Introduction
The emergence of focal therapy as a potential and major shift in the way we manage localised prostate cancer is a highly debated area of research in uro-oncology. Whilst the concept of the index lesion being the primary driver for progressive prostate cancer seems attractive and its ablation seems logical, the wider acceptance of focal therapy, even using such ablative modalities as cryotherapy and high-intensity focused ultrasound (HIFU) seems limited and questioned. This is especially an issue when regulatory approvals are being considered for device approvals or an approval for an additional indication including the focal ablation of a lesion [
1,
2]. However, focal therapy has been increasing in popularity, which is highlighted by the increasing number of focal cryotherapy cases registered in the Cryo Online Database (COLD). The number of focal procedures performed per year has increased from 1 in 1999 to 293 in 2007 [
3‐
5].
Currently, there is an array of treatment modalities being evaluated for the focal treatment of localised prostate cancer. Whilst the two in pole position are HIFU and cryotherapy, having been used for focal therapy for a longer period of time, others under investigation include photodynamic and vascular-targeted photodynamic therapy (PDT/VTP) [
6,
7], irreversible electroporation [
8], radiofrequency ablation [
9], magnetic thermotherapy [
10], convective thermal water vapour [
11], injectable toxins [
12] and focal brachytherapy [
13]. The aim of all these technologies is to treat the index lesion within the prostate whilst leaving behind healthy non-cancerous tissue or non-significant disease.
One of the significant areas of debate relates to how success or failure of focal therapy is assessed, regardless of which ablative modality is being delivered. The arguments for and against a number of outcome measures follow a common theme in all studies and considerations given for acceptance and approval for focal therapy. Established prostate-specific antigen (PSA) follow-up criteria used for whole-gland treatments such as ASTRO, Phoenix and the Stuttgart criteria, are difficult to apply to a patient after focal therapy as the untreated normal prostatic tissue will continue to secrete PSA. PSA kinetics and PSA nadir may play a role as PSA secretion from the cancerous lesion is larger than healthy prostate tissue, and Ahmed et al. [
14] noted an 80 % decrease in PSA at 3 months after focal HIFU. This decrease persisted at 12 months. The more solid endpoints of metastases and death would require over a decade of follow-up due to the long natural history of even clinically significant prostate cancer. Insisting on such outcomes for changes in clinical practice or regulatory approvals would inevitably stifle innovation. Not only would these findings have limited external validity, reported after 10–15 years but would be prohibitively expensive and resource heavy. Other outcome measures, such as histological outcomes, are clearly needed.
In this paper, we aim to first discuss the existing state-of-the-art with respect to histological outcomes after focal therapy and how best to interpret these in clinical practice. The literature is mainly based on the two most widely used modalities of HIFU and cryotherapy. Secondly, we will discuss histological endpoints from a trial design point of view that might be met with wider acceptance or at the very least, form the premise for further debate.
Discussion and focal therapy trial design
Knowing this data, the questions that need be asked before designing an appropriate trial are “when to biopsy?”, “How to biopsy?” and “How to interpret the results?”.
It is accepted that a well-designed RCT is high-level evidence upon which treatment decisions can be based. However, designing an appropriate prospective trial to assess the follow-up of patients after focal therapy carries its own challenges. Successfully randomizing men to different surgical treatments in prostate cancer can be challenging.
The initial entry point would be from diagnosis of disease. The methodology for diagnosis is important. Pre-operative TRUS biopsy has been shown to be inaccurate particularly when assessing patients for focal therapy. For instance, 23 % of patients having a template biopsy after a previous TRUS biopsy had upgrading, whilst 60 % were found to have bilateral disease [
41]. Systematic transperineal and MRI-guided biopsies have been shown to be equally accurate in detecting cancer [
42,
43]. Thus, a similar principle should be applied to post-treatment biopsies. Most still perform systematic TRUS biopsies after treatment that may under-sample the treated area and lead to detection of insignificant disease from untreated tissue.
