Asherman’s Syndrome: it may not be all our fault

Abstract

Asherman’s Syndrome (AS) is an acquired condition defined by the presence of intrauterine adhesions (IUA) that cause symptoms such as menstrual abnormalities, pelvic pain, infertility, recurrent miscarriage, abnormal placentation and attendant psychological distress. Classically, AS is considered an iatrogenic disease triggered by trauma to the pregnant uterus. Different factors can cause the destruction of the endometrium, thus affecting the endometrial stem cell niche and creating IUAs. Curettage of the pregnant uterus appears to be the most common source of this destruction. Nevertheless, some AS cases have been associated with congenital uterine abnormalities and infections, and there are some idiopathic cases without any prior surgical procedures, suggesting a putative constitutional predisposition to IUA. Factors reported to cause AS share an underlying inflammatory mechanism leading to defective endometrial healing and vascularization. Interestingly, distinct genetic profiles have been observed in the endometrium of AS patients. These data suggest that AS might not just be an iatrogenic complication, but also the result of a genetic predisposition. Elucidating the possible physiopathological processes that contribute to AS will help to identify patients at risk for this condition, providing an opportunity for prevention.

Asherman’s Syndrome, intrauterine adhesions, iatrogenesis, endometritis, curettage, congenital uterine abnormalities

Introduction

A syndrome is a group of signs and symptoms that coincide with and characterize a particular abnormality or condition. According to this definition, Asherman’s Syndrome (AS) is considered to be an acquired condition, defined by the presence of intrauterine adhesions (IUAs). AS occurs when bona fide endometrium is replaced by fibrotic tissue, causing the uterine walls to adhere to one another and resulting in symptoms such as menstrual abnormalities, pelvic pain, infertility, recurrent miscarriage, abnormal placentation and attendant psychological distress (Dmowski and Greenblatt, 1969). The term ‘AS’ should be applied when such signs and symptoms are present in women with established intrauterine adhesions (Zupi et al., 2015), although currently it may also be used when only referring to intrauterine adhesions, scarring or synechiae or Fritsch Syndrome (Fritsch, 1894) indistinctly.

The first case of intrauterine adhesions was reported by Heinrich Fritsch in 1894, but it was the Israeli gynecologist Joseph Asherman who, in 1948, first described patients with partial or complete obliteration of the uterine cavity and/or cervical canal, resulting in menstrual abnormalities, infertility and recurrent pregnancy loss. At the time, these cases were described as ‘traumatic amenorrhea’ (Asherman, 1948; Roy et al., 2010). Hundreds of articles have since been published, but the prevalence and incidence of this disease is difficult to establish due to limited high-quality epidemiological data and methodological issues such as populations with differences in degrees of physician awareness, numbers of therapeutic and illegal abortions, and rates of infections (tuberculosis or endometritis) (Al-Inany, 2001).

Recently, the European Medicine Agency (EMA) has established the prevalence of AS in the EU as 4 in every 10 000 individuals (COMP, 2017). Notably, this estimated value is below the threshold established by the European Agency for a condition to be considered as a rare disease, i.e. the prevalence is lower than 5 in 10 000. Based on these criteria, AS is considered to be a ‘rare disease’ and is included in the Orphanet database under the registry ORPHA137686. Furthermore, AS is classically considered to be an iatrogenic disease, although recent epidemiological, clinical and basic findings challenge the assertion that AS is exclusively of iatrogenic origin.

Aberrant endometrial repair and adhesion formation: cause or consequence?

The endometrium has a remarkable capacity to regenerate the functional layer from its basalis during the reproductive period, and throughout the entire life-span if hormonal supplementation is administered (Mutlu et al., 2015), due to the existence of endometrial stem cells. The specific physiological microenvironment in which stem cell proliferation and cell fate decisions are regulated is known as the niche (Schofield, 1978). In humans, the endometrial stem cell niche seems to be located at the endothelium of the spiral arterioles in the basal layer, providing support to both the epithelial and stromal compartments (Cervello et al., 2007; Schwab and Gargett, 2007; Murakami et al., 2014).

Recent studies have reported that bone marrow-derived stem cells (BMDSCs) contribute to the repair and regeneration of tissues and organs (Pittenger et al., 2008), including the murine endometrium (Bratincsák et al., 2007; Faivre et al., 2009; Du et al., 2012; Morelli et al., 2013) and human endometrium (Taylor, 2004; Du and Taylor, 2007; Mints et al., 2008; Ikoma et al., 2009; Cervelló et al., 2012). According to these reports, BMDSCs have the capacity to differentiate into fully functional stromal and epithelial endometrial cells. However, the specific BMDSC subpopulation that promotes the repair of the endometrium remains unknown.

