ARTICLE OUTLINE Abstract Introduction Methods Results Discussion Acknowledgement References Tables

Background: Jewish women have been reported to have a higher risk for familial breast cancer than non-Jewish women and to be more likely to carry mutations in breast cancer genes such as BRCA1. Because BRCA1 mutations also increase women's risk for ovarian cancer, we asked whether Jewish women are at higher risk for familial ovarian cancer than non-Jewish women.

Methods: To determine the effects of 1) Jewish religion and 2) ovarian cancer in a first-degree relative on women's risk for epithelial ovarian cancer, we used data from a population-based, case-control study conducted in 8 geographic regions in the United States from 1980 through 1982. The study group included 471 cases and 4,025 controls.

Results: Jewish women were more likely to have familial ovarian cancer than non-Jewish women (OR = 8.4, 95% CI = 2.6 - 28). The risk of having ovarian cancer appeared to be greater in Jewish women having a first-degree relative with ovarian cancer (OR = 8.81, 95% CI = 2.02 - 38.23) than in non-Jewish women having a first-degree relative with ovarian cancer (OR = 3.01, 95% CI = 1.61 - 5.64), but differences between Jewish and non-Jewish women were not statistically significant. Jewish women with no first-degree relative with ovarian cancer had no increased risk for ovarian cancer (OR = 1.27, 95% CI = 0.74 - 2.91) compared to non-Jewish women.

Conclusions: These results suggest that Jewish women may have a higher rate of familial ovarian cancer than non-Jewish women, but because the results are based on a small number of Jewish women with familial ovarian cancer, the results need to be confirmed in larger studies.

Germline mutations in the BRCA1 gene on chromosome 17 have been implicated as causes of familial breast and ovarian cancers [Shattuck-Eidens et al., 1995]. One BRCA1 mutation (185delAG) has been estimated to occur in 1% of Eastern European (Ashkenazi) Jews, a rate believed to be at least three times higher than the rate for the general U.S. population [Struewing et al., 1995]. Another BRCA1 mutation (5382insC) has also been identified in the Jewish population, but it occurs less frequently than 185delAG [Roa, 1996; Tonin et al., 1995]. Other mutations in BRCA1 and mutations in other genes such as BRCA2 and MSH2 may also contribute to the occurrence of familial ovarian cancer in this population [Oddoux, 1996; Roa, 1996; Szabo and King, 1995; Orth et al., 1994]. The 185delAG BRCA1 mutation has been estimated to cause 16% of the breast cancer cases and 39% of the ovarian cancer cases among Ashkenazi Jewish women whose cancer is diagnosed before they are 50 years old. (In contrast, this mutation is estimated to cause less than 4% of the breast cancer cases and about 3% of the ovarian cancer cases among women in the general U.S. population [Struewing et al., 1995]. Because one germline mutation (185delAG) is found at such a high rate in Jewish women, and other mutations are likely to contribute to the occurrence of ovarian cancer in this population, we asked whether Jewish women were at higher risk for familial ovarian cancer. To answer this question, we used data from a large population-based, case-control study conducted in the United States from 1980 through 1982.

Data for this study are derived from the Cancer and Steroid Hormone (CASH) study, a large, population-based, case-control study designed to assess the association between oral contraceptive use and cancer of the breast, ovary, and endometrium in women 20 to 54 years of age. Details of the study have been published elsewhere [Wingo et al., 1988]. Briefly, women with newly diagnosed, histologically confirmed ovarian cancer diagnosed from December 1, 1980 through December 31, 1982, were identified from eight population-based cancer registries that are part of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute [Young et al., 1981]. Controls were selected by random digit dialing from the same geographic regions and during the same period that the cases were selected.

Prior to our analysis, we excluded women with histories of ovarian, breast, or endometrial cancer, and women with an unknown number of ovaries or no ovaries. We We also excluded all women whose maternal medical history was unknown (23 women with and 128 women without cancer). We determined the distributions for the following characteristics in 471 case and 4,025 control women: birth year, menopause status, age at menopause, total pregnancies, education, anovulatory cycles (excluding those caused by oral contraceptives), and oral contraceptive (OC) use. We calculated the odds ratios (ORs) and 95% confidence intervals (CIs). We used multiple logistic regression to estimate the odds ratio for ovarian cancer among Jewish women while controlling for age, education, OC use, menopause status, age at menopause, and anovulatory cycles. We also used logistic regression to estimate the independent and combined effects of Jewish religion and family history of cancer on the odds ratio for cancer. Because of the small number of cases, we were unable to control for potential confounding factors in our interaction assessment. Familial ovarian cancer was defined as ovarian cancer in a woman who reported ovarian cancer in a mother, sister or daughter, and non-familial ovarian cancer as ovarian cancer in a woman who reported no such family history. Although family history of cancer was defined as reported ovarian cancer in a mother, sister, or daughter, there were no cases of daughters with ovarian cancer.

