Que peut-on attendre d’une IRM prostatique non suspecte ? Une revue de la littérature

29 décembre 2020

Auteurs : G. Fiard, J.M. Norris, T.A. Nguyen, V. Stavrinides, J. Olivier, M. Emberton, C.M. Moore
Référence : Prog Urol, 2020, 16, 30, 986-999



Prostate cancer diagnosis has considerably evolved in the past decade thanks to the progress of multiparametric magnetic resonance imaging (mpMRI) and the development of MRI-targeted prostate biopsies. In the presence of level 1 evidence, there is now consensus toward using mpMRI before the initial or repeat biopsy [1].

However, there is still disagreement on the attitude toward non-suspicious mpMRI (PI-RADS or Likert score 1-2). Therefore, systematic biopsies are still often used, driving a potential for overdetection of insignificant cancers [1]. Some centres have already chosen to biopsy only in case of mpMRI-visible lesion, but this attitude is not supported by all recommendations [2].

A systematic review and meta-analysis from the European Association of Urology Prostate Cancer Guidelines Panel including studies published between 2005 and 2015 estimated a combined negative predictive value between 67% and 88%, depending on the overall prostate cancer prevalence [3]. A very recent review considering various definitions of non-suspicious mpMRI and clinically significant cancer provided an estimated NPV of 90.8% of a PIRADS/Likert 1-2 MRI for ISUP grade≥2 disease in biopsy-naïve patients [4].

Should we offer biopsy to patients with a non-suspicious mpMRI? What additional factors might help in deciding which patients should still have a biopsy? Can the non-visibility of the cancer be explained and what are the implications? We conducted a review of the most recent literature to help provide answers to these questions.

Materials and methods

Search strategy

We conducted a narrative review of the literature based on a non-structured search strategy in Medline databases, going back to 2015. The initial search terms were ("Magnetic Resonance Imaging" or "MRI") and "prostate cancer" and ("diagnosis" or "biopsy") and ("non-suspicious" or "negative" or "invisible"). For the section on mpMRI invisible cancer, a second search including the terms "prostate cancer MRI visible" was performed. All English titles and abstracts were reviewed and included if providing sufficient available data to answer the study questions. References of the selected articles were screened for additional articles.

Definition of clinically significant cancer

Unless stated otherwise, clinically significant cancer was defined as ISUP≥2.


Should we continue to offer biopsy to patients with a non-suspicious mpMRI?

To answer to this question, we will look at the negative predictive value of mpMRI, taking into account that it will vary depending on the population considered (biopsy-naïve versus previous negative biopsy, prostate cancer prevalence), the definition of clinically significant cancer used, and the type of confirmatory exam (density of biopsy sampling) performed.

Transperineal template biopsy as reference test: NPV 76-92%

Transperineal biopsy has been used as a confirmatory reference test to evaluate the performance of mpMRI to diagnose clinically significant prostate cancer. Details are provided in Table 1.

Biopsy-naïve patients

Two prospective multicentre studies provide us with a high level of evidence in this population. In the PROMIS study, 576 biopsy-naïve patients had a systematic pre-biopsy mpMRI. Multiparametric MRI findings were compared to a 5-millimetre transperineal template biopsy mapping (reference test), and a 12-core transrectal systematic biopsy. Using such an extensive and thorough confirmatory measure, the negative predictive value of a non-suspicious mpMRI (Likert 1-2; n =158) was 76% (95% CI: 69-82) to rule out any cancer with ISUP grade≥2 [5]. Reassuringly, a recent post-hoc analysis of the PROMIS data showed that cancers overlooked by MRI were of lower grade and volume [6].

Hansen and colleagues recruited 807 consecutive biopsy-naïve patients in a multicentre prospective outcome study. Among them, 236 had no suspicious lesion on multiparametric mpMRI. Comparison with systematic sampling according to the Ginsburg protocol (median: 22 cores) demonstrated an NPV of 80% for mpMRI [7].

Another retrospective study, looking at 288 patients, including 131 with non-suspicious mpMRIs biopsied by a transperineal approach, reported an NPV of 84%, where clinically significant cancer included any grade with a maximum cancer core length of≥4mm. These results have to be interpreted in light of a lower number of cores taken (14 cores) [8]. Similarly, the NPV of 85% demonstrated by Otti and colleagues accounts for the fact that transperineal biopsies (10 cores) were taken to sample anterior regions only, otherwise, transrectal sextant biopsies were performed [9].

Thompson and colleagues included 364 patients in a prospective cohort study including 79 men with a non-suspicious mpMRI, demonstrating an NPV of 92% for ISUP grade≥2 disease with transperineal template mapping (median 30 cores) as reference test. Of note, double reporting of all mpMRIs was done by two expert radiologists (>1000 prostate mpMRIs each) [10].

Biopsy-naïve men and those who have had a previous negative TRUS biopsy

Distler and colleagues retrospectively analysed the performance of mpMRI in a mixed population (42% previous negative biopsy) using transperineal template biopsy mapping with a median of 27 cores as reference test. Based on the analysis of 1040 patients including 344 with non-suspicious mpMRIs (PIRADS 1-2), the NPV was 79% [11].

Repeat biopsy (including prior positive biopsies)

In a population with a high prevalence of prostate cancer, Simmons and colleagues reported an NPV of 91.4% of a non-suspicious mpMRI in ruling out clinically significant disease (ISUP grade≥2) using transperineal template biopsy mapping as a reference test (median 49 cores) in the PICTURE trial [12].

Transrectal systematic biopsy as reference test: NPV 82-99%

Looking at studies using a less extensive reference test numbers logically provides higher NPVs as non-visible cancer is less likely to be found where the sampling density is lower. Detailed results are presented in Table 2 (biopsy-naïve and prior negative biopsy) and Table 3 (prior positive biopsy).

Biopsy-naïve patients

In the French MRI-FIRST trial, 45 of the 275 men enrolled had no suspicious lesion on mpMRI (PIRADSv2 or Likert 1-2) and transrectal systematic biopsy led to the diagnosis of clinically significant prostate cancer (ISUP grade≥2) in 5 (11%) [13]. Among the 643 patients recruited in the 4M study, 309 received systematic biopsies (12 cores) following a non-suspicious mpMRI. The NPV of mpMRI, taking into account one significant cancer diagnosed at 1-year follow-up, was 96% [14].

In another recent cohort of 833 biopsy-naïve men undergoing upfront mpMRI, systematic sampling was performed for 140 patients with a Likert 1-2 score using a transrectal (12 cores - 65%) or transperineal (24 cores-35%) approach. The NPV of a non-suspicious mpMRI was 92.8%. Of note, 344 patients with a non-suspicious mpMRI did not undergo biopsy [15].

