Chapter 7 – Indications for Hysteroscopy


Improved hysteroscopic technology has enabled clinicians not only to perform diagnostic outpatient hysteroscopy but also to treat uterine lesions. Bipolar diathermy electrodes have reduced the risk of fluid overload, allowing the excision of larger fibroids at a single setting and thus reducing the need for abdominal operations. Novel indications for hysteroscopy have become available, such as hysteroscopic tubal occlusion. Concomitant developments in hysteroscopic equipment and ultrasound technology have influenced how these tools are used in the diagnosis and treatment of intrauterine conditions.

Chapter 7 Indications for Hysteroscopy

Dimitrios Mavrelos , Ertan Saridogan , Mary E. Connor and T. Justin Clark

7.1 Introduction

Improved hysteroscopic technology has enabled clinicians not only to perform diagnostic outpatient hysteroscopy but also to treat uterine lesions. Bipolar diathermy electrodes have reduced the risk of fluid overload, allowing the excision of larger fibroids at a single setting and thus reducing the need for abdominal operations. Novel indications for hysteroscopy have become available, such as hysteroscopic tubal occlusion. Concomitant developments in hysteroscopic equipment and ultrasound technology have influenced how these tools are used in the diagnosis and treatment of intrauterine conditions.

7.2 Development of Other Diagnostic Modalities

Alongside hysteroscopy, the second half of the twentieth century saw rapid developments in other diagnostic modalities in the field of gynaecology. These directly influenced the role of hysteroscopy in the diagnosis and management of uterine diseases. The earliest descriptions of transvaginal ultrasound probes date to 1955. However the real-time endovaginal probe did not become widely available until the Austrian company KretzTechnik produced the first in 1985. Since then, the use of high-frequency probes has been adopted by most clinics worldwide. In the last three decades of the twentieth century, the value of transvaginal ultrasound for the assessment of women with menorrhagia, post-menopausal bleeding and pelvic pain, and for the diagnosis of uterine fibroids, endometrial polyps and other gynaecological conditions, has been established. More recently, 3D-capable transvaginal probes have been used to accurately diagnose congenital uterine anomalies.

7.3 Diagnostic Indications for Hysteroscopy

7.3.1 Abnormal Uterine Bleeding

Changes in the regularity and/or intensity of menstrual flow, or bleeding after the menopause, are some of the commonest reasons for seeing a gynaecologist [1]. These complaints have been grouped together by the International Federation of Obstetricians and Gynaecologists (FIGO) as abnormal uterine bleeding (AUB) [2]. FIGO summarised the causes of AUB using the acronym PALM-COEIN [2] (Table 7.1). Structural causes of AUB affecting the uterine cavity will be detectable and may be treatable by hysteroscopy.

Table 7.1 PALM-COEIN classification of abnormal uterine bleeding

Pathology Abbreviation Comment
Polyps AUB-P Endometrial and endocervical polyps may cause AUB; often asymptomatic
Adenomyosis AUB-A As determined by ultrasound views
Leiomyoma AUB-L Submucosal fibroids likely to be the most relevant, especially if completely within the cavity; often asymptomatic
Malignancy or hyperplasia AUB-M Uncommon but may occur in pre- or post-menopausal women; risk factors are recognised for atypical hyperplasia and endometrial cancer
Coagulopathy AUB-C Includes bleeding on anticoagulant therapies; 13% of women with HMB have von Willebrand’s disorder
Ovulatory dysfunction AUB-O PCOS and perimenopause
Endometrial disorders AUB-E No other abnormality apparent
Iatrogenic AUB-I Includes hormonal therapies
Not yet classified AUB-N Includes chronic endometritis, AV malformations, myometrial hypertrophy

AUB, abnormal uterine bleeding; AV, arteriovenous; HMB, heavy menstrual bleeding; PCOS, polycystic ovary syndrome

Data from [2]

For many gynaecologists, the first line of investigation for women with AUB is, when possible, a transvaginal ultrasound examination. This is reliable for the diagnosis of structural uterine abnormalities such as fibroids [3] and adenomyosis, but endometrial polyps are less reliably diagnosed and might not be differentiable from thickened endometrium [4]. Other means of imaging the uterus may provide more information but are less readily available. Three-dimensional transvaginal ultrasound enhanced by saline infusion into the endometrial cavity (3D-SIS) is equivalent in accuracy to hysteroscopy for the diagnosis of submucous fibroids [5]. However, because it can be used to globally assess the uterus and objectively measure fibroid dimensions, 3D-SIS may provide a better estimation of the likelihood of complete hysteroscopic resection [6].

