16.1.1 Definition and Diagnosis

With the rising rate of caesarean sections (CS) and the increasing capabilities of ultrasound and hysteroscopy, there is growing interest in the caesarean scar in the non-pregnant woman. A caesarean scar is visible with transvaginal ultrasound (TVS) (Figure 16.1), hysterosalpingography (HSG) (Figure 16.2), magnetic resonance imaging (MRI) (Figure 16.3), sonohysterography (SHG) (Figure 16.4) or hysteroscopy (Figure 16.5).

Figure 16.1 Transvaginal ultrasound. Sagittal view of the uterus with a defect in the anterior wall of the cervix with niche formation (yellow arrow). Blue arrow, bladder; pink arrow, endometrium.

Figure 16.2 Hysterosalpingography. (a) The balloon of the catheter is located in the cervix, with dye deposited in the anterior wall defect in the cervix and a polyp in the left cornua. (b) After emptying the uterine cavity, there is still dye in the anterior wall defect of the cervix.

Figure 16.3 Magnetic resonance image showing caesarean section niche at tip of the arrow.

Figure 16.4 Caesarian section niche. (a) Sonohysterographic view. (b) Corresponding hysteroscopic view.

Figure 16.5 Hysteroscopic views. (a) Anterior wall defect of the cervix. (b) Caesarean scar niche.

A niche in the scar is defined as a hypoechogenic triangular area at the presumed site of the scar of a non-pregnant woman, seen by ultrasound. Different terminology has been used (caesarean scar defect, scar dehiscence, scar insufficiency, isthmocele, pouch) and the definition of the niche is still under debate [1, 2]. Some of the terms used imply a deficient scar, although there is not yet a proven relationship between the appearance of the scar and its function [2].

The prevalence of a niche in populations of women who have had a CS when randomly sampled is 24–70% with TVS, 56–84% with SHG and 33–75% with HSG [2, 3]. SHG gives a higher prevalence of niches than TVS in the same population; Osser and colleagues found a niche prevalence of 84% on SHG and 70% on TVS in the same population of women [4]; likewise, Bij de Vaate et al. and van der Voet et al. found a higher percentage with SHG than TVS (56% vs 24%, and 64% vs 50%, respectively) [5, 6]. SHG is reported to be more accurate owing to better delineation of the borders of a niche [4].

There are no data for prevalence investigated by hysteroscopy in random samples of women from similar populations. Two studies correlated the prevalence by TVS/SHG with hysteroscopy in women with gynaecological symptoms and found a prevalence of 31% and 88%, respectively [7, 8].

A sonographic description of the CS niche has been proposed [9] that, when used with very specific measurements of different parts of the niche, could prove useful for both research and clinical management. Using the suggested terminology, a niche must consist of an indentation of at least 2 mm, with a simple niche consisting of a single indentation. More complex niches include one or more ‘branches’, where a branch is a thinner part of the main niche directed towards the serosa (Figure 16.6). It has yet to be determined whether differences in niche complexity are of clinical significance.

Figure 16.6 Diagrammatic illustration of niche complexity. (a) Simple niche with a single indentation. (b) Complex niche with a branch (blue) extending towards the serosal surface from the main niche (red, green). The main niche is wider than the branch. (Modified from [9].)

16.1.2 Risk Factors

The reported predisposition for developing a niche, as seen with TVS in an unselected population, was multiple CSs [10]. Risk factors for large niches in a prospective study were the presenting part of the fetus below the pelvic inlet, cervical dilatation of more than 5 cm, and duration of labour of more than five hours. One randomised controlled trial comparing full thickness and split thickness suturing at CS found a reduced incidence of niches with full thickness suturing [11]. No significant differences were found between one- and two-layer closure, but one-layer closure resulted in a thinner residual myometrium [12].

16.1.3 Related Symptoms

Reported complications of a niche include caesarean scar pregnancies, malplacentation, perforation of the scar when inserting an intrauterine contraceptive device, and possible infertility [13, 14]. Other reported symptoms in women with a niche were dysmenorrhoea (53.1%), chronic pelvic pain (36.9%) and dyspareunia (18.3%) [15]. A recent meta-analysis reported that CS on average reduced the probability of subsequent pregnancy by 10% (relative risk, 0.91; 95% confidence interval (CI), 0.87–0.95) in comparison to a vaginal delivery [16]. None of these studies evaluated the relationship between the presence of a niche and subsequent fertility. However, intrauterine fluid, or cervical mucus or blood accumulation in association with a niche, are expected to hamper the penetration of sperm cells or impair embryo implantation [3].

