Abstract
We previously demonstrated that integrin β3 Leu33Pro homozygotes have an increased risk of cancer, possibly most pronounced for ovarian cancer. We now test the latter hypothesis in case-control and prospective studies. We genotyped 463 Danish women with ovarian cancer, and 4291 women from the Danish general population. Calculation of odds ratios by conditional logistic regression was performed in the case-control study (n=463 + 3543), and of ovarian cancer incidence, log-rank statistics and hazard ratios by Cox regression in the prospective study (n=4291) with 9.5-year follow-up. In the case-control study matched for age and marital status, the odds ratio for ovarian cancer in homozygotes versus non-carriers was 1.6 (95% confidence interval: 1.0–2.6). In the prospective study with 28 incident ovarian cancers, non-carriers and homozygotes had incidences of 7 (4–11) and 30 (10–92) per 10 000 person-years (log-rank P=0.02). The age-adjusted hazard ratio for ovarian cancer in homozygotes versus non-carriers was 3.9 (1.1–13). Risk of ovarian cancer did not differ between heterozygotes and non-carriers in either study. Integrin β3 Leu33Pro homozygotes have an increased risk of ovarian cancer.
Introduction
It is thought that genetic factors explain roughly 22% (95% confidence interval: 0–41%) of all ovarian cancer (Lichtenstein et al. 2000). Mutations in BRCA1 and BRCA2 and in the genes involved in hereditary non-polyposis colorectal cancer have been identified as causes of familial aggregation of ovarian cancer; however, these mutations explain only about 40% of the familial cases (Gayther et al. 1999) and only 2–5% of all ovarian cancer (Pharoah & Ponder 2002). Therefore, other, presumably low-penetrant ovarian cancer susceptibility loci probably constitute the main bulk of the genetic contribution to ovarian cancer.
In a recent large prospective study of the general population, we demonstrated that germline integrin β3 Leu33Pro homozygotes have an increased risk of cancer, which in a post hoc explorative analysis appeared to be most pronounced for ovarian cancer (Bojesen et al. 2003). Biologically this seems plausible for the following reasons:
-
β3 integrin expression is higher in ovarian cancer than in benign ovarian tumours (Liapis et al. 1997).
-
In ovarian cancer cells, β3 integrins interact with their ligands vitronectin (Hapke et al. 2003) and periostin (Gillan et al. 2002), altering the intracellular levels of factors essential for the adhesive, migratory and proliferative phenotype of human ovarian cancer cells (Rangaswami et al. 2004).
-
A third ligand of β3 integrins is osteopontin (Panda et al. 1997), recently identified as a potential diagnostic marker for ovarian cancer (Kim et al. 2002); upon binding to β3 integrin, osteopontin releases migratory responses, increased in cells expressing the Pro33 version of β3 integrin compared with the Leu33 version (Sajid et al. 2002).
-
The Pro33 versus Leu33 version of β3 integrins also enhances integrin-mediated activation of mitogen-activated, protein-kinase pathways (Vijayan et al. 2003), crucial for the malignant potential of cancer cells (Johnson & Lapadat 2002).
We therefore directly tested the hypothesis that integrin β3 Leu33Pro homozygosity is associated with an increased risk of ovarian cancer.
First, we tested the hypothesis in a case-control study of 463 Danish women with ovarian cancer compared with 3543 Danish control women from the general population matched for age and marital status. In this study, we also investigated whether increased risk of ovarian cancer among integrin β3 Leu33Pro homozygotes depends on parity and menopausal status. Finally, we re-examined our prospective study of 4291 women from the general population (Bojesen et al. 2003), but in a version improved in several ways, now extending the follow-up for more than 3 years until the end of 2002.
Materials and methods
Case-control study
The subjects were 463 women with invasive ovarian cancer from the population-based MALOVA (MALignant OVArian) study (Høgdall et al. 2003a,b, Glud et al. 2004). The MALOVA study was a multidisciplinary Danish study on ovarian cancer, covering epidemiology, biochemistry and molecular biology. From 1 December 1994 to 31 May 1999, at the gynaecological departments in a certain area of Denmark (municipality and county of Copenhagen, municipality of Frederiksberg, and counties of Roskilde, Western Zealand, Funen, and Southern and Northern Jutland), all available women aged 35–79 years and diagnosed with an ovarian tumour were included. A blood sample for DNA analyses was drawn prior to surgery, and after the operation patients were interviewed concerning medical, marital, menopausal and reproductive history. According to the protocol, women with borderline and benign ovarian tumours were also enrolled in the MALOVA study, but for the present study, only women with invasive ovarian cancer were included. The control group of the case-control study comprised women from the Danish general population in the prospective study who did not develop ovarian cancer (see below), matched with the MALOVA women on age at blood sampling and on marital status (n=3543).
