Magnetic resonance-guided focused ultrasound (MRgFUS) compared with abdominal hysterectomy for treatment of uterine leiomyomas
To compare women undergoing magnetic resonance-guided focused ultrasound (MRgFUS) to a group of contemporaneously recruited women undergoing total abdominal hysterectomy. Patient demographics, safety parameters, quality of life outcomes and disability measures are reported.
One hundred and nine women were recruited in seven centers for MRgFUS treatment and 83 women who underwent abdominal hysterectomy were recruited in seven separate centers to provide contemporaneous assessment of safety. The adverse-event profile and disability parameters were prospectively assessed. Patients were also screened at baseline and at 1, 3 and 6 months using the SF-36 health survey questionnaire.
There were no life-threatening adverse events in either group. Overall, the number of significant clinical complications and adverse events was lower in women in the MRgFUS group compared to women undergoing hysterectomy. MRgFUS was associated with significantly faster recovery, including resumption of usual activities. At 6 months of follow-up, there were four (4%) treatment failures in the MRgFUS arm. Regarding SF-36 subscale scores, at 6 months there was improvement in all SF-36 subscales for both treatment groups. However, most of the SF-36 subscale scores were significantly better at this stage in the hysterectomy group than in the MRgFUS group. Women undergoing MRgFUS had steady improvement in all parameters throughout the 6-month follow-up period, despite the fact that they continued to have myomatous uteri and menstruation, which at baseline had given them significant symptomatology.
The results of this study show that MRgFUS treatment of uterine leiomyomas leads to clinical improvement with fewer significant clinical complications and adverse events compared to hysterectomy at 6 months' follow-up. Copyright © 2009 ISUOG. Published by John Wiley & Sons, Ltd.
Leiomyomas (fibroids or myomas) are benign myometrial neoplasms and represent the primary indication for hysterectomy in the USA1, 2. Alternatives to hysterectomy have been introduced to treat these tumors with decreased perioperative morbidity but with a high rate of leiomyoma recurrence3. Furthermore, laparoscopic and hysteroscopic myomectomies are applicable only for fibroids in particular locations and of a certain size, and uterine artery embolization (UAE) is associated with a significant incidence of post-embolization syndrome4–8.
A second important factor driving innovation in fibroid therapies is the magnitude of healthcare costs. Recent reports suggest that healthcare costs are higher for women with fibroids than for unaffected women and disability costs are substantial, probably because surgical therapy is the major treatment option9–11. Furthermore hysterectomy accounts for most of the costs, recently estimated to exceed $ 2.1 billion annually in the USA12.
Magnetic resonance-guided focused ultrasound (MRgFUS) surgery was approved by the United States Food and Drug Administration (FDA) as a non-invasive treatment for uterine leiomyomas in 200413. MRgFUS surgery for leiomyomas shows treatment efficacy for at least 24 months following treatment in both single-institution series and multicenter collaborative trials3, 14–21. Furthermore, MRgFUS represents a cost-effective treatment option for symptomatic leiomyomas in some healthcare systems22.
While the results of the cohort of women who underwent MRgFUS in the trial that led to FDA approval of this technique have been reported, the outcomes for contemporaneously-recruited women undergoing hysterectomy have never been published14. In this study we examined women undergoing MRgFUS or abdominal hysterectomy and compared the incidence of significant clinical complications and adverse events, as well as quality of life outcomes, during the first 6 months following treatment.
Study participants and enrollment
We conducted the study in 14 medical centers in the United States, Israel, the UK and Germany. Premenopausal women with symptomatic uterine fibroids meeting the same enrollment criteria were recruited for MRgFUS or hysterectomy. The institutional review board or ethics committee of each institution approved the study protocol.