When we compared demographic and risk-factor characteristics among women with familial ovarian cancer (n = 16) with those among women with nonfamilial ovarian cancer (n = 455) and control women (n = 4,025), we found that women with familial ovarian cancer were older than women with nonfamilial ovarian cancer and control women, but the differences were not statistically significant (Table 1). Women with familial ovarian cancer were more likely to have experienced natural menopause than women with nonfamilial ovarian cancer (chi-square = 6.1, p = 0.108) or control women (chi-square = 10.2, p =0.017). Women with familial ovarian cancer were more likely to have had more than three pregnancies than women with nonfamilial ovarian cancer, but the results were not statistically significant (chi-square = 5.9, p = 0.117). Women with nonfamilial ovarian cancer were slightly more likely to have reported no use of oral contraceptives than control women (chi-square = 43.0, p = 0.000). Women with familial ovarian cancer had a similar history of oral contraceptive use to that of control women (chi-square = 1.32, p = 0.51).

The odds ratio for epithelial ovarian cancer for Jewish women compared with non-Jewish women, regardless of the women's family history of cancer was slightly, but not significantly, increased. The crude estimate of the OR was 1.40

(95% CI = 0.84 - 2.33). When we controlled for age, use of OCs for any duration, menopause status, age at menopause, education, and number of anovulatory cycles, the OR was 1.63 (95% CI = 0.96 - 2.77).

We classified case women as having either familial or nonfamilial ovarian cancer. The control women had no personal history of breast or ovarian cancer, but may have had a family history of breast, ovarian cancer, or endometrial cancer. We chose not to exclude women who had a family history of breast, ovarian, or endometrial cancer from the control group in order to maintain a representative sample of the general population. We found that the OR for familial ovarian cancer was 8.4 for Jewish women compared with non-Jewish women (Table 2). We present crude ORs because the numbers were too small for adjusted analysis. In contrast to the large increase in risk for familial ovarian cancer, we found only a small, nonsignificant, increased risk (unadjusted OR = 1.25) for nonfamilial ovarian cancer among Jewish women. When we adjusted for age, age at menopause, OC use, anovulatory cycles, education, and menopause status the OR increased slightly, but remained statistically nonsignificant OR = 1.42 (95% CI = 0.8 - 2.52). Among the control women, Jewish women were three times more likely to have a mother, sister, or daughter with ovarian cancer than non-Jewish women (OR = 2.9,

95% CI = 0.9 - 9.7). When we adjusted for age, education, and any duration of OC use, the OR increased to 3.6 (95% CI = 1.03 - 12.54). Jewish women were also more likely to have ever used OCs than were non-Jewish women (72% vs 62%,

chi square = 4.6, p = 0.032).

Finally, we examined the independent and combined effects of Jewish religion and family history on women=s risk for epithelial ovarian cancer (Table 3). The OR for ovarian cancer among non-Jewish women with a family history of epithelial ovarian cancer was three times higher than that among non-Jewish women with no family history. However, among Jewish women, the OR associated with having a family history of ovarian cancer was increased at least eight times higher than non-Jewish women with no family history. Although the risk for ovarian cancer among Jewish women with a family history of ovarian cancer (OR = 8.81) appeared to be substantially greater than among non-Jewish women with familial ovarian cancer

(OR = 3.01), these risks were not statistically different from each other, perhaps due to the small numbers of Jewish women with familial ovarian cancer.

