Sonographic classification and reporting system for diagnosing adenomyosis

In 2015, the international Morphological Uterus Sonographic Assessment (MUSA) group published a consensus on which terminology to use when describing myometrial lesions seen on ultrasonography1. The use of MUSA terminology to describe ultrasound images of fibroids, including their location according to the International Federation of Gynecology and Obstetrics (FIGO), is relatively straightforward to implement2, 3. However, even though the MUSA consensus statement suggests which terms should be used to describe ultrasound images of adenomyosis, it does not provide guidelines on how to classify morphological types or the extent of adenomyosis1. We propose a uniform reporting system for ultrasound findings of adenomyosis. The opinion presented herein is based on a thorough discussion among all authors, including a Delphi procedure (Appendix S1). Images and videos of cases typical of the different morphological variations of adenomyosis were used in the debates. The gold standard for the diagnosis of adenomyosis is histological examination of a hysterectomy specimen. Because only a small, selected group of women undergo hysterectomy, an accurate estimation of the prevalence of the disease cannot be established4. The introduction of imaging techniques such as transvaginal sonography and magnetic resonance imaging (MRI) has allowed non-invasive diagnosis of adenomyosis5-8. Ultrasound imaging is widely available in an office setting, is relatively inexpensive, requires no preparation, has no contraindications and is relatively accurate in expert hands, making it the imaging modality of choice in gynecology7. Although adenomyosis is usually diagnosed in women between 40 and 60 years of age, it is also described in younger women, in whom any surgery performed on the uterus might adversely affect child-bearing9. The treatment of choice for adenomyosis is primarily hormonal (e.g. levonorgestrel intrauterine device; oral progestins)10. Patient management is often based on ultrasound diagnosis alone. This highlights the importance of a uniform, reproducible and clinically relevant reporting system for ultrasound findings of adenomyosis. Uniform reporting also facilitates studies on the prevalence, etiology and clinical implications of adenomyosis and on the effectiveness of therapies. There are various theories about the pathogenesis of adenomyosis. It is commonly thought to originate from direct contact between the endometrium and the underlying myometrium, which allows the formation of ectopic endometrial glands and stroma. However, the precise pathophysiological pathway is not known. Reported risk factors for adenomyosis include multiparity11, 12 and previous uterine surgery13-16 (curettage and Cesarean delivery), suggesting a possible role of damage to the endometrial–myometrial junction. Such damage allowing the growth of ectopic endometrial glands and stroma into the myometrium may explain ultrasound findings of subendometrial lines and buds with expansion to hyperechogenic islands in the myometrium (Figure 1)1. On the other hand, Kishi et al.17 reported infiltration of endometriosis from outside the uterus, with disruption of the serosa and infiltration of the external myometrium inducing another subtype of adenomyosis. Extrauterine adenomyosis has also been reported, for example in the rectovaginal septum18. Other theories involve infolding of endometrium which then penetrates into the myometrium, basement membrane damage through single nucleotide polymorphisms, and initiation/progression of adenomyosis modulated by vascular factors and estrogen receptors18-20. The histologic diagnosis of adenomyosis is made by observing the presence of endometrial stroma and glands in the myometrium. However, there is no consensus among pathologists, and various histopathological definitions have been reported, including: (1) disruption of the normal boundary between the endometrium and myometrium21, 22, (2) presence of ectopic endometrium that is basal-type non-secretory tissue with a direct connection to the basalis layer5, (3) myometrial invasion by endometrium > 4 mm below the basalis layer23, (4) myometrial invasion by endometrium > 2.5 mm below the basalis layer5, (5) endometrial invasion to > 25% of the thickness of the uterine musculature, as measured from the endometrial–myometrial junction18. On histology, adenomyosis is classified as focal if there are circumscribed nodular aggregates of endometrial glands and stroma surrounded by normal myometrium, and diffuse if there are endometrial glands and stroma distributed diffusely throughout the myometrium24. Adenomyomas are a subgroup of focal adenomyosis surrounded by hypertrophic myometrium25. Different histological disease severity classifications have been suggested, but without international consensus17, 26-30. The ultrasound characteristics of adenomyosis reflect the histological features. Different morphological types, seen on ultrasound examination or histological examination, may reflect different stages in the development of the disease, and may have different clinical significance with respect to symptomatology, fertility, obstetric outcome and therapeutic options. Clearly, there is a need for an internationally accepted, uniform classification of adenomyosis that, preferably, can be achieved using ultrasound. Seven items should be assessed when examining and describing a uterus with adenomyosis. First, the myometrium should be classified as normal or as abnormal, and, in the latter case, according to whether it manifests signs of adenomyosis, myoma or sarcoma, using MUSA terminology1. Other myometrial lesions to be considered in the differential diagnosis are accessory cavitated uterine masses, also reported as juvenile cystic adenomyoma, and postoperative uterine scarring, including focal loss of myometrium and fistulae31. MUSA features typical of a uterus with adenomyosis include an enlarged globular uterus, asymmetrical thickening of the myometrium, myometrial cysts, echogenic subendometrial lines and buds, hyperechogenic islands, fan-shaped shadowing, an irregular or interrupted junctional zone and translesional vascularity on color Doppler ultrasound examination (Figure 2)1. The definitions of these features can be found in