EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 available at www.sciencedirect.com journal homepage: www.eu-openscience.europeanurology.com Prostatic Disease Improving the Early Detection of Clinically Significant Prostate Cancer in Men in the Challenging Prostate Imaging-Reporting and Data System 3 Category Juan Morote a,b,c,*, Miriam Campistol b, Marina Triquell b, Anna Celma a,b, Lucas Regis a,b, Inés de Torres b,d,e, Maria E. Semidey b,c,d, Richard Mast c, Anna Santamaria b, Jacques Planas a,b, Enrique Trilla a,e a Department of Urology, Vall d’Hebron Hospital, Barcelona, Spain; b Prostate Cancer Research Group, Vall d’Hebron Research Institute, Barcelona, Spain; c Department of Radiology, Vall d’Hebron Hospital, Barcelona, Spain; d Department of Pathology, Vall d’Hebron Hospital, Barcelona, Spain; e Universitat Autònoma of Barcelona, Barcelona, Spain Article info Abstract Article history: Background: Prostate Imaging-Reporting and Data System (PI-RADS) category 3 is Accepted December 21, 2021 a challenging scenario for detection of clinically significant prostate cancer (csPCa) and some tools can improve the selection of appropriate candidates for prostate Associate Editor: biopsy. Guillaume Ploussard Objective: To assess the performance of the European Randomized Study of Screening for Prostate Cancer (ERSPC) magnetic resonance imaging (MRI) model, Keywords: the new Proclarix test, and prostate-specific antigen density (PSAD) in selecting Clinically significant prostate candidates for prostate biopsy among men in the PI-RADS 3 category. cancer Multiparametric magnetic Design, setting, and participants: We conducted a head-to-head prospective anal- resonance imaging ysis of 567 men suspected of having PCa for whom guided and systematic biopsies Proclarix were scheduled between January 2018 and March 2020 in a single academic insti- Prostate-specific antigen density tution. A PI-RADS v.2 category 3 lesion was identified in 169 men (29.8%). European Randomized Study of Outcome measurement and statistical analysis: csPCa, insignificant PCa (iPCa), and Screening for Prostate Cancer unnecessary biopsy rates were analysed. csPCa was defined as grade group 2. predictive model Receiver operating characteristic (ROC) curves, decision curve analysis curves, and clinical utility curves were plotted. Results and limitations: PCa was detected in 53/169 men (31.4%) with a PI-RADS 3 lesion, identified as csPCa in 25 (14.8%) and iPCa in 28 (16.6%). The area under the ROC curve for csPCa detection was 0.703 (95% confidence interval [CI] 0.621–0.768) for Proclarix, 0.657 (95% CI 0.547–0.766) for the ERSPC MRI model, and 0.612 (95% CI 0.497–0.727) for PSAD (p = 0.027). The threshold with the highest sensitivity was 10% for Proclarix, 1.5% for the ERSPC MRI model, and 0.07 ng/ml/cm3 for PSAD, which yielded sensitivity of 100%, 91%, and 84%, respectively. Some 21.3%, 26.2%, and 7.1% of biopsies would be avoided with Proclarix, PSAD, and the ERSPC MRI * Corresponding author. Department of Urology, Vall d’Hebron Hospital, Passeig Vall d’Hebron, 119–129, 08035 Barcelona, Spain. Tel. +34 2746009. E-mail address: jmorote@vhebron.net (J. Morote). https://doi.org/10.1016/j.euros.2021.12.009 2666-1683/Ó 2022 The Author(s). Published by Elsevier B.V. on behalf of European Association of Urology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 39 model, respectively. Proclarix showed a net benefit over PSAD and the ERSPC MRI model. Both Proclarix and PSAD reduced iPCa overdetection from 16.6% to 11.3%, while the ERSPC MRI model reduced iPCa overdetection to 15.4%. Conclusions: Proclarix was more accurate in selecting appropriate candidates for prostate biopsy among men in the PI-RADS 3 category when compared to PSAD and the ERSPC MRI model. Proclarix detected 100% of csPCa cases and would reduce prostate biopsies by 21.3% and iPCa overdetection by 5.3%. Patient summary: We compared three methods and found that the Proclarix test can optimise the detection of clinically significant prostate cancer in men with a score of 3 on the Prostate Imaging-Reporting and Data System for magnetic reso- nance imaging scans. Ó 2022 The Author(s). Published by Elsevier B.V. on behalf of European Association of Urology. This is an open access article under the CC BY-NC-ND license (http://creative- commons.org/licenses/by-nc-nd/4.0/). 