A Multicentre Study of 5-year Outcomes Following Focal Therapy in Treating Clinically Significant Nonmetastatic Prostate Cancer

,


Introduction
The therapeutic approach to nonmetastatic prostate cancer is an outlier compared to strategies for other solid organ cancers.Regardless of the burden or location of cancer within the gland, treatment is directed at the whole gland using surgery or radiotherapy.While both approaches are effective, they can be associated with urinary incontinence and erectile dysfunction, with radiotherapy occasionally causing rectal side effects [1][2][3].While current trends demonstrate that radical therapy is increasingly used among the patients most likely to benefit from treatment with respect to better cancerspecific survival [4], there is still a need to reduce treatmentrelated side effects, particularly as the survival benefit conferred by radical treatment is often seen over 10-15 yr when compared to a strategy of active monitoring [5].
The aim of focal therapy is to reduce side effects and maintain cancer control by targeting areas of known cancer in a similar approach to that for other solid organ cancers [6][7][8].This concept of tissue preservation has come about through improvements in disease localisation using multiparametric magnetic resonance imaging (mpMRI) and mapping biopsy techniques [9].However, many studies have hitherto been small, based in expert centres, or subject to short follow-up [10][11][12][13][14][15][16][17][18].The only randomised study to evaluate focal therapy recruited patients with low-risk disease [19], a population known to have little to no chance of metastases or cancer-related mortality.
We now report on medium-term cancer control outcomes in a large multicentre patient cohort with clinically significant nonmetastatic prostate cancer treated with focal therapy using high-intensity focused ultrasound (HIFU).

Institutional review board exemption was granted by University College
London Hospital for our health technology assessment programme, which followed guidelines for evaluating surgical interventions [20,21] and high-risk groups [24].
Disease was localised using mpMRI combined with targeted and systematic biopsies, or transperineal mapping biopsies.Targeted biopsies involved taking three to six biopsy cores from all lesions with a Likert score of 3-5, as well as systematic biopsies (transperineal or transrectal).Mapping biopsies involved taking biopsy cores every 5 mm.
Intermediate-and high-risk cases also underwent a radioisotope bone scan and/or cross-sectional computed tomography to rule out distant metastases, depending on local institution guidelines.
All surgeons underwent a rigorous period of training involving online learning, observation of five cases on two separate occasions in an expert centre, and on-site proctoring by an expert urologist for five to ten cases, and were followed by an expert clinical applications specialist for all subsequent treatments.
There were various forms of focal HIFU permitted (Fig. 1).Up to two retreatments with focal HIFU were allowed.All men were advised to undergo 3-6-monthly PSA testing, with mpMRI routinely performed at 1 yr and every 1-2 yr.Two rises in PSA after the nadir, without predefining the magnitude of the rise, were investigated using prostate biopsy or mpMRI followed by biopsy if the mpMRI was suspicious.We have previously reported on the high negative predictive value of mpMRI following focal-HIFU [23,25].
Repeat HIFU was offered when either (1) clinically significant cancer on biopsy occurred in field or out of field and mpMRI staging indicated that the disease was still localised, or (2) when mpMRI demonstrated clear in-field recurrence (mpMRI Likert score 5) associated with a rising PSA.Patients were routinely offered the option of radical prostatectomy or radical radiotherapy.
Physicians assessed postoperative adverse events during follow-up clinic visits.Functional outcomes were assessed via patient-reported outcome measures using validated questionnaires collected at 1-2 and 2-3 yr after focal HIFU treatment.Validated questionnaires included the International Prostate Symptom Score [26] and the Expanded Prostate Cancer Index Composite (EPIC) urinary continence domain [27].All data were audited and quality controlled by two data managers (N.M. and F.H.J.).

Primary outcome
Failure-free survival (FFS) was defined as avoidance of local salvage therapy (surgery or radiotherapy), systemic therapy, metastases, and prostate cancer-specific death.This excluded PSA kinetics, as there are no kinetic measures that are valid in this setting.
(98%) achieved complete pad-free urinary continence and none required more than 1 pad/ d.Limitations include the lack of long-term follow-up.Conclusions: Focal therapy for select patients with clinically significant nonmetastatic prostate cancer is effective in the medium term and has a low probability of side effects.Patient summary: In this multicentre study of 625 patients undergoing focal therapy using high-intensity focused ultrasound (HIFU), failure-free survival, metastasis-free survival, cancer-specific survival, and overall survival were 88%, 98%, 100%, and 99%, respectively.Urinary incontinence (any pad use) was 2%.Focal HIFU therapy for patients with clinically significant prostate cancer that has not spread has a low probability of side effects and is effective at 5 yr.© 2018 The Authors.Published by Elsevier B.V. on behalf of European Association of Urology.This is an open access article under the CC BY license (http://creativecommons. org/licenses/by/4.0/).

