Expression of steroid hormone receptors in BRCA1-associated ovarian carcinomas

Gynecologic Oncology 97 (2005) 16 – 25 www.elsevier.com/locate/ygyno

Expression of steroid hormone receptors in BRCA1-associated ovarian carcinomas Morteza Aghmesheha,b,c,d,*, Lyndal Edwardse, Christine L. Clarked,f,g, Karen Bythg, Benita S. Katzenellenbogenh, Pamela J. Russella,c, Michael Friedlanderb,d,i, Katherine M. Tuckerb,d,i, Anna de Faziof,g,j a

Oncology Research Centre, Prince of Wales Hospital, Randwick, NSW 2031, Australia Hereditary Cancer Clinic, Prince of Wales Hospital, Randwick, NSW 2031, Australia c School of Medicine, University of New South Wales, Kensington, NSW 1465, Australia d The Kathleen Cuningham Consortium for Research into Familial Breast Cancer (http://www.kconfab.org) e Department of Pathology, Prince of Wales Hospital, Randwick, NSW 2031, Australia f Westmead Institute for Cancer Research, Westmead Hospital, Westmead, NSW 2145, Australia g University of Sydney at Westmead Millennium Institute, Westmead Hospital, Westmead, NSW 2145, Australia h Department of Molecular and Integrative Physiology, University of Illinois, Urbana, IL 61801, USA i Institute of Oncology, Prince of Wales Hospital, Randwick, NSW 2031, Australia j Department Gynaecological Oncology, Westmead Hospital, Westmead, NSW 2145, Australia b

Received 26 February 2004

Abstract Objective. BRCA1 mutations predispose to cancer in hormone responsive tissues. A predominance of estrogen receptor (ER)-negative breast cancers in BRCA1 mutation carriers and potential interactions between ERa and BRCA1 suggest a link between hormones and BRCA1. However, the expression pattern of ERa and other hormone receptors in BRCA1-associated ovarian cancer was unknown. Methods. Twenty-two BRCA1-associated ovarian cancer cases were matched with sporadic cases (no family history of ovarian or breast cancer) for FIGO stage, grade, histologic subtype, and patient age and hormone receptor expression was measured immunohistochemically. Results. ERa expression was similar in BRCA1-associated ovarian cancer compared with matched sporadic counterparts, in contrast with previous findings in BRCA1-linked breast cancer. There was also no significant difference in expression of progesterone receptors and androgen receptor between the matched cases in the two groups. However, differences were noted in the relative expression of receptor isotypes, in particular, levels of ERa and ERh were positively correlated in sporadic tumors but inversely related in BRCA1-associated tumors. Conclusion. Similar hormone receptor expression in BRCA1-associated ovarian cancer and matched sporadic counterparts may be further evidence that at least a proportion of sporadic ovarian tumors and BRCA1-associated tumors develop through similar pathways. D 2005 Elsevier Inc. All rights reserved. Keywords: Estrogen receptor; Progesterone receptor; Androgen receptor; BRCA1; Ovarian carcinoma

Introduction Various functions have been attributed to the BRCA tumor suppressor genes including regulation of cell cycle progression, induction of apoptosis, involvement in DNA * Corresponding author. Oncology Research Centre, Prince of Wales Hospital, Randwick, NSW, 2031, Australia. Fax: +61 2 9382 2629. E-mail address: [email protected] (M. Aghmesheh). 0090-8258/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2004.12.030

repair, and the maintenance of genomic integrity [1]. However, these generic bcaretakerQ activities are not celltype specific and do not explain the association of germline mutations in BRCA1 and BRCA2 with predisposition to cancer in hormone responsive tissues, breast, and ovary. Ovarian hormones are implicated in the aetiology of both breast and ovarian cancers, and there is increasing evidence for a link between steroid hormone action and BRCA1. BRCA1 inhibition of estrogen receptor a (ERa) transcrip-

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

tional activity has been shown in vitro [2,3] and an interaction between the amino-terminal region of BRCA1 (amino acid 1–300) and the conserved carboxyl-terminal activation function (AF-2) domain of ERa has been reported in vivo and in vitro [4]. BRCA1 interacts physically with androgen receptor (AR) resulting in enhanced AR-dependent transactivation in transfected prostate and breast cancer cell lines [5] and BRCA1 has also been reported to play a positive role in androgen induced cell death in prostate cancer cells [6]. In addition, the length of the CAG repeat polymorphism in AR may influence penetrance of breast cancer in BRCA1 mutation carriers [7] providing support for a link between BRCA1 and androgen action. Increased BRCA1 and BRCA2 mRNA expression in mouse mammary tissue during puberty and pregnancy, when circulating hormone levels are high, provides further evidence of potential interactions between BRCA1 and BRCA2 and sex hormones [8]. BRCA1-associated breast cancers are predominantly ERa-negative and progesterone receptor (PR)-negative, in contrast with sporadic disease [9,10] and distinct characteristic histology (high mitotic count, presence of lymphocytic infiltrate, and the presence of bpushing bordersQ) was observed in BRCA1-associated compared with sporadic breast cancer [11,12]. Most studies have reported that papillary serous adenocarcinoma is the predominant ovarian cancer type to occur in BRCA1 mutation carriers and that the tumors are of high grade [13–16]. However, other investigations have reported that these features occurred with similar frequency in control groups [17,18]. It has also been suggested that BRCA1 mutation status may influence patient survival [14,19,20] however this has not been found in all studies [18,21,22]. Therefore, while some studies suggest that ovarian cancer may develop through a different tumorigenic pathway in BRCA1 mutation carriers compared with sporadic cases [23,24], the pathways are not well understood and not all findings agree [25]. The causes of ovarian cancer are not well understood, however there is increasing evidence that steroid hormones play an important role [26]. Many hormone related factors have been either positively or negatively associated with ovarian cancer risk [27,28]. Extended duration of ovulation (e.g. early menarche and late menopause without anovulation due to pregnancy, oral contraceptive use, and breast feeding) and hormone replacement therapy have been positively associated, while ovarian cancer risk is reduced by more live births, long-term breast feeding, and oral contraceptive use [29–31]. Risch hypothesized that estrogen, and more importantly progesterone and androgen have a critical role in the etiology of ovarian carcinoma and that progesterone has a possible protective role in tumorigenesis [26]. Nuclear receptors for estrogen, progesterone, and androgens are expressed in ovarian epithelial tumor cells. The majority of ovarian cancers express ER whereas far fewer tumors express PR [32–37]. Despite high expression of ERa,