Heavy Menstrual Bleeding

Heavy menstrual bleeding (HMB) is common and affects 20–30% of women who menstruate [1]. Some women will be managed satisfactorily within primary care, their clinical history and transvaginal ultrasound result having excluded many of the causes identified by the PALM-COEIN classification [2] (Table 7.1). Referral to secondary care is required for those who respond poorly to first-line treatment such as the levonorgestrel-releasing intrauterine system (LNG-IUS), or where a significant structural uterine abnormality has been found or serious uterine disease needs to be excluded [7].

Hysteroscopy is often reserved for women with AUB who have a suspicion of structural uterine pathology on transvaginal ultrasound. For example, 10–40% of women with AUB will have an endometrial polyp on scan [8]. In these women, hysteroscopy is indicated both to confirm the diagnosis and to perform treatment. Despite the widespread use of ultrasound, current evidence from studies evaluating the overall diagnostic performance of hysteroscopy in women with AUB confirms that hysteroscopy remains the gold standard for diagnosing endometrial abnormalities. The effectiveness of hysteroscopy in accurately assessing the uterine cavity was confirmed in a meta-analysis [9]. In addition, the overall success rate of diagnostic hysteroscopy was estimated to be 96.9% (standard deviation 5.2%, range 83–100%), meaning that the test will be feasible in the large majority of patients.

A recent economic analysis found outpatient hysteroscopy (OPH), alone or combined with endometrial sampling, to be the most cost-effective option when modelling a variety of diagnostic tests and combination testing strategies [10]. The various strategies were compared to a strategy of no further investigation and simple treatment with the LNG-IUS; in addition, OPH in a one-stop ‘see-and-treat’ clinic was the most cost-effective option given the scenario of unresolved HMB with LNG-IUS previously managed in primary care in women who wished to remain fertile [10].

The National Institute for Health and Care Excellence (NICE) has recently published its updated HMB evidence-based guidelines [7]. The 2018 update of the original 2007 guidance places a greater emphasis on outpatient hysteroscopy in diagnostic workup and removes any treatment hierarchy in secondary care for women with HMB and significant fibroids (any fibroid >3 cm in maximum diameter) (Figure 7.1).

Figure 7.1 Diagnostic pathway for HMB and the role of hysteroscopy [7].

Transvaginal ultrasound was considered the first-line diagnostic test in the evaluation of HMB because it is accessible and acceptable to women. Blind outpatient endometrial biopsy was generally considered the second-line test to evaluate the endometrium in secondary care because the equipment is cheap (although the subsequent pathological analysis is expensive) and can be easily integrated into a clinical pelvic examination. OPH was relegated to a third-line test in HMB because it requires a specific infrastructure that not all hospitals can provide and was considered invasive and expensive. However, many of these assumptions are inaccurate when subject to clinical [9, 11] and economic [7, 10] analysis and qualitative patient-centred research [12].

NICE conducted systematic reviews, network meta-analyses and economic evaluations to guide its decision-making, with the consequence that the usefulness of some previously considered investigations were challenged and the following points were made:

  • Transvaginal scans: the information provided by these scans does not inform clinical decision-making over and above clinical assessment in the absence of a palpable or non-informative clinical examination or where women complain of additional and related symptoms such as pelvic pain.

  • Endometrial biopsy: this may be more painful than outpatient hysteroscopy and also not as accurate as previously thought in detecting serious endometrial disease because of non-representative sampling, although this is more relevant in a post-menopausal population [13, 14]. In addition, unselected sampling or sampling restricted to women over 45 years of age [7] has a low diagnostic yield for serious endometrial disease, endometrial hyperplasia, pre-malignant disease (atypical endometrial hyperplasia) or malignancy (endometrial cancer).