Very similar results were obtained by Falagario and colleagues based on a retrospective cohort of 266 men among whom 117 had non-suspicious mpMRIs (PIRADS 1-2). A clinically significant cancer was detected in 9 patients (NPV 92%) [16]. Similarly, an NPV of 92% was calculated by Zhang and colleagues looking at 273 non-suspicious mpMRIs in patients undergoing 24-core systematic biopsy [17].

Prior negative biopsy

Higher NPVs are achieved in this category of patients, presumably because they are at lower risk of prostate cancer, especially in the most recent studies where patients likely had an MRI before their initial biopsy.

Boesen and colleagues focused their retrospective analysis on the 3-year follow-up of 194 patients with a history of negative biopsy. Patients were included if the mpMRI performed at the beginning of the study was non-suspicious. In this population (prior negative biopsy+non-suspicious mpMRI), the NPV was 95% [18]. Kotb and colleagues in another study including 81 non-suspicious mpMRIs found an NPV of 87.6% calculated [19].

Lo and colleagues retrospectively reviewed the mpMRIs of 138 patients including 73 that were non-suspicious. Even with long-term follow-up (median 6.7 years), the NPV remained at 96% [20]. Subgroups of patients with previous history of negative biopsy in studies reporting the results of mixed populations showed respective NPVs of 83.3%, 92%, 97%, 100% and 100% [21, 22, 23, 24, 25].

Mixed populations

The results of studies including biopsy-naïve, prior negative biopsy and prior positive biopsy/active surveillance patients are difficult to interpret. Nonetheless, these studies provide us with pooled NPVs reflecting daily practice. Of note, all data shown here are retrospective.

Two recent retrospective studies including 222 and 391 cancer-naïve patients with non-suspicious mpMRIs reported NPVs of 96% and 89.8%, respectively [21, 26]. Interestingly, in the second study by De Visschere and colleagues, only 47% of mpMRIs were still classified as non-suspicious in retrospect, in light of clinical data revealing a diagnosis of prostate cancer at 2-year follow-up in 124/391 (32%) patients [26]. With 4-year follow-up, Panebianco and colleagues from a large series comprising 1255 non-suspicious MRIs provided an NPV of 96% among patients with a previous history of prostate biopsy and 91% in biopsy-naïve patients [27].

Adding patients with prior positive biopsy, the retrospective evaluation of a mixed cohort including 27/135 patients with prior positive biopsy performed by Oishi and colleagues provided an NPV of 82% [22]. Wang and colleagues similarly assessed a mixed cohort of 464 patients including 84 negative mpMRIs whose NPV was 87% [23].

Conversely, three retrospective studies reported NPVs ranging from 96.5% to 99% [24, 25, 28]. However, although all three studies included a significant number of men with non-suspicious mpMRIs (114, 100 and 201, respectively), the NPVs described were likely driven by the majority of patients with previous negative biopsies and the very low rate of significant cancers diagnosed among them.

Active surveillance

In patients on active surveillance for ISUP 1 prostate cancer, the NPV of mpMRI to rule out clinically significant prostate cancer is lower, presumably because this is a population with a higher prevalence of prostate cancer.

The ASIST trial randomised patients between confirmatory biopsy versus mpMRI plus confirmatory biopsy. Among the 137 men in the mpMRI arm, 45 had a non-suspicious mpMRI and the NPV was 85% [29]. The same estimate was obtained by Chesnut and colleagues in a retrospective evaluation including 207 men with a 3-year follow-up. Multiparametric MRI at 3 years was non-suspicious in 72 men, and a clinically significant cancer was detected in 11 (15%) [30].

In a recent retrospective study including 111 patients on active surveillance, 48 had a non-suspicious mpMRI and a significant cancer was detected by systematic transrectal biopsy in 5 (10%) [31].

What predictive factors can we use to help decide which patients with a non-suspicious mpMRI should have a biopsy?

Patients with non-suspicious mpMRIs may avoid prostate biopsy, but 5-20% of clinically significant prostate cancer may be missed [1, 5]. The NPV of a non-suspicious mpMRI is dependent on the prevalence of the disease, therefore the performance of MRI can be deceiving in a very high-risk population and risk-stratification appears critical before omitting biopsy [3]. The prevalence of significant cancer will vary by patient group (biopsy naïve>active surveillance or previous negative biopsy), but also will vary within the health care context e.g., whether PSA testing is commonly done at a population level or not.

The NPV will also vary with the threshold set for clinical significance, with a higher NPV as the threshold is raised. Some groups use ISUP grade alone to set a significance threshold, whilst others will also incorporate a proxy for tumour volume, such as maximum cancer core length.

The optimal approach to identify patients with non-suspicious mpMRI but a higher risk of significant prostate cancer is not yet well established. In an editorial on PRECISION, Nzenza and colleagues proposed that in case of non-suspicious mpMRI, "systematic biopsy still has a role in patients with red flags based on factors such as family history, PSA density, PSA velocity, BRCA mutation, and abnormal digital rectal exam" [32]. Combining mpMRI score with additional predictive factors may increase the negative predictive value (NPV) of mpMRI and reduce the number of prostate biopsies for men whose risk of significant cancer is low, as well as detecting cancers in men with equivocal or non-suspicious MRIs. Few studies reported predictive factors of clinically significant cancer in patients with non-suspicious mpMRIs.

Clinical factors and PSA density

Factors like family history, abnormal digital rectal exam or ethnicity have been studied and failed to prove significant predictive factor of clinically significant cancer in non-suspicious mpMRI populations [22, 24, 27, 33]. Studies looking at age, PSA and prostate volume showed conflicting results [22, 23, 24, 27, 33]. It is acknowledged that the prostate changes with age. Young men often have ‘bright young prostates' which can harbour low volume ISUP 2 disease, which is less readily seen than in older men with greater differentiation between the transition and peripheral zones. Age, total PSA and prostate volume alone were found to be significant predictive factor of clinically significant prostate cancer in only one study [27], 2 studies [23, 27] and 2 studies [22, 33] respectively.

Previous negative biopsy status was reported as a significant predictive factor of clinically significant cancer in 2 studies [22, 27]. Oishi and colleagues reported an OR=5.2 (1.6-16.5), P =0.005 when no previous negative biopsy has been done [22].