1. Introduction usefulness of these three tools for appropriate selection of candidates for prostate biopsy in the challenging setting of Early detection of clinically significant prostate cancer PI-RADS category 3. (csPCa) decreases PCa-specific mortality [1]. Currently, sus- picion of PCa is still based on detection of elevated serum levels of prostate-specific antigen (PSA) and/or an abnormal 2. Patients and methods digital rectal examination (DRE) [2]. Suspected PCa has typi- 2.1. Design, setting, and participants cally been confirmed via systematic biopsies of the prostate, but this approach results in a high rate of unnecessary biop- This was a prospective head-to-head study in which the likelihood of sies and overdetection of insignificant PCa (iPCa) [3]. True csPCa was assessed using the Proclarix test, PSAD, and the ERSPC MRI improvement in the early detection of csPCa has come from model. A cohort of 567 men with a suspicion of PCa because of PSA >3 multiparametric magnetic resonance imaging (mpMRI) and ng/ml and/or abnormal DRE underwent prebiopsy 3-T mpMRI and had guided biopsies [2]. At present, the negative predictive value guided and systematic prostate biopsies scheduled between January of mpMRI can reach 95%, so prostate biopsies can usually be 12, 2018 and March 15, 2020 at a single academic institution. A PI- avoided in men with a Prostate Imaging-Reporting and Data RADS v.2 category 3 lesion was identified in 169 men (29.8%). Men with PCa on active surveillance and those with symptomatic benign prostatic System (PI-RADS) score <3 [4,5]. By contrast, most clinicians hyperplasia treated with 5a-reductase inhibitors were previously recommend prostate biopsy in men with PI-RADS >3 because excluded. Two- to three-core transrectal ultrasound-guided biopsies of the likelihood of PCa ranges from 62% to 92% [6]. PI-RADS cat- suspected lesions and 12-core systematic biopsies were performed in egory 3 is the most challenging scenario: 60–85% of prostate all men using the transrectal approach. The study was approved by biopsies are unnecessary and up to 60% of PCa cases detected our institutional ethics committee (PR-AG129/2020). are insignificant [7,8]. Therefore, tools such as PSA density (PSAD), modern markers, and predictive models are recom- mended for appropriate selection of candidates for prostate 2.2. Testing biopsy [2]. Proclarix is a new blood-based test that estimates the Blood was obtained immediately before prostate biopsy and frozen likelihood of csPCa by computing the risk according to serum was stored locally at 80 °C for 63–811 d (C. 0003439; https:// measurement results for thrombospondin-1 (THBS1), biobancs.isiii.es) and then shipped on dry ice to Proteomedix (Zurich- cathepsin D (CTSD), total PSA, and percentage free PSA in Schlieren, Switzerland). Processing of serum samples, the ELISA kit, and calculation of the risk score by laboratory technicians were per- serum, as well as patient age [9]. Recent studies have formed blind before any clinical information was available. THBS1 and suggested that Proclarix can improve csPCa detection by CTSD were measured using a Proclarix kit (Proteomedix, Zurich- reducing unnecessary biopsies in men with or without Schlieren, Switzerland) according to the kit instructions [15]. Serum total mpMRI [10]. However, data on the behaviour of Proclarix PSA and free PSA were reanalysed for all samples using a Roche Cobas by PI-RADS category are lacking. Meanwhile, PSAD has immunoassay system (Roche Diagnostics, Rotkreuz, Switzerland). Pro- become relevant as prostate volume can be accurately mea- clarix risk calculation was performed according to the instructions and sured on MRI [11–13]. The externally validated European the results ranged from 0% to 100%. PSAD was calculated from the pros- Randomized Study of Screening for Prostate Cancer (ERSPC) tate volume measured on MRI and the total PSA value from the