The primacy of multiparametric MRI in men with suspected prostate cancer

Background Multiparametric MRI (mpMRI) became recognised in investigating those with suspected prostate cancer between 2010 and 2012; in the USA, the preventative task force moratorium on PSA screening was a strong catalyst. In a few short years, it has been adopted into daily urological and oncological practice. The pace of clinical uptake, born along by countless papers proclaiming high accuracy in detecting clinically significant prostate cancer, has sparked much debate about the timing of mpMRI within the traditional biopsy-driven clinical pathways. There are strongly held opposing views on using mpMRI as a triage test regarding the need for biopsy and/or guiding the biopsy pattern. Objective To review the evidence base and present a position paper on the role of mpMRI in the diagnosis and management of prostate cancer. Methods A subgroup of experts from the ESUR Prostate MRI Working Group conducted literature review and face to face and electronic exchanges to draw up a position statement. Results This paper considers diagnostic strategies for clinically significant prostate cancer; current national and international guidance; the impact of pre-biopsy mpMRI in detection of clinically significant and clinically insignificant neoplasms; the impact of pre-biopsy mpMRI on biopsy strategies and targeting; the notion of mpMRI within a wider risk evaluation on a patient by patient basis; the problems that beset mpMRI including inter-observer variability. Conclusions The paper concludes with a set of suggestions for using mpMRI to influence who to biopsy and who not to biopsy at diagnosis. Key Points • Adopt mpMRI as the first, and primary, investigation in the workup of men with suspected prostate cancer. • PI-RADS assessment categories 1 and 2 have a high negative predictive value in excluding significant disease, and systematic biopsy may be postponed, especially in men with low-risk of disease following additional risk stratification. • PI-RADS assessment category lesions 4 and 5 should be targeted; PI-RADS assessment category lesion 3 may be biopsied as a target, as part of systematic biopsies or may be observed depending on risk stratification. Electronic supplementary material The online version of this article (10.1007/s00330-019-06166-z) contains supplementary material, which is available to authorized users.


Introduction
In 2012, the European Society of Urogenital Radiology (ESUR) prostate committee promoted the use of multiparametric MRI (mpMRI) in the routine management of men with suspected or confirmed prostate cancer [1]. That proposal has gained widespread acceptance. The debate has now moved to when mpMRI should be used.
Expressions of interest were sought from the 58 members of the ESUR Prostate MRI Working Group at the European Congress of Radiology (ECR) in March 2017 in contributing to a position statement on the use of mpMRI in prostate cancer diagnosis. Each of the 7 initial positive respondents was invited to contribute but based on the relative contributions, the final author list was revised to 5 ensuring due representation of the group's European composition; a sixth contributor joined at ECR, March 2018. The final contributors were from the UK, France, The Netherlands, Denmark, Italy, and Belgium.
The approach was to review published evidence, supplemented by knowledge of completed cohort studies in the process of being published (having been accepted for publication). In this way, there was very little intra-author disagreement. On the specific topic of biopsy planning-the pros and cons of systematic versus targeted-driven approach-there was some variation in how strongly the argument for targeted biopsy over systematic biopsy could be worded. As the paper neared completion, the results from on-going studies became known to the author group, such that a consensus position was reached.