17

the response rate to hormonal therapy, such as tamoxifen, in ovarian cancer is low and the lack of PR expression in ERpositive ovarian tumors has been suggested to indicate impaired estrogen action [35]. It has also been suggested that expression of estrogen receptor h (ERh) may contribute to determination of sensitivity to hormonal agents. ERa and ERh have distinct patterns of tissue expression and can exist as homo- or heterodimers that may have different responses to estrogens and antiestrogens [38,39]. ERh is expressed in the normal ovary and most studies have found expression to either be similar or to decrease in ovarian tumors [35,38]. PRA and PRB, two isoforms of PR, are products of a single gene but under the control of distinct promoters [40]. Few studies have measured the two PR isoforms separately in ovarian cancer [32,41]. While there have been several studies, mentioned above, examining expression of hormone receptors in ovarian cancer, there are no data comparing the profile of steroid hormone receptor expression in hereditary and sporadic ovarian carcinomas nor have expression of steroid receptor isotypes been examined. Steroid hormones are thought to play a critical role in the carcinogenic pathway of ovarian cancer, however the role of functional BRCA1 inactivation in this pathway remains unknown. The aim of this study was to determine the relative expression of steroid hormone receptors, ERa, ERh, PRA, PRB, and AR, in human ovarian cancer and to compare the expression in tumors from known BRCA1 mutation carriers compared with dsporadicT ovarian tumors obtained from patients with no family history of breast or ovarian cancer, matched for clinical and histopathological features.

Materials and methods Sample collection Ovarian tumor tissue from 22 pathogenic BRCA1 mutation carriers was collected from various hospitals in Sydney (Prince of Wales, Westmead, Nepean, and Royal Prince Alfred) and from kConFab, the Australasian Consortium for Research into Familial Breast Cancer. The BRCA1 mutations were detected at a number of different Australian molecular genetics laboratories with appropriate accreditation from the National Association of Testing Authorities of Australia (NATA). This group was matched for histological subtype, grade, stage, and patient age with 22 ovarian cancer patients selected for no family history of breast or ovarian carcinoma, termed bsporadic Q cases. The cases were matched by L. Edwards (Gynecological Oncology Pathologist), using the integrated SEALS pathology database at the Royal Hospital for Women, Randwick, NSW. All samples were then deidentified and the study was performed with approval from the institutional Human Research Ethics Committee. Approval was not given to test the bsporadicQ group for BRCA1 mutation and data from consecutive series

18

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

of ovarian cancers suggest that around 10% of apparently sporadic ovarian cancers have BRCA1 mutations [42]. Immunohistochemical staining Paraffin-embedded tissue specimens were sectioned (5 Am), adhered to Mayer’s Egg Albumin (equal volume of egg white and glycerol) coated Superfrost Plus slides (Menzel-Glaser, Germany), and stained for ERa, ERh, PRA, PRB, and AR using routine immunohistochemical techniques optimized for each antibody. All incubations were at room temperature unless otherwise specified. PRA antigen was retrieved by incubation of sections for 2 min and 30 s at high power followed by 10 min incubation at 40% of power in an 800 W microwave. The other antigens were retrieved by autoclaving as previously described [43]. Endogenous peroxidase activity was quenched in 1.0% H2O2/PBS and non-specific protein binding was blocked in 0.3% casein/0.5% Tween 20/PBS. The sections were incubated with primary antibody in a humid chamber. The primary antibodies used for immunostaining were as follows: For detection of ERa a monoclonal antibody, NCLER-6F11 obtained from NovoCastra (New Castle upon Tyne, UK) was used at 1:50 dilution and slides were incubated at room temperature for 1 h ERh CWK-F12 (2.5 Ag/ml, overnight incubation at 48C) monoclonal antibody raised against the hormone-binding domain of human ERh [44]. PRA (NovoCastra, Clone 16, catalogue number NCL-PGR-312, 1:200 dilution, 1 h incubation) and PRB (hPRa6, 1:50 dilution, overnight incubation at 48C) [45] were detected using monoclonal antibodies specific for each isoform [43]. NCL-AR-318, a mouse monoclonal antibody recognizing the N-terminus of the human androgen receptor, obtained from NovoCastra was used at 1:50 dilution and incubated overnight at 48C. All primary antibodies were diluted in 1% BSA/PBS. Antigens were visualized after incubation with biotinylated secondary antibodies (anti-mouse IgG and anti-goat IgG raised in horse and anti-mouse IgG raised in goat (Vector Laboratories, Burlingame, CA)) and color development (DAB liquid substrate–chromagen system, DAKO, Carpinteria, CA). The sections were counterstained with hematoxylin. Analysis of hormone receptor expression The expression of hormone receptors in normal and malignant ovarian tissues was heterogenous both in the level of expression (intensity) and the proportion of tumor cells stained (extent). Therefore, both intensity and extent of staining in each section were considered using a semi-quantitative scaling system [46]. The whole section of each case was examined by high power (400) light microscopy and assigned a modified histoscore (MH) from 0 to 4 as following: 0. No nuclear staining 1. Focal, weak nuclear staining of tumor cells

2. Strong nuclear staining of b25% or moderate staining of less than or equal 80% tumor cells 3. Strong nuclear staining of 25–50% or moderate staining of N80% tumor cells 4. Strong nuclear staining of N50% tumor cells A case with a score of 2 or higher was considered positive. The slides were coded and evaluated in a bblindedQ fashion by at least two independent observers. Statistical analysis Conditional logistic regression analysis was used to identify any potential difference in expression of hormone receptors between BRCA1-associated and sporadic ovarian carcinomas. Chi-squared test was used for analysis of the association between steroid receptors expression (positive or negative), corrected for small cell size where it was appropriate (Fisher’s Exact Test). Kendall’s correlation test was used to quantify the rank correlation between steroid receptor expression levels within each group. In a comparison of breast cancer in sporadic cases and BRCA1-mutation carriers, ERa was expressed in 65% of sporadic cases and in only 10% of BRCA1-associated cases [12]. If these distributions were similar for ovarian cancer, then a sample size of 22 per group has over 90% power to detect a statistically significant difference assuming a 5% significance level in a two-sided test (nQuery Advisor version 4.0).

