American Society for Therapeutic Radiology and Oncology, EBRT, external beam radiotherapy
biochemical disease-free survival.
In recent years, high-intensity focused ultrasound (HIFU) has been harnessed as an energy source to destroy tissue with increasing precision. The ability to apply ultrasound transrectally to the prostate gland for both visualization and therapy has led to the development of this technology for treating localized prostate cancer, with clinical follow-up data now reaching 10 years. Attractions include a treatment that can be carried out in a day-case setting with a PSA nadir reached 3–4 months after the procedure. HIFU has been applied for the primary treatment of localized disease and more recently as a salvage therapy for patients in whom radiotherapy has failed.
Ultrasound energy causes mechanical stress, thermal stress (from absorption of the ultrasound waves by the tissue), and cavitation effects (cavitation is the development of bubbles followed by implosive collapse and local heating). The overall result is that ultrasound energy is transformed into intense local heat (≈ 85 °C) with subsequent coagulative necrosis of tissue. There is a sharp decline in energy around the lesion, allowing precise targeting .
Lesions are created from a focal point that is remote from the probe, and propagate into an ellipse of a similar size and shape as a long grain of rice. Structures which are in the pathway from probe to focal point, such as the rectal wall, are undamaged. Lesions (350–1000 per gland, depending on the size) must be continually produced next to each other to cover the entire prostate. Energy is related to the frequency, power, and duration of each pulse.
Animal studies using the highly metastatic Dunning R3327 model in rats showed that HIFU could be used to destroy prostatic carcinoma without causing metastasis [2,3]. Subsequent canine studies showed that foci could be applied transrectally to the prostate without damaging the rectal wall . Initial in vivo human studies showed defined margins of tissue necrosis in benign tissue before a Millin prostatectomy , and studies on cancerous prostates several days before radical prostatectomy showed delineated areas of coagulative necrosis in the treated areas [6,7]. HIFU is repeatable, and it is not uncommon for more than one treatment session to be needed for a satisfactory response. The gland must be small enough for the anterior part to be reached by the lesion (<40–50 mL), and with no large calcifications (>5 mm) that might interfere with the ultrasound signal.
In 1996, Gelet et al. reported their early experiences with HIFU for prostate carcinoma, treating 14 men who had T1-T2 disease deemed unsuitable for radical surgery. Half of these men achieved a PSA level of <4 ng/mL and negative biopsies after HIFU.
Two commercially available devices are currently in use for HIFU treatment of prostate cancer; both have undergone modifications and technical improvements since first produced, and are likely to continue to do so. A common principle is to allow simultaneous imaging with treatment, using a probe housed within a degassed fluid-filled balloon that cools the rectum. The main differences are in the patient positioning and the degree of manual control of power. The prostate is divided into ‘blocks’, with each block planned and then treated by computer-driven probe movements and firing.
The SonablateTM (Focus Surgery, Inc., Indianapolis, IN, USA), mostly trialled in Japan, allows treatment in the lithotomy or supine position on a standard operating table. Various transducers are available with differing focal lengths (30–40 mm), depending on the prostate size to be treated. A single piezoelectric crystal allows both imaging and treatment (4 MHz) with real longitudinal imaging available. Lesions are 10 mm long and 3 mm wide. Treatment occurs in layers of 10-mm thickness from apex to base in typically three target zones: anterior, mid and posterior (Fig. 1) , and the operator has some control over the power of each lesion. Degassed water circulates within the probe balloon to cool the rectum, and a urinary catheter is placed at the end of the procedure.