Proclarix predictive model has recently incorporated the PI-RADS test. The ERSPC MRI risk of high-grade PCa was calculated for every man version 1 score and age in the ‘‘3+DRE’’ and ‘‘4+DRE’’ risk using the Prostate Cancer Research Foundation (Reeuwijk, The Nether- calculators [14]. However, no specific analyses of its beha- lands) web application at www.prostatecancer-riskcalculator.com. The viour regarding PI-RADS categories have been carried out. ERSPC MRI risk calculator includes PSA (0.5–50 ng/ml), repeat biopsy In this study we compare the behaviour of PSAD, Pro- (yes/no), DRE (normal/abnormal), prostate volume (10–110 cm3, which clarix, and the ERSPC MRI predictive model according to can now be obtained from MRI), age (50–75 years), and PI-RADS version PI-RADS categories. Our main objective was to analyse the 1 score [14]. We introduced the MRI-based prostate volume and the PI- 40 EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 RADS version 2 categories. When the observed values were not within yr and the median serum PSA was 6.0 ng/ml (Table 2). the acceptable range, we entered the minimum or maximum acceptable The rate of abnormal DRE was 7.1%, the rate of PCa family value, whichever was closer to the observed value. history was 6.5%, and the rate of men with prior negative biopsy results was 28.4%. Prostate biopsy showed benign 2.3. Outcome measurements tissue in 116 men (68.6%) and PCa in 53 men (31.4%), of The main outcome measured was the rate of csPCa detection. csPCa was whom 25 (47.2%) had csPCa and 28 (52.8%) had iPCa. The defined as International Society of Urological Pathology grade group 2 csPCa and iPCa detection rates were 14.8% and 16.6%, [16]. The rate of prostate biopsies avoided and the rate of iPCa overdetec- respectively (p = 0.573). csPCa was detected in both guided tion were secondary outcome measurements. and systematic biopsies in 16 men (64%), and exclusively in five systematic biopsies (20%) and four guided biopsies 2.4. Statistical analysis (16%; p = 0.236). Quantitative variables are presented as the median and interquartile range (IQR). Qualitative variables are presented as the frequency and 3.3. Efficacy, net benefit, and clinical utility of Proclarix, proportion. Associations between variables were analysed using the PSAD and MRI-ERSPC predictive model for csPCa detection in PI- Mann-Whitney U test and the Kruskal-Wallis test. Associations between RADS 3 variables were also analysed with the v2 test. The odds ratio and 95% ROC curves for csPCa detection according to Proclarix, PSAD, confidence interval (CI) were estimated. Receiver operating characteris- and the ERSPC MRI model are presented in Fig. 2A. The AUC tic (ROC) curves were constructed and areas under the ROC curve (AUCs) was 0.703 (95% CI 0.621–0.768) for Proclarix, 0.612 (95% CI were estimated and compared with the DeLong test. The PSAD, Proclarix, 0.497–0.727) for PSAD, and 0. 657 (95% CI 0.547–0.766) for and ERSPC MRI model thresholds were selected to analyse the optimal sensitivity for csPCa. Sensitivity, specificity, positive and negative predic- the ERSPC MRI model (p = 0.027). DCA showed a net benefit tive values (PPV and NPV), accuracy, and rates of biopsies avoided, for Proclarix versus PSAD and the ERSPC MRI model at low overdetection of iPCa, and misdiagnosis of csPCa were estimated. Deci- thresholds within 0.09% versus 0.17% and 0.2%, respectively sion curve analysis (DCA) was carried out to assess the net benefits. Clin- (Fig. 2B). CUCs showing the rates of csPCa missed and biop- ical utility curves (CUCs) were used to check the correlation of rates of sies avoided in relation to the thresholds for the three tools csPCa misdiagnosis and biopsies avoided regarding the thresholds on a are presented in Fig. 2C. Analysis of Proclarix scores, PSAD continuous basis. A p value of <5% was considered significant. SPSS ver- values, and ERSPC MRI likelihood values showed that the sion 25 (IBM Corp., Armonk, NY, USA) and R version 3.3.1 (R Foundation thresholds with the highest sensitivity for csPCa were 10%, for Statistical Computing, Vienna, Austria) were used for statistical 0.07 ng/ml/cm3, and 1.5%, which yielded sensitivity of analyses. 