Different factors can cause destruction of the endometrium through to its basal compartment, affecting the local cellular niche (Cervelló et al., 2010, 2011; Cervello et al., 2015). The damaged endometrium then cannot be properly repaired, leading to the formation of fibrotic synechiae across the cavity in which the glands are replaced by inactive cubocolumnar endometrial epithelium unresponsive to hormonal stimulation (Conforti et al., 2013). The newly formed fibrotic tissue is usually avascular, although thin-walled telangiectatic vessels can be observed and the glands may be sparse and inactive or cystically dilated, causing reduced blood perfusion and atrophy (Yaffe et al., 1978).

Histologically, these IUAs can be formed by endometrial tissue, connective tissue or muscular tissue. The connective tissue is composed of thin collagen bundles or can form dense fibrous strips. When muscular tissue is present, it has similar histologic characteristics as the normal myometrium although, the myometrium in AS patients usually tends to show an increase in thickness due to myohyperplasia (Foix et al., 1966).

The endometrium represents a highly dynamic tissue that undergoes regular cycles of repair and breakdown where angiogenesis (the formation of new blood vessels from pre-existing vasculature) plays a pivotal role (Demir et al., 2010). There is a paucity of blood circulation in women with adhesions, indicating occlusion damage of the uterine arteries concomitant with fibrosis (Polishuk et al., 1975; Schenker and Margalioth, 1982; Conforti et al., 2013). Arterial fibrosis also contributes to failed endometrial growth as it is unresponsive to hormonal treatment and yields to tissue hypoxia, triggering a cascade of events leading to the creation of adhesions, as demonstrated by in-vitro studies in which human fibroblasts exposed to hypoxia manifested phenotypical and irreversible changes characteristic of fibrosis (Fletcher et al., 2008). Furthermore, hypoxia seems to negatively modulate all pathways involved in tissue formation, with effects such as an increase in the expression of collagen 1 mRNA in fibroblasts or proliferation induction along with apoptosis inhibition in fibroblasts, favouring firm adhesion formation (Saed and Diamond, 2002).

Impaired homeostasis of the extracellular matrix (ECM) is also involved in the creation of adhesions (PCASRM and Practice Committee of the American Society for Reproductive Surgeons, 2007). The ADAM family of proteins degrade ECM anchoring proteins which are strictly regulated under physiological conditions. ADAM-15 and ADAM-17 are significantly upregulated in endometrial tissue from patients with moderate and severe AS, favouring adhesiogenesis (Liu et al., 2013). Further, dysregulation of TGF-β1, SMAD3 and SMAD7 in patients with IUAs as well as in the rabbit model induced by endometrial curettage has been reported (Salma et al., 2016). Therefore, an imbalance between TGFβ1 and SMAD3 (abnormally upregulated) and SMAD7 (downregulated) might be relevant in the development of IUA and AS. More recently, impaired NF-κB signaling has also been implicated in AS. Such signaling is a marker of aberrant inflammation that can create a vicious fibrotic cycle, as demonstrated in other tissue types where fibrosis associates with aberrant inflammation (Wang et al., 2017). Additionally, a recent study (Zhao et al., 2017) has reported an upregulation of ΔNp63 in women suffering from AS, causing endometrial stem cell quiescence and leading to adhesion formation. Interestingly, autologous bone marrow mononuclear cells have the capacity to reverse these inhibitory effects on endometrial epithelial cells. In general, these data support a model in which aberrant endometrial repair and/or fibrotic processes result in massive adhesion formation and AS, thus suggesting putative molecular targets for potential therapies. What remains unclear is whether such processes are the cause or the effect of the syndrome.

Adhesiogenesis seems to be the culmination of an abnormal response to inflammation, resulting in increased extracellular matrix production associated with diminished matrix degradation combined with decreased fibrinolytic activity due to hypoxia (Saed and Diamond, 2002) of adhesions, in response to trauma, ischemia or infection. The menstrual period represents a model of self-limiting inflammation, ECM degradation and tissue repair as well as fibrinolytic activity, hypoxia (Maybin and Critchley, 2015) and angiogenesis (Girling and Rogers, 2005). In the menstrual cycle, endometrial breakdown and repair occur simultaneously, side by side, under a carefully regulated balance (Evans and Salamonsen, 2012) that have been termed ‘orderly inflammation’ (Nathan and Ding, 2010). The balance among these processes seems to be the key to controlling the inflammatory processes at menstruation and the complete understanding of how this balance may be influenced may reveal important information that could enable full control over this process (Evans and Salamonsen, 2012).

