Examining the Evidence
Q Does prophylactic oophorectomy raise the risk of death?
<huc>A</huc> Maybe. Women who undergo bilateral oophorectomy before the age of 45 have significantly higher mortality, especially when no estrogen is given, than women who do not have their ovaries removed. However, it is unclear whether the relationship between bilateral oophorectomy and increased mortality is causal or merely a marker of underlying risk.
A longstanding controversy in gynecologic practice is whether the ovaries should be removed at the time of abdominal hysterectomy. Depending on the patient’s age, this question requires the clinician to weigh the risk of subsequent ovarian cancer against the benefit of protection against cardiovascular disease and osteoporosis. During my training in the late 1970s, prophylactic oophorectomy was recommended at the time of hysterectomy if the patient was older than 40 years. The most recent guidelines (1999) from the American College of Obstetricians and Gynecologists1 state: “The decision to perform prophylactic oophorectomy should not be based only on age; it should be a highly individualized decision that takes into account several patient factors and choices.”
High risk of heart disease versus low risk of ovarian cancer
Given that the number of women older than 40 who will die of heart disease is vastly greater than the number who will die of ovarian cancer, even a small protective effect against heart disease from the retention of estrogen-producing ovaries might outweigh the potential risk of ovarian cancer. Other variables influence the outcome—eg, ovarian conservation reduces hip fracture but increases breast cancer—but the main drivers of overall outcome are heart disease and ovarian cancer.
How this study explored the issue
One way to address the question of risk is to model various outcomes using simulation methods, as discussed in a recent issue of Obg Management.2 Another way is to analyze retrospectively the survival across time of women who did or did not undergo oophorectomy. The study by Rocca and colleagues is such an analysis, using data from women residing in Olmstead County, Minnesota, between 1950 and 1987.
Women were included as cases if they had oophorectomy between 40 years and menopause; controls were age-matched women in the database who had survived without oophorectomy to the same index year as the case. Using these criteria, almost 1,100 women were identified who underwent bilateral oophorectomy between 40 years and menopause.
Life-table analysis revealed no difference between cases and controls in survival across time (hazard ratio, 1.05; 95% confidence interval, 0.92–1.20). However, a subgroup of 79 women who had bilateral prophylactic oophorectomy between 40 and 45 years, but who were not given estrogen, were estimated to have twice the mortality of their controls across time. Similarly, a group of 183 women who underwent bilateral oophorectomy for benign disease between 40 and 45 years had a 50% higher mortality. None of the subgroups of women who underwent bilateral oophorectomy after 45 years had increased mortality in comparison with controls.
What this means for clinical practice
It is safe to say that data from a randomized, controlled trial will not be forthcoming anytime soon. Therefore, we must rely on the careful analysis of retrospective data, as in the study by Rocca et al. From this analysis, 2 recommendations can be drawn:
- After age 45. It appears that oophorectomy after age 45 will not alter the subsequent overall mortality risk. For this age group, the decision to remove or retain the ovaries should be made on an individual basis depending on the risk profile and informed patient choice, as suggested by American College of Obstetricians and Gynecologists guidelines.1
- Prior to age 45. If bilateral oophorectomy is performed, estrogen replacement should be strongly considered. In contemporary practice, if alternatives to estrogen are desired, the patient should be monitored for evidence of preclinical cardiovascular disease and osteoporosis, and appropriate treatment should be initiated, if indicated.