100% (25/25), 84% (21/25), and 96% (24/25), respectively. The corresponding specificity was 5% (36/144) for Proclarix, 3. Results 28.5% (41/144) for PSAD, and 7.6% (11/144) for the ERSPC MRI model. The NPV and PPV were 100% (36/36) and 3.1. Efficacy of Proclarix, PSAD, and the ERSPC MRI 19.8% (25/133) for Proclarix, 91.1% (41/45) and 16% predictive model for csPCa detection across the PI-RADS (21/124) for PSAD, and 91.7% (11/12) and 15.3% (24/157) categories for the ERSPC MRI model, respectively. The diagnostic accu- racy was 36.1% (61/169) with Proclarix, 36.7% (62/169) with The distribution of PCa, csPCa, and iPCa among the 567 par- PSAD, and 20.7% (35/169) with the ERSPC MRI model. In ticipants by PI-RADS category is presented in Table 1. csPCa terms of clinical efficacy, Proclarix would avoid 21.3% was detected in 6% of men with PI-RADS <3, 14.8% of men (36/169) of prostate biopsies and reduce overdetection of with PI-RADS 3, 54.7% of men with PI-RADS 4, and 88% of iPCa from 16.6% to 11.2% (19/169) without misdiagnosing men with PI-RADS 5 findings (p < 0.001). Fig. 1 shows the csPCa. PSAD would avoid 26.2% (45/169) of prostate biop- efficacy of Proclarix, PSAD, and the ERSPC MRI predictive sies, reduce overdetection of iPCa from 16.6% to 11.2% model for the overall population and by PI-RADS category. (19/169), but misdiagnose 16% (four out of 25) of csPCa cases. The ERSPC MRI predictive model would avoid only 3.2. Characteristics of the cohort of men with PI-RADS 3 7.1% (12/169) of prostate biopsies, reduce overdetection of findings iPCa from 16.6% to 15.3% (26/169), and misdiagnose 4% Men with a PI-RADS 3 lesion represented 29.8% of all men (two out of 25) of csPCa cases, as shown in Table 3. The per- with suspected PCa for whom prebiopsy mpMRI data were formance of Proclarix, PSAD, and the ERSPC MRI model available (Table 1). The median age for this group was 66 according to biopsy status (biopsy-naïve vs repeat biopsy) Table 1 – Distribution of men with suspected PCa by PI-RADS category and the corresponding rates of PCa, csPCa, and iPCa PI-RADS category Men, n (%) PCa, n (%) csPCa, n (%) iPCa, n (%) 1 77 (13.6) 10 (13.0) 4 (5.2) 6 (7.8) 2 23 (4.1) 5 (21.7) 2 (8.7) 3 (13.0) 3 169 (29.8) 53 (31.4) 25 (14.8) 28 (16.6) 4 190 (33.5) 129 (67.9) 94 (54.7) 25 (13.2) 5 108 (19.0) 99 (91.7) 95 (88.0) 4 (3.7) All 567 (100) 296 (52.2) 230 (40.6) 66 (11.6) PI-RADS = Prostate Imaging-Reporting and Data System; PCa = prostate cancer; csPCa = clinically significant PCa; iPCa = insignificant PCa. EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 41 Fig. 1 – Efficacy of Proclarix, PSAD, and the ERSPC MRI predictive model for detection of clinically significant prostate cancer in the overall population and by PI-RADS category. Receiver operating characteristic curves and area under the curve (AUC) for (A) the overall population, (B) men with PI-RADS <3, (C) men with PI-RADS 3, (D) men with PI-RADS 4, and (E) men with PI-RADS 5 findings. PSAD = prostate-specific antigen density; ERSPC = European Randomized Study of Screening for Prostate Cancer; MRI = magnetic resonance imaging; PI-RADS = Prostate Imaging-Reporting and Data System; CI = confidence interval. Table 2 – Characteristics of men in PI-RADS category 3 observed differences in utility for the three tools analysed across the PI-RADS categories. We noted a progressive Characteristic Result increase in the csPCa detection rate across the PI-RADS cat- Number of cases 169 Median age, yr (IQR) 66 (60–72) egories, as well as differences in efficacy. Currently, most Median total PSA, ng/ml (IQR) 6.0 (3.6–10.2) clinicians accept that prostate biopsies can be avoided for Abnormal digital rectal examination, n (%) 12 (7.1) men with negative mpMRI because of its high NPV [4.5]. Median free PSA, ng/ml (IQR) 1.1 (0.7–1.6) Median prostate volume, ml (IQR) 66 (45–85) In the present study, the NPV of 94% observed represents Median percentage free PSA, % (IQR) 16.4 (11.5–20.7) an overall csPCa misdiagnosis rate of 2.6% and a 17.6% Median PSA density, ng/ml/cm3 (IQR) 0.11 (0.07–0.16) reduction