Prospective study
In the 1991–4 examination of the Copenhagen City Heart Study, 5111 women from the Danish general population of Copenhagen participated. More than 99% were white of Danish descent. Questionnaire information included medical, marital, menopausal and reproductive history. A blood sample for DNA analyses was drawn. Women answering ‘yes’ to the question, ‘Did your menstruation cease due to an operation?’ (n=806), were excluded, because we could not be sure whether they had had their ovaries removed and consequently no longer were at risk of developing ovarian cancer. An additional 14 women with a diagnosis of ovarian cancer occurring outside the follow-up period were excluded, leaving 4291 participants for the present prospective study. Diagnoses of ovarian cancer (World Health Organisation (WHO) International Classification of Diseases (7th edn), codes 175.0, 175.1, 175.2, 175.3, 175.5, 176.9, 375.0, 475.0), were obtained from the Danish National Cancer Registry covering the period from 1 January 1946 until 31 December 2002. Follow-up started at the time of blood sampling and ended at death or 31 December 2002, whichever came first.
The re-examination of the prospective study included in the present paper has been improved in the following ways in addition to the extended follow-up:
-
To reduce any bias due to genotype-dependent survival, follow-up now starts at the day of blood sampling in 1991–4, rather than at study entry in 1976–8.
-
Our previous explorative study (Bojesen et al. 2003) used the classification ‘ovarian cancer etc.’ including cancers of the fallopian tube, vagina and vulva according to WHO (Bray et al. 2002), but in the present study only invasive ovarian cancers were included.
-
Risk of ovarian cancer was examined as a function of age, and not as a function of follow-up time as in the previous study (Bojesen et al. 2003).
-
Women who might have had their ovaries removed prior to entry were excluded (n=806).
Ethics approval
The ethics committee of Copenhagen and Frederiksberg approved the prospective study and the MALOVA studies (100.2039/91 and KF01-384/ 95). The studies were also approved by Herlev University Hospital.
Genotyping
Participants were genotyped as described earlier (Zimrin et al. 1990) by genomic DNA isolated from leukocytes from peripheral blood. In short, the Leu33Pro polymorphism is a T→C substitution in exon 3 at position 176 in the integrin β3 gene (GenBank accession no. NM_000212.1), which introduces an MspI recognition site. The assay also included a second MspI recognition site always cleaved, which served as a control site for the digestion reaction. A 268 bp fragment of exon 3 was amplified from genomic DNA with flanking intronic primers, cleaved with MspI, run on a 3% agarose gel and visualized by staining with ethidium bromide. Genotypes were determined independently by an experienced laboratory technician and an author (S E B). Control sequencing confirmed genotype in selected samples.
Statistical analysis
We used the statistical software package STATA (2004). Two-sided probability values of <0.05 were considered significant. Mann–Whitney U test and Pearson’s chi-square test were used. Reference population was Leu33Pro non-carriers. Heterozygotes and homozygotes were each compared with reference.
In the case-control study, genotype specific risk of ovarian cancer was expressed as an odds ratio, using conditional logistic regression. We matched cases with controls for marital status (ever/never married or living with a partner) and age at blood sampling in 2-year strata. Marital status as a matching parameter was chosen to reduce any distortion due to the higher frequency of single-living women without children among the control population from Copenhagen city than among the case population from the whole country. Controls in strata without cases were excluded from the analyses. This resulted in 40 strata with an average of 7.7 controls per case. This model was run unadjusted and adjusted for parity and menopausal status. Interaction between genotype (homozygosity vs non-carrier) and parity/menopausal status was tested by introducing two-factor interaction terms in models already including genotype and parity/ menopausal status.
In the prospective study, we tested differences in ovarian cancer incidence as a function of left truncated age between genotypes by log-rank statistics. Genotype specific risk of ovarian cancer was expressed as a hazard ratio, using a Cox regression model with left truncated age as time scale. As age is the underlying time variable in this model, age is automatically adjusted for. The multifactorial adjusted model also included parity (nulliparous vs parous) and menopausal status (pre- vs postmenopausal) at the examination date.