Results of the MRgFUS arm of this trial have been reported in detail14. Briefly, 109 women underwent treatment and were recruited at seven sites: Brigham and Women's Hospital, Boston, MA (n = 24); Sheba Medical Center, Tel-Hashomer, Israel (n = 20); St. Mary's Hospital, London (n = 19); Johns Hopkins Hospital, Baltimore, MD (n = 16); Mayo Clinic, Rochester, MN (n = 12); Charité Hospital, Berlin (n = 9); and Hadassah Medical Center, Jerusalem, Israel (n = 9)14.
To avoid biased referral between groups, women undergoing hysterectomy for symptomatic fibroids were recruited from different medical centers: University of Texas Southwestern in Dallas, TX (n = 35); Ha'Emek Medical Center, Afula, Israel (n = 20); William Beaumont Hospital, Royal Oak, MI (n = 16); John Radcliffe Hospital, Oxford, UK (n = 6); Texas Institute of Clinical Research, Ft Worth, TX (n = 3); Wayne State University, Detroit, MI (n = 2); and Frauenklinik Universitätsklinikum Benjamin Franklin, Berlin (n = 1). Thus, the group of women that underwent hysterectomy comprised 83 women.
All the women were at least 18 years old and did not want children in the future. Exclusion criteria included women with a uterus larger than 24 weeks' gestational size, a hematocrit < 25%, a positive pregnancy test, any contraindication to surgery or contraindication to magnetic resonance imaging (MRI) such as aneurysm clip, deep brain stimulator, cardiac pacemaker or implantable defibrillator. All patients had pretreatment MRI to determine imaging eligibility.
Symptom assessment was carried out using the Uterine Fibroid Symptom and Quality-of-Life Questionnaire (UFS-QOL), a validated disease-specific instrument23. The UFS-QOL consists of an eight-item Symptom Severity Scale (SSS) and six dimensions of health-related quality of life (HRQL). A raw SSS of at least 21 out of a possible 40 points was required for entry into the study so as to be certain that the women had significant leiomyoma symptoms at enrollment. In addition, the Short Form-36 health survey questionnaire (Medical Outcomes Trust) was administered to provide an additional method of assessment of HRQL at baseline, and at 1, 3 and 6 months of follow-up24.
Technical details for MRgFUS treatment have been published elsewhere3, 14, 21, 25, 26. Briefly, the MRgFUS system (ExAblate 2000, InSightec, Haifa, Israel) works in conjunction with a standard 1.5-Tesla MRI system (GE Medical Systems, Milwaukee, WI, USA). The focused ultrasound treatment transducer is incorporated in the patient table and thus when a subject is in the prone position her abdomen is directly above the FUS transducer. Women were treated in the MRI suite as outpatients with light intravenous conscious sedation allowing communication with the treating physician. Prophylactic antibiosis was not used before MRgFUS.
T2-weighted MR images acquired immediately before treatment were used for treatment planning. For women undergoing MRgFUS, leiomyomas were selected for treatment based on size, symptomatology and the accessibility of the fibroid to the focused ultrasound beam. Most of the women were treated for a single fibroid, however, treatment of multiple leiomyomas was allowed within the protocol guidelines. In consultation with the FDA, the guidelines for this initial protocol included a margin from the region of treatment of 1.5 cm that was maintained at both the serosal and mucosal borders of the uterus. Coagulation volume was limited up to 100 cm3 per myoma and 150 cm3 per treatment. Total treatment time was limited to 3 h.
Abdominal hysterectomy was performed per the physicians' standard of care and the method was not specified. Prophylactic antibiotics were administered preoperatively in all hysterectomy cases. Women in both treatment groups were then followed to ascertain significant clinical complications and adverse events.
Strict reporting of adverse events and, particularly, serious adverse events (SAE) was in compliance with the Standard Code of Federal Regulations. In fact, because of the novel nature of this treatment, we elected to adhere to a stricter level of reporting for adverse events such as those used during pharmaceutical trials rather than device-trial regulations14. Prior studies of complications following hysterectomy were used as a model for the design of this study but updated to take into account current shorter lengths of stay27. Significant clinical complications (SCC) were defined as fever > 38 °C on any 2 post-treatment days, blood transfusion, unintended major surgical procedure, discharge to a rehabilitation facility, discharge with an appliance such as a drain or urinary catheter, outpatient interventional treatment, rehospitalization, life-threatening event or death within 42 days of treatment.
