Clinical outcomes of magnetic resonance-guided focused ultrasound surgery for uterine myomas: 24-month follow-up




To assess the volume reduction ratio, symptom improvement and reintervention rate following magnetic resonance-guided focused ultrasound surgery (MRgFUS) for uterine myomas.


A total of 91 Japanese women with symptomatic myomas underwent MRgFUS between June 2004 and June 2008 using the ExAblate 2000 system. The volume change ratio was calculated at 6, 12 and 24 months following MRgFUS based on T2-weighted magnetic resonance images. The symptom severity score (SSS) was examined before and after the treatment (at 3, 6, 12 and 24 months). Additional post-MRgFUS treatments, such as hysterectomy, myomectomy, uterine artery embolization or repeat MRgFUS, were recorded and the reinterventional treatment rates were compared according to the signal intensity of pretreatment T2-weighted magnetic resonance images of the myomas.


The mean volume change ratios of low- and intermediate-intensity (Type 1/2) myomas were −36.5% 6 months post-procedure and −39.5% 24 months post-procedure. The mean ± SD SSS value for patients with Type 1/2 myomas before MRgFUS was 35.1 ± 21.0, and the values diminished significantly during the 24-month follow-up period to a mean value of around 15.0. High-intensity (Type 3) myomas were not observed to have decreased in size 6 months after MRgFUS. Of the 45 Type 1/2 myoma patients with complete follow-up, seven required reinterventional treatment within 24 months. The reintervention rates were 14.0% for Type 1/2 patients and 21.6% for Type 3 patients at 24 months post-treatment.


Moderate volume reductions of Type 1/2 myomas were noted following MRgFUS, and the reduction in SSS values and the relatively low reintervention rates observed are encouraging. We found MRgFUS to be an appropriate treatment method for Type 1/2 uterine myomas. Copyright © 2009 ISUOG. Published by John Wiley & Sons, Ltd.


Many minimally invasive treatments for uterine myoma are currently available, including hormonal therapy, endoscopic surgery, uterine artery embolization (UAE)1–7 and magnetic resonance-guided focused ultrasound surgery (MRgFUS)8–19. MRgFUS involves generation of a high-intensity ultrasound beam, which can be focused at the desired location. All processes are carried out entirely under magnetic resonance imaging (MRI) guidance, and the operator can concurrently monitor and control the precise location and temperature elevation at the focal point. MRgFUS is a minimally invasive treatment for symptomatic myomas. It has a low risk for complications, and avoids hospitalizations and radiation exposure. Moreover, many studies have shown that MRgFUS significantly improves clinical symptoms in 70–80% of women treated for uterine myomas8, 9.

We have previously reported early to mid-term clinical effects of MRgFUS for uterine myomas10, 11, 15. In this study, we sought to estimate the long-term effects based on 24-month follow-up because we have observed that some ablated myomas begin to regrow, especially after a longer follow-up period. The study design included a prospective evaluation of symptom improvement and the volume change ratio of treated uterine myomas, and a retrospective evaluation of the reintervention rate up to 24 months after MRgFUS.


After receiving approval from the local ethics committee of Shinsuma General Hospital for MRgFUS treatment of symptomatic uterine myomas, we treated patients with uterine myoma with an MRgFUS system using a conventional diagnostic 1.5-T MRI scanner (Signa, GE Medical Systems, Milwaukee, WI, USA) integrated with a focused ultrasound delivery system (ExAblate 2000 version 4.1 or version 4.2, InSightec, Tirat Carmel, Israel)10, 12.

Patients who would otherwise have been offered conventional surgery were considered for eligibility. Excluded from this treatment were pregnant and lactating women, dialysis patients, patients receiving anticoagulation therapy, and patients with active pelvic infection, hemolytic anemia, a history of cerebrovascular disease, unstable cardiac status, pelvic mass outside the uterus, standard contraindications for MRI, known intolerance to MRI contrast agents, intolerance of the required prolonged stationary position during the treatment or extensive abdominal scars directly anterior to the area of treatment.

Before MRgFUS, all patients were examined using standard T2- and T1-weighted MRI and contrast-enhanced T1-weighted MRI after intravenous injection of a gadolinium-based contrast agent, gadopentetate dimeglumine (Magnescope®, Terumo Corporation, Tokyo, Japan), at 0.1 mmol per kg body weight (Figure 1). These images were used for the diagnosis of myomas and to determine whether a clear pathway to the lesion for the ultrasound beam was present. We classified the myomas into three types on the basis of the signal intensity of pretreatment T2-weighted images: Type 1, a very low-intensity image comparable to that of skeletal muscle; Type 2, an image intensity lower than that of the myometrium and higher than that of the skeletal muscle; and Type 3, an image intensity equal to or higher than that of the myometrium10.

