Pregnancy outcomes after uterine artery embolisation for fibroids
Department of Obstetrics and Gynaecology, Institute for Women's Health, University College London, 86–96 Chenies Mews, London WC1E 6HX, UK Email: firstname.lastname@example.org
- • Data on pregnancy following uterine artery embolisation (UAE) are scarce, with just over 200 pregnancies reported to date.
- • The weight of retrospective data and two small prospective trials of UAE versus surgical intervention suggest increased levels of adverse pregnancy outcomes following fibroid embolisation.
- • In the absence of more robust evidence, caution should be exercised in recommending UAE to women who retain reproductive ambitions.
- • To increase awareness of the potential reproductive risks of UAE.
- • To appreciate the limitations of the existing evidence.
- • Caution should be exercised before advocating UAE for intractable fibroid-related symptoms in women desirous of future pregnancy.
Please cite this article as: Homer H, Saridogan E. Pregnancy outcomes after uterine artery embolisation for fibroids. The Obstetrician & Gynaecologist 2009;11:265–270.
Uterine artery embolisation (UAE) is a nonsurgical approach to treating uterine fibroids. There is compelling level A evidence (including the EMMY [EMbolisation versus hysterectoMY] and the REST [Randomised trial of Embolisation versus Surgical Treatment for fibroids] trials) that UAE is a safe and effective alternative to hysterectomy for alleviating fibroid-related disorders such as menorrhagia and pressure symptoms.1–3 However, there is concern regarding the use of UAE in women who have not yet completed their families. With the prevalence of fibroids being as high as 20–40% among women of reproductive age,4 this is a pressing issue as much of the treatment population are potentially fertile.
The most comprehensive review on pregnancy outcome following UAE5 reported on 53 such pregnancies and found that they were significantly more likely to be complicated by malpresentation and preterm delivery than pregnancies that arose following laparoscopic myomectomy. There was also a nonsignificant trend towards higher rates of spontaneous miscarriage and postpartum haemorrhage. Since this paper, further series have analysed the course of at least a further 150 pregnancies that have arisen post-UAE. Here we provide an overview of the UAE technique, review recent evidence regarding pregnancy outcome and conclude by exploring some of the mechanisms by which UAE could influence reproductive performance.
Rationale and technical aspects of UAE
Uterine artery embolisation was first described over a decade ago6 and is proposed to exploit a differential recovery ability between fibroids and normal uterine tissues following ischaemia at the arteriolar level.7 Using angiographic catheters inserted percutaneously using a common femoral artery approach, embolic agents such as polyvinyl alcohol particles or trisacryl gelatin spheres are injected into both uterine arteries under X-ray guidance until the blood flow becomes sluggish.8 The procedure takes about an hour to complete and is usually performed under conscious sedation. A day-case or overnight stay is usually sufficient, with the primary postoperative sequela being ischaemic pain for the first 12 hours, which usually subsides over the ensuing 12 hours.9 Subsequent recovery is usually brief, with 4–5 days of uterine cramping and constitutional embolisation syndrome (low-grade fever, fatigue, nausea and malaise) and a return to normal activities anticipated within 1–2 weeks.
A relatively frequent complication of UAE (up to 10%) is expulsion of an infarcted fibroid, usually within 6 months of the procedure; this is often associated with submucosal fibroids or intramural fibroids with a submucosal component. Occasionally, a partially infarcted fibroid may remain firmly attached to the uterine wall and require hysteroscopic resection or dilatation and curettage. A more serious complication is uterine infection, reported in fewer than 1% of cases, which could lead to sepsis and the need for hysterectomy.9 Deaths from overwhelming sepsis have been reported.10
Prior reports of pregnancy after UAE
Data on pregnancies after UAE continue to accrue and demonstrate that women can and do conceive and have uncomplicated deliveries of healthy offspring. However, it is widely believed that there is an increased risk of complications in pregnancy following UAE. Based largely on earlier reports,5,11 the joint Standards of Practice Committee of the Cardiovascular and Interventional Radiological Society of Europe and the Society of Interventional Radiology considered the desire to maintain childbearing potential to be a relative contraindication to UAE.12
The earliest review11 intimated that in pregnancies post-UAE women experienced higher rates of adverse outcome when compared with accepted figures for the general population. This is not entirely surprising since, by virtue of having fibroids, the UAE population will inevitably constitute a high-risk group. These limitations notwithstanding, this early report served to alert clinicians to the types of risks that may be associated with UAE pregnancies.