Results The study group comprised 44 epithelial ovarian cancers, 22 from women with a confirmed BRCA1 mutation and 22 from women with no family history of breast or ovarian carcinoma, termed bsporadicQ ovarian carcinomas. Of the entire study group with tumor stages ranging from II to IV, the majority were serous adenocarcinomas, stage III (75%, 33/44) and all were of moderate to high grade (Table 1). The two study groups were matched for tumor stage, grade, histologic subtype, and patient age. Mean age of diagnosis in hereditary and sporadic groups was 46.7 (a range of 35–67 years old) and 50.2 (41–67 years old), respectively. Since age was a criterion for matching, the age distribution of our sporadic group was younger than sporadic ovarian cancers reported in other studies [47]. The majority of the BRCA1 mutations were truncating mutations with stop codons ranging from amino acid 39 to 1829 and all have been previously cited in the Breast Cancer Information Core (B.I.C) database (http://research.nhgri.nih.gov/bic/). Estrogen receptor a Relatively high expression of ERa was observed in most of the ovarian carcinoma specimens examined in this study (Fig. 1A). The staining pattern was heterogenous between the

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25 Table 1 Clinical characteristics of patient cohorts

19

and sporadic ovarian carcinomas in ERh positivity (P value = 0.5, OR = 0.7, 95%CI = 0.2–2.4).

No. of cases Stage I II III IV Grade 1 2 3 Histological subtypes Serous Mixed mullerian Serous + endometroid Serous + clear cell Age 30–39 40–49 50–59 60–69

BRCA1-associated

Sporadic

Total (%)

Progesterone receptor A

0 3 16 3

0 3 17 2

0 6 (14) 33 (75) 5 (11)

0 8 14

0 9 13

0 17 (39) 27 (61)

20 1 1 0

20 1 0 1

40 2 1 1

(91) (5) (2) (2)

4 10 7 1

0 12 8 2

4 22 15 3

(9) (50) (34) (7)

PRA staining was localized in the cell nuclei and was largely heterogenous (Fig. 1C). PRA was expressed in 36% (8/22) of sporadic but only 9% (2/22) of the BRCA1associated cases, and the difference between matched BRCA1-associated and sporadic ovarian carcinomas approached statistical significance (P value = 0.08, OR = 0.3, 95%CI = 0.1–1.2). There was no strong nuclear staining (score 3 or 4) in the hereditary ovarian carcinomas, and few (2/22) in the sporadic cases while the positive control (human myometrium) and internal positive control (epithelial cells of the fallopian tube) when present, were strongly positive.

Distribution of stage, grade, histologic subtype, and patient age in 22 BRCA1-associated and 22 sporadic ovarian carcinomas.

cells from the same tumor section. Staining was mainly confined to the nucleus, however weak cytoplasmic staining was also observed in some cases. Staining was considered as positive only if nuclear staining was observed. Human ovarian surface epithelial cells, where present, expressed this receptor weakly. ERa was expressed in most ovarian carcinomas and the proportion of positive cases was the same in BRCA1associated (77%; 17/22) and sporadic ovarian carcinomas (Table 2). Univariate conditional logistic regression analysis of cases matched for tumor stage, grade, histologic subtype, and patient age showed no statistically significant difference between hereditary and sporadic ovarian carcinomas in ERa expression determined by modified histoscore (P value = 1.0, odds ratio (OR) = 1.0, 95% confidence interval (CI) = 0.2–5.0) (Table 2). Estrogen receptor b There has been some discrepancies in reports of immunohistochemical detection of ERh and controversy regarding the specificity of some commercially available ERh antibodies for application on paraffin-embedded tissues [48]. The CWK-F12 antibody against the hormone-binding domain of human ERh has been shown to specifically detect ERh in reproductive tissues [44]. In this study, normal ovary was used as positive control and specific nuclear staining was detected in the granulosa cells of the ovary. ERh expression was identified in only 18% (4/22) of BRCA1-associated and 27% (6/22) of sporadic ovarian carcinomas (Table 2) (Fig. 1B). There was no statistically significant difference between matched BRCA1-associated

Progesterone receptor B PRB staining was localized in the nuclei (Fig. 1D). None of the cases expressed PRB strongly (scores 3 or 4). PRB was expressed in 32% (7/22) and 23% (5/22) of hereditary and sporadic ovarian carcinomas, respectively and was not different between the two groups (P value = 0.5, OR = 1.7, 95%CI = 0.4–7.0). Androgen receptor AR expression was identified in the nuclei (Fig. 1E) and the staining pattern was very heterogenous (from one area of the tumor section to another). The level of AR expression was similar between the two matched study groups. AR was expressed in 27% (6/22) and 32% (7/22) of BRCA1associated and sporadic ovarian carcinoma, respectively (P value = 0.8, OR = 0.9, 95%CI = 0.3–2.6). Association of steroid receptor expression and BRCA1 mutation ERa has been shown to interact with the amino-terminus of BRCA1 (amino acid 1–300). The pattern of hormone receptor expression was therefore compared with specific BRCA1 mutations. However, no clear association was found between BRCA1 mutation (N-terminal or C-terminal of BRCA1) and hormone receptor expression. Indeed, ovarian tumors from different individuals with the same BRCA1 mutation showed divergent patterns of hormone receptor expression. Cases 11 and 14 both have a 3875– 3878 del GTCT mutation resulting in a premature stop codon at 1262: however, case 11 was ERa and AR-positive, whereas case 14 is only positive for ERh. Similarly, cases 28, 59, and 30; and 1 and 98 share the same mutation but have different receptor expression profiles. Only one mutation, 3450–3453 del CAAG (stop 1115) showed two cases (case 22 and 33) with the same receptor expression,

20

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

Fig. 1. Immunohistoochemical staining of steroid hormone receptors. Paraffin-embedded ovarian cancer tissue sections were stained with antibodies to (A) ERa, (B) ERh, (C) PRA, (D) PRB, and (E) AR. Immunoreactivity was visualized using DAB resulting in brown staining and the sections were counterstained with hematoxylin (blue staining) (original magnification 400).