Interestingly, most cases of AS display these features. However, not all women undergoing a miscarriage or dilation and curettage (D&C) develop AS, and there are other cases where IUAs develop even with no intervention (Panayotidis and Ranjit, 2004; Fernandez et al., 2006, 2012). Hence, with the present data, we hypothesize that some constitutional or transitory factors may contribute to an imbalance in the processes affecting the ECM regeneration and vascularization, leading to defective endometrial repair and vascularization, which might result in a major predisposition to developing AS for some patients.

Is Asherman’s Syndrome entirely our fault?

Miscarriage is the commonest complication of pregnancy affecting 15% of clinically recognized pregnancies; 5% of women will experience two or more miscarriages during their lifetime (Rai and Regan, 2006). Conservative or pharmacological management of miscarriage with misoprostol represents a less-invasive therapeutic option since ~50% of cases evacuate within 2 weeks of diagnosis (Wieringa-De Waard et al., 2002). When successful, it is a cost-effective strategy with an insignificant rate of IUA (Hooker et al., 2013). However, most miscarriages are still treated by D&C (Hooker et al., 2016).

Trauma from procedures such as curettage, cesarean section or myomectomy to a gravid or non-gravid uterus is typically the cause of IUAs (Schenker and Margalioth, 1982; Yu et al., 2008; Panayotidis et al., 2009; Conforti et al., 2013). These procedures cause damage to the basalis layer of the endometrium and promote IUA. Pregnancy seems to play an adjuvant role since 15–20% of patients receiving curettage due to an induced or spontaneous abortion and 21–40% with post-partum curettage develop IUA (Hooker et al., 2013; Gilman et al., 2016); this rate decreases to 1.6% after diagnostic curettage in gynecological conditions and 1.3% after abdominal myomectomy (Deans and Abbott, 2010). Women with repeated curettages as well as women with large uteri or multiple pregnancies show an increased risks of IUA (Gilman et al., 2016). Further, women with two or more miscarriages show an increased risk of IUA (Hooker et al., 2013).

The surgical technique used in the procedure also seems to be relevant since sharp curettage or mechanical suction imposes increased risk for the development of IUA compared to manual vacuum aspiration (Gilman et al., 2016). As such, medical awareness seems to be important in minimizing the risk of IUA. As a matter of fact, the number of D&C in Germany decreased from 77, 115 in 2005 to 51, 342 in 2015, representing a reduction of 28%, although the number of diagnosed AS cases only dropped from 181 to 161, which is only an 8% decrease (DESTATIS, 2017).

Certain constitutional factors such as Müllerian duct malformations are associated with a higher incidence of adhesion formation. In 1985, Stillman and Asarkof (1985) found a significant correlation in infertile patients between AS and Müllerian duct malformations, especially in patients with septate uteri. Among the 43 infertile patients with Müllerian duct malformations, 7 (16%) had AS. More recent studies with improved diagnostic techniques have observed a low rate of IUA (2–3%) in infertile patients with no Müllerian duct defects (Hinckley and Milki, 2004). Therefore, it seems that congenital uterine abnormalities such as septate or bicornuate uterus predispose to adhesion formation (Hooker et al., 2013). The prevalence of uterine malformations in the general population is estimated to be 3.8% to 9.6% (Zupi et al., 1996; Fedele et al., 2006) and, according to certain studies, some degree of IUA can be visualized during hysteroscopic investigation of Müllerian abnormalities (Panayotidis et al., 2009). Specific causes of these malformations remain unknown although in some cases they might be associated with vaginal or renal tract malformations (Oram et al., 2010). While these findings seem sporadic in most of the cases, familial recurrences have also been reported (Jacquinet et al., 2016). In this sense, rare variations of the HOXA 10 gene have been related to Mullerian abnormalities (Ekici et al., 2013) whereas mutations in patients with Mullerian malformations are associated with renal malformations (Oram et al., 2010). Moreover, a high rate of spontaneous abortions are reported in pregnancies in septate uteri since the septum seems to be poorer vascularized (Cohen et al., 2001). A study reported the location of implantations sites in 12 pregnancies in septate uteri by ultrasound. Interestingly, all of the pregnancies that did not miscarriage were implanted in the lateral uterine walls whereas the eight pregnancies that miscarriaged showed mostly septal implantation (Fedele et al., 1989).