Results
Characteristics of participants are shown in Table 1. In women from the general population (n=4291), 3018 (70%) were non-carriers, 1157 (27%) heterozygotes and 116 (3%) homozygotes. This genotype distribution was in Hardy–Weinberg equilibrium (P=0.69). Neither in the case-control study nor in the prospective study did age or parous status of cases differ from that of controls; however, in both studies, more cases than controls were postmenopausal (P=0.006 and P=0.03).
In the matched case-control study, the odds ratio for ovarian cancer in homozygotes versus non-carriers was 1.6 (1.0–2.6) (Table 2). Adjustment for parity and menopausal status produced an equivalent odds ratio of 1.6 (0.9–2.7); because these covariates were unknown for 93 cases, statistical power was reduced in this multifactorial adjusted model versus the matched only model. Risk of ovarian cancer in heterozygotes did not differ from that in non-carriers.
We next looked for context-dependent associations between genotype and risk of ovarian cancer in the case-control study (Table 3). Odds ratios for ovarian cancer in homozygous versus non-carriers were 4.6 (1.6–13) and 1.2 (0.6–2.4) among nulliparous and parous women respectively; genotype by parity interaction was borderline significant (P=0.06, not adjusted for multiple comparisons). Menopausal status did not influence risk of ovarian cancer for homozygotes versus non-carriers (Table 3); genotype by menopausal status did not interact in relation to risk of ovarian cancer (P=0.84).
In the prospective study, the total follow-up time was 38 642 person-years with a median follow-up of 9.5 years. Twenty-eight incident ovarian cancer events were recorded corresponding to an incidence of 7 (95% confidence interval: 5–10) per 10 000 person-years. Non-carriers, heterozygotes and homozygotes had incidences of 7 (4–11), 6 (3–13) and 30 (10–92) per 10 000 person-years respectively. The incidence was higher in homozygotes than in non-carriers, but similar in heterozygotes and non-carriers (log-rank P=0.02 and P=0.61 respectively). Age-adjusted and multifactorially adjusted hazard ratios for ovarian cancer in 33Pro/Pro homozygotes versus non-carriers were 3.9 (1.1–13) and 3.9 (1.1–13) (Table 2). Due to lack of statistical power, it was not possible to look for context-dependent associations in the prospective study. Risk of ovarian cancer in heterozygotes did not differ from that in non-carriers.
Discussion
In women from the Danish general population, we observed an increased risk of ovarian cancer in integrin β3 Leu33Pro homozygotes versus non-carriers in both case-control and prospective study designs. The present finding of increased risk of ovarian cancer in Leu33Pro homozygotes agrees with our previous results from an explorative epidemiological study (Bojesen et al. 2003). In that study, the primary hypothesis was that Leu33Pro homozygosity influences risk of all cancer, as it did, while risk of site-specific cancers was examined as post hoc analyses. Although the risk of ovarian cancer, breast cancer and melanoma all appeared to be increased in Leu33Pro homozygotes in those analyses (Bojesen et al. 2003), after correction for multiple comparison, only the increased risk of ovarian cancer remained significant.
In a population-based, age-matched case-control study of 240 ovarian cancer case patients versus 426 age-matched control subjects, Wang-Gohrke and Chang-Claude (2005) observed an odds ratio in integrin β3 Leu33Pro homozygotes versus non-carrier women of 1.2 (0.4–3.7), which at first hand would appear to contrast with our results of an equivalent odds ratio in the case-control study of 1.6 (1.0–2.6) and an equivalent hazard ratio in the prospective study of 3.9 (1.1–13). In contrast to our study, they included borderline tumours, adjusted for first-degree family history of ovarian cancer, and had less statistical power than our case-control study of 463 cases and 3543 controls, so we cannot exclude that their results are compatible with ours (Bojesen et al. 2005). Importantly, however, because the two case-control studies had lower risk estimates than our prospective study, it is likely that the magnitude of the association between integrin β3 Leu33Pro homozygosity and the risk of ovarian cancer is less than that originally observed in our exploratory analysis (Bojesen et al. 2003). Interestingly, Wang-Gohrke and Chang- Claude also observed a higher proportion of Leu33Pro carriers among ovarian cancer patients with adverse prognostic markers than those without, suggesting that the integrin β3 Leu33Pro polymorphism is involved in the metastasis of, and therefore is an indicator of, the malignant potential of ovarian cancer.
