Best Current Practice and Research Priorities in Active Surveillance for Prostate Cancer—A Report of a Movember International Consensus Meeting

Background: Active surveillance (AS) is recommended for low-risk and some intermediate-risk prostate cancer. Uptake and practice of AS vary signiﬁcantly across different settings, as does the experience of surveillance—from which tests are offered, and to the levels of psychological support. Objective: To explore the current best practice and determine the most important research priorities in AS for prostate cancer. Design, setting, and participants: A formal consensus process was followed, with an international expert panel of purposively sampled participants across a range of health care professionals and researchers, and those with lived experience of prostate cancer. Statements regarding the practice of AS and potential research priorities spanning the patient journey from surveillance to initiating treatment were developed. Outcome measurements and statistical analysis: Panelmembersscoredeachstatementon a Likert scale. The group median score and measure of consensus were presented to participants prior to discussion and rescoring at panel meetings. Current best practice and future research priorities were identiﬁed, agreed upon, and ﬁnally ranked by panel members. Results and limitations: There was consensus agreement that best practice includes the use of high-quality magnetic resonance imaging (MRI), which allows digital rectal examination (DRE) to be omitted, that repeat standard biopsy can be omitted when MRI and prostate-speciﬁc antigen (PSA) kinetics are stable, and that changes in PSA or DRE should prompt MRI ± biopsy rather than immediate active treatment. The highest ranked research priority was a dynamic, risk-adjusted AS approach, reducing testing for those at the least risk of progression. Improving the tests used in surveillance, ensuring equity of access andexperience across different patients andsettings, and improving information and communication between and within clinicians and patients were also high priorities. Limitations include the use of a limited number of panel members for practical reasons. Conclusions: The current best practice in AS includes the use of high-quality MRI to avoid DRE and as the ﬁrst assessment for changes in PSA, with omission of repeat standard biopsy when PSA and MRI are stable. Development of a robust, dynamic, risk-adapted approach to surveillance is the highest research priority in AS for prostate cancer. Patient summary: A diverse group of experts in active surveillance, including a broad range of health care professionals and researchers and those with lived experience of prostate cancer, agreed that best practice includes the use of high-quality magnetic resonance imaging, which can allow digital rectal examination and some biopsies to be omitted. The highest research priority in active surveillance research was identiﬁed as the development of a dynamic, risk-adjusted approach. (cid:1)


Introduction
Active surveillance (AS) for localised prostate cancer aims to defer or avoid treatment and its side effects, using repeat testing to detect disease that would benefit from treatment. The uptake and practice of surveillance vary significantly between different countries [1] and settings [2,3].
This variation starts with differences in the approach to prostate cancer detection, which impacts the number and proportion of men diagnosed with lower-risk prostate cancer who could be considered for AS. In the UK, where prostate-specific antigen (PSA) testing can be requested by patients but is not offered routinely, magnetic resonance imaging (MRI) before biopsy was performed in 80% of men in 2016/2017 [4], and biopsy is often omitted in men with negative MRI and low PSA density. The 2018-2019 National Prostate Cancer Audit data show that low-risk prostate can-cer comprises 6% of new diagnoses [5]. However, in the USA, where PSA testing is much more widespread, low-risk disease comprises one-third of new diagnoses, and Gleason 6 disease accounts for more than half of the diagnoses of localised prostate cancer in a series assessing the risk profile of prostate cancer at diagnosis over 10 years [6]. The United States Preventative Services Taskforce recommendations in 2008 [7] led to a reduction in the proportion of men seen with low-risk disease, according to the National Cancer Database, with a reduction from 38% in 2004 to 27% in 2014.
The biopsy and PSA thresholds for entry to AS vary, and there is also significant variation between published protocols and usual practice, particularly in the uptake of routine biopsies after the 1st year or 2nd year of AS. Although all established guidelines include the use of clinical staging by digital rectal examination (DRE), it is common in some centres for MRI to be used for staging, with DRE not done where E U R O P E A N U R O L O G Y O N C O L O G Y X X X ( X X X X ) X X X -X X X MRI is used. We know that there is significant variation in adherence to different protocols [8].
The use of MRI, and subsequent MRI-targeted biopsies, before enrolment and during follow-up differs significantly between different settings. Even for those who regularly use MRI, there are no agreed MRI features that denote magnetic resonance characteristics unsuitable for surveillance.
Movember is a global men's health charity that has invested in prostate cancer research since 2005. Movember's investment in AS to date has largely focussed on the Global Action Plan (GAP3) Prostate Cancer Active Surveillance consortium and database [9][10][11][12][13].
Movember commissioned this work to identify best practice and research priorities in AS, using a recognised structured consensus framework.