Conversely, PSA density (PSAd)<0.15ng/mL/mL was found to be a significant predictive factor in 8 studies [7, 8, 9, 11, 19, 22, 34, 35]. The NPV of the combination of mpMRI and PSAd<0.15ng/mL/mL ranged from 84% to 100% in these studies. Panebianco and colleagues found that PSAd>0.15ng/mL/mL had a hazard ratio>7 for the diagnosis of clinically significant disease in patients with a non-suspicious mpMRI [27].

A recent meta-analysis by Pagniez and colleagues showed that PSAd was the most studied and relevant predictive factor of clinically significant cancer. Indeed, when associated with a PSAd<0.15ng/mL/mL, the NPV of mpMRI in cancer-naïve, biopsy-naïve and previous negative biopsy groups increased from 83.7%, 82.7%, 88.2% to 90.4%, 88.7% and 94.1%, respectively [36].

Biomarkers and risk calculators

Biomarkers such as PCA3, PHI, 4K score or risk calculators have been studied as potential prognostic factors. Perlis and colleagues presented results based on 128 non-suspicious mpMRIs. No patient with a non-suspicious mpMRI and a normal PCA3 score had significant cancer on biopsy (NPV=100% [P <0.0001]) [37]. Gnanapragasam and colleagues reported the results of a PHI test combined with a non-suspicious mpMRI in 94 patients. In patients with a PHI score≥35, the NPV for significant disease was 97% (84-100%) whereas the NPV of mpMRI alone was 75.5% [38]. Falagario and colleagues presented results of the 4K score combined with non-suspicious mpMRI in 117 patients. NPV was 96.9% and 97.1% for patients with low or intermediate 4Kscore risk [16]. These biomarkers seem interesting when combined with mpMRI to increase the NPV, but further validation studies are needed before they can be established in clinical practice.

Risk calculators also appear of interest to determine the individual risk of each patient. On multivariate analysis, the Prostate Cancer Prevention Trial Risk Calculator (PCPT-RC) was a significant predictive factor of clinically significant cancer in the study conducted by Wang and colleagues (OR 1.01, P <0.01) [23]. A new tool from the European Randomized Study of Screening for Prostate Cancer risk calculator (ERSPC-RC) comprising mpMRI results has been published recently and may help selecting patients for biopsy. Falagario and colleagues presented the results of ERSPC-RC in 117 patients. Negative predictive value was 98.7% for ERSPC-RC<2% [16].

How does mpMRI-invisible prostate cancer differ from mpMRI-visible cancer (and is this clinically useful)?

Perhaps the most intriguing question about prostate cancer undetected by mpMRI is - how does this differ from disease that we detect? Early evidence suggests that mpMRI-invisible cancer differs from its visible counterpart in almost every way, from the molecular-level through to the population-level. Furthermore, these differences may account for the mechanisms that underpin mpMRI-visibility and potentially the clinical utility afforded by mpMRI-invisibility.

Firstly, and reassuringly, tumour visibility on mpMRI is positively associated with cancer grade, volume and stage. In other words, the larger and more aggressive the cancer, the greater the probability of detection [5, 27, 39, 40, 41]. This characteristic of prostate mpMRI is fundamental and is likely the reason for incorporation of mpMRI into national guidelines [2]. In a post hoc analysis of PROMIS, mpMRI-invisible cancers had significantly lower overall and maximum Gleason scores (P =0.0007 and P <0.0001, respectively) and shorter maximum cancer core lengths (median difference: 3mm [5 vs. 8], P <0.0001; 95% CI 1-3) compared to mpMRI-visible cancers [6]. Indeed, this makes radiobiological sense; smaller tumours are likely to fall below the threshold of spatial resolution of mpMRI (and therefore, will be invisible) and lower grade peripheral zone tumours are likely to have sparse tissue density (and therefore, less restricted diffusion on mpMRI).

Secondly, there are additional histopathological features (beyond tumour grade and size) that account for mpMRI conspicuity [42]. Miyai and colleagues recently compared mpMRI-visible and mpMRI-invisible tumours on whole-mount radical prostatectomy; they found that visible tumours had higher architectural density with increased proportions of cancer cells (60.9% vs. 42.7%, P <0.0001), decreased stromal proportions (33.8% vs. 45.1%, P =0.00089) and decreased luminal proportions (5.2% vs. 12.2%, P <0.0001) [39]. These results confirm previous findings demonstrating that restricted diffusion was more strongly linked to the distribution of epithelial, luminal and stromal components than to glandular differentiation [43, 44]. Again, association of increased tissue density and tumour visibility has biological plausibility - diffusion of water on mpMRI would be further restricted in tissue with higher density, thus increasing tumor conspicuity on diffusion. In a similar study, Borren and colleagues found that mpMRI-visible tumours had higher cellular and microvessel density compared to mpMRI-invisible tumours (cell density: 3560 cells/mm2 vs. 2910 cells/mm2; microvessel density: 115 vessels/mm2 vs. 90 vessels/mm2) [40]. In this case, tumour visibility is explained by higher microvessel density, which would generate higher dynamic contrast enhanced signal through higher tissue concentrations of gadolinium within microvessels. Interestingly, there are two (potentially aggressive) prostate cancer sub-types that have reduced detection on mpMRI, namely ductal and cribriform [45, 46]. However, this claim is contested; Tonttila and colleagues recently examined a cohort of men undergoing radical prostatectomy and found that preoperative mpMRI identified 90.5% (86/95) of tumours containing any cribriform or ductal pattern (95% CI 82.5-95.6), and so clearly this area warrants future research [47].

Thirdly, there has been significant recent interest in uncovering the genetics that underpin different mpMRI phenotypes [48]. Purysko and colleagues correlated the Decipher Genomic Classifier with mpMRI appearances for men undergoing radical prostatectomy and found significantly higher Decipher scores in mpMRI-visible tumours (mean difference 0.22, 95% CI 0.13-0.32; P <0.001). This suggests enrichment of high-risk cancer genes (for early metastasis) in mpMRI-visible disease; however, the authors, crucially, did not control for tumour grade and size [49]. In contrast, Houlahan and colleagues specified inclusion of only ‘clinically significant cancer' (in their case, above Gleason grade 3+4 and 1.5cm diameter) and through genomic and transcriptomic profiling showed again, that even in a matched cohort comparison, mpMRI-visible tumours were enriched with aggressive molecular and microenvironmental features, which they described as "nimbosus" (gathering of stormy clouds, Latin) [41, 50]. However, not all of the extant genomic evidence has confirmed lack of aggressive molecular features in invisible disease. For example, Parry and colleagues compared mpMRI-visible and mpMRI-invisible tumours (obtained at radical prostatectomy) using low-pass whole genome sequencing, methylation arrays and RNAseq. Somewhat concerningly, they found that three of six cores obtained from mpMRI-invisible tumours harboured genetic alterations observed in metastatic castration-resistant prostate cancer, however, this finding was heavily limited by the associated biases of radical prostatectomy and their small sample size (n =6) [51].