The second and most recent comprehensive review5 on pregnancies following UAE for fibroids was undertaken in 2004. An additional objective of this study was to compare pregnancy outcomes in women who had had UAE with those in women having laparoscopic myomectomy for the treatment of their fibroids. The authors obtained all the data available at the time regarding pregnancy outcome following UAE, representing a total of 53 pregnancies. The laparoscopic myomectomy group was derived from three of the largest published series of pregnancy after laparoscopic myomectomy, constituting a total of 139 pregnancies.5 When compared with pregnancies post-laparoscopic myomectomy, the authors found that in pregnancies post-UAE there was significantly higher risk of preterm delivery (P = 0.008; odds ratio [OR] 6.2; CI 1.4–27.7) and malpresentation (P = 0.046; OR 4.3; CI 1.0–20.5). There was also a trend towards higher rates of miscarriage (P = 0.175; OR 1.7; CI 0.8–3.9), postpartum haemorrhage (P = 0.093; OR 6.3; CI 0.6–71.8) and caesarean delivery (P = 0.662; OR 1.2; CI 0.5–2.6). Based on this, the authors concluded that pregnancies post-UAE are at increased risk compared with pregnancies post-laparoscopic myomectomy, reinforcing the viewpoint of their earlier paper11 that UAE impacts negatively on pregnancy success.
Important caveats need to be borne in mind when interpreting these data, however. As the authors themselves document, women undergoing UAE were significantly older (37.6 ± 4.4 versus 34.2 ± 3.9; P = 0.001) and possessed significantly larger fibroids (8.2 ± 3.3 cm versus 5.4 ± 2.4 cm; P = 0.001) than women having laparoscopic myomectomy,5 two features anticipated to exert independent adverse effects on pregnancy outcomes. Overall, therefore, the authors identified women having UAE as a poorer prognosis group but it is uncertain how much is attributable to UAE and what proportion is purely the result of confounding factors such as advanced maternal age and differences in fibroid characteristics.
Pregnancy after UAE revisited
Since the publication of the latest comprehensive review,5 a number of important papers have reported on pregnancy outcomes following UAE. Two years after the publication of its initial results in 2003, the Ontario Multicenter Trial13 reported on pregnancies that were identified during its 5-year prospective follow-up interviews. The Ontario Multicenter Trial was a prospective multicentre clinical study designed to evaluate the safety, effectiveness and durability of UAE as an alternative to hysterectomy for symptomatic leiomyomata.14 It incorporated the practices of 11 interventional radiologists in 8 Ontario hospitals and involved 555 women undergoing UAE. During the follow-up period there were 24 pregnancies among 21 women, of which 2 were electively terminated. The remaining 22 pregnancies resulted in 18 live births (81.8%) and 4 spontaneous miscarriages (18.2%). Among the 18 live births, 14 (77.8%) were term, 4 (22.2%) were preterm and 9 (50%) were delivered by caesarean section. There was one case of malpresentation (5.6%) and 4 infants (22.2%) that were small for gestational age/intrauterine growth restricted (5th centile). Unique to this study was an unusually high rate of abnormal placentation (3 [16.7%]), all of which resulted in postpartum haemorrhage.
Walker and McDowell (2006)15 published the single largest series, which was a retrospective analysis of pregnancies following UAE in a UK setting. They identified 56 pregnancies that accrued from among a total of 1200 UAE cases. Of these 56 pregnancies, 3 were terminated, 1 was an ectopic and 2 resulted in stillbirths. Among the remaining 50 pregnancies there were 17 miscarriages (34%) and 33 live births (66%). Of the live births, 27 (81.8%) delivered at term, 6 (18.2%) were preterm and 6 (18.2%) were complicated by postpartum haemorrhage, which was very similar to the rates observed in the Canadian study. Twenty-four of the 33 (72.7%) infants were delivered abdominally and included 3 (9.1%) malpresentations, which was higher than the rates observed by Pron et al.13 Unlike the latter study, however, there was only one case of abnormal placentation (3%) and one of fetal growth restriction (3%).
A prospective, cohort-controlled study conducted in the Czech Republic reported on the outcomes of 28 pregnancies among 20 women treated with UAE for symptomatic fibroids.16 After excluding 1 ectopic pregnancy, 2 elective terminations and 1 continuing pregnancy, there were 14 miscarriages (58.3%) and 10 live births (41.7%). Similar to the above reports, there were high rates of preterm delivery (n = 2 [20%]), caesarean section (n = 8 [80%]), malpresentation (n = 2 [20%]) and postpartum haemorrhage (n = 2 [20%]). One infant was small for gestational age (n = 1 [10%]) and one case was complicated by abnormal placentation (n = 1 [10%]).
