Elsevier

Maturitas

Volume 72, Issue 3, July 2012, Pages 214-219
Maturitas

Review
Potential mechanisms of postmenopausal endometriosis

https://doi.org/10.1016/j.maturitas.2012.04.010Get rights and content

Abstract

Endometriosis is a chronic gynaecological disorder, the cause of which remains a subject of controversy. Oestrogen dependence is considered central to development and progression, and endometriosis is widely viewed as a disease of the premenopausal years, which normally regresses during the menopause. Increasingly however, reports of cases of postmenopausal endometriosis challenge our current understanding of the pathophysiology and raise further questions concerning the processes involved. Exploring the limited evidence available on postmenopausal disease we attempt to draw comparisons with pre-menopausal endometriosis, and in doing so to propose mechanisms for postmenopausal disease that are compatible with our current general understanding of the condition.

Introduction

Endometriosis is a common, chronic gynaecological disorder affecting an estimated 10% of women of reproductive age [1]. Defined morphologically by the presence of endometrial glands and stroma outside of the uterus, endometriosis is a cause of chronic pelvic pain and infertility [2], [3]. To date there is still no single, unifying theory to explain the existence of endometriosis in all its various forms [4], [5]. However, oestrogen dependence is considered central to the pathophysiological process, and the persistence of these lesions [6]. This concept, combined with clinical data demonstrating reduction in disease burden and symptoms with oestrogen lowering medication, has lead to the widely held belief that endometriosis is a disease of premenopausal women and ‘cured’ by menopause [7], [8].

Several theories have been proposed to attempt to explain the pathophysiology of premenopausal disease. Of these Sampson's theory of retrograde menstruation and implantation is the most widely accepted [9]. However, the presence of endometrial cells in the abdominal cavity is frequently observed in women during menses [10], and this theory therefore fails to explain why some women develop endometriosis while others do not [8]. It is likely that additional factors determine the ability of these endometrial deposits to implant, proliferate and persist. It has been suggested that molecular aberrations in the eutopic endometrium of women with endometriosis [11] are likely to result in the activation of oncogenes [12], and progesterone resistance [13]. It is possible that alterations in the immune system resulting in failure to clear peritoneal implants might also play a central role in the pathophysiology [14], [15]. Additionally it has been proposed that endometriosis is an angiogenesis-dependent disease, an imbalance in pro- and anti-angiogenic factors facilitating the growth and establishment of lesions [16], [17], [18].

In keeping with the concept of oestrogen dependence, high levels of aromatase and steroidogenic acute regulatory protein (StAR) expression and enzyme activity have been demonstrated in cultured stromal cells taken from endometriotic lesions, under prostaglandin E2 (PGE2) stimulation. In contrast aromatase activity was undetectable in eutopic endometrial stromal cells from disease-free controls [8], [19], [20]. StAR facilitates the entry of cholesterol into the mitochondrion, thus initiating oestrogen production [8]. In fact in vivo studies have indicated that endometriotic lesions express the full compliment of enzymes required for oestrogen synthesis [21], and it is possible that locally produced oestrogen further drives the disease process in endometriosis through paracrine effects [6].

Contrary to popular belief, post-menopausal endometriosis is thought to affect up to 2–4% of women [22]. However, the data on postmenopausal disease is currently limited, and largely confined to case reports and retrospective studies. Reports of disease arising in post-menopausal women challenge current concepts and raise important questions regarding the mechanisms by which these lesions might develop and/or persist during the post-menopausal years. Central to this debate are general issues regarding the natural history of premenopausal disease, and specifically whether or not endometriosis can be considered progressive in nature. In one study disease progression at 6 years was demonstrated in less than 10% of women with asymptomatic, untreated, laparoscopically confirmed rectovaginal endometriosis [23]. Furthermore it has been suggested that mild endometriosis is not a disease at all, rather a transient phenomenon occurring in healthy individuals, which is capable of resolving spontaneously [2]. In a double-blind placebo-controlled trial of treatment with oral gestrinone, infertile women with laparoscopic evidence of asymptomatic endometriosis were allocated to receive either oral gestrinone or a placebo and followed-up with a second-look laparoscopy at 24 weeks. Remarkably even among women in the placebo group, disease appeared to be eliminated in 24% and showed improvement in 29% [24]. It is important to note however that high inter- and intra-observer variability has been reported when disease severity is staged by visual inspection [25], [26].

