Chapter 4 – Diagnostic Hysteroscopy: Accuracy and Interpretation of Findings


Hysteroscopy can be used to diagnose endometrial and structural cavity pathologies associated with abnormal uterine bleeding (AUB) and reproductive failure. Optimal interpretation of hysteroscopic findings requires an understanding of the limitations of the technology and incorporation of information obtained from the preceding clinical history and examination. Ideally, hysteroscopy should be avoided during the menstrual period, as the obtained views are likely to be compromised. Although normal endometrial appearances during the secretory phase could potentially be misinterpreted (e.g. as polyps or hyperplastic endometrium), with experience the likelihood of this is small, and so timing the procedure to coincide with the proliferative phase of the menstrual cycle is not necessary or indeed practical

Chapter 4 Diagnostic Hysteroscopy: Accuracy and Interpretation of Findings

Ayesha Mahmud and T. Justin Clark

4.1 Introduction

Hysteroscopy can be used to diagnose endometrial and structural cavity pathologies associated with abnormal uterine bleeding (AUB) and reproductive failure. Optimal interpretation of hysteroscopic findings requires an understanding of the limitations of the technology and incorporation of information obtained from the preceding clinical history and examination. Ideally, hysteroscopy should be avoided during the menstrual period, as the obtained views are likely to be compromised [1]. Although normal endometrial appearances during the secretory phase could potentially be misinterpreted (e.g. as polyps or hyperplastic endometrium), with experience the likelihood of this is small, and so timing the procedure to coincide with the proliferative phase of the menstrual cycle is not necessary or indeed practical [1].

4.2 Accuracy and Utility

To gain the most out of hysteroscopy, the proficient practitioner should be aware of the indications for hysteroscopy and understand how to conduct a systematic and thorough endoscopic inspection of the lower and upper genital tract. Formulating a hysteroscopic diagnosis requires an assessment of the endometrium and the uterocervical cavity. The endometrial parameters and uterocervical features that need to be assessed are summarised in Box 4.1.

Box 4.1 Uterine Assessment

Endometrial Parameters

  1. 1. Endometrial thickness

  2. 2. Endometrial surface

  3. 3. Endometrial colour

  4. 4. Vasculature

  5. 5. Glandular openings

Uterocervical cavity features

  1. 1. Cervical canal

  2. 2. Uterine fluid

  3. 3. Uterine axis

  4. 4. Uterine shape

  5. 5. Uterine size

  6. 6. Uterine focal lesions or foreign bodies

Data from [1]

Hysteroscopy is accurate and represents the gold standard test for diagnosing intrauterine structural abnormalities, which may be congenital (e.g. uterine septae) or acquired (e.g. polyps, fibroids, adhesions) [2]. Discriminating between normal and abnormal endometrium at hysteroscopy is more difficult because there are no agreed standardised criteria for hysteroscopic diagnosis. The difficulty establishing such criteria stems from the fact that there is considerable overlap between the hysteroscopic features of the functional and pathological endometrium. In short, macroscopic visual inspection of the uterine cavity is fine for seeing architectural problems (i.e. structural lesions) but is limited in its ability to allow more subtle ‘histological’ diagnoses to be made. It has been proposed that broadly classifying hysteroscopic endometrial diagnoses into ‘normal’, ‘abnormal (thickened)’, ‘abnormal (suspicious)’ or ‘abnormal (cancerous)’ by evaluating the endometrial parameters outlined in Box 4.1 is more practical, reproducible and consistent with the capability of the technology [1].

Several reviews have reported evidence of the superiority of hysteroscopy in discriminating types of endometrial and structural uterine pathology [36]. Compared to saline infusion sonography (SIS) and transvaginal ultrasound, hysteroscopy has greater sensitivity and specificity [37]. Farquhar et al. compared all three modalities for investigating abnormal uterine bleeding in pre-menopausal women, reporting that transvaginal ultrasound had a higher false-negative rate than SIS and hysteroscopy [4]. In addition, they found that hysteroscopy was far superior to SIS in diagnosing submucosal fibroids [4].