Secondary outcomes
Our main secondary outcomes included metastasis-free survival and prostate cancer-specific and overall mortality.We also report biopsy outcomes when carried out, as well as adverse events and side effects.
Urinary continence was defined as being completely pad-free and socially continent (0-1 pads/d).The latter definition is commonly used in many series reporting radical prostatectomy outcomes [28].We also report on complete pad-free, leak-free urinary continence.In addition, we evaluated whether certain baseline factors might predict for FFS.
Fig. 1 -Types of focal therapy using high-intensity focused ultrasound (HIFU) carried out in men with non-metastatic prostate cancer.

Statistical analysis
As this was a cohort registry study, there were no a priori sample size

Baseline demographics
A total of 625 patients were treated with focal HIFU, of whom 599 reached at least 6-mo follow-up and 505 (84%) had intermediate-or high-risk prostate cancer (Table 1).Table 2 lists the biopsy characteristics.When excluding patients with an event (n = 60) the median follow-up was 56 mo (IQR 35-70).

Secondary outcomes
Following focal HIFU, eight patients transitioned to salvage radical prostatectomy, 36 had salvage external beam radiotherapy, and one received androgen deprivation therapy.Ten patients experienced metastases during the follow-up, of whom two had low-risk, four had intermediate-risk, and four had high-risk disease; three had metastases after the second HIFU treatment.Kaplan-Meier estimates for metastasis-free survival at 1, 3, and 5 yr were 99.7% (95% CI 99-100%), 99% (95% CI 98-100%), and 98% (95% CI 97-99%), respectively (Table 3).There were seven deaths, none of which was related to prostate cancer.Kaplan-Meier estimates for overall survival at 1, 3, and 5 yr were 100% (95% CI 99-100%), 99% (95% CI 98-100%), and 99% (95% CI 97-100%), respectively.At least one repeat focal HIFU treatment was performed in 121 patients (one repeat HIFU in 112 and two repeat HIFU in 9).A total of 222 patients underwent biopsy following focal HIFU.Of these, 111 were within our three earlier phase 1/2 studies [13][14][15] and 111 had for-cause biopsies as a result of  rising PSA and/or suspicious mpMRI after focal HIFU.Overall, 29 men had histological in-field on biopsy, while 16 had histological evidence of out-of-field de novo cancer or progression of untreated low-risk disease.A further 11 patients had both in-field and out-of-field disease on biopsy.
No patients had bleeding requiring intervention or transfusion.Within 6 mo of treatment, postoperative urinary tract infection and epididymo-orchitis occurred in 53/625 (8.5%) and 12/625 cases (1.9%), respectively.Endoscopic interventions for lower urinary tract symptoms at any time point were required in 60/625 men (9.6%).There were two (0.3%) recto-urethral fistulae.One of these cases healed following urinary diversion using urethral and suprapubic catheters for 6 mo but no bowel diversion.
The other required open reconstructive surgery because of failure of conservative management (Table 4).
Baseline urinary continence status was reported by 421 men using the EPIC urinary domain questionnaire.At 1-2 and 2-3 yr after focal HIFU, pad-free status was available at both baseline and follow-up for 313 and 247 men, respectively; 304 (97%) and 241 (98%) were pad-free (0 pads) at these two time points.No men required more than 1 pad/d, so social continence was achieved in 100%.At 1-2 and 2-3 yr after focal HIFU, pad-free, leak-free status at both baseline and follow-up was available for 250 and 195 men, respectively; 208 (83%) and 156 (80%) were pad-free, leak-free continent at these two time points (Table 5).