The HOPEFUL study17 is a multicentre retrospective cohort study comparing the efficacy and safety of hysterectomy and UAE for the treatment of uterine fibroids across 18 National Health Service hospital Trusts in the UK. Data regarding 37 pregnancies that were recorded among 27 women in the HOPEFUL study were published in a separate paper.18 After excluding 2 ectopic pregnancies and 1 termination, there were 15 miscarriages (44.1%) and 19 live births (55.9%) among the remaining 34 pregnancies. Of these 19 pregnancies, the vast majority (n = 15 [78.9%]) were, again, delivered by caesarean section with no information regarding other obstetric outcomes such as gestation at delivery, malpresentation and so on.
The year 2008 saw the publication of reproductive results of the only randomised controlled trial involving UAE that specifically targeted women who were desirous of future pregnancy.19 The purpose of the study was to compare the efficacy and safety of UAE versus myomectomy (laparoscopic and open) for the treatment of symptomatic fibroids as well as their impact on fertility. The mean length of follow-up was 2 years, in which time only 26 women in the UAE cohort pursued pregnancy and half conceived a total of 17 pregnancies. Of these, 1 was an ectopic pregnancy, 1 was electively terminated and 1 pregnancy was still continuing at the time of publication. Of the residual 14 pregnancies, 9 ended in miscarriage (64.3%) and 5 resulted in live births (35.7%), all of which were at term. The trend towards high abdominal delivery rates seen in other studies was also reflected in this paper, as 3 (60%) were delivered by caesarean section. There was 1 case (20%) of postpartum haemorrhage and none of the fetuses were growth restricted. It must be emphasised, however, that only a very small number of women (n = 5) went past the first trimester in this cohort, making it difficult to make definitive statements regarding obstetric outcomes. In comparison, of the 25 completed pregnancies in the laparoscopic myomectomy arm, there were 19 live births (76%) and 6 miscarriages (24%), outcomes that were significantly better than in the UAE group. Preliminary results from this study also suggest benefits of laparoscopic myomectomy over UAE in terms of likelihood of conceiving, as 31 of 40 women (77.5%) desirous of pregnancy post-laparoscopic myomectomy achieved a pregnancy, versus 50% in the UAE arm.
In 2008 there were two further publications addressing pregnancy following UAE. Kim et al.20 prospectively followed up 87 women treated with UAE over a 3-year period. Fifteen pregnancies were recorded among 19 women attempting to conceive, of which 5 resulted in terminations and 1 was ectopic. Of the remaining 9 pregnancies, 3 ended in miscarriage (33.3%) and all but one of the 6 live births were delivered abdominally (83.3%). Finally, the most recent series is a prospective observational study of pregnancies following UAE that was undertaken in 100 women between January 2002 and June 2006.21 There was a total of 11 pregnancies, from which figures could be extracted for miscarriage rates (3 [27.3%]), prematurity (1/8 [12.5%]), fetal growth restriction (0%) and caesarean delivery (4/8 [50%]).
Based on the tabulated figures (Table 1), it can be seen that following on from the 51 pregnancies for which there were pregnancy outcome data in the last review,5 there are now at least 215 such pregnancies following UAE in the world literature (Table 1 and Table 2). There are similarities between the profiles of the pregnancies in the Goldberg review5 and the current pool of pregnancies. These include relatively high preterm delivery rates, often in the region of 15–20%, and a strikingly high caesarean delivery rate of up to 80% (Table 1). Interestingly, the miscarriage rates appear even higher in our current analysis (35.8%) compared with that published by Goldberg et al.5 (23.5%) (Table 2). This is accompanied by a concomitant fall in live birth rates from 76.5% to 64.2% (Table 1). In addition, there appear to be increases in the rates of fetal growth restriction (from 4.5% to 7.3%) and postpartum haemorrhage (from 5.7% to 13.9%) (Table 1). Overall, therefore, our updated data do not reveal any improvements in pregnancy prospects for women following UAE. On the contrary, many aspects appear to be worsened, most notably the risks of miscarriage and postpartum haemorrhage (Table 1). In keeping with an apparent detrimental impact of UAE on early pregnancy, we have found significantly higher miscarriage rates following UAE for fibroids than for women with untreated fibroids (OR 2.9; 95% CI 2.1–4.0).22
Obstetric outcomes in pregnancy after UAE
Rates of miscarriage after UAE
When our cumulative data are compared with figures for the general population (Table 3), many trends continue as were noted previously.5 These are represented by a trend towards increased rates of adverse pregnancy outcomes, including early pregnancy catastrophe and later obstetric complications such as preterm delivery, malpresentation and caesarean delivery. Also similar to the previous review,5 our analysis suggests that fetal growth restriction rates are comparable to the general population. Notably, however, unlike the Goldberg paper,5 which found that postpartum haemorrhage was no more likely in pregnancies following UAE than in the general population, our figures suggest that postpartum haemorrhage occurs with over twice the frequency following UAE.