In the context of diagnosing endometrial pathology, a prospective comparative study of these modalities reported significantly better performance of hysteroscopy both as a diagnostic tool and in terms of precision for the diagnosis of intracavitary masses [7]. A systematic review and meta-analysis of the accuracy and feasibility of hysteroscopy for diagnosing intrauterine anomalies in AUB estimated the overall accuracy rate of hysteroscopy to be 97% [5]. The review showed that hysteroscopy was useful for both identifying and excluding these intrauterine structural abnormalities (polyps and fibroids) because estimates of the positive and negative likelihood ratios substantially altered the probability of abnormality following testing, thereby being of value in informing subsequent clinical management [5]. Regarding endometrial pathology, a large systematic quantitative review by Clark et al. evaluated over 26 000 cases examining serious endometrial conditions, such as endometrial hyperplasia and cancer, in women with AUB. The majority of these cases represented post-menopausal bleeding (PMB). The review reported a high diagnostic accuracy of hysteroscopy for confirming disease but a lower (but moderate) accuracy for ruling out disease [6]. Thus, endometrial sampling would still be advised in PMB where the prevalence of premalignant and malignant endometrial disease is high. Table 4.1 shows the accuracy estimates (sensitivity and specificity, positive and negative likelihood ratios) and the post-test probabilities (pre-test probabilities assumed to be the underlying prevalence reported in the source reviews) with hysteroscopy for the diagnosis of common endometrial pathology [5, 6]

Table 4.1 Accuracy of hysteroscopy for common endometrial pathology

Endometrial pathology Sensitivity % Specificity % Likelihood ratio Pre-test probability, % (prevalence) Post-test probability, %
Positive Negative Positive Negative
Endometrial polyp [5] 95.4 (87.4–98.4) 96.4 (93.7–98.0) 12.9 (8.0–20.9) 0.09 (0.06–0.14) 29.8 0.85 (0.77–0.90) 0.04 (0.03–0.06)
Submucosal fibroid [5] 97.0 (89.8–99.2) 98.9 (93.3–99.8) 24.7 (9.0–68.2) 0.16 (0.09–0.27) 23.4 0.88 (0.73–0.95) 0.05 (0.03–0.08)
Endometrial disease (hyperplasia/cancer) [6] 78.0 (76.3–79.6) 95.8 (95.6–96.1) 10.4 (9.7–11.1) 0.24 (0.22–0.25) 10.6 (0.2–11.0) 55.2 (52.4–57.8) 2.8 (2.4–3.0)
Endometrial cancer [6] 86.4 (84.0–88.6) 99.2 (99.1–99.3) 60.9 (51.2–72.5) 0.15 (0.13–0.18) 3.9 (3.7– 4.2) 71.8 (67.0–76.6) 0.6 (0.5–0.8)

Data presented as median (95% CI)

The diagnostic limitations of the technology can be overcome by the correct interpretation of visual findings at hysteroscopy, optimising subsequent actions based on the visual analysis. For example, hysteroscopy allows structural lesions to be excised and directed biopsies of suspicious areas taken for histological assessment, as well as enhancing the interpretation of global ‘blind’ biopsies, especially if they are non-diagnostic. In the presence of congenital abnormalities or submucous fibroids, further testing with MRI or ultrasound can be arranged. Thus, while hysteroscopy can be used in certain circumstances as a stand-alone test in conjunction with information obtained from the clinical history and examination, it also complements other investigative modalities such as imaging, microbiological screening and endometrial biopsy.

Structural lesions can be diagnosed at hysteroscopy with the requisite knowledge, practice and experience to interpret the appearance of the endometrium, which includes discriminating endometrial pre-malignancy (atypical endometrial hyperplasia) and malignancy from benign endometrial conditions. The following mini-atlas details the key features of common diagnosable conditions.

4.3 A Mini-Atlas of Hysteroscopy

4.3.1 Structural Anomalies

Endometrial Polyps

Endometrial polyps are benign pedunculated protrusions arising from the uterine endometrium that can affect pre- and post-menopausal women. They commonly present with a history of AUB (heavy menstrual, intermenstrual or PMB), detected as part of a subfertility workup or as an incidental finding on imaging [811]. Although the majority of these polyps are benign, up to 6% may show pre-malignant or malignant changes in post-menopausal women, compared to 2% in pre-menopausal women [12, 13]. Polyps may also interfere with fertility by hindering implantation of the fertilised embryo [14]. Therefore, hysteroscopic polypectomy is performed with the aim of alleviating abnormal uterine bleeding, aiding fertility and excluding pre-malignancy.

