Uterine sarcoma is a mesenchymal tumor that can originate from smooth muscle cells (leiomyosarcoma) or from endometrial stroma (mixed Müllerian tumor or endometrial stroma sarcoma). The leiomyosarcoma (LMS) may appear similar to the leiomyoma (fibroid) and is therefore a confusing element in the diagnosis of fibroids. In general it was hypothesized that LMS is not originating in a fibroid but develops from a primary malignant cell line, but recently four cases of LMS originating in a normal fibroid were reported [32]. The risk of finding a sarcoma in a presumed fibroid ranges from 0.2 to 0.49 %. See Table 8.1 [33–37].

The most important criterion in the diagnosis of LMS is the number of mitoses found in active parts of the smooth muscle tumor. A common cut off level is over 5 mitoses in 10 high power (400×) fields [32, 38]. Other common characteristics of LMS are tissue necrosis, nuclear pleomorphism and vascular invasion. Patients with LMS are more often older and have larger tumours which are mostly solitary compared to patients with fibroids. Rarely the diagnosis is made with endometrial tissue sampling, but only so in case of abnormal uterine bleeding. Contrary to common belief, rapidly growing uterine tumours bear no additional risk of LMS as was reported by Parker in 1994 [35].

Ultrasound greyscale imaging of potential LMS is focussed on the lesion’s margins and the presence of lacunae, which is a sign of necrosis (Fig. 8.7a). With these criteria up to 95 % of cases was reported to be detected [39]. The 2D power Doppler criteria pulsatility and resistance index and the peak systolic velocity, though changed in sarcomata, seem to have poor prognostic value [40]. However the finding of increased periferal and central vascularisation resulted in high sensitivity (100 %) at the cost of specificity (88 %), resulting in a high negative predictive value (100 %) and a predictive value of a positive test of 18 % [41] (Fig. 8.7b). The 3D PD has not yet been studied in the distinction between LMS and fibroids, but in a recent report the vascular indices correlated with cellularity in the fibroid, which may be associated with LMS and STUMP [43].

The value of MRI in the diagnosis of LMS is well established although absolute certainty cannot be achieved. As the incidence of LMS is low, most clinical studies of the diagnostic accuracy of MRI are underpowered. In general LMS shows more high signal areas on the T2 (Fig. 8.7c) than fibroids and a characteristic of LMS is the well demarcated, unenhanced pockets in dynamic MRI with the use of contrast. This feature was seen in 9/12 patients with LMS or STUMP and in 0/12 patients with fibroids [43].

The Positrone Emission Tomography has a place in the diagnostic armentarium of presumed fibroids. In PET scanning a radionuclide (tracer) on a biologically active molecule is visualized. In imaging of fibroids usually fluodeoxyglucose (FDG) is used, but also other molecules, such as deoxyfluorothymidine (FLT) or alphafluorobetaoestradiol (FES) have been reported. In general the uptake of FDG in a fibroid is associated with the estrogen status, cellularity and the presence of malignancy. The result of the PET-scan is displayed in a three dimensional image, often combined with a CT overlay and expressed in Standard Uptake Value (SUV) (Fig. 8.8). In one study the mean SUV in fibroids was higher (1.39 ± 0.65) than in myometrium (1.24 ± 0.33) [44]. In another study [45] of 5 sarcomas the mean SUV was 4.5 ± 1.3. Generally a SUV of 2.0–3.0 is considered the cut off level for presumed fibroid becoming suspicious for sarcoma. FES may be more accurate in distinguishing LMS from fibroids than FDG, with an accuracy of resp. 93 and 81 % [46]. By calculating the rate between two different uptake molecules, e.g. FDG and FES, the test characteristics of PET can improve as well [47]. Only the retrospective study of Umesaki [45] compares different imaging techniques in case of suspected uterine sarcoma. Of the 5 sarcomata all were detected by FDG PET, 4 by dynamic MRI and 2 by Power Doppler ultrasound. Nevertheless, in case of clinical suspicion of uterine sarcoma the less invasive imaging modalities are first choice: grey scale ultrasound, dynamic MRI and PET. The role of image guided biopsies is not completely clear. Although the predictive value of a negative biopsy might be expected to be low because of the large area’s of necrosis, an excellent negative predictive value is reported using MRI guided needle biopsies and Bell’s classification on histology [48] with a cut off level of 2 [49]. No data are found on the possible spreading of sarcoma cells by multiple puncturing of the sarcoma.

