival as determined by log-rank test (P = .018; Fig 4). In addition, protein expression, adjusted for debulking, also showed a significant correlation with poor survival (HR = 1.20, 95% CI, 1.01 to 1.44; P = .042).

View larger version (13K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 4. Predicted survival, as determined by (A) mRNA and (B) protein expression, by Kaplan-Meier analysis using median expression levels as a cutoff.

FGF-1 Expression Levels Correlated With Microvessel Densities
Since previous studies demonstrated that FGF-1 is a modifier of angiogenesis, we therefore evaluated whether microvessel densities correlated with FGF-1 expression. The result showed a significant positive correlation between the number of CD31-positive blood vessels and FGF-1 protein expression in ovarian tumor tissue samples as determined by Spearman's rho (P = .024; Fig 5). CD31 expression levels also showed a strong correlation with survival, as determined by Cox regression analysis (HR, 1.02; 95% CI, 1.01 to 1.04; P = .009).

View larger version (72K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 5. Immunolocalization of fibroblast growth factor 1 (FGF-1) and CD31 on the same specimen showing high levels of FGF-1 expression and abundant CD31+ blood vessels. For FGF-1, sections were incubated in ABC solution (Vectastain ABC kit; Vector Laboratories, Burlingame, CA) and an alkaline phosphatase-conjugated polymer (EnVision System; Dako Cytomation, Carpinteria, CA), for FGF-1 and CD31, respectively. Sections were then incubated with the Fast Red substrate system (Dako Cytomation), and counterstained using Mayer's hematoxylin (Sigma, St Louis, MO).

Biologic Effects of FGF-1 on HUVEC
To explore the mechanistic basis for the association of FGF-1 amplification and poor survival in patients with advanced stage papillary serous tumors of the ovary, we tested the effects of recombinant FGF-1 on HUVEC biology. Treatment of HUVECs with FGF-1 demonstrated a dose dependent increase in the migration of HUVECs (Fig 6A). Further, treatment with FGF-1 at 10 ng/mL resulted in a 32% increase in cellular proliferation (P < .001) and a 35% increase in cell survival under low serum conditions (P < .001; Figs 6B and 6C). The net result of these biologic effects is consistent with the observed increase in microvessel density observed in tumors with FGF-1 amplification.

View larger version (14K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 6. (A) Effect of exogenous fibroblast growth factor 1 (FGF-1) on human umbilical vein endothelial cells (HUVECs) motility through a polyethylene (PET) membrane in a modified Boyden chamber. Cells that migrated through the pores of the membrane to the bottom chamber after 2.5 hours were stained with hematoxylin for 5 minutes at 37°C. (B) Effect of exogenous FGF-1 on HUVECs growth. Cell number was quantified via MTT assay. The data are the mean ± standard deviation (SD) for three independent experiments. (C) Effect of exogenous FGF-1 on HUVECs survival in 0.2% fetal calf serum (FCS) with control buffer or with purified FGF-1 protein at 10 ng/mL. The data are the mean ± SD for three independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
In this study, we identify FGF-1 as an amplified and overexpressed gene which can serve as a prognostic marker that can be used to predict survival in patients with high-grade, advanced-stage serous ovarian adenocarcinoma using a 22K 60-mer oligonucleotide aCGH platform, and DNA isolated from microdissected tumor tissue samples, combined with multivariate Cox regression analysis.

The spotted 60-mer oligonucleotide array platform used in this study, together with other commercially available oligonucleotide microarray platforms, has been successfully used for aCGH.^12-16 Using this platform, we identified a total of 176 unique chromosomal segments showing CNA in high-grade advanced-stage serous ovarian adenocarcinomas. Multivariate Cox regression followed by standard leave one out procedure identified an amplicon on chromosome segment 5q31-35 associated with patient survival. We did not apply any arbitrary cutoff so that we could directly treat the DNA copy number as a continuous variable in the Cox regression model. Subsequent validation studies using independent sets of cases showed that FGF-1 DNA and RNA copy numbers as well as protein levels significantly associated with survival time. These data suggest that FGF-1 is a candidate prognostic marker for high-grade late-stage ovarian cancer.