i.e. ERa-positive only, however this is likely to have occurred by chance as it was a common phenotype. Correlation of expression of estrogen receptors, progesterone receptors, and androgen receptor In most estrogen target tissues, including the breast, expression of PR is known to be transcriptionally regulated by estrogen and expression of PR and AR is tightly associated with ERa [49,50]. However, in ovarian cancer, particularly serous ovarian cancer, a strong association between ERa/ERh and PR expression has not been

observed thus far [34,35]. Notwithstanding, BRCA1 can interact with steroid hormone receptors, inhibiting ERdependent transcriptional activity [2,3] and enhancing ARdependent transcription [5], and there may be further, as yet uncharacterized, interactions between BRCA1 and hormone receptors. Therefore, the correlation between ER isoform expression levels, PR isoforms, and AR were evaluated separately in BRCA1-associated and sporadic ovarian cancer groups. Modified Histoscores for each of the receptors were ranked and analyzed using Kendall’s rank correlation test. In addition, cases were defined as positive or negative (based on Modified Histoscore: 0 and 1 were

Table 2 Expression of steroid hormone receptors in BRCA1-associated and matched sporadic ovarian carcinomas

ERa ERh PRA PRB AR

BRCA1-associated (n = 22)

Sporadic (n = 22)

% (Positive cases/total)

% (Positive cases/total)

77 18 9 32 27

77 27 36 23 32

(17/22) (4/22) (2/22) (7/22) (6/22)

(17/22) (6/22) (8/22) (5/22) (7/22)

P valuea

Odds ratioa

95%CIa

1.0 0.5 0.08 0.5 0.8

1.0 0.7 0.3 1.7 0.9

0.2–5.0 0.2–2.4 0.1–1.2 0.4–7.0 0.3–2.6

Receptor expression was measured by immunohistochemistry and results are based on a modified histoscore. Univariate conditional logistic regression analysis of BRCA1-associated cases and sporadic cases matched for tumor stage, grade, histologic subtype, and patient age.

a

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

21

defined as negative, z2 was defined as positive) and analyzed using the chi-squared test. Correlation between ER and PR No significant associations were observed between ER and PR isoform expression in either the BRCA1-associated or sporadic groups (Table 3) and the relationship between ERa and total PR was found to be similar between the two groups. ERa+/PR+ was detected in 32% (7/22) of BRCA1associated and 41% (9/22) of sporadic ovarian cancers. The predominant phenotype was ERa+/PR (46% in BRCA1linked and 36% in sporadic group) and the least common was ERa /PR+ (5%, 1/22) in BRCA1-associated and ERa /PR (9%, 2/22) in sporadic ovarian carcinomas. Correlation between ER and AR AR expression was identified only in the presence of ERa in both BRCA1-associated and sporadic ovarian cancers. The correlation between expression of ERa and AR reached statistical significance only in BRCA1-associated tumors by Kendall’s rank correlation (P value = 0.005). There was a trend toward the association of AR and ERa positivity in sporadic ovarian cancers by chi-squared test (P value = 0.08).

Fig. 2. Relationship between ERa and ERh expression in BRCA1associated and sporadic ovarian carcinomas. ERa and ERh were measured by immunohistochemistry and staining was scored using a modified histoscore. Cases with a modified histoscore of 2 or higher, were considered positive. (A) The correlation between the level of expression of ERa and ERh in BRCA1-associated and (B) sporadic ovarian carcinomas. (C) The proportion of positive cases expressing ERa only, is represented by the open bars, ERa and ERh black bars, ERh only hatched bars (Fisher’s Exact Test P = 0.02, between BRCA1-associated and sporadic ovarian carcinomas).

Correlation between ERa and ERb BRCA1-linked and sporadic ovarian tumors were similar in that the majority expressed ERa and few expressed ERh, however the relationship between expression of ERa and ERh was found to be different between the two groups of tumors. ERa was inversely correlated with expression of ERh in BRCA1-associated ovarian cancer (Fig. 2A) (Kendall’s rank correlation test, P value = 0.09, Table 3; chi-squared test, P value = 0.006). Whereas, in sporadic ovarian cancer, the opposite was observed, ERa expression was positively associated with expression of ERh (Fig. 2B) (Kendall’s rank correlation test, P value = 0.05, Table 3; chisquared test, P value = 0.1). This was reflected in the co-expression of ERa and ERh, which was significantly different between BRCA1-associated and sporadic cases (Fisher’s Exact Test P = 0.02, Fig.

2C). Few ER-positive BRCA1-associated cases expressed both ERa and ERh (5%, 1/20) compared with sporadic cases (35%, 6/17) and the ERa-negative: ERh-positive phenotype was only seen in BRCA1-associated cases. Expression of PRA and PRB BRCA1-associated cases that were PR-positive expressed predominantly PRB (6/8, 75%) in contrast with

Table 3 Rank correlation of ERa, ERh, PRA, PRB, and AR in BRCA1-associated and sporadic ovarian carcinomas BRCA1-associated ERa ERa ERh PRA PRB AR

1 0.3 0.2 0.3 0.5

Sporadic ERh

P P P P

= = = =

0.09 0.3 0.1 0.005b

1 0.2 P = 0.4 0.2 P = 0.4 0.3 P = 0.2

PRA

1 0.2 P = 0.4 0.4 P = 0.08

PRB

1 0.3 P = 0.07

AR

1

ERa 1 0.4 P = 0.05a 0.3 P = 0.2 0.14 P = 0.5 0.3 P = 0.2

ERh 1 0.08 P = 0.7 0.2 P = 0.2 0.2 P = 0.5

PRA

1 .06 P = 0.8 0.05 P = 0.8

PRB

AR

1 0.2 P = 0.4

1

Expression levels of hormone receptors were measured by immunohistochemistry and evaluated by modified histoscore. Kendall’s rank correlation test was applied separately in the two ovarian cancer groups. Within group rank correlations between expression level of steroid receptor protein tested by Kendall’s rank correlation. Kendall correlation coefficient, T(0 to F1), with the corresponding P values. a Significant positive correlation, P b 0.05 (2-tailed). b Significant positive correlation, P b 0.01 (2-tailed).

22

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

sporadic cases, which expressed either PRA (7/12, 58%) or PRB (4/12, 33%). Few tumors, either BRCA1-associated or sporadic, expressed both PRA and PRB (Fig. 3). However, the overall number of PR-positive cases was low and the difference in PRA and PRB expression between sporadic and BRCA1-associated cases did not reach significance (Fisher’s Exact Test P = 0.09).