Evaluating clinically significant prostate cancer
Evaluating clinically significant prostate cancer in the pre-MRI era Urologists and oncologists gauge prostate cancer aggressiveness by combining DRE findings, serum PSA levels and data derived from systematic biopsy findings.
Men with suspected prostate cancer are categorised into risk groups (see EAU risk classification in EAU guidelines on prostate cancer [2]). This classification is based on the grouping of patients with a similar risk of biochemical recurrence after radical treatment [3]. Tables and nomograms have been developed to predict the likelihood of extraprostatic spread, seminal vesicle invasion and lymph node involvement, and some even state recurrence-free survival rates at 3 and 5 years [4][5][6][7][8].
There is, however, still no consensus of what constitutes a clinically significant prostate cancer (csPCa) [9]. Current argument centres on Gleason category 7 pattern 3 or 4 dominance, ISUP grades 2 and 3, respectively. While the 2014 grading system differentiates Gleason 7 by dominant pattern, all Gleason 7 is classed intermediate risk, albeit with the qualification that emerging clinical data support the distinction between favourable (ISUP grade 2) and unfavourable-risk (ISUP grade 3) patient categories within the intermediate-risk group [2,10,11].
Evaluating csPCa in the post-MRI era mpMRI can detect and localise cancers with a Gleason score ≥ 7 more easily than lower-grade cancers [12][13][14][15], relying on the lower signal intensity of higher-grade cancers on T2weighted imaging (T2w), more impeded diffusion on diffusion-weighted imaging (DWI), early enhancement on dynamic contrast sequences (DCE), and (previously) higher choline over citrate ratios on spectroscopic imaging [16][17][18]. mpMRI evaluates lesion volume with reasonable accuracy, at least for aggressive tumours [19]. In one study, correlations between lesion volume estimated on T2-weighted images, ADC maps, and DCE-MR images with pathology were 0.91 and 0.93, respectively [20].
In 2012, the ESUR proposed a standardised reporting tool called 'PI-RADS' (Prostate Imaging Reporting and Data System) [1] in an attempt to align mpMRI findings with the risk of having csPCa. In 2015, an updated version (PI-RADS v2) was published in collaboration with the American College of Radiology and the AdMeTech Foundation [21,22] (Table 1). PI-RADS 2 has been validated in a meta-analysis of 21 studies including over 3857 patients. This demonstrated a pooled sensitivity of 89% and a pooled specificity of 73% [23].
The Eligibility criteria to have an mpMRI (in place of biopsy) should be based on the EAU and/or National current recommendations for biopsy referrals. Following suitable clinical evaluation for acute or chronic reasons not to be investigated, the reasons to offer mpMRI would mirror those currently used to offer biopsy.

Pre-biopsy mpMRI advantages
The problem with any approach dominated by TRUS-guided systematic biopsy (SBx) is that it is organ rather than lesion based, introducing two major limitations: overdiagnosis of clinically insignificant prostate cancers (cisPCa) and underdiagnosis of csPCa.
The case for excluding men from biopsy based on mpMRI Avoiding or deferring biopsy (possibly indefinitely) if mpMRI suggests low likelihood of csPCa would reduce the burden to men and to their health systems of initial diagnostic workup and low-grade prostate cancer follow-up. Such an approach may also improve the cost efficiency of the diagnostic workup [30]. Accepting that the results are subject to assumptions around test costs, sensitivity of mpMRI-influenced biopsies, and long-term outcomes of men with PCa, recent analysis of a UK population concluded mpMRI first followed by up to 2 rounds of biopsy is more cost effective than current practice [31]. This strategy requires a high negative predictive value (NPV) of mpMRI in excluding csPCa.
A recent systematic review (9613 men) in conjunction with the EAU-ESTRO-ESUR Prostate Cancer Guidelines panel revealed a median mpMRI NPV of 82% (interquartile range (IQR), 69-92%) for overall cancer exclusion and of 88% (IQR, 86-92%) for csPCa exclusion [32]. The critical issue highlighted in this review is that the reported range of the NPV for mpMRI is extreme and varies according to definitions and risk categorisation used.
A key variable of the NPV is the prevalence of cancer within the population being monitored: when the prevalence doubles from 30 to 60%, the NPV of mpMRI (scores 1-2 taken as 'negative') falls from 88 to 67% (for any cancer grade) ( Table 2) [32]. NPV therefore is bound to be variable as it depends on whether mpMRI is being used in a low-risk screening setting or in a selected high-risk cohort. Furthermore, the prevalence will alter according to the definition of csPCa (Table 3). Three pivotal multicentre trials on the use of mpMRI in biopsy-naïve men inform this review: PROMIS, PRECISION [33,34], and the 4M study by van der Leest [35].
Similar results were reported in the more recent multicentre, randomised, noninferiority PRECISION trial, in which 500 biopsy-naïve men were randomised to undergo either mpMRI with or without targeted biopsy, or standard transrectal ultrasonography-guided biopsy [34]. Using mpMRI as a triage test could have avoided 28% of primary biopsies, while detecting 12% more csPCa (defined as Gleason score ≥ 3 + 4) than SBx and 'missing' 13% of cisPCa. These results were obtained in 25 centres (academic and non-academic) with mixed experience in both mpMRI and MR-targeted biopsy, and without restrictions on the use of a 1.5-T or a 3.0-T system, endorectal coil, or biopsy technique (visual registration, software-assisted registration or in-bore).
In a clinical follow-up study (median follow-up of 41 months) of a mixed population of biopsy-naïve, repeat biopsies, and active surveillance (n = 300), who had undergone a negative in-bore MR-guided biopsy for PI-RADS 3-5 lesions, only 1.7% (5/300) had csPCa subsequently diagnosed by any kind of follow-up histology in 82 men (any biopsy or radical prostatectomy), and in 218 without any histology confirmation [37]. In another cohort of 1255 patients with negative mpMRI, the csPCa-free survival rates at 48 months were 95% in originally biopsy-naïve patients and 96% in patients with a prior negative biopsy [38].