Hysteroscopically, an intrauterine polyp appears as a discrete outgrowth of the endometrium, attached by a pedicle, which moves with the flow of the distension medium (Figure 4.1). These localised overgrowths of endometrial tissue can occur anywhere in the uterine cavity, be either single or multiple, and vary in size from a few millimetres to several centimetres. Endometrial polyps contain variable amounts of glands, stroma and blood vessels that are covered by a layer of endometrium. The intracavity formation is usually soft (glandular), but may be firm and fibrous. The covering endometrium may be thin (appearing translucent and revealing a delicate vascular network) or thick (obliterating surface vascularity) and should be carefully assessed to exclude features suggestive of hyperplasia (see below).

Figure 4.1 Examples of benign endometrial polyps from six different women illustrating the variety of shapes and sizes.

(Source of images: Mary E. Connor.)

Vascular polyps typically have a reddish-pink appearance in contrast to more fibrous polyps, which are whitish-grey. Most polyps have a single feeding vessel that branches in a regular pattern. Their surface may be regular or irregular and contain areas of translucent cyst formation, necrosis (yellow appearance) or petechial haemorrhage. They may appear as elongated ‘pedunculated’ narrow-based lesions or broader-based ‘sessile’ mucosal protrusions. Usually, they can be distinguished from submucosal fibroids by their movement pattern and softer texture with hysteroscopic indentation and uterine distension. Diagnosis is easier when the background endometrium is thinner, such as in the proliferative phase of the menstrual cycle or where the endometrium is inactive, as in post-menopausal women.

At present, there is no widely adopted or validated hysteroscopic classification system to determine the nature of a polyp and make a reliable estimate from collections of certain features as to whether there is a high probability of a polyp containing atypical hyperplasia or frank malignancy (Figure 4.2). However, despite this lack of consensus, the key features of polyps to record are the quantity, size, location and external appearance, which includes their colour, surface regularity, vascularity (number of feeding vessels and characteristics of the plexus, such as regularity, irregular branching and prominence) and other features such as areas of necrosis or cystic spaces.

Figure 4.2 Examples of malignant endometrial polyps from two women, demonstrating irregular surface blood vessels.

(Source of images: Mary E. Connor.)

Submucous (Intracavity) Fibroids

Fibroids are benign growths arising from the smooth muscle layer of the uterus. They vary in size, location and position within the uterine myometrium. Some 5–10% of fibroids are submucosal, with some degree of protrusion into the uterine cavity. Clinically, they usually present with symptoms of AUB especially heavy menstrual bleeding, and may also be associated with infertility, recurrent miscarriage or pain [1517].

Hysteroscopically, they appear as protrusions into the uterine cavity of solid, dense, and usually regular, whitish tissue. The degree to which they project into the uterine cavity varies, as does their surface area, but the pathognomonic feature is the whitish, vascular appearance caused by the usually thick pink overlying endometrium being ‘stretched’ over the projecting fibroid surface, revealing an array of thin, fragile surface blood vessels (Figure 4.3). Although submucosal fibroids by definition have some intracavity component, wholly intramural fibroids adjacent to the endometrial mucosal surface without causing distortion (International Federation for Obstetrics and Gynaecology (FIGO) type 3 fibroid, described in more detail in the next paragraph and in Section 12.4) can be suspected by the appearance of fragile surface vessels just underneath the endometrial surface. In the absence of discrete submucous projection, multiple intramural fibroids within the uterus can be suspected as they often impart a gentle undulating appearance to the uterine cavity.

Figure 4.3 Submucosal fibroids from six different women. (a) FIGO type 0. (b) FIGO type 1. (c) FIGO type 2. Type 0 fibroids in particular are a cause of abnormal uterine bleeding and they are also associated with reduced fertility.

(Source of images: Mary E. Connor.)

FIGO has defined three types of submucosal fibroids based on the proportion protruding into the endometrial cavity: type 0, completely within the cavity; type 1, ≥50% within the cavity; type 2, <50% within the cavity [18]. This can help determine the ease with which the fibroid can be resected hysteroscopically; the greater the intramural component, the more difficult the resection. The intracavitary projection of the fibroid (FIGO type) can be estimated hysteroscopically by observing the angle of the fibroid in relation to the endometrium at the attachment with the uterine wall; the greater the angle, the lower the proportion within the cavity.

The STEP-W classification is a more comprehensive pre-operative classification of submucous fibroids, where a score is generated to indicate the degree of surgical complexity [19]. In addition to recording the FIGO degree of myometrial penetration, STEP-W classification considers other parameters such as the extension of the base of the fibroid with respect to the wall of the uterus, the size of the visible intracavity fibroid in centimetres, the surface area of the base of the fibroid with respect to the uterine wall, and its location.

