Ultrasound Findings After Uterine Artery Embolisation: Pictorial Essay and Clinical Validation

Matthew Lukies: conceptualization, investigation, writing – original draft, writing – review and editing, formal analysis, methodology, validation, visualization, project administration, data curation, supervision. Warren Clements: conceptualization, writing – review and editing. James M. Chan: investigation, methodology. Olivia Darby: investigation, writing – original draft, methodology. Jessica Allford: writing – review and editing. Georgina Venn: writing – review and editing, conceptualization. Suhail Wani: investigation, methodology.

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Approval was obtained by the Alfred Human Research and Ethics Committee prior to performing this study, including waiver of individual patient consent. This article does not contain any studies with animals performed by any of the authors.

This study was not supported by any funding.

To validate the typical ultrasound findings, a retrospective review of 107 primary uterine artery embolisation cases over a 3‐year period was performed after ethical review board approval. Patients were excluded if they underwent UAE for an indication other than symptomatic uterine fibroids and/or adenomyosis, such as malignant uterine bleeding or were enrolled in another trial ( n  = 4). Clinical success was defined as the patient no longer requiring or seeking further treatment on clinical follow‐up at 6 or 12 months post UAE, namely, clinically significant improvement in heavy menstrual bleeding or mass‐effect symptoms. Ultrasound parameters examined at a median of 6 months (IQR 5–11 months) post UAE were recorded and correlated with clinical success (Chi‐Square test or two‐sample t ‐test). In the cohort ( n  = 107; Table  1 ) treated with UAE for symptomatic fibroids ( n  = 95) and/or adenomyoma ( n  = 19), 88% had a clinically successful outcome, 86% of fibroids and adenomyomas demonstrated central avascularity, 64% were hypoechoic, mean reduction in fibroid or adenomyoma volume was 53%, and the mean overall uterine volume reduction was 41%. Endometrial thickness pre‐ and post‐UAE was similar (median of 7 mm and 6 mm, p  = 0.46), adjusted for menstrual cycle dates if recorded at the time of scanning. Based on the 93 cases with stored vascularity images, there was a significant correlation between persistent vascularity in a fibroid or adenomyoma and seeking further treatment ( p  < 0.01). Significant correlation was also demonstrated between post‐treatment volume of a dominant fibroid or adenoma volume ( p  < 0.01) and seeking or requiring further treatment, with a similar finding for overall uterine volume ( p  = 0.02). No significant correlation was demonstrated between hypoechoic appearance of a fibroid or adenomyoma and seeking or requiring further treatment ( p  = 0.16). Patient age, pathology, symptomatology, imaging features pre and post uterine artery embolisation, and clinical outcome. Includes patients with both pathologies/symptomatologies.