Figure 1.

Magnetic resonance images before and after magnetic resonance-guided focused ultrasound surgery (MRgFUS) in a 44-year-old woman with a Type 2 myoma and hypermenorrhea. (a) T2-weighted image before MRgFUS. Myoma volume was 228.9 cm3 and the treated area ratio was 66.8%. Targeted myoma is indicated by arrowheads. (b) T2-weighted image 24 months after MRgFUS, showing 81.1% volume reduction from pretreatment volume. (c) Enhanced T1-weighted image 24 months after MRgFUS. The center of the myoma was not perfused. EM, endometrium; NP, non-perfused area.

The study group comprised 91 premenopausal Japanese women with 141 symptomatic uterine myomas who underwent MRgFUS treatment between June 2004 and June 2008 (Table 1). Patients who had received gonadotropin-releasing hormone analog at least 6 months before MRgFUS were excluded from this study. Written informed consent was obtained from all patients. The study included 11 Type 3 myoma patients treated in our institution at an initial stage in the introduction of MRgFUS. However, we no longer treat patients with Type 3 myomas because of poor response to MRgFUS10, 11.

Table 1. Myoma volumes and treated area ratios of the study population
ParameterType 1Type 2Type 3
  • Data expressed as mean ± SD (range), n or n (%).

  • *

    P < 0.05 vs. Type 1 and vs. Type 2;

  • P < 0.05 vs. Type 1.

Number of patients255511
Number of myomas547512
Myoma volume (cm3)129.0 ± 145.2 (14.4–724.1)211.7 ± 173.0 (14.9–757.7)180.8 ± 154.5 (34.6–501.9)
Non-perfused volume (cm3)70.7 ± 79.4 (0.3–318.3)119.4 ± 114.1 (1.4–535.4)67.3 ± 69.6 (1.1–244.0)
Treated area ratio (%)54.7 ± 22.654.0 ± 18.736.2 ± 28.6*
Number of myomas with33 (61.1)48 (64.0)2 (16.7)
 treated area ratio > 50%   

The therapeutic ultrasound beam was delivered to a targeted area where the acoustic energy was converted to thermal energy, raising the local temperature to 60–90 °C in a matter of seconds and producing thermal coagulation of the focal mass. The focal volume was smaller than that of a cigarette and therefore patients with large or multiple myomas either received a second procedure (11 patients in this study) or were excluded from MRgFUS treatment. Myomas 4–10 cm in diameter were deemed to be the appropriate size for treatment. The treated area was mapped on the system software and the operator repeatedly sonicated until an adequate volume of myoma was ablated. In view of safety considerations, the operator monitored the location and temperature rise of the focal point and patients were provided with a ‘stop button’ with which they could instantly stop the sonication.

Immediately after the treatment, the post-treatment non-perfused volume (NPV) was calculated using a sum of slices method; the specific NPV is outlined in each sequential slice and the volumes are summed10. We defined the treated area ratio as the percentage of NPV to targeted myoma volume. The patients were asked to return for repeat MRI examinations 6, 12 and 24 months after MRgFUS treatment. The volume change ratio was calculated by the same method based on T2-weighted images during these examinations. Myomas that were previously treated more than once with MRgFUS were excluded from the volume change results because the treatment intervals varied considerably (from 1 to 9 months).

We employed the symptom severity score (SSS) analysis method20, which was designed to assess symptom severity of uterine myomas. All eligible patients completed an eight-item questionnaire before and after the treatment, and at 3-, 6-, 12- and 24-month intervals for SSS assessment. The questionnaires were collected by postal mail or when patients returned for outpatient examination. Responses for each item were scored from 1 (no symptoms) to 5 (major symptoms). The sum of the scores was transformed into a 0–100 scale for comparison. Higher scores indicated worse symptoms. The control score for Japanese women was calculated from the women (n = 46) who came for medical check-up and in whom no gynecological problems were confirmed by internal examination and transvaginal sonography.

Measured values were expressed as mean ± SD and/or median and range. The statistical significance was evaluated using one-way ANOVA for the changes in myoma volume, treated area ratio and SSS values, and Student's t-test was used for comparing the groups defined by myoma type. P < 0.05 was considered statistically significant.



This study included 91 patients with 141 uterine myomas. Mean ± SD age of the patients was 40.4 ± 4.6 years. Sixty-four patients were treated for single myomas and 27 were treated for two or more myomas simultaneously. Mean sonication time was 2 h 50 min ± 1 h 36 min. Table 1 shows myoma volumes and treated area ratios by myoma type. The treated area ratio of Type 3 myomas was significantly smaller than that of Type 1 and Type 2 myomas.