Several differently designed studies on ovarian cancer and integrins suggest that this association between integrin β3 Leu33Pro polymorphism and ovarian cancer incidence and/or prognosis may not be a coincidence:
-
In contrast to normal ovaries, expression of β3 integrins is a common feature of epithelial ovarian cancer (Bridges et al. 1995, Cannistra et al. 1995), where they promote proliferation and survival of ovarian cancer cells (Cruet-Hennequart et al. 2003).
-
Integrins are crucial to cell invasion (Hood & Cheresh 2002), and inhibition of integrin β3 function inhibits migration of human ovarian tumour cells (Szaniawska et al. 2001).
-
The Pro33 versus Leu33 allele enhances the activation of intracellular signalling pathways relevant to cancer development (Vijayan et al. 2003) as well as increases adhesive properties of the cells (Feng et al. 1999, Vijayan et al. 2000, Bennett et al. 2001).
In view of all this evidence, we therefore speculate that the adhesive properties of the Pro33 version of β3 integrins increase the probability that premalignant ovarian cells with this genotype have an advantage with regard to migration, survival and/or adhesion to the extracellular matrix, and thereby facilitate neoplastic growth and/or metastasis.
Our studies have some limitations. First, the rarity of 33Pro/Pro homozygosity status limits our statistical power to explore interaction between homozygosity and parity and/or menopausal status in detail. This is certainly the case in the prospective study with only three homozygotes with incident ovarian cancer. In the case-control study, we observed a markedly increased risk of 4.6 (1.6–13) for homozygosity versus non-carriers among nulliparous women, but as this finding included only 6 cases and 15 controls among homozygotes, we cannot exclude that the magnitude of this risk estimate arose by chance.
Second, even though we adjusted for both parity and menopausal status, in the prospective study and in the controls of the case-controls study, we do not have information on some other commonly recognized risk factors for ovarian cancer like duration of oral contraceptive use, duration of hormone replacement therapy, and familial history of ovarian cancer. This is an important limitation since it limits our ability to make inferences of the risk according to genotype in different contexts. However, because of Mendelian randomization (Smith & Ebrahim 2004), it is unlikely that use of oral contraceptives or hormone replacement is associated with genotype. Therefore, it is not likely that adjustment for these factors would significantly change our results on genotype and risk of ovarian cancer. Our lack of information on familial history of ovarian cancer limits our ability to infer the overall genetic make-up of cases and controls. On the other hand, adjusting for family history in studies examining association between genetic factors and disease risk might eliminate the very association examined, because genetic variation is inherited.
Third, we used a case-control analysis and a cohort analysis to test the hypothesis of a previously observed association between integrin β3 Leu33Pro homozygosity and ovarian cancer risk. These two analyses used practically the same controls, since the controls were drawn from our original Danish cohort study. The cohort analysis was based on an extended follow-up of our original Danish cohort. Therefore, these two analyses cannot be strictly considered a totally independent testing of the hypothesis developed in the Danish cohort study. Nevertheless, independent population-based cases were employed for the case-control analysis, and the results suggest that integrin β3 Leu33Pro homozygosity is associated with ovarian cancer risk. However, the magnitude of the association needs to be estimated and confirmed in other population samples.
