J. Appl. Genet. 45(2), 2004, pp.

255-263

Ovarian cystadenoma as a characteristic feature of families with hereditary ovarian cancers unassociated with BRCA1 and BRCA2 mutations
Janusz MENKISZAK1, Marek BRZOSKO3, Bohdan GRSKI2, Jacek FLICISKI3, Anna JAKUBOWSKA2, Dariusz EBIEOWICZ5, Jacek GRONWALD2, Tomasz HUZARSKI2, Tomasz BYRSKI2, Leszek TERESISKI4, Maria CHOSIA4, Izabella RZEPKA-GRSKA1, Jan LUBISKI2
1 Chair and Department of Surgical Gynecology and Gynecological Oncology of Adults and Adolescents and 2Hereditary Cancer Center, Department of Genetics and Pathology and 3 Department of Rheumatology and 4Department of Pathology and 5Chair and Department of Obstetrics and Perinatology, Pomeranian Medical University, Szczecin, Poland

Abstract. The study aimed to determine whether hereditary ovarian cancers that are not caused by BRCA1/BRCA2 constitutional mutations are associated with a predisposition to cystadenoma. The study consisted of two parts. Part one concerned the incidence of ovarian cystadenoma in females from families with hereditary ovarian cancer unassociated with BRCA1 mutations. The study group included 62 female patients from 29 families, without any previously diagnosed malignancy, with no proven constitutional mutation of the BRCA1 gene. The first control group was composed of 62 female patients from 53 families, without any previously diagnosed malignancy, with an identified constitutional mutation of the BRCA1 gene. The second control group comprised 124 female patients for whom the only reason for the examination was a prophylactic check-up. All studied women were subjected to intravaginal ultrasonographic investigations. In 8 patients with benign and/or borderline ovarian cystadenoma, a complete sequencing of coding fragments of the BRCA2 gene from the peripheral blood DNA was performed. Part two of this study concerned the incidence and pattern of malignant tumors in the families of female patients with ovarian cystadenoma. The final study group included 117 patients who had 726 I0 relatives (359 females and 367 males). We concluded that cystadenoma is likely to be a characteristic feature of the subgroup of families with hereditary ovarian cancers unassociated with BRCA1/BRCA2 constitutional mutations.
Key words: hereditary ovarian cancer, ovarian cystadenoma. Received: January 20, 2003. Revised: December 17, 2003. Accepted: February 4, 2004. Correspondence: J. MENKISZAK, Chair and Department of Surgical Gynecology and Gynecological Oncology of Adults and Adolescents, Pomeranian Medical University, Al. Powstacw Wielkopolskich 72, 70-111 Szczecin, Poland, e-mail: janusz@sz.home.pl

256

J. Menkiszak et al.

Introduction
Cystadenoma, a benign epithelial tumor of the ovary, may undergo malignant transformation in some cases, which leads to cystadenocarcinoma (DUTT, BERNEY 2000, ELTABBAKH et al. 1999, HONG et al. 1998, SEIDMAN, KURMAN 1996). So far, it has been confirmed that hereditary ovarian cancers develop most frequently in BRCA1 or BRCA2 gene mutation carriers. However, in a substantial proportion of familial ovarian cancers, no cases of constitutional mutations of BRCA1 or BRCA2 genes have been found (MENKISZAK et al. 2003). Moreover, the majority of patients with cystadenoma do not carry any BRCA1 or BRCA2 mutations (RISCH et al. 2001). This study aimed to determine whether the predisposition to cystadenoma is related to hereditary ovarian cancers unassociated with BRCA1/BRCA2 constitutional mutations.

Material and methods

Part one incidence of ovarian cystadenoma in females from families with hereditary ovarian cancer unassociated with BRCA1mutations

The study group included 62 female patients from 29 families, without any previously diagnosed malignancy, with no identified constitutional mutation of the BRCA1 gene. The first control group comprised 62 female patients from 53 families, without any previously diagnosed malignancy, with an identified constitutional mutation of BRCA1 gene. The second control group was composed of 124 female patients from various workplaces of the city of Szczecin, for whom the only reason for the examination was a prophylactic check-up. The patients from the study group and from the first control group were consecutive patients of the Hereditary Cancer Center in Szczecin. The mean age in the study group was 43.6 years. The study group and both control groups were matched with respect to age 2 years, with group size ratios 1 : 1 and 1 : 2, respectively. All studied women were subjected to intravaginal ultrasonographic investigations with a 7.0 MHz frequency transducer and 3535 type medical diagnostic ultrasound system (serial no. 1635620, Bruel & Kjaer, Denmark), and in cases of suspected pathology additionally with a C125 transducer (color Doppler) and 128XP/10 model medical diagnostic ultrasound system (serial no. 19990 Acuson Corporation, USA). The investigations were performed in the Ultrasonography Department, the Chair and Department of Surgical Gynecology and Gynecological Oncology for Adults and Adolescents, and the Chair and Department of Obstetrics and Perinatology, Pomeranian Medical University. In cases of ovarian lesions (detected by ultrasonography) that most probably represented a benign pattern (SPACZYSKI, SPACZYSKI 2000), follow-up imag-

