Premature or early bilateral oophorectomy: a 2021 update

Abstract

In this invited review, we discuss some unresolved and controversial issues concerning premature (<40 years) or early (40–45 years) bilateral oophorectomy. First, we clarify the terminology. Second, we summarize the long-term harmful consequences of bilateral oophorectomy. Third, we discuss the restrictive indications for a bilateral oophorectomy in premenopausal women to prevent ovarian cancer that are justified by the current scientific evidence. Fourth, we explain the importance of estrogen replacement therapy when the bilateral oophorectomy is performed. Hormone replacement therapy is indicated after bilateral oophorectomy until the age of expected natural menopause like in premature or early primary ovarian insufficiency. Fifth, we discuss the relationship between adverse childhood experiences, adverse adult experiences, mental health, gynecologic symptoms, and bilateral oophorectomy. The acceptance and popularity of bilateral oophorectomy over several decades, and its persistence even in the absence of supporting scientific evidence, suggest that non-medical factors related to sex, gender, reproduction, cultural beliefs, and socioeconomic structure are involved. We discuss some of these non-medical factors and the need for more research in this area.

Introduction and terminology

In this invited review, we discuss some unresolved and controversial issues concerning premature or early bilateral oophorectomy. This is not a systematic review, and we have based our comments on our direct clinical or research experience. Part of the confusion surrounding the practice of bilateral oophorectomy in premenopausal women arises from the imprecise use of words. If we consider the cessation of ovarian follicular activity occurring naturally at age 46 or older as physiologic or ‘normal’, a cessation of activity occurring before age 46 can be considered premature (<40 years) or early (40–45 years). Premature or early cessation of ovarian function may be primary, when no cause is known (idiopathic), or secondary to a known disease or to a medical or surgical intervention. Bilateral oophorectomy causes an abrupt and complete loss of ovarian function, and may have more extreme consequences than a premature or early ovarian insufficiency occurring naturally.

Bilateral oophorectomy is often incorrectly referred to as surgical menopause. Because menopause refers to 12 months or more of continuous lack of menstruation, the term menopause incorrectly implies that the uterus is the key organ [1,2]. The term menopause was introduced historically when the only way to infer the functioning of the ovaries was through their effect on the uterus (endometrium). Indeed, if the uterus is removed but the ovaries are conserved, the woman will undergo surgical menopause, but not cessation of ovarian function. The time of cessation of ovarian function may be difficult to determine in the absence of the uterus and of menstruation [2]. Therefore, a clear separation of the cessation of uterine function from the cessation of ovarian function may be a useful clarification for both clinicians and the lay public.

We use the term bilateral oophorectomy rather than bilateral salpingo-oophorectomy (BSO) because BSO implies simultaneous removal of the ovaries and the fallopian tubes, and we believe that surgical removal of each organ must be justified separately by scientific evidence. In more recent times, it has been advocated to remove the tubes rather than the ovaries to prevent ovarian cancer [3,4]. However, the cancer prevention efficacy of this so-called ‘opportunistic salpingectomy’ remains controversial [5,6]. We will focus our discussion on bilateral oophorectomy with or without the removal of other pelvic organs (uterus, tubes, or other structure) because we are most interested in the disruption of the endocrine part of the reproductive system.

As with most biological events, the transition from premature to early, and from early to physiologic (or ‘normal’) ovarian insufficiency is continuous rather than discrete. Therefore, all cut-off ages are arbitrary and based on the conventional use of 5-year age groups in the medical literature. In agreement with most authors, we use the term premature ovarian insufficiency when the insufficiency occurs before the age of 40 years (<40). On the other hand, we use the age of ≤45 years as a cut-off between ‘early’ and ‘physiologic’ menopause. Some authors use the cut-off of <45 years, and in some studies it is unclear whether the 45th year is included or not. We invite the authors reporting future studies to be precise about the cut-off that they use.

We use the term estrogen replacement therapy (ERT) in the context of bilateral oophorectomy. If a woman experiences surgically induced cessation of ovarian function and receives hormonal therapy (most commonly, unopposed estrogen if the uterus was also removed), the therapy should be considered a ‘replacement’ of the estrogen that was artificially removed with the oophorectomy. The use of estrogen in these women is equivalent to giving replacement thyroid hormone to a woman who underwent a removal of the thyroid gland to treat cancer. By contrast, the term estrogen therapy (ET) may be more correctly used when a woman has reached natural menopause at age 46 or older and is prescribed estrogen to treat menopausal symptoms [7,8].