Despite the fact that ovarian cancer is one of the leading causes of cancer death in women (Ferlay et al. 2004), our current capability of predicting risk of ovarian cancer is mainly limited to estimates based on age, family history of ovarian cancer, nulliparity, long menstrual life, no use of oral contraceptive pills and use of hormone replacement therapy (Glud et al. 2004). Identification of specific genetic variations in genes, such as TP53, progesterone and androgen receptors, STK15, human leucocyte antigens (Agorastos et al. 2004, Agoulnik et al. 2004, Dicioccio et al. 2004, Monos et al. 2005, Terry et al. 2005) or integrin β3, associated with increased risk of ovarian cancer in the general population may contribute to risk-profiling of women with increased risk of ovarian cancer for further preventive procedures. Population-attributable risk (Levin 1953) for ovarian cancer of β3 Leu33Pro homozygosity was estimated to be 8% in the prospective study, with hazard rates of 3.9% and 2% estimated in the case-control study with an odds ratio of 1.6. As other studies have also shown that the frequency of β3 Leu33Pro homozygosity is 2–3% in Caucasian populations in Germany, the USA, France and Italy (Di Castelnuovo et al. 2001), our results are likely to apply to other women in the affluent world. Genotyping for integrin β3 Leu33Pro might be useful not only before ovarian cancer diagnosis, but also for prognostic purposes, perhaps in combination with measurement of its ligand osteopontin (Brakora et al. 2004, Coppola et al. 2004, Schorge et al. 2004). Finally, in the future, genotyping at the time of diagnosis might also assist in drug selection: blockade of integrin β3-mediated signal transduction inhibits growth of human ovarian cancer cell lines (Szaniawska et al. 2001, Cruet-Hennequart et al. 2003), and interaction between integrin β3 and its agonists and antagonists depends on Leu33Pro genotype (Michelson et al. 2000, Boncler et al. 2002, Wheeler et al. 2002, Angiolillo et al. 2004).
In conclusion, in a case-control study and a prospective study, we demonstrated that women homozygous for the Leu33Pro polymorphism of the β3 integrin subunit have an increased risk of ovarian cancer.
Funding
This work was supported by the Danish Medical Research Council, the Danish Heart Foundation, Overlæge Johan Boserup og Lise Boserups Fond, and Copenhagen County. The funding sources are public or non-profit organizations that support science in general. They had no role in gathering, analysing, or interpreting the data or in approving the submitted manuscript. No conflicts of interests prejudicing the impartiality of this paper are declared.
Characteristics of participants
| Case-control study ( n =4006) | Prospective study ( n =4291) | |||||
|---|---|---|---|---|---|---|
| Ovarian cancer cases | Controls | P value | Ovarian cancer cases | Other participants | P value | |
| Values represent median (2.5–97.5 percentiles) or frequencies (95% binomial confidence interval). | ||||||
| Ovarian cancer cases versus controls/other participants on Mann-Whitney U test or Pearson’s chi-square test. | ||||||
| Cases and controls were matched on age and marital status in the case-control study. | ||||||
| Number | 463 | 3543 | 28 | 4263 | ||
| Age (years) | 61 (38–78) | 61 (35–79) | 0.36 | 63 (37–77) | 60 (26–83) | 0.15 |
| Nulliparous (%) | 18 (14–22) | 18 (17–20) | 0.79 | 25 (8–42) | 16 (15–17) | 0.72 |
| Postmenopausal (%) | 83 (79–87) | 77 (75–78) | 0.006 | 86 (72–100) | 66 (65–68) | 0.03 |
Risk of ovarian cancer by integrin β3 Leu33Pro genotype
| Non-carriers 33Leu/Leu | Heterozygotes 33Leu/Pro | Homozygotes 33Pro/Pro | |||
|---|---|---|---|---|---|
| *Additionally adjusted for parity (nulliparous vs parous) and menopausal status (pre- vs postmenopausal). | |||||
| Case-control study | Cases | Controls | Odds ratio (95% confidence interval) | ||
| Matched for age and marital status | 463 | 3543 | 1.0 | 1.1 (0.9–1.4) | 1.6 (1.0–2.6) |
| Matched for age and marital status and multi-factorially adjusted* | 370 | 3528 | 1.0 | 1.0 (0.8–1.3) | 1.6 (0.9–2.7) |