All data were entered into an internet-based database designed by ClickFind (Bryan, TX, USA). The data entry application has a robust audit trail and all data were verified and double-checked before analysis. Data were then exported to an Excel spreadsheet (Microsoft, Redmond, WA, USA) for transfer to SAS statistical software (SAS Institute, Cary, NC, USA) for analysis.
Comparisons between the two treatment groups were conducted using chi-square tests for categorical variables or t-tests for continuous variables. For the significant clinical complications analysis we used Fisher's exact test owing to the low counts per cell. For categorical variables, the frequency and percentage of patients within each category are reported; for continuous variables, the mean and standard deviation are reported. Race was analyzed as a discrete variable. All tests were two-tailed and P < 0.05 was considered statistically significant in all statistical analyses.
Baseline characteristics for women from both treatment groups are presented in Table 1. A total of 192 women were included in the study, of whom 109 underwent MRgFUS and 83 underwent total abdominal hysterectomy. The study population was typical of premenopausal women with symptomatic fibroids, with most women being in their forties and having an increased body mass index (BMI). Women in the hysterectomy group were less likely to be Caucasian (54 vs. 80%; P < 0.001) and had higher BMI on average (29.9 ± 6.0 vs. 25.8 ± 5.2 kg/m2; P = 0.001).
Table 1. Baseline patient characteristics of the treatment groups
|Age (years)||44.8 ± 4.9||44.4 ± 5.6||NS|
|Body mass index (kg/m2)||25.8 ± 5.2||29.9 ± 6.0||< 0.001|
|Race|| || ||0.001|
| Caucasian||87 (80)||45 (54)|| |
| Black||12 (11)||28 (34)|| |
| Asian||3 (3)||2 (2)|| |
| Other||7 (6)||8 (10)|| |
|UFS-QOL baseline scores|| || || |
| Symptom severity||61.7 ± 15.2||69.6 ± 8.1||0.001|
| Total HRQL score||47.0 ± 18.6||38.4 ± 23.8||0.008|
|Medication use for fibroids|| || || |
| None||76 (70)||35 (42)||< 0.0001|
| NSAIDs||18 (17)||24 (29)||0.04|
| Depo-Provera||1 (1)||8 (10)||0.005|
| Oral progesterone||5 (5)||12 (14)||0.02|
|SF-36 baseline scores|| || || |
| Physical functioning||72.8 ± 23.9||56.1 ± 30.0||< 0.0001|
| Physical role||45.2 ± 41.5||33.4 ± 39.1||0.048|
| Bodily pain||52.0 ± 22.4||40.1 ± 26.9||0.0001|
| General health||66.0 ± 19.7||58.1 ± 22.8||0.001|
| Vitality||41.3 ± 20.8||37.2 ± 20.2||NS|
| Social functioning||61.5 ± 27.6||54.2 ± 29.2||NS|
| Emotional role||57.8 ± 40.2||39.4 ± 42.9||< 0.0001|
| Mental health||63.0 ± 16.9||55.1 ± 21.6||0.007|
|Disability assessment baseline scores|| || || |
| Days of work missed||0.8 ± 1.9||3.9 ± 6.3||0.001|
| Days late for work||1.1 ± 3.3||2.0 ± 5.1||NS|
| Days spent in bed||4.0 ± 6.6||6.5 ± 8.5||< 0.0001|
| Days kept from normal activities||1.2 ± 2.3||5.3 ± 8.1||0.03|
Both groups had significant levels of fibroid symptomatology and health-related impairment of quality of life at baseline, although the hysterectomy group had higher levels of symptoms (Table 1). Additionally, women undergoing hysterectomy had significantly worse function on several subsections of the SF-36 questionnaire including physical functioning, bodily pain, vitality, emotional role and mental health (Table 1).
