PubMed1996
If instead of the title "Prophylactic Oophorectomy: Reducing the U.S. Death Rate from Epithelial Ovarian Cancer," the title were "Drug X Reducing the U.S. Death Rate from Epithelial Ovarian Cancer," there would be great media and medical attention worldwide to such a report. Correctly so. Regrettably, there probably is no new Drug X in the foreseeable future that will significantly reduce the death rate from ovarian cancer, be it Taxol®, taxotere, topotecan, gemcitabine, or liposomal doxorubicin-although each may result in significant responses and some prolongation of median survival. Epithelial ovarian cancer is a much more complex disease than anyone envisioned, when it was believed that extensive debulking surgery and the newest cytotoxic chemotherapy would radically reduce the death rate from ovarian cancer in the United States. Over 20 years after the first patient was treated with cisplatin for epithelial ovarian cancer, the annual death rate from ovarian cancer continued to increase. Just in the past decade, the number of women in the United States dying from ovarian cancer has increased 18% (Fig. 1) [1]. Although ovarian cancer is estimated to account for 26,700 cases and 14,800 deaths in 1996, it is a low-prevalence disease in comparison with breast cancer, which in 1996 is estimated to account for 185,700 cases and 44,560 deaths. Inexplicably, similar to breast cancer, the lifetime risk for ovarian cancer in the United States continues to increase. The most recent Surveillance, Epidemiology and End Results (SEER) calculations of lifetime risk for ovarian cancer are that 1 in 55 women will develop ovarian cancer over their lifetime, or 1.8%, up from the 1970 figures of 1 in 70, or 1.4% [2]. The 1.8% baseline lifetime risk for the general population is used to estimate the lifetime risk of known ovarian cancer risk factors (Table 1). Even utilizing what are now believed to be two of the most effective cytotoxic drugs against stage III and IV epithelial ovarian cancer, Taxol® and cisplatin, researchers reported that this resulted in an increase in the median disease-free survival of only five months, as compared with those women allocated to receive cisplatin and cyclophosphamide (median disease-free survival of 18 and 12.9 months, respectively), felt then to be the standard therapy [3]. Patients treated with Taxol® and cisplatin survived a median of 14 months longer than those treated with cisplatin and cyclophosphamide. These results may improve in women whose cancers were optimally debulked to </=1 cm residual disease, since the current study included only women with residual cancer greater than 1 cm. Given these sobering statistics, the public health issue is whether prophylactic oophorectomy for two select groups of women may be one measure in reducing the mortality from ovarian cancer in the United States. The two groups of women are: A) women age 40 or older who undergo hysterectomy for non-cancerous uterine conditions and B) those with a family history of ovarian cancer. PROPHYLACTIC OOPHORECTOMY AT THE TIME OF HYSTERECTOMY IN WOMEN AGE 40 OR OLDER: Concerning the first group of women, researchers from the University of Miami reported that 4.5%-14.1% of women develop ovarian cancer after prior hysterectomy for non-ovarian conditions (Table 2) [4-9]. Utilizing their University of Miami experience and the three other similar series in the literature, they were able to determine the number of ovarian cancer patients who had previously undergone hysterectomy at age 40 or older for non-cancerous uterine conditions. Of the 2,632 ovarian cancer cases, they predicted that 5.2% (138) of the ovarian cancers could have been prevented if prophylactic oophorectomy were performed in women 40 years or older at the time of hysterectomy for benign disease. Confirming these results is a recent survey of the American College of Surgeons which reported that of 12,316 cases of ovarian cancer studied, 18.2% of these women had had a previous hysterectomy for benign disease with ovarian preservation, 57.4% of which were performed in women over the age of 40 [10]. Prophylactic oophorectomy in such cases has not been a uniformly accepted practice because it has been assumed that preserved ovaries continue to function normally and thus delay both the onset of osteoporosis and menopause with its effect on the cardiovascular system. However, as many as 30% of women have postmenopausal symptoms within 24 months of hysterectomy with preservation of the ovaries [11]. In addition, in an important cross-sectional study of bone density of premenopausal women who had undergone hysterectomy with ovarian conservation compared with a matched group which had not had hysterectomy, investigators