  • OPH: this is well tolerated by most women (however, choice of setting and anaesthesia should be offered to all women) and is the gold standard investigation for diagnosing the most relevant structural pathologies (polyps and submucosal fibroids) that underlie HMB, treatment of which (surgical removal) can be effective in alleviating symptoms [12, 1517]. The inherent ability of hysteroscopy to most accurately diagnose and effectively treat in an efficient, minimally invasive manner drives the economic analyses that support the notion that diagnostic workup strategies based upon initial testing with OPH are the most cost-effective [10].

In light of these considerations, among others, NICE (2018) [7] has made recommendations regarding the indications for hysteroscopy in HMB, detailed in Box 7.1.

Box 7.1 National Institute for Health and Care Excellence (NICE) Heavy Menstrual Bleeding (HMB) Guidelines 2018

  • Take into account the woman’s history and examination findings when deciding whether to offer hysteroscopy or ultrasound as the first-line investigation.

  • Offer outpatient hysteroscopy to women with HMB if their history suggests submucosal fibroids, polyps or endometrial pathology because:

    • they have symptoms such as persistent intermenstrual bleeding or

    • they have the following risk factors for endometrial pathology:

      1. persistent intermenstrual bleeding

      2. persistent irregular bleeding

      3. taking tamoxifen

      4. infrequent heavy bleeding in those who are obese or have polycystic ovary syndrome.

  • Ensure that outpatient hysteroscopy services are organised and the procedure is performed according to best practice, including:

    • advising women to take oral analgesia before the procedure

    • vaginoscopy as the standard diagnostic technique, using miniature hysteroscopes (3.5 mm or smaller)

    • ensuring that hysteroscopy services are organised to enable progression to ‘see-and-treat’ hysteroscopy in a single setting if feasible

    • explaining to women with HMB who are offered outpatient hysteroscopy what the procedure involves and discussing the possible alternatives.

  • If a woman declines outpatient hysteroscopy:

    • offer hysteroscopy under general or regional anaesthesia [7]

    • consider pelvic ultrasound, explaining the limitations of this technique for detecting uterine cavity causes of HMB.

  • Consider endometrial biopsy at the time of hysteroscopy for women who are at high risk of endometrial pathology (see risk factors above).

  • Obtain an endometrial sample only in the context of diagnostic hysteroscopy. Do not otherwise offer ‘blind’ endometrial biopsy to women with HMB.

Data from [7]

Intermenstrual Bleeding

Intermenstrual (IMB) and post-coital bleeding may accompany a complaint of HMB in as many as 36% of women, as they may arise for the same reasons [18]. IMB as an isolated symptom is often correctly attributed to current contraceptive therapy or to physiology. However, investigation is indicated to exclude endometrial pathologies if it is persistent. Submucosal fibroids, endometrial polyps and other endometrial abnormalities (Figure 7.2) are all possible causes, as suggested by the PALM-COEIN classification system [2] (Table 7.1).

Figure 7.2 Hysteroscopic images of endometrial pathologies. (a) Endometrial polyps. (b) Endometrial polyp growing on top of type 1 submucosal fibroid. (c) Partially removed type 2 fibroid with large necrotic tissue remnant prior to repeat hysteroscopic resection.

(Source of images: Mary E. Connor.)