Ovarian cystadenoma in hereditary ovarian cancers families

257

ing was scheduled after the consecutive menstruation. In cases of persistent changes recorded by ultrasonography, the lesions were surgically excised during laparotomy/laparoscopy, and then histologically verified. In the study group and in the first control group, the incidence of the three BRCA1 mutations that are prevalent in Poland (5382insC, C61G, 4153delA) have been estimated (GRSKI et al. 2000). The above changes constitute almost 90% of BRCA1 mutations in Polish families with aggregation of breast/ovarian cancer (GRSKI et al. 2004). The BRCA1 test was performed in peripheral blood or paraffin specimens of ovarian tissues In each family (I0 or II0 relatives) of patients from the study group and the first control group, there were two or more ovarian cancers, or an ovarian cancer and breast cancer in individuals aged under 50 years of life. In 8 patients with both benign and borderline ovarian cystadenoma, a complete exon-by-exon sequencing of coding fragments of the BRCA2 gene from the peripheral blood DNA was performed (JAKUBOWSKA et al. 2002, 2003). The study group and both control groups were statistically compared with regard to incidence of cystadenoma and other pathological lesions within sex organs by using Stata Statistical Software 5 (Stata Corporation, College Station, TX). Odds ratio (OR), confidence interval (CI), level of significance (p), specificity, and sensitivity were calculated.
Incidence and pattern of malignant tumors in families of female patients with ovarian cystadenoma

Between 1998 and 2001, 152 consecutive patients with histologically confirmed ovarian cystadenoma were recorded in the histopathology laboratories of the Department of Pathology and Department of Genetics and Pathology, Pomeranian Medical University. For 19 of the patients some information was missing, so our initial study group included 133 female patients with known personal data. Out of these 133 patients, 7 could not provide information on I0 relatives, and 9 refused cooperation. Consequently, the final study group included 117 patients who had 726 I0 relatives (359 females and 367 males). Serous ovarian cystadenoma was diagnosed in 66 patients, including 15 cases of borderline type. Mucinous tumors were diagnosed in 48 patients, including 5 borderline cases. In one case, in a 35-year-old patient, serous cystadenoma of borderline malignancy and mucinous cystadenoma were diagnosed simultaneously. There were also 2 cases of endometrioid cystadenoma. Pedigree data were obtained on the basis of interviews. The following parameters were evaluated: year of birth, age at onset, location of the malignant lesion, and present age (or age at death). The incidence and type of the malignancy in the study group were compared with the expected values basing on epidemiological data for malignancy-related morbidity in Poland in 1999 (DIDKOWSKA et al. 2002). BRCA1 testing for occurrence of germline 5382insC, C61G and 4153delA was performed in cystadenoma patients with first-degree relatives

258

J. Menkiszak et al.

affected by breast or ovarian cancers. The statistical significance of the achieved results was evaluated with regard to OR, CI, and p according to ROTHMAN and GREENLAND (1998).

Results
Incidence of ovarian cystadenoma in females from families with hereditary ovarian cancer unassociated with BRCA1mutations

The results are shown in Tables 1 and 2. In the study group, cystadenoma was found in 8 patients (12.9% of the group), including 4 serous and 4 mucinous cases. Three of them were borderline malignancy cases. No cystadenoma cases were found in the two control groups, including patients with constitutional mutations of the BRCA1 gene. Constitutional mutations of the BRCA2 gene were not detected by complete exon by exon sequencing of DNA extracted from peripheral blood of patients with cystadenoma (JAKUBOWSKA et al. 2002, 2003). Comparison of incidence of other pathology within sex organs between the study and control groups revealed no statistically significant differences except for an increased frequency of uterine myomas in the study group, in comparison to BRCA1 mutation carriers (OR 4.72; p = 0.0132).
Incidence and pattern of malignant tumors in families of female patients with ovarian cystadenoma

The observed and expected incidences of malignant tumors are shown in Table 3. An insignificantly higher number of malignant tumors was found in the study group. The highest risk increase was found for ovarian cancer (OR- 4.07; p = 0.056). Moreover, the risk of early-onset breast cancer (diagnosed under the age of 50) increased approximately two-fold (OR-1.68; but p = 0.48). No BRCA1 mutation was found among 17 cystadenoma patients with first-degree relatives affected by breast or ovarian cancer. The observed number of lung cancer cases in males was only slightly higher than the expected value. No increased risk for colorectal cancer was found. The numbers of the remaining types of malignancies were too small to perform a reliable statistical analysis and were excluded from the comparison.