Lastly, and possibly most importantly, tumour visibility (and invisibility) on mpMRI appears to confer clinical utility. PROMIS demonstrated this, in that the most significant cancers (i.e., aggressive disease, that we wish to detect and treat) were visible, and the least significant cancers (i.e. indolent disease, that we do not wish to detect and treat) were invisible [5, 6]. This was then validated in the randomised PRECISION trial, in which men who were assigned to the mpMRI-directed arm had favourable diagnostic pathological outcomes [1]. In a longitudinal trial, Panebianco and colleagues grouped men with baseline negative mpMRI according to their biopsy status; firstly, biopsy-naïve men, and secondly, men with prior negative biopsy. They showed that 95% of biopsy-naïve men, and 96% of men with previous negative biopsy, remained free of clinically significant prostate cancer after four years of follow-up; showing the important utility held by negative mpMRI [27]. Finally, it is worth noting that we may soon be in a position in which our threshold for ‘significant cancer' (liable to cause cancer-specific death) actually aligns with mpMRI visibility phenotypes. In a large cohort study of MRI-based active surveillance, Stavrinides and colleagues showed that the outcomes of non-visible Gleason grade 2 (3+4) cancers were similar to Gleason grade 1 (3+3) lesions, whilst visible Gleason 3+4 cancers were more likely to progress [52].

In the 29-year follow-up of the SPCG-4 trial, Bill-Axelson and colleagues found that in men treated with radical prostatectomy, Gleason score 3+4 disease was not associated with prostate-cancer-related death; however, they found that Gleason score 4+3 disease (or worse) was in fact associated with a prostate-cancer-related death (adjusted relative risk 5.73; 95% CI 1.59-20.67). This potentially suggests a new threshold for clinically significant disease, which remarkably aligns with conspicuity of prostate cancer on mpMRI, given that in PROMIS, no men with overall Gleason score 4+3 had mpMRI-invisible disease [53]. Therefore, peripheral evidence suggests that there is utility in invisibility, however, this does need further validation with bespoke MRI-correlated longitudinal studies.


Upfront mpMRI has been included in clinical practice by many institutions, often in anticipation with the latest guidelines, and is now supported by the results of high-quality international studies.

The negative predictive value of non-suspicious mpMRI has proven to be excellent (90-95% in biopsy-naïve patients), and it is fascinating to see the initial results obtained by small retrospective studies confirmed in large, multicentre cohorts and randomised control trials [14, 27, 54]. Those who still doubt the ability of MRI to rule out clinically significant prostate cancer and put a higher confidence in systematic transrectal biopsy must remember that this biopsy scheme has been proven inferior to MRI in a randomised clinical trial [1]. However, as often in clinical practice, uncertainty remains, and the answer to the question - should we biopsy in the presence of a non-suspicious mpMRI? - depends on how well we are dealing with this uncertainty. And how confident we are in explaining this uncertainty to our patients and not turn it into anxiety.

We are in an opening era of personalised medicine, and the field of prostate cancer diagnosis makes no exception. No test is perfect, particularly when applied in very "extreme" conditions, i.e. in populations with a very high prevalence of the disease considered. The accuracy of MRI is such that many clinical factors, biomarkers and risk calculators have failed to demonstrate a significant added value, especially when assessed in a population with a low prostate cancer prevalence. The PSA density with a threshold at PSAd<0.15ng/mL/mL was able to increase the performances of mpMRI in biopsy-naïve patients and patients with previous negative prostate biopsies. Known biomarkers (PCA3, PHI, 4Kscore) and risk calculators (PCPT-RC, ERSPC-RC) still have to be validated prospectively in multicentric larger cohorts before considering their inclusion in clinical practice. New blood or urine biomarkers are crucially needed to better define each patient's individualised risk and document the decision to avoid - or proceed with - biopsy. When deciding to proceed with biopsy, in these patients deemed at higher risk of prostate cancer despite a non-suspicious mpMRI, an adjustment of the biopsy strategy should be considered. Indeed, this review also outlines the superior diagnostic yield of a transperineal mapping biopsy approach compared to systematic transrectal biopsy, suggesting that it may be useful in these patients to increase sampling density.

The progress made in understanding the biology and genetics behind mpMRI visibility is still in its early phase, and future data obtained from MRI-correlated longitudinal studies will prove critical in determining the potential impact of this knowledge and translating the findings into clinical practice.

There are limitations of our review. The majority of included studies were conducted after mpMRI was already introduced in daily practice with many men already avoiding biopsy. It is therefore very likely that all retrospective studies reporting results of biopsies performed in patients with non-suspicious mpMRIs suffer of a degree of selection bias as these patients had already been judged at higher risk to warrant biopsy rather than surveillance. Furthermore, evidence cited here has frequently been produced by highly-trained, expert centres, which potentially limits generalisability to smaller centres. However, we believe that every centre can become expert with a strong collaboration between urologists, radiologists, pathologists, and the possibility for those who report mpMRIs to obtain feedback and discuss cases with those who do the biopsies.


Non-suspicious mpMRI has a high negative predictive value in ruling out significant cancer, especially in patients with previous negative biopsies. Recent studies support that mpMRI-invisible tumours differ from a histopathological and genetic point of view and we are getting closer to understanding the mechanisms of mpMRI visibility. Biopsy-naïve men with increased PSA density over 0.15 should be offered biopsy as they are at higher risk of harbouring mpMRI-invisible lesions. Further work is needed to validate the best sampling approach and strategy in this situation.

Disclosure of interest

The authors declare that they have no competing interest.


Gaelle Fiard receives funding from the Fondation de France and the European Urology Scholarship Program. Joseph Norris is funded by the Medical Research Council (MRC) (Grant Reference: MR/S00680X/1). Vasilis Stavrinides is funded by the Medical Research Council (MRC) (Grant Reference: MR/S005897/1). Mark Emberton receives research support from the United Kingdom's National Institute of Health Research (NIHR) UCLH/UCL Biomedical Research Centre. He was appointed to an NIHR Senior Investigator position in 2013. Caroline M. Moore is supported by the UK NIHR, Movember, PCUK, and the EAU Research Foundation.

