Chapter 17 – MRI-Guided Ultrasound Lysis of Fibroids


Magnetic-resonance-guided focused ultrasound surgery (MRgFUS) is the only truly non-invasive procedure for the treatment of uterine fibroids. In MRgFUS, high-frequency ultrasound beams target specific tissue, in this case fibroids, causing increased temperatures leading to destruction of the tissue by coagulative necrosis. The resultant fibroid shrinkage occurs under the guidance of real-time magnetic resonance imaging (MRI) for accuracy and preservation of healthy surrounding myometrium [1]. This chapter will begin with an overview of the history of MRgFUS, then examine the technical parameters of the procedure and measurement of success, discuss patient eligibility and selection, preparation for the procedure, recovery, potential complications, and finally, outlooks for fertility and other outcomes.

Chapter 17 MRI-Guided Ultrasound Lysis of Fibroids

Jessica Gelman and Charles E. Miller

17.1 Introduction

Magnetic-resonance-guided focused ultrasound surgery (MRgFUS) is the only truly non-invasive procedure for the treatment of uterine fibroids. In MRgFUS, high-frequency ultrasound beams target specific tissue, in this case fibroids, causing increased temperatures leading to destruction of the tissue by coagulative necrosis. The resultant fibroid shrinkage occurs under the guidance of real-time magnetic resonance imaging (MRI) for accuracy and preservation of healthy surrounding myometrium [1]. This chapter will begin with an overview of the history of MRgFUS, then examine the technical parameters of the procedure and measurement of success, discuss patient eligibility and selection, preparation for the procedure, recovery, potential complications, and finally, outlooks for fertility and other outcomes.

17.2 History

MRI has the best resolution and sensitivity for fibroids compared to other modalities, such as ultrasound or CT scanning, and allows for real-time three-dimensional thermal mapping [2]. The Food and Drug Administration (FDA) approved MRgFUS for the treatment of fibroids in 2004 [1]. At that time, the treatment could only be used to ablate 33% of a fibroid, as it was yet unclear whether this procedure was safe for the surrounding non-target tissue. In 2006, the FDA approved ablation of 50% of any given fibroid, and by 2009 ablation of 100% of a fibroid was approved, with the caveat that the fibroid had to be more than 10 mm from the serosal surface [3]. Today, the FDA has approved this technique for the ablation of uterine fibroids in pre- or perimenopausal women with symptomatic uterine fibroids desiring a procedure that spares the uterus, and whose uterus size is less than 24 weeks.

The original system that the FDA approved for MRgFUS was the ExAblate 2000 (InSightec, Haifa, Israel). Newer platforms, including the ExAblate 2100 (InSightec, Haifa, Israel) and the Sonalleve MR-HIFU (Philips Healthcare, Vantaa, Finland), offer novel sonication techniques allowing for more efficient fibroid lysis [4].

17.3 How It Works

High-frequency ultrasound beams cause heating between 56°C and 70°C which subsequently leads to protein degradation, cell death, and coagulative necrosis that is eventually absorbed by the body [1, 5]. More specifically, temperatures between 50°C and 52°C for 4–6 minutes cause cellular damage that is irreversible; between 60°C and 100°C they will cause desired protein degradation and coagulative necrosis. Heating past 105°C will cause carbonization and vaporization of tissue, which can worsen ablation potential by impeding US wave transmission [69].

17.4 Measuring Treatment Success

The amount of tissue that has been effectively ablated after MRgFUS is termed non-perfused volume (NPV). Studies have shown NPV to be the best predictor of success for this procedure. Increased NPV correlates with better symptom improvement as measured in several studies by the Symptom Severity Score (SSS), which is included within the Uterine Fibroid Symptom and Quality of Life (UFS-QOL) scale, a validated questionnaire. Higher NPV is also associated with increased fibroid shrinkage and decreased rates of reintervention [46, 1012].

17.5 Patient Selection

When MRgFUS was first approved for use on uterine fibroids, it was difficult to find eligible patients. In addition to the ablation percentage restrictions that were initially present as mentioned earlier in this chapter, several other factors were considered contraindications to this procedure. For example, MRgFUS was previously contraindicated for anyone who had loops of bowel or abdominal scars in the pathway of the US beam. One study showed that bowel interposition affected 60.4% of women, and thus only 38.9% of women were eligible for MRgFUS, while 99.2% were eligible for uterine artery embolization (UAE) [13]. Today’s techniques, which circumvent many of these obstacles, have increased the number of eligible patients immensely.