The Cancer and Steroid Hormone (CASH) study has addressed multiple issues related to ovarian cancer in more than 12 published reports. In these reports, oral contraceptive use and parity were found to decrease women's risk for ovarian cancer [Gross and Schlesselman, 1994; Gross et al., 1992; Irwin et al., 1990; Gwinn et al., 1990; Irwin et al., 1991], and a family history of breast or ovarian cancer was found to increase their risk [Thompson et al., 1991; Bernstein et al., 1992; Amos et al., 1992; Schildkraut et al., 1989]. The odds ratio for epithelial ovarian cancer among women with first degree relatives with ovarian cancer was reported to be 3.6

(95% CI 1.8 - 7.1) and among women with second degree relatives with ovarian cancer it was reported to be 2.9 (95% CI 1.6 - 5.3) [Schildkraut and Thompson, 1988].

Our analysis of CASH study data shows that the increase in risk for ovarian cancer among Jewish women compared with non-Jewish women

(adjusted OR = 1.63) is not statistically significant. Our estimate of risk for familial ovarian cancer is derived from only 3 cases of familial ovarian cancer among Jewish women and must be considered preliminary to further research. The small overall increase in risk for ovarian cancer among Jewish women may be the result of a possible eightfold increase in risk for familial ovarian cancer. Since only 5% to 10% of epithelial ovarian cancers are believed to be hereditary [Szabo and King, 1995; Schildkraut and Thompson 1994; Narod et al., 1994] a large increase in familial epithelial ovarian cancer could account for a small overall increased risk of epithelial ovarian cancer in Jewish women.

One mutation that is believed to confer a 40% to 60% lifetime risk for ovarian cancer, the 185del AG, has been estimated to occur in 1% of Eastern European (Ashkenazi) Jews [Struewing et al, 1995]. This mutation has been estimated to cause 39% of the ovarian cancers in Ashkenazi Jewish women ages 50 and under and 12% of the ovarian cancers in non-Ashkenazi women ages 50 and under [Struewing et al., 1995] if the penetrance of the mutation is 90%. Women in this age group are well-represented in the CASH study, which included women ages 54 and under. The threefold increase in ovarian cancer caused by the 185delAG in a population-based survey [Struewing et al, 1995] may not account for the eightfold increase in familial ovarian cancer that we found among Jewish women. But because of the small number of Jewish women with familial ovarian cancer used to calculate the risk, the increased risk could have been as low as 2.6 or as high as 26.1. Other mutations to BRCA1, including the 5382insC and mutations to other genes such as BRCA2 and MSH2 [Szabo and King, 1995; Orth et al, 1994], as well as environmental exposures common to women in the same family may account for some portion of the cases of familial ovarian cancer among Jewish women.

Because of the small number of Jewish women in the study, we had limited power to detect differences between groups and limited ability to control for confounding factors. Because the average age of onset for familial ovarian cancer is approximately 47 years and for sporadic ovarian cancer approximately 59 years [Amos et al, 1992], women with sporadic cancer may be under-represented in the CASH study which had an upper age limit of 54 years.

In this analysis, we included only first-degree relatives in the definition of family history of cancer. The decision not to include second-degree relatives with cancer may have caused the family history of some women to be misclassified, but we did not believe that family history could be as reliably ascertained in second-degree relatives as in first-degree relatives. Family history can also be biased by family size, because a small family size does not afford an opportunity for expression of the number of cases that might be predicted and which would be needed to confirm probability estimates [Khoury et al., 1993].

Jewish women with a family history of epithelial ovarian cancer had an approximate eightfold increased risk for that cancer and non-Jewish women with such a family history had a threefold increased risk compared to control women. Although the number of cases is small, these results suggest an interaction between Jewish religion and family history. Confirmation of this interaction in larger studies would indicate that some factor or factors specifically related to family history of cancer among Jewish women (e.g. genetic or environmental influences) cause their increased risk for epithelial ovarian cancer to be greater than that among non-Jewish women with a similar family history of cancer.

OCs are known to protect against epithelial ovarian cancer [Whittemore et al, 1992], and there are data to suggest that OCs are protective against familial ovarian cancer [Piver et al., 1993]. We found 1) that despite the higher incidence of familial ovarian cancer, Jewish women were more likely to have used OCs than non-Jewish women; 2) that women with familial ovarian cancer used OCs at a rate similar to control women; and 3) that women with familial ovarian cancer used OCs at a rate higher than women with nonfamilial ovarian cancer. These three findings suggest that OCs do not protect against familial ovarian cancer in the same way that they have been shown to protect against ovarian cancer in general. Results of another analysis of the same data for effect modification by OCs suggested that the effect of family history of cancer was stronger among women who had used OCs for 3 months than among women who had not used OCs [Schildkraut and Thompson, 1988]. This effect modification is opposite to the protective effect of OCs on risk of ovarian cancer in general, supporting the suggestion that OCs were not protective against familial ovarian cancer in this population. A similar situation was found for parity. Increasing parity has been shown to decrease risk of ovarian cancer (Whittemore et al., 1992). The data analyzed here, however, showed that the magnitude of the odds ratios for family history of ovarian cancer increased by 40 percent with each pregnancy that a woman had. Others have reported similar results for the interaction between family history and parity [Narod et al, 1995].