Improving detection of csPCa
Pre-biopsy mpMRI in men with suspected prostate cancer is justified further if it improves the detection of csPCa through targeted biopsies of any suspicious lesion suggested by mpMRI.
Radiologic-pathologic correlations with whole-mounts have shown that mpMRI is highly sensitive for locating aggressive cancers, with 80-86% of Gleason 7 and 93-100% of Gleason ≥ 8 detected [12]. Correlation studies of mpMRI with TBx or radical prostatectomy specimens performed after the introduction of PI-RADS showed that the location of the index lesion was correctly assessed by mpMRI in 95% of patients [ A recent systematic review of mostly retrospective studies showed that TBx performed under MR/TRUS fusion detected more csPCa than SBx, with a median detection rate of 33.3% (range, 13.2-50%) versus 23.6% (range, 4.8-52%), respectively. The absolute difference in the detection rates between the two approaches was a median of 6.8% (range, 0.9-41.4%) and The combination of SBx and MRI-targeted biopsies (TBx) comes at a cost: over-detection of cisPCa [51, 52]. In a systematic review of 16 studies comparing TBx and SBx in mixed populations of biopsy-naïve men and men with previous negative biopsies, this overdiagnosis was almost halved by omitting SBx [48]. In the PRECISION trial, 13% (95% CI, − 19 to − 7%; p < 0.001) fewer men were diagnosed with cisPCa in the MRI-targeted biopsy group (in total 9%) than in the standard biopsy group (in total 22%); again, this diagnosis was more than halved by omitting SBx. Likewise, in a prospective nonrandomised trial of 1003 men who underwent both TBx and SBx, adding SBx to TBx identified an additional 103 (22%) prostate cancers, 83% of which were low-grade [53].
Excluding SBx in the MRI-negative men in the 4M would have avoided biopsies in 49% in this study population at a small expense of missing csPCa [35].
Technology that allows mapping exactly the biopsy needle path has suggested that the median number of systematic cores sampling the same region selected for target is 2. If these 'isometric' systematic cores are disregarded, systematic biopsy has a modest benefit only: 3% cancer detection instead of benign diagnosis and~1% from cisPCa to csPCa [55].
There are few data comparing TBx to TTP Bx. TTP Bx should be considered in patients at high risk with negative mpMRI and in some patients at low-risk with persistently elevated PSA and a negative MRI.

Biopsy strategy in the 'mpMRI first' era
It is impossible to brush aside that the range of published mpMRI NPV figures is broad and that csPCa prevalence (i.e. pre-MRI probability of csPCa) has a major impact on NPV [9, 32]. Therefore, there is a need to refine the biopsy planning process by incorporating the mpMRI findings within a larger nomogram containing clinical data to determine an individual's likelihood of having csPCa.
In a recent multivariate logistic regression analysis to predict likelihood of csPCa for biopsy-naive and previously biopsied men, the PI-RADS classification contributed significantly to a newly developed risk model (p < 0.001) in combination with the ERSPC-RCs (www.prostatecancerriskcalculator.com) based on the European Randomised Study of Screening for PCa (ERSPC) [56,57]. For biopsynaive men, the risk model reached a higher AUC (0.83), compared with ERSPC-RC3 (0.81), refitted RC3 (0.80), and PI-RADSv1.0 (0.76). The risk model AUC was comparable with that of ERSPC-RC3 + PI-RADSv1.0 (0.84). Likelihood ratio test results similarly showed that the risk-models may perform significantly better compared with (refitted) ERSPCs and PI-RADS alone. Others have confirmed these results [58].