Discussion
Firstly, the results of the study revealed an increased incidence of cystadenoma in female patients from families with aggregation of ovarian/breast cancers

Table 1. Comparison of incidence rates of cystadenoma and other pathological lesions within sex organs between the study group (BCRA1) and first control group (BCRA1+)
Incidence rate Study group (BRCA1-) n = 62 1st control group (BRCA1+) n = 62 OR CI p 0/62 3.23% (2/62) 0.49 2.63 1.0 3.10 4.72 1.0 2.03 0.43 1.34 16.40 0.02 51.20 0.26 0.02 51.20 0.56 3.23% (2/62) 1.61% (1/62) 1.61%(1/62) 4.84% (3/62) 1.61%(1/62) 0/62 1.61%(1/62) 0/62 0 3.86 0.5589 0.2431 0.3154 0.3094 0.0132 0.3154 0.5589 16.20 1.08 73.60 0.0035 12.00 1.61 8.06 0.81 4.84 19.35 0.81 3.23 12.9 (8/62) 1.61% (1/62) 8.06% (5/62) 0/62 4.84% (3/62) 19.35% (12/62) 0/62 0/62 3.23% (2/62) 0/62 Sensitivity Specificity

Type of lesion

Ovarian cystadenoma and borderline cystadenoma (serous/mucinous)

99.19 96.77 96.77 99.29 98.39 95.16 99.29 98.39

Ovarian cancer

Simple/serous/follicular/endometrial/lutein cyst of ovary

Cystic teratoma of ovary

Functional ovarian cyst or polycystic ovary

Uterine myoma

Endometrial hypertrophy/hyperplasia

Endometrial/cervical polyp

Hydrosalpinx

Other lesions within sex organs (congenital defects or fluid in uterine cavity or retrouterine pouch)

Table 2. Comparison of incidence rates of cystadenoma and other pathological lesions within sex organs between the study group (BCRA1-) and second control group
Incidence rate Study group (BRCA1) n = 62 2nd control group n = 124 0/124 0/124 2.01 1.12 0.37 3.50 0.16 2.04 0.39 1.74 0 7.83 0 4.75 0.36 46.12 4.1 1.0 0 7.83 0.58 0.82 1.00 0.66 7.26% (9/124) 0/124 8.06% (10/124) 22.58% (28/124) 1.61% (2/124) 2.42% (3/124) 0.81%(1/124) 1.61% (2/124) 0 0.6152 0.8442 0.4160 0.61372 1.0 0.7208 0.2169 1.0 18.22 2.86 0.0003 OR CI p 12.9 (8/62) 1.61% (1/62) 8.06% (5/62) 0/62 4.84% (3/62) 19.35% (12/62) 0/62 0/62 3.23% (2/62) 0/62 12.90 1.61 8.06 4.84 19.35 1.61 1.61 3.23 1.61 Sensitivity Specificity

Type of lesion

Ovarian cystadenoma and borderline cystadenoma (serous/mucinous)

99.19 99.19 92.74 91.94 77.42 98.39 97.58 99.19 98.39

Ovarian cancer

Simple/serous/follicular/endo-metrial/lutein cyst of ovary

Cystic teratoma of ovary

Functional ovarian cyst or polycystic ovary

Uterine myoma

Endometrial hypertrophy/hyperplasia

Endometrial/cervical polyp

Hydrosalpinx

Other lesions within sex organs (congenital defects or fluid in uterine cavity or retrouterine pouch)

Ovarian cystadenoma in hereditary ovarian cancers families

261

Table 3. The observed and expected numbers of malignant tumors total, breast cancers, ovarian cancers, lung cancers and colorectal cancers among I0 relatives* of patients with cystadenoma
Type of malignancy Total Breast cancer (regardless of age) Breast cancer (<50 years) Ovarian cancer Lung cancer Colorectal cancer
* n = 726 (females n = 359, males n = 367) # n=359 ^ n=367