Discussion To our knowledge, this is the first study to investigate the profile of steroid hormone receptors in BRCA1-associated ovarian carcinomas. BRCA1-associated cases were matched with sporadic ovarian cancer for clinical criteria that may influence steroid receptor expression such as histological subtype, tumor stage, and grade. The expression of individual hormone receptors was similar in BRCA1associated compared with sporadic ovarian cancers, unlike the case in BRCA1-associated breast cancer [9,10]. Our data showing similar expression of individual steroid receptors between BRCA1-associated and sporadic ovarian carcinomas is supported by data from cDNA microarray analysis showing that expression profiles of sporadic ovarian tumors overlapped with BRCA1 and BRCA2linked tumors [25]. Furthermore, established risk modifying factors such as oral contraceptive use and tubal ligation apply to both BRCA1-linked and sporadic ovarian cancer [51,52] suggesting that they develop through similar carcinogenic pathways. Together, these data suggest that in at least a proportion of sporadic cases the carcinogenic pathways may result from somatic or epigenetic aberrations of BRCA1 or BRCA2 or their downstream effectors. Somatic BRCA1 mutations are found in around 10% of all ovarian cancers [53–55] and transcriptional silencing of BRCA1 expression due to promoter hypermethylation has been reported [56,57]. These data combined suggest that

BRCA1 is dysfunctional or non-functional in at least a proportion of sporadic ovarian cancers. Indeed, reduced or loss of BRCA1 protein expression in the majority of sporadic ovarian cancers has been previously reported [58,59]. Loss of BRCA1 protein expression was detected in over 80% of the sporadic cases in the cohort described in the present study (publication in preparation). This evidence suggests that BRCA1 dysfunction may occur in a high proportion of sporadic ovarian cancers and may explain the similarities found between BRCA1-linked and sporadic ovarian cancers. Relative levels of expression of ERa and PR in ovarian cancer Our study found that PR expression was low in both BRCA1-associated and sporadic cases. Auersperg et al. [60] showed that progesterone receptor mRNA is expressed in human ovarian surface epithelial cells and down-regulated in ovarian cancer cells. Most previous studies have found low PR expression in serous ovarian tumors [33,61] which is in agreement with our results. Lau et al. found loss of expression of PR transcripts in most ovarian cancer cell lines and primary cultures and suggested that loss of PR mRNA expression may contribute to neoplastic transformation in ovarian epithelial cells [35]. A positive association between ERa and PR has been well characterized in breast cancer and has also been reported in ovarian cancer. However, in ovarian cancer, the relationship between ER and PR, while significant in endometrioid ovarian cancer, is not seen in serous ovarian cancer [34]. The ovarian tumors in this study were predominantly serous ovarian cancer, the expression of PR was low and no significant correlation was found between ERa and PR, which are consistent with previous studies [34]. Relative levels of expression of ERa and AR in ovarian cancer

Fig. 3. Expression of PRA and PRB in BRCA1-associated and sporadic ovarian carcinomas. PRA and PRB expression was measured by immunohistochemistry and staining was scored using a modified histoscore. Cases with a modified histoscore of 2 or higher were considered positive. The proportion of positive cases expressing PRA only is represented by the black bars, PRA and PRB, open bars, PRB only, hatched bars.

While some AR immunoreactivity was observed in majority of cases, significant AR expression was only observed in ~30% of the total cohort and this was concordant with previous findings [37]. A positive correlation between ERa and AR has been reported in both breast and ovarian cancers and cancer cell lines [32–34]. This study further demonstrated a significant correlation between ERa and AR in ovarian cancer. AR expression was only observed in ERa-positive ovarian cancers and although the positive association between ERa and AR was statistically significant only in the BRCA1-associated group, the trend was the same in both groups. BRCA1 interacts in vitro with ERa and AR and may have an important role in receptor signaling, however we found no evidence that the relationship between ERa and AR expression was disrupted in BRCA1-associated cancers compared with sporadic disease.

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

Relative levels of expression of ERa and ERb in ovarian cancer The majority of tumors in both study groups expressed ERa. Several studies have shown higher ER content in ovarian carcinomas compared with normal ovary or benign ovarian tumors [38,62]. Use of Premarin and especially with diethylstilbestrol and menopausal estrogen replacement may be associated with increased risk of ovarian cancer [26]. This evidence suggests that ERa signaling may be important in ovarian carcinogenesis. ERa and ERh have distinct patterns of tissue expression and can exist as homo- or heterodimers. In vitro studies have shown that, depending on the nature of ligand and response element on DNA with which the activated receptors interact, ERa and ERh can have opposite effects on gene transcription [63]. In this study we found, for the first time, a distinctive correlation between ERa and ERh in BRCA1-associated compared with sporadic ovarian carcinomas. A positive correlation between the level of expression of ERa and ERh was found in sporadic ovarian cancer cases. In contrast, ERa and ERh were inversely expressed in BRCA1-associated ovarian cancer. Previous reports have shown ERa and ERh to be coexpressed in the majority of ovarian tumors [62,64]. It has been suggested that an altered ratio of ERa relative to ERh mRNA may be important in ovarian carcinogenesis. Pujol et al. [62] hypothesized that the overexpression of ERa relative to ERh mRNA may be a marker of ovarian carcinogenesis and it has been suggested that the balance between ERa and ERh is critical in maintaining normal cellular function in ovarian epithelial cells and that loss of ERh expression may lead to uncontrolled cellular proliferation [64,65]. In BRCA1-associated cases, ERh was found predominantly in tumors expressing low or no ERa. This may indicate that when ERh is co-expressed with significant levels of ERa, BRCA1-induced carcinogenesis is less likely to occur. However, the number of ERh positive cases is small and further studies are required to confirm this result and to determine whether the inverse relationship between ERh and ERa is related to the cause of BRCA1associated ovarian carcinogenesis or whether it is a consequence of this process. Relative levels of expression of PRA and PRB in ovarian cancer Normal epithelial cells in hormone responsive tissues including the breast [66], endometrium [67] and ovary [68] express both PRA and PRB at equivalent levels. However, it has previously been shown in the breast [66] and the endometrium [67], that progression to malignancy is accompanied by loss of one, or other isoform, leading to an imbalance in PR isoform expression and presumably altered progestin responsiveness. However, in ovarian