Furthermore, the persistence of trophoblastic tissue after blind D&C seems to be more common in cases of Müllerian duct defects such as didelphic uterus or septate uterus (Cohen et al., 2001; Faivre et al., 2009). Recently, it has been shown that immunological factors are also involved in the ethiopathogenesis of missed abortions (Li et al., 2017) and miscarriage of karyotypically normal embryos may be caused by abnormal inflammation at the feto-maternal interface (Christiansen et al., 2006). Therefore, this residual trophoblastic tissue can expose the uterus to inflammation which may lead to scarring and finally IUAs. At the same time, these IUAs may cause stagnation of menstrual debris that may subsequently provoke chronic inflammation inside the uterine cavity. Therefore, the risk of developing IUAs seems to be specially high in the patients with residual trophoblastic tissue and the management selective hysteroscopic surgery without electrodissection (Faivre et al., 2009) or even hysteroscopic morcellation (Hamerlynck et al., 2013) should be considered.

In this line, there is a significantly higher rate of IUA after D&C if the miscarriage was a missed abortion (30.9%) versus an early miscarriage (6.4%) (Adoni et al., 1982) or spontaneous abortion (60%) versus induced abortion (17.8%). (Lancet and Kessler, 1988). In the study of Adoni et al., a missed abortion was defined histologically with images displaying fibrotic villi without blood vessels; in contrast, an early miscarriage is typically indicated by villi that contain blood vessels and are not fibrotic. Moreover, uterine anomalies and especially partial uterine septum have been reported in up to 25% of patients diagnosed of missed abortion. However, early evacuation of the uterus in missed abortions within the first 4 days of diagnosis does not seem to increase the incidence of IUAs (Friedler et al., 1993). Further, in a series of 15 patients diagnosed with AS after post-partum curettage, the curettage was performed during the second, third or fourth week post-partum, whereas immediate post-partum curettage does not seem to cause traumatic IUA. It is supposed that the endometrium seems very susceptible to denudation during this critical period as observed by the amount of basalis and myometrium observed in the specimens removed in the curettage (Jensen and Stromme, 1972).

Infections represent another established cause of IUA. Mycobacterium tuberculosis infection of the genital tract results in severe intrauterine adhesions (Bukulmez et al., 1999), even without any previous surgery (Gupta et al., 2007), with higher recurrence of IUAs and poor prognosis after hysteroscopic surgery (Sharma et al., 2008). There are several early reports from the 1950s and 1960s (not available) cited in other papers highlighting the link between tuberculosis and adhesions, naming the condition Netter syndrome or Musset-Netter syndrome (Schenker and Margalioth, 1982; Conforti et al., 2013). One case report suggests that Schistosomiasis may also cause AS through a granulomatous reaction to the schistosome eggs that causes fibrosis and pain and ultimately adhesions and amenorrhea (Krolikowski et al., 1995). Consistent with this, a recent study concluded that 40% of patients affected by IUAs present symptoms or findings of chronic endometritis (CE) and a higher recurrence of adhesion, indicating that chronic inflammation may play a role in the development and recurrence of IUAs (Chen et al., 2017). Earlier studies exploring the relationship between endometritis and the formation of adhesions were less conclusive due to the heterogeneity and small size of the populations investigated (Polishuk et al., 1975; Schenker and Margalioth, 1982; Conforti et al., 2013).

Less common are idiopathic cases of AS without any previous related congenital abnormality, injury or infection. The presence of AS in a 52-year-old virgin menopausal woman without any significant predisposing factors illustrates the possibility of having IUAs without a known cause (Panayotidis and Ranjit, 2004). Similarly, other authors have reported several cases of patients affected by AS with an idiopathic origin. Polishuk et al. (1975) reported three cases of pregnant women who developed AS without any surgical intervention. Subsequently, Fernandez et al. reported two cases: one with an adenomyotic uterus diagnosed by magnetic resonance imaging (MRI) (Fernandez et al., 2012) and the other with a spontaneous miscarriage without any curettage or surgery (Fernandez et al., 2006). A Canadian study (Dawood et al., 2010) reported four cases with stage I and stage II adhesions in the absence of a known causative factor. Finally, in a study performed by our group (Santamaria et al., 2016), one patient (Patient 9) was diagnosed with AS without any previous surgical procedure. Therefore, although AS has been considered iatrogenic since its initial description, all these facts suggest that not only iatrogenic causes and infections, but also anatomical or constitutional predisposition, might contribute to this pathology. Although residual, the association of genetic mutations with certain Mullerian malformations, does not fully exclude a potential genetic predisposition to developing IUAs in these patients.