Table 1 - Negative predictive value of mpMRI based on studies using transperineal biopsy as reference test.
Reference  Inclusion period  Design  Entry criteria  MRI type  Definition non-suspicious MRI  Main biopsy type (N cores)  Patients having mpMRI (N)  Non-suspicious mpMRI and biopsy (N)  Clinically significant cancers N (%)  Definition clinically-significant cancera,b  NPV (%) 
Biopsy-naïve patients 
Ahmed 2017 (PROMIS) [5 2012-2015  Prospective  Raised PSA, abnormal DRE or family history  1.5T  Likert 1-2  TP (50)  576  158  38 (24%)  ISUP≥ 76 
Hansen 2018 [7 2012-2016  Prospective  Raised PSA, abnormal DRE or family history  1.5 or 3T  PIRADS 1-2  TP (22)  807  236  48 (20%)  ISUP≥ 80 
Washino 2017 [8 2010-2014  Retrospective  mpMRI+biopsy  1.5 or 3T  PIRADS 1-2  TP (14)  288  131  21 (16%)  ISUP≥2 +  84 
Otti 2019 [9 2013-2016  Retrospective  mpMRI+biopsy  1.5T  PIRADS 1-2  TP (10)  1023  348  52 (15%)  ISUP≥ 85.1 
Thompson 2016 [10 2012-2014  Prospective  Age>40, abnormal PSA or DRE  1.5 or 3T  PIRADS 1-2  TP (30)  364  79  6 (8%)  ISUP≥ 92 
Biopsy-naïve+prior negative biopsy 
Distler 2017 [11 2012-2015  Retrospective  PSA>4ng/ml or abnormal DRE  3T  PIRADS 1-2  TP (27)  1040  344  71 (20.6%)  ISUP≥ 79.4 
Repeat biopsy (including prior positive biopsy) 
Simmons 2017 (PICTURE) [12 2012-2014  Prospective  Indication for repeat biopsy  3T  Likert 1-2  TP (49)  249  35  3 (8.6%)  ISUP≥ 91.4 

Légende :
TP: transperineal.

When several definitions are used in the study, we present here the results for ISUP≥2.
ISUP≥2+ means a broader definition is used.

Table 2 - Negative predictive value of mpMRI based on studies using transrectal biopsy as reference test for biopsy-naïve patients and patients with prior negative biopsy.
Reference  Inclusion period  Design  Entry criteria  Follow-up  MRI type  Definition non-suspicious MRI  Main biopsy type (N cores)  Patients having mpMRI (N)  Non-suspicious mpMRI and biopsy (N)  Clinically significant cancers N (%)  Definition clinically-significant cancera,b  NPV (%) 
Biopsy-naïve patients 
Rouvière 2018
(MRI first) [13
2015-2016  Prospective  Raised PSA, abnormal DRE or family history    1.5 or 3T  Likert 1-2  TRUS (12)  275  45  5 (11%)  ISUP≥ 89 
Van der Leest 2019 (4M) [14 2015-2017  Prospective  Age 50-75 and PSA≥3ng/mL  1 year  3T  PIRADS 1-2  TRUS (12)  643  309  13 (4%)c  ISUP≥ 96 
Barrett 2019 [15 2015-2018  Retrospective  Referral to prostate cancer diagnostic clinic    1.5 or 3T  Likert 1-2  TRUS (12)  833  140  11 (7.9%)  ISUP≥ 92.8 
Falagario 2019 [16 2016-2019  Retrospective  mpMRI+4K score+prostate biopsy    3T  PIRADS 1-2  TRUS (12)  266  117  9 (7.7%)  ISUP≥ 92 
Zhang 2019 [17 2012-2018  Retrospective  mpMRI+biopsy    3T  PIRADS 1-2  TRUS (24)  491  273  20 (7%)  ISUP≥ 93 
Prior negative biopsy 
Boesen 2017 [18 2011-2017  Retrospective  Non-suspicious mpMRI or negative MRI-targeted biopsy  3 years  3T  PIRADS 1-2  TRUS (10)  289  194  10 (5%)c  ISUP≥ 95 
Kotb 2018 [19 2015-2016  Retrospective  mpMRI+ repeat biopsy    3T  PIRADS 1-2  TRUS (12)  228  81  10 (12.4%)  ISUP≥ 87.6 
Lo 2019 [20 2004-2009  Retrospective  mpMRI+repeat biopsy  6.7 years  1.5 or 3T  MRI report with no visible focus  TRUS (NA)  138  73  3 (4%)c  ISUP≥ 96 
Mixed biopsy-naïve+prior negative biopsy 
Regis 2019 [21 2012-2017  Retrospective  PSA>4ng/mL or abnormal DRE    3T  PIRADS 1-2  TRUS (12)  1128  222  12 (5%)  ISUP≥ 96 
Prior negative biopsy 
8 (9%)
4 (3%) 
De Visschere 2016 [26 2002-2014  Retrospective  mpMRI  2 years  1.5T  MRI score 1-2 (4-point scale)  TRUS (12)  830  391  40 (10.2%)c  ISUP≥ 89.8 
Panebianco 2018 [27 2010-2015  Retrospective  Non-suspicious mpMRI  2 years  3T  PIRADS 1-2  TRUS (12-18)  1255  NA  60 (4.8%)c  ISUP≥2+  95.2 
Prior negative biopsy 
36 (9%)
24 (4%) 

Légende :
TRUS: transrectal ultrasound-guided; NA: not available.

When several definitions are used in the study, we present here the results for ISUP≥2.
ISUP≥2+ means a broader definition is used.
Including follow-up.