23

cancer, it is still controversial in which loss of PRB [41] and PRA [32] as well as PRB overexpression [68] have been reported. In this study, we found PRA to be lost preferentially in BRCA1-associated ovarian cancer cases compared with sporadic cases, where either PRA or PRB expression was lost. Few cases, either BRCA1-associated or sporadic, expressed both PRA and PRB. This expression pattern may be a further example of the phenotype of BRCA1-associated tumors being distinct but overlapping with a subset of sporadic tumors. A limitation of the current study is the modest cohort size however similar small datasets have previously shown a difference in response to platinum-based chemotherapy in BRCA1-associated ovarian cancer compared with sporadic cancer [19] and were sufficient to determine the overlap in gene expression profiles of BRCA1 and BRCA2-linked ovarian cancers with subgroups of sporadic ovarian cancers [25]. The absence of a major difference between the two ovarian groups in the present study could reflect the fact that, in matching for age, histology, and stage this study inadvertently biased the sporadic cases to those most likely to have mutations in BRCA1 in the absence of family history (i.e. young with serous histology). In conclusion, we demonstrated that the association between ERa and ERh and relative expression of PRA and PRB in BRCA1-associated ovarian cancer was different to that found in the sporadic group. However, individual steroid hormone receptor expression in BRCA1-associated ovarian carcinoma was similar to expression in matched ovarian cancers from women with no family history of breast or ovarian cancer. This is different from the findings in breast cancer where BRCA1-associated tumors tend to be more ER and PR-negative compared with sporadic breast cancers and to have distinct histological features. This may indicate that the role of ovarian hormones in the aetiology of breast and ovarian cancer is distinct and/or that the functional interaction between BRCA1 and ovarian hormone receptors is different in the two tissue types. The overlapping but not identical hormone receptor profiles of BRCA1-associated and sporadic tumors may be further evidence that epigenetic inactivation of BRCA1 may be involved in carcinogenesis in a proportion of sporadic ovarian tumors.

Acknowledgments The authors wish to thank Georgia Chenevix-Trench (Queensland Institute for Cancer Research and kConFab), Jennifer Leary (Familial Cancer Clinic, Westmead Hospital), Mirella Daja and Kim Ow (Oncology Research Centre, POWH), Michael Buckley (Cytogenetics Department, POWH), and Irena Kotchetkova (Hereditary Cancer Clinic, POWH) for their expert contributions and helpful discussions. We also thank The Kathleen Cuningham Consortium for Research into Familial Breast Cancer (kConFab) for

24

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

providing some of the BRCA1-associated ovarian tumor specimens for the present study. This study was supported by substantial donations from Mr. Steve Eckowitz and the Farmoz company and in part by the National Institutes of Health (USA) NIH CA 18119. kConFab has been funded by the Kathleen Cuningham Foundation, National Breast Cancer Foundation, National Health and Medical Research Council, Anti Cancer Council of Victoria, Anti Cancer Foundation of South Australia, Cancer Foundation of Western Australia, Queensland Cancer Fund and NSW Cancer Council. References [1] Chen Y, Lee W, Chew HK. Emerging roles of BRCA1 in transcriptional regulation and DNA repair. J Cell Physiol 1999;181:385 – 92. [2] Fan S, Wang JA, Yuan R, Ma Y, Meng Q, Erdos MR, et al. BRCA1 inhibition of estrogen receptor signaling in transfected cells. Science 1999;284:1354 – 6. [3] Zheng L, Annab LA, Afshari CA, Lee WH, Boyer TG. BRCA1 mediates ligand-independent transcriptional repression of the estrogen receptor. Proc Natl Acad Sci U S A 2001;98:9587 – 92. [4] Fan S, Ma YX, Wang C, Yuan RQ, Meng Q, Wang JA, et al. Role of direct interaction in BRCA1 inhibition of estrogen receptor activity. Oncogene 2001;20:77 – 87. [5] Park JJ, Irvine RA, Buchanan G, Koh SS, Park JM, Tilley WD, et al. Breast cancer susceptibility gene 1 (BRCA1) is a coactivator of the androgen receptor. Cancer Res 2000;60:5946 – 9. [6] Yeh S, Hu YC, Rahman M, Lin HK, Hsu CL, Ting HJ, et al. Increase of androgen-induced cell death and androgen receptor transactivation by BRCA1 in prostate cancer cells. Proc Natl Acad Sci U S A 2000; 97:11256 – 61. [7] Rebbeck TR, Kantoff PW, Krithivas K, Neuhausen S, Blackwood MA, Godwin AK, et al. Modification of BRCA1-associated breast cancer risk by the polymorphic androgen-receptor CAG repeat. Am J Hum Genet 1999;64:1371 – 7. [8] Rajan JV, Marquis ST, Gardner HP, Chodosh LA. Developmental expression of BRCA2 colocalizes with BRCA1 and is associated with proliferation and differentiation in multiple tissues. Dev Biol 1997; 184:385 – 401. [9] Huang WY, Newman B, Millikan RC, Schell MJ, Hulka BS, Moorman PG. Hormone-related factors and risk of breast cancer in relation to estrogen receptor and progesterone receptor status. Am J Epidemiol 2000;151:703 – 14. [10] Loman N, Johannsson O, Bendahl PO, Borg A, Ferno M, Olsson H. Steroid receptors in hereditary breast carcinomas associated with BRCA1 or BRCA2 mutations or unknown susceptibility genes. Cancer 1998;83:310 – 9. [11] Lakhani SR, Jacquemier J, Sloane JP, Gusterson BA, Anderson TJ, van de Vijver MJ, et al. Multifactorial analysis of differences between sporadic breast cancers and cancers involving BRCA1 and BRCA2 mutations. J Natl Cancer Inst 1998;90:1138 – 45. [12] Lakhani SR, van de Vijver MJ, Jacquemier J, Anderson TJ, McGuffog L, Easton DF, et al. The pathology of familial breast cancer: predictive value of immunohistochemical markers estrogen receptor, progesterone receptor, HER-2, and p53 in patients with mutations in BRCA1 and BRCA2. J Clin Oncol 2002;20:2310 – 8. [13] Lakhani SR, Manek S, Penault-Llorca F, Flanagan A, Arnout L, Merrett S, et al. Pathology of ovarian cancers in BRCA1 and BRCA2 carriers. Clin Cancer Res 2004;10(7):2473 – 81. [14] Rubin SC, Benjamin I, Behbakht K, Takahashi H, Morgan MA, Li Volsi VA, et al. Clinical and pathological features of ovarian cancer in women with germ-line mutations of BRCA1. N Engl J Med 1996;335(19):1413 – 6.