| Prospective study | Events | Participants | Hazard ratio (95% confidence interval) | ||
| Adjustment for age | 28 | 4291 | 1.0 | 0.8 (0.3–2.0) | 3.9 (1.1–13) |
| Multifactorially adjusted* | 28 | 4291 | 1.0 | 0.8 (0.3–2.0) | 3.9 (1.1–13) |
Risk of ovarian cancer by integrin β3 Leu33Pro genotype stratified for other risk factors in the case-control study
| Pro33 allele frequency (%) | Non-carriers 33Leu/Leu | Heterozygotes 33Leu/Pro | Homozygotes 33Pro/Pro | ||||
|---|---|---|---|---|---|---|---|
| Cases/Controls | Cases/Controls | Odds ratio | Cases/Controls | Odds ratio(95%CI) | Cases/Controls | Odds ratio (95%CI) | |
| Cases and controls were matched for age and marital status. CI=confidence interval. | |||||||
| For 56 non-carrier cases and 9 and 6 non-carrier controls, parity and menopausal status were unknown. | |||||||
| For 34 heterozygous cases and 4 and 5 heterozygous controls, parity and menopausal status were unknown. | |||||||
| For 3 homozygous cases and 1 homozygous control, parity and menopausal status were unknown. | |||||||
| Nulliparous | 24.2/15.9 | 40/458 | 1.0 | 20/176 | 1.2 (0.7–2.2) | 6/15 | 4.6 (1.6–13) |
| Parous | 16.4/16.4 | 214/2016 | 1.0 | 80/781 | 0.9 (0.7–1.2) | 10/83 | 1.2 (0.6–2.3) |
| Premenopausal | 15.1/15.5 | 46/591 | 1.0 | 15/213 | 0.7 (0.4–1.4) | 2/21 | 1.3 (0.3–5.8) |
| Postmenopausal | 18.4/16.6 | 208/1886 | 1.0 | 85/743 | 1.1 (0.8–1.4) | 14/77 | 1.5 (0.8–2.8) |
The authors thank Hanne Damm and Mette Refstrup for excellent technical assistance and the participants of the Copenhagen City Heart Study for their willingness to participate. Fruitful discussions with biostatistician Henrik Scharling are appreciated.
References
Agorastos T, Masouridou S, Lambropoulos AF, Chrisa. S, Miliaras D, Pantazis K, Constantinides TC, Kotsis A & Bontis I 2004 P53 codon 72 polymorphism and correlation with ovarian and endometrial cancer in Greek women. European Journal of Cancer Prevention 13 277–280.
Agoulnik IU, Tong XW, Fischer DC, Korner K, Atkinson NE, Edwards DP, Headon DR, Weigel NL & Kieback DG 2004 A germline variation in the progesterone receptor gene increases transcriptional activity and may modify ovarian cancer risk. Journal of Clinical Endocrinology and Metabolism 89 6340–6347.
Angiolillo DJ, Fernandez-Ortiz A, Bernardo E, Alfonso F, Sabate M, Fernandez C, Stranieri C, Trabetti E, Pignatti PF & Macaya C 2004 PlA polymorphism and platelet reactivity following clopidogrel loading dose in patients undergoing coronary stent implantation. Blood Coagulation and Fibrinolysis 15 89–93.
Bennett JS, Catella-Lawson F, Rut AR, Vilaire G, Qi W, Kapoor SC, Murphy S & FitzGerald GA 2001 Effect of the PlA2 alloantigen on the function of beta 3 integrins in platelets. Blood 97 3093–3099.
Bojesen SE, Tybjærg-Hansen A & Nordestgaard BG 2003 Integrin beta 3 Leu33Pro homozygosity and risk of cancer. Journal of the National Cancer Institute 95 1150–1157.
Bojesen SE, Tybjaerg-Hansen A & Nordestgaard BG 2005 Re: Integrin beta3 Leu33Pro homozygosity and risk of cancer. Journal of the National Cancer Institute 97 779–780.
Boncler MA, Golanski J, Paczuski R & Watala C 2002 Polymorphisms of glycoprotein Ib affect the inhibition by aurintricarboxylic acid of the von Willebrand factor dependent platelet aggregation. Journal of Molecular Medicine 80 796–801.
Brakora KA, Lee H, Yusuf R, Sullivan L, Harris A, Colella T & Seiden MV 2004 Utility of osteopontin as a biomarker in recurrent epithelial ovarian cancer. Gynecologic Oncology 93 361–365.
Bray F, Sankila R, Ferlay J & Parkin DM 2002 Estimates of cancer incidence and mortality in Europe in 1995. European Journal of Cancer 38 99–166.
Bridges JE, Englefield P, Boyd IE, Roche WR & Thomas EJ 1995 Expression of integrin adhesion molecules in normal ovary and epithelial ovarian tumors. International Journal of Gynecological Cancer 5 187–192.
Cannistra SA, Ottensmeier C, Niloff J, Orta B & DiCarlo J 1995 Expression and function of beta 1 and alpha v beta 3 integrins in ovarian cancer. Gynecologic Oncology 58 216–225.