Volume change ratio

Figure 2 shows the volume change ratio of Type 1/2 myomas 6–24 months after MRgFUS. Compared with pretreatment volume, the mean volume change ratio for Type 1/2 myomas significantly decreased from −36.5% 6 months post-treatment to −39.5% 24 months post-treatment (P < 0.001). When the two study groups were considered separately, a similar tendency was observed for Type 1 and Type 2 myomas (data not shown). All well ablated Type 1/2 myomas (treated area ratio > 50%) decreased in volume, with a mean volume change ratio of around −50% at 24 months' follow-up. Type 3 myomas did not show a reduction in size 6 months after MRgFUS (−9.1 ± 44.8%, n = 9), but were found to be smaller at 12 and 24 months' follow-up. This may be explained by the higher reintervention rate in this group.

Figure 2.

Volume change ratios of myomas during follow-up. Type 1/2 myoma volumes (equation image) decreased significantly from pretreatment volumes. Myomas that had been previously treated more than once with magnetic resonance-guided focused ultrasound surgery (MRgFUS) were excluded from this analysis. Type 1/2 myomas that were well ablated (treated area ratio > 50%) showed a greater mean reduction in volume (equation image). Type 3 myomas did not show a reduction in size 6 months after MRgFUS, but were smaller at 12 and 24 months' follow-up (equation image). Whiskers show the SD for each group at each follow-up point. *P < 0.001 vs. pretreatment volume (one-way ANOVA).

Improvement of clinical symptoms

Subjective symptoms were evaluated using the SSS questionnaire, in which a higher score indicates worse symptoms20. The mean SSS value of Type 1/2 patients before treatment was 35.1 ± 21.0 (n = 72), which diminished significantly during follow-up for 3–24 months after treatment (Figure 3). The control score from healthy Japanese women was 12.1 ± 11.8 (n = 46). The SSS values for Type 3 patients could not be evaluated because of the high reintervention rate and low response rate.

Figure 3.

Improvement of symptom severity score (SSS) during follow-up for patients with Type 1/2 myomas. The number of women who returned the questionnaire at each follow-up point is shown below the graph and whiskers show the SD. The normal control score from healthy Japanese women was 12.1 ± 11.8 (n = 46). *P < 0.001 vs. pretreatment score (one-way ANOVA).

Reintervention rate

Of 80 Type 1/2 patients, 12 required reinterventional treatment because of unsuccessful outcomes after MRgFUS (median follow-up period, 33.9 (range, 6–54) months). We excluded patients who underwent surgical treatment for indications other than myomas. For example, one patient underwent an abdominal hysterectomy concomitant with colonic cancer therapy and another underwent transcervical resection (TCR) for endometrial hyperplasia; these cases were included in the ‘lost to follow-up’ cell. One patient underwent hysterectomy, five patients underwent myomectomy, one patient underwent TCR of myomas and five patients underwent repeated MRgFUS. Four of 80 patients achieved pregnancy (two live term birth and two first-trimester miscarriage). One patient became menopausal within this period; she was 44 years old (Table 2).

Table 2. Follow-up results of patients with Type 1/2 myomas
ParameterFollow-up time (months)
361224> 24
  • *

    Values in this row decrease because women joining in the final 18 months of the study could not be followed up for the full period. Data given as n. MRgFUS, magnetic resonance-guided focused ultrasound surgery.

Number of patients*8080745757
 Hysterectomy   1 
 Myomectomy 11 3
 Transcervical resection   1 
 Repeat MRgFUS   32
Pregnancy1 21 
Menopause  1  
Lost to follow-up46912 
Number evaluated76746545 

The reintervention rates are presented in Table 3 according to myoma type. A greater proportion of patients with Type 3 myomas received reinterventional treatment than those with Type 1/2 myomas. Two of 11 Type 3 patients required reinterventional therapy (median follow-up period, 46.4 (range, 40–53) months).

Table 3. Reintervention rates at 6, 12 and 24 months' follow-up, by myoma type
Myoma typeReintervention rate (%)
6 months12 months24 months
  1. Rates calculated as cumulative across each time period to take into account loss to follow-up.

Type 1/21.42.914.0
Type 39.121.621.6


MRgFUS was introduced as a minimally invasive treatment for symptomatic uterine myomas12. Myoma tissue is destroyed via thermal coagulative necrosis by focused ultrasound energy, leaving normal uterine muscle undamaged12. The outcome of MRgFUS should be comparable to that of a conventional myomectomy, UAE, laser ablation or cryotherapy. However, no prospective comparative study has been reported to date. Although our study is a single-center experience, we would like to compare clinical outcomes with other methods in order to determine the relative value of MRgFUS treatment.