All the reported etiological factors trigger an aberrant inflammatory response (Fig. 1). Inflammation seems to play a pivotal role in the pathogenesis of AS, causing damage to the endometrial niche and the release of factors into the intrauterine environment that stimulate the formation of fibrotic tissue and decrease vascularization following endometrial trauma (Sharma et al., 2008). Therefore, we propose the following hypothesis for the development of AS: infection (endometritis, genital tuberculosis or schistosomiasis) may cause defective endometrial healing as well as inflammation, while miscarriages and Müllerian duct defects, with or without iatrogenic factors such as curettages, may trigger an inflammatory response that can provoke an imbalance in endometrial tissue breakdown, affecting the ECM regeneration and vascularization and leading to defective endometrial healing and vascularization. Recent studies have determined a characteristic endometrial signature when IUAs are present (Liu et al. 2013; Salma et al. 2016; Wang et al. 2017). Whether this signature is the cause or consequence needs to be further investigated.

Figure 1

Proposed hypothesis for the development of Asherman’s Syndrome (AS). This hypothesis highlights the importance of the inflammatory pathway in the pathogenesis of AS, resulting in damage to the endometrial niche and the release of factors into the intrauterine environment, which stimulate the formation of fibrotic tissue and decrease vascularization after endometrial trauma.

A successful approach to AS is presented in the acronym PRACTICE (PRevention, Anticipation, Comprehensive therapy, Timely surveillance of subsequent pregnancies, Investigation and Continuing Education) (March, 2011). Despite the significant increase in conservative management of miscarriages in recent years, the number of diagnosed AS cases has not decreased in comparative terms, suggesting the existence of other factors involved in the development of this pathology.

Prevention of IUA is paramount and starts by optimizing the diagnostic strategy (Hooker et al., 2016). In this line, special attention has to be given to patients affected by Müllerian abnormalities, missed abortions, persistence of trophoblastic tissue, post-partum complications and/or uterine infections such as gynecological tuberculosis, schistosomiasis or chronic endometritis in order to anticipate a comprehensive approach. Therefore, it seems reasonable to initially consider conservative or pharmacological management in these patients. However, when surgical management is performed, early vacuum evacuation of the uterus in missed abortions within the first 4 days of diagnosis is advised as well as selective hysteroscopic surgery without electrodissection or morcellation in persistence of trophoblastic tissue. Curettage performed during the second, third, or fourth week post-partum seems to carry a higher risk of developing IUA. Finally, several post-operative preventive measures have also been effective. Application of auto-crosslinked hyaluronic acid (SCP) after D&C reduces the incidence and severity of IUA in high risk patients (Hooker et al., 2017) while other authors also suggest second-look hysteroscopy after hysteroscopic removal of trophoblastic tissue (Faivre et al., 2009). Since hyaluronic acid has the capacity to interact with stem/progenitor cells supporting their survival and differentiation (Chai and Leong, 2007), it might not only exert its effect as a physical barrier, but also support the niche to regenerate the damaged endometrium. Finally, there is little evidence regarding the use of antibiotic therapy to prevent IUA. However, antibiotic therapy has to be considered when a diagnosis of chronic endometritis has been established (Chen et al., 2017).

Conclusions

AS is a pathological condition that is caused by a variety of factors that mostly trigger inflammation and can be defined as a rare disease as recently recognized by the EMA (COMP, 2017). It is recognized and classified by different international medical societies and organizations as a rare disease with different treatment guidelines (AAGL, 2010).

AS may develop even in the absence of any prior surgical intervention, suggesting that it may not be just a simple iatrogenic complication as initially described. Accumulated evidence suggests a predisposition in some women to developing this pathological condition. Therefore, proactive attitudes in order to identify patients at a higher risk of developing AS are strongly recommended.

Further studies to better understand the physiopathology of this syndrome are needed, not only to develop effective cures to regenerate and re-establish the normal function of the endometrium, but also to anticipate and personalize the most comprehensive medical treatment for patients at high risk of developing AS.

Authors’ roles

Xavier Santamaria (XS) contributed in the conception and design of the paper, aquisition of data and interpretation of data, and drafting of the article and approved the final version, Keith Isaacson (KI) contributed in revising the manuscript for relevant intellectual content and gave final approval. Carlos Simon (CS) critically revised the article, helped in the design of the manuscript and provided final approval.

Funding

No funding was used for this paper.

Conflict of interest

The authors declare no conflict of interest.