Intrauterine Adhesions

Intrauterine adhesions (synechiae) are associated with damage to the basal layer of the endometrium, with the pregnant endometrium being particularly vulnerable [13]. Uterine infection and trauma, such as from excessive curettage or instrumentation, are thought to contribute to adhesion formation [20]. At worst, the uterine cavity is constricted or even obliterated by fibrosis, rendering the endometrium inactive and resulting in amenorrhoea and infertility. These associated symptoms are referred to as ‘Asherman’s syndrome’. Uterine adhesions are usually diagnosed at hysteroscopy, which shows several fibrous bands of scar tissue coursing through the uterine cavity (Figure 4.4). They may appear as isolated or extensive bands of white or grey tissue, which lead to varying degrees of uterine cavity distortion. The adhesions may be mild and ‘filmy’ or more dense and fibrous in nature. On occasion, the uterine walls may be fused together with no evidence of a cavity. The uterine fundus and cornual recesses may be partially or completely occluded. Several hysteroscopic classification systems have been developed, but there is no overall consensus on the preferred nomenclature. The main components of the classifications are descriptions of the nature, site and extent of the adhesions, and some correlate these with the menstrual pattern.

Figure 4.4 Dense intrauterine adhesions can significantly compromise the uterine cavity and are a cause of infertility and abnormal uterine bleeding.

(Source of images: Luis Alonso Pacheco.)

Clinically, patients usually present with a history of infertility, recurrent miscarriage or menstrual disturbance (amenorrhoea or oligomenorrhoea) [20]. Successful restoration of the uterine cavity shape by hysteroscopic adhesiolysis, and without recurrence of adhesions, has been reported in up to 95% of cases [21]. However, while it is true that hysteroscopic adhesiolysis can restore menstrual activity and increase chances of conception, robust data on long-term fertility outcome are lacking [22]. Moreover, the more severe the adhesions, the worse the outcome – with a higher risk of intraoperative perforation, incomplete adhesiolysis and reformation of adhesions – despite measures to minimise these risks. In more severe cases, the endometrium may not reactivate even if the cavity is restored. It is important to be aware of which women are at risk, and to take preventative measures. Further research is required into reducing fibrosis after surgical treatment and stimulating endometrial regeneration.

Congenital Malformations (Müllerian Duct Anomalies)

Hysteroscopy can be used to evaluate congenital uterine anomalies arising from abnormal development or fusion of the Müllerian (paramesonephric) ducts during embryonic development. Müllerian duct abnormalities may interfere with fertility. Depending on the type of abnormality, women may present with symptoms of recurrent miscarriage, infertility or fetal malpresentation [23]. The nature of the anomaly depends upon the type and timing of the abnormal embryonic development of the Müllerian duct. Many classifications have been proposed. The most common anomalies encountered are arcuate and bicornuate uteri or the presence of a uterine septum, which may be complete (septate) or incomplete (subseptate) [24]. A unicornuate uterus or an underdeveloped (hypoplastic) uterus may also be encountered. The hysteroscopic characteristics of these abnormalities are described below [2526].

Arcuate Uterus

This anatomic variant arises as a result of near-complete resorption of the uterovaginal septum. The fundus appears to gently bulge into the uterine cavity so that the cornual recesses are more pronounced, producing a space (<1 cm) between the tubal ostia and the deepest median point of the fundus (Figure 4.5). Some regard this as a normal physiological variant rather than an abnormality linked with suboptimal reproductive outcomes.

Figure 4.5 An arcuate uterus is regarded by some as a minor Müllerian anomaly with (a) a concave contour at the fundus and (b) an extended cornual recess of ≤1 cm. Whether this anomaly is related to fertility problems remains debated.

(Source of images: (a) Alexandra Sutcliffe and (b) T. Justin Clark.)

Bicornuate Uterus

Partial non-fusion of the Müllerian ducts characterises this class. A two-chambered uterus is discovered upon hysteroscopy, with a central division extending into the uterine cavity (Figure 4.6).

Figure 4.6 Bicornuate uterus with atrophic endometrium. A post-menopausal woman may well be aware of her uterine anomaly. The thin endometrium allows a clear view of the uterine architecture.

(Source of image: Mary E. Connor.)