To improve upon, these ultrasound–MRI fusion techniques have been developed. There seems to be potential benefit of this technique over systematic TRUS biopsies with 67 % more significant disease detected; however, the benefit is less clear when comparing cognitive MRI targeting versus a fusion technique [
44‐
46]. Wysock et al. [
45] did not show a significantly increased cancer detection rate between the two techniques, but did find improved accuracy for smaller lesions with fusion biopsies. Whilst Cool et al. [
46] showed a 100 % increased accuracy for sampling a clinically significant tumour with fusion targeting versus cognitive. Five millimetres of transperineal mapping biopsies is arguably still the gold standard though, with up to 95 % accuracy when compared to RP specimens [
47,
48]. Nonetheless, these require a general anaesthetic and due to the higher number of cores taken have a potentially higher side effect profile.
Subsequently, the clinical significance of disease needs to be reviewed before any treatment is offered. Gleason 6 disease has been shown to rarely if ever lead to metastases and thus death [
49‐
51]. Similarly, metastases are very rare with tumour volume <0.5 cc [
52]. Combining this data from results from AS series where few deaths occur, it can be reasoned that low-volume Gleason 6 disease may not benefit from interventional treatment [
36]. Thus, the ideal focal therapy candidate is likely to be men with Gleason 7 and above disease or high-volume Gleason 6 cancer.
These same factors play a role following focal therapy. If a patient’s disease has been accurately classified pre-operatively, then follow-up can consist of imaging and biopsies only targeted to the treated area. As previously mentioned, although various criteria exist for follow-up, all effectively conclude that a rising PSA can be used to guide further investigation for recurrence and timing of the post-treatment biopsy. An additional factor to consider is the use of mpMRI to follow up treated lesions and thus triggers a biopsy if residual or recurrent disease is detected.
After focal ablation, there is a residual inflammation, necrosis and eventual fibrosis. Along with some of the data in our own series, Biermann et al. [
53] found that detection of grading of cancer in biopsies taken 6 months post-HIFU was possible (unlike with radiotherapy); thus, it seems reasonable that biopsies can be performed at least as early as 6 months.
With respect to mpMRI a consensus meeting found that 77 % of panellists felt that this is a reliable tool for follow-up [
54]. From the series of 42 men by Ahmed et al., nine patients were found to have a positive mpMRI, seven of whom had subsequent positive biopsies. Overall, there is some evidence to support the use of mpMRI as a method for follow-up, with it appearing to form an important part of multi-modal follow-up [
55‐
58].
Subsequently, the clinical significance of post-procedural biopsies needs to be considered. There is limited evidence to suggest that low-grade low-volume residual in-treatment field cancer behaves in a similar way to patients with primary disease of the same pathology. Similarly, there is no data on the natural history of secondary low-risk lesions. As mentioned previously, the safest option would be to apply the same principles as for AS and follow up these patients with repeated imaging with biopsies as deemed necessary. From the reviewed cryotherapy and HIFU papers, redo focal treatment and AS appear to be the most common management options selected for patients with positive biopsies.
We subsequently reviewed the results of three recently published consensus statements released discussing many of these issues concerning focal therapy (Table
3) [
59‐
61]. All consensus meetings and expert opinions are considered as level five evidence, and their findings should be considered with this in mind. However, in the absence of high-level evidence, they present important opinions from experts in the field and highlight areas of uncertainty.
