Uterine Artery Embolization for Heavy Menstrual Bleeding

Abstract

Uterine artery embolization (UAE) as a treatment option for fibroids was first reported by Ravina in 1995. Although rapidly adopted by enthusiasts, many were skeptical and its introduction varied widely across the globe. It was not until randomized controlled trials and registries were published and national guidance statements issued that UAE was accepted as a safe and proven treatment for fibroids. The technique is now established as one of the treatment options to be discussed with patients as an alternative to surgery for fibroid-associated heavy menstrual bleeding. Research is on-going to evaluate the relative merits of UAE compared with other medical and surgical treatment options for heavy menstrual bleeding, particularly for women wishing to maintain their fertility.

Indications & contraindications

The indications for uterine artery embolization (UAE) are similar to those for surgical treatment, namely significant fibroids associated bulk or menstrual symptoms uncontrolled by medical or less invasive surgical treatments. Adenomyosis can also be treated by UAE and this indication has been recently reviewed by NICE and given ‘normal arrangements’. The main reasons for patients choosing UAE include a preference for a minimally invasive, nonsurgical treatment option with reduced morbidity and faster recovery compared with surgery and a wish to retain the uterus. In women with significant risk factors for surgery such as obesity and intra-abdominal adhesions, gynecologists may be reluctant to undertake myomectomy or hysterectomy, because of lack of feasibility or concerns over the potential for serious operative complications.

There are very few absolute contraindications to UAE but several where there is controversy. These include:

• Suspected pregnancy;

• Recent or on-going pelvic sepsis;

• Fibroid noted to be already necrotic on imaging;

• Severe allergy to radiographic contrast media.

Although any fibroid can be embolized it may not always be the best option, for example, a submucosal fibroid is probably more easily dealt with by hysteroscopic resection. Although size (>10 cm) and pedunculated subserosal lesions were originally thought to be a contraindication there are several published series showing that the complication rate is no higher in these groups. For women wishing to maintain their fertility there is much debate and many polarized views. The evidence to make a clear choice is currently lacking but a large trial (FEMME) is underway in the UK [1]. Until this randomized controlled trial (RCT) reports, UAE should probably be reserved in this group for those unable or unwilling to have a myomectomy.

Pretreatment workup

Establishing the correct diagnosis is paramount for UAE because in contrast to open or laparoscopic surgery, the abdominal cavity is not inspected nor tissue extracted which can be pathologically analyzed. Pelvic ultrasound and particularly transvaginal ultrasound can identify fibroids and other pelvic pathologies with high accuracy and is a safe, acceptable and convenient outpatient test. However, while MRI is less easily accessible, the technology has many advantages over ultrasound such as the ability to image in any plane, identify the uterine vessels, assess vascularity and minimize operator-dependent performance. Other pathologies such as adenomyosis are seen more easily although the complete exclusion of the rare leiomyosarcoma remains a challenge unless advanced at the time of imaging.

Magnetic resonance imaging protocols vary but a minimum schedule should include T1/T2 imaging in the transverse and coronal oblique planes. Gadolinium contrast should be routinely administered to determine vascularity as on occasion a fibroid is found to be already dead and clearly UAE would have no role in that situation. Many units now add a magnetic resonance angiogram sequence which can help determine the origin of the uterine vessels (Figure 1). While MRI is an excellent preoperative imaging modality which helps identify and characterize uterine fibroids as well as exclude other relevant pathology such as adenomyosis and necrotic fibroids, the information obtained is of less value in predicting clinical outcomes.

Figure 1.

Magnetic resonance angiogram acquired during the pre-uterine artery embolization MRI scan. Both uterine arteries are hypertrophied and clearly seen (arrows).

Severe anemia should be corrected although the blood loss from UAE is usually minimal. If the patient is taking a gonadotrophin releasing hormone agonist, there is some debate as to whether it should be stopped for several weeks as it may make the procedure more difficult due to a reduction in uterine artery caliber. Although the evidence to support this is nothing more than anecdotal it seems prudent to stop the drug if at all possible. The newer selective progesterone receptor modulators (e.g., ulipristal acetate) may be a better option to control bleeding prior to treatment but there is no evidence that they improve the results of UAE or indeed myomectomy or hysterectomy [2]. Ideally, an intrauterine device should be removed before the procedure as it could act as a source of infection.