Candidate selection is based upon the likelihood of achieving an acceptable NPV. Part of determining a patient’s likelihood of success involves pre-screening all patients with an MRI to assess for certain fibroid features, namely size, number of fibroids, vascularity, accessibility, and texture. Assessing for other uterine or other pelvic disorders is also prudent.

17.5.1 Size Restrictions

A maximum diameter of 10 cm has been suggested for best success of MRgFUS. This number is dependent on the time it takes to ablate a fibroid of such girth. For example, an 8 cm fibroid takes about 3 hours to achieve an ideal NPV. For patient procedure tolerance (as the procedure requires lying still in a confined space, which can be uncomfortable), as well as deep vein thrombosis (DVT) risk, 3–4 hours is the time limit for any given ablation session [6].

In addition to maximum size restrictions, there are also minimum fibroid sizes that are considered safe for MRgFUS. A single focal ablation is about 2.5 cm in length along the anterior–posterior (A-P) direction. Thus, attempting to sonicate a fibroid less than 3 cm in diameter may result in undesired over-ablation of healthy tissue. For this reason, treating fibroids less than 3 cm with MRgFUS should be reserved for those with severe symptoms in whom the benefit of ablating such a fibroid outweighs the risks of injuring healthy intra- or extrauterine tissue [5].

17.5.2 Location

Each ablation system also has a finite range or distance of tissue that it is able to reach. Thus, fibroids that fall outside of this range are unsuitable for treatment with this modality [5].

Fibroids may also be in dangerous proximity to other vital structures that might be at risk of being sonicated. Because high-density tissues absorb ultrasound energy more readily, low doses of energy can significantly heat bone, like the sacrum. Overheating the sacrum can in turn damage nearby nerves, causing pain or damage to that nerve. For this reason, fibroids lying close to the bone and specifically sacrum may not be good candidates for MRgFUS. A distance of at least 4 cm from the spine or nerve roots has been recommended [14]. Other high-density areas, such as abdominal scars, can pose a challenge. Scars often have increased energy absorption and thus are at risk for thermal injury with less than expected energy levels. Alternatively, low-density matter, like air, reflects US rays. Thus, a mass of intestines between the transducer and the fibroid is a relative contraindication to MRgFUS [5].

Some studies argue that limits on proximity to endometrium, such as >1.5 cm away, may be prudent, but this is not standardly recognized or practised as of yet [15].

17.5.3 Vascularity

Fibroids that are very vascular may not reach therapeutic temperatures because the high volume of blood flow carries heat away from the tissue of interest. High vascularity can be detected if the intensity of the fibroids seen on T2W MRI is increased relative to the wall of the uterus [5]. Hyperintense fibroids, which are more common in younger women, are associated with decreased NPV and worse post-treatment SSS [4].

17.5.4 Indications

Anything that distorts the focal point of the beam, such as copious adipose tissue or distorted muscle, may cause suboptimal treatment or pose an additional risk to non-target structures [5].

There are other factors to consider in selecting for optimal patient outcome aside from the likelihood of achieving high NPVs. For one, fibroids that are seemingly optimal for MRgFUS based on size, vascularity, and accessibility may not be appropriate targets if they are not relevant to a patient’s symptoms. For example, a woman with abnormal uterine bleeding may not experience relief of symptoms from ablation of a subserosal fibroid. However, ablation of a submucosal fibroid invading the endometrium may very well decrease her bleeding. Alternatively, attempting to ablate a subserosal fibroid in a woman having bowel or bladder symptoms may be appropriate. Furthermore, if a woman’s uterus is overrun by numerous small fibroids, it is unlikely that one or a few of these fibroids alone are causing her symptoms. It is thus recommended that treatment is concentrated on the “dominant” fibroid, which is the one most likely associated with the patient’s symptoms [5].

While no cases have yet been reported, pedunculated subserosal fibroids have a theoretical risk of amputation, followed by dislodgement into the intraperitoneal cavity [5]. One study retrospectively looked at stalk-sparing MRgFUS treatments in nine women. Results showed that with an average NPV of 67%, mean fibroid volume had decreased by 30% and stalk diameter by 13% at 6-month follow-up. This translated into average baseline SSS of 30 diminishing to 14.6, with no cases of stalk separation or untoward outcomes [4].

African-American women, who are more likely to have symptomatic and early-onset fibroids, are also less likely to be eligible for the MRgFUS procedure, as one study found. These authors noted the most likely reasons for failed eligibility were multiple small fibroids or significant abdominal wall scarring. However, African-American women, when eligible, have equivalent outcomes to others undergoing MRgFUS [16].