Patients and methods
The aim was to explore consensus for best practice in AS and then identify the highest research priorities. We used expert experience to explore the gap between published guidelines and common practice.
The University of Maryland initiative [21] focussed on designing novel research studies relevant to AS, using a Population, Intervention, Comparator, Outcome, and Timeline framework.
In order to build on these initiatives, key participants from each panel were invited to this consensus group (Supplementary Table 1). The unique approach of this project was to identify research priorities, for future investment by Movember, and the research community more widely.

Study design
The RAND-UCLA appropriateness method for assessing the agreement in areas of uncertainty [22] was used (Fig. 1). This includes the following: 1. An evidence review comprising an overview of recent systematic reviews in AS 2. Formation of an expert panel of health care and research professionals (HCP panel) and one of those with lived experience of prostate cancer (LE panel) 3. Iterative development of discussion statements informed by the evidence review and modified by the panels 4. Independent scoring of statements for the strength of agreement from 1 (strongly disagree) to 9 (strongly agree) 5. An analysis of group median scores (where 1-3 show disagreement, 4-6 uncertainty, and 7-9 agreement) 6. Calculation of group consensus (yes, no, or uncertain) based on the interpercentile range adjusted for symmetry [22] After scoring, panellists received a summary of their own results, group median scores and consensus status, and distribution of anonymised scores. Online discussions were held, where statements could be removed, added, or modified, prior to rescoring. Weighted summed scores were used to analyse these data.
Given the importance of hearing the distinct voice of those with lived experience, dedicated meetings were held for each panel, with an iterative process of sharing outcomes between the two panels ( Fig. 1).
An initial list of potential research priorities was identified from the literature. These were then discussed at each panel, following the scoring of the statements across patient journey. A revised list of ten research priority areas was developed, and all panel participants were asked to rank their top five priorities from these ten areas.

Setting and participants
The health care professional expert panel included expertise across urology, oncology, radiology, pathology, translational science, population health, psycho-oncology (including behavioural science), general practice, nursing, clinical trials, health care, and health services research. First, potential participants were purposively sampled via author listings on papers identified in the evidence review. We also aimed to include diversity of thought, gender, profession, and race. Additional participants were invited based on track record in AS research in areas complementary to the other panellists. Urologists represented a spectrum of those working within a more general practice with prostate cancer, to those with a specialist academic practice in prostate cancer, with a range of surgical options offered across the group including radical prostatectomy and focal treatment for prostate cancer ( Fig. 2 and Supplementary Table 1). The lived experience expert panel was purposively sampled to include participants across a range of experience of AS (new to AS, established on AS, previously on AS, and now had active treatment) and international representation. Many of the lived experience panellists participated as representatives of patient advocacy organisations and so could represent a broad range of views.

Results
The initial HCP survey contained 234 statements, formulated from the evidence review, with 117 of these sent to the LE panel.
The two panels removed a total of seven statements, added 74 statements, and rephrased 31 statements prior to final scoring. The additions were to address the panels' per-ceived gaps in the statements derived from the narrative review, and the rephrasing was necessary to clarify mean-ing. A total of 307 statements underwent final scoring and analysis across both panels.
The statements were discussed and are reported according to the patient journey from diagnosis to active treatment. The statements and panel scores are shown in Table 1.

3.1.
Who should be offered AS?
There was consensus and agreement across the two panels that a number of factors, including clinical factors, comorbidities, and patient preferences, should be used to determine whether someone should be offered AS. Gleason grade and MRI findings were the most important criteria, with PSA density and PSA being the next most important ones (Fig. 3A). There was much discussion on the use of other parameters, including cancer core length (in millimetres), tumour visibility, and tumour volume on MRI, none of which are recognised in established guidelines. There was agreement that percentage cancer in a biopsy core is a flawed concept as it does not consistently reflect absolute tumour volume and that numbers of positive cores should not be considered indicative of tumour burden in the context of MRI-targeted biopsies.