Observed number of cases 69* 9# 5# 8# 9^ 5*

Expected number of cases 53* 6# 2.6# 1.8# 6.8^ 4.7*

OR 1.33 1.51 1.68 4.07 1.29 1.0

CI 0.92-1.93 0.55-4.12 0.44-6.39 0.97 0.49-3.38 0.31-3.264

p 0.1301 0.4337 0.4770 0.0560 0.6132 1.0

and without constitutional mutations of the BRCA1 gene. Secondly, the results indicate a higher incidence of ovarian cancer and breast cancer diagnosed under the age of 50 years in relatives of cystadenoma patients. The above-mentioned observations indicate the presence of a familial predisposition for development of ovarian and breast cancers under the age of 50, which are associated with constitutional molecular abnormalities resulting also in cystadenoma development. For years cystadenomas have been highly suspected to be a precancerous condition for ovarian cancers. The multilocular structure of the tumor, the preserved cystadenoma pattern in cystadenocarcinomas, and the similar type of molecular abnormalities in benign cystadenomas, borderline malignancy cystadenomas and cystadenocarcinomas, all support this hypothesis (SEIDMAN, KURMAN 1996, SINGER et al. 2002, TIBILETTI et al. 2001). Bilateral and/or familial cases of cystadenoma support the hypothesis on the genetic pattern of the predisposition to cystadenoma development. Similarly, multilocular ovarian cancers frequently tend to be bilateral or familial. We have not found in the literature any references describing the risk of ovarian cancer and early onset breast cancer in female relatives of patients with cystadenoma. The only exception was a Finnish study (AURANEN et al. 1996), but those authors limited their studies to borderline malignance cases and failed to find a statistically increased risk for the chosen malignancies in relatives of patients with cystadenoma. The correlations observed in this study are logically consistent with the existing data. The highly increased risk of malignancy in families with aggregation of cystadenomas, ovarian cancers or early-onset breast cancers (<50 years of life) among I0 and/or II0 relatives may serve as an additional proof: initial results of our analysis of a series of 22 consecutive cystadenoma cases fulfilling the above-men-

262

J. Menkiszak et al.

tioned criteria revealed an over 20-fold higher risk of ovarian cancer and over 2-fold higher risk of breast cancer diagnosed under the age of 50. It seems that the most probable scheme of carcinogenesis in these families includes development of cystadenoma on the basis of a defect caused by a single gene and transformation of some cystadenomas to cancers. The Carneys Syndrome is an example of a higher incidence of ovarian cystadenomas coexisting with endocrine glands malignancies (STRATAKIS et al. 2000). The growth of cystadenoma results in hormonal abnormalities, thus increasing the risk of breast cancer (HONDA et al. 2000). Another well-known risk factor for breast cancer the expression of 17-beta-hydroxysteroid dehydrogenase responsible for estradiol synthesis was found in immunohistochemical tests in cases of ovarian cystadenoma (SASANO et al. 1996). So far we do not know the rate of progression from cystadenoma to cancer. The majority of ovarian cancers reveal a multilocular pattern on macroscopic examination. Mutation of the BRCA1 gene cannot be diagnosed in more than 50% of cases of familial ovarian cancers (MENKISZAK et al. 2003), and a higher incidence of cystadenoma cases is found in families with aggregation of ovarian/breast cancers. Thus it is likely that a high percentage of ovarian cancers may originate from cystadenomas that developed due to a hereditary predisposition. The relationship found in our study seems to be very important from the point of view of clinical practice. Identification of high-risk families enables an early diagnosis and excision of yet asymptomatic cystadenomas. That approach might reduce the risk of development of ovarian/breast cancers in this group to the level of a few percent. Identification of the gene/genes associated with the described predisposition for development of ovarian cystadenoma will be important for progress in diagnosis of the studied families. Verification of data achieved in our study on a larger number of samples and by independent centers will enable application of these results in clinical practice.
REFERENCES AURANEN A., GRENMAN S., MAKINEN J., PUKKALA E., SANKILA R., SALMI T. (1996). Borderline ovarian tumors in Finland: epidemiology and familial occurrence. Am. J. Epidemiol. 144: 548-553. DIDKOWSKA J., WOJCIECHOWSKA U., TARKOWSKI W., ZATOSKI W. (2002). Cancer in Poland in 1999. Warszawa, Centrum Onkologii: 85-88. DUTT N., BERNEY D.M. (2000).Clear cell carcinoma of the ovary arising in a mucinous cystadenoma. J. Clin. Pathol. 53: 938-939. ELTABBAKH G.H., NATARAJAN N., PIVER M.S., METTLIN C.J. (1999). Epidemiologic differences between women with borderline ovarian tumors and women with epithelial ovarian cancer. Gynecol. Oncol. 74: 103-175. GRSKI B., BYRSKI T., HUZARSKI T., JAKUBOWSKA A., MENKISZAK J., GRONWALD J., PUANSKA A., BEBENEK M., FISCHER-MALISZEWSKA L., GRZYBOWSKA E.,