There are few absolute contraindications to this procedure, but patients with suspected uterine or gynaecologic malignancy, current pregnancy, and acute inflammatory diseases should not undergo MRgFUS for fibroids. One must recognize the increased risk of leiomyosarcoma in women beginning in their 40s [17]. Any contraindications to having an MRI, such as implantable metal devices (intrauterine devices, cardiac pacemakers, etc.), allergy to contrast agents, or morbid obesity over limits for the MRI machine, are also contraindications to this procedure [18].

17.6 Optimizing Patient Selection

Today, certain methods have been demonstrated to circumvent several of the barriers mentioned above and thus increase patient eligibility for this procedure.

17.6.1 Interposed Bowel

As mentioned previously, the presence of bowel loops in the path of the transducer is a contraindication to MRgFUS because the low-density air can reflect beams; or worse, unpredictable bowel contents can result in thermal damage or even bowel perforation. While this factor is one of the most common barriers, it is also one of the most modifiable. One way to displace bowel is through bladder filling, whereby a Foley catheter is inserted and the bladder is backfilled with normal saline. The Foley catheter is then clamped. Urine production over the course of the procedure can cause bladder distension and distortion of the mapped area, so small drainages from the catheter should be done frequently with MRI re-verification of the anatomy.

Rectal filling is another method to mechanically shift interposed bowel. Using ultrasound gel transrectally or a rectal balloon can shift the uterus forward. Not only can this displace the anterior bowel loops, but it can also decrease the distance between the anterior abdominal wall and the uterus, and increase the space between the uterus and the spinal nerves and bones. Perhaps most simply, a convex gel pad can be placed on the table directly under the patient’s abdomen to apply central pressure and shift the bowel. One study used vaginal pessaries to manipulate bowel. Any combination of these methods can be trialled to optimize patient eligibility [6].

17.6.2 Scars

Abdominal scars are other common barriers to eligibility for this procedure. Scar tissue is often more dense than non-scarred tissue and thus has increased US absorption. Also, scar tissue is often less vascular than other tissue, and thus experiences decreased cooling and risk for underestimation of accumulated heat. Since sensation may be decreased in scarred tissue, patients may not recognize and verbalize that injury is happening as readily as with normal tissue. Plus, once a burn has progressed to second or third degree, even less sensation is appreciable due to thermal nerve damage.

Thus, visualizing scars on MRI is imperative but not always easy. Older scars are much more difficult to visualize, but fortunately more safe to sonicate through. Newer, thick, pronounced scars are often the densest and thus the worst to attempt to treat through. Scars of the epidermis, dermis, and subcutaneous tissue are usually much easier to visualize than scars on muscle or uterus through MR imaging. Special MRI sequences or even MRI paint (like nail varnish or dedicated contrast with paramagnetic iron oxide particles) can help pick up scars on imaging. Reflective acoustic patches, which are visible on MRI, suitable for scars of many shapes, water-resistant, and easily adherent to skin, are another newer and viable option. One clinical study found a small risk of hyperemic changes on skin and muscle but otherwise no major adverse events. These may reflect energy too well, though, and can damage the machine’s transducer [19].

Positioning the patient to remove the scar from the field of interest is also possible through tilting or bladder filling [6]. Beam shaping, which is possible with some platforms, can disable one part of the beam to avoid the scarred tissue.

Sonicating through a scar can sometimes be appropriate, and certain patient factors make this process safer. For example, attempting to sonicate a deep fibroid means the focus of energy will be less intense close to the skin, leaving lower risk of overheating a scar. Conversely, fibroids that are very vascular or cellular that will require high energies for ablation are not suitable for treatment if there is an unavoidable scar. If the decision is made to sonicate through a scar, real-time mapping, ample cooling time, and constant patient feedback are crucial.

17.6.3 Pretreatment with GnRH

Pretreatment with GnRH agonists has been shown to shrink fibroids prior to the procedure and improve outcomes of MRgFUS [2023]. Other studies have shown that GnRH agonists can help reduce fibroid vascularity [24]. While there have been no specific studies to date examining the effect of GnRH and reduced vascularity in relation to MRgFUS outcomes, one can predict that decreased vascularity through GnRH treatment would improve the success rate of MRgFUS since, as mentioned above, MRgFUS is superior for fibroids with lower vascularity [6].

17.6.4 Vascularity

Other methods have been used for optimizing conditions with highly vascular fibroids. Vessel targeting can specifically ablate vessels that are feeding fibroids by taking a non-enhanced MR angiography (MRA) image immediately before the procedure. While this technique has shown extremely promising results, the outermost circumference of the fibroid must still not be treated for the safety of surrounding healthy tissue [19]. Using oxytocin as a pretreatment has also been shown to decrease energy level requirements and sonication time [25]. With extreme caution, some practitioners have been using higher temperatures than average to effectively treat very vascular fibroids.