The safety profile of MRgFUS compared favorably with that of hysterectomy, with fewer SCCs occurring in the MRgFUS arm compared to the hysterectomy arm (14 events vs. 33 events; P < 0.0001) (Table 2). The rate of febrile morbidity (fever > 38 °C on any two post-treatment days) was substantially lower in the MRgFUS group than in the hysterectomy group (3 events vs. 12 events; P = 0.005). Unintended surgical procedures related to treatment (removal of foreign body from the bladder, surgical repair of hernia, revision of enterotomy and surgical repair of an iatrogenic colonic lesion) were performed in four (5%) women from the hysterectomy group; no unintended surgical procedures were required in women from the MRgFUS group. There were no deaths or life-threatening events in either group (Table 2).
Table 2. Summary of significant clinical complication events (SCC) for the magnetic resonance-guided focused ultrasound (MRgFUS) and hysterectomy groups
|Total number of SCC events||14||33||< 0.0001|
|Fever > 38 °C on any 2 post-treatment days||3||12||0.005|
|Unintended surgical procedures related to treatment||0||4||NS|
|Discharge with appliance*||0||1||NS|
|Rehospitalization with a duration > 24 h||8||8||NS|
At least one adverse event was reported by 88 (81%) women from the MRgFUS group and by 82 (99%) women from the hysterectomy group (P < 0.0001) (Table 3). Adverse events related to pain or discomfort, the gastrointestinal tract, dermatological conditions and nervous system were reported significantly less frequently by women from the MRgFUS group compared to women from the surgical group (Table 3). At 6 months' follow-up nine (8%) SAEs were recorded in the MRgFUS group and eight (10%) SAEs were recorded in the hysterectomy group. Of note, one of the SAEs reported in the MRgFUS group was a pre-existing brain tumor and one of the SAEs reported in the hysterectomy group was ventricular ectopy. Both events were reported as being unlikely to be related to treatment.
Table 3. Summary of adverse and serious adverse events related to treatment procedures in both patient groups
|At least one adverse event||88 (81)||82 (99)||< 0.0001|
|Adverse event*|| || || |
| Pain/discomfort||68 (62)||79 (95)||< 0.0001|
| Gynecological||30 (28)||23 (28)||NS|
| Urinary tract||26 (24)||16 (19)||NS|
| Gastrointestinal tract||22 (20)||56 (67)||< 0.0001|
| Dermatological system||17 (16)||29 (35)||0.002|
| Nervous system||8 (7)||17 (20)||0.007|
| Cardiovascular system||3 (3)||7 (8)||NS|
| Respiratory tract||0 (0)||2 (2)||NS|
| Systemic events†||19 (17)||17 (20)||NS|
| Other||0 (0)||6 (7)||0.004|
|Serious adverse event||9 (8)||8 (10)||NS|
The most serious complication following MRgFUS was the development of a significant but reversible sciatic nerve palsy in one patient14. Assessment of far-field energy absorption has subsequently been incorporated into treatment planning, and no further similar complications have been reported3. Because the MRgFUS treatment is captured electronically and can be reviewed at a later time, thermal map analysis was able to determine that the nerve was not injured by direct sonication. Instead the pelvic bone received an amount of energy that would not have directly damaged soft tissue but which led to nerve injury by passive heat transfer beyond the focal volume. The subject also had significant pain in the regions innervated by the sciatic nerve (buttocks and posterior leg) during the procedure, which she did not communicate to the treating physician.
Four subjects in the MRgFUS arm (4%) did not complete the study protocol in order to pursue surgical or interventional treatment (three hysterectomies and one uterine artery embolization) for continued or recurrent leiomyoma-related symptoms during the first 6 months of follow-up. Surgical or interventional treatment occurred within a mean (range) of 10 (8–16) weeks following MRgFUS treatment.