The Ablatherm® (EDAP SA, Lyon, France) has been widely used in Europe. It has a dedicated treatment bed that incorporates a robotic arm into which a specially designed endorectal probe is inserted (Fig. 2). The probe incorporates a firing transducer (3 MHz) with a focal distance of 45 mm and an imaging transducer (7.5 MHz), and is surrounded by a specially designed degassed fluid (Ablasonic®) in a balloon. Captured transverse images are computed into longitudinal images. The patient lies in the right lateral position for treatment, which allows gas bubbles to remain out of the firing line before being removed in the circulating Ablasonic fluid. This fluid is cooled in a 5 °C bath, producing a peri-probe temperature of 16–18 °C to cool the rectal wall as protection against inadvertent heating. A further safety feature is an infrared motion detector that stops the treatment if triggered. The lesion size can be altered by the operator from 19 to 24 mm long, with a constant diameter of 1.7 mm. The current technique incorporates a TURP immediately before treatment and during the same anaesthetic; this reduces the indwelling catheter time after treatment and debulks the gland for technically improved treatment . The TURP is followed by HIFU treatment of the gland in six treatment blocks, or four if small or as a salvage treatment (Fig. 3). A urinary catheter is placed at the end of the procedure.
Table 1[9,11–16] summarizes published data from clinical studies on HIFU for prostate cancer. For the Sonablate, Uchida et al.[9,11] reported on their own experience and on a multicentre study in Japan comprising 72 men with previously untreated T1c-T2 N0m0 disease, a PSA level of <20 ng/mL and prostate volumes of <50 mL. Most of the men (55/70) had a Gleason score of 5–7. The median follow-up was 14 months and biopsies were taken at 6 months; 68% of these were negative. The biochemical disease-free survival (BDFS) predictions (based on the definition of the American Society for Therapeutic Radiology and Oncology, ASTRO) were 78% at 1 year and 76% at 2 years. The other study had a longer follow-up of 22 months and a similar group of patients but five of the 63 had a PSA level of >20 ng/mL; 72% of the men achieved a PSA nadir of <1.0 ng/mL at 4–8 weeks after treatment, and this predicted BDFS. The median operative duration was 149 min, and the men were catheterized urethrally for a median of 14 days.
Table 1. A summary of pertinent clinical studies of HIFU for prostate cancer; PSA values are ng/mL
The Ablatherm device is associated with several studies across Europe; the European multicentre study recruited 652 patients from six centres between 1995 and 2000; 402 of these, who were unsuitable for radical surgery, were treated with monotherapy for T1-T2 localized disease using Ablatherm HIFU. This group was analysed in 1999 and the results published in 2003 . Several modifications were made during this time, including increasing the treatment frequency from 2.25 to 3 MHz, increasing the shot duration from 4.5 to 5 s, and reducing the planned number of treatment sessions from two to one. Of 288 men for whom biopsy data were assessable, 87% had negative sextant biopsies. In all, 212 patients reached 6-month follow-up and a median PSA level of 0.6 ng/mL. A lower PSA nadir was associated with a prostate volume of <40 mL, complete treatment rather than partial, and modern treatment protocols. The biopsy results and PSA nadir showed no significant differences between stratified risk groups.
Gelet and colleagues in Lyon have extensive experience with Ablatherm HIFU; in 2001  they reported on 102 patients with a mean follow-up of 19 months, including eight in whom external beam radiotherapy (EBRT) had failed, and eight who had been treated previously with neoadjuvant hormonal therapy. Failure was defined as positive biopsies or three successive PSA rises with a PSA velocity of ≥ 0.75 ng/mL/year. Overall, in 66% there was no failure. Significant prognostic indicators included a PSA level of ≤ 10 ng/mL (73% survival vs 50%); Gleason score ≤ 6 (81% vs 46%); and four or fewer positive sextant biopsy samples (68% vs 40%). A further study from that group in 2003  considered 120 men with localized disease and a PSA level of ≤10 ng/mL; 77 had a Gleason score of 2–6. Amongst this latter group, the 5-year Kaplan–Meier predicted success rate using the same criteria was 85% (77% overall, 61% for Gleason score ≥ 7). The 7-year results for 190 men (mean follow-up 40 months) with a PSA level of ≤ 15 ng/mL and Gleason score ≤ 7 showed a 61% success rate with the above criteria and a PSA level of <1 ng/mL .