[15] Chang J, Fryatt I, Ponder B, Fisher C, Gore ME. A matched control study of familial epithelial ovarian cancer: patient characteristics, response to chemotherapy and outcome. Ann Oncol 1995;6(1): 80 – 2. [16] Chen X, Feng Y. Effect of progesterone combined with chemotherapy on epithelial ovarian cancer. Chin Med J (Engl) 2003;116(3):388 – 91. [17] Johannsson OT, Idvall I, Anderson C, Borg A, Barkardottir RB, Egilsson V, et al. Tumour biological features of BRCA1-induced breast and ovarian cancer. Eur J Cancer 1997;33 3:362 – 71. [18] Pharoah PD, Easton DF, Stockton DL, Gayther S, Ponder BA. Survival in familial, BRCA1-associated, and BRCA2-associated epithelial ovarian cancer. United kingdom coordinating committee for cancer research (UKCCCR) familial ovarian cancer study group. Cancer Res 1999;59(4):868 – 71. [19] Cass I, Baldwin RL, Varkey T, Moslehi R, Narod SA, Karlan BY. Improved survival in women with BRCA-associated ovarian carcinoma. Cancer 2003;97:2187 – 95 [New York, NY, United States]. [20] Boyd J, Sonoda Y, Federici MG, Bogomolniy F, Rhei E, Maresco DL, et al. Clinicopathological features of BRCA-linked and sporadic ovarian cancer. JAMA 2000;283(17):1 – 18. [21] Buller RE, Shahin MS, Geisler JP, Zogg M, DeYoung BR, Davis CS. Failure of BRCA1 dysfunction to alter ovarian cancer survival. Clin Cancer Res 2002;8(5):1196 – 202. [22] Johannsson OT, Ranstam J, Borg A, Olsson H. Survival of BRCA1 breast and ovarian cancer patients: a population-based study from southern Sweden. J Clin Oncol 1998;16(2):397 – 404. [23] Ramus SJ, Pharoah PDP, Harrington P, Pye C, Werness B, Bobrow L, et al. BRCA1/2 mutation status influences somatic genetic progression in inherited and sporadic epithelial ovarian cancer cases. Cancer Res 2003;63(2):417 – 23. [24] Patael-Karasik Y, Daniely M, Gotlieb WH, Ben-Baruch G, Schiby J, Barakai G, et al. Comparative genomic hybridization in inherited and sporadic ovarian tumours in Israel. Cancer Genet Cytogenet 2000; 121:26 – 32. [25] Jazaeri AA, Yee CJ, Sotiriou C, Brantley KR, Boyd J, Liu ET. Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers. J Natl Cancer Inst 2002;94:990 – 1000. [26] Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis concerning the role of androgens and progesterone. J Natl Cancer Inst 1998;90:1774 – 86. [27] La Vecchia C, Decarli A, Franceschi S, Regallo M, Tognoni G. Age at first birth and the risk of epithelial ovarian cancer. J Natl Cancer Inst 1984;73:663 – 6. [28] Makar AP. Hormone therapy in epithelial ovarian cancer. Endocr Relat Cancer 2000;7:85 – 93. [29] Hildreth NG, Kelsey JL, LiVolsi VA, Fischer DB, Holford TR, Mostow ED, et al. An epidemiologic study of epithelial carcinoma of the ovary. Am J Epidemiol 1981;114:398 – 405. [30] Key TJA, Chen J, Wang DY, Pike MC, Boreham J. Sex hormones in women in rural China and in Britain. Br J Cancer 1990;62:631 – 6. [31] Pike MC, Spicer DV. Contraception. In: Haseltine DSF, editor. Oral Contraceptives and Cancer. New York7 Springer-Verlag; 1993. p. 67 – 84. [32] Akahira J, Inoue T, Suzuki T, Ito K, Konno R, Sato S, et al. Progesterone receptor isoforms A and B in human epithelial ovarian carcinoma: immunohistochemical and RT-PCR studies. Br J Cancer 2000;83:1488 – 94. [33] Cardillo MR, Petrangeli E, Aliotta N, Salvatori L, Ravenna L, Chang C, et al. Androgen receptors in ovarian tumors: correlation with estrogen and progesterone receptors in an immunohistochemical and semiquantitative image analysis study. J Exp Clin Cancer Res 1998;17:231 – 7. [34] Langdon SP, Gabra H, Bartlett JMS, Rabiaz GJ, Hawkins RA, Tesdale AL, et al. Functionality of the progesterone receptor in ovarian cancer and its regulation by estrogen. Clin Cancer Res 1998;4:2245 – 51. [35] Lau KM, Mok SC, Ho SM. Expression of human estrogen receptor-a and-h, progesterone receptor, and androgen receptor mRNA in normal

M. Aghmesheh et al. / Gynecologic Oncology 97 (2005) 16–25

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47] [48]

[49]

[50]