Adenomyosis consists of endometrial glands in the myometrium that do not communicate with the uterine cavity. The endomyometrium is more hypoechoic than the periferal myometrium and is called the junctional zone. Diffuse adenomyosis is characterized by a thickened junctional zone (cut off level in general >12 mm) around the uterine cavity (Fig. 8.9). The uterine adenomyoma is a mass of focal adenomyosis not necessarily making contact with the junctional zone. The adenomyoma in particular may sometimes difficult to distinguish from the uterine fibroid as it has a similar appearance (Fig. 8.10). It may cause similar symptoms as heavy menstrual bleeding and features such as enlargement of the uterus and deformation of the uterine cavity. Adenomyosis is a common cause of failure of the endometrial ablation and of myomectomy if mistaken for a fibroid. It has been reported that 28 % of hysterectomy specimens in premenopausal patient contain adenomyosis [50, 51]. Patients with symptomatic adenomyosis have a high risk of hysterectomy and ideally the diagnosis should be made before any minimal invasive therapeutic attempt is undertaken.

No imaging technique can diagnose adenomyosis with absolute certainty. Ultrasound, including elastography and MRI however may demonstrate the different features of the adenomyoma and the fibroid.

The typical features in imaging of the uterine fibroid comprise a homogenous structure with circumscript smooth borders and often a round shape and a distinct vascular capsule, which is visible in power Doppler as well as in elastoscan as a soft rim. The vascularity within the fibroid – not in the capsule – is generally lower than in the myometrium. In contrast, adenomyosis in ultrasound imaging is characterized by ill-defined margins, by heterogeneity with typical hypoechoic striation caused by echogenic foci next to hypoechogenic elements formed by pockets of blood (‘lacunae’) [52, 53]. The shape is often not round but ellipsoid and there is little mass effect on the uterine cavity if any. The lesion is conjunct to the junctional zone that may be variably thickened if diffuse adenomyosis is present. In the coronal plane with 3D ultrasound a higher sensitivity for adenomyosis is reported (cut off level 8 mm) than in the sagittal and transversal plane in 2D ultrasound [54]. The myometrium is asymmetrically thickened [53] or the uterine configuration may be globular. With power Doppler (PD) ultrasound random like vessels are seen in the adenomyotic lesion as in the myometrium but in higher numbers. The use of intracavitary contrast with saline or gel may give more accurate information on the protrusion of the myometrial lesion and support the distinction of adenomyosis and submucous fibroids as the latter tend to protrude more into the cavity [55]. In elastography the amount of pressure applied by the probe affects the brightness and colouring of the acquired images. A steady state image should be obtained to interpret an elastography image. A steady state image of the ute‑rus is established when a clear delineation of the serosa is visible. Adenomyosis is in general softer than myometrium, visualized as several bright colours within the dark myometrium (i.e. bright yellow, green or light blue) (Fig. 8.11). While fibroids are clearly delineated, the shape of adenomyosis is irregular. The feasibility of elastography for the discrimination of intrauterine fibroids and adenomyosis, has been reported in literature [29, 56]. There are certain artefacts the examiner has to be aware of. A probe contact artefact is visible in most elastographic images and shows an illuminated area at the point of contact with the ultrasound transducer mimicking softer areas. This bright (yellow) area should not be confused with the presence of adenomyosis. Due to this artefact it is difficult to study the existence of adenomyosis in the low anterior part of the uterus.

In case of excessive pressure or movements of the uterus bright colours will appear in the image defined as movement artefacts. Various bright colours on different locations can be induced even in normal myometrium resembling the appearance of adenomyosis. Both the arcuate plexus and the AV malformations are visible with power and colour Doppler and should not be confused with adenomyosis [28, 56, 57].

In those studies that compare transvaginal ultrasound and MRI in the diagnosis of adenomyosis no significant difference in diagnostic accuracy could be demonstrated [58, 59].The sensitivity and specificity are ranging from 70 to 89 % resp. 65 to 98 % in both techniques. The area under the curve (AUC) for the diagnostic performance of MRI in the detection of myomas and adenomyosis is reported 0.81 and 0.73, respectively [57].