Objective
To summarize the evidence on short-, mid-, and long-term results up to 5 years of uterine artery embolization in comparison to surgery.
Study Design
We searched the CENTRAL, MEDLINE and EMBASE databases for randomized clinical trials comparing uterine artery embolization with hysterectomy/myomectomy in premenopausal women with heavy menstrual bleeding caused by symptomatic uterine fibroids, written from September 1995 to November 2010. Two reviewers independently assessed methodologic quality and extracted data from included trials.
Results
Four randomized controlled trials with a total of 515 patients were included. On the short-term, uterine artery embolization showed fewer blood loss, shorter hospital stay, and quicker resumption of work. Mid- and long-term results showed comparable health-related quality of life results and a higher reintervention rate in the uterine artery embolization group, whereas both groups were equally satisfied.
Conclusion
Uterine artery embolization has short-term advantages over surgery. On the mid- and long-term the benefits were similar, except for a higher reintervention rate after uterine artery embolization.
During the last decade, uterine artery embolization (UAE) has been studied as a minimally invasive treatment alternative to surgery in reducing symptoms of heavy menstrual bleedings caused by fibroids. UAE was first described in 1995 by Ravina et al and was reported to be safe and effective as an alternative treatment option, although no randomized controlled trials (RCTs) were available at that time. Afterward, several randomized controlled trials comparing UAE and surgery occurred. A Cochrane systematic review that summarized the short-term results (until 6 months of follow-up) of these trials showed a reduction of total length of hospital stay, quicker resumption of daily activities, more minor complications after UAE, and similar major complications and satisfaction. Despite this evidence the use of UAE in the treatment of symptomatic uterine fibroids is relatively low. One of the reasons for this could be the absence of knowledge on the long-term effectiveness of UAE compared with hysterectomy or myomectomy. After the publication of the abovementioned systematic review, several of the trials comparing UAE and surgery have been published describing mid- and long-term results until 5 years of follow-up. The aim of the current review is to summarize all results of all eligible trials comparing UAE with hysterectomy and myomectomy with regard to procedural results, postprocedural results, hospital stay, resumption of activities, pain, and other symptoms, reinterventions, hemoglobin level, follicle-stimulating hormone (FSH) level, menopausal complaints, health-related quality of life (HRQOL), and satisfaction.
See Journal Club, page 389
Materials and Methods
Literature search
We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, and EMBASE from September 1995 to November 2010, using the following search terms: fibroid, (leio)myoma, heavy menstrual bleeding, menorrhagia, uterine artery/fibroid embolization, hysterectomy, myomectomy, clinical trial, and RCT. The search strategy is described in detail in Table 1 . We also searched the metaRegister of Controlled Trials, a major international searchable database of ongoing randomized controlled trials in all areas of health care, built by combining registers and held by public, charitable, and commercial sponsors of trials ( http://controlled-trials.com.easyaccess1.lib.cuhk.edu.hk/mrct/ . Accessed Nov. 1, 2010). In addition to this, the citation lists of relevant publications, review articles, and included studies were searched and authors of published and unpublished trials for additional information were contacted when necessary. Non-RCTs (letters, reviews, editorials, and commentaries) were excluded. The bibliographies of identified studies and relevant reviews were reviewed for additional studies. No language restriction was made. The methods of analysis and inclusion criteria were specified in advance and documented in a protocol.