We noted that Type 3 myomas were not amenable to MRgFUS treatment; both the treated area ratio and the volume reduction ratio for Type 3 myomas were statistically inferior to those achieved for Type 1 and Type 2 myomas in our previous mid-term follow-up study10, 11. Lénárd et al. reported that low signal intensities on pretreatment T2-weighted magnetic resonance images and NPV immediately after MRgFUS independently predict myoma volume reduction at a 12-month follow-up examination13. Moreover, in this study, we calculated the reintervention rate, which was high for Type 3 patients.

The most common and conventional uterus-preserving therapy for myoma is a myomectomy, which allows the operator to remove all palpable myomas during surgery. UAE and MRgFUS are not intended to remove myomas. These procedures destroy myoma tissue and induce reduction of their volume leading to symptom improvement. However, myomas persist in most cases. The volume change ratios after UAE have been reported to be −31.7 to −58.7% 6 months post-procedure1, −33.9 to −43.2% 4 months post-procedure2 and −60% 2 years post-procedure3. These ratios were somewhat higher than that of MRgFUS in our study14, 16. Well ablated myomas shrink more than poorly ablated myomas10, 18. In order to achieve larger volume reduction following MRgFUS, one should make an effort to achieve a large treated area ratio in the treatment stage.

We examined SSS before and after treatment with MRgFUS. Our cases demonstrated relatively low scores before treatment compared with those reported in previous European and American studies4, 8, 17, 18, 20. However, a normal control group of a sample Japanese population was found to have a lower SSS than Europeans and Americans, and SSS change over time was comparable to that of UAE reports4. SSS dropped immediately after MRgFUS and remained low for at least 24 months after treatment18. Uterine myoma is a non-malignant neoplasm and the goal of treatment is symptom relief. At this point, the symptom improvement after MRgFUS is encouraging, although longer-term follow-up is required.

The reported reintervention rates following myomectomy vary from 3.2% (2-year follow-up)1 to 23.5% (5-year follow-up)21, and from 6.9%–23.5% (2-year follow-up)1, 3 to 14.4% (3-year follow-up)4 following UAE. It has been reported in European and American studies that 28% of patients sought alternative treatment within 12 months following MRgFUS8, 17. Morita et al. reported an 8% reintervention rate at 12 months that included treatment by medication16. In our study, Type 1/2 patients were found to have a reintervention rate of 2.9% at 12 months and 14.0% at 24 months. Furthermore, our study suggested that reintervention rates were related to the myoma type. For precise comparison, reintervention rates should be based on the same condition.

Stewart et al. reported 2-year follow-up results of SSS, shrinkage of total myoma volume and the probability of undergoing an additional myoma treatment following MRgFUS treatment18. With a highly selected group of patients (Type 1/2 myomas), our study presents moderate volume reduction of targeted myomas, moderate improvement in myoma-related symptoms and relatively low reintervention rates. Taking these points into consideration, our results correspond approximately with those obtained by Stewart et al.

Incompletely infarcted myomas begin regrowing after UAE2, 5. Therefore, the clinical outcome following any treatment is likely to depend upon whether the myoma is completely infarcted. As the initial aim of MRgFUS treatment is to ablate the interior of the myoma, the peripheral portion of the targeted myoma remains viable. For this reason, it might be suspected that the volume reduction ratio and reintervention rate of MRgFUS treatment would be inferior to those of UAE with long-term follow-up. In our study, however, some myomas without a large treated area ratio gradually shrank over a long period of time. UAE induces volume reduction of myomas through ischemic necrosis, whereas MRgFUS induces coagulative necrosis through thermal ablation. The cells adjacent to the focal point increase in temperature to some extent, but not above the threshold of heat necrosis. However, this might result in apoptosis of these cells22. More specifically, ablated vessels will also interrupt the blood supply, resulting in tissue infarction at the end of the artery23.

The MRgFUS procedure did not lead to a requirement for emergency hysterectomies or life-threatening adverse effects. Ovarian function following MRgFUS should be evaluated. Only one patient reached menopause within the follow-up period in our study. This small number does not indicate ovarian dysfunction secondary to MRgFUS treatment. We propose that MRgFUS is a safe treatment method compared with myomectomy or UAE4–7, 19, 24.

In conclusion, we evaluated volume reduction ratio, symptom improvement and reintervention rate up to 24 months following MRgFUS. The reintervention rate was low (< 25%3) for Type 1/2 myoma patients, and symptom improvement was encouraging with moderate shrinkage of treated myomas. We therefore recommend MRgFUS as one of the appropriate treatment methods for Type 1/2 uterine myomas. As previously stated, we currently exclude Type 3 myoma patients from MRgFUS treatment. However, new or enhanced procedural and software or hardware advances may improve the outcome for these patients. Longer-term follow-up of MRgFUS-treated patients would facilitate the evaluation of this technique and might lead to broadening of the indications of MRgFUS for uterine myomas.