Keywords

1 Introduction

Intrauterine adhesions (IUA) are rare and they are almost always preceded by pregnancy-related intrauterine procedures like a miscarriage curettage or a puerperal curettage for placental remnants. IUA never occur spontaneously. In case of hypomenorrhoea or amenorrhoea, hysteroscopic treatment is necessary if the patient is waiting for a pregnancy. There is only very limited evidence related to the treatment of Asherman’s disease, so the different treatment modalities are based on individual experience. Since the condition is so rare, looking for evidence for the primary prevention of AS seems rather irrational as the number needed to treat will probably always be very high. However, the search for adequate measures to prevent recurrent adhesions after treatment seems logical and trials, preferably randomized, are urgently needed.

2 History

The first description of the treatment of intrauterine adhesions by Heinrich Fritsch was published in 1894 [1]. In the original articles of Joseph Asherman, he described only sounding and gradual dilatation of the cervix until difficulty was met [2,3,4]. If the first attempt was unsuccessful, a period of observation was advised since many patients still had a restoration of menstruation. When the patient or the physician was not satisfied, a second attempt was tried following a period of observation. If no further recovery occurred, a laparotomy and anterior hysterotomy was advised. The adhesions were gently removed by the index finger and the uterine cavity was palpated. A ‘medium-soft catheter’ was inserted and passed out through the cervix into the vagina. The uterine wall was sutured in two layers and covered by bladder peritoneum or the ligamenta rotunda. The catheter was removed after 3 days. At that time it was already recognized that curettage was ineffective and it was stated that ‘it needed to be investigated whether the operation could be carried out with the eye as a guide by means of hysteroscopy’.

In the seventies of the last century, the first use of hysteroscopy for the treatment of IUA appears in the international literature from different parts in the world [5,6,7,8,9,10].

An example of one of those early documents, 30 years later, describing the systematic use of the hysteroscope for the treatment of AS with follow-up on the outcome is a publication of Sugimoto from Kyoto, Japan in 1978 [9]. During diagnostic hysteroscopy in 192 patients, intrauterine adhesions were found. Central adhesions appearing as a bridge-like tissue connecting the anterior and posterior walls of the uterus and marginal adhesions appearing as wedge-like tissue projected from a part of the uterine side wall were described separately. All of the central adhesions and most of the marginal adhesions were removed by pushing with the tip of the outer sleeve of the hysteroscope under visual control. Furthermore, the hysteroscope was used to get an impression of the severity and extend of the pathology by judging the force needed for separation, and the appearance of the stumps of adhesions removed. The effects of adhesiolysis were evaluated by describing menstrual blood flow and pregnancies. Amongst the 192 patients, 143 recovered previous menstrual flow, the others remained hypomenorrhoeic. In 79 pregnant patients, 45 were delivered of viable infants. The publication ends with this statement: ‘Improvement of endometrial malfunction in advanced cases of intrauterine adhesions is a problem awaiting a solution’. The value of this statement, nowadays almost 40 years later, did not loose much of its strength.

3 Hysteroscopical Equipment and Techniques

With the introduction of minimal access surgical techniques like hysteroscopy, the ability was created to remove intrauterine disorders selectively with great patient satisfaction and important advantages to both the patient and the surgeon. In such cases a hysterectomy can be prevented and especially for women who wish to retain their fertility this is of extreme importance.

Hysteroscopy is the process of viewing and operating in the endometrial cavity from a transcervical approach. The basic instrument is a long, narrow telescope connected to a light source to illuminate the area to be visualized. With a patient in the lithotomy position, the cervix is visualized and the distal end of the telescope is passed into a cervical canal, and, under direct visualization, the instrument is advanced into the uterine cavity. A camera is attached to the proximal end of the hysteroscope to broadcast the image onto a video monitor. Given their safety and efficacy, diagnostic and operative hysteroscopy have become standards in gynaecologic practice. Over the past few decades, refinements in optic and fibre-optic technology and inventions of new surgical accessories have dramatically improved visual resolution and surgical techniques in hysteroscopy [11]. Many hysteroscopic procedures have replaced old and/or invasive techniques like dilatation and curettage and hysterectomy.