17.6.5 MRgFUS Platforms

Certain MRgFUS systems can modify the size, energy level, and transmission of each sonication beam, such as the ExAblate. This can help optimize treatment by tailoring each ablation attempt to the precise tissue of interest. For example, a long area of ablation can be targeted if the energy level is increased parallel to the beam, while increasing the level perpendicular to the beam will create a thicker, deeper spot [5]. The Sonalleve system has the option for volumetric ablation (as opposed to the standard point-by-point method), which is more energy-efficient and thus can sonicate a larger fibroid in a shorter period of time. It does so by ablating in a circular manner, thus pre-heating the next area to be treated and limiting time needed to reach therapeutic temperatures [19].

17.6.6 Large Fibroids

To circumvent the issue of a fibroid being too large, physicians can pre-plan for two separate treatment sessions, one for the superior aspects of the fibroid and one of the inferior aspects, for example. Of note, it may be prudent to avoid sonicating parts of the fibroids that have already been treated, as those now drier areas may absorb too much energy. Sometimes, though, such a procedure may be beneficial and thus intentionally planned. For instance, a fibroid that is too close to the sacrum could have one ablation anteriorly, followed by another session targeting the rest of the fibroid once it has moved further from the bone [5].

17.7 Patient Preparation

All patients should have a screening gadolinium-enhanced MRI. This will assess for eligibility and allow for patient-specific preprocedure planning.

On the day of the procedure, the patient should arrive having fasted for at least 6 hours. Some recommend that bowel cathartics, glucagon, and a low-residue diet prior to fasting may also minimize bowel activity during the procedure [3]. The abdominal wall should be shaved from the umbilicus down to the pubic symphysis. Patients can be asked to do this the night before the procedure. The abdomen should also be free of creams or lotions, as sonicating through these can result in skin burns. An intravenous line will then be placed so conscious sedation can be administered. This helps with patient comfort, decreased anxiety, and minimized movement. The patient should, however, be able to converse with the operators and verbalize if they are feeling burning, pain, or cramping. Some highly compliant and relaxed patients may not require any sedation. A Foley catheter is also placed to keep the bladder empty, unless bladder filling is desired for bowel manipulation as described above, in which case the Foley will be used for filling and then clamped [14, 18, 26].

17.8 Procedure

The MRgFUS system is integrated with a special table that can be docked to a compatible MRI machine. On top of this table is a degassed water tank coupled to a gel pad upon which the patient lies prone. There should be ultrasound gel on both sides of the solid gel pad. The gel and water help to propagate ultrasound waves, as any air between the beam and the patient can cause beam reflection. Then the T2-weighted MRI is used to locate and map fibroids in relation to surrounding structures in three orthogonal planes. After mapping, a low-energy test dose is performed. During the procedure, there is real-time, quantitative temperature mapping that identifies minute temperature changes in non-target tissues before irreversible damage can occur. The full ablation session then ensues with constant MRI monitoring and modified sonication as necessary. Between each sonication, down time is allowed for adequate cooling, as thermal buildup can damage surrounding tissues. The two newest platforms, the ExAblate 2100 and the Sonalleve, automatically pause for cooling. At the end of the procedure, the new area of NPV is measured with gadolinium-enhanced MRI. The results seen immediately post-treatment may be a good indicator of the thermal dose received for a small treatment area; for larger treatment volumes, however, there will likely be a delay in visualizing the full treatment effect. This is thought to be due to vessel occlusion causing a slow downstream effect, or perhaps from underestimation of temperature at the edges of the field that are too distant from the transducer to detect the small changes [4, 14, 26].

17.9 Recovery

There is minimal time needed for immediate postoperative recovery as patients have not been under general anaesthesia and have had an incision-free procedure. Thus, patients usually do not require hospitalization after MRgFUS, can often leave 1 hour after the procedure, and can even return to work within 1–2 days [27]. Most people note symptomatic improvement within 6 months with an average improvement in Symptom Severity Score (SSS) to 50% of pretreatment values [4].

17.10 Complications

Initial studies during clinical phase trials showed no unintended second procedures, no life-threatening events or deaths after MRgFUS [28]. To date, there have been no reports of death or hysterectomy as direct outcomes of this procedure [2]. One of the most serious complications reported is sciatic nerve injury, but this resolved without intervention by 1 year post-procedure [2]. Risks of urinary tract infection and minor skin burns have been reported to be less than 2%, and risk of major skin burn requiring further treatment is less than 0.5% [29]. All reports of major skin burns have been through a previously existing abdominal scar [2]. DVTs are infrequent and bowel perforation is an extremely rare complication of this procedure [18]. More commonly, a patient may experience lower abdominal pain, leg pain, or buttocks pain after the procedure [27].