FGF-1 or acidic FGF belongs to a family of structurally related fibroblast growth factor polypeptides that have been shown to have mitogenic activities on a broad range of cell types and be to mediators of a wide spectrum of developmental and pathophysiological processes.¹⁷ FGF-1 in particular, has been shown to regulate the growth and migration of the endothelial cells.¹⁸ Because tumor growth and progression depend on angiogenesis,^19-22 we hypothesized that tumor cells secreting high levels of FGF-1 may stimulate endothelial cell growth, and motility of the endothelial cells. Our in vitro data clearly demonstrate increased proliferation, survival, and motility of endothelial cells by physiologic levels of FGF-1. We predict that this biologic activity should produce an increase in microvessel densities and the development of more aggressive ovarian tumors. Our results showed that tumors expressing high levels of FGF-1 indeed had significantly higher CD31-positive microvessel densities, suggesting that FGF-1 secreted by ovarian tumor cells stimulates angiogenesis, which may lead to poor patient survival rates. In fact, miocrovessel density has been shown to be a prognostic marker for different types of tumors.^23-26

Like other FGFs, the effect of FGF-1 is mediated primarily through its binding to FGFRs.²⁷ To date, four FGFRs (FGFR1, FGFR2, FGFR3, and FGFR4) have been shown to bind to the FGF-1 ligand.¹⁷ To evaluate whether FGF-1 secreted by ovarian cancer cells can stimulate ovarian cancer growth, motility, or invasive activity through an autocrine loop, FGFRs expression patterns were determined. The results showed that ovarian cancer cells expressed predominantly FGFR1 and FGFR4, suggesting that secreted FGF-1 may also activate ovarian cancer cells through binding to its receptors on the cancer cell surface. The biologic significance of this autocrine loop is reflected in our in vitro assays, which showed that FGF-1 increased the motility, survival, and proliferation of ovarian cancer cell lines. The observed biologic effects are diverse among the ovarian cancer cell lines and there was no correlation between receptor status and biologic response to FGF-1. This is consistent with the complexity of the FGF ligand where biologic specificity is dependent in addition to ligand, and receptor also downstream signaling factors.²⁷ These biologic effects would likely lead to the development of more aggressive tumors.

The clinical implications of these findings are important. This is one of a very limited number of amplicons identified in ovarian cancer with prognostic value.³ The prognostic stratification of patients with advanced disease has not been possible to date. Further, the identification of a druggable target (within the amplicon), which stratifies patients into prognostic groups, allows for this finding to be incorporated into the design of phase III clinical trials. This is particularly important as antiangiogenesis therapy has recently been shown to be effective in ovarian cancer.

In conclusion, oligonucleotide-based aCGH has identified amplification of FGF-1 on 5q31 as a prognostic marker for high-grade advanced-stage serous ovarian adenocarcinomas. Results from validation and subsequent correlation studies suggest that FGF-1 overexpression may lead to increased angiogenesis, resulting in poorer overall patient survival. These findings imply that FGF-1 may be used as a therapeutic target for ovarian cancer.

	AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
The authors indicated no potential conflicts of interest.

	AUTHOR CONTRIBUTIONS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
Conception and design: Michael J. Birrer, Ke Hao, Kwong-Kwok Wong, Dong-Choon Park, Samuel C. Mok

Provision of study materials or patients: William R. Welch, Ross S. Berkowitz

Collection and assembly of data: Michael J. Birrer, Michael E. Johnson, Ke Hao, Samuel C. Mok, Aaron Bell

Data analysis and interpretation: Michael J. Birrer, Michael E. Johnson, Ke Hao, Kwong-Kwok Wong, Dong-Choon Park, Samuel C. Mok