Age and life expectancy
There was consensus in the LE panel that no upper age limit (eg, 75 or 80 yrs old) should be applied, and there was uncertainty regarding a lower age limit. The HCP panel was uncertain about an upper age limit but agreed that there should be no lower age limit. There were also uncertainty and disagreement in the LE panel about whether life expectancy (eg, 10 or 15 yr) should be used as an eligibility criterion. Discussion centred on the need to offer a full set of choices to men depending on their likelihood of benefitting from treatment, avoiding artificial barriers based on biological age or estimated life expectancy, which can lead to age discrimination. The HCP panel agreed that life expectancy of 10 yr should be considered within the eligibility criteria for AS, with a shorter life expectancy requirement for intermediate-risk prostate cancer.

3.2.
Who is more likely to choose AS?
It was agreed that clinician confidence in AS, time to explain all the options, and the person or team informing patients of their diagnosis increase the likelihood of a choice of AS.

Psychological impact of AS
The HCP panel agreed that AS can lead to anxiety in some men, especially when awaiting test results, although the LE panel members were uncertain about this, suggesting that some men experience relief and positive psychological effects when able to have surveillance versus active treatment.
It was agreed that clinicians should aim to identify those with low-risk disease who might have, or be at risk of, negative psychological consequences of AS and offer psychological support, rather than immediate active treatment for the disease.

3.4.
How should AS be carried out?

Communication, education, and support
There was strong agreement that newly diagnosed patients should be offered separate appointments for diagnosis and decision-making, allowing time to process feelings, learn about the disease, and think of relevant questions. There was also strong agreement that better, clearer, and more consistent information is an unmet need for patients and their partners and families, that support groups can be an important support for men on AS, and that there is a need for clinician education to help men stay on AS when it is warranted.
There was strong agreement across both panels that patients should have an open line of communication with the health care team.
There was agreement in the LE panel that well-being, diet, exercise, and mindfulness are important during AS. There was also consensus in the HCP panel that men on AS are more likely to die from cardiovascular disease than prostate cancer and should be counselled about lifestyle modifications, with favourable results in AS populations (eg,            There was strong agreement that, once the AS decision has been made, active treatment should not be offered routinely at each subsequent visit, as this causes unnecessary anxiety and increased rates of acceptance of active treatment when tests results are stable. There was agreement in the HCP panel that interventions designed to improve trust in AS should be utilised to improve adherence to AS.

Triggers for investigation and treatment
There was HCP agreement that PSA should be monitored every 3-6 mo, and that this could be done in primary care if there are mechanisms and protocols in place to facilitate shared care. There was consensus disagreement with the use of routine DRE but a recognition that DRE may be done for reasons other than assessment of disease progression.
There was strong agreement that DRE is unnecessary if multiparametric MRI or other routine imaging (eg, transrectal ultrasonography) is being carried out during AS.
There was agreement that MRI should be done routinely during surveillance, but less frequently than annually. Triggers for requesting MRI include a significant PSA rise, although no agreement about a numerical threshold for this was reached. There was consensus agreement that prebiopsy MRI should be performed before any reclassification biopsies and that MRI-targeted biopsy should be added to standard cores when there is an MRI target. There was agreement that MRI quality should be reported (eg, PIQUAL criteria [23]) and that repeat MRI on AS should be reported in a standardised manner (eg, PRECISE [24]). There was agreement that MRI targeting can cause ''grade inflation'', although there was HCP uncertainty that this would lead to overtreatment.
There was agreement that a change in PSA kinetics, PSA density, or DRE should lead to MRI ± biopsy, rather than a treatment discussion. There was also agreement that further research is warranted into the role of PSA velocity as a noninvasive predictor of histological progression.
There was overall agreement that a confirmatory biopsy should be done in AS, but that negative MRI, in combination with other stable negative predictors (eg, stable PSA and low PSA density), can support the decision to omit additional prostate biopsies, at least on an individual basis with adequate counselling.
It was agreed that repeat biopsy should not be performed annually, and there was uncertainty that it should be done routinely, with recognition that protocols that include scheduled biopsies are more likely to have higher rates of conversion to active treatment.