Table 3
Consensus meeting results
Multi-stage Delphi process | 3 + 3 with “substantial cancer” 3 + 4 >10-year life expectancy PSA > 15 with caution | Clinically insignificant disease (volume < 0.5 cc) | MRI targeted or fusion with systematic | Negative 12-month biopsies Targeted and systematic | 3 monthly but not sufficient as an end point | Alterations in MRI not sufficient as end point | Any cancer in-field of treatment Low-grade, low-volume (Gleason 6, <3 mm) out of field is not considered failure | Acceptable on one occasion |
RAND/UCLA appropriateness methodology | NCCN intermediate risk disease 3 + 3 with >3–5 mm MCCL >10-year life expectancy Multifocal disease included (secondary lesion of Gleason 6, ≤5 mm can be left untreated) | <5-year life expectancy and those <40 and >80 years with caution WHO performance status 3–4 | MRI targeted or template mapping (if no MRI available) | 12-month biopsy Targeted only Uncertainty about systematic sampling | Rising PSA may trigger biopsy | Suspicious MRI may trigger biopsy | Cancer in field of equivalent or higher than pre-operative grade Low-grade, low-volume (Gleason 6, <3 mm) in field is not considered failure | ≤20 % retreatment rates |
Delphi process | | | | 12 month biopsy Targeted and systematic | 3 monthly No consensus on role of PSA | 1st MRI 6 months post-treatment Suspicious MRI should lead to biopsy Further biopsies after 12 months only if suspicious MRI | | |
Donaldson et al. and Van den Bos et al. separately discussed inclusion and exclusion criteria and both consensus groups agreed that high-volume Gleason 6 (>5 mm MCCL) and Gleason 7 disease are the optimal candidates. They discussed the importance of accurate pre-operative assessment using MRI-targeted or fusion biopsy methods prior to offering a patient treatment. They recommend protocol biopsies at 12 months post-treatment and that a rising PSA or suspicious MRI should also trigger a biopsy. The ultimate end point was a negative 12-month biopsy. However, there was differing opinion on whether these should only be targeted or also systematic or whether residual low-grade and low-volume cancer is considered treatment failure. Donaldson et al. mentioned that biopsies should be targeted rather than systematic in order to reduce sampling of untreated tissue, whilst Muller et al. (another consensus group) commented that systematic biopsies are useful for surveillance of the untreated gland. Re-treatment was considered acceptable and Donaldson et al. commented that overall re-treatment rate should be below 20 % since it was argued that retreatment was a positive attribute of the strategy.
The final point to consider when designing such a trial is the comparator arm. The two most common radical whole-gland treatments are radiotherapy or RP. Neither has been assessed head-to-head in a RCT. Retrospective reviews have shown surgery to have better oncological outcomes; however, even with the best matching, bias can never be fully excluded [
62,
63].
The role of focal therapy will ultimately help decide the most appropriate comparator arm. We feel that the current role of focal therapy is similar to that of tissue preservation in almost all other solid organ cancers (bar ovarian), in that, it leads to a minimal decrease in quality of life for the patient whilst providing acceptable cancer control. At least in the short to medium term, the data suggest that focal therapy may be meeting these aims. However, few studies have long-term follow-up and a systematic review showed a varying bDFS of between 86.2 % at 8 years and 60 % at 5 years [
64].
Thus, in the first instance, patients with localised intermediate- to high-risk disease should be compared to those undergoing radical treatment with either EBRT or RRP. The aim of this study would be to show non-inferiority for short- to medium-term oncological outcomes with a superior side effects profile. Repeat treatments are possible with focal therapy, and a second treatment would not be considered a treatment failure unless there was a significant reduction in the two outcome measures mentioned.
Acknowledgments
M. Emberton and H. U. Ahmed would like to acknowledge funding from the Medical Research Council (UK), the Pelican Cancer Foundation Charity, Prostate Cancer UK, St Peters Trust Charity, Prostate Cancer Research Centre the Welcome Trust, National Institute of Health Research-Health Technology Assessment Programme and the US National Institute of Health-National Cancer Institute. M. Emberton receives funding in part from the UK National Institute of Health Research UCLH/UCL Comprehensive Biomedical Research Centre. M. Emberton and H. U. Ahmed receive funding from USHIFU, GSK and Advanced Medical Diagnostics for clinical trials. M. Emberton is a paid consultant to Steba Biotech and USHIFU. Both have previously received consultancy payments from Oncura/GE Healthcare and Steba Biotech.