Coppola D, Szabo M, Boulware D, Muraca P, Alsarraj M, Chambers AF & Yeatman TJ 2004 Correlation of osteopontin protein expression and pathological stage across a wide variety of tumor histologies. Clinical Cancer Research 10 184–190.
Cruet-Hennequart S, Maubant S, Luis J, Gauduchon P, Staedel C & Dedhar S 2003 alpha(v) integrins regulate cell proliferation through integrin-linked kinase (ILK) in ovarian cancer cells. Oncogene 22 1688–1702.
Di Castelnuovo A, de Gaetano G, Donati MB & Iacoviello L 2001 Platelet glycoprotein receptor IIIa polymorphism PlA1/PlA2 and coronary risk: a meta-analysis. Thrombosis and Haemostasis 85 626–633.
Dicioccio RA, Song H, Waterfall C, Kimura MT, Nagase H, McGuire V, Høgdall E, Shah MN, Luben RN & Easton DF 2004 STK15 polymorphisms and association with risk of invasive ovarian cancer. Cancer Epidemiology Biomarkers and Prevention 13 1589–1594.
Feng DL, Lindpaintner K, Larson MG, Rao VS, O’Donnell CJ, Lipinska I, Schmitz C, Sutherland PA, Silbershatz H & D’Agostino RB 1999 Increased platelet aggregability associated with platelet GPIIIα PlA2 polymorphism – the Framingham Offspring Study. Arteriosclerosis Thrombosis and Vascular Biology 19 1142–1147.
Ferlay J, Bray F, Pisani P & Parkin DM 2004 GLOBOCAN 2002 Cancer incidence, mortality and prevalence worldwide. IARC CancerBase No.5.version 2.0 2004, www-depdb.iarc.fr/globocan/GLOBOframe.htm.
Gayther SA, Russell P, Harrington P, Antoniou AC, Easton DF & Ponder BA 1999 The contribution of germline BRCA1 and BRCA2 mutations to familial ovarian cancer: no evidence for other ovarian cancer-susceptibility genes. American Journal of Human Genetics 65 1021–1029.
Gillan L, Matei D, Fishman DA, Gerbin CS, Karlan BY & Chang DD 2002 Periostin secreted by epithelial ovarian carcinoma is a ligand for alpha(V)beta(3) and alpha(V)beta(5) integrins and promotes cell motility. Cancer Research 62 5358–5364.
Glud E, Kjaer SK, Thomsen BL, Høgdall C, Christensen L, Høgdall E, Bock JE & Blaakaer J 2004 Hormone therapy and the impact of estrogen intake on the risk of ovarian cancer. Archives of Internal Medicine 164 2253–2259.
Hapke S, Kessler H, Luber B, Benge A, Hutzler P, Hofler H, Schmitt M & Reuning U 2003 Ovarian cancer cell proliferation and motility is induced by engagement of integrin alpha(v)beta3/Vitronectin interaction. Biological Chemistry 384 1073–1083.
Høgdall EV, Christensen L, Kjær SK, Blaakær J, Bock JE, Glud E, Nørgaard-Pedersen B & Høgdall CK 2003a Distribution of HER-2 overexpression in ovarian carcinoma tissue and its prognostic value in patients with ovarian carcinoma: from the Danish MALOVA Ovarian Cancer Study. Cancer 98 66–73.
Høgdall EV, Kjær SK, Glud E, Christensen L, Blaakær J, Vuust J, Bock JE, Nørgaard-Pedersen B & Høgdall CK 2003b Evaluation of a polymorphism in intron 2 of the p53 gene in ovarian cancer patients. From the Danish ‘Malova’ Ovarian Cancer Study. Anticancer Research 23 3397–3404.
Hood JD & Cheresh DA 2002 Role of integrins in cell invasion and migration. Nature Reviews. Cancer 2 91–100.
Johnson GL & Lapadat R 2002 Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases. Science 298 1911–1912.
Kim JH, Skates SJ, Uede T, Wong KK, Schorge JO, Feltmate CM, Berkowitz RS, Cramer DW & Mok SC 2002 Osteopontin as a potential diagnostic biomarker for ovarian cancer. Journal of the American Medical Association 287 1671–1679.
Levin ML 1953 The occurrence of lung cancer in man. Acta – Unio Internationalis Contra Cancrum 9 531–541.
Liapis H, Adler LM, Wick MR & Rader JS 1997 Expression of alpha(v)beta3 integrin is less frequent in ovarian epithelial tumors of low malignant potential in contrast to ovarian carcinomas. Human Pathology 28 443–449.