Several studies reported an association between abnormal uterine bleeding and a niche [5, 6, 8, 15, 17]; post-menstrual spotting, especially a brownish, bloody discharge after menstruation, is associated with a niche. Two recent prospective studies reported that patients with a niche have more post-menstrual abnormal uterine bleeding (PAUB) than those without (odds ratio (OR), 3.1; 95% CI, 1.5–6.3 [5] and OR, 6.13; 95% CI, 1.74–21.63) [6].

The volume of the niche is also correlated with post-menstrual spotting, with large niches (depth more than 50% of the adjacent myometrium) producing more frequent PAUB than smaller niches [5, 6]. Wang et al. described significantly wider niches in women with PAUB, dysmenorrhoea or chronic pelvic pain than women with none of those symptoms. Also, the prevalence of AUB and prolonged menstruation was higher with a larger diameter niche [15].

16.1.4 Hypothesis

AUB may be due to the retention of menstrual blood in the niche, which is intermittently expelled after menstruation has almost completely ceased, causing post-menstrual spotting and pain [18, 19]. The presence of fibrotic tissue below the niche may impair the drainage of menstrual flow [17]. Additional, newly formed, fragile vessels in the niche may also contribute to the accumulation of blood produced in situ [20].

16.1.5 Therapy

The first choice of treatment for PAUB symptoms is conservative medical treatment to reduce the amount of blood during menstruation, and this includes (continuous) oral contraceptives, continuous progestogens (oral or systemic) or the levonorgestrel intrauterine device. If conservative treatment is not effective, then surgical options can follow.

Laparoscopic repair (with or without robotic assistance), vaginal repair and hysteroscopic resection are reported therapies. A recent review evaluated 12 studies of minimally invasive therapy in 455 women; eight of these studies investigated hysteroscopic resection [21]. AUB improved in 87–100% of patients without severe complications. However, the methodological quality of the selected papers was considered to be moderate to low, and there is a lack of randomised trials comparing hysteroscopic therapy with conservative therapies, so no solid conclusions can be drawn as yet.

16.1.6 Hysteroscopic Appearance of the Niche

A hysteroscopic classification of a niche is lacking, and different definitions exist: a cavity at the scar site with a superior and inferior fibrotic ring [8], a pouch-like defect [22], a diverticulum with or without mucosa [23], a dome-shaped scar defect with nodules of endometrial hyperplasia and vascular hyperplasia [24], an anterior outpouching in the canal and exposed dilated vessels [25], a pouch on the anterior uterine segment [18]. Prospective data investing correlations between hysteroscopic appearances and clinical symptoms are also lacking. Two studies correlate TVS and hysteroscopy findings in a population of patients with abnormal uterine bleeding or infertility; all niches found with TVS were also detected with hysteroscopy [7, 17]. One study performed diagnostic hysteroscopy in 20 patients with a history of CS and PAUB and described an enlargement followed by retraction of the anterior wall of the cervix resulting in a pseudocavity in 18 patients. Within this cavity, in some patients, polyps and brown bloody discharge were visible [26].

The best way to investigate a caesarean scar during hysteroscopy is to withdraw the scope from the uterine cavity back into the cervical canal with continuous inspection of the anterior wall of the lower uterine segment and the cervix (Figure 16.5). A discontinuation of the normal lining of the myometrium or cervical tissue will be visible if there is a niche. The use of liquid distension medium during hysteroscopy allows for the niche to be cleaned and rinsed. Large niches, especially in the retroverted uterus, can be misinterpreted as the uterine cavity. Niches can be seen in the cervix or in the low anterior uterine segment, depending on the progress of cervical dilatation at the time of the caesarean section [23, 27]. The niche lining may consist of fibrotic tissue, polyp-like structures, endometrium with dilated or abnormal vessels, or serosa. Histological samples taken by hysteroscopic resection showed inflammation in 78%, fibrotic tissue in 7–17%, hyperplastic polyps in 28%, and abnormal endometrium in 54% [25, 28]. Niches often contain mucus or dark haematinic material.

16.1.7 Hysteroscopic Resection

Hysteroscopic resection of the niche is a minimally invasive procedure that can be performed as a day care procedure. The resection aims to facilitate drainage of the menstrual blood and to reduce in situ production of blood by coagulating the niche vessels.