If endometriosis can still be considered a progressive disease, the detection of postmenopausal lesions does not directly contradict Sampson's retrograde menstruation and implantation theory because lesions arising in this way during the premenopausal years might simply progress and persist into the menopause. In contrast, if endometriotic lesions arise de novo in postmenopausal women in the absence of a functioning endometrium, it may be assumed that these lesions develop as a result of alternative processes. In this review, we will attempt to explore the pathophysiological processes responsible for postmenopausal disease, and based on current evidence, we propose potential mechanisms for post-menopausal endometriosis.

There is evidence to suggest that post-menopausal endometriosis is morphologically similar to pre-menopausal disease, albeit less extensive and less active [7]. This evidence is derived from a retrospective study on excised ectopic endometrial tissue from over 100 women, grouped according to age as ‘likely pre-menopausal (<50 years) and ‘likely post-menopausal’ (≥50 years). The similar distributions of oestrogen and progesterone receptors in ectopic endometrial tissue taken from across age groups, suggests that endometriotic lesions might have the potential to reactivate given the appropriate hormonal stimulus [7], [27]. Although unusual in its approach, particularly in grouping patients by age rather than by confirmed menopausal status, the study implies that post-menopausal endometriosis is likely to share some of the features normally associated with pre-menopausal disease.

However, lesions from women in the ‘likely post-menopausal’ group were found to be less haemorrhagic and less extensive compared to women in the ‘likely pre-menopausal’ category. These findings are in keeping with the concept that endometriotic lesions should normally regress after the menopause. This forms the basis of gonadotrophin-releasing hormone (GnRH) agonist therapy for the management of endometriosis, which is thought to suppress disease activity in affected individuals by generating an artificially induced menopausal state. It is supported by studies demonstrating reduced inflammation and angiogenesis, and increased apoptosis in endometriotic lesions derived from women receiving systemic GnRH agonists [28].

Several cases of post-menopausal endometriosis described in the literature identify a history of pre-menopausal disease. A post-menopausal 53 year-old woman presenting with frank haematuria was discovered to have a retroperitoneal mass compressing the right distal ureter [29]. 8 years previously, during her pre-menopausal years, the patient had undergone a total abdominal hysterectomy and bilateral salpingo-oophorectomy (BSO) for uterine fibroids and extensive pelvic endometriosis. Pelvic adhesions involving the small bowel and right adnexa had been identified at the time [29]. When surgical resection of the newly identified retroperitoneal mass was attempted, severe and extensive endometriosis was once again identified, this time involving the distal ileum, caecum, and inferior vena cava, with widespread adhesions of the anterior abdominal wall, small bowel, sigmoid colon and rectum [29].

In another report a 57 year-old woman who developed severe and recurrent post-menopausal disease had already undergone a hysterectomy and bilateral salpingo-oophorectomy (BSO) for confirmed endometriosis at the age of 35 [30]. It is possible that lesions arising prior to the onset of the menopause might account for the discovery of disease during the post-menopausal years, even in the absence of a functioning endometrium providing a source of new endometrial fragments. If all post-menopausal disease is indeed associated with pre-menopausal disease, Sampson's theory for the pathophysiology of endometriosis [9] is not necessarily incompatible.