reported that bone density in the hysterectomy group was significantly reduced [12]. They concluded that premenopausal women who had undergone hysterectomy with ovarian preservation will have a significantly low bone density, somewhat negating the value of ovarian preservation. Women age 40 or over who undergo prophylactic oophorectomy at the time of hysterectomy for benign uterine conditions can be started immediately on hormone replacement therapy, thus helping to prevent loss of bone density and the onset of osteoporosis. PROPHYLACTIC OOPHORECTOMY IN WOMEN WITH A FAMILY HISTORY OF OVARIAN CANCER: Since familial (also known as site-specific or hereditary) ovarian cancer is considered to be an autosomal dominant inheritance with variable penetrance, sisters and daughters of families with a history of familial ovarian cancer (two or more first-degree relatives [mother, sister, daughter-who share one-half of one's genes]) may have as high as a 50% chance of inheriting the deleterious gene. Easton has estimated that for women with a BRCA1 mutation, the lifetime risk for developing ovarian cancer by age 70 is 63% [13]. The lifetime risk for BRCA2 mutations has not yet been calculated [14]. We have included women with two or more first-degree or a first- and second-degree relative with epithelial ovarian cancer as having a history of familial ovarian cancer. However, the Breast Cancer Linkage Consortium stated that "families with three or more cases of ovarian cancer are generally considered to be examples of hereditary ovarian cancer" [15]. In 1982, we first recommended prophylactic oophorectomy in women with a history of familial ovarian cancer. This definition included any woman with two or more first-degree or first- and second-degree relatives with epithelial ovarian cancer [16]. Because of what appeared to be an apparent increase in frequency of familial ovarian cancer in the 1970s, the Familial Ovarian Cancer Registry was formed in 1981 to further evaluate familial ovarian cancer syndrome. The Registry, renamed the Gilda Radner Familial Ovarian Cancer Registry in 1989 in memory of the late comedienne after her death from ovarian cancer, has registered as of June 1996, 1,376 families with ovarian cancer. A report on the first 1,000 families was recently published [17]. Previously, it was estimated that 5%-10% of all ovarian cancers are hereditary. However, scientists from the University of Pennsylvania believed the estimate of ovarian cancer cases that are hereditary may be as high as 20%, based on their recent report of germline BRCA2 mutations in women with ovarian cancer but without a family history of ovarian cancer [18]. In the CASH study, researchers compared 493 women with ovarian cancer to 2,465 matched controls [19]. The relative risk for ovarian cancer for any first-degree relative was 3.6, resulting in a lifetime risk of 6.5% (Table 1). However, if the mother was the first-degree relative with ovarian cancer, the relative risk was 4.3 and the lifetime risk was 7.5%. Although the latter risk is almost four times that of the general population, it is currently not recommended that women with only one first-degree relative with ovarian cancer undergo prophylactic oophorectomy. This recommendation would also be applicable to women with a strong family history of breast cancer but only one case of ovarian cancer in their family and to hereditary non-polyposis colorectal cancer (Lynch Syndrome II) if there were only one case of ovarian cancer. However, if a first-degree relative in a family with only one case of ovarian cancer carries the same BRCA1 or BRCA2 mutation as the germline mutation in the patient with ovarian cancer, prophylactic oophorectomy would be indicated. Using complex segregation analysis in evaluating the observed number of cases of ovarian cancer as compared with the expected number of cases during the same time period in England and Wales, researchers from the Ovarian Cancer Screening Clinic at Kings College Hospital in the United Kingdom studied 391 ovarian cancer family pedigrees [20]. The overall risk for any first-degree relative (8.1%) was almost exactly the same as that of the CASH study (Table 1). However, if the mother developed ovarian cancer before the age of 45, the relative risk was 14.2 with a calculated lifetime risk of 25%. This young age is consistent with an inherited cancer even with only one first-degree relative. If this study can be confirmed by other researchers, then prophylactic oophorectomy in women who have one first-degree relative who developed ovarian cancer before the age of 45 would indeed be indicated. As previously stated, we have used the definition of familial ovarian cancer as any family with two or more first-degree (mother, sister, daughter) or first- and second-degree relatives (grandmother, aunt) with epithelial ovarian cancer, and, further, that these women are candidates for prophylactic oophorectomy after age 35 if they have completed their family and desire the surgery. In contrast, the Breast Cancer Linkage Consortium studying inherited ovarian and breast cancer stated that "families with three or more cases of ovarian cancer are generally considered to be examples of hereditary ovarian cancer, and families with a total of five or more breast or ovarian cancers in first- or second-degree relatives qualify" [15]. In a study by one member of the Breast Cancer Consortium of hereditary or familial ovarian cancer in southern Ontario, Canada, the author stated that "our distinction between hereditary and familial ovarian cancer is arbitrary. We have included families with three ovarian cancers or five or more cases of ovarian cancer or breast cancer in the hereditary group" [21]. The U.S. NIH Consensus Conference on Ovarian Cancer in its 1990 report did not define site-specific ovarian cancer in its statement that "less than 0.5% of women are at a significant risk (for the development of ovarian cancer) because of three hereditary ovarian cancer syndromes: (1) breast/ovarian cancer syndrome; (2) site-specific ovarian cancer; and (3) hereditary non-polyposis colorectal cancer (Lynch Syndrome II), colorectal, endometrial, GI, kidney, and ovarian cancer" [22]. With the "arbitrary" definition of site-specific ovarian cancer by the Breast Cancer Consortium and the absence of a definition by the NIH Consensus Conference on Ovarian Cancer, we continue to define familial or site-specific ovarian cancer as two or more first-degree or first- and second-degree relatives with ovarian cancer. As seen in Table 3, it can be very difficult to assess the risk of familial ovarian cancer. PRIMARY PERITONEAL CARCINOMA AFTER PROPHYLACTIC OOPHORECTOMY FOR A FAMILY HISTORY OF OVARIAN CANCER: In 1982, scientists from the National Cancer Institute reported that three women with a familial history of ovarian cancer developed widespread adenocarcinoma "indistinguishable histopathologically from ovarian cancer" 1, 5, and 11 years after undergoing prophylactic oophorectomy [23]. These three cases were from 16 families in which 28 women had undergone prophylactic oophorectomy. This important report brought into question the value of prophylactic oophorectomy for women in such families. To study this problem utilizing members of the Gilda Radner Familial Ovarian Cancer Registry, we evaluated the women in the Registry from its initiation in 1981 through July 1992 [24]. At that time, the Registry had registered 931 families totaling 2,221 cases of familial ovarian cancer. Of the 324 women who had undergone prophylactic oophorectomy, 103 had two or more first-degree relatives, 166 had a first- and a second-degree or more relative, 41 had two or more second-degree relatives, and 14 had other relatives with ovarian cancer. Six, or 1.8%, of these 324 women developed primary peritoneal carcinoma indistinguishable from ovarian cancer 1, 2, 5, 13, 15, and 27 years after prophylactic oophorectomy. Of these six women, whose ages ranged from 43 to 75, five had two or more first-degree relatives with ovarian cancer and the sixth had three second-degree relatives with ovarian cancer (Figure 2). The Registry was able to confirm primary peritoneal carcinoma similar to ovarian adenocarcinoma and previously normal ovaries in all six cases. To date, there have been 12 cases of primary peritoneal carcinoma after prophylactic oophorectomy for a family history of ovarian cancer (Table 4). However, there have also been five cases of stage IA ovarian cancer diagnosed in women undergoing prophylactic oophorectomy for a family history of ovarian cancer. Although prophylactic oophorectomy will provide 100% protection against the development of ovarian cancer, it is clear that a very small fraction of such women will subsequently develop primary peritoneal carcinoma. Whether the inherited BRCA1 and BRCA2 mutations that result in familial ovarian cancer also cause primary peritoneal carcinoma is as yet unknown. All such women are followed closely post prophylactic oophorectomy by physical examination and CA125 blood testing. In conclusion, it is posited that prophylactic oophorectomy in women with a history of familial ovarian cancer who have completed their family by age 35 and desire prophylactic oophorectomy and in women age 40 or older who undergo hysterectomy for benign uterine conditions may result in a significant decrease in the death rate from ovarian cancer - a disease in search of a highly sensitive screening test(s) and improved therapy.