Long-term harmful consequences of bilateral oophorectomy

The practice of ‘elective’, ‘incidental’, or ‘prophylactic’ bilateral oophorectomy at the time of hysterectomy was very common in the 20th century and in the first decade of the 21st century. The rationale for the surgery was that the ovaries are the source of ova, and are not needed beyond reproduction. The mistake was to consider the ovaries primarily a reproductive organ, rather than an endocrine organ. The ovarian hormones, primarily estrogen and progesterone, were considered important only for their effects on the reproductive organs (e.g. uterus or breast). The effects of estrogen and progesterone on tissues, organs, and systems outside of the reproductive system were not considered important (e.g. effects on the cardiovascular system, on the bones, or on the brain). In addition, the effects of ovarian androgens were not considered at all, neither before nor after menopause [6].

Because ovarian cancers are often detected late, when the subsequent prognosis is poor, prophylactic removal of the ovaries was considered justified, even if the ovaries were completely healthy. If a bilateral oophorectomy before menopause had no negative endocrine sequelae, the surgery would have an excellent cost-to-benefit ratio. However, around 2005, Parker and colleagues raised the possibility that the cost-to-benefit ratio was against bilateral oophorectomy because there was a significant reduction in overall survival following oophorectomy [9,10]. The harmful sequelae of bilateral oophorectomy have subsequently been documented in three major studies, listed here in chronological order: (1) the Mayo Clinic Cohort Studies of Oophorectomy and Aging (2006 onward), (2) the Nurses’ Health Study (2009 onward), and (3) the national UK database (2017) [11–16]. Even though these three studies had important methodological differences, they produced a broad body of consistent evidence against the use of prophylactic bilateral oophorectomy in women at average risk of ovarian cancer.

The Mayo Clinic project included two separate studies, the Mayo Clinic Cohort Study of Oophorectomy and Aging 1 and 2 (MOA-1 and MOA-2). MOA-1 included women who underwent bilateral oophorectomy in 1950–1987 and MOA-2 in 1988–2007 [17]. In both studies, both women who underwent the bilateral oophorectomy and the referent women were representative of a geographically defined population (Olmsted County, MN, USA). Referent women were a random sample of all women in the general population who had not undergone bilateral oophorectomy at the time of sampling (index date). Therefore, the majority of the referent women had not undergone any gynecologic surgery. Details about indication, actual performance of the surgery, and pathology were abstracted from medical records [17,18].

The MOA-1 study showed that bilateral oophorectomy at age <45 years was associated with a significantly increased overall mortality, cardiovascular mortality, and mortality for neurological and mental disorders [11,19,20]. In addition, bilateral oophorectomy was associated with an increased incidence of cognitive impairment or dementia (≤48 years), parkinsonism, depressive symptoms, anxiety symptoms, and glaucoma (<43 years) [21–24]. The risk was partly reduced by ERT for overall mortality, cardiovascular mortality, and incidence of cognitive impairment or dementia, but not for incidence of parkinsonism, depressive symptoms, anxiety symptoms, or glaucoma (suggestive differences; the sample was inadequate for formal statistical testing).

The MOA-2 study showed that bilateral oophorectomy at age <46 years was associated with an increased risk of incident depression, hyperlipidemia, cardiac arrhythmias, coronary artery disease, arthritis, asthma, chronic obstructive pulmonary disease, and osteoporosis. In addition, bilateral oophorectomy was associated with an accelerated rate of accumulation of 18 chronic conditions considered together as a multi-morbidity score [12,13]. Several of these associations were partly reduced in women who received ERT up to age 46 years or longer (suggestive differences; the effect was statistically significant only for osteoporosis). Analyses considering chronic conditions separately showed an association with chronic kidney disease and restless leg syndrome [25,26]. The association was partly reduced in women who received ERT up to age 46 years or longer for chronic kidney disease but not for restless leg syndrome (suggestive differences). Finally, bilateral oophorectomy was associated with medial temporal lobe structural abnormalities in a brain imaging study [27].