Comparison of pregnancy outcomes in the general population and after UAE
Possible mechanisms by which UAE could influence pregnancy progression
Our updated data suggest that early pregnancy may be particularly vulnerable following UAE. This is reinforced by our findings, which indicate that UAE imposes additional risks for miscarriage over and above those known to be associated with fibroids.22 The implication is that UAE exerts deleterious effects on the intrauterine milieu, making it hostile to the early conceptus. One possibility is that endometrial ischaemia at the time of UAE induces a long-term effect that negatively alters the quality of potential implantation sites. In support of this, there have been case reports of endometrial atrophy following UAE23 and even of uterine necrosis requiring hysterectomy.24
Uterine artery embolisation may alter the embryo-endometrium interface via other mechanisms. It does not eliminate fibroids completely but instead induces a reduction in fibroid volume of about 50–60%9 accompanied by fibroid migration towards the endometrial cavity that could result in distortion of the endometrial contour. Among 51 women who were followed up by hysteroscopy following UAE for intramural fibroids, there was intrauterine protrusion of the fibroid remnant in 37% (19), intrauterine or cervical adhesions in 14% (7) and a communication between the myoma and the endometrial cavity in 10% (5); indeed, in only 37% (19) of cases was hysteroscopy completely normal.25 Any distortion of the endometrial cavity as a consequence of UAE in combination with relative endometrial ischaemia is likely to escalate the risk of early pregnancy failure.
Based on our analysis, UAE pregnancies may be at heightened risk of a wide range of antenatal complications (Table 1 and Table 3). A lower threshold for elective caesarean section is perhaps not surprising as practitioners may adopt a precautionary approach in women in whom myometrial integrity is potentially compromised. It must also be remembered that UAE does not remove fibroids and the mere presence of such tumours and a tendency for more severe fibroid burden in women following UAE could sway the decision away from vaginal delivery. In the largest single series of completed pregnancies post-UAE, 72.7% (24/33) of live births were delivered abdominally, of which 54.2% (13/24) were elective with fibroids being the indication in 69.2% (9/13) of these.15 Additionally, the increased risk of malpresentation associated with pregnancies in women with fibroids26 appears to be preserved post-UAE which in turn will strongly influence mode of delivery.
Whilst not an effect on pregnancy per se, it is important to note that UAE can adversely affect fertility by inducing transient or permanent amenorrhoea accompanied by other symptoms of ovarian failure in up to 5% of women. This is felt to be the consequence of non-target ovarian embolisation by utero-ovarian collaterals resulting in ovarian ischaemia and depletion of ovarian follicles.27 Most cases appear to occur in women over 45 years,28 but younger women are not immune from this devastating sequela.29
Given the efficacy of UAE for relieving fibroid-related symptoms, women with reproductive ambitions and symptomatic fibroids will undoubtedly be drawn to this minimally invasive approach for fibroid treatment. This is especially so since myomectomy carries substantial risks such as bleeding, infection and adhesion formation, which are themselves threats to fertility. However, our data suggest that in pregnancies following UAE there is an increased risk of miscarriage during the early stages and, from an obstetric perspective, that they are more susceptible to preterm delivery, caesarean section, malpresentation and postpartum haemorrhage compared with the general population.
An obvious limitation of this study is that it combines the results of several individual series of varying experimental rigour and design (mostly retrospective) and reports aggregate outcomes. It must nevertheless be emphasised that, on the whole, there are very few studies looking at pregnancies post-UAE and even fewer that set out to examine pregnancy outcome in women following UAE prospectively and specifically. Indeed, there is only one randomised controlled trial19 and this study lacked sufficient power to provide definitive statements because of small numbers of completed pregnancies. In spite of this, the majority of studies are strikingly consistent in reporting increased pregnancy risk.
Although the two comparative studies5,19 involving UAE pregnancies sought to compare the radiological approach with myomectomy, it is important to bear in mind that surgery may not always be a viable option for treating symptomatic fibroids; for instance, because of a lack of technical feasibility or because of a woman's choice to avoid surgery. For such women, UAE may ultimately be chosen and it will be important for caregivers to be aware of the potential risks associated with ensuing pregnancies highlighted in this review. Based on our cumulative figures, and in the absence of more robust prospective data, we would advocate that practitioners continue to adopt a cautious approach when considering UAE for treating symptomatic fibroids in women desiring future pregnancy.