Given particle embolisation is a distal technique, and utilises the hyperdynamic flow of abnormal tissue, it is expected that hypervascular fibroid and adenomyotic tissue will diminish in vascularity on imaging after treatment. The endometrium is spared as radial arteries are typically < 200 μm in calibre and do not allow passage of appropriately‐sized embolic material. Through careful prevention of over‐embolisation, the remaining myometrial vascularity is preserved [ 9 ]. This is supported by preservation of the main uterine artery flow and the rich collateral arterial uterine supply, which may dilate after treatment (e.g., via ovarian and round ligament arteries). Fibroids and adenomyomas typically do not effectively re‐establish sufficient arterial flow and undergo infarction [ 7 ]. However, fibroids may promote neoangiogenesis through the release of angiogenic cytokines such as TNF‐alpha, particularly if they are not sufficiently treated or treated with an embolic agent that does not penetrate sufficiently distally. As such, it is expected that well‐treated uterine fibroids and adenomyomas should have no internal vascularity on ultrasound Doppler assessment after UAE [ 8 , 9 , 10 , 11 ] (Figures  1 , 2 , 3 , 4 ). Whilst some peripheral or marginal vascularity around the fibroid may persist [ 8 ], this is typically due to normal myometrium that surrounds the fibroid and not the fibroid itself. If a dominant fibroid demonstrates persistent internal or centripetal vascularity, it may have been undertreated or regenerated a new arterial supply, and thus be a potential target for future treatment if there are ongoing symptoms. Alternatively, a new fibroid may have developed [ 12 ]. Ultrasound images pre (a–c) and 6 months post (d–g) uterine artery embolisation for fibroids with heavy menstrual bleeding. Symptoms resolved post‐treatment. (a, b) Transabdominal transverse and long axis ultrasound views of the uterus showing a dominant anterior bilobed intra/submucosal heterogeneously isoechoic and hypoechoic uterine fibroid, measuring 163 cc in volume. (c) Colour Doppler assessment of the fibroid showing central vascularity. (d, e) Transabdominal transverse and long axis ultrasound views of the uterus showing the post‐treatment dominant fibroid decreased in size (51 cc) with heterogeneous hypoechoic and hyperechoic areas. (f, g) Colour Doppler assessment of the treated fibroid showing no central vascularity. As fibroids undergo infarction post UAE, there is cellular breakdown analogous to hyaline degeneration and formation of small cystic spaces [ 9 , 13 ]. This gradual decrease in cellularity causes the treated fibroid to commonly, but not universally, appear more hypoechoic on ultrasound than baseline and relative to the surrounding myometrium [ 13 ] (Figures  2 and 3 ). There may also be formation of a hyperechoic or calcific rim at the fibroid margin as a final stage of hyaline degeneration and/or due to aggregation of particle embolic material in the fibroid periphery [ 13 , 14 ]. In some instances, calcification of fibroids may have been present prior to treatment from a previously degenerated lesion. This may make echotexture and vascularity difficult to determine. Given UAE treats all fibroids, including seedlings, which are often poorly visible when small, there may be new regions of discrete low echogenicity through the myometrium after treatment, and this is expected. Ultrasound images pre (a–c) and 6 months post (d–f) uterine artery embolisation for fibroids with mass effect symptoms. Symptoms resolved post‐treatment. (a, b) Transabdominal transverse and long axis ultrasound views of the uterus showing a dominant fundal intramural heterogeneously isoechoic uterine fibroid, measuring 664 cc in volume. (c) Colour Doppler assessment of the fibroid showing central vascularity. (d, e) Transabdominal transverse and long axis ultrasound views of the uterus showing the post‐treatment dominant fibroid decreased in size (223 cc) with a more hypoechoic appearance compared to pre‐treatment. (f) Colour Doppler assessment of the treated fibroid showing no central vascularity. Ultrasound images pre (a, b) and 6 months post (c–e) uterine artery embolisation for fibroids with heavy menstrual bleeding. Symptoms resolved post‐treatment. (a, b) Transvaginal long and transverse axis ultrasound views of the uterus showing a dominant posterior intramural/submucosal predominantly isoechoic uterine fibroid, measuring 98 cc in volume. (c) Transvaginal long‐axis ultrasound views of the uterus showing the post‐treatment dominant fibroid decreased in size (44 cc) with a more hypoechoic appearance compared to pre‐treatment, including cystic areas. (d, e) Colour Doppler assessment of the treated fibroid showing no central vascularity. Ultrasound images pre (a–c) and 6 months post (d–g) uterine artery embolisation for fibroids with heavy menstrual bleeding. Heavy menstrual bleeding persisted post, and the patient was seeking further treatment. (a, b) Transvaginal transverse and long axis ultrasound views of the uterus showing a dominant submucosal heterogeneously isoechoic uterine fibroid, measuring 3.8 cc in volume. (c) Colour Doppler assessment of the fibroid showing central vascularity. (d, e) Transabdominal transverse and long axis ultrasound views of the uterus showing the post‐treatment fibroid modestly decreased in size (2.9 cc). (f) Colour Doppler assessment of the treated fibroid showing persistent central vascularity. Ultrasound images pre (a–c) and 6 months post (d–f) uterine artery embolisation for adenomyosis with heavy menstrual bleeding. Symptoms resolved post‐treatment. (a, b) Transvaginal long and transverse axis ultrasound views of the uterus showing posterior myometrial thickening, “Venetian blind” striated shadowing, and an ill‐defined hypoechoic adenomyoma. (c–e) Transabdominal transverse and long axis ultrasound views of the uterus showing resolution of previous posterior myometrial thickening, resolution of the “Venetian blind” striated pattern, and a small peripherally hyperechoic/calcified avascular lesion at the site of previous adenomyoma. Following the UAE, a dominant non‐calcified fibroid can be expected to decrease in volume by approximately 40%–70% over 6–12 months (Figures  1 , 2 , 3 ), with ongoing volume reduction seen up to 2 years post‐operatively [ 8 , 9 , 10 ]. This can be measured in three dimensions on transabdominal or transvaginal ultrasound examination using a prolate ellipse formula. If a dominant fibroid has not decreased in volume appropriately and retains central vascularity, it may be a potential target for future treatment if there are ongoing symptoms. On occasion, more than one adjacent fibroid may have appeared as a single lesion or cluster on pre‐UAE ultrasound imaging, but on follow‐up, appear as separate, discrete fibroids, leading to overestimation of volume reduction. Care should be taken to ensure that the same dominant fibroid in the same location is assessed like‐for‐like on follow‐up imaging to avoid miscalculation. Given that the volume of the uterine fibroids in a fibroid uterus disproportionally contributes to the overall uterine volume, there is typically a concordant decrease in overall uterine volume following the uterine artery embolisation of 25%–60% [ 8 , 9 , 13 ]. The majority of this overall volume decrease is likely attributable to fibroid infarction rather than true myometrial atrophy. Uterine volume is an important measurement after UAE, as maximal diameter often underestimates volume reduction, and it is typically more reproducible than accurately measuring fibroid mean diameter in 3‐planes. The majority of studies of uterine artery embolisation for adenomyosis have presented post‐operative imaging findings on MRI [ 4 , 7 , 15 ], describing a decrease in junctional zone thickness and overall uterine volume (Figure  5 ). Following first principles, it would also be expected that a completely embolised adenomyoma would also be avascular on Doppler ultrasound assessment, gradually decrease in volume and develop hypoechoic or hyperechoic/calcific foci due to infarction and hyaline degeneration. Infarction of the secretory glandular tissue at the myometrial‐endometrial junctional zone may also cause loss of the classical “Venetian blind” striated shadowing appearance, as an indirect feature according to the Morphological Uterus Sonographic Assessment (MUSA) consensus [ 7 , 15 , 16 ].