The results of the SF-36 questionnaire at 1, 3 and 6 months of follow-up are shown in Table 4. At 6 months there was improvement in all SF-36 subscales for both treatment groups. At 1 month women from the MRgFUS group had significantly greater improvement scores than the surgery group for the physical-function, physical-role and social-function components. At 3 and 6 months these differences were not significant (Table 4).
Table 4. SF-36 health survey questionnaire and disability assessment scores for magnetic resonance-guided focused ultrasound (MRgFUS) and hysterectomy groups at 1, 3 and 6 months post-treatment
| Physical functioning||80.3 ± 19.9||58.0 ± 20.1||< 0.0001||81.7 ± 18.4||77.8 ± 22.6||NS||82.5 ± 15.9||86.9 ± 19.2||NS|
| Physical role||55.3 ± 34.8||20.0 ± 36.0||< 0.0001||67.4 ± 33.5||66.1 ± 41.0||NS||68.3 ± 31.0||80.0 ± 37.6||0.05|
| Bodily pain||64.6 ± 21.9||49.3 ± 22.6||< 0.0001||67.2 ± 20.1||77.6 ± 24.3||0.009||69.1 ± 19.2||79.5 ± 22.6||0.004|
| General health||68.2 ± 15.9||71.3 ± 15.9||NS||69.7 ± 15.1||74.4 ± 18.4||NS||69.3 ± 15.1||75.3 ± 18.4||0.04|
| Vitality||53.9 ± 19.9||44.5 ± 20.1||0.004||58.0 ± 16.7||63.1 ± 20.9||NS||59.1 ± 15.1||65.6 ± 20.1||0.04|
| Social functioning||74.9 ± 23.9||56.3 ± 24.3||< 0.0001||78.7 ± 17.6||83.2 ± 21.8||NS||79.5 ± 18.4||84.8 ± 22.6||NS|
| Emotional role||65.4 ± 40.8||48.7 ± 41.0||0.01||81.3 ± 31.0||71.9 ± 36.8||NS||75.0 ± 30.1||78.1 ± 36.8||NS|
| Mental health||71.8 ± 15.9||74.1 ± 15.9||NS||73.3 ± 12.5||79.1 ± 15.9||0.02||73.3 ± 11.7||79.6 ± 15.1||0.008|
|Disability assessment scores|
| Lost work days||1.2 ± 0.8||19.2 ± 0.9||< 0.0001||0.9 ± 0.6||3.9 ± 0.8||0.005||0.2 ± 0.6||1.7 ± 0.7||NS|
| Days late for work||0.6 ± 0.5||2.1 ± 0.7||NS||0.2 ± 0.1||0.2 ± 0.1||NS||0.1 ± 0.0||0.0 ± 0.0||NS|
| Days spent in bed||1.3 ± 0.7||9.9 ± 0.8||< 0.0001||0.9 ± 0.4||1.8 ± 0.5||NS||0.5 ± 0.2||0.5 ± 0.2||NS|
| Days kept from usual activities||2.7 ± 0.9||17.4 ± 1.1||< 0.0001||1.8 ± 0.7||4.3 ± 0.9||0.04||1.4 ± 0.5||1.7 ± 0.6||NS|
The differences in recovery trajectory are supported by data on return to normal activities. At 1 month, women undergoing MRgFUS reported 1.2 lost working days and 2.7 days kept from usual activities compared with 19.2 and 17.4, respectively, for women undergoing hysterectomy (P < 0.0001). Women undergoing MRgFUS had a steady improvement in all parameters throughout the follow-up period, despite the fact that they continued to have a myomatous uterus and menstruation, which at baseline had given them significant symptomatology (Table 4).