There is, however, evidence to contradict this idea. Extensive transmural intestinal endometriosis was discovered in a 74 year-old woman presenting with symptoms of bowel obstruction, who reported no premenopausal history of symptoms consistent with endometriosis [31]. However, the patient had not undergone any procedures during the pre-menopausal period, which could have provided imaging or direct visualisation of the pelvic cavity prior to the onset of the menopause. In another case, a 65 year-old female who presented with haematuria and pelvic pain had developed post-menopausal vesical endometriosis, which was severe and recurrent [32]. The patient had undergone a total abdominal hysterectomy during the pre-menopausal years for uterine fibroids. There was no mention of a diagnosis of endometriosis at the time; however, this is not explicitly stated in the report [32]. It is therefore possible that post-menopausal endometriosis might not be universally associated with pre-menopausal disease. However, as these cases demonstrate, it is not always easy to establish a clear history of pre-menopausal disease in affected women.

It has been suggested that endometriosis might be detected laparoscopically in up to 7% of multiparous, asymptomatic women [33]. Because the presence of endometriotic lesions can be symptomless in some women, endometriosis can be an incidental finding at surgery for other indications [33]. Therefore, if symptoms were to arise for the first time during the post-menopausal years, a lack of pre-menopausal endometriosis might be incorrectly assumed particularly in the absence of pre-menopausal investigations. Laparoscopic inspection of the pelvic cavity remains the gold standard investigation for the diagnosis of endometriosis due to a lack of biomarkers with sufficient sensitivity and specificity [34], [35]. Women with post-menopausal disease who presented during the premenopausal years with symptoms consistent with endometriosis, but did not undergo laparoscopic investigation at the time, form a group in whom pre-menopausal disease might be suspected but cannot be confirmed retrospectively. Furthermore, while the natural history of endometriosis is poorly understood, there is strong evidence to suggest that lesions might arise as a transient phenomenon in some women [2], [23]. The discovery of lesions in a pre-menopausal woman does not guarantee their persistence up to the age of onset of the menopause. Likewise the exclusion of pelvic disease by laparoscopic visualisation at any given time does not preclude the later development of lesions prior to the menopause. Based on this evidence it is plausible that all observed cases of post-menopausal endometriosis could have arisen in patients with a pre-menopausal history of the disease.

Regardless of the origins of the lesions themselves, endometriosis is widely considered to be an oestrogen dependent disorder [6]. In women of reproductive age, cyclic ovarian oestradiol production from circulating androstenedione and testosterone, catalysed by granulosa cell aromatase activity, provides the major source of serum oestrogen [6]. In the post-menopausal woman the skin and adipose tissue become the main sites of oestrogen production, once again catalysed by the aromatase enzyme, utilising adrenal androstenedione as the substrate [6]. This endogenous post-menopausal production is enhanced in obese individuals, and may result in elevated serum oestradiol levels [6]. Additional sources of endogenous oestrogen during the post-menopausal years may include oestrogen-producing ovarian tumours [7]. Post-menopausal women might also be exposed to exogenous oestrogen sources, classically from hormone replacement therapy (HRT) and phyto-oestrogens [7].

A clear source of excessive oestrogen has been identified in several cases of post-menopausal endometriosis. A prospective randomised study of 172 women implies a link between HRT and post-menopausal disease recurrence. All women included in the study had histological confirmation of endometriosis, and had all undergone surgery for BSO, leading to a surgically induced menopause [36]. Of the 172 women included, 115 were randomly allocated to receive HRT in the form of controlled-release transdermal oestradiol patches at a dose of 50μg/day, and an orally administered progesterone. All women underwent 6 month follow-up by clinical examination, hormonal evaluation and transvaginal ultrasound. Although statistical significance was not reached, of the women allocated to receive HRT the incidence of endometriosis recurrence was 0.91 per 100 woman years compared to a complete lack of disease recurrence in the non-HRT controls [36]. Risk factors for post-menopausal disease appeared to include previous severe peritoneal disease and failure to perform total hysterectomy at the time of BSO, the latter reaching statistical significance [36]. Taken together the evidence suggests a link between residual pre-menopausal lesions and post-menopausal recurrence under HRT stimulation. However it is important to mention that although women included in the study were subject to 6-month hormonal evaluation, mean serum oestradiol levels were not published in the report [36]. Further to this, repeat laparoscopy, not ultrasound, would be considered the gold standard for detecting disease recurrence [34] as lesions on the peritoneum might have been missed by ultrasound alone. However, there are clearly ethical considerations precluding the exposure of these women to an invasive procedure such as a diagnostic laparoscopy.