Discussion
Our study shows that following focal HIFU, failure-free survival was 88% at 5 yr.Metastasis-free survival was 98% and cancer-specific survival was 100% at 5 yr.Only 2% had urinary incontinence requiring use of one daily pad and no men required more than 1 pad/d.Bowel complications were rare (0.3%).
The strengths of our study lie in its large and multicentre nature, with medium-term follow-up data prospectively collected in a nationally mandated registry and with independent quality control of data entry against source records.We predominantly treated patients with clinically significant prostate cancer, for whom it is widely accepted that treatment is required because of a higher risk of progression on active surveillance.
While it is accepted that whole-gland radical prostatectomy and radical radiotherapy are effective in treating clinically significant nonmetastatic prostate cancer, these modalities involve significant probabilities of urinary and sexual side effects, and some bowel toxicity in the case of radiotherapy.Focal therapy, by treating known areas of cancer with a margin, aims to preserve prostate tissue and minimise damage to the neurovascular bundles, bladder neck, external urethral sphincter, and rectum.We have shown that this strategy has a low probability of treatmentrelated side effects and achieves good cancer control in the medium term.Despite the absence of long-term follow-up, we find these results reassuring and acceptable considering that we predominantly used focal HIFU in intermediateand high-risk cancer.
Focal therapy using various energy modalities has been evaluated in single-arm retrospective and prospective studies.Recent systematic reviews have shown that focal therapy has a minimal impact on quality of life, and while oncological effectiveness is yet to be established, genitourinary function is well preserved [8,29].
Most of our low-risk cases were historical from a time when active surveillance for such disease was still questioned and it was deemed appropriate to offer these men focal HIFU as an alternative to radical therapy.Nonetheless, these cases still required a minimum amount of Gleason 6 cancer to qualify.Our UK focal HIFU programme now does not allow treatment in instances of low-volume Gleason 6 disease because the probability of progression for such lesions is rare, unless there are extenuating circumstances such as psychological distress to the patient due to being on active surveillance or family history [30].This stance contrasts with some other studies that have predominantly treated very low-risk or low-risk disease with focal therapy [16,17,19].
While long-term data are awaited, comparative studies among patients with intermediate-risk cancers are currently being piloted [31].However, because of previous problems in maintaining physician and patient equipoise in 11 failed randomised comparative trials comparing different interventions for localised prostate cancer, this might not be possible to deliver [32].Even if randomisation could be delivered, the sample size would have to be based on a noninferiority design that might require between 2000 and 8000 patients recruited, randomised and followed up over 10-15 yr, depending on the noninferiority margin used.
However, other treatments such as prostate brachytherapy and robotic prostatectomy have been approved for  clinical use without or before the completion of randomised comparative studies.Partial nephrectomy for the treatment of renal cancer another example.Such changes in practice were often based on medium-term outcomes from cohort studies because the long natural history of prostate cancer and small renal tumours made it unfeasible to deliver randomised controlled trials.In this context, physicians and health care organisations will need to consider whether it is justified to insist on randomised comparative data for focal therapy compared to radical therapy powered on mortality and metastases.On the basis of our data, patients currently diagnosed with prostate cancer that is suitable for focal therapy may prefer to have the option to choose wholegland radical therapy or focal therapy.

Limitations
We accept that there are some limitations to our study.First, since our prospective registry was embedded in clinical care, not all patients were routinely biopsied after treatment.We validated the role of mpMRI for follow-up after focal-HIFU compared to prostate biopsy and showed that the negative predictive value for mpMRI in this setting is 95% for clinically significant prostate cancer [25].Second, in light of not having a validated and accepted cancer control measure, we considered a clinically meaningful composite outcome measure that reflects the recent Intermediate Clinical Endpoint in Carcinoma of the Prostate consensus findings [33].Finally, validated questionnaire data were not available for all patients owing to our reliance on postal return of questionnaires.While we used the International Index of Erectile Function (5-item) questionnaire, these data were unavailable at the time of analysis.

Conclusions
Focal therapy using HIFU could be offered to select patients with clinically significant nonmetastatic prostate cancer as it is effective in the medium term and has a low probability of urinary and rectal side effects.This study was presented in May 2016 at the American Urology Association Annual meeting.
calculations.Our decision to analyse and publish these data was based on the registry being open for 10 yr and the availability of median follow-up of approximately 5 yr for the cohort.Baseline characteristics are presented as the median (interquartile range [IQR]) or proportion, as appropriate.Kaplan-Meier estimates of time-to-event outcomes are described with 95% confidence interval (CI) for those men with at least 6mo follow-up.Adverse events are reported as proportions only, and for these the entire cohort was included to reduce selection bias.Urinary continence was evaluated using the EPIC urinary continence domain with a pad-free definition and a socially continent definition (0-1 pads).

7 )Table 2 -
HIFU = high-intensity focused ultrasound; IQR = interquartile range; PSA = prostate-specific antigen.a Some cumulative percentages are higher than 100% because of rounding.Pre-HIFU biopsies Type of biopsy/detection Patients, n (%) Median number of cores, n (IQR) -intensity focused ultrasound; IQR = interquartile range; NA = not applicable/available. a Data missing for six patients.b Data missing for 12 patients.c Data missing for eight patients.

Table 1 -
Baseline characteristics for patients undergoing focal HIFU for nonmetastatic prostate cancer a

Table 3 -
Kaplan-Meier estimates of freedom from repeat HIFU, overall survival, metastasis-free survival, and overall failure-free survival following focal HIFU therapy among men treated for nonmetastatic prostate cancer