Fibroids are an ever present problem in gynecology and may in some way affect as many as 25% of women of childbearing age1. Traditionally, treatment for fibroids has been divided into the conservative approach, involving nothing other than administration of iron tablets and support, and the surgical approach, involving removal of the fibroids by either hysterectomy or myomectomy. There is an immense disparity between these two regimens. Over the last decade many women have, quite rightly, found this polarization of options unacceptable, and are demanding other alternatives that lie somewhere between the extremes to treat without recourse to surgery what is, after all, a benign disease.
Several procedures have emerged which fill the gap and are starting to allow active fibroid treatment in an effective yet minimally invasive manner, without the need for significant surgery. Uterine artery embolization (UAE) was the first major paradigm shift in this field and is now widely practiced2; it can be extremely effective by causing infarction of fibroids. UAE is associated with a short in-patient hospital stay but does have a variety of associated complications, plus a substantial amount of postprocedural pain which usually requires continuous opiate analgesia for at least 24 h3. This technique, however, has assumed an important role in minimally invasive uterine fibroid treatment and is starting to contribute immensely to this area.
The newest treatment for uterine fibroids, which is currently undergoing worldwide assessment, is MR-guided focused ultrasound (MRgFUS)4. This technique is entirely non-invasive, consisting of the use of high-intensity focused ultrasound targeted at selected fibroids using an anterior acoustic window through the anterior abdominal wall. It causes heating within the fibroid and thermal ablation of the fibroid tissue. Magnetic resonance imaging (MRI) is used to control the whole process, including targeting of fibroids, online real-time temperature mapping of the target site as the energy is applied and assessment of the extent of treatment achieved at the end of the procedure.
MRgFUS is not an entirely new procedure; the USA Food and Drug Administration (FDA) approved it for limited uterine fibroid treatment in late 2004 and, prior to this, substantial work in the field had been underway5. It is a technology that has been slowly developing and evolving, with improving results as practitioners learn how to use the technology more effectively and appropriately, and begin to understand which patients are most likely to benefit from it and, more importantly, which patients not to treat. This is particularly well illustrated in three articles in this issue of the Journal, each describing different but complementary aspects of this emerging field with respect to the treatment of uterine fibroids. Between them, they investigate the procedure from its early pre-FDA approval studies to a more mature utilization with up to 24 months' post-treatment follow-up.
Taran et al.6 provide us with a valuable direct comparison between the most traditional and common treatment for uterine fibroids, hysterectomy, and MRgFUS in a matched group of patients recruited contemporaneously at other institutions. Patients were not randomized at one particular institution to one therapy or the other according to the regulatory authorities, but were recruited at matched institutes using the same entry criteria. The results produce a very clear picture. The non-invasive, outpatient MRgFUS procedure was associated with significantly fewer procedural and periprocedural adverse events in comparison with surgical hysterectomy. Similarly, recovery times were much faster for MRgFUS, with an average of only 1 day away from work as opposed to 6 weeks for hysterectomy. When postprocedural disability was measured by disability questionnaires, the MRgFUS procedure was associated with significantly lower scores than was hysterectomy, with substantially less associated disability until 3–6 months postprocedure, after which time the two groups equilibrated.
Comparing postprocedure fibroid-related symptoms between the two techniques is of course much more difficult, for several reasons. First, patients do not normally have further fibroid-related problems following hysterectomy, and this is not easily represented in symptomatic questionnaires. Second, the main aim of the initial studies carried out under the FDA's egis was to assess the safety of this new procedure, with efficacy measurements being a secondary consideration. Therefore, the size of ablated areas that could be produced during ablation procedures and the length of the procedures themselves were strictly limited in early studies. Therefore, although these early studies7 showed quite effective symptomatic relief over 6 months, long-term results are hard to interpret from early studies since the treatment sizes, as measured by non-perfused volume (NPV) of treated fibroids (percent of ablated targeted fibroid relative to remaining perfused tissue), were relatively low overall. One of the incidental findings of these early studies4, 7 was that the greater the percentage of the targeted fibroid that was destroyed by the ablation procedure, the better and more longer-lasting the symptomatic result. This was hardly a surprising finding and is one which the groups investigating UAE had discovered years previously; it was nevertheless an important finding in relation to thermal ablation modalities and has influenced substantially subsequent therapeutic approaches, with treatment now aimed at producing the largest possible area of destruction within the targeted fibroid.
The second paper, by Okada et al.8, takes this finding as its main theme. Japanese groups started to treat fibroids with MRgFUS in 2003. Initially they treated patients mainly using the criteria from the previous FDA-overseen studies, with their relatively low NPV. Over the course of the next 4 years they relaxed these restrictions progressively and now utilize a procedure that involves targeting for ablation as large an area of fibroid as possible. In their retrospective study of patients who had undergone MRgFUS, they divided the population chronologically into two equal groups, each with 12 months of postprocedural follow-up. The early group had a mean NPV of 39% whereas the later group had a mean NPV of 54%. Interestingly, the number of procedural complications fell as the NPV rose, indicating that there is a clear learning curve in understanding how to minimize treatment-associated problems, and that increasing the size of the area of fibroid ablated does not compromise the safety of the procedure. As a measure of the success of their procedure, they used primarily the percentage of patients that required further treatment for residual fibroid symptomatology following MRgFUS and found that 12% of the early group required subsequent fibroid therapy compared with only 5% of the later group, a very significant statistical reduction. These data, derived from centers changing over time the extent of fibroid area treated, which allowed them to subsequently act as their own crossover controls, are extremely useful. They confirm initial perceptions that ablating large areas of fibroid is not only safe but improves the symptomatic outcome and should be the treatment goal in order to achieve optimal therapeutic response.