It is possible that excessive endogenous oestrogen might also play a causative role in post-menopausal endometriosis. A 69 year-old female presenting with pelvic pain, constipation and weight loss of 30 kg was discovered to have a sigmoid colon endometrioma. The disease appeared to involve the uterus and bladder and was managed surgically by colonic resection and a hysterectomy [37]. At the time of diagnosis the patient's body mass index (BMI) was within normal limits, while prior to the recent weight loss her BMI had been 31.6. Analysis of the endometrium following uterine excision revealed evidence of pseudocystic hyperplasia consistent with residual oestrogen stimulation [37]. However due to the nature of her advanced disease and significant associated weight loss, evidence of prior exposure to excessive endogenous oestrogen could not be confirmed by serum oestradiol levels at the time of presentation.

A case of post-menopausal cutaneous endometriosis in a 58 year-old female receiving HRT over a five year period challenges a direct causative link between excessive oestrogen exposure and post-menopausal disease. The woman had been postmenopausal for 8 years, and had been exposed to 625 μg of oestrogen daily over the latter 5 years, which she had been taking for the relief of post-menopausal symptoms [38]. Three years prior to presentation she had undergone a total abdominal hysterectomy and BSO for an ovarian cystic teratoma and chronic cervicitis. She later developed several erythematous plaques and papules on her back, confirmed to be endometriosis on histological analysis [38]. Although the cutaneous lesions resolved spontaneously within one month of cessation of HRT, there was no evidence of disease recurrence within 6 months of re-starting hormone therapy [38].

Endometriotic lesions classically arise on the ovaries and pelvic peritoneum, but reports of endometriosis in extra-pelvic locations including the bowel, renal tract, the abdominal wall, lung and pleura are relatively common [39]. In fact extra-pelvic disease is thought to occur in up to 12% of cases of endometriosis [38]. Given the lack of continuity of these sites with the pelvic cavity, the discovery of extra-pelvic disease is considered a contradiction of Samson's theory of retrograde menstruation and implantation [9]. If cases of extra-pelvic endometriosis also arise during the post-menopausal years, it is conceivable that post-menopausal, and likewise pre-menopausal disease, does not rely on a functioning endometrium at any stage. If so, it is possible that another theory for the pathophysiology of endometriosis is more compatible. For example the coelomic metaplasia hypothesis suggests that mesothelial cells of the ovary and peritoneal cavity may undergo metaplasia to form endometrial-like tissue [40] perhaps under the influence of hormonal stimulation [41]. However this theory could only account for disease arising on the ovaries and peritoneal serosa [42] and could not explain other forms of extra-pelvic disease.

Alternatively it has been suggested that endometrial tissue from the uterine cavity might be transported to distant sites via the veins or lymphatics [8], [9], perhaps accounting for other extra-pelvic manifestations of endometriosis, such as cutaneous deposits. In fact, endometriotic lesions have been detected in the pelvic lymph nodes of women with severe disease [43], [44] suggesting the possibility that dissemination of eutopic endometrium via the lymphatics might contribute to disease spread. This theory, however, still appears to rely on the presence of a functioning endometrium. Intriguingly, in the case of post-menopausal cutaneous endometriosis in a 58 year-old discussed previously [38], the patient had undergone a total hysterectomy and BSO three years prior to the appearance of the lesions. If endometriosis does arise secondary to venous or lymphatic transport of endometrial fragments, it is possible that the development of post-menopausal cutaneous disease could have been the result of dissemination of the endometrium at the time of the previous surgery [38]. Regardless of the origins of these lesions, the persistence of endometriotic deposits is generally considered to be oestrogen dependent [6]. None of the theories discussed so far can therefore account for the presence of active disease in the absence of serum oestradiol levels in excess of the normal post-menopausal range [29], [38].