Manuscript writing: Michael J. Birrer, Dong-Choon Park, William R. Welch, Ross S. Berkowitz, Samuel C. Mok, Michael E. Johnson

Final approval of manuscript: Michael J. Birrer, Michael E. Johnson, Ke Hao, Kwong-Kwok Wong, Dong-Choon Park, William R. Welch, Ross S. Berkowitz, Samuel C. Mok

	Appendix

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 

View larger version (15K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig A1. Expression of fibroblast growth factor-1 (FGF-1) and FGF-1 receptors (FGRFR) FGFR1 and FGFR4 mRNA in human umbilical vein endothelial cells (HUVECs) and microdissected ovarian tumor tissue samples. Quantitative real-time polymerase chain reaction (qRT-PCR) was performed using TaqMan probes specific for FGF-1, FGFR1, and FGFR4, and the 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA) as described by Kim et al (Kim JH, Skates SJ, Uede T, et al: JAMA 287:1671-1679, 2002).

	NOTES

 
Supported in part by Dana-Farber/Harvard Cancer Center Ovarian Cancer SPORE Grants No. P50CA105009, R33CA103595, and M.D. Anderson Ovarian SPORE Grant No. P50CA083639 from National Institutes of Health, Department of Health and Human Services, Gillette Center for Women's Cancer, Adler Foundation Inc, Edgar Astrove Fund, the Ovarian Cancer Research Fund Inc, the Morse Family fund, the Natalie Pihl fund, the Ruth N. White research fellowship, Friends of Dana Farber Cancer Institute, the Robert and Deborah First fund, and the intramural research program of the National Cancer Institute.

Presented in part at the 97th Annual Meeting of the American Association for Cancer Research, April 1-5, 2006, Washington, DC.

M.J.B. and M.E.J. contributed equally to this study.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS Appendix REFERENCES

 
1. Wu X, Groves FD, McLaughlin CC, et al: Cancer incidence patterns among adolescents and young adults in the United States. Cancer Causes Control 16:309-320, 2005[CrossRef][Medline]

2. McGuire WP, Hoskins WJ, Brady MF, et al: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334:1-6, 1996[Abstract/Free Full Text]

3. Cheng KW, Lahad JP, Kuo WL, et al: The RAB25 small GTPase determines aggressiveness of ovarian and breast cancers. Nat Med 10:1251-1256, 2004[CrossRef][Medline]

4. Tsuda H, Ito YM, Ohashi Y, et al: Identification of over-expression and amplification of ABCF2 in clear cell ovarian adenocarcinomas by cDNA microarray analyses. Clin Cancer Res 15: 2005

5. Pollack JR, Perou CM, Alizadeh AA, et al: Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat Genet 23:41-46, 1999[Medline]

6. Pollack JR, Sorlie T, Perou CM, et al: Microarray analysis reveals a major direct role of DNA copy number alteration in the transcriptional program of human breast tumors. Proc Natl Acad Sci U S A 99:12963-12968, 2002[Abstract/Free Full Text]

7. Bourdon V, Naef F, Rao PH, et al: Genomic and expression analysis of the 12p11-p12 amplicon using EST arrays identifies two novel amplified and overexpressed genes. Cancer Res 62:6218-6223, 2002[Abstract/Free Full Text]

8. Olshen AB, Venkatraman ES, Lucito R, et al: Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics 5:557-572, 2004[Abstract]

9. Kim JH, Skates SJ, Uede T, et al: Osteopontin as a potential diagnostic biomarker for ovarian cancer. JAMA 287:1671-1679, 2002[Abstract/Free Full Text]

10. Wang TL, Maierhofer C, Speicher MR, et al: Digital karyotyping. Proc Natl Acad Sci U S A 99:16156-16161, 2002[Abstract/Free Full Text]

11. Bonome T, Lee JY, Park DC, et al: Expression profiling of serous low malignant potential, low-grade, and high-grade tumors of the ovary. Cancer Res 65:10602-10612, 2005[Abstract/Free Full Text]