3.5.
Switch to active treatment Both the HCP and the LE panel agreed that switching to active treatment should be done on the basis of a combination of changes in test results and patient discussion, although there was agreement that upgrading to Gleason 4 + 3 should act as a trigger for recommending active treatment. The ranking of factors determining the need for active  The bold text indicates discordance between HCP and LE (group medium category/group consensus); italics indicate additional statements or modifications proposed live during consensus meetings; and italics *NEW indicates additional statements proposed during review period and R1.
treatment was similar to that for determining eligibility (Fig. 3B).

Graduation to watchful waiting
There was HCP agreement that those who reach a point where they would no longer be recommended treatment for localised disease should be graduated to less intensive monitoring or watchful waiting (WW), although not based on age or life expectancy alone. The LE panel was in consensus disagreement that graduation to WW should occur at age 75 or 80 yr or for life expectancy of <10 yr, with uncertainty about whether this should occur for life expectancy of <5 yr.

Research priorities
The panel discussions on the AS pathway helped identify, shape, and reach agreement on the final list of ten AS research priorities, across the patient journey (Fig. 4). The most important priority for research, across both the LE and the HCP panel, is the development of a personalised, dynamic, risk-adapted approach according to an agreed framework, with less testing in men at the lowest risk of progression. This risk-adapted approach contrasts with the current guideline-approved standardised approach.
There was a significant concern raised by the LE panel that a personalised approach might lead to patients being at the discretion of an individual clinician, leading to variable standards of care.
There were also concerns raised about the variation in offering AS across patients in the same health care system and across different health care systems, especially regarding access to MRI and newer biopsy approaches.

4.1.
Consensus agreement on best practice that differs from current international guidelines A number of differences between the current consensus opinion (often based on clinical practice) and current guideline recommendations were evident in this work: 1. DRE can be omitted when MRI is used routinely in AS. 2. In men with stable MRI and other stable parameters (PSA kinetics and density), consideration can be given to omitting routine biopsy, with discussion between the clinician and the patient. 3. A change in PSA or DRE (if done) should lead to MRI with the option of a biopsy after this, rather than immediate biopsy or discussion of active treatment. 4. A change to active treatment should be based on a combination of clinical parameters, and the discussion between the clinician and the patient rather than on any single parameter. 5. Men who are suitable for AS but experience, or are at a risk of, significant psychological consequences of AS should be offered additional support rather than immediate active treatment.
It may seem surprising that these findings differ from current guidelines. Part of the reason for this is that guidelines, particularly the more robust ones such as those from the EAU and the UK National Institute for Health & Care Excellence are, quite understandably, based on published data. They also often recommend what is most feasible across a wide range of settings, rather than the latest advances. This can mean that the guidelines take some time to catch up with the most modern practice, started at a small number of pioneering centres, which were well represented in this consensus panel.
The explicit point of expert consensus is to give opinion in areas where data may be less robust and in the process of changing, based on expertise of the consensus panel. The expertise aggregated here is broad, across those with clinical practice across a range of professions (urologists, oncologists, radiologists, pathologists and qualitative researchers) and including those with expertise by lived experience of localised prostate cancer, from those on AS, those who have had active treatment, and family members of those with localised prostate cancer.
Expert opinion is also able to draw on common practice that may not yet be reflected in published data and hence in subsequent guidelines. In addition, for some areas, for example, omission of DRE when MRI is performed, there Y X X X ( X X X X ) X X X -X X X are data from the area adjacent to AS that can be brought into the discussion; for example, diagnostic pathways that use MRI before biopsy may perform the MRI first and omit DRE, whilst still accepting an abnormal DRE in primary care as an entry route to the pathway. The clinical relevance of this expert consensus process is that the wider community can be reassured that common practice (eg omission of DRE when MRI is done) is supported by expert agreement.
In terms of modern risk stratification, this consensus process acknowledged that the use of Gleason grade and MRI findings, which give an indication of the volume and aggressivity of the tumour, is ranked as the most important criterion in determining the eligibility for and continuation on AS. This reflects a current gap in the guidelines where, whilst MRI is recommended before biopsy and therefore MRI data are widely available in clinical practice, the granular details of MRI data (eg tumour volume) are not specified in risk stratification systems yet. New risk stratification approaches are likely to incorporate MRI data to address this, but widespread data on MRI parameters including tumour volume need to be collected and published.
The use of DRE ranked lowest for both determining eligibility for and continuation on AS, due to its poor positive predictive value and impact on the patient.
In fact, the use of DRE to initiate either additional tests (such as biopsy or MRI) or a treatment choice scored lowest among all tests, including PSA density, which is not commonly included in any of the guidelines. When community practice of AS is considered, according to both European and US studies, the reduction in biopsy frequency as men progress through surveillance can clearly be seen [1,25], as both patients and clinicians see less value in biopsies where PSA, DRE, or MRI findings are stable. This expert consensus is able to recognise this.