In conclusion, the results of our analysis indicate that Jewish women are significantly more likely than non-Jewish women to have familial ovarian cancer. When we compared Jewish women and non-Jewish women who all had family histories of ovarian cancer, the Jewish women were more likely to develop ovarian cancer than the non-Jewish women, but differences were not statistically significant. These findings are similar to recent findings that Jewish women with a family history of breast cancer are at greater risk for breast cancer than non-Jewish women with a similar family history [Egan et al, 1996]. In both cases, we suspect that if cancer rates for familial breast and ovarian cancer are higher among Jewish women, then the increase may be due to their having higher rates of BRCA1 and BRCA2 mutations than women in the general population.

Because not all women with dominant-acting gene mutations, to BRCA1 for example, will develop ovarian cancer, understanding environmental and other genetic factors that interact with these mutations may prove to be important in prevention of these cancers in the future.

The authors wish to acknowledge the contributors to the Cancer and Steroid Hormone Study:

Study Design and Coordination: The Division of Reproductive Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia

Principal Investigator:
George L. Rubin, M.B., F.R.A.C.P.

Project Director:
Phyllis A. Wingo, Ph.D.

Project Associates:
Nancy C. Lee, M.D., Michele G. Mandel, B.A., Herbert B. Peterson, M.D.

Data Collection Centers Principal Investigators:
Atlanta: Raymond Greenberg, M.D.; Connecticut: J. Wister Meigs, M.D., and W. Douglas Thompson, Ph.D.; Detroit, Michigan: G. Marie Swanson, Ph.D.; Iowa: Elaine Smith, Ph.D.; New Mexico: Charles Key, M.D., and Dorothy Pathak, Ph.D.; San Franscisco: Donald Austin, M.D.; Seattle: David Thomas, M.D.; Utah: Joseph Lyon, M.D. and Dee West, Ph.D.

Pathology Review Principal Investigators:
Fred Gorstein, M.D., Robert McDivitt, M.D., and Stanley J. Robboy, M.D.

Project Consultants:
Lonnie Burnett, M.D., Robert Hoover, M.D., Peter M. Layde, M.D., M.Sc., Howard W. Ory, M.D., M.Sc., James J. Schlesselman, Ph.D., David Schottenfeld, M.D., Bruce Stadel, M.D., Linda A. Webster, M.S.P.H., and Colin White, M.B.B.S.

Pathology Consultants:
Walter Bauer, M.D., William Christopherson, M.D., Deborah Gersell, M.D., Robert Kurman, M.D., Allen Paris, M.D., and Frank Vellios, M.D.

The Cancer and Steroid Hormone Study was supported by interagency agreement 3-Y01-HD-8-1037 between the Centers for Disease Control and the National Institute of Child Health and Human Development, with additional support from the National Cancer Institute.