Table 3 - Negative predictive value of mpMRI based on studies using transrectal biopsy as reference test for patients with prior positive biopsy.
Reference  Inclusion period  Design  Entry criteria  Follow-up  MRI type  Definition non-suspicious MRI  Main biopsy type (N cores)  Patients having mpMRI (N)  Non-suspicious mpMRI and biopsy (N)  Clinically significant cancers N (%)  Definition clinically-significant cancera  NPV (%) 
Mixed including prior positive biopsy 
Oishi 2019 [22 2011-2017  Retrospective  Non-suspicious mpMRI    3T  PIRADS 1-2  TRUS (12)  395  135  24 (18%)  ISUP≥ 82 
Wang 2016 [23 2012-2015  Retrospective  Non-suspicious mpMRI    3T  PIRADS 1-2  TRUS (12)  464  84  11 (13%)  ISUP≥ 87 
Prior negative biopsy
Prior positive biopsy 
4 (10%)
5 (16%)
2 (13%) 
An 2018 [24 2013-2017  Retrospective  Non-suspicious mpMRI    3T  PIRADS 1  TRUS (12)  NA  114  4 (3.5%)  ISUP≥ 96.5 
Prior negative biopsy
Prior positive biopsy 
4 (9.8%) 
Lu 2017 [25 2012-2016  Retrospective  Non-suspicious mpMRI    3T  Low suspicion on 3-point Likert scale
TRUS (12)  670  100  3 (3%)  ISUP≥ 97 
Prior negative biopsy
Prior positive biopsy 
1 (2.6%)
2 (6.9%) 
Rozas 2019 [28 2015-2016  Retrospective  mpMRI    3T  PIRADS 1-2  TRUS (18)  342  201  2 (1%)  ISUP≥ 99 
  Prior negative biopsy
Prior positive biopsy 
Active surveillance 
Klotz 2019
(ASIST) [29
2011-2015  Prospective  ISUP 1 prostate cancer+indication for repeat biopsy    3T  Likert 1-2  TRUS (12)  137  45  5 (11%)  ISUP≥ 85 
Chesnut 2020 [30 2013-2016  Retrospective  ISUP 1 prostate cancer+2 mpMRIs  3 years  3T  PIRADS 1-2  TRUS (14)  207  72  11 (15%)b  ISUP≥ 85 
Osses 2020 [31 2013-2019  Retrospective  ISUP 1 prostate cancer+mpMRI    3T  PIRADS 1-2  TRUS (8-12)  111  48  5 (10%)  ISUP≥ 90 

Légende :
TRUS: transrectal ultrasound-guided; NA: not available.

When several definitions are used in the study, we present here the results for ISUP≥2.
Including follow-up.