Lichtenstein P, Holm NV, Verkasalo PK, Iliadou A, Kaprio J, Koskenvuo M, Pukkala E, Skytthe A & Hemminki K 2000 Environmental and heritable factors in the causation of cancer – analyses of cohorts of twins from Sweden, Denmark, and Finland. New England Journal of Medicine 343 78–85.
Michelson AD, Furman MI, Goldschmidt-Clermont P, Mascelli MA, Hendrix C, Coleman L, Hamlington J, Barnard MR, Kickler T & Christie DJ 2000 Platelet GP IIIa PlA polymorphisms display different sensitivities to agonists. Circulation 101 1013–1018.
Monos DS, Pappas J, Magira EE, Gaughan J, Aplenc R, Sakkas L, Freedman R, Reveille JD & Platsoucas CD 2005 Identification of HLA-DQalpha and -DRbeta residues associated with susceptibility and protection to epithelial ovarian cancer. Human Immunology 66 554–562.
Panda D, Kundu GC, Lee BI, Peri A, Fohl D, Chackalaparampil I, Mukherjee BB, Li XD, Mukherjee DC & Seides S 1997 Potential roles of osteopontin and alpha Vbeta 3 integrin in the development of coronary artery restenosis after angioplasty. PNAS 94 9308–9313.
Pharoah PD & Ponder BA 2002 The genetics of ovarian cancer. Best Practice and Research. Clinical Obstetrics and Gynaecology 16 449–468.
Rangaswami H, Bulbule A & Kundu GC 2004 Nuclear factor-inducing kinase plays a crucial role in osteopontin-induced MAPK/IkappaBalpha kinase-dependent nuclear factor kappaB-mediated promatrix metalloproteinase-9 activation. Journal of Biological Chemistry 279 38921–38935.
Sajid M, Vijayan KV, Souza S & Bray PF 2002 PlA polymorphism of integrin beta 3 differentially modulates cellular migration on extracellular matrix proteins. Arteriosclerosis, Thrombosis, and Vascular Biology 22 1984–1989.
Schorge JO, Drake RD, Lee H, Skates SJ, Rajanbabu R, Miller DS, Kim JH, Cramer DW, Berkowitz RS & Mok SC 2004 Osteopontin as an adjunct to CA125 in detecting recurrent ovarian cancer. Clinical Cancer Research 10 3474–3478.
Smith GD & Ebrahim S 2004 Mendelian randomization: prospects, potentials, and limitations. International Journal of Epidemiology 33 30–42.
STATA/SE for Windows, rel. 8.2. USA: Stata Corporation 2004.
Szaniawska B, Trembacz H, Miloszewska J, Lipkowski AW, Misicka A, Ostrowski J & Janik P 2001 Peptide analog of fibronectin that inhibits cell migration and ERK 1/2 activity. Peptides 22 1949–1953.
Terry KL, De Vivo I, Titus-Ernstoff L, Shih MC & Cramer DW 2005 Androgen receptor cytosine, adenine, guanine repeats, and haplotypes in relation to ovarian cancer risk. Cancer Research 65 5974–5981.
Vijayan KV, Goldschmidt-Clermont PJ, Roos C & Bray PF 2000 The PlA2 polymorphism of integrin β3 enhances outside-in signaling and adhesive functions. Journal of Clinical Investigation 105 793–802.
Vijayan KV, Liu Y, Dong JF & Bray PF 2003 Enhanced activation of mitogen-activated protein kinase and myosin light chain kinase by the Pro33 polymorphism of integrin beta 3. Journal of Biological Chemistry 278 3860–3867.
Wang-Gohrke S & Chang-Claude J 2005 Re: Integrin beta3 Leu33Pro homozygosity and risk of cancer. Journal of the National Cancer Institute 97 778–779.
Wheeler GL, Braden GA, Bray PF, Marciniak SJ, Mascelli MA & Sane DC 2002 Reduced inhibition by abciximab in platelets with the PlA2 polymorphism. American Heart Journal 143 76–82.
Zimrin AB, Gidwitz S, Lord S, Schwartz E, Bennett JS, White GC & Poncz M 1990 The genomic organization of platelet glycoprotein IIIa. Journal of Biological Chemistry 265 8590–8595.