17.11 Fertility Outlook

MRgFUS is believed to be a suitable option for women with infertility related to intrauterine fibroids, as this procedure is uterus- and endometrium-sparing, non-invasive, and does not use radiation [18].

While there have been no studies on women who have infertility as their main complaint, to date there have been 54 documented pregnancies in 51 women who have previously been treated with MRgFUS for uterine fibroids. Each of these women had reported being family complete, as desiring future fertility was previously a contraindication to having this procedure. The study that describes these cases reports the mean age for the women who became pregnant to be 37.2 years and the miscarriage rate as 26%. The average time between MRgFUS treatment and pregnancy was 8.2 months. At the time of publication, only 42% of the women had delivered, but 64% of them had normal spontaneous vaginal deliveries with a mean birth weight of 3.3 kg. There were two cases of placenta praevia reported and one serious complication relating to a myomectomy that was done at the time of the caesarean section; however, this woman went on to have a second uncomplicated pregnancy [30]. Other studies have shown that the caesarean section rate after MRgFUS treatment and the rate of low birth weight or stillborn infants (of which there were zero) were lower than after uterine artery embolization [30].

Currently, there is one randomized controlled trial offering MRgFUS treatment to women whose chief complaint is infertility. The first published report coming from this study demonstrated a successful conception 3 months following two sessions of MRgFUS. This patient went on to have an uncomplicated pregnancy and full-term normal spontaneous vaginal delivery [31]. There is also a case report showing the first successful case of in vitro fertilization after MRgFUS for fibroids [32].

At the time of this publication, there is still much to be learned and researched regarding the role of MRgFUS in infertility and pregnancy. Thus, it is important to follow patients closely who become pregnant after MRgFUS treatment. As of yet, there is no evidence of worsening fertility after MRgFUS, but the appropriate period between treatment and conception is still to be determined [18].

17.12 Outcomes

A meta-analysis of 38 studies found that MRgFUS is safe, cost-effective, efficient, and improves quality of life and infertility. When focusing in on quality of life, this analysis notes improvements on the validated Uterine Fibroid Symptoms Quality of Life Questionnaire (UFS-QOL) of between 15 and 44 points at 6 months and between 21 and 47 points at 12 months post-procedure [33]. One particular study that pre-treated with GnRH for 3 months found 28.35 points of improvement at 6 months and 30.3 points at 12 months [34]. Another study of 359 women treated with MRgFUS used the UFS-QOL over time and found improved quality of life scores that persisted for over 2 years [11]. When comparing quality of life scores for MRgFUS to abdominal hysterectomy, one study concluded that while the scores for improvement in quality of life were not significantly different at 6 months post-procedure, MRgFUS had significantly fewer complications [35]. Another study, however, comparing MRgFUS to UAE, found that UAE had superior total health-related quality of life and symptom severity scores [36]. Importantly, 60% of women who were informed about the different treatment modalities for fibroids chose MRgFUS as their top choice [37].

A recent study notes that an NPV greater than 80% is necessary to define clinical success [38]. Alternatively, one study notes that having greater than 45% NPV correlated with a 15% reintervention rate at 2 years, while an NPV between 10% and 20% was associated with a 40% risk of reintervention [39]. Another study looking at 5-year follow-up found that the overall reintervention rate was 58.64%. When stratified for NPV, those with more than 50% NPV post-procedure had a 50% reintervention rate at 5 years [2, 23]. Newer technologies are reporting NPVs at a range of 33–100% [2].

17.13 Conclusion

MRgFUS is a newer, non-invasive technique for ablation of uterine fibroids in symptomatic women. Maximizing patient eligibility using various methods is crucial for the success of this safe and effective treatment. Outcomes of using MRgFUS in women with a chief complaint of infertility are looking positive, though much more research is necessary.


1.Carranza-Mamane B, Havelock J, Hemmings R. The management of uterine fibroids in women with otherwise unexplained infertility. J Obstet Gynaecol Can 2015;37(3):277–85. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

2.Quinn SD, Gedroyc WM. Thermal ablative treatment of uterine fibroids. Int J Hyperthermia 2015;31(3):272–9. CrossRef | Find at Chinese University of Hong Kong Findit@CUHK Library | Google Scholar | PubMed

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Dec 29, 2020 | Posted by in GYNECOLOGY | Comments Off on Chapter 17 – MRI-Guided Ultrasound Lysis of Fibroids

Full access? Get Clinical Tree

Get Clinical Tree app for offline access