Kasivisvanathan V., Rannikko A.S., Borghi M., Panebianco V., Mynderse L.A., Vaarala M.H., et al. MRI-targeted or standard biopsy for prostate-cancer diagnosis N Engl J Med 2018 ;  378 : 1767-177710.1056/NEJMoa1801993
Rozet F., Hennequin C., Beauval J.-B., Beuzeboc P., Cormier L., Fromont-Hankard G., et al. [French ccAFU guidelines - Update 2018-2020: Prostate cancer] Prog Urol 2018 ;  28 : S79-S13010.1016/j.purol.2018.08.011 [inter-ref]
Moldovan P.C., Van den Broeck T., Sylvester R., Marconi L., Bellmunt J., van den Bergh R.C.N., et al. What Is the negative predictive value of multiparametric magnetic resonance imaging in excluding prostate cancer at biopsy? A systematic review and meta-analysis from the European Association of Urology Prostate Cancer Guidelines Panel Eur Urol 2017 ;  72 : 250-26610.1016/j.eururo.2017.02.026 [cross-ref]
Sathianathen N.J., Omer A., Harriss E., Davies L., Kasivisvanathan V., Punwani S., et al. Negative Predictive value of multiparametric magnetic resonance imaging in the detection of clinically significant prostate cancer in the prostate imaging reporting and data system era: a systematic review and meta-analysis Eur Urol 2020 ;  78 : 402-41410.1016/j.eururo.2020.03.048 [cross-ref]
Ahmed H.U., El-Shater Bosaily A., Brown L.C., Gabe R., Kaplan R., Parmar M.K., et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study Lancet 2017 ;  389 : 815-82210.1016/S0140-6736(16)32401-1 [cross-ref]
Norris J.M., Carmona Echeverria L., Bott S., Brown L.C., Burns-Cox N., Dudderidge T., et al. Which prostate cancers are overlooked by mpMRI? An analysis from PROMIS J Urol 2020 ;  203 : e1243-e1244
Hansen N.L., Barrett T., Kesch C., Pepdjonovic L., Bonekamp D., O'Sullivan R., et al. Multicentre evaluation of magnetic resonance imaging supported transperineal prostate biopsy in biopsy-naïve men with suspicion of prostate cancer BJU Int 2018 ;  122 : 40-4910.1111/bju.14049 [cross-ref]
Washino S., Okochi T., Saito K., Konishi T., Hirai M., Kobayashi Y., et al. Combination of prostate imaging reporting and data system (PI-RADS) score and prostate-specific antigen (PSA) density predicts biopsy outcome in prostate biopsy naïve patients BJU Int 2017 ;  119 : 225-23310.1111/bju.13465 [cross-ref]
Otti V.C., Miller C., Powell R.J., Thomas R.M., McGrath J.S. The diagnostic accuracy of multiparametric magnetic resonance imaging before biopsy in the detection of prostate cancer BJU Int 2019 ;  123 : 82-9010.1111/bju.14420 [cross-ref]
Thompson J.E., van Leeuwen P.J., Moses D., Shnier R., Brenner P., Delprado W., et al. The diagnostic performance of multiparametric magnetic resonance imaging to detect significant prostate cancer J Urol 2016 ;  195 : 1428-143510.1016/j.juro.2015.10.140 [cross-ref]
Distler F.A., Radtke J.P., Bonekamp D., Kesch C., Schlemmer H.-P., Wieczorek K., et al. The value of PSA density in combination with PI-RADSâ„¢ for the accuracy of prostate cancer prediction J Urol 2017 ;  198 : 575-58210.1016/j.juro.2017.03.130 [cross-ref]
Simmons L.A.M., Kanthabalan A., Arya M., Briggs T., Barratt D., Charman S.C., et al. The PICTURE study: diagnostic accuracy of multiparametric MRI in men requiring a repeat prostate biopsy Br J Cancer 2017 ;  116 : 1159-116510.1038/bjc.2017.57 [cross-ref]
Rouvière O., Puech P., Renard-Penna R., Claudon M., Roy C., Mège-Lechevallier F., et al. Use of prostate systematic and targeted biopsy on the basis of multiparametric MRI in biopsy-naive patients (MRI-FIRST): a prospective, multicentre, paired diagnostic study Lancet Oncol 2019 ;  20 : 100-10910.1016/S1470-2045(18)30569-2 [inter-ref]
van der Leest M., Cornel E., Israël B., Hendriks R., Padhani A.R., Hoogenboom M., et al. Head-to-head comparison of transrectal ultrasound-guided prostate biopsy versus multiparametric prostate resonance imaging with subsequent magnetic resonance-guided biopsy in biopsy-naïve men with elevated prostate-specific antigen: a large prospective multicenter clinical study Eur Urol 2019 ;  75 : 570-57810.1016/j.eururo.2018.11.023 [cross-ref]
Barrett T., Slough R., Sushentsev N., Shaida N., Koo B.C., Caglic I., et al. Three-year experience of a dedicated prostate mpMRI pre-biopsy programme and effect on timed cancer diagnostic pathways Clin Radiol 2019 ;  74 : 10.1016/j.crad.2019.06.004[894.e1-894.e9].
Falagario U.G., Martini A., Wajswol E., Treacy P.-J., Ratnani P., Jambor I., et al. Avoiding unnecessary magnetic resonance imaging (MRI) and biopsies: negative and positive predictive value of MRI according to prostate-specific antigen density, 4Kscore and risk calculators Eur Urol Oncol 2019 ; 10.1016/j.euo.2019.08.015[S2588-9311(19)30141-5].
Zhang Y., Zeng N., Zhang F., Huang Y., Tian Y. How to make clinical decisions to avoid unnecessary prostate screening in biopsy-naïve men with PI-RADs v2 score≤3? Int J Clin Oncol 2020 ;  25 : 175-18610.1007/s10147-019-01524-9 [inter-ref]
Boesen L., Nørgaard N., Løgager V., Thomsen H.S. Clinical outcome following low suspicion multiparametric prostate magnetic resonance imaging or benign magnetic resonance imaging guided biopsy to detect prostate cancer J Urol 2017 ;  198 : 310-31510.1016/j.juro.2017.02.073 [cross-ref]
Kotb A.F., Spaner S., Crump T., Hyndman M.E. The role of mpMRI and PSA density in patients with an initial negative prostatic biopsy World J Urol 2018 ;  36 : 2021-202510.1007/s00345-018-2341-4 [cross-ref]
Lo G., Burton K.R., Haider M.A., Fleshner N., Finelli A., Ghai S. Negative predictive value of prostate multiparametric magnetic resonance imaging among men with negative prostate biopsy and elevated prostate specific antigen: a clinical outcome retrospective cohort study J Urol 2019 ;  202 : 1159-116510.1097/JU.0000000000000388 [cross-ref]
Regis L., Celma A., Planas J., Lopez R., Roche S., Lorente D., et al. The role of negative magnetic resonance imaging: can we safely avoid biopsy in P.I.-R.A.D.S 2 as in P.I.-R.A.D.S. 1? Scand J Urol 2019 ;  53 : 21-2510.1080/21681805.2018.1551243 [cross-ref]
Oishi M., Shin T., Ohe C., Nassiri N., Palmer S.L., Aron M., et al. Which patients with negative magnetic resonance imaging can safely avoid biopsy for prostate cancer? J Urol 2019 ;  201 : 268-27610.1016/j.juro.2018.08.046
Wang R.S., Kim E.H., Vetter J.M., Fowler K.J., Shetty A.S., Mintz A.J., et al. Determination of the role of negative magnetic resonance imaging of the prostate in clinical practice: is biopsy still necessary? Urology 2017 ;  102 : 190-19710.1016/j.urology.2016.10.040 [cross-ref]
An J.Y., Sidana A., Holzman S.A., Baiocco J.A., Mehralivand S., Choyke P.L., et al. Ruling out clinically significant prostate cancer with negative multi-parametric MRI Int Urol Nephrol 2018 ;  50 : 7-1210.1007/s11255-017-1715-7 [cross-ref]
Lu A.J., Syed J.S., Nguyen K.A., Nawaf C.B., Rosoff J., Spektor M., et al. Negative multiparametric magnetic resonance imaging of the prostate predicts absence of clinically significant prostate cancer on 12-Core template prostate biopsy Urology 2017 ;  105 : 118-12210.1016/j.urology.2017.01.048 [cross-ref]
De Visschere P.J.L., Naesens L., Libbrecht L., Van Praet C., Lumen N., Fonteyne V., et al. What kind of prostate cancers do we miss on multiparametric magnetic resonance imaging? Eur Radiol 2016 ;  26 : 1098-110710.1007/s00330-015-3894-x [cross-ref]
Panebianco V., Barchetti G., Simone G., Del Monte M., Ciardi A., Grompone M.D., et al. Negative multiparametric magnetic resonance imaging for prostate cancer: What's next? Eur Urol 2018 ;  74 : 48-5410.1016/j.eururo.2018.03.007 [cross-ref]