Recognition of wider social and psychological factors
Guidelines often concentrate on the clinical aspects of disease management, whilst this consensus process explicitly sought to acknowledge important social and psychological determinants of entry into and maintaining participation in AS. Having a dedicated lived experience panel was particularly helpful in this regard. The concept of ''surveillance fatigue'' was discussed where men and their families become tired of the process and uncertainty of surveillance, and choose active treatment despite the stability of objective disease parameters. It was acknowledged that the inclusion of routine biopsies in surveillance increases surveillance fatigue, as does difficulty in accessing communication with the health care team [8]. Kinsella and colleagues [26] showed that the use of an educational intervention reduced dropout rates at 5 yr, without evidence of progression from 41.5% at 5 yr to 21.7%. By contrast, an MRI-led AS programme, with no routine biopsies, showed dropout rates at 5 yr in the absence of progression at <1% [27], highlighting that education, support, and a more acceptable surveillance schedule all have an impact.
Similarly, this consensus process identified agreement that men experiencing adverse psychological consequences from surveillance should be offered additional support, rather than immediate active treatment.
The theme that communication and support are of paramount importance throughout surveillance emerged clearly. Furthermore, addressing both the need for high-quality, personalised psychological support, and communication and information gaps with clinicians, patients, families, and communities was identified as a key research priority.
One priority, that of working to address inequities, including self-identified race, age, and socioeconomic factors, to ensure inclusion and retention of all who are eligible to be offered AS, was felt by many expert panel members to be a priority that ought to be seen as ''cutting across'' or ''threading through'' all the others. For instance, improving our understanding of MRI must be approached in a way that takes into account potential disparities in access to the most high-quality imaging, both at diagnosis and later during surveillance.
Each of the identified research priorities will require specific development into a dedicated research project. The topmost priority, across both expert panels, is the development of a personalised, dynamic, risk-adapted approach according to an agreed framework, with less testing in men at the lowest risk of progression. This represents a fundamental break with guideline-based practice in AS, not just in terms of practice, but in terms of the philosophy of care as well. As such, this priority calls for a major programme of research, and Movember has set out a funding call to address this.

Limitations
There are many more experts in AS for prostate cancer than could be included, and a different panel may have had different outcomes. However a significant effort was made to accurately reflect diverse international thinking and practice across those countries where the conduct and practice of AS are reported widely. This included those with experience of different settings for managing patients with AS-in rural and urban settings, and those offering a broad range of treatments. Similarly, efforts were made to include men with different lived experience of AS, including active treatment.

Conclusions
The current best practice of AS includes the use of highquality MRI to be able to omit routine DRE, and to avoid routine repeat biopsy in those with stable MRI findings and PSA kinetics. In addition, for those settings where MRI access is limited, the use of MRI to determine the need for biopsy is favoured over standard transrectal biopsy promoted by a rise in PSA or a change in DRE.
The importance of an additional support for those on AS was recognised as an important alternative to immediate active treatment, especially for those with a significant risk of greater psychological impact of surveillance.
Working to address inequities, including self-identified race, age, and socioeconomic factors, to ensure inclusion and retention of all who are eligible to be offered AS was viewed as a research priority and as a potential theme that could inform work on all the other research priorities in the years ahead.
This report outlines future research priorities in AS agreed on by diverse health care professional and lived experience experts. Movember has published an expression of interest for participants in a collaborative research programme on the highest ranked priority of personalised, dynamically adapted AS in 2022.
Author contributions: Caroline M. Moore had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.