PI-RADS 3 lesions and risk stratification and biopsy
PI-RADS assessment category 3 is assigned when the probability of prostate cancer is uncertain. The percentage of patients assigned a PI-RADS assessment category of 3 is extremely variable in the different published cohorts [59]. As expected, the biopsy positivity rate is also highly variable in these lesions; the cumulative total of high-grade PCa (GS ≥ 3 + 4) in the PI-RADS category 3 has been reported 21% (range 4-27%) in biopsy-naïve men and 16% (range 10-19%) men with previous negative biopsies [59]. Another paper lists overall cancer detection 16-67% and proportion of Gleason ≥ 7 cancers 0-43% [46]. Based on these data, the authors concluded that the prevalence of PI-RADS 3 index lesions in the diagnostic workup is not negligible, varying between one in five (22%) and one in three (32%) men, depending on patient cohort of the first biopsies or previous negative biopsies. The actual prevalence of csPCa after TBx in PI-RADS category 3 lesions varies between patient groups from one in five (21%) and one in six (16%), depending on previous biopsy status. Although this prevalence is lower in comparison to PI-RADS category 4 and PI-RADS category 5 lesions, still a considerable proportion of men harbour significant disease.

Biopsy strategy and patient risk
Biopsy decisions should first be based on mpMRI findings, favouring avoiding biopsy in 'negative' (any/all lesions PI-RADS 2 or less) studies and targeting PI-RADS 4 or 5 lesions. For some 'negative' studies and most PI-RADS 3 lesions, a second assessment incorporating clinical (age, DRE, family history for example) and biochemical (PSA density and velocity) parameters should be applied to see if systematic biopsy alone or in addition to targets to the low-grade PI-RADS lesions are indicated.
In patients stratified into a priori low-to-intermediate risk (Table 1), the mpMRI NPV is probably sufficiently high to avoid SBx in case of negative mpMRI [34]. Systematic biopsy results would be expected to be less influential on patient management in the setting of a positive mpMRI with a suspicious lesion than in a negative mpMRI. In high-risk patients, however, patients with negative mpMRI will probably still need SBx [34]; even in expert centres, mpMRI may 'miss' 10-12% csPCa. The recommended shift towards an MRIdirected and risk-stratified approach to seeking csPCa is captured in the flowchart (Fig. 1). A shift in emphasis away from SBx with additional MRI-TBx to MRI-TBx for highprobability mpMRI examinations is recommended, accepting the modest price of missing csPCa [33, 68-75].  Fig. 1 Proposed flowchart for investigating men suspected having prostate cancer, beginning with mpMRI. Using mpMRI as the primary investigation in prostate cancer diagnostic workup following clinical suspicion, men will be stratified into PI-RADS assessment categories 1-2, 3, and 4-5. Capitalising on the high negative predictive value of mpMRI, assessment category 1-2 may indicate clinical follow-up avoiding systematic biopsy, or indicate further risk stratification with developing risk calculators (nomograms). Assessment category 3 may indicate MR-targeted biopsy (TBx) combined with systematic biopsy (SBx) to gain maximal diagnostic yield. Alternatively, risk stratification may sub-differentiate these men into high-risk and low-risk; the low-risk group may defer systematic biopsy. Assessment category 4-5 may indicate MR-targeted biopsy. Systematic biopsy could be performed in direct combination or secondary, depending on biopsy workflow. In assessment category 5, the added value of systematic biopsy would be limited. When prostate cancer has not been identified, additional risk stratification could be performed to indicate or avoid additional systematic and targeted biopsy. Green arrows, low-risk; red arrows, intermediate-/high-risk. Dotted lines indicate research in progress. PCa, prostate cancer; MRI, magnetic resonance; PI-RADS, prostate imaging reporting and data system = suspicion MRI score (1-5); TBx, MRI-targeted biopsy; SBx, transrectal/transperineal ultrasound-guided systematic biopsy; AS, active surveillance