Rozas G., de Q., Saad L.S., Melo H.J., de F.E., Gabrielle H.A.A., Szejnfeld J. Impact of PI-RADS v2 on indication of prostate biopsy Int Braz J Urol 2019 ;  45 : 486-49410.1590/S1677-5538.IBJU.2018.0564 [cross-ref]
Klotz L., Loblaw A., Sugar L., Moussa M., Berman D.M., Van der Kwast T., et al. Active Surveillance Magnetic Resonance Imaging Study (ASIST): Results of a Randomized Multicenter Prospective Trial Eur Urol 2019 ;  75 : 300-30910.1016/j.eururo.2018.06.025 [cross-ref]
Chesnut G.T., Vertosick E.A., Benfante N., Sjoberg D.D., Fainberg J., Lee T., et al. Role of changes in magnetic resonance imaging or clinical stage in evaluation of disease progression for men with prostate cancer on active surveillance Eur Urol 2020 ;  77 : 501-50710.1016/j.eururo.2019.12.009 [cross-ref]
Osses D.F., Drost F-JH, Verbeek J.F.M., Luiting H.B., van Leenders G.J.L.H., Bangma C.H., et al. Prostate cancer upgrading with serial prostate MRI scans and repeat biopsy in men on active surveillance: are confirmatory biopsies still necessary? BJU Int 2020 ;  126 : 124-13210.1111/bju.15065
Nzenza T., Murphy D.G. PRECISION delivers on the PROMIS of mpMRI in early detection Nat Rev Urol 2018 ;  15 : 529-53010.1038/s41585-018-0046-5 [cross-ref]
Numao N., Yoshida S., Komai Y., Ishii C., Kagawa M., Kijima T., et al. Usefulness of pre-biopsy multiparametric magnetic resonance imaging and clinical variables to reduce initial prostate biopsy in men with suspected clinically localized prostate cancer J Urol 2013 ;  190 : 502-50810.1016/j.juro.2013.02.3197 [cross-ref]
Hansen N.L., Kesch C., Barrett T., Koo B., Radtke J.P., Bonekamp D., et al. Multicentre evaluation of targeted and systematic biopsies using magnetic resonance and ultrasound image-fusion guided transperineal prostate biopsy in patients with a previous negative biopsy BJU Int 2017 ;  120 : 631-63810.1111/bju.13711 [cross-ref]
Bryant R.J., Hobbs C.P., Eyre K.S., Davies L.C., Sullivan M.E., Shields W., et al. Comparison of Prostate Biopsy with or without Prebiopsy Multiparametric Magnetic Resonance Imaging for Prostate Cancer Detection: An Observational Cohort Study J Urol 2019 ;  201 : 510-51910.1016/j.juro.2018.09.049 [cross-ref]
Pagniez M.A., Kasivisvanathan V., Puech P., Drumez E., Villers A., Olivier J. Predictive factors of missed clinically significant prostate cancers in men with negative MRI: a systematic review and meta-analysis J Urol 2020 ;  204 : 24-3210.1097/JU.0000000000000757[101097JU0000000000000757].
Perlis N., Al-Kasab T., Ahmad A., Goldberg E., Fadak K., Sayyid R., et al. Defining a cohort that may not require repeat prostate biopsy based on PCA3 score and magnetic resonance imaging: the dual negative effect J Urol 2018 ;  199 : 1182-118710.1016/j.juro.2017.11.074 [cross-ref]
Gnanapragasam V.J., Burling K., George A., Stearn S., Warren A., Barrett T., et al. The Prostate Health Index adds predictive value to multi-parametric MRI in detecting significant prostate cancers in a repeat biopsy population Sci Rep 2016 ;  6 : 3536410.1038/srep35364
Miyai K., Mikoshi A., Hamabe F., Nakanishi K., Ito K., Tsuda H., et al. Histological differences in cancer cells, stroma, and luminal spaces strongly correlate with in vivo MRI-detectability of prostate cancer Mod Pathol 2019 ;  32 : 1536-154310.1038/s41379-019-0292-y [cross-ref]
Borren A., Groenendaal G., Moman M.R., Boeken Kruger A.E., van Diest P.J., van Vulpen M., et al. Accurate prostate tumour detection with multiparametric magnetic resonance imaging: dependence on histological properties Acta Oncol 2014 ;  53 : 88-9510.3109/0284186X.2013.837581 [cross-ref]
Houlahan K.E., Salmasi A., Sadun T.Y., Pooli A., Felker E.R., Livingstone J., et al. Molecular hallmarks of multiparametric magnetic resonance imaging visibility in prostate cancer Eur Urol 2019 ;  76 : 18-2310.1016/j.eururo.2018.12.036 [cross-ref]
Norris J.M., Carmona Echeverria L.M., Simpson B.S., Allen C., Ball R., Freeman A., et al. Prostate cancer visibility on multiparametric magnetic resonance imaging: high Gleason grade and increased tumour volume are not the only important histopathological features BJU Int 2020 ;  126 : 237-23910.1111/bju.15085 [cross-ref]
Chatterjee A., Watson G., Myint E., Sved P., McEntee M., Bourne R. Changes in epithelium, stroma, and lumen space correlate more strongly with Gleason pattern and are stronger predictors of prostate ADC changes than cellularity metrics Radiology 2015 ;  277 : 751-76210.1148/radiol.2015142414 [cross-ref]
Kwak J.T., Sankineni S., Xu S., Turkbey B., Choyke P.L., Pinto P.A., et al. Prostate cancer: a correlative study of multiparametric MR imaging and digital histopathology Radiology 2017 ;  285 : 147-15610.1148/radiol.2017160906 [cross-ref]
Schieda N., Coffey N., Gulavita P., Al-Dandan O., Shabana W., Flood T.A. Prostatic ductal adenocarcinoma: an aggressive tumour variant unrecognized on T2 weighted magnetic resonance imaging (MRI) Eur Radiol 2014 ;  24 : 1349-135610.1007/s00330-014-3150-9 [cross-ref]
Truong M., Feng C., Hollenberg G., Weinberg E., Messing E.M., Miyamoto H., et al. A comprehensive analysis of cribriform morphology on magnetic resonance imaging/ultrasound fusion biopsy correlated with radical prostatectomy specimens J Urol 2018 ;  199 : 106-11310.1016/j.juro.2017.07.037 [cross-ref]
Tonttila P.P., Ahtikoski A., Kuisma M., Pääkkö E., Hirvikoski P., Vaarala M.H., Multiparametric M.R.I. prior to radical prostatectomy identifies intraductal and cribriform growth patterns in prostate cancer BJU Int 2019 ;  124 : 992-99810.1111/bju.14812 [cross-ref]
Norris J.M., Simpson B.S., Parry M.A., Kasivisvanathan V., Allen C., Ball R., et al. Genetic correlates of prostate cancer visibility (and invisibility) on multiparametric magnetic resonance imaging: it's time to take stock BJU Int 2019 ;  125 : 340-34210.1111/bju.14919
Purysko A.S., Magi-Galluzzi C., Mian O.Y., Sittenfeld S., Davicioni E., du Plessis M., et al. Correlation between MRI phenotypes and a genomic classifier of prostate cancer: preliminary findings Eur Radiol 2019 ;  29 : 4861-487010.1007/s00330-019-06114-x [cross-ref]
Chua M.L.K., Lo W., Pintilie M., Murgic J., Lalonde E., Bhandari V., et al. A Prostate Cancer "Nimbosus": Genomic Instability and SChLAP1 Dysregulation Underpin Aggression of Intraductal and Cribriform Subpathologies Eur Urol 2017 ;  72 : 665-67410.1016/j.eururo.2017.04.034 [cross-ref]
Parry M.A., Srivastava S., Ali A., Cannistraci A., Antonello J., Barros-Silva J.D., et al. Genomic evaluation of multiparametric magnetic resonance imaging-visible and -nonvisible lesions in clinically localised prostate cancer Eur Urol Oncol 2019 ;  2 : 1-1110.1016/j.euo.2018.08.005 [cross-ref]
Stavrinides V., Giganti F., Trock B., Punwani S., Allen C., Kirkham A., et al. Five-year outcomes of magnetic resonance imaging-based active surveillance for prostate cancer: a large cohort study Eur Urol 2020 ;  78 : 443-45110.1016/j.eururo.2020.03.035 [cross-ref]
Bill-Axelson A., Holmberg L., Garmo H., Taari K., Busch C., Nordling S., et al. Radical prostatectomy or watchful waiting in prostate cancer - 29-year follow-up N Engl J Med 2018 ;  379 : 2319-232910.1056/NEJMoa1807801 [cross-ref]
Ahdoot M., Wilbur A.R., Reese S.E., Lebastchi A.H., Mehralivand S., Gomella P.T., et al. MRI-targeted, systematic, and combined biopsy for prostate cancer diagnosis N Engl J Med 2020 ;  382 : 917-92810.1056/NEJMoa1910038 [cross-ref]

Joint senior authors.

© 2020 
Elsevier Masson SAS. Tous droits réservés.