the MUSA Consensus Opinion1. The location of the adenomyosis should be described as anterior, posterior, lateral left, lateral right or fundal. To determine the exact location, the uterus should be examined in both sagittal and transverse planes. The additional value of three-dimensional ultrasound to examine the coronal plane of the uterus needs to be established in future studies. In each location, it should be determined whether the adenomyosis is focal or diffuse, by estimating the relative proportions of the lesion and the surrounding normal myometrium on a sagittal section through the uterus where the adenomyotic lesion appears to be at its largest (Figure 3). We propose that an adenomyotic lesion should be defined as focal if > 25% of the circumference of the lesion is surrounded by normal myometrium. Adenomyosis is classified as diffuse if 50% affected). If there are adenomyotic lesions in different locations, the sum of volumes of the different lesions should be estimated when describing the extent of the disease. The estimated extent of the disease might not be associated with the type or severity of symptoms, but may be useful for research purposes. The largest diameter of the adenomyosis lesion(s) should be measured. In clinical situations, this should be done in the plane of the largest diameter of the largest lesion. In the research setting, we advise measurement of the largest diameter of each focal lesion. In the case of a diffuse lesion, the myometrial wall thickness should be measured and the site involved noted. Future studies are needed to estimate the additional value of assessing lesion size in all three orthogonal planes. We propose a consensus-based practical classification of adenomyosis, on the basis of ultrasound findings (Figure 6), that consists of: (1) identification of the presence of adenomyosis, using the MUSA criteria1; (2) determination of location of the adenomyosis; (3) differentiation between focal and diffuse disease; (4) discrimination between cystic and non-cystic lesion; (5) determination of myometrial layer involvement; (6) classification of disease extent as mild, moderate or severe; (7) measurement of size of lesion. Cycle day and current hormonal use should always be recorded. Several examples of classification are given in Table 1. Sliding viscera: yes JZ (inner myometrium) and middle myometrium (Type 1–2) 2 cm Focal Type-1–2 cystic adenomyosis in anterior wall Extent: mild Max diam: 2 cm Sliding viscera: no JZ (inner myometrium), middle myometrium and outer myometrium (subserosa)(Type 1–3) Diffuse Type-1–3 cystic adenomyosis in posterior wall Extent: moderate Max diam: 3 cm Diffuse Type-1–2 adenomyosis in anterior wall Extent: severe Max diam: 7 cm Mixed-type cystic adenomyosis: Type-1–2 adenomyoma in anterior wall and Type-3 focal adenomyosis in left lateral side Extent: mild Max diam: 1. Adenomyoma: 3 cm 2. Focal lesion: 2 cm Mixed-type cystic adenomyosis: Type-2 focal adenomyosis in left lateral wall and diffuse Type-1–3 adenomyosis in anterior wall Extent: Severe Max diam: 1. Focal lesion: 4 cm 2. Diffuse lesion: 7 cm Although this consensus was developed using MUSA terms and definitions, we have proposed some changes. For example, using the original MUSA terminology, an ‘ill-defined lesion’ may be localized or diffuse, the latter being a lesion involving at least 50% of the total uterine volume1. We suggest reporting adenomyosis as focal or diffuse in each location. For example, it should be possible to report focal adenomyosis in the anterior wall and diffuse adenomyosis in the posterior wall, using the definition of focal disease outlined above. A second slight adjustment is the specification of the myometrial layer involved. This classification should facilitate consistent reporting of adenomyosis, but it needs to be validated and refined in future studies. For example, the feasibility and reproducibility of the differentiation between the middle and outer myometrial layers using power Doppler needs to be assessed. Likewise, the additional value of the use of transverse and/or coronal planes (the latter using three-dimensional ultrasound) for the assessment of location, extent and size of adenomyotic lesions needs to be investigated further. Research is also needed to assess the accuracy of ultrasound examination in the diagnosis of focal vs diffuse adenomyosis. In a prospective case series assessing uterine lesions in hysterectomy specimens, some adenomyosis lesions appeared to be much more extensive than was suspected at ultrasonography or on macroscopic examination36. The possibility of diffuse disease not being detected at ultrasound examination should be borne in mind when planning management of patients with adenomyosis and when planning studies. Several recent studies have shown that the number of morphological features of adenomyosis is associated with clinical symptoms and success of fertility treatment37-39. However, more research is needed to evaluate the importance of different ultrasound features. In our clinical experience, some women with small lesions may present with severe symptoms of pain and uterine bleeding, whereas others with larger lesions may be asymptomatic. Clearly, the classification we have suggested cannot be used on its own to decide on treatment. The proposed classification might need to be amended after external validation and based on the results of future studies evaluating the relationships between ultrasound features, clinical symptoms, histological findings and possibly also MRI findings. There are still uncertainties regarding the clinical importance of myometrial cysts, the relevance of discriminating between the uterine layers and the reliability of estimating disease extent. A study addressing the intra- and interobserver variability using the proposed reporting is also needed. Our suggested approach should be considered only as a first step towards an internationally accepted classification and reporting system. Finally, we recognize that some aspects of the suggested reporting and classification system might require extensive ultrasound skills. After validation and optimization of this proposed classification, it would be reasonable to develop an e-learning program for less experienced ultrasonographers. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.