Blana et al. published their 5-year experience of 146 patients (presenting PSA level of <15 ng/mL and Gleason score ≤ 7), reporting a mean PSA nadir of 0.07 ng/mL at 3 months; 87% of the men had a constant PSA level of <1 ng/mL and 93% had negative control biopsies. However, 43% of this group had neoadjuvant hormonal treatment. The Kaplan–Meier estimate of 5-year disease-free survival, when failure was defined as a positive control biopsy and/or a PSA increase of >0.2 ng/mL, was 54%, with the curve flattening by 30 months. Operating times were similar to the Sonablate experience  and the mean suprapubic catheter time was 13 days.
Prolonged urinary catheterization is an issue with both the Sonablate and Ablatherm approaches to HIFU because of obstruction caused by oedema and necrotic tissue; suprapubic catheters are associated with longer indwelling times than urethral catheters (Table 1). Up to a third of patients have required surgical intervention after HIFU . A TURP immediately before HIFU has the potential to reduce this morbidity, remove calcification, and reduce the prostate size and anteroposterior diameter, making HIFU technically easier; the defect is compressed by the endorectal balloon and haemorrhage is reduced by the HIFU treatment . In a series of 30 cases, the mean total operating time for TURP plus Ablatherm HIFU was 2.8 h . Chaussy and Thuroff  reported on 96 patients who had HIFU alone and 175 who had TURP and HIFU; the mean resection volume was 15.7 (2–110) g. There were no differences in disease control in terms of PSA nadir, control biopsies or biochemical failure. A quarter of the HIFU-only group required re-treatment, compared to 4% for those who had TURP. A previous TURP reduced the mean indwelling suprapubic catheter time from 45 days to 14 days, and UTI rates from 48% to 11%. Surgery after treatment, for bladder neck stenosis, was required in 8% who had had a previous TURP (vs 27%), and the mean IPSS improved from 6.7 to 3.4. TURP before ablation is now recognized as part of the standard protocol for Ablatherm HIFU.
Most patients have mild LUTS after HIFU, partly due to necrotic debris, but without lasting change in IPSS [9,11,14,15]. UTIs are common, but less so in patients who have a TURP as part of the procedure (11% vs 48%); this is probably due to the shorter catheterization time. Similarly fewer patients receive antibiotic prophylaxis (17% vs 58%) [16,19].
Recto-urethral fistula is probably the most devastating of all complications associated with transrectal HIFU and is reported in up to 2%[9,12–15] for primary treatments, and more frequently when used after radiotherapy failure (6%) . Many cases are treated with prolonged catheterization or fibrin glue, but some, particularly after radiotherapy, require stoma diversion of urine and feces. Many of the fistulae reported occurred before the introduction of rectal cooling and rectal safety margins with autodetection-driven alarms, and in patients who would not now be offered HIFU because of abnormal rectal anatomy or a rectal wall that is too thick (>6 mm) to be cooled effectively. It is expected that this complication will reduce significantly in the future . Chaussy and Thuroff  found a reduction from 3.5% in 96 treatments before 1997 to 0.5% after the introduction of safety features.
Stress incontinence is related to the treatment of apical tissue. It is usually transient, but a substantial proportion of men in early series required artificial sphincters, e.g. four in the European multicentre trial . By incorporating a 5-mm safety margin, which is now standard in Ablatherm procedures, Chaussy and Thuroff  reported a reduction from almost a quarter of treated men to 3.9% with mild stress incontinence. Although some are concerned that the heat diffusion from the lesion might not be enough to fully treat the gland in an area where positive margins are often found in surgical experience , the authors assert that there was no increase in apical residual cancer.
Erectile dysfunction affects two-thirds of men who are potent before treatment . A nerve-sparing protocol is described for men who only have positive biopsies on one side, in which a 5-mm lateral margin is not treated on the contralateral side; this halves the rate of erectile dysfunction. However, the required re-treatment rate (due to positive control biopsies) is increased by 15%.
Other complications include transient perineal pain , and epididymitis, balanoposthitis and prostatitis, in the Sonablate experience [9,11].