Technique

The procedure is carried out under conscious sedation in an angiographic theater. For an experienced interventional radiologist (IR), UAE is a relatively simple, straightforward procedure usually completed in less than 1 h. The uterine arteries are accessed through the femoral artery using a super selective technique and a coaxial 2–3 French catheter placed in each uterine artery in turn. Care must be taken to ensure a good safe stable position before embolization to avoid targeting nonuterine branches of the internal iliac artery, for example, vesical artery. The use of a 2–3 French coaxial catheter is strongly recommended to avoid spasm which then severely compromises the embolization procedure. The aim is to allow pulsatile arterial flow to carry the embolic particles to the microcirculation of the fibroids. Particles are injected into each uterine artery until there is complete or near complete arterial stasis (Figure 2). It is essential to embolize both uterine arteries as cross flow collateralization will prevent infarction if one is left patent. The puncture site is usually controlled by simple manual pressure. Like many technical procedures the majority are straightforward but occasionally it can be very challenging to select the uterine artery due to uterine distortion and differently shaped catheters may be required. An up-to-date angiographic machine with the latest imaging software allowing the vascular anatomy to be overlaid on the live fluoroscopic image (road mapping) is essential to minimize the radiation burden to the patient.

Figure 2.

Selective internal iliac angiogram. (A) Selective angiogram of the left internal iliac artery and (B) A co-axial catheter placed deeply into the left uterine artery immediately prior to embolization. The tortuosity and then medial course of the uterine artery is very typical.

There are several embolic products that can be used for this procedure and it is not clear which is best. The exact nature of the chosen embolic material probably makes little difference to the outcome. Many IRs use the well-established polyvinyl alcohol (PVA) particles, whereas others prefer the newer agents such as Embospheres (Merit Medical, UT, USA) and Embozene (Celonova, San Antonio, TX, USA). Some even advocate the use of Gelfoam (Ethicon, NJ, USA) which is regarded as a temporary embolic agent frequently used for postpartum hemorrhage control. Although the procedure itself is not painful the resulting ischemia to the uterus and fibroids certainly is. There should be a written pain protocol in place and normally this would include premedication, periprocedural and postoperative opioid analgesia. A single overnight stay is a normal practice in most centers with the patient discharged the following day on oral analgesia. There is consensus that the procedure should be covered with perioperative antibiotics.

Follow-up

After hospital discharge, normal practice is to review the patient at 4–6 weeks and again at 6 months with some form of imaging at the 6-month visit. The author's preference is MRI and the expectation is to find completely infarcted fibroids with a typical uterine volume reduction of 50–70% (Figure 3). Imaging studies carried out up to 5 years show a small further volume reduction beyond the first 6 months [3].

Figure 3.

Sagittal magnetic resonance gadolinium enhanced image pre and post embolisation. (A) Contrast-enhanced T1-weighed sagittal pelvic MRI scans prior to uterine artery embolization showing a large intramural fibroid with avid contrast enhancement. (B) Repeat MRI scan at 6 months post uterine artery embolization showing marked volume reduction and no contrast enhancement indicating fibroid infarction.

The clinical response to UAE, namely, reduction in HMB and fibroid size, is not immediate and women seldom experience any noticeable improvement in the first month. However, by 6 months there should be a good clinical response and this may continue beyond this time frame because of further fibroid shrinkage. Women should on average expect to return to work and resume normal activities 2 weeks after UAE.

Results including trial evidence

UAE has been reasonably well investigated and tested using the rigors of RCT methodology. In this respect, UAE has been better evaluated than either myomectomy or hysterectomy. There have been seven RCTs comparing UAE with either hysterectomy (5) or myomectomy (2), two systematic reviews and a Cochrane review (2006 and 2011) [4–13]. A national, multicenter RCT comparing the effect of UAE compared with myomectomy on health-related quality of life (FEMME) is underway in the UK. Table 1 summarizes the five major studies published to date.

Table 1.

Summary of five of the uterine artery embolization randomized trials.

Primary outcome measure	PINTO (n = 60)		MARA (n = 121)		EMMY (n = 177)		REST (n = 157)		FUME (n = 163)
	HMB		Pregnancy		HMB		QoL		QoL
Procedure	UAE	Surgery (H)	UAE	Surgery (M)	UAE	Surgery (H)	UAE	Surgery (H/M)	UAE	Surgery (M)
Randomized allocation	n/a†	n/a†	58	63	88	89	106	51	82	81
Technical success (%)	90	100	90	92	83	100	97	98	97	98
Minor complications (%)	72‡	45‡	21‡	16‡	64§	56§	47#	28#	13¶	11¶
Major complications/Adverse events (%)					5§	3§	19#	25#	3¶	8¶
QoL	n/a	n/a	n/a	n/a	NS#	NS#	NS#	NS#	NS¶	NS¶
Re-intervention rates (%)	10	n/a	33¶	3¶	28#	11#	32#	4#	14¶	3¶

^†Zelen randomization.

^‡Minor and major complications at 1 month.

^§6 weeks.

^¶2 years.

^#5 years.

n/a: Not applicable; NS: Not significant; H: Hysterectomy; HMB: Heavy menstrual bleeding; M: Myomectomy; QoL: Quality of life.

Reproduced from [12] with permission from Springer.