12. van den Ijssel P, Tijssen M, Chin SF, et al: Human and mouse oligonucleotide-based array CGH. Nucleic Acids Res 33:e192, 2005[Abstract/Free Full Text]

13. Hermsen MA, Tijssen M, Acero IH, et al: High resolution microarray CGH and MLPA analysis for improved genotype/phenotype evaluation of two childhood genetic disorder cases: Ring chromosome 19 and partial duplication 2q. Eur J Med Genet 48:310-318, 2005[CrossRef][Medline]

14. Buffart TE, Coffa J, Hermsen MA, et al: DNA copy number changes at 8q11-24 in metastasized colorectal cancer. Cell Oncol 27:57-65, 2005[Medline]

15. Brennan C, Zhang Y, Leo C, et al: High-resolution global profiling of genomic alterations with long oligonucleotide microarray. Cancer Res 64:4744-4748, 2004[Abstract/Free Full Text]

16. Barrett MT, Scheffer A, Ben-Dor A, et al: Comparative genomic hybridization using oligonucleotide microarrays and total genomic DNA. Proc Natl Acad Sci U S A 101:17765-17770, 2004[Abstract/Free Full Text]

17. Friesel RE, Maciag T: Molecular mechanisms of angiogenesis: Fibroblast growth factor signal transduction. Faseb J 9:919-925, 1995[Abstract]

18. Folkman J: Angiogenesis and breast cancer. J Clin Oncol 12:441-443, 1994[Medline]

19. Folkman J: The role of angiogenesis in tumor growth. Semin Cancer Biol 3:65-71, 1992[Medline]

20. Folkman J, Ingber D: Inhibition of angiogenesis. Semin Cancer Biol 3:89-96, 1992[Medline]

21. Blood CH, Zetter BR: Tumor interactions with the vasculature: Angiogenesis and tumor metastasis. Biochim Biochim Biophys Acta 1032:89-118, 1990[Medline]

22. D'Amore PA: Capillary growth: A two-cell system. Semin Cancer Biol 3:49-56, 1992[Medline]

23. Chavez-Macgregor M, Aviles-Salas A, Green D, et al: Angiogenesis in the bone marrow of patients with breast cancer. Clin Cancer Res 11:5396-5400, 2005[Abstract/Free Full Text]

24. Korkolopoulou P, Thymara I, Kavantzas N, et al: Angiogenesis in Hodgkin's lymphoma: A morphometric approach in 286 patients with prognostic implications. Leukemia 19:894-900, 2005[CrossRef][Medline]

25. Canoglu A, Gogus C, Beduk Y, et al: Microvessel density as a prognostic marker in bladder carcinoma: Correlation with tumor grade, stage and prognosis. Int Urol Nephrol 36:401-405, 2004[CrossRef][Medline]

26. Ribas C, Colleoni GW, Silva MR, et al: Prognostic significance of vascular endothelial growth factor immunoexpression in the context of adverse standard prognostic factors in multiple myeloma. Eur J Haematol 73:311-317, 2004[CrossRef][Medline]

27. Johnson DE, Williams LT: Structural and functional diversity in the FGF receptor multigene family. Adv Cancer Res 60:1-41, 1993[Medline]

Submitted September 6, 2006; accepted March 1, 2007.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				D. Etemadmoghadam, A. deFazio, R. Beroukhim, C. Mermel, J. George, G. Getz, R. Tothill, A. Okamoto, M. B. Raeder, AOCS Study Group, et al. Integrated Genome-Wide DNA Copy Number and Expression Analysis Identifies Distinct Mechanisms of Primary Chemoresistance in Ovarian Carcinomas Clin. Cancer Res., February 15, 2009; 15(4): 1417 - 1427. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplementary Data Publisher's Note Erratum (v25,p3184) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Birrer, M. J. Articles by Mok, S. C. Search for Related Content PubMed PubMed Citation Articles by Birrer, M. J. Articles by Mok, S. C. Social Bookmarking                            What's this?