The results of this study show that MRgFUS treatment of uterine leiomyomas leads to clinical improvement with fewer significant clinical complications and adverse events compared to hysterectomy during the 6-month follow-up period. Furthermore, MRgFUS treatment was associated with significantly faster recovery, including resumption of usual activities. The level of morbidity seen in the hysterectomy group was identical to that seen in prior studies27.
A limitation of the current study is that the two groups of patients differed in important parameters. The women undergoing hysterectomy had increased BMI, were less likely to be Caucasian, had higher symptom severity scores and had an increased use of medication for fibroid-related symptoms. All these factors are consistent with more severe disease, which is consistent with women who have elected to undergo definitive therapy28–32. Furthermore, the patients' preferences for either treatment were not evaluated.
Comparing a novel treatment approach for uterine leiomyomas (MRgFUS) with a standard treatment (hysterectomy) may result in potential data-collection bias through a more complete documentation of significant clinical complications and adverse events occurring in the novel-treatment group. Furthermore, comparison of the non-invasive MRgFUS with other less invasive alternatives to hysterectomy will be necessary to inform future decision-making.
All patients from the MRgFUS group were treated as outpatients. In addition to these obvious direct medical benefits, the economic impact is also likely to be significant, with mean time of return to work of approximately 1 day, compared to approximately 14 days following UAE and approximately 6 weeks following abdominal surgery8, 14, 33, 34.
The results of this study show that the use of MRgFUS treatment as an alternative to hysterectomy for symptomatic leiomyomas requires certain trade-offs. The advantages of MRgFUS—an outpatient procedure with fewer significant clinical complications and adverse events and faster recovery—have to be weighed against the risk of treatment failure requiring possible subsequent surgery. Nevertheless MRgFUS, by monitoring temperature and imaging beyond the targeted organ, gives the surgeon more information than he/she has with vision and tactile sensation alone. Treatment data are also captured to allow retrospective analysis in a novel way35.
The choice of hysterectomy for the control group also limits comparisons. However, at the outset of the study, although both myomectomy and uterine artery embolization were discussed as possible controls, the FDA felt that women undergoing hysterectomy would provide the appropriate comparison for safety parameters for this novel treatment. A randomized clinical trial comparing MRgFUS to another alternative to hysterectomy is clearly required to ascertain how MRgFUS fits into fibroid treatment and to identify ideal candidates for this innovative treatment option. Optimizing efficacy and decreasing future interventions are important goals for all myoma treatments3. Nonetheless, having a safe and effective non-invasive treatment for women with uterine leiomyomas represents an important development in the field.
The MRgFUS Group:
Jaron Rabinovici (Sheba Medical Center, Tel-Hashomer, Israel); Jonathan Hindley, Wladyslaw M. Gedroyc (St. Mary's Hospital, London, United Kingdom); Hyun S. Kim, Jean-François H. Geshwind (Johns Hopkins School of Medicine, Baltimore, MD, USA); Gina Hesley, Bobbie Gostout (Mayo Clinic, Rochester, MN, USA); Tilman Ehrenstein, Susanne Hengst (Charité Medical Center and Humboldt University, Berlin, Germany); Simcha Yagel, John M. Gomeri (Hadassah Medical Center, Jerusalem, Israel); Nathan McDannold, Kullervo Hynynen, Ferenc A. Jolesz, Louise Greenberg (Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA); David L. Hemsell (UT Southwestern Medical Center, USA); D. Alan Johns (Texas Institute of Clinical Research, Forth Worth, TX, USA); Stephen Kennedy (University of Oxford, Oxford, United Kingdom); Shabtai Romano (Ha'Emek Medical Center, Afula, Israel); Gene McNeely (Wayne State University, Detroit, MI, USA); Craig Hartrick (William Beaumont Hospital, Royal Oak, MI, USA); Christiane Richter-Ehrenstein (Universitaetsklinkum Benjamin Franklin, Berlin, Germany); Matthias David (Charité Medical Center, Berlin, Germany).