The rigid, direct optical hysteroscope , which is derived from the cystoscope, uses fluid or gas distending media to obtain a wide-angle view of the uterine cavity. This hysteroscope takes advantage of a system of lenses and prisms to give the operator a well-illuminated image with excellent contrast and resolution. The most widely used optical hysteroscopes have an outer diameter (OD) of 3–4 mm, but thinner rigid scopes with fibre-optics and an OD of less than 2 mm have been developed.

Fibre-optic hysteroscopy was developed as a simple mechanism for viewing the uterine cavity with the use of gas as distending media. Fibre-optic hysteroscopy is inferior to the direct optical hysteroscopy for evaluation of the uterine cavity because of a lower contrast and resolution of the images that are a composite of individual fibres.

Both optical and fibre-optic hysteroscopes are monocular and provide little depth perception. They are available with different viewing angles, from 0 to 70°. Optical 30° scopes are most commonly used for diagnostic and operative procedures. The hysteroscope is inserted, securely fastened or permanently affixed into a metallic sleeve or sheath. Diagnostic sheaths generally have an OD from 2.5 to 4.5 mm and operative sheaths have an OD from 3.5 to 6.5 mm.

Modern diagnostic and operative sheaths have isolated, dual ports and provide continuous laminar flow of distending media. Continuous flow with separated in- and outflow channels guarantees the optimal irrigation and imaging of the uterine cavity and allows optimal intrauterine pressure and degree of cavity distension. In- and outflow can be regulated with separate stopcocks. Operative sheaths are designed to allow the passage of one or more operative instruments.

New developments in hysteroscopes and sheath were generally dominated by decreasing outer diameter without losing the quality of the image. For hysteroscopic adhesiolysis , this is extremely important since the operating field in the (obstructed) uterine cavity in patients with AS is often very limited in space. Although newer hysteroscopes provide separate in- and outflow channels, the inflow channel is often also designed as the working channel for introduction of surgical instruments. In case of a very small OD, the small fluid inflow channel can be mostly obstructed by the introduced instruments in such cases causing impaired visualization due to limited flow. Therefore further downsizing the OD of hysteroscopes should not impair channel size any further; more can be expected from the introduction of optical chip technology in hysteroscopes.

Various delivery systems are designed to suit the media used for uterine distention and to accurately record volumes of inflow and outflow. This recording is important because fluid can leave the uterus through the hysteroscope, by cervical or tubal leakage, or by means of intravasation. Pumps are available to monitor pressure and volume of liquid media.

Normal saline and lactated Ringer solution are isotonic, conductive, low-viscosity fluids that can be used for diagnostic hysteroscopy and for mechanical operative procedures. Excessive intravasation during hysteroscopic adhesiolysis is rather rare but it may occur; the isotonic fluid overload is generally readily treatable with diuretics. Therefore a higher amount of intravasation during surgery can be accepted. Generally, most protocols and guidelines mention approximately 1500 mL as the upper limit of saline intravasation.

Hypotonic, non-conductive, low-viscosity fluids like 5% mannitol, 3–5% sorbitol, and 1.5% glycine should only be used with electrical monopolar operative procedures.

One of the most important factors in performing operative hysteroscopy is maintenance of a clear operative field at the lowest intrauterine pressure. This can only be accomplished by the use of a sheath that provides continuous laminar flow of media. Media then usually flow into the uterine cavity through an inner sheath around the hysteroscope. A perforated outer sheath is used for collection or outflow of media. Whether the outflow tubing is connected to a vacuum collecting system or whether the outflow is passive is a matter of taste. Distention and visualization are improving with higher intrauterine pressure and flow; however, the lowest pressure and flow in which distention and visualization are acceptable is advised. This is to prevent fluid loss as much as possible and the original anatomy of the cavity and the extent of adhesions are recognized easier and better. An accurate assessment of the fluid deficit and lost media in the drapes, towels, or on the operating room floor ideally by a fluid management system is mandatory during extensive operative procedures. During diagnostic or minor operative procedures, fluid measurements are not that important because the risk of intravasation is very low during procedures that take only minutes to perform.

The uterus is a very vascular organ and its vascularity may be enhanced by uterine pathology. A low pressure and high flow usually creates a better visualization than a high pressure and low flow. However, intraoperative bleeding is generally not a significant problem because a rapid liquid flow will clean the image and after the procedure uterine contractions will diminish bleeding rapidly. Bladder catheterization is only indicated in case of high fluid loss.