The Nurses’ Health Study (NHS) included a large sample of volunteer women from throughout the United States. The women who underwent hysterectomy with bilateral oophorectomy were compared to referent women who underwent hysterectomy with ovarian conservation. The type of gynecologic surgery was self-reported by the women. The average follow-up was 24 years. Bilateral oophorectomy was significantly associated with an increased risk of total mortality and fatal plus non-fatal coronary heart disease. The authors also reported a significantly reduced incidence of breast cancer, ovarian cancer, and total cancer; however, the incidence of lung cancer and total cancer mortality was increased. Women who underwent bilateral oophorectomy < 50 years and did not receive ERT experienced a significantly increased risk of all-cause mortality, coronary heart disease, and stroke [14].

In a second report from the NHS focusing on long-term mortality (28 years of follow-up), the authors reported an association with lower risk of ovarian cancer mortality and breast cancer mortality (<47.5 years). However, bilateral oophorectomy at all ages was associated with a significant increase in mortality for all causes, lung cancer, colorectal cancer, total cancer, coronary heart disease, and cardiovascular disease. In women who underwent bilateral oophorectomy <50 years and never used ERT, there was an increased risk of death for all causes [15]. Several of the associations were attenuated by ERT (suggestive differences).

The UK study was nationwide and used the English Hospital Episode Statistics database and national registers of death and deprivation. The women who underwent bilateral oophorectomy with hysterectomy were compared to referent women who underwent hysterectomy with conservation of at least one ovary (ovarian conservation). Women with ovarian conservation had a lower risk of hospital admission for ischemic heart disease and for cancer. In addition they experienced lower mortality for all causes, heart disease, and cancer [16]. Data about ERT were not available.

In conclusion, bilateral oophorectomy in premenopausal women is a unique iatrogenic event that causes an abrupt and complete endocrine disruption. It has been postulated that the abrupt loss of ovarian hormones causes an alteration of several fundamental aging processes at the cellular, tissue, organ, and system levels, leading to clinical manifestations such as multi-morbidity, frailty, and reduced survival. It has been shown clinically and confirmed using DNA biomarkers (DNA methylation) that estrogen deprivation in a premenopausal woman accelerates the overall aging processes [6,8,12,13,28–32].

When should bilateral oophorectomy be performed?

Our recommendation is that bilateral oophorectomy should not be considered in women who have not reached natural menopause, and in particular in women who are younger than 46 years, with the exception of women who have a primary or metastatic ovarian cancer, and in women who carry a high genetic risk of ovarian cancer. The most common high-risk variants are in the BRCA1 or BRCA2 genes. Fortunately, these variants are rare in the general population. Having a positive family history for ovarian cancer is not considered sufficient to justify oophorectomy, but it may justify undergoing genetic testing. Figure 1 provides a simple summary of the steps involved in reaching a decision about bilateral oophorectomy. Three key questions remain partly unanswered:

1. Who should be tested?

2. Who should undergo a bilateral oophorectomy?

3. Who should receive estrogen replacement therapy [31]?

Figure 1.

Flow chart of decision making for women considering bilateral oophorectomy for the prevention of ovarian cancer.

Screening all women in the general population for BRCA-1 or BRCA-2 variants is controversial and currently not recommended. Indeed, in 2014, the US Preventive Services Task Force recommended against genetic testing and counseling for women without a positive family history [8,33]. Unfortunately, the definition of positive family history is imprecise. For example, a positive family history among first-degree relatives depends on the number of relatives at risk, and in the case of a disease restricted to women, on the number of sisters. Some authors have recommended the use of a clear operational definition of positive history, for example, two or more first-degree relatives with confirmed epithelial ovarian cancer [34]. Unfortunately, women may not know the precise age at diagnosis or the histologic type. It often remains unclear whether a woman with a reported positive family history of breast or ovarian cancer should or should not be tested, and personal characteristics of the woman (e.g. reproductive history, marital status, education, personal preference) or the preference of the physician (e.g. training, acceptance of published guidelines, experience from previous patients, or financial incentives) may influence the decision.

If the genetic test is positive, the woman must decide whether or not to take a preventive action against the risk of breast and ovarian cancer (Figure 1). It is difficult for a woman to take action against a hypothetical disease (a percent risk) rather than a present disease. A woman with the BRCA-1 variant has a 72% risk of breast cancer and a 44% risk of ovarian cancer by age 70 years; with the BRCA-2 variant the risks are 69% for breast cancer and 17% for ovarian cancer [8,31]. A bilateral oophorectomy performed at the recommended ages of 35–40 years for BRCA-1 and 40–45 years for BRCA-2 may significantly reduce the risk of cancer. Unfortunately, women who undergo bilateral oophorectomy before age 46 are the women at highest risk for long-term morbidity and mortality sequelae other than cancer [12,13,29,31].