| Database | MEDLINE | EMBASE | The Cochrane Central Register of Controlled Trials |
|---|---|---|---|
| Search term 1 | (randomi?ed controlled trial.pt. OR controlled clinical trial.pt OR randomised controlled trials/ OR random allocation/ OR double-blind method/ OR single-blind method/ OR clinical trial.pt. OR exp clinical trials/ OR (clin$ adj25 trial$).tw. OR ((singl$ or doubl$ or treb$ or tripl$) adj25 (blind$ or mask$)).tw. OR placebos/ OR placebo$.tw. OR random$.tw. OR research design/) | (Controlled study/ or Randomised Controlled Trial/ OR Double Blind Procedure/ OR Single Blind Procedure/ OR Crossover Procedure/ OR Drug Comparison/ OR Placebo/ OR Random$.tw. OR latin square.tw. OR crossover.tw. OR cross-over.tw. OR placebo$.tw. OR ((doubl$ or singl$ or tripl$ or trebl$) adj5 (blind$ or mask$)).tw. OR (comparativ$ adj5 trial$).tw. OR (clinical adj5 trial$).tw.) | |
| AND | AND | ||
| Search term 2 | (menstrual$ adj5 (bleed$ OR blood).tw OR heavy NEXT menstrual$.tw OR . menorrhagia.tw OR hypermenorrh$.tw OR fibroid$.tw OR leiomyom$.tw OR hysteromyom$.tw) | (menstrual$ adj5 (bleed$ OR blood).tw OR heavy NEXT menstrual$.tw OR menorrhagia.tw OR hypermenorrh$.tw OR fibroid$.tw OR leiomyom$.tw OR hysteromyom$.tw) | (uterine artery emboli*ation OR uterine fibroid emboli*ation OR UAE) |
| AND | AND | AND | |
| Search term 3 | uterine artery emboli?ation.tw OR uterine fibroid emboli?ation.tw OR UAE.tw OR fibroid adj5 emboli?ation.tw OR uter$ adj5 emboli?ation.tw) | (uterine artery OR emboli?ation.tw OR uterine fibroid emboli?ation.tw OR UAE.tw OR fibroid adj5 emboli?ation.tw OR uter$ adj5 emboli?ation.tw) | (fibroma* OR fibroid* OR leiomyom* OR menstrua* NEXT (bleed* OR blood) OR heavy NEXT menstrua* OR menorrhagia OR hypermenorrhea) |
Selection of trials
The titles and abstracts of all potentially relevant studies were screened by 2 reviewers independently (S.M.K. and S.B.) and eventually in consensus. Full-text articles were obtained if the following selection criteria were met:
Types of studies: randomized controlled trials (RCTs).
Types of participants: women with symptomatic uterine fibroids.
Types of intervention/comparison: UAE compared with surgical treatment (hysterectomy or myomectomy).
Duration of follow-up: all time of follow-up up to 5 years.
Outcome measures: procedural results, postprocedural results, hospital stay, resumption of activities, pain and other symptoms, complications, reinterventions (interventions after UAE/surgery), hemoglobin-level, menopausal status, pelvic floor problems, HRQOL, and satisfaction.
Data collection
From each selected trial data were extracted independently by the same reviewers. Methodologic qualities of the studies were assessed using the Delphi list. The following items of the Delphi lists were scored: (1) whether a method of randomization was performed and whether the treatment allocation was concealed, (2) whether the groups were similar at baseline regarding the most important prognostic indicators, (3) whether the eligibility (inclusion/exclusion) of criteria were specified, (4) whether relevant data were presented with confidence intervals, (5) whether the data analysis was performed according to intention-to-treat principle, and (6) whether drop-outs were adequately reported. In all trials, neither the outcome assessor nor the care providers and patients could be blinded for the received treatment. Therefore, these items of the Delphi list were not scored.
Data analysis
Dichotomous data were presented as odds ratio (OR) with 95% confidence intervals (CIs) and continuous outcome measures by weighted mean differences and 95% CIs. In addition, if more than 1 dataset was available for an outcome, fixed effects and the random effects models were used to pool results and to present these in Forest plots. The I 2 test was used to test for heterogeneity between studies. I 2 ranges from 0–100% and measures the degree of inconsistency across studies in metaanalysis. If no statistically significant heterogeneity was found, the fixed-effects model was used to pool results. In the presence of statistical heterogeneity, a random-effects model was applied for metaanalysis. Statistical significance was set at P = .10 for testing heterogeneity, and at P = .05 for all other analyses. Data were entered and analyzed using Cochrane RevMan software (version 5.0.; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark).
Results
Search strategy and study selection
The search strategy revealed 196 articles; 182 articles were excluded after reading titles and abstracts because of not being prospective, not being RCTs or not comparing UAE with surgery; 3 articles were excluded after reading the whole article; 1 appeared not to be randomized; and 2 were retrospective studies. All remaining 11 articles, originating from randomized controlled trials, reported on the outcomes of UAE compared with those of surgery (hysterectomy or myomectomy) in the treatment of symptomatic uterine fibroids, and were included in our analysis. The summary of the search strategy and study selection is described in Table 1 and in the flowchart ( Figure 1 ).
