Survival Patterns after Oophorectomy in Premenopausal Women: A Population-Based Cohort Study
Abstract & Commentary
By Sarah L. Berga, MD, James Robert McCord Professor and Chair, Department of Gynecology and Obstetrics, Emory University School of Medicine, Atlanta, is Associate Editor for OB/GYN Clinical Alert.
Dr. Berga is a consultant for Pfizer, Organon, and is involved in research for Berlex and Health Decisions, Inc.
Synopsis: Risk of death was increased in women who had a bilateral oophorectomy before the age of 45 years but not in those who had a unilateral oophorectomy. The increased risk of death was also not seen in women who received estrogen treatment up to age 45 years.
Source: Rocca WA, et al. Survival patterns after oophorectomy in premenopausal women: a population-based cohort study. Lancet Oncol. 2006;7:821-827.
This study was made possible by the mayo Clinic Cohort Study of Oophorectomy and Aging, the main aim of which is to detect new cases of age-related diseases, especially Parkinson's disease and dementia. This report focused on the mortality outcomes. The cohort included women who underwent bilateral or unilateral oophorectomy and a group of referent women who did not undergo oophorectomy. All study participants were born before 1962 (were at least 40 years old by January 1, 2002) and resided in Olmsted County, Minnesota, which is mainly middle class, well-educated, and has excellent access to health care. Cases were carefully matched with referent women. The indication for oophorectomy had to be non-cancer and included benign ovarian conditions such as cysts, endometriosis, and benign tumors; about half had an oophorectomy prophylactically.
The follow-up procedures were exhaustive. Ascertainment of outcomes was high and participants were asked about hormone use. The final groups included 1091 women who had a bilateral oophorectomy, 1274 with a bilateral oophorectomy, and 1755 referent women. The Kaplan-Meier survival curves show decreased survival across time for women who underwent a bilateral oophorectomy. The hazard ratio for mortality in those who underwent bilateral oophorectomy was 1.93 (1.25-2.96, 79 cases) for women younger than 45 years, 1.02 (0.78-1.32, 243 cases) for those between 45 and 50 years, and 0.90 (0.68-1.19, 170 cases) for those older than 50 years. In contrast, the hazard ratio for mortality for those who had a unilateral oophorectomy younger than age 45 was 0.94 (0.65-1.37, 218 cases).
Further, analyses stratified by age at estrogen deficiency showed that the increased risk of death was restricted to women who had bilateral oophorectomy before the age of 45 years and did not receive estrogen treatment up to this age. Women who had estrogen deficiency before age 50 had a smaller but significantly increased risk of death. Women who underwent unilateral oophorectomy with hysterectomy had a survival advantage compared to referent women or women who underwent only a unilateral oophorectomy. Although the numbers are small, mortality for neurological or mental disorders was significantly increased in women who underwent bilateral oophorectomy before age 45 years, hazard ratio, 6.28 (1.83-21.5, 12 cases; P = 0.003). When this analysis was confined to those who had estrogen deficiency before age 45, the hazard ratio was 2.34 (9 cases, nonsignificant statistically).
This is a fascinating and provocative study that lends fuel to the hormone therapy fire. I look forward to future analyses, especially those which explore the role of estrogen deficiency upon the risk of dementia. The current results seem relatively straightforward. Bilateral prophylactic oophorectomy in unselected women (those not identified by molecular screening to be at high risk for breast and ovarian cancer) reduces survival if done in women younger than age 45 years, presumably because most women do not take hormones afterward and are thus exposed to estrogen deficiency. Said another way, premature estrogen deficiency seems to promote mortality.
The study is limited by the small number of cases, but it is otherwise carefully conducted, makes a valuable contribution to our fund of knowledge, and should hopefully alter practice patterns. For years, there have been two camps regarding the practice of prophylactic oophorectomy in women undergoing hysterectomy for benign indications. The glib answer was that the ovaries should be removed after age 35 years and the women placed on hormones, thus promoting survival and freeing them from the risk of ovarian cancer. However, because compliance is low, this strategy appears to not work as intended.
Further, the data herein fail to show a protective effect of prophylactic oophorectomy before menopause even when estrogen is instituted and continued until the anticipated time of menopause. It would appear that the only women who might benefit from a prophylactic oophorectomy are BRCA carriers. Bilateral oophorectomy with hysterectomy might be rarely needed for women with endometriosis, but this point remains more controversial. The present data do not directly address this group, although the authors noted that "survival was improved for women who underwent unilateral oophorectomy for endometriosis." The authors also pointed out that the Women's Health Initiative seems to have led some patients and physicians to conclude that women under age 50 who have hypoestrogenism due to ovarian removal or failure should not use hormones and thus these women may be unwisely exposed to considerable durations of hypoestrogenism. They finished by noting that "estrogen might be protective before menopause, fairly protective during and soon after menopause, and have no or negative effects if introduced 10-15 years after menopause." Thus, the hormone controversy rages and the wise practitioner is advised to avoid dogmatism. If future studies support the notion that estrogen is neuroprotective when begun at the time of menopause, it will tip the balance for many in favor of longer-term use.