HIFU AS SALVAGE TREATMENT FOR RADIOTHERAPY FAILURE
Gelet et al. reported a series of 71 cases of radiotherapy failure that were treated with salvage HIFU, in Paris and Lyon. All had biochemical failure after primary treatment, and histological confirmation was by biopsies (34 showed poorly differentiated disease). The evaluation of the patients established negative nodal and metastatic status; 30% received hormonal therapy, which was stopped at the time of HIFU treatment. The mean (range) follow-up was 15 (6–86) months with 3-monthly PSA tests and control biopsies at 3 months. Of the patients, 80% had negative biopsies (73% actuarial at 30 months) and the median PSA level was 0.2 ng/mL. The predicted 30-month actuarial disease-free rate, defined as negative biopsies and absence of biochemical failure using the ASTRO criteria, was 38%. Just over half of the patients required adjuvant therapy (hormones or hormones plus chemotherapy), nine had metastatic disease and four died from prostate cancer. There were four cases of recto-urethral fistula, although the authors state that these occurred early in the series before specific features were introduced to prevent this severe complication, and none have occurred since . There were 12 episodes of bladder neck stenosis, treated with urethrotomy, and incontinence was experienced by 25 patients, four of whom required artificial sphincters.
In an update on 118 men with a mean follow-up of 16.4 months, disease progression (defined as the requirement for salvage androgen deprivation) was strongly correlated with the initial Gleason score before HIFU. Survival rates were 58%, 44% and 14% in patients with Gleason scores of ≤ 6, 7 and ≥ 8, respectively .
Published clinical data on HIFU for prostate cancer are limited to case series with a generally short mean follow-up; many reports include the same patient datasets at different times and overlap into multicentre studies. Clinical outcomes are generally assessed by biopsy results after HIFU or by biochemical progression, neither of which guarantee a cancer-free status. When the ASTRO definition for BDFS is used (designed specifically for radiotherapy follow-up), three consecutive rises in serum PSA are required with ≥ 3 months between each sample and subsequent backdating of the failure; it is recommended that ≥ 2 years of observation is required for this to be meaningful and to reduce artefactual flattening of the Kaplan–Meier curves . Longer-term results are needed for Kaplan–Meier estimates of the ASTRO-defined failure in HIFU for prostate cancer.
It is disappointing that no further results have been published from the European multicentre trial after the interim analysis in 1999, as this would provide an impressive dataset with 10 years of follow-up, and more recent data subsequent to technological developments. Further clinical studies with a longer follow-up and consistent reporting of data will hopefully provide a more objective analysis of the effectiveness of this treatment. Unfortunately this is currently lacking.
Table 2 provides a brief assessment of the suitability of patients for HIFU.
Table 2. Suitability criteria for HIFU as primary treatment
Normal, rectal wall ≤ 6 mm
<5 mm (can be removed by TUR)
T1-T2 N0 M0
Can manage 4 h anaesthetic – spinal or general
Suitability for prostatectomy
Unsuitable or refuses
Current research is assessing the value of ‘visually directed HIFU’, i.e. adjusting the power levels of treatment to control the delivered energy based on ultrasonographic appearances, and the use of HIFU as a salvage treatment for patients who fail permanent prostate brachytherapy. Improved understanding of the effects on tissues and how to control these, with optimal targeting of lesions and rigorous data collection, has the potential to translate into an effective treatment for prostate cancer in the future.
With developments in protocols, hardware and software, the safety profile of HIFU for prostate cancer appears to be improving and is becoming well tolerated. However, current data are not long-term enough to justify its use as a replacement for well-established curative treatment options (i.e. EBRT, brachytherapy and surgery). The short learning curve needed (10 cases for those familiar with TRUS of the prostate)  might be tempting to the ‘office urologist’, but it should be borne in mind that European groups reserve HIFU for men who are unsuitable for or refuse radical surgery. HIFU might have a place in deferring hormonal therapy, but there are no data to support this at present. As a salvage treatment after radiotherapy failure, HIFU seems to be a promising option.