and malignant ovarian epithelial cells. Proc Natl Acad Sci U S A 1999;96:5722 – 7. Munstedt K, Steen J, Knauf AG, Buch T, von Georgi R, Franke FE. Steroid hormone receptors and long term survival in invasive ovarian cancer. Cancer 2000;89:1783 – 91. van Doorn HC, Burger CW, van der Valk P, Bonfrer HMG. Estrogen, progesterone, and androgen receptors in ovarian neoplasia: correlation between immunohistochemical and biochemical receptor analyses. J Clin Pathol 2000;53:201 – 5. Brandenberger AW, Tee MK, Jaffe RB. Estrogen receptor alpha (ER-a) and beta (ER-h) mRNAs in normal ovary, ovarian serous cystadenocarcinoma and ovarian cancer cell lines: down-regulation of ER-h in neoplastic tissues. J Clin Endocrinol Metab 1998;83:1025 – 8. Katzenellenbogen BS. Mechanisms of action and cross-talk between estrogen receptor and progesterone receptor pathways. J Soc Gynecol Investig 2000;7(Suppl 1):S33–7. Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gronemeyer H, et al. Two distinct estrogen-regulated promoters generate transcripts encoding the two functionally different human progesterone receptor forms A and B. EMBO J 1990;9:1603 – 14. Havrilesky LJ, McMahon CP, Lobenhofer EK, Whitaker R, Marks JR, Berchuck A. Relationship between expression of coactivators and corepressors of hormone receptors and resistance of ovarian cancers to growth regulation by steroid hormones. J Soc Gynecol Investig 2001; 8:104 – 13. Risch HA, McLaughlin JR, Cole DEC, Rosen B, Bradley L, Kwan E, et al. Prevalence and penetrance of germline BRCA1 and BRCA2 mutations in a population series of 649 women with ovarian cancer. Am J Hum Genet 2001;68:700 – 10. Mote PA, Johnston JF, Manninen T, Tuohimaa P, Clarke CL. Detection of progesterone receptor forms A and B by immunohistochemical analysis. J Clin Pathol 2001;54:624 – 30. Choi I, Ko C, Park-Sarge OK, Nie R, Hess RA, Graves C, et al. Human estrogen receptor beta-specific monoclonal antibodies: characterization and use in studies of estrogen receptor beta protein expression in reproductive tissues. Mol Cell Endocrinol 2001;181:139 – 50. Clarke CL, Zaino RJ, Feil PD, Miller JV, Steck ME, Ohlsson-Wilhelm BM, et al. Monoclonal antibodies to human progesterone receptor: characterization by biochemical and immunohistochemical techniques. Endocrinology 1987;121:1123 – 32. Yang JL, Ow KT, Russell PJ, Ham JM, Crowe PJ. Higher expression of oncoproteins c-myc, c-erbB-2/neu, PCNA, and p53 in metastasizing colorectal cancer than in nonmetastasizing tumors. Ann Surg Oncol 1996;3:574 – 9. Yancik R. Ovarian cancer. Age contrasts in incidence, histology, disease stage at diagnosis, and mortality. Cancer 1993;71(Suppl 2):S517–23. Skliris GP, Carder PJ, Lansdown MRJ, Speirs V. Immunohistochemical detection of ER-h in breast cancer: towards more detailed receptor profiling? Br J Cancer 2001;84:1095 – 8. Hall RE, Lee CSL, Alexander IE, Shine J, Clarke CL, Sutherland RL. Steroid hormone receptor gene expression in human breast cancer cells: inverse relationship between estrogen and glucocorticoid receptor messenger RNA levels. Int J Cancer 1990;46:1081 – 7. Horwitz KB, Koseki Y, McGuire WL. Estrogen control of progesterone receptor in human breast cancer: role of estradiol and antiestrogen. Endocrinology 1978;103:1742 – 51.

25

[51] Narod SA, Risch H, Moslehi R, Dorum A, Neuhausen S, Olsson H, et al. Oral contraceptives and the risk of hereditary ovarian cancer. Hereditary Ovarian Cancer Clinical Study Group. N Engl J Med 1998;339:424 – 8. [52] Narod SA, Sun P, Ghadirian P, Lynch H, Isaacs C, Garber J, et al. Tubal ligation and risk of ovarian cancer in carriers of BRCA1 or BRCA2 mutations: a case-control study. Lancet 2001;357:1467 – 70. [53] Berchuck A, Heron K-A, Carney ME, Lancaster JM, Fraser EG, Vinson VL, et al. Frequency of germline and somatic BRCA1 mutations in ovarian cancer. Clin Cancer Res 1998;4:2433 – 7. [54] Lallas TA, Buekers TE, Buller RE. BRCA1 mutations in familial ovarian cancer. Mol Genet Metab 1999;67:357 – 63. [55] Merajver SD, Pham TM, Caduff RF, Chen M, Poy EL, Cooney KA, et al. Somatic mutations in the BRCA1 gene in sporadic ovarian tumors. Nat Genet 1995;9:439 – 43. [56] Baldwin RL, Nemeth E, Tran H, Shvartsman H, Cass I, Narod S, et al. BRCA1 promoter region hypermethylation in ovarian carcinoma: a population-based study. Cancer Res 2000;60:5329 – 33. [57] Geisler JP, Hatterman-Zogg MA, Rathe JA, Buller RE. Frequency of BRCA1 dysfunction in ovarian cancer. J Natl Cancer Inst 2002;94: 61 – 7. [58] Russell PA, Pharoah PDP, De Foy K, Ramus SJ, Symmonds I, Wilson A, et al. Frequent loss of BRCA1 mRNA and protein expression in sporadic ovarian cancers. Int J Cancer 2000;87:317 – 21. [59] Zheng W, Luo F, Lu JJ, Baltayan A, Press MF, Zhang Z, et al. Reduction of BRCA1 expression in sporadic ovarian cancer. Gynecol Oncol 2000;76:294 – 300. [60] Auersperg N, Edelson MI, Mok SC, Johnson SW, Hamilton TC. The biology of ovarian cancer. Semin Oncol 1998;25:281 – 304. [61] Chadha S, Rao BR, Slotman BJ, van Vroonhoven CCJ, van der Kwast TH. An immunohistochemical evaluation of androgen and progesterone receptors in ovarian tumors. Hum Pathol 1993; 24:90 – 5. [62] Pujol P, Rey JM, Nirde P, Roger P, Gastaldi M, Laffargue F, et al. Differential expression of estrogen receptor-a and-h messenger RNAs as a potential marker of ovarian carcinogenesis. Cancer Res 1998;58:5367 – 73. [63] Paech K, Webb P, Kuiper GGJM, Nilsson S, Gustafsson JA, Kushner PJ, et al. Differential ligand activation of estrogen receptors ERa and ERh at AP1 sites. Science 1997;277:1508 – 10. [64] Rutherford T, Brown WD, Sapi E, Aschenazi S, Munoz A, Mor G. Absence of estrogen receptor-h expression in metastatic ovarian cancer. Obstet Gynecol 2000;96:417 – 21. [65] Bardin A, Hoffmann P, Boulle N, Katsaros D, Vignon F, Pujol P, et al. Involvement of estrogen receptor beta in ovarian carcinogenesis. Cancer Res 2004;64(16):5861 – 9. [66] Mote PA, Bartow S, Tran N, Clarke CL. Loss of co-ordinate expression of progesterone receptors A and B is an early event in breast carcinogenesis. Breast Cancer Res Treat 2002;72:163 – 72. [67] Arnett-Mansfield RL, deFazio A, Wain GV, Jaworski RC, Byth K, Mote PA, et al. Relative expression of progesterone receptors A and B in endometrioid cancers of the endometrium. Cancer Res 2001;61: 4576 – 82. [68] Li AJ, Baldwin RL, Karlan BY. Estrogene and progesterone receptor subtype expression in normal and malignant ovarian epithelial cell cultures. Am J Obstet Gynecol 2003;189:22 – 7.