1. Amos CI, Shaw GL, Tucker MA, Hartge P. Age at onset for familial epithelial ovarian cancer. JAMA 1992;268:1896-99.
2. Bernstein JL, Thompson WD, Risch N, Holford TR. The genetic epidemiology of second primary breast cancer. Am J Eipdemiol 1992;136:937-48.
3. Egan KM, Newcomb PA, Longnecker MP, Trentham-Deitz A, Baron JA, Trichopoulos D, Stampfer MJ, Willett WC. Jewish religion and risk of breast cancer risk. Lancet 1996;347:1645-46.
4. Gross TP, Schlesselman JJ. The estimated effect of oral contraceptive use on the cumulative risk of epithelial ovarian cancer. Obstet Gynecol 1994;83:419-24.
5. Gross TP, Schlesselman JJ, Stadel BV, Yu W, Lee NC. The risk of epithelial ovarian cancer in short-term users of oral contraceptives. Am J Epidemiol 1992;136:46-53.
6. Gwinn ML, Lee NC, Rhodes PH, Layde PM, Rubin GL. Pregnancy, breast feeding, and oral contraceptives and the risk of epithelial ovarian cancer. J Clin Epidemiol 1990;43:559-68.
7. Irwin KL, Wingo PA, Lee NC. Agreement of self-reported ovarian number following gynecologic surgery with medical record reports. J Clin Epidemiol 1990;43:181-87.
8. Irwin KL, Weiss NS, Lee NC, Peterson HB. Tubal sterilization, hysterectomy, and the subsequent occurrence of epithelial ovarian cancer. Am J Epidemiol 1991;134:362-69.
9. Khoury MJ, Beaty TH, Cohen BH. Fundamentals of genetic epidemiology. Oxford University Press. New York, 1993:166-69.
10. Narod SA, Goldgar D, Cannon-Albright L, Weber B, Moslehi R, Ives E, Lenoir G, Lynch H. Risk modifiers in carriers of BRCA1 mutations. Int J Ca (1995)64:394-8.
11. Narod SA, Madlensky L, Bradley L, Cole D, Tonin P, Rosen B, Rixch HA. Hereditary and familial ovarian cancer in southern Ontario. Cancer 1994;74:2341-46.
12. Oddoux C, Struewing JP, Clayton CM, Neuhausen S, Brody LC, Kuback M, Haas B, Norton L, Borgeu P, Jahnwar S, Goldgar D, Ostrer H, Offit K. The carrier frequency of the BRCA2 6174delT mutation among Ashkenazi Jewish individuals is approximately 1%. Nature Genetics. 1996;14:188-90.
13. Orth K, Hung J, Gazdar A, Bowcock A, Mathis JM, Sambrook J. Genetic instability in human ovarian cancer cell lines. Proc Natl Acad Sci USA 1994;91:9495-91.
14. Piver MS, Baker TR, Jishi MF, Sandecki Am, Tsukada Y, Natarajan N, Mettlin CJ, Blake CA. Familial ovarian cancer. A report of 658 families from the Gilda Radner Familial Ovarian Cancer Registry, 1981-1991. Cancer 1993;71:582-88.
15. Roa BB, Boyd AA, Volcik R, Richards CS. Ashkenazi Jewish population frequencies for common mutations in BRCA1 and BRCA2 . Nature Genetics. 1996;14:185-7.
16. Schildkraut JM, Risch N, Thompson WD. Evaluating genetic association among ovarian, breast, and endometrial cancer: evidence for a breast/ovarian cancer relationship. Am J Hum Genet 1989;45:521-29.
17. Schildkraut JM, Thompson WD. Familial ovarian cancer: a population-based case-control study. Am J Epidemiol 1988;128:456-66.
18. Shattuck-Eidens D, McClure M, Sinard J, et al. A collaborative survey of 80 mutations in the BRCA1 breast and ovarian cancer susceptibility gene. JAMA 1995;273:535-41.
19. Struewing JP, Abeliovich D, Peretz T, Avishai N, Kaback MM, Collins FS, Brody LC. The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nat Genet 1995;11:198-200.
20. Szabo CI, King MC. Inherited breast and ovarian cancer. Hum Mol Genet 1995;4:1811-17.
21. Thompson, WD, Schildkraut JM. Family history of gynecologic cancers: Relationship to the incidence of breast cancer prior to age 55. Int J Epidemiol 1991;20:595-602.
22. Tonin P, Serova O, Lenoir G, Lybnch H, Durocher F, Simard J, Morgan K, Narod S. BRCA1 mutations in Ashkenazi Jewish women. Am J Hum Genet 1995;57:189.
23. Whittemore AS, Harris R, Itnyre J, Collaborative Ovarian Cancer Group. Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. Am J Epidemiol 1992;136:1184-1203.
24. Wingo PA, Ory HW, Layde PM, Lee NC, and the Cancer and Steroid Hormone Group. The evaluation of the data collection process for a multicenter, population-based, case-control study design. Am J Epidemiol 1988;128:206-17.
25. Young JL Jr, Percy CL, Asire AJ. Surveillance, epidemiology, and end results: incidence and mortality data, 1973-1977. Bethesda, MD: Department of Health and Human Services, 1981 (NIH publication no. 81-2330).