Harnessing the interaction of ultrasound with tissue for therapeutic benefit: high‐intensity focused ultrasound

Most readers of this Journal will be used to the idea that diagnostic ultrasound, when used prudently and in accordance with published guidelines, is safe and does not harm the embryo or fetus. The reason for the need for constant vigilance and the concern about, for example, the safety of performing Doppler examinations in the first trimester of pregnancy, is that it is well known that ultrasound can lead to biological change, and care must be taken to ensure that such effects do not occur under circumstances that will lead to lasting harm. The possibility exists, however, of harnessing this capacity to produce tissue effects for therapeutic benefit, as is done with physiotherapy applications, most widely known for the treatment of sports injuries. A new form of ultrasound therapy is rapidly gaining clinical acceptance. This is known as high-intensity focused ultrasound (HIFU) or focused ultrasound surgery (FUS)1. The technique relies on the ability to bring an ultrasound beam to a tight focus at a distance from the transducer. This is possible because of the millimeter wavelengths of ultrasound at the megahertz frequencies most commonly used in medicine. The principle of HIFU is that when sufficient power emanates from the focusing transducer, the energy concentrated in the focal volume is sufficient to raise the tissue temperature to cytotoxic levels (> 56 °C for 1 s) so rapidly that no heat is lost by thermal conduction or vascular perfusion. At 1.5 MHz, in typical soft tissues, this requires focal peak intensities in excess of 1000 W/cm2, 10 000 times those used in diagnostic ultrasound imaging. In practice, it has been shown that thermal ablation (coagulative necrosis), can be achieved solely within the focal volume, with no damage to surrounding or overlying tissues. The principle is illustrated in Figure 1, and the insets show the appearance of a HIFU ‘lesion’ created in ex-vivo liver tissue. The size and shape of the HIFU focal volume depends on the ultrasound frequency and transducer geometry, but is usually ellipsoidal, 1–2 cm along the long axis and 2–3 mm in diameter. It has been shown histologically that the lesion margin is very sharp, the boundary between dead and unaffected tissue being only a few cell layers in thickness2. It is thus possible to use focused ultrasound beams from a distant (extracorporeal or intracavitary) source in a minimally invasive technique, to target selectively deep-seated masses such as malignant tumors of the liver or kidney, or benign tumors of the breast or uterus. A single HIFU exposure (lasting 1–5s) can be used to destroy a small target, but more often some method of thermally ablating larger tissue volumes is required for successful clinical treatments. This is usually achieved by mechanical motion of the transducer, by electronic beam steering, or by a combination of both. Schematic diagram showing the principle of high-intensity focused ultrasound (HIFU). A high-power ultrasound transducer is used to bring the beam to a tight focus, which is positioned within the target tissue. Inset A shows a HIFU ‘lesion’ in ex-vivo bovine liver tissue—a region of coagulative necrosis with a sharp boundary between live cells and dead cells, as shown in the histological section shown in inset B. While the potential to selectively destroy tissue at depth in a single session, without damaging overlying organs and, using an extracorporeal approach, without breaching the skin, has considerable appeal, its success depends critically on being able to place the HIFU focus accurately at the target site, and to monitor the treatment as it progresses. While HIFU was first investigated for clinical usage in the 1950s, it was the lack of this ability that hampered its more widespread development at that time3, 4. The advent of sophisticated imaging techniques, however, has now made clinical implementation feasible, and both magnetic resonance (MR) and ultrasound imaging have been integrated into HIFU devices. Magnetically compatible ultrasound HIFU transducers have been developed which allow MR imaging to be used both to aid targeting and to monitor treatment as it progresses (MRgHIFU, MRgFUS)5. An advantage of MRgHIFU is that currently available thermometry sequences can be used to monitor the treatment and to assess tissue damage. Thermal dose is used as an indicator of ablation in this context6. This parameter was developed for hyperthermic cancer treatments, and reflects both the temperature achieved and the time for which it is maintained. A temperature of 43 °C held for 240 min is deemed to have resulted in thermal necrosis for HIFU treatments5. While it is not yet possible to display ultrasound-based temperature maps in real time with ultrasound-guided treatments (USgHIFU), tissue damage is inferred from the appearance of bright echoes in the focal region7. Each monitoring method has inherent advantages and disadvantages. Ultrasound- and MR-based elastography, which map tissue stiffness, are now also being investigated for their potential application with HIFU, since thermal ablated tissues are palpably harder than their surroundings8, 9. To date, MR guidance has been used almost exclusively for extracorporeal treatments of uterine fibroids10, 11, 12, whereas ultrasound guidance has been used more widely for cancer applications and for transrectal HIFU treatments of the prostate13. Only a limited amount of clinical data from HIFU treatments is available in the published literature, despite the large numbers of patients that have been treated. By the end of 2007, more than 15 000 prostate treatments, more than 20 000 ablations of abdominal tumors, osteosarcoma and uterine fibroids and more than 400 000 treatments of the cervix had been delivered. The majority of patients have been Chinese, although all prostate treatments and most uterine fibroid procedures have been carried out in the West (mainly Europe and the USA). Small-scale trials of HIFU application in the treatment of breast14-18, pancreatic19, renal20, 21 and bone and liver20 cancer have been reported. The two areas in which the largest trials have been conducted are prostate cancer13 and the treatment of uterine fibroids10-12. Brain tumors have also been targeted, although the results have not yet been published outside conference proceedings. A specialist handheld device has been developed in China for the treatment of cervicitis and vulvar dystrophy22, although this has not yet been used in the West. Uterine fibroid treatments have been the subject of a multicenter Food and Drug Administration (FDA)-approved study and results of Phase I, II and III clinical trials, carried out in seven sites in Europe (two sites), the USA (three sites) and Israel (two sites) have been reported23, 24. These clinical trials involved patients who were over 18 years old, had no desire for future pregnancies and, for the Phase I and II trials, had no dominant fibroid > 10 cm in diameter. Phase III studies are underway, with fibroid volume reduction and quality of life outcomes being assessed. Women with extensive abdominal scars or scars lying in the path of the focused ultrasound beam are excluded from the trial, as scar tissue absorbs ultrasound strongly, and it has been shown that this may result in a skin burn. Patient weight is limited to < 113 kg (the upper limit for the MR imaging gantry) and co-existent significant pelvic or systemic diseases result in exclusion from the study. Early results from the first 108 patients showed that 79.3% of women reported significant improvement in fibroid-related symptoms. The mean fibroid volume reduction at 6 months was 13.5%, but as there was a persistent non-enhancing volume on contrast MR imaging, it was thought that further shrinkage might still occur. At the end of the treatments, only approximately 25% of the treated fibroid was non-perfused (indicating successful ablation), which might explain the modest tumor shrinkage observed. This modest shrinkage was, in part, dictated by the FDA, who required that wide outer margins of the fibroid should remain untreated. 75% of patients experienced no pain after the procedure, but mild to moderate pain during the treatment was reported by 66% of patients during treatment. A further 16% reported severe pain during treatment but this only persisted after treatment in 1%. Patients with larger treated fibroid volumes reported greater symptom improvement, and once restrictions on safety margins imposed by the regulatory bodies are lifted, it seems likely that as treatment volumes increase, so will symptom relief. Adverse effects must be closely monitored. In one treatment, leg and buttock pain developed immediately after treatment, and MR images showed the sciatic nerve to be in the far field of the ultrasound beam. However, there was no nerve damage, and the patient made a complete recovery. This led to a change in practice, and the focus was then kept at least 4 cm away from any major nerve bundles positioned near bone. The reports of the 6- and 12-month results from this Phase III trial showed that 71% of women at 6 months and 54% at 12 months had achieved the targeted 10-point reduction in the symptom severity scale of the Uterine Fibroid Symptoms Quality of Life questionnaire24. However, tumor shrinkage remained modest, which is probably why 28% of women sought alternative treatment at 12 months. 5% of patients suffered skin burns which were thought to be in areas of abdominal wall with incomplete hair removal. No urgent surgical interventions or bowel injuries were reported. Larger fibroids have been treated with a combination of MRgHIFU and a gonadotropin releasing hormone (GnRH) analog. These patients were excluded from earlier trials because of the time required to treat them with HIFU alone. GnRH is known to reduce fibroid size. Patients with fibroids > 10 cm in diameter were given GnRH analogs for 3 months prior to MRgHIFU to reduce the fibroid size. This group achieved a significantly greater reduction in fibroid volume than did the control group (MRgHIFU alone). 89% of patients pretreated with GnRH analogs had a > 10-point reduction in symptom scores at 12 months, and an average reduction in fibroid volume of 37%25, 26. Thus, HIFU treatment of uterine fibroids appears to be a safe treatment option. As treatment delivery is refined, and larger volume fractions are ablated, it may find its place amongst other, more conventional, therapies. Randomized control trials are needed to determine the patient groups that would be best suited to such a non-invasive treatment method. One of the easiest organs for HIFU access is the breast. The acoustic window is unimpeded by bone, tissue movement is not a problem, and ultrasonic and MR imaging of the breast is well-established. If this non-invasive HIFU technique proves to be useful in the treatment of breast cancer, it will have the added advantage of excellent cosmesis, as surgery will have been avoided. Fibroadenomas and malignant breast tumors have been treated using MRgHIFU14, 15. Of 20 patients with breast cancer treated with HIFU, 53% had negative core biopsies 6 months later and 66% of the remainder were tumor-free following retreatment15. In a study of breast cancers following radical mastectomy 7–14 days after HIFU treatment, histology showed complete coagulative necrosis of the tumor and the 1.8 ( ± 0.6)-cm normal tissue margin treated in all patients16. In their follow-up study, Wu et al.17 reported a 5-year disease-free survival rate of 95%. Apart from the possibility of inducing widespread volume ablation with these high-power, highly focused ultrasound beams, localized targeting of sites is possible. A number of applications of this type can be envisaged, such as, for example, the selective destruction of nerves for pain relief. It is now well-established that HIFU can be used to seal blood vessels, stopping flow within them27-32. This requires a higher power beam than that used to cause coagulation necrosis, with the range of intensities used being 400–6500 W/cm2. Not only does this have potential application, for example, in the management of bleeding following damage to the spleen or liver and in the sealing of blood vessels following injury or catheterization, but also, the selective occlusion of blood vessels may be useful in the treatment of fetofetal transfusion syndrome (with the shunt vessel between the fetuses being targeted) or in the treatment of cancer, where it might be useful to cut off feeder vessels to the tumor32-34. It seems clear that HIFU is finding useful clinical roles, although its application (and technical development) is still in its infancy. The potential to destroy a cellular volume to an accuracy of only a few cells, while leaving surrounding tissue unaffected, has considerable appeal. Future developments will involve speeding up treatments (currently it takes c. 1 hour to ablate 3 cm3 of tissue) and improving treatment targeting and monitoring. Apart from the uterine fibroid and breast cancer treatments, obstetric and gynecological applications have remained largely unexplored. Since HIFU may be applied without general anesthetic, and as a day-patient procedure, it seems evident that it could become an attractive option for this specialty.