in prostate biopsies, so it is acceptable to avoid Repeat biopsy, n (%) 48 (28.4) Family history of PCa, n (%) 11 (6.5) prostate biopsies in these men. By contrast, many clinicians Overall PCa detection, n (%) 53 (31.4) would not accept a test that does not guarantee 100% sensi- csPCa detection, n (%) 25 (14.8) tivity for csPCa in men with PI-RADS >3 lesions. Thus, it iPCa detection, n (%) 28 (16.6) makes sense to focus our attention on men with the recog- PI-RADS = Prostate Imaging-Reporting and Data System; IQR = interquar- tile range; PSA = prostate-specific antigen; PCa = prostate cancer; nised challenging PI-RADS 3 category [7,8]. csPCa = clinically significant PCa; iPCa = insignificant PCa. The incidence of PI-RADS 3 findings in our series of patients with suspected PCa was 29.8%, which is within is presented in Table 4. Proclarix outperformed PSAD and the range of 14–46% reported in the literature. This inci- the ERSPC MRI model, detecting all csPCa cases in both sub- dence mainly depends on the proportions of biopsy-naïve sets and avoiding 24.8% of prostate biopsies in the biopsy- men and men with a prior negative biopsy, which is approx- naïve group and 12.5% in the repeat biopsy group. imately 30% in mixed samples [7]. The incidence of csPCa detected among the 169 men with PI-RADS 3 findings (14.8%) is also within the wide range (5–30%) reported in 4. Discussion the literature [8]. We observed a net benefit of Proclarix over PSAD and the ERSPC MRI predictive model for deter- The efficacy of diagnostic tools changes in relation to the mining the likelihood of csPCa, especially at low thresholds, incidence of the disease in question. This is why we for which high sensitivity is observed. This finding is consis- Fig. 2 – Analysis of the efficacy, net benefit, and clinical utility of Proclarix, PSA density, and the ERSPC MRI predictive model for detection of clinically significant prostate cancer in men with a PI-RADS 3 lesion. (A) Receiver operating characteristic curves and area under the (AUC), (B) decision curve analysis, and (C) clinical utility curves. PSA = prostate-specific antigen; ERSPC = European Randomized Study of Screening for Prostate Cancer; MRI = magnetic resonance imaging; PI-RADS = Prostate Imaging-Reporting and Data System. 42 EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 Table 3 – Performance of Proclarix, PSAD, and the ERSPC MRI Modern markers have been analysed in the context of predictive model for csPCa detection in men in the PI-RADS 3 the current pathway for csPCa diagnosis and are intended category using the most sensitive thresholds to avoid mpMRI scans and subsequent prostate biopsies or Parameter csPCs detection, n/N (%) to select appropriate candidates for prostate biopsy after Proclarix PSAD ERSPC MRI RC mpMRI [17]. Some of these markers are combined with clin- (cutoff = 10%) (cutoff = 0.07 (cutoff = 1.5%) ical independent predictors in predictive models [15,18]. ng/ml/cm3) The Prostate Health Index (PHI) and PCA3 [19], PHI Sensitivity 25/25 (100) 21/25 (84.0) 24/25 (96.0) [20,21], 4K [22–25], and the Stockholm 3 test [26] have Specificity 36/144 (25.0) 41/144 (28.5) 11/144 (7.6) Negative 36/36 (100) 41/45 (91.1) 11/12 (91.7) been analysed, although their specific behaviours regarding predictive PI-RADS categories has never been reported as a main value research objective. Data for PHI and SelectMDx could be Positive 25/133 (19.8) 21/124 (16.0) 24/157 (15.3) predictive extracted from analyses of overall series published in the value literature [18,27,28] and one specific series of men with Accuracy 61/169 (36.1) 62/169 (36.7) 35/169 (20.7) PI-RADS 3 findings [29]. Supplementary Table 1 summarises Prostate 36/169 (21.3) 45/169 (26.2) 12/169 (7.1) biopsies the clinical utility of PHI [27] and SelectMDx [18,28,29] for avoided csPCa detection in men with suspected PCa and a PI-RADS 3 Decrease in overdetection 9/169 (5.3) iPCa lesion in comparison to Proclarix in the present study. Fan 9/169 (5.3) 2/169 (1.2) et al. [27] analysed the PHI performance for a group of 56 Misdiagnosis of 0/25 (0) 4/25 (16.0) 1/25 (4.0) men and observed a rate of