Women and their doctors must trade one set of risks (cancer) for another set of risks (non-cancer morbidity and mortality, primarily cardiovascular disease and dementia). From the perspective of the overall population, it can be argued that the preventive practice should be designed to maximize the reduction of cancer, of other chronic disease outcomes, or of both. These objectives can be combined in a formal cost-effectiveness analysis [35]. However, at the individual level, it remains difficult and counterintuitive to translate a percent risk into a decision to undergo a surgery that has the potential to cause major long-term problems. For example, is 17% risk of ovarian cancer by age 70 years in a woman with a BRCA-2 variant sufficient to justify a bilateral oophorectomy at age 40–45 years? At what percent risk does a bilateral oophorectomy became cost-effective [6]?

In this complicated landscape, it would be useful to individualize the decision to perform bilateral oophorectomy by identifying additional factors that predict which women with a BRCA-1 or BRCA-2 variant will develop cancer, and by identifying additional factors that predict major morbidity other than cancer after the surgery [8,29,31]. Alternatively, a more radical solution is the development of an early and non-invasive biomarker of early-stage ovarian cancer (e.g. a very sensitive blood marker or imaging technique). Development of such biomarkers is an important area for future work.

Is estrogen replacement therapy recommended?

If the woman and her doctor have decided to perform the bilateral oophorectomy, they need to implement a preventive program for the harmful sequelae of the abrupt loss of ovarian hormones. One approach is to utilize an adequate dose of ERT for an adequate length of time (Figure 1). However, it may be difficult for a woman to commit to taking ERT for 10 or more years to prevent a disease that is not present, and that may or may not develop. The obstacles to accepting the therapy may be cost, interference with lifestyle, side effects, and fear that ERT may cause other diseases. The two major problems with using ERT are increased risk of thromboembolism and cancer, primarily breast cancer. If a positive family history of cancer or a general fear of cancer were major determinants of the bilateral oophorectomy, these same fears may be obstacles to ERT compliance.

Unfortunately, the results of the Women’s Health Initiative (WHI) clinical trials have had a profound and lasting effect on the use of ET even in women who underwent premature or early menopause [36,37]. Moreover, the generalization of the results of the WHI estrogen plus progestin trial to the unopposed estrogen trial conducted in women who underwent hysterectomy has obscured the data. Indeed, unopposed estrogen initiated before age 60 years in women who underwent bilateral oophorectomy and hysterectomy reduced overall long-term mortality and produced a favorable safety profile in the WHI trial [38]. Nevertheless, confusion and uncertainty remain, and long-term compliance may be low [39–41].

In addition, scientific evidence about the optimal dose, route of administration, or length of treatment for ERT is not available. With some major simplification, a common recommendation is to treat women after bilateral oophorectomy with transdermal 17β-estradiol at 100 μg/day (or higher), or with a therapeutic equivalent, starting immediately after surgery and continuing up to age 50–51 years (see Figure 1) [32]. The objective of dosing ERT is to achieve a blood level of estradiol approximating the physiologic level of the early follicular phase. However, the optimal dose may vary across women (individualized optimal dose) [32,42,43].

These recommendations are based on judgement and extrapolation, but have not been tested with adequate clinical trials. In addition, studies are needed to identify genetic variants, or other personal characteristics that predict the optimal dose of ERT for the prevention of harmful sequelae (individualized prevention). The cost-to-benefit ratio of continuing estrogen treatment after age 50–51 years remains less clear and falls under the general guidelines for treatment of menopausal symptoms after natural menopause [7,8,31].

Is mental health related to bilateral oophorectomy?

Despite the growing body of evidence that bilateral oophorectomy has severe sequelae and that the cost-to-benefit balance is unacceptable for women at average risk for cancer, many women and their physicians continue to elect bilateral oophorectomy [6,9,10,29]. Therefore, it is important to understand the historic justification of the practice and why it persists even in the absence of scientific justification.