An assortment of rigid, semi-rigid and flexible instruments has been developed or adapted for hysteroscopic surgery. The rigid and semi-rigid instruments include scissors, grasping forceps and biopsy forceps. Special care should be taken when handling them as the handle, shaft and tips can be easily damaged. Different hysteroscopical techniques have been described. In order to avert further damage of the endometrial lining, only conventional instruments should be used. High frequency electrical surgical instruments (electrodes or resectoscopes) should not be used in the treatment of patients with AS. The local and lateral spread of heat and electricity could eventually destroy parts of original vital endometrial lining.

Uterine adhesions are incised using a standard rigid operative hysteroscope. A blind dilatation should never be performed as it can ruin or destroy details like dark areas that absorb more light indicating an entrance to a (part of a) cavity or differences in colour or level of depth in the obstruction that might lead to the right way to find the remaining parts of the cavity. Following placement of the scope into the entrance of the endometrial cavity, hysteroscopic scissors of forceps are advanced through the operative channel and used to divide any noted adhesions (Figs. 63.1, 63.2. and 63.3). Filmy adhesions can be ruptured often by touching them with the sheath of the hysteroscope or even only by the pressure of the inflow of distension and irrigation fluid. Adhesiolysis should begin with the most centrally located adhesions and proceed to those located at the periphery of the cavity. Especially myometrial vasculature , showing the margins of the original cavity, is best recognized by the lowest intrauterine pressure and distension that allows visualization. With the use of conventional instruments like scissors and forceps, further damage to the remaining endometrial lining is prevented. It is not necessary, even in case of dense adhesions, to use monopolar or bipolar electrocautery electrodes, resectoscopic electrocautery needles or loops nor a NdYAG laser. Further damaging of the endometrial lining with electrocautery or laser should always be prevented. In cases involving lysis of extensive or dense adhesions with significant obliteration of the endometrial cavity, it may be beneficial to perform the procedure with concurrent guidance by fluoroscopy, ultrasonography or even laparoscopy.

Fig. 63.1
figure 1

Hysteroscopic image of intrauterine adhesions

Full size image
Fig. 63.2
figure 2

Conventional hysteroscopic adhesiolysis with cutting forceps

Full size image
Fig. 63.3
figure 3

Hysteroscopic image after successful adhesiolysis

Full size image

Hysteroscopic adhesiolysis is classified as the most difficult (ESGE and AAGL Standards and Guidelines) procedure of hysteroscopic surgery. The learning curve is long and this type of surgery is not generally implemented in daily practice. Besides, surgical trainees and gynaecologists perform less surgery than in the past because of limited work hours. This might have an important effect on the surgical volume of physicians in general. In a study we examined the case volume of surgeons by looking at the number of hysteroscopic myomectomies performed across the years [12]. High volume surgeons were significantly more effective than low volume surgeons, without any significant difference in complications.

The risk of infection after operative hysteroscopy is low, thus prophylactic antibiotics are not routinely administered during operative procedures. The only RCT ever performed to study the effect of prophylactic antibiotics on the incidence of bacteraemia following hysteroscopic surgery demonstrated no convincing evidence that antibiotics are of value in this setting [13].

4 Simultaneous Guidance

In the need of further guidance for the hysteroscopic adhesiolysis, fluoroscopic control can be used but there are different modalities like ultrasound or laparoscopy. There is no golden standard nor is there any evidence that one method is superior to the other. Different authors reported on the guidance used in their series. In Table 63.1, some of these series and the guidance used are listed.

Table 63.1 A selection of studies describing simultaneous guidance
Full size table

The benefit of abdominal or transvaginal ultrasonography is rather limited. Sometimes it can be used to demonstrate an area of proliferated endometrium in the upper part of the uterine cavity.

The advantage of fluoroscopy is that it can identify ‘blocked’ areas of original cavity and endometrium behind or above adhesions that are obscured hysteroscopically [15, 22] (Fig. 63.4). The contrast medium will flow through minor opening in the adhesions through which the hysteroscope cannot pass and will provide further information on the area’s beyond (Fig. 63.5a–f). Furthermore, tubal patency can be assessed during the procedure; this is of major importance as recognizing (at least one) tube(s) is a landmark of a proper anatomical restoration of the uterine cavity and it is reassuring for future fertility. Another benefit is the early detection of a false route or perforation.