Biopsy Naïve Men
Prevalence for overall cancer, unless specified otherwise (2) Randomly selected from a cohort of 118 pts. with a single lesion prospectively scored Likert ≥ 3/5; these patients were re-evaluated by two readers (3) Patients with at least one lesion prospectively scored PI-RADS V1

Inter-reader variability
There are three difficulties with the widespread introduction of pre-biopsy mpMRI: the variable NPV of mpMRI, the variable accuracy of using mpMRI with TRUS to target suspicious lesions regardless of their location within the prostate gland, and inter-reader variability. The results from studies addressing the variability between 2013 and 2017 are summarised in Table 4. The conclusions from the studies are succinctly captured by Hansen et al [88]: (1) mpMRI exams are more often called negative in subspecialist reads (41% vs 20%); (2) second readings of prostate mpMRI by subspecialist uroradiologists significantly improve NPV and PPV; (3) reporter experience may reduce overcalling and avoid over targeting of lesions; and (4) greater education and training of radiologists in prostate mpMRI interpretation are advised. Many European countries are, in collaboration with ESUR and EAU, working to address this training need.

Conclusion
Current national guidelines in Europe highlight the worth of mpMRI in the management of men with suspected PCa. The case for using mpMRI to help in selecting which men with suspected PCa should have a biopsyand which need not-and to then select the regions of the prostate to biopsy (and which regions can be ignored) is compelling. The evidence base, including level 1 studies, is overpowering as are the arguments for patient benefit, avoiding either biopsy or overdiagnosis of clinically insignificant cancer.
Patients contemplating a biopsy are becoming aware that imaging by means of mpMRI may permit avoidance of biopsy in some cases and targeting in others. These patients will be understandably anxious to avoid the risks of biopsy, or at least minimise the risks by having fewer biopsy samples. Going away from 'default' SBx to premeditated TBx judiciously and selectively complimented by SBx using a two-step risk evaluation offers the best compromise to reduce biopsy rates and reduce overdiagnosis of cisPCa while minimising the chances of missing csPCa. The evidence to expect to avoid SBx altogether even in the era of pre-biopsy mpMRI is weak [54].
Our summary suggestions are presented in Table 5.
Funding The authors state that this work has not received any funding.

Compliance with ethical standards
Guarantor The scientific guarantor of this publication is Jonathan Richenberg, MA BM Bch MRCP FRCR Hon Sen Lect BSMS, on behalf of ESUR Prostate Working Group.

Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.

Statistics and biometry
No complex statistical methods were necessary for this paper.

Informed consent No human subjects involved.
Ethical approval Institutional Review Board approval was not required because this is a review article of evidence published.

Methodology
• Literature review and evidence analysis.  1. mpMRI should be the first investigation in the workup of men with suspected prostate cancer (Fig. 1) 2. PI-RADS assessment categories 1 and 2 have a high predictive value in excluding significant disease, and systematic biopsy may be postponed, especially in men with low-risk of disease following additional risk stratification (see 7 below) 3. PI-RADS assessment category lesions 4 and 5 should be targeted 4. PI-RADS assessment category lesion 3 may be targeted and systematic biopsied depending on risk stratification 5. Targeted biopsy (cognitive, MRI/US fusion, or 'in-bore') should be available for biopsy of focal lesions 6. Systematic biopsies in addition to targeted biopsy should be used judiciously rather than as a default, for example in cases being considered for focal therapy or nerve-sparing surgery 7. Where clinical risk parameters including age, family history, DRE findings, PSA velocity, and PSA density are of concern, SBx should be considered even in the setting of a negative mpMRI radiation therapy, or interstitial radiation therapy for clinically lo- In vivo assessment of prostate cancer aggressiveness using magnetic resonance spectroscopic imaging at 3 T with an endorectal coil.