The diagnosis should be suspected when only one tubal ostium or cornual recess is seen in a narrow, cylindrical uterus [26]. This appearance could represent a unicornuate or hypoplastic uterus, but a careful exploration for an additional entrance into another uterine horn must be carried out. A bicornuate uterus cannot be distinguished hysteroscopically from a completely septate uterus without additional investigation with ultrasound, MRI or laparoscopy (Figure 4.7). However, the central ‘divide’ in a bicornuate uterus is generally wider and pinker than the narrower, whiter apex seen with an incomplete uterine septum.

Figure 4.7 Septate or bicornuate uterus? These two images are from women with a septate uterus, but differentiating this from a bicornuate cavity can be difficult with hysteroscopy alone and requires additional imaging or laparoscopy. One horn of a bicornuate or double uterine cavity can be mistaken for a unicornuate uterus; careful withdrawal of the hysteroscope may reveal a second cavity.

(Source of images: Luis Alonso Pacheco.)

Unicornuate Uterus

This anomaly results from complete or near-complete arrested development of one of the Müllerian ducts. A unilaterally formed uterus can be visualised at hysteroscopy where only one ‘half’ of the uterus is accessible through the single cervical canal, and one tubal ostium will be seen on the left or right according to which communicating uterine horn has developed. The hysteroscopist should withdraw the hysteroscope carefully and move it towards the contralateral side to look for any communication, i.e. a channel into the other ‘half’ of the uterus. Similarly, a careful vaginoscopic inspection of the upper vagina in conjunction with a pelvic examination should be undertaken to avoid overlooking a complete uterine septum or uterine didelphys. The latter is a consequence of complete non-fusion of the Müllerian ducts, resulting in a ‘double’ uterus with only one ostium in each cavity; there may even be a double cervix with a complete vaginal septum [26] (Figure 4.8).

Figure 4.8 A complete vaginal septum with double uterus and double cervix: views show (a) the vaginal septum, (b) the right cavity, (c) left cavity, (d) left cervix (with yellow arrow on edge of divided vaginal septum) and (e) right cervix (with white arrow on identifying mark on anterior lip enabling confirmation of the double cervix).

(Source of images: Mary E. Connor.)

Uterine Septum

Hysteroscopy is particularly useful for both the diagnosis and the treatment of a septate uterus [26]. A septate uterus develops when the final fibrous septum between the two Müllerian ducts partially or fully fails to resorb and is one of the most common forms of congenital uterine malformation. An incomplete uterine septum (subseptate uterus) partially separates the uterine cavity into two chambers. The septum is covered with endometrial mucosa, and the more the septum projects into the uterine cavity, the more easily the other uterine chamber can be overlooked. The septum is usually thin proximally, becoming wider towards the fundus. A complete uterine septum (septate uterus) separates the uterine cavity completely (Figure 4.9) so that only one tubal ostium and cornual recess is seen [2526]. Careful withdrawal of the hysteroscope and exploration for an entrance into another cavity is necessary to bypass any intervening tissue.

Figure 4.9 The septate uterus is a congenital anomaly caused by complete or incomplete absence of involution of the central portion of the Müllerian ducts as they fuse during embryological development. The persisting fibromuscular septum is associated with early pregnancy loss and infertility. (a) View of cavity on the right, (b) view of septum and (c) view of left cavity.

(Source of images: Luis Alonso Pacheco.)

Removal of the uterine septum or septoplasty may be associated with better reproductive outcomes with a lower associated risk of uterine rupture [27]. However, high-quality study data are lacking, and the results of a randomised controlled trial comparing hysteroscopic septoplasty versus expectant management in women with either subfertility or recurrent miscarriage are eagerly awaited (TRUST trial 2017) [28].

Uterine Hypoplasia

Some degree of developmental failure of the Müllerian ducts results in an underdeveloped or hypoplastic uterus. Hysteroscopically a small, cylindrical uterine cavity (typical uterine length <6 cm) is detected [26] (Figure 4.10). The contracted uterus is frequently T-shaped, with normally formed cornual recesses, but a tubular uterine body with thickened myometrium, making the ostia difficult to access. Hysteroscopic metroplasty procedures to expand a hypoplastic (dysmorphic) uterus have been described, which involve using a miniature bipolar electrode to incise or ‘score’ the uterine sidewalls [29].

Sep 17, 2020 | Posted by in GASTROENTEROLOGY | Comments Off on Chapter 4 – Diagnostic Hysteroscopy: Accuracy and Interpretation of Findings

Full access? Get Clinical Tree

Get Clinical Tree app for offline access