Study characteristics
The 11 included articles reported data from 4 RCTs: the EMMY trial, the REST trial, the trial by Mara et al, and the trial by Pinto et al; 6 articles contained results from the EMMY trial, 3 reported data from the REST trial, 1 from the trial by Pinto et al, and 1 from the trial by Mara et al. In the Cochrane review a long-term paper of Pinto et al was being referred to as an ongoing study, with 2 years follow-up results to be awaited. When contacting Pinto et al, follow-up appeared not to have exceeded the 6 months mentioned in their publication. Study characteristics are described in detail in Table 2 . The results of the Delphi score (methodologic quality) of the 4 trials are presented in Table 3 . All trials met the Delphi criteria, except for the trial by Pinto et al, in which analysis was not done according to the intention-to-treat principle.
| Study | Description |
|---|---|
| EMMY trial | |
| Study design | Randomized controlled trial. Attending gynecologist contacted the trial bureau by telephone, where the participant was registered and randomly assigned (1:1) to UAE or hysterectomy, using a computer-based minimization scheme (‘balancing procedure’), and stratified for study center. The randomization result was recorded electronically. |
| Participants | Women with symptomatic fibroids. Inclusion criteria: the clinical diagnosis of uterine fibroids confirmed by ultrasonography; menorrhagia (subjectively reported by the patient as increased or prolonged menstrual blood loss which caused dysfunction in daily life) was their predominant complaint, among other possibly fibroid-related signs and symptoms; they were premenopausal; and they were to be scheduled for a hysterectomy. Exclusion criteria: preservation of the uterus was warranted for future pregnancy; renal failure (creatinine >150 mmol/L), active pelvic infection, or clotting disorders were clinically established; they were allergic to contrast material; uterine malignancy was suspected; submucosal fibroids with 50% of their diameter within the uterine cavity or dominant pedunculated serosal fibroids were present. |
| Interventions | 177 women were randomized: 88 were allocated to UAE and 89 to hysterectomy. UAE was successfully performed in 72 of 81 women, 10 of whom had a unilateral procedure because of single-sided arterial blood flow to the fibroid or technical failure and 4 women (4.9%) with bilateral UAE failure underwent a hysterectomy. Control group: after randomization, 75 (85.4%) underwent hysterectomy with 84% through an abdominal route. The mean age in the UAE group was 44.6 y (SD 4.8) and in the hysterectomy group 45.4 y (SD 4.2). |
| Outcomes | The outcomes of this trial were reported in 6 articles. Peri- and postprocedural results, pain and recovery to daily activities, clinical results (ie, changes in menstrual bleeding) and reinterventions up to 2 y (Volkers et al, 2007), loss of ovarian reserve and menopausal symptoms up to 2 y, HRQOL, urinary and defecatory function, and satisfaction about the received treatment up to 2 y and reintervention, HRQOL, menopausal symptoms 5 y after treatment, urinary and defecation function, and satisfaction about the received treatment and recommendation to a friend. |
| REST trial | |
| Study design | Randomized controlled trial. Patients were randomly assigned to study groups according to a computer-generated schedule (permuted blocks) held by the trial coordinator. Randomization was stratified by center and was performed in a 2:1 ratio, with twice as many patients allocated to the UAE group as to the surgical group. |
| Participants | Inclusion criteria were: women with symptomatic fibroids, age 18 y or older, 1 or more fibroids of more than 2 cm in diameter that could be adequately visualized with the use of MRI, caused symptoms (such as menorrhagia or pelvic pain and pressure), and considered by the patient’s physician to justify surgical treatment. Exclusion criteria included a contraindication to MRI, severe allergy to iodinated contrast media, subserosal pedunculated fibroids, recent or ongoing pelvic inflammatory disease, pregnancy, and any contraindication to surgery. There was no upper limit on the size or number of fibroids. |
| Interventions | A total of 157 women were randomly assigned to study groups: 106 to undergo UAE and 51 to undergo surgery, including 43 hysterectomies and 8 myomectomies. All the hysterectomies and myomectomies were performed through an abdominal incision. The mean age in the UAE group was 43.6 y (SD 5.5) and in the surgical group 43.3 y (SD 7.1). |