Ovarian cystadenoma in hereditary ovarian cancers families

263

NAROD S.A., LUBINSKI J. (2000). Founder mutations in the BRCA1 gene in Polish families with breast-ovarian cancer. Am. J. Hum. Genet. 66: 1963-1968. GRSKI B., JAKUBOWSKA A., HUZARSKI T., BYRSKI T., GRONWALD J., GRZYBOWSKA E. et al. (2004). A high proportion of founder BRCA1 mutations in Polish breast cancer families. Int. J. Cancer submitted. HONDA K., ISHIKO O., WAKASA K., KAWABATA S., NISHIMURA S., OGITA S. (2000). Estrogen-producing ovarian mucinous cystadenomas in postmenopausal women. Oncol. Rep. 7: 803-808. HONG S.R., CHUN Y.K., KIM Y.J., LIM K.T., KIM H.S. (1998) Ovarian mucinous cystadenoma with mural nodule of anaplastic carcinoma. J. Korean Med. Sci. 13: 680-684. JAKUBOWSKA A., NEJ K., HUZARSKI T., SCOTT R.J., LUBINSKI J. (2002). BRCA2 gene mutations in families with aggregations of breast and stomach cancers. Br. J. Cancer 87: 888-891. JAKUBOWSKA A., SCOTT R., MENKISZAK J., GRONWALD J., BYRSKI T., HUZARSKI T. et al. (2003). A high frequency of BRCA2 gene mutations in Polish families with ovarian and stomach cancer. Eur. J. Hum. Genet. 11: 955-958. MENKISZAK J., GRONWALD J., GRSKI B., JAKUBOWSKA A., HUZARSKI T., BYRSKI T. et al. (2003). Hereditary ovarian cancer in Poland. Int. J. Cancer 106: 942-945. RISCH H.A., MCLAUGHLIN J.R., COLE D.E.C., ROSEN B., BRADLEY L., KWAN E., JACK E., VESPRINI D.J., KUPERSTEIN G., ABRAHAMSON J.L.A., FAN I., WONG B., NAROD S.A. (2001). Prevalence and penetrance of germline BRCA1 and BRCA2 mutations in a population series of 649 women with ovarian cancer. Am. J. Hum. Genet. 68: 700-710. ROTHMAN K.J., GREENLAND S. (1998) Modern Epidemiology. 2nd edn., Lippincott-Raven Publishers. SASANO H., SUZUKI T., NIIKURA H., KAGA K., SATO S., YAJIMA A., RAINEY W.E., NAGURA H. (1996). 17 beta-hydroxysteroid dehydrogenase in common epithelial ovarian tumors. Mod. Pathol. 9: 386-391. SEIDMAN J.D., KURMAN R.J. (1996). Subclassification of serous borderline tumors of the ovary into benign and malignant types: a clinicopathologic study of 65 advanced stage cases. Am. J. Surg. Pathol. 20: 1331-1345. SINGER G., KURMAN R.J., CHANG H.W., CHO S.K., SHIH Ie.M. (2002). Diverse tumorigenicpathwaysinovarianserouscarcinoma.Am. J. Patrol.160(4):1223-1228. SPACZYSKI M., SPACZYSKI S. (2000). Ultrasound in early detection of ovarian cancer. Pol. Przegl. Ginekol. 1: 9-15. STRATAKIS C.A., PAPAGEORGIOU T., PREMKUMAR A., PACK S., KIRSCHNER L.S., TAYMANS S.E., ZHUANG Z., OELKERS W.H., CARNEY J.A. (2000). Ovarian lesions in Carney complex: clinical genetics and possible predisposition to malignancy. J. Clin. Endocrinol. Metab. 85: 4010-4012. TIBILETTI M.G., BERNASCONI B., FURLAN D., BRESSAN P., CERUTTI R., FACCO C., FRANCHI M., RIVA C., CINQUETTI R., CAPELLA C., TARAMELLI R. (2001). Chromosome 6 abnormalities in ovarian surface epithelial tumors of borderline malignancy suggest a genetic continuum in the progression model of ovarian neoplasms. Clin. Cancer Res. 7: 3404-3409.