The third paper, by Funaki et al.9, presents 24 months' of follow-up data on a group of 91 patients treated with MRgFUS. Interestingly, these researchers divided fibroids into different categories according to their appearance on MRI. In the relatively recent past, clinicians and imagers were mostly oblivious to the range of appearances that fibroids can exhibit on MRI. Classical textbook descriptions predominated, indicating that fibroids are well-defined muscle masses of low signal intensity on T2-weighted imaging and that are isointense on T1 sequences. As soon as investigators started to image fibroids regularly using MRI it became very clear that this textbook description was completely inadequate. Many fibroids show a substantial diffuse high-intensity signal on T2 imaging or may display mixed high- and low-intensity signals or many gradations of signal in between. These differences are independent of additional necrosis and red degeneration, both of which can be clearly visualized separately. It seems likely that the hyperintense T2 fibroids are completely different genetic strains from the more traditionally described low-intensity T2 signal variety. Of greater importance is the fact that the response to MRgFUS is quite different between these two different types of fibroid10. Low-intensity T2 signal fibroids respond extremely well to thermal ablation, with large areas of destruction being easily produced which lead to large NPVs. In comparison, fibroids of high-intensity T2 signal respond much less well to MRgFUS. Much greater power is required to produce a comparable rise in temperature and a much higher power density of sonication is required to produce significant ablation volumes. Even when moderate areas of NPV are achieved in these fibroids the effects are not as long-lasting as are the effects in low-intensity signal T2 fibroids. Why this should be the case is not clear; higher vascular flow is the most common suggestion for these differences, but this is largely conjectural.
Funaki et al.9 found in their patient group that hyperintense T2 fibroids had not shrunk at all 6 months post-therapy, whereas the other fibroid groups had decreased in volume by 36.5% at 6 months and by 39.5% at 24 months' post-therapy. Symptom scores had also fallen significantly by 3 and 6 months and these median falls were maintained at the same level at 24 months post-therapy in their patient group. T2 hyperintense fibroids were not included in this analysis because they had such a high retreatment rate that they could not be scored adequately in this type of assessment.
Four pregnancies were achieved incidentally in the Funaki study9. More than 55 pregnancies have been described following MRgFUS treatment worldwide which, bearing in mind that in the initial studies only women who had completed their families were treated, is very reassuring. There appears to be no excess of pregnancy complications described11 in patients who fall pregnant after MRgFUS, suggesting that pregnancy may be safe after this therapy and that, unlike many of the other treatments utilized in this field, MRgFUS probably does not impair fertility. Whether MRgFUS can actually improve fertility in women with fibroids is as yet unknown. Much attention will be given to this area in the immediate future by researchers in the field to try to determine whether it could be a useful therapy for women who have fibroid-related infertility; in fact, multinational studies in this field are starting to get underway.
Our understanding of how to apply this new technology in the treatment of uterine fibroids is rapidly expanding. We now know that MRgFUS works quickly with very few side effects and that patients can return to normal function almost immediately after treatment. We know that the larger the area of ablation that can be produced in the targeted fibroids, the better the symptomatic response and the less the need for other fibroid interventions, and that when the early symptomatic response is highly effective, long-term improvement of symptoms is almost invariably seen. It is also becoming extremely clear that hyperintense T2 fibroids respond poorly to MRgFUS and there is a reasonable argument for not treating these types of fibroids at all.
The non-invasiveness of the MRgFUS procedure and its ability to rapidly restore patients to an improved function are very strong advantages and the conversion of hospital inpatient stays into outpatient procedures is very attractive economically given the rapid gain in quality of life post-therapy, with few complications. What issues therefore still remain that limit the uptake of this type of technology as a widespread and commonly utilized therapeutic modality for uterine fibroids, other than the simple financial issue of acquiring the machinery?
MRgFUS still takes quite a long time to perform, particularly in patients with large fibroids. These are very problematic because of the daunting size of total tissue volumes involved, which are hard to cover in a sensible time frame. The most effective way of addressing this problem will be by further technological developments which will allow larger areas of destruction to be produced more quickly, and research into these types of improvements is underway. Randomized comparisons of MRgFUS with other treatment modalities, such as UAE, must be carried out in the relatively near future to set this new therapeutic modality in an appropriate context relative to other more established therapies, although there is a reasonable argument for waiting until further technological improvements have taken effect in the near future, prior to undertaking such expensive and time-consuming studies. Without effective randomized studies, however, this field will not move forward as rapidly as it should and we must see these being carried out in the next few years. Fibroids that are hyperintense on T2 and multiple fibroids studded throughout a very large uterus remain difficult problems and these types of configurations may be much better treated with UAE at the moment. The position of a fibroid in the uterus or whether it is pedunculated are much less of a problem for MRgFUS as long as there is a reasonable acoustic window to reach the targeted area. Probably the most important area of development in this field is the understanding of which therapeutic modality is most appropriate for which patient, given the multifactorial problems with which such patients often present.
The potential of this type of therapy, however, is huge. The ability to destroy deep tissue in the body under image guidance with thermal feedback, allowing safe and effective application of destructive energy, is very exciting and may well, in the coming decade, change our understanding of how we apply therapy in many body areas.