avoidable biopsies of 67.9% csPCa while misdiagnosing one out of 16 men (6.3%) men with PI-RADS = Prostate Imaging-Reporting and Data System; PCa = prostate detected csPCa using a cutoff point of 50. Maggi et al. [18] cancer; iPCa = insignificant PCa; csPCa = clinically significant PCa; PSAD =prostate-specific antigen density; ERSPC = European Randomized Study reported that SelectMDx with a threshold of 13% misdiag- of Screening for Prostate Cancer; MRI = magnetic resonance imaging; nosed 12/14 men (85.7%) diagnosed with csPCa among a RC = risk calculator. sample of 54, avoiding 33.3% of prostate biopsies. Hendriks et al. [28] found that SelectMDx with a threshold of 13% misdiagnosed 7/9 men (77.8%) diagnosed with csPCa among tent with the ROC curves, according to which Proclarix out- a sample of 38 men, avoiding 40.2% of prostate biopsies. We performed PSAD and the ERSPC MRI predictive model, recently observed that SelectMDx with a threshold of 13% showing 100% sensitivity, although its AUC of 0.703 seems misdiagnosed four out of six men (66.7%) with csPCa in a suboptimal. This is positive for clinicians, given that Pro- sample of 62 men, avoiding 40.6% of prostate biopsies clarix can detect all csPCa cases while avoiding 21.3% of [29]. The present study shows that Proclarix is very sensi- all prostate biopsies (24.8% in biopsy-naïve men and 12.5% tive for csPCa, making it reliable enough to reassure clini- in those scheduled for repeat biopsy). Although PSAD seems cians. PHI seems more sensitive than SelectMDx but less slightly more specific than Proclarix, avoiding 26.3% of biop- sensitive than Proclarix. Multicentre validation studies sies, it misdiagnoses 16% of csPCa cases, which is hardly should be performed to confirm the effectiveness of any acceptable. This PSAD performance is similar to previous marker and cost-benefit studies regarding the quality- findings [11–13]. The accuracy of the ERSPC MRI predictive adjusted life years gained are desirable [30]. model was notably lower compared to Proclarix and PSAD, The present study was carried out on the largest pub- except at high thresholds, at which the sensitivity is very lished sample of men with suspected PCa and PI-RADS 3 low. Finally, the reduction in overdetection of iPCa was findings; however, the sample size may still be a limitation 5.3% for Proclarix and PSAD (decrease from 16.6% to because of the low incidence of csPCa. Unfortunately, we 11.2%) and 1.2% for the ERSPC MRI model. found no way to estimate the appropriate cohort size to Table 4 – Performance of Proclarix, PSAD, and the ERSPC MRI model for csPCa detection using the most sensitive threshold in biopsy-naïve men and men undergoing repeat biopsy Parameter csPCs detection, n/N (%) Proclarix (cutoff = 10%) PSAD (cutoff = 0.07 ng/ml/cm3) ERSPC MRI RC (cutoff = 1.5%) Initial Bx Repeat Bx Initial Bx Repeat Bx Initial Bx Repeat Bx Sensitivity 18/18 (100) 7/7 (100) 15/18 (83.3) 6/7 (85.7 17/18 (94.4) 7/7 (100) Specificity 30/103 (29.1) 6/41 (14.6) 33/103(32.0) 8/41 (19.5) 6/103 (5.8) 5/41 (12.2) Negative predictive value 30/30 (100) 6/6 (100 33/36 (91.7) 8/9 (88.9) 6/7 (85.7) 5/5 (100) Positive predictive value 18/91 (19.8) 7/42 (16.7) 15/85 (17.6) 6/39 (15.4) 17/114 (14.9) 7/43 (16.3) Accuracy 48/121 (39.7) 13/48 (27.1) 48/121(39.7) 14/48 (29.2 23/121 (19.0) 13/48 (27.1) Prostate biopsies avoided 30/121 (24.8) 6/48 (12.5) 36/121 (29.8) 9/48 (18.8) 7/121 (5.8) 5/48 (10.4 Decrease in iPCa overdetection 7/121 (5.8) 2/48 (4.2) 7/121 (5.8) 2/48 (4.2) 1/121 (0.8) 1/48 (2.1) Misdiagnosis of csPCa 0/18 (0) 0/7 (0) 3/18 (16.7) 1/7 (14.3) 1/8 (5.6) 0/7 (0) PSAD = prostate-specific antigen density; PCa = prostate cancer; csPCa = clinically significant PCa; iPCa = insignificant PCa; Bx = biopsy; ESPRC = European Randomized Study of Screening for Prostate Cancer; MRI = magnetic resonance imaging; RC = risk calculator. EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 43 assess the efficacy of certain tools owing to the lack of pre- References vious data. Although all studies use the same definition of [1] Hugosson J, Roobol MJ, Månsson M, et al. A 16-yr follow-up of the csPCa