Post-menopausal Bleeding: Endometrial Cancer, Endometrial Hyperplasia

Post-menopausal bleeding (PMB) requires investigation as the prevalence of endometrial cancer or pre-malignant disease in this population is at least 5% [19]; in the UK, it is the fourth most common cancer in women [20]. Referral to a fast-track clinic to be seen within two weeks is advocated, especially in women 55 years of age and older, but should also be considered in younger post-menopausal women, particularly if they have risk factors for endometrial cancer [21]. These factors include obesity, nulliparity, hypertension, diabetes mellitus, history of polycystic ovaries or prolonged anovulation, tamoxifen therapy and hereditary non-polyposis colorectal cancer (HPNCC or Lynch Syndrome) [19]. Hysteroscopy can be useful in this situation, but published guidelines do not recommend it as a first-line investigation [2224]. Instead, initial transvaginal scan (TVS) measurement of the endometrial thickness (ET) is advised. Subsequent endometrial sampling is indicated when the ET is over the threshold of 3–5 mm, or when the endometrial view on TVS is unsatisfactory. This was shown to be the most cost-effective strategy for determining the presence of endometrial cancer [10], and use of such a protocol can avoid the need for an invasive test in some 30–40% of patients [25]. Endometrial sampling using a Pipelle device has a sensitivity of 99.6% for detecting endometrial carcinoma [26]. However, endometrial cancer can occasionally occur even with a very thin endometrium, so repeat or further investigation is warranted when PMB persists or recurs [25].

Hysteroscopy becomes particularly important when a biopsy cannot be obtained, or a previous biopsy contains no endometrium when some could be expected, or with recurrent PMB [25]. The hysteroscopic appearance of endometrial cancer can be obvious, with chaotic endometrium and irregular and abnormal looking blood vessels; sometimes it may be much more subtle and confined to a small area of endometrium (Figure 7.3). Cancer concealed within an endometrial polyp will not be apparent.

Figure 7.3 Endometrial cancer as seen hysteroscopically.

(Source of image: Mary E. Connor.)

The accuracy of the hysteroscopic diagnosis of cancer or endometrial hyperplasia is high, as shown in a large systematic quantitative review [27]. The probability of cancer with a positive hysteroscopy is increased from a pre-test level of 3.9% to 71.8%, with a positive likelihood ratio (LR) of 60.9 (95% CI 51.2–72.5); with a negative hysteroscopy result, the probability is reduced to 0.6% (95% CI 0.5–0.8), with a negative LR of 0.15 (95% CI 0.13–0.18). Hysteroscopy with the positive LR of 60.9 is accurate for revealing endometrial cancer but less accurate for excluding it, as the negative LR of 0.15 is insufficiently low to avoid other investigations.

There is no clear cost-effective strategy for diagnosing the other causes of bleeding in the remaining majority of women [10]. Hysteroscopy is again useful when the endometrial biopsy with a thickened endometrium has returned with no evidence of endometrial cancer. Polyps can be found in 15% of all women presenting with PMB, and in nearly 38% in those without cancer with an ET >5 mm [25]. Offering the patient a single hysteroscopic examination, with the option of immediate endometrial polypectomy, particularly as an outpatient if appropriate, has advantages for patients and service providers [28, 29].

Thickened, cystic endometrium on TVS may be an indication of endometrial hyperplasia, whether simple or atypical, rather than an endometrial polyp or frank malignancy [3]. It may occur in both peri- and post-menopausal women; it is a consequence of intrinsic or extrinsic oestrogen unmodified by progesterone [30]. Women who are at risk include those with a high body mass index (BMI), due to the peripheral conversion within adipose tissue of androgens to oestrogen, those with anovulatory cycles due to the perimenopause or polycystic ovary syndrome, those with oestrogen-secreting tumours, such as a granulosa cell tumour and those taking tamoxifen or oestrogen-containing hormone replacement therapy, though continuous-combined therapy is less of a risk [30]. Subsequent investigation is by endometrial biopsy, although hysteroscopy is indicated, as for PMB, if there is difficulty obtaining a sample [30]. Hysteroscopy is also indicated if abnormal bleeding persists or TVS shows irregular endometrium, as endometrial sampling may miss discrete areas of endometrial abnormality.

Hysteroscopic diagnosis of endometrial hyperplasia may not be obvious as it can be difficult to differentiate it from endometrium stimulated by progestogerone therapy or endometrial cancer, so endometrial sampling with directed biopsies as well as a global sample is indicated [30] (Figure 7.4).

Figure 7.4 Hysteroscopic view of complex endometrial hyperplasia without atypia.

(Source of image: Mary E. Connor.)