There is however new evidence to suggest that locally produced oestrogen plays a significant role in the pathophysiology of endometriosis, perhaps negating the requirement for elevated circulating oestrogen levels during the post-menopausal years [6]. High levels of aromatase and StAR expression and enzyme activity have been demonstrated in cultured stromal cells derived from endometriotic lesions, and it has been suggested that PGE2 production might further enhance localised oestrogen syntheses [8], [19], [20]. This is likely to form a positive feedback loop, whereby oestrogen produced within lesions further promotes proliferation and disease progression through autocrine and paracrine effects, resulting in more inflammation and localised PGE2 production, thus further enhancing oestrogen synthesis [6], [19]. If this picture is accurate, it is possible that endometriosis is a self-perpetuating condition, the lesions themselves driving local oestrogen production and disease progression, even in the absence of elevated serum oestradiol as the stimulus. It is therefore conceivable that post-menopausal disease might arise in cases where pre-menopausal lesions persist and remain active into the post-menopausal years via local mechanisms and paracrine effects. This theory might explain cases of post-menopausal endometriosis where a source of excessive circulating oestrogen cannot be identified [32], [45].

In keeping with this idea, it appears that aromatase inhibitors provide an effective treatment for post-menopausal endometriosis [6], [22]. In the case of a 61 year-old, obese postmenopausal female presenting with recurrent abdominal wall endometriosis, serum oestradiol concentrations were found to be elevated in excess of the normal postmenopausal range at 39 pg/ml (pre-menopausal > 20 pg/ml), while FSH levels were consistent with the post-menopausal status at 44 mIU/ml (pre-menopausal < 40 mIU/ml) [22], [29]. Following aspiration of the abdominal wall endometrioma, oestradiol concentrations within the cystic fluid of the lesion itself were recorded. Local oestradiol levels within the aspirate were found to be elevated in excess of circulating oestradiol, at 89 pg/ml [22]. Initial treatment with the aromatase inhibitor letrozole administered orally at 5 mg/day resulted in a decrease in cystic fluid oestradiol levels to 28 pg/ml [22]. Following 4 weeks of treatment with the aromatase inhibitor (with the addition of medroxyprogesterone acetate at 10 mg/day PO during the final week) the abdominal wall lesion had reduced significantly in size and could be fully reduced following aspiration, while serum and cystic fluid oestradiol levels had decreased to <20 pg/ml and 22 pg/ml respectively [22]. This case demonstrates the potential relevance of local oestradiol production in post-menopausal endometriosis and, related to this, the efficacy of aromatase inhibitors in treating post-menopausal disease.

Section snippets

Conclusion

Endometriosis is a common gynaecological disorder, which has traditionally been considered a disease of the pre-menopausal years [7]. For pelvic disease alone three clinically distinct forms have been described, including superficial implants on the pelvic peritoneum and ovaries, ovarian endometriotic cysts, and rectovaginal nodules [8]. Further to this, rare extra-pelvic disease has been reported in the literature [39]. Of the theories proposed so far, there is not a single one that is capable

Funding source

This work was supported by an MRC New Investigator Award (G0601458; C.M.B.) and the Oxford Partnership Comprehensive Biomedical Research Centre with funding from the Department of Health's NIHR Biomedical Research Centres scheme (C.M.B.).

The funding source had no role in writing the paper.

Contributors

C.L.B. and C.M.B. wrote the paper and approved the final version.

Competing interests

No conflicts of interest to declare.

Provenance and peer review

Commissioned and externally peer reviewed.

References (53)

  • M.D. Hornstein et al.

    The reproducibility of the revised American Fertility Society classification of endometriosis

    Fertility and Sterility

    (1993)
  • O. Buchweitz et al.

    Interobserver variability in the diagnosis of minimal and mild endometriosis

    European Journal of Obstetrics, Gynecology and Reproductive Biology

    (2005)
  • A.P. Bailey et al.