UAE compared with hysterectomy

All the trials showed a significantly reduced hospital stay with a more rapid return to normal activities with UAE compared with hysterectomy. The UK REST trial [4] used quality of life assessed using the generic measure SF36 instrument as the primary outcome measure. There was no difference in this outcome at both 1 and 5 years, with both the UAE and surgical groups returning to normalized scores. Satisfaction scores were very high in both groups (UAE 90% and surgery 87%) at 5 years [14]. However, the cumulative need for re-intervention at 5 years was significantly higher in the UAE group (32%) than the surgery group (4%). This was either due to re-intervention for a complication or more commonly recurrent or persistent symptoms. This re-intervention included hysterectomy, myomectomy or repeat UAE. The adverse event rate between surgery and UAE was similar.

It seems therefore that UAE carries some advantages and disadvantages when compared with hysterectomy. Assuming the patient is suitable for either treatment, the informed individual should herself choose between surgery and UAE. The economic analysis conducted alongside the REST trial showed surgery and UAE to have comparable cost–effectiveness at 5 years [14].

UAE compared with myomectomy

Only two trials (MARA and FUME) have studied this group of patients [7,8] and the largest excluded patients wishing to become pregnant [8]. Both trials showed a more rapid recovery period and reduced hospital stay in the UAE group compared with myomectomy. The FUME trial showed a similar improvement in quality of life using the fibroid specific instrument (UFS-QoL) with no significant difference in major complications (UAE 3%, myomectomy 8%) [8]. At 2 years, there were significantly more re-interventions in the UAE arm (14%) than the myomectomy arm (3%). The MARA trial [7] is the only one to report on pregnancy outcomes and during the mean follow-up period of 25 months, there were more pregnancies (33), labors (19) and fewer abortions (6) in the myomectomy group compared with UAE pregnancies (17), labors (5) and abortions (9). It was concluded the myomectomy group had a greater chance of pregnancy. No significant differences were seen in technical success rate, symptom control, re-intervention for fibroid recurrence or complications between the groups in the MARA trial. The Cochrane review noted that “the single trial looking at fertility issues limits any conclusion to be made regarding the effects of UAE on fertility and future pregnancy.” [13].

In view of the uncertainty surrounding the use of UAE in women who would ordinarily be offered a myomectomy it is hoped that the FEMME trial [1], which has now closed to recruitment, will help provide more clarity. This RCT is comparing UAE with myomectomy assessing adverse events and pregnancy outcomes in addition to quality of life.

Complications & management

Minor complications are fairly common with a reported incidence from the trials of 20–30% which is very similar to that of hysterectomy (Table 1). Minor complications include:

• ‘Postembolization syndrome’ — fever, pain, systemic upset, raised inflammatory markers;

• Infection – fever, pain, systemic upset, purulent vaginal discharge;

• Nonpurulent vaginal discharge;

• Delayed return of menses.

Major complications occur in 8–12% which again is very similar to hysterectomy. The time at which complications appear is different to surgery, however, where most complications occur within the traditional 30-day postoperative window. The trials found that for UAE most major complications occur beyond the 30 days; indeed a fibroid expulsion has been reported 4 years after UAE. Major complications include:

• Fibroid expulsion (8%);

• Premature ovarian failure (1–2%);

• Infection leading to hysterectomy <1%;

• Deep venous thrombosis.

The management of these complications is usually expectant observation, for example:

Nonpurulent vaginal discharge is managed expectantly. It can be difficult to separate postembolization syndrome from infection and both antibiotics and analgesia should be administered if there is clinical doubt. Usually treatment can be on an outpatient basis but admission to hospital is occasionally required for observation and parenteral drug administration. Persistent pain should raise the possibility of fibroid expulsion and sometimes it may be necessary to dilate the cervix to allow passage of a necrotic fibroid. Severe infection requiring rescue hysterectomy was reported with an incidence of 1% in the early days of UAE. With better early recognition of infection and perhaps the use of prophylactic antibiotics, salvage surgery should only be needed in exceptional circumstances.

Ovarian function

There has always been understandable concern as to whether UAE could impair ovarian function. It is known that the ovarian and uterine arteries communicate and therefore it is possible at least in theory for embolic material to enter the ovarian circulation. However, it has been difficult to measure and quantify any objective damage. Ovarian function is classically assessed by a standard gonadotrophin hormone profile (follicle stimulating hormone [FSH], luteinizing hormone [LH] and estradiol [E2]). However, these markers only become abnormal as the ovary finally fails. The REST, EMMY and MARA trials have all reported gonadotrophin outcomes [7,15–16]. Anti-mullerian hormone (AMH) is a newer assay which gives a surrogate assessment of follicular reserve and is much more sensitive in detecting potential ovarian damage. Unfortunately, only the EMMY trial [15] used this assay and then only in a subgroup (n = 63).