The systematic review analysing the accuracy of hysteroscopy for diagnosing endometrial cancer previously discussed also looked at the accuracy for diagnosing endometrial disease, defined as endometrial hyperplasias and endometrial cancer [27]. It concluded that hysteroscopy is only modestly accurate in these circumstances; the pre-test probability of 10.6% was raised to 55.2% (positive LR 10.4) with a positive hysteroscopy result and lowered to 2.8% (negative LR 0.24) with a negative result, emphasising the need for additional investigations such as endometrial sampling.

Most patients will find vaginal examination with an endometrial biopsy and see-and-treat hysteroscopy acceptable as outpatient procedures, but not all; the option of an inpatient procedure should be readily available and patients made aware of this from the outset [3133].

Irregular Bleeding on Hormonal Therapy: HRT, Contraceptives and Tamoxifen

Bleeding unexpectedly when taking any hormonal treatment causes concern and, at the very least, inconvenience. Most concern is expressed when this occurs after a period of amenorrhoea while on no-bleed hormone replacement therapy (HRT); there is worry, too, if bleeding occurs between monthly bleeds on cyclical HRT. The causes are most likely to be related to the hormonal preparation, but more sinister causes need to be excluded. Women with a uterus taking unopposed oestrogen are particularly at risk of endometrial hyperplasia and malignancy [30]. The same pathway can be followed as for the investigation of PMB, with TVS as the first-line investigation, with an endometrial biopsy only indicated for thickened endometrium. The ET threshold depends upon the type of HRT [22, 27] and remains open to debate. The role of hysteroscopy is again limited to exploring the cavity if a biopsy cannot be obtained or if indicated by the suggestion of a uterine cavity abnormality.

Irregular bleeding in pre-menopausal women using hormonal contraceptives is very common, especially with progesterone-only preparations and during the first few months of use. It usually resolves without intervention and often needs no investigation, though infections and pregnancy need to be excluded. A persistent problem that fails to settle may be due to an intrauterine lesion or a vascular abnormality within the endometrium [18]. Hysteroscopic intervention is often considered unnecessary unless indicated by TVS findings, though this approach may change with the new NICE 2018 guidelines [7].

Unscheduled bleeding in women taking tamoxifen therapy, usually as adjuvant therapy for breast cancer, always needs investigating urgently, as the risk of endometrial hyperplasia and cancer is elevated in this population [30]. The risk of endometrial cancer (OR 2.4; 95% CI 1.8−3.0) increases with duration of treatment (≥5 years OR 3.6; 95% CI 2.6−4.8, Ptrend <0.001), lasts for over five years beyond cessation of therapy and does not seem to differ between pre- and post-menopausal women [34]. Benign endometrial polyps are very common in this population (30−60%) and may be the cause of the bleeding [35]. Polyps in tamoxifen-treated women tend to be multiple and larger than in healthy women on HRT, with median sizes of 2.9 cm (range 0.3−11.0 cm) and 1.05 cm (range 0.3−2.0 cm) respectively [36]. TVS with a threshold endometrial thickness of >5mm for hysteroscopy and endometrial biopsy is advised [37], along with endometrial polypectomy if indicated. Malignancy can occasionally occur within a tamoxifen-related polyp even if an endometrial sample is clear [36]. However, a thin endometrium on TVS is reassuring that endometrial cancer is unlikely to be present, though persistent bleeding would still warrant more invasive investigation [25].

7.3.2 Fertility Concerns


A significant proportion (40–50%) of women with infertility will have intrauterine pathology that could contribute to their condition [38]. Despite this high prevalence of lesions, routine use of hysteroscopy for all women being treated for infertility is still debated. The first diagnostic test for these patients should be a TVS, possibly enhanced with saline infusion (SIS) to increase the diagnostic yield [39]. Hysteroscopy is indicated, particularly when the scan suggests an abnormality, to confirm the diagnosis and to offer treatment. It is now established that submucous fibroids should be treated, either to improve the chance of spontaneous conception or before assisted reproductive techniques [40] (Figure 7.5).

Sep 17, 2020 | Posted by in GASTROENTEROLOGY | Comments Off on Chapter 7 – Indications for Hysteroscopy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access