Fig. 63.4
figure 4

Set-up of hysteroscopic adhesiolysis with fluoroscopic guidance

Full size image
Fig. 63.5
figure 5

(a–f) Series of fluoroscopic guidance images during hysteroscopic adhesiolysis

Full size image

Some authors favour laparoscopy to decrease the risk of perforation; however, laparoscopy will not prevent perforation, and it might reduce the risk of damage to the intestines.

Perforation of the uterus is a well-known complication of difficult hysteroscopic surgery. A prospective study in the Netherlands showed a cumulative incidence of 0.6% perforations in diagnostic and therapeutic hysteroscopies [23]. Other studies from the USA report an incidence of 1.5% for perforation [24]. In almost all studies, hysteroscopic adhesiolysis as treatment of AS is mentioned as the procedure with the highest incidence of perforation. When a perforation occurs during the introduction of the hysteroscope or by a not activated surgical conventional instrument like forceps and scissors used in a (non-blind) visual way, intra-abdominal bleeding and intestinal perforation are very rare and an expectative management seems justified. However when a perforation is caused by an activated instrument (either electrosurgical or morcellator), an emergency laparoscopy should follow to exclude intra-abdominal haemorrhage and further damage to the intestine.

5 Surg.ical Outcome

There is no consensus about how outcome should be defined. A few studies describe normalizing the uterine cavity anatomy as successful outcome. Fedele et al. reported, in 1986, 31 cases of IUA with 27 cases of complete adhesiolysis [25]. After several months, a repeat hysterography and hysteroscopy was performed and showed complete normalization of the uterine cavity in 62.5%. Pace et al. reported in 2003 anatomic normalization of the cavity in 70 out of 75 (93.3%) patients after surgery in AS [26]. Fernandez et al. reported, in 2006, 71 cases with stage 3 and 4 AS; reconstruction of a normal uterine cavity was achieved in 31 (43.7%) cases after the first attempt [27]. Capella and Allouc reported in 1999 about patients with severe intrauterine adhesions and in 51.6% of the 31 cases they described complete restoration of the cavity [20].

From 1986 onwards, studies were published with success defined as returning of normal or improved menses after surgery. A study from Fedele et al. from 1986 reported of 21 cases with a success rate of 67.7% [25]. Valle and Sciarra described in 1988 169 patients, their success rate was 88.2% [18]. Pabuccu et al. reported 34 cases with a success rate of 81% [21]. In 2000, Preutthipan et al. described 73 cases with a success rate of 90.9% [28]. In 2004, Zikopoulos et al. reported on 46 cases with a success rate 92.4% [29]. Yasmin et al. in 2007 described that in 96% of the 20 patients had a normal menstruation returned after hysteroscopic adhesiolysis [30]. In 2007, Thomson et al. described a success rate of 81.5% in 30 cases [31]. Finally, Robinson et al. in 2009 reported on 24 cases of which 95% had improvement of menstrual flow [32].

In a Dutch series of patients treated between 2003 and 2013, 97.8% restoration of menses was achieved [14]. The high success rate achieved in this group of patients probably reflects the effect of concentration of AS care in the Netherlands. It is a relative small country of 37.354 km2 with a population of almost 17 million people (density 450 km−2). Travel distances are relatively short and therefore centralization of difficult cases is common and easy accomplished. Asherman syndrome is a rather rare disease and its treatment is reviewed as the most difficult of hysteroscopic surgery (ESGE and AAGL Standards and Guidelines). Furthermore incompletely or inadequately restoration of the uterine cavity is a serious threat for future childbearing. Centralization of AS surgery seems of essential importance, not only because of the positive effect on the relation between volume and outcome, but also because of the improvement of support of women with AS related to patient-education, awareness and advocacy.

The last 25 years AS care was more or less concentrated in one centre in the Netherlands. Some surgeons have performed more than two thousand procedures (MHE).