The mainstay of follow‐up evaluation after UAE for symptomatic fibroids and adenomyosis is clinical assessment of symptoms by the treating Interventional Radiologist. The role of imaging is complementary to clinical assessment, and ultrasound plays an important role in many jurisdictions despite the granular advantages of MRI. Typical ultrasound features at follow‐up include decreased volume, centrally avascular appearance, and hypoechoic consistency of treated dominant fibroids and adenomyomas, which may aid in clinical decision‐making, particularly if clinical response is suboptimal and further treatment such as repeat embolisation is being considered.

Uterine artery embolisation (UAE) is a level‐1 evidence‐based interventional radiology treatment option for symptomatic uterine fibroids in all locations [ 1 , 2 , 3 ], and has a growing level‐2 evidence base as a treatment for symptomatic adenomyosis [ 4 , 5 ]. As an increasing number of women access and undergo UAE treatment, understanding of the typical post‐operative imaging findings is increasingly important for all practitioners involved in medical imaging, as well as knowing when findings may be informative for the treating Interventional Radiologist (IR). The assessment of clinical outcome following UAE is centred on clinical assessment of symptomatic improvement in the outpatient clinic setting [ 6 ]. Imaging findings play a complementary role to clinical assessment, particularly when there has been less‐than‐expected improvement in symptoms such as heavy menstrual bleeding or mass‐effect symptoms, and further treatments, including repeat UAE, might be considered. Whilst Magnetic Resonance Imaging (MRI) is performed in many centres for imaging follow‐up of patients undergoing UAE and has been the subject of numerous studies over recent decades [ 7 ], ultrasound is widely available and significantly less costly than MRI. In addition, MRI is not funded in many jurisdictions, leaving patients or insurers to bear the costs, which are often prohibitive. For these reasons, many practices, particularly in regions with government single‐payer healthcare systems, utilise ultrasound as the main modality of imaging pre and post UAE [ 8 ]. Along with the trend towards MRI imaging, there have been a few recent publications outlining ultrasound findings following UAE. This article outlines the typical ultrasound imaging features post‐uterine artery embolisation, with clinical validation via a retrospective cohort of real‐world treated patients.

The authors declare that they have no conflicts of interest. Warren Clements is an editorial board member of JMIRO and co‐author of this article. To exclude bias, he was excluded from decision‐making regarding the acceptance of this article for publication.