Fernandez and colleagues were the first to describe this technique in 1996 at an annual meeting of the American Association of Gynecologic Laparoscopists (AAGL) [29]. Since then, several authors have reported the use of this technique in case series or cohort studies using resectoscopes varying in diameter from 8 to 12 mm. Large prospective cohort studies and randomised controlled trials on the effectiveness and safety of the hysteroscopic resection are lacking. Nine cohort studies and case studies reporting on hysteroscopic resection are summarised in Table 16.1. Shih and colleagues reported on the hysteroscopic resection technique with an instruction video but without providing patient data and so this is not included in this table [30].

Most studies in Table 16.1 were performed using a monopolar resectoscope, with rollerball coagulation of the niche and sorbitol/mannitol for irrigation under general anaesthetic in day care. Diagnoses were made by ultrasound combined with diagnostic hysteroscopy.

16.1.8 Hysteroscopic Technique

Before the start of a therapeutic resection, transvaginal sonography or sonohysterography is advised to measure the minimal thickness of the anterior uterine wall above the niche and to identify any lateral branches at the margins of the niche [9]. Subsequently, a diagnostic hysteroscopy can be considered prior to cervical dilatation to reduce the risk of perforation; it also facilitates the evaluation of the entire niche with full assessment of the lateral aspects and branches [22–25].

After cervical dilatation to 9–12 mm, depending on the diameter of the resectoscope used, a monopolar or bipolar resectoscope is introduced. The proximal anterior rim of the niche is removed with a cutting loop on a pure cutting current (60–80 W) (Figure 16.7a). The cutting loop is applied to the roof of the niche and then withdrawn towards the external os in a line parallel to the axis of the cervical canal. Polyp-like structures and inclusion cysts are removed as well. The surface of the remaining niche is coagulated or fulgurated with a rollerball to coagulate the vessels and endometrium-like glands to prevent production of mucus and blood in the niche itself [22–25]. Some surgeons combine the resection of the distal rim with a resection of the proximal rim in order to equalise the niche [23, 27, 31] (Figure 16.7b and Video 16.1). However, its effect on cervical incompetence or preterm delivery of a subsequent pregnancy has not yet been evaluated.

Figure 16.7 (a) Resection of the proximal rim of the niche. (b) Resection of proximal and distal rims.

To prevent possible coagulation injury to the bladder it is advisable to conduct the hysteroscopic niche resection under simultaneous ultrasound control [22]. In addition, perioperative installation of methylene blue dye in the bladder can be considered in order to identify bladder injuries. Different cut-off levels of the depth of residual myometrium between the bladder and the niche are offered to prevent bladder injuries. Considering a coagulation depth of approximately 3 mm, it seems logical to take a cut-off level of at least 3 mm. Most reports describe a minimal residual myometrium of 2.5 or 4 mm [23, 32]. One publication took 2 mm as the cut-off [25]; others did not report on the minimal thickness. Bladder injuries have not been reported so far. Seven studies report on the evaluation of complications [22–25, 27, 31, 32]; two do not [19, 28].

The reported duration of resection varies between 8 and 30 minutes [23, 25, 32]. All studies report an anatomical success rate of 100%, but there were no standardised measurements of this effect. A control hysteroscopy was performed three months after resection in two studies, and the niche surface was found to be covered by cuboidal epithelium in all 60 treated patients [23, 27]. Most publications reported a 100% effect without any clear definition of success. Only two studies evaluated inter- or post-menstrual bleeding after the procedure and compared it to the pre-operative situation [22, 24]. In these studies, the mean reduction in post-menstrual spotting or menstrual bleeding following hysteroscopic niche resection was 2–4 days. Satisfaction rates varied between 59% and100% (Table 16.1).

Pain symptoms were reduced in 97% of patients [23, 27, 31]. Pregnancy outcome was reported in 52 women after hysteroscopic resection, though four patients had a miscarriage. No other adverse outcomes were reported; however, selection bias cannot be excluded and detailed information about these pregnancies is lacking. All reported CSs delivered pregnancies at term [21]. Duration of subfertility and details of previous diagnostic workup in relation to the therapies received were reported in one case series [23]. This study included 41 subfertile women, all with fertility problems of unknown aetiology and of more than two years’ duration. All patients were reported to have conceived within two years without additional therapy.