| Outcomes | The primary outcome measure was HRQOL measured by SF-36. Secondary outcomes included an assessment of findings on the EuroQol-5D questionnaire, an instrument used to measure preferences for certain health outcomes, including hysterectomy. A satisfaction score measuring whether patients would recommend the procedure to a friend; a linear-analogue pain score at 24 h; the presence or absence of complications; and treatment failure, defined as the need for subsequent intervention for symptom control, including hysterectomy or repeated UAE. Ovarian function according to FSH levels was measured after 1 y and described in a separate article. After 5 y was measured: HRQOL, reinterventions, recommendation to a friend, pregnancies, and economic analysis. |
| Pinto et al | |
| Study design | Randomized controlled trial. Patients were assigned to 1 of 2 groups: patients who were offered the option of undergoing either UAE or hysterectomy (group 1) and patients who were not informed of the alternative treatment—that is, UAE (group 2). |
| Participants | The study population consisted of women with bleeding uterine fibroids who were candidates for hysterectomy according to the criteria approved by the American College of Obstetricians and Gynecologists. Patients who wished to maintain fertility, had fibroids larger than 10 cm in diameter, had contraindications to surgery, or had sensitivity to iodine-based contrast material were excluded. |
| Interventions | A total of 40 patients underwent UAE and 20 underwent hysterectomy, all abdominal. Three of the patients who underwent hysterectomy had undergone UAE first. The mean age in the UAE group was 46.4 y (SD 4.4) and in the hysterectomy group 44.6 y (SD 5.0). |
| Outcomes | The primary variables that were considered for evaluation of the effectiveness, efficiency, and safety of the 2 procedures were, respectively, bleeding cessation, total length of hospital stay, and resulting complications. The lengths of hospital stay in the 2 study arms were compared on an intent-to-treat basis. Owing to crossover between the treatment arms, however, effectiveness and safety were evaluated on the basis of the actual treatment received. Follow-up of the patients was 1 y. |
| Mara et al | |
| Study design | Randomized controlled trial. Every newly recruited patient was randomly assigned a computer-generated integral number from 1 to 100 (using a random number generating program available at http://www.random.org ). |
| Participants | Inclusion criteria were: US-verified intramural fibroid at least 4 cm at its largest diameter (in the case of multiple fibroids, at least 1 with a size of 4 cm); age 40 y; serum FSH concentration >10 IU/L (on the third day of menstrual cycle); and planned pregnancy. Intramural fibroids were defined using US as a uterine wall expansion of typical echo structure with the prevalent part of its volume inside the myometrium. Exclusion criteria included; nonintramural localization of the main fibroid (submucosal and subserous); size of the dominant myoma 12 cm in its largest diameter (according to US) or uterus enlarged to the size corresponding to 4 mo of pregnancy (according to bimanual pelvic examination); previous myomectomy, embolization, or hormonal therapy of fibroids with gonadotropin releasing hormone agonists or danazol; suspected uterine sarcoma or diffuse adenomyosis (according to US or MRI); and serious disease contraindicating gravidity. |
| Interventions | A total of 58 embolizations and 63 myomectomies (42 laparoscopic, 21 open) were performed. One hundred eighteen patients have finished at least a 12-mo follow-up; the mean follow-up in the entire study population was 24.9 mo. The mean age in the UAE group was 32.4 y and in the myomectomy group 32.0 y. |
| Outcomes | Outcome measures were technical success, symptomatic effectiveness, postprocedural FSH levels, number of reinterventions for fibroid recurrence or regrowth, complication rates, gestations, and obstetric and perinatal results. |
| Studies | EMMY trial | REST trial | Pinto et al trial | Mara et al trial |
|---|---|---|---|---|
| 1. Treatment allocation | ||||
| a. Was a method of randomization performed? | YES | YES | YES | YES |
| b. Was the treatment allocation concealed? | YES | YES | YES | YES |
| 2. Were the groups similar at baseline regarding the most important prognostic indicators? | YES | YES | YES | YES |
| 3. Were the eligibility criteria specified? | YES | YES | YES | YES |
| 4. Are relevant data presented with confidence intervals? | YES | YES | YES | YES |
| 5. Is data analysis performed according to intention-to-treat principle? | YES | YES | NO | YES |
| 6. Is reporting drop-outs adequate? | YES | YES | YES | YES |
Outcome measures
Outcome measures measured in the trials are summarized in detail in Table 4 . The data were collected at different time-points: (1) intraprocedural results (at the time of the procedure), (2) early postprocedural results (first 24 hours, during hospital stay or day 1 to day 30), (3) late postprocedural results (between 30 days and 6 months), (4) follow-up at 6 months, (5) follow-up at 1 year, (6) follow-up at 2 years, (7) follow-up at 3 years, (8) follow-up at 4 years, and (9) follow-up at 5 years.