From the perspective of the gynecologists, their training and established practice have been primarily focused on other gynecologic diseases (e.g. ovarian cancer). Additionally, because gynecology is a surgical specialty, and the long-term sequelae may develop more than a decade after surgery, these women would report their sequelae to community medicine, internal medicine, psychiatry, or menopausal clinics [6]. Historically, gynecologists have not been aware of the long-term sequelae of their practice.

From the perspective of the women, several factors may play a role in their willingness to undergo a major surgery and the removal of their reproductive organs. Some of these factors are non-medical, and may be related to sex, gender, reproduction, mental health, cultural beliefs, and socioeconomic structure. For example, there may be some interaction between mental health issues and gynecologic surgeries. One study showed that women who experienced adverse childhood events were at higher risk of undergoing bilateral oophorectomy [44,45]. Similarly, anxiety, depression, somatoform disorders, and personality disorders, alone or in combination, were found associated with a higher risk of undergoing a bilateral oophorectomy [46].

Unfortunately, the women may or may not be aware of the possible link between mental health and gynecologic health. In addition, very often, the women are in a state of experiential avoidance of their past traumatic experiences. This condition is called disavowal in the psychoanalytic literature [44]. Also, very often the primary-care doctors, the gynecologists, and the psychiatrists are not adequately trained and experienced in recognizing the link between emotional experiences, mental health, and gynecologic health. Based on our work, we suggest that there may be four major groups of women who undergo bilateral oophorectomy in the general population, as illustrated in Figure 2.

Figure 2.

Identification of four subgroups of women with different combinations of mental health conditions and gynecologic symptoms. Group 1: neither adverse childhood experiences nor psychiatric diagnoses. Group 2: both adverse childhood experiences and psychiatric diagnoses. Group 3: adverse childhood experiences but not psychiatric diagnoses. Group 4: no adverse childhood experiences but psychiatric diagnoses.

In a first group (Figure 2, group 1), women have no history of recognized psychiatric conditions or adverse childhood experiences and may reach the decision to undergo the surgery because they genuinely experience severe symptoms that did not respond to the available treatments, or were not treated adequately. Some of these women have a confirmed uterine or pelvic condition that is believed to explain the symptoms (e.g. a fibroma or endometriosis). In these situations, the surgery may be perceived as the only solution to the symptoms. However, the possible control of gynecologic symptoms must be weighed against the risk of long-term somatic and mental sequelae.

In a second group (Figure 2, group 2), women suffered one or more childhood adverse events, and some women later also experienced adult adverse events. For example, women who were exposed to emotional, physical, or sexual violence during childhood may later in life select partners who perpetrate the same violence. Some of these women may later develop psychiatric disorders, and the combination of childhood or adult adverse events and psychiatric disorders may lead to the development of gynecologic symptoms, to an exaggerated response to gynecologic symptoms (e.g. excessive bleeding), or to a more general lower threshold to pain.

In a third group (Figure 2, group 3), women with adverse childhood events did not develop psychiatric disorders, or the disorders were not recognized by the care providers. In these women the association appears to be directly between the childhood events and development of gynecologic symptoms, exaggerated response to gynecologic symptoms, or reduced tolerance to pain, without the mediation of psychiatric conditions. Many of these women may not report having suffered from the adverse childhood events when they are asked directly at the time of referral for gynecologic symptoms or pain. Some of the adverse events may only be reported in the context of a major psychiatric experience, such as a major depressive episode or an attempted suicide. Therefore, these events may only be recognized through review of historical medical records [44,45].

Finally, in a fourth group (Figure 2, group 4), women did not experience childhood adverse events, or the events were never reported. Some of these women later developed psychiatric conditions, and these conditions may have led to the development of gynecologic symptoms, to an exaggerated response to the symptoms, or to a lower threshold to pain. This mechanism is well illustrated by women who developed a somatoform disorder and who repeatedly sought treatment from several care providers for symptoms that did not correspond to a recognized clinical condition. A number of these women underwent repeated surgeries in the abdomen and pelvis [46].

Conclusions

The history of bilateral oophorectomy is a good example of the continuous need to evaluate the practice of medicine and surgery over time [6]. A medical or surgical treatment that may be considered adequate today may become obsolete, or be recognized as harmful tomorrow. Unfortunately, the history of medicine includes a long list of these ‘errors’, and women with gynecologic disorders should be offered safer and less invasive treatment options [10].