Between 2003 and 2013, a total of 770 procedures were performed in 638 patients. After the first adhesiolysis, a normal uterine cavity was restored in 512 (80.3%) patients. However, in 46 of these patients, although their uterine cavity was completely normalized, only one tubal ostium could be visualized. The remaining 126 (19.7%) patients had an incomplete adhesiolysis after the first attempt. Of these patients, 114 choose to have a second procedure. Twelve patients refused a new attempt for various reasons. In the second attempt, 80 (70.2% of this remaining group) patients had a successful operation with a normalization of the uterine cavity; however, again in 19 patients only one tubal ostium could be visualized. After the second attempt, 34 (29.8%) patients were left with still an unrestored uterine cavity. A further 13 of these patients denied any further treatment for various reasons. In the remaining 21 patients, a third procedure was performed. In 14 patients, this attempt was successful; however, again in seven only one tubal ostium could be visualized after normalization of the cavity. So, in seven patients (this is 1.1% of the initial group of 638), the uterine cavity could not be restored after three procedures. See flowchart Fig. 63.6. It seems logical to conclude that the number of procedures should not be limited to one even if success rates after consecutive attempts are declining. This is consistent with findings in the literature [33].

Fig. 63.6
figure 6

Flowchart of treatment (HA hysteroscopic adhesiolysis )

Full size image

The perforation rate was in the first procedure 2.5%, in the second attempt 3.5% and in the third 4.8%. This perforation rate is lower than reported in the literature. Such a low complication rate could be contributed (amongst other reasons) by the use of fluoroscopic control as method of guidance. At the control visit hysteroscopy 2 months after surgery, the menstrual pattern was evaluated. Of all 638 patients, menstrual pattern was restored in 624 women (97.8%). Menses could not be restored (amenorrhoea) in 14 (2.2%) patients. A restored normal cavity anatomy, free of adhesions, with hysteroscopic visualization of a least one tubal ostium was accomplished in 606 patients. If the 24 patients, who refused a new attempt after incomplete or unsuccessful adhesiolysis, are taken into account the overall success rate (restore of menses and cavity anatomy) was 95.0%.

The mean age of all patients was 34.1 (SD ± 4.30, range 21–50). A majority of 497 (77.9%) patients had menstrual disorders. They suffered from amenorrhoea or hypomenorrhoea (64.4% respectively 30.1%). In 44 (6.9%) cases patients primarily suffered from subfertility and in 97 (15.2) cases patients had both. A total of 276 (43.3%) patients had been pregnant only once. A majority of 362 (56.7%) patients had more than one pregnancy and 320 (50.2%) patients had at least one life birth. In 99.8% of all patients, AS was preceded by a pregnancy-related intrauterine operative procedure. In more than half of all cases this was a first trimester curettage for miscarriage, termination of pregnancy or mola pregnancy (n = 371, 58.2%), in 243 (38.1%) patients this was a post-partum procedure and in 23 (3.6%) patients this was a caesarean section. Only one patient (0.2%) had a post-partum endometritis that presumably caused AS.

Repeated curettages are often mentioned as a high risk for the occurrence of AS [34, 35]. It’s important to distinguish between a repeated curettage because of a new pregnancy and a repeated curettage because of retained products in the same pregnancy [36]. This increased risk was not observed in the first group of patients. Subgroup analysis for first trimester procedures (miscarriage, termination of pregnancy and mola) shows that in 373 patients AS occurred in 239 (64.1%) patients after their first curettage, in 73 (19.6%) patients after a curettage in a new pregnancy and in only 61 (16.4%) after repeated procedures for persisted products of conception of the same pregnancy. Furthermore, patients with a first trimester curettage were more prone to have more mild adhesions (grade 1, 2 and 2a adhesions). Patients with post-partum problems had significant more type 3, 4 and 5 adhesions (Fig. 63.7).

Fig. 63.7
figure 7

Relation of pregnancy-related procedure and grade of adhesions (asterisk, ESGE/ESHRE classification)

Full size image

Patients with a first trimester curettage in their medical history had a better chance of a successful hysteroscopical or transcervical adhesiolysis (TCA) than patients with post-partum procedures. The last are more prone to cause more severe adhesions (grade 3, 4 and 5) and therefore lower the chance of a successful TCA. Furthermore, they have a higher chance of spontaneous recurrence of IUA. We advocate that post-partum remnants should be carefully approached and if possible hysteroscopic techniques should be used instead of blind curettages in order to minimalize the damage to the uterus. Especially hysteroscopic morcellation techniques seem to be favourable in preventing the formation of adhesions in the uterine cavity [37, 38].