| (1) Procedural results | |
|---|---|
| Outcome | Definitions |
| Duration of procedure | In minutes |
| Blood loss | Intra- or postprocedural blood loss in mL |
| Need for blood transfusion | Blood loss more than acceptable, defined by the attending physician |
| Complications | Intraprocedural, early post-procedural (up to 30 days) and late post-procedural (up to 6 months) complications |
| Pain | Several pain scores, ie, visual analog scale, numeric rating scale |
| Febrile morbidity | Determined by the attending physician |
| Need for antibiotics | Signs of infection according to the attending physician |
| Length of hospital stay | In hours or in days |
| Readmission rate | Patients readmitted to the hospital |
| Unscheduled visits | Visits to the outpatients clinic after discharged that were not scheduled |
| Resumption to activities | Normal to various activities |
| (2) Effectiveness | |
| Menorrhagia | Any question about the improvement of menorrhagia |
| Reinterventions | Reinterventions after the first intervention |
| Hemoglobin level | Hemoglobin level in g/dL |
| Volume of fibroids | Measured with echo/MRI |
| Volume of uterus | Measured with echo/MRI |
| (3) Menopause | |
| FSH >10 | Blood sample, in IU/L |
| Wiklund Vasomotor Symptom Subscale score | Rating the severity of 13 menopausal symptoms (0 [none] to 10 [severe]). |
| (4) Health-related quality of life (HRQOL) | |
| Medical Outcome Study Short Form 36 (SF-36) | The SF-36 questionnaire comprises 36 questions summarized in the following 8 domain scores of physical and mental functioning: physical functioning, role functioning physical as well as emotional, social functioning, bodily pain, mental health, vitality and general health perceptions. The 8 domain scores can be summarized in 2 measures: the physical component summary score (PCS) and the mental component summary score (MCS), which are derived using weighted averages of the individual domain scores. The PCS and MCS provide scores ranging from 0–100 (100 indicating the optimal score). |
| Health Utility Index-3 (HUI-3) | HUI3 consists of the following 8 dimensions: vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain. Each question has 5 or 6 possible answers (levels), varying from highly impaired to normal. HUI3 was designed to have full structural independence, meaning that any combination of levels among attributes is logically possible. The 8 questions yield a score ranging from 0 (∼ dead) to 1 (∼ perfect health). |
| Euro Quality of Life 5D (EuroQoL 5) | The EuroQol 5 describes health status in terms of 5 dimensions (ie, mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is described using 3 functional levels (ie, 1 = no problems, 2 = moderate problems, and 3 = severe problems). These functional levels are transformed into a score ranging from 0 to 1.0 (1.0 being the optimal score) by an algorithm that adjusts for age. |
| (5) Pelvic Floor Problems | |
| Urogenital Distress Inventory (UDI) | Questions on urination and pelvic floor problems. The UDI ranges from 0–100. A higher score indicates a less favorable outcome. |
| Incontinence Impact Questionnaire (IIQ) | Comprises of 7 items from the original 30 items and ranges from 0–100. A higher outcome indicates a worse outcome. The IIQ score was only administered to patients who reported complaints of urinary incontinence. |
| Defecation Distress Inventory (DDI) | Questions on defecation and pelvic floor problems. The DDI score ranges from 0–100. A higher DDI score indicates a less favorable outcome |
| (6) Satisfaction | |
| Satisfaction with received treatment | Patients could indicate how satisfied they were with the treatment they received. Options were: very satisfied, satisfied, fairly satisfied, neither satisfied nor unsatisfied, fairly unsatisfied, unsatisfied, or very unsatisfied. |
| Recommendation of received treatment to a friend | A satisfaction score measuring whether patients would recommend the procedure to a friend. |
Intraprocedural results ( Table 5 )
The duration of the procedure was significantly shorter in the UAE group, compared to both hysterectomy (79 vs 95 minutes, P = .009) and myomectomy (59 vs 109 minutes, P < .0001). Intraprocedural blood loss was significantly less in the UAE group than in the hysterectomy group (31 vs 436 mL, P < .00001). There were no significant differences regarding intraprocedural complications between UAE and hysterectomy ( P = .17) and between UAE and myomectomy.
| Outcome | UAE | Surgery (hysterectomy/myomectomy) | Weighted mean difference (95% CI) or odds ratio (95% CI) | |||
|---|---|---|---|---|---|---|
| Study | n | Mean (SD) n (proportion in %) | n | Mean (SD) n (proportion in %) | ||
| Duration of procedure | Hehenkamp et al, 2005 | 81 | 79 (30.5) | 75 | 95.4 (30.9) |
|
| Duration of procedure d | Mara et al, 2008 | 58 | 59.2 (23.1) | 63 | 108.90 (27.7) |
|
| Intraprocedural blood loss | Hehenkamp et al, 2005 | 81 | 30.9 (23.8) | 75 | 436.10 (474.5) |
|
| Intraprocedural complications | Hehenkamp et al, 2005 a | 81 | 7 (8.6%) | 75 | 2 (2.7%) |
|
| Pinto et al, 2003 b | 40 | 10 (25.0%) | 20 | 4 (20.0%) | ||
| Intraprocedural complications d | Mara et al, 2008 c | 58 | 4 (6.9%) | 6 | 5 (7.9%) |
|
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