The REST & EMMY trials used an FSH level >40 IU/l and MARA FSH >20 IU/l to indicate ovarian damage. Results from REST and EMMY when pooled showed an elevation in these hormone levels but there was no group difference between UAE and hysterectomy [15,16]. The small subgroup having AMH levels measured showed a reduction after both UAE and hysterectomy with partial recovery in the hysterectomy group. There was no significance between group differences. The MARA trial showed no significant hormonal changes from baseline in either group and no between group differences.

A small prospective cohort study (n = 36) of women aged 26–39 years undergoing UAE were compared with a matched cohort of 36 control women over a 5-year period [17]. Standard gonadotropin assays and ultrasound ovarian volume and follicle measurements were taken. Although there was a significant loss of ovarian volume and function over time, there was no group difference suggesting UAE does not lead to accelerated decline in ovarian function.

It would appear therefore that ovarian function deteriorates after both hysterectomy and UAE. The etiology is unclear; interruption to the ovarian blood supply is a possibility with both procedures but the natural history cohort study suggests it may simply be due to natural age related ovarian attrition. Data from the MARA trial showed no evidence of ovarian decline after either UAE or myomectomy. Thus it has been difficult to assess ovarian function against a background of natural ovarian decline. While controversy continues regarding the impact of UAE on ovarian function, larger trials such as FEMME which is measuring AMH in several hundred women, should provide further information.

Pregnancy & future fertility

In 2004, the American College of Obstetricians and Gynecologists recommended that UAE should be considered investigational or relatively contraindicated in women wishing to retain fertility. The latest guidance from NICE in 2010 states “patients contemplating pregnancy should be informed that the effects of the procedure on fertility and on pregnancy are uncertain.” [18].

There has been one RCT (MARA) designed to look at fertility outcomes and it was small (n = 121) [7]. This trial reported that 50% of women who tried to conceive after UAE became pregnant compared with 78% after myomectomy. The live birth rate was 19% in the UAE group and 48% in the myomectomy group. The rate of spontaneous abortion was 64% (UAE group) and 23% (myomectomy group). These differences were all statistically significant. However, this trial has been criticized and contained several flaws not least of which was that the number of women trying to get pregnant was higher in the myomectomy group. Furthermore in the UAE group if the fibroid had not shrunk by a predetermined size then myomectomy was carried out. The Cochrane review stated that the MARA trial limits any conclusion to be made regarding the effects of UAE on fertility and future pregnancy [13].

A systematic literature review compared 227 pregnancies reported after UAE with a group (matched for age and fibroid location) who became pregnant but had untreated fibroids [19]. This review showed a significantly higher miscarriage rate in those who had undergone UAE (35%) compared with the untreated fibroid group (16%; p < 0.001). However, there was no difference in obstetric complications rates which included preterm delivery, malpresentation and intrauterine growth retardation.

There remains considerable uncertainty as to the optimal treatment for women with fibroids who wish to maintain their fertility. Concerns have raised regarding abnormal placentation post UAE possibly leading to higher risks of early miscarriage and postpartum hemorrhage. However, when assessing the putative risks of UAE, it is important to do so in the context that myomectomy has not been studied in the same depth as UAE. Myomectomy carries risks such as adhesions and uterine scarring which may impair fertility. Both UAE and myomectomy are used in this patient group without good evidence of clinical effectiveness and this is an unsatisfactory situation for women. The UK FEMME trial is currently randomizing 216 women to try and answer the question of which procedure is best.

National guidance conclusion

The latest guidance from NICE for UAE for fibroids was issued in 2010 (IPG367) [18]. It concludes that the “current evidence on uterine artery embolization for fibroids shows that the procedure is efficacious for symptom relief in the short and medium term for a substantial proportion of patients. There are no major safety concerns. Therefore this procedure may be used provided that normal arrangements are in place for clinical governance and audit”. In addition to quality assuring clinical practice we hope that current and future research will hopefully better define the relative risks and benefits of UAE compared with other, particularly surgical treatments for fibroids. This is particularly important for women with symptomatic fibroids contemplating future pregnancy.

More recently (2013) NICE has reviewed UAE for adenomyosis looking at safety and efficacy (IPG473). There were no safety issues and it was noted that symptoms may return and further procedures may be needed.

It is generally accepted that the results of UAE for adenomyosis are less well understood than for fibroids and no clinical trials have yet been conducted.

Executive summary

• Uterine artery embolisation is an established alternative treatment for heavy menstrual bleeding in women not wishing surgery.

• Clinical trials have shown similar quality of life and satisfaction rates compared with surgery.

• The re-intervention rate is higher that after hysterectomy.

• Comparison with myomectomy is less clear and currently being studied in a trial (FEMME).

• Uterine artery embolisation should be discussed with patients as an alternative to surgery.