6 Adjuvant Measures and Postoperative Care

Spontaneous recurrence of adhesions is reported in the literature by almost all authors. Variations from 3% to 60% reflect the enormous heterogeneity of all series described.

There is also a large variation in measures to prevent these recurrences.

An overview of the effect of some of these measures is listed in Table 63.2.

Table 63.2 A selection of studies describing measures to prevent spontaneous recurrences of adhesions
Full size table

Of 606 patients, who had a successful hysteroscopic adhesiolysis in our centre, adhesion spontaneously recurred in 28.7% (n = 174). Six (3.4%) patients did not want a procedure to resolve the recurrence, in 42 patients (24.1%) the adhesions were mild so they could be removed immediately during the second-look office hysteroscopy and in 47 patients (27.0%) this was accomplished only by sounding and/or dilation of the isthmic area. A new operative hysteroscopic adhesiolysis was needed in 68 (39.1%) patients and in 11 (6.3%) patients even a second procedure was needed to resolve these recurrent adhesions.

The spontaneous reformation of adhesions seems related to the initial grade of AS. If the initial grade of adhesions was moderate, grade 1–2a, the chance of spontaneous recurrence of adhesion varied from 20.8% (grade 1) to 25.0% (grade 2a). Grade 3 had a recurrence rate of 29.1%, grade 4 of 38.5% and grade 5 of 41.9%. A multivariate analysis of these results showed that apart from the grade of adhesions (p = 0.021) age also significantly influences the chance of a spontaneous recurrence (p = 0.044).

In the majority of our patients, standard measures to prevent recurrences were taken by placing an IUCD (of which the cupper was removed) and a therapy of adjuvant hormonal medication, both for six weeks. Already in 1964 Wood and Pena described oestrogen therapy to stimulate regeneration of the endometrium and improve its re-epithelialization [45]. A levonorgestrel containing IUCD was never used as it causes atrophy of the endometrium. This is also true for oral contraceptives; the continuous use of progesteron containing pills does not stimulate the endometrium but, on the contrary, it causes deep atrophy. Two months after surgery, a control or second-look hysteroscopy is always performed. If opposed endometrial surfaces were ‘sticking’ and the uterine cavity could be easily distended by fluid or by passing the hysteroscope sheath only, this was not classified as a recurrence.

Different methods are described for the (secondary) prevention of adhesions: postoperative placement of IUCDs or (special) foley balloon catheters, anti-adhesive agents and adjuvant hormonal treatment are the most mentioned and used options. One retrospective study suggests that the intrauterine balloon and IUD are more effective than the use of hyaluronic acid gel [40]. Intrauterine (spray) gels have been tested also [46]. Very recently a RCT was published, comparing IUDs (n = 80) and balloons (n = 82). The adhesion reformation rates (balloon group, 30%; IUD group, 35%) were not significantly different between the two groups [41].

In our experience, this spontaneous recurrence seems to be one of the most important problems with the various therapeutic approaches to intrauterine adhesiolysis. More RCTs are needed to demonstrate the efficacy of measures to prevent a recurrence after treatment for AS.

A RCT to evaluate the efficacy of adjuvant continuous oestrogen and sequential progesteron containing medication for six weeks to reduce the incidence of recurrent adhesions is undertaken at our centre. A report of the findings is expected in near future.

The external validation of the above-mentioned results can be lower than expected because surgery was performed by a low number of surgeons. To lower this uncertainty, very strict definitions are used.

The emerging field of tissue engineering has recently been applied in the reproductive tissues like the endometrium with elegant studies involving humans and animal models. The role of cell therapy with bone marrow-derived stem cells in treating endometrial dysfunction in Asherman syndrome and/or endometrial atrophy is described also [47].

In most of the patients with AS (95%), a normal uterine cavity can be restored with resumption of the menstruation. A real unsolved problem, however, is how to prevent the common occurrence of spontaneous recurrences. Future research should be focused on lowering this recurrence rate preferably by demonstrating the efficacy of the use of adjuvant measures in RCTs. For the primary prevention of AS, future research should be focussed on how to treat post-partum remnants, because their removal causes more severe IUA.

Furthermore, centralization of the treatment of AS is advocated because outcome and results are related with surgical volume and research is easier optimized because of a lower chance that RCTs suffer from lack of recruitment.