    Florid endometriosis in a postmenopausal woman

    Fertility and Sterility

    (2010)
  • K. Takayama et al.

    Treatment of severe postmenopausal endometriosis with an aromatase inhibitor

    Fertility and Sterility

    (1998)
  • R. Matorras et al.

    Recurrence of endometriosis in women with bilateral adnexectomy (with or without total hysterectomy) who received hormone replacement therapy

    Fertility and Sterility

    (2002)
  • B. Deval et al.

    Sigmoid endometriosis in a postmenopausal woman

    American Journal of Obstetrics and Gynecology

    (2002)
  • S.W. Choi et al.

    A case of cutaneous endometriosis developed in postmenopausal woman receiving hormonal replacement

    Journal of the American Academy of Dermatology

    (1999)
  • H. Suginami

    A reappraisal of the coelomic metaplasia theory by reviewing endometriosis occurring in unusual sites and instances

    American Journal of Obstetrics and Gynecology

    (1991)
  • D. Vinatier et al.

    Theories of endometriosis

    European Journal of Obstetrics, Gynecology and Reproductive Biology

    (2001)
  • S. Mechsner et al.

    Immunohistochemical evaluation of endometriotic lesions and disseminated endometriosis-like cells in incidental lymph nodes of patients with endometriosis

    Fertility and Sterility

    (2010)
  • M.H. Moen et al.

    EMAS position statement: managing the menopause in women with a past history of endometriosis

    Maturitas

    (2010)
  • G.M. Zanetta et al.

    Hyperestrogenism: a relevant risk factor for the development of cancer from endometriosis

    Gynecologic Oncology

    (2000)
  • P.R. Koninckx et al.

    Deeply infiltrating endometriosis is a disease whereas mild endometriosis could be considered a non-disease

    Annals of the New York Academy of Sciences

    (1994)
  • D. Wagstaff et al.

    Do we know what causes endometriosis?

    British Journal of Hospital Medicine (London)

    (2008)
  • S.E. Bulun et al.

    Estrogen production and metabolism in endometriosis

    Annals of the New York Academy of Sciences

    (2002)
  • J. Cumiskey et al.

    A detailed morphologic and immunohistochemical comparison of pre- and postmenopausal endometriosis

    Journal of Clinical Pathology

    (2008)
  • Cited by (31)

    • Total laparoscopic hysterectomy in patients with deep endometriosis: Different technical and postoperative considerations, in comparison with a procedure performed for other benign indications

      2022, Taiwanese Journal of Obstetrics and Gynecology
      Citation Excerpt :

      It is estimated that postmenopausal endometriosis affects 2–4% of the female population. Recurrence of symptoms, may occur even in the absence of identifiable anatomic relapse, but it is difficult to define the exact incidence as many patients complaining of symptoms suggestive of a relapse, do not undergo laparoscopic evaluation [20]. Incomplete surgery, ovarian conservation, and postoperative administration of HRT are among the most important factors for postmenopausal recurrence of pain [7].

    • The prevalence of endometrioma and associated malignant transformation in women over 40 years of age

      2020, Journal of Gynecology Obstetrics and Human Reproduction
      Citation Excerpt :

      The most widely accepted theory is the Sampson’s theory. The occurrence of endometriosis in advanced age could be explained with extra-ovarian production of estrogen by adipose tissue and/or endometriotic tissue itself or the exposure to iatrogenic hormones [15,16]. Endometrioma over 40 years of age develops in a state of relatively low estrogen levels.

    • Endometriosis: A life cycle approach?

      2013, American Journal of Obstetrics and Gynecology
      Citation Excerpt :

      If, on the other hand, endometriotic lesions can arise de novo in postmenopausal women, then an alternative processes must be in place. It is unlikely that one single theory can account for all postmenopausal lesions that have been described.116 It has been suggested that in untreated postmenopausal women endometriosis may arise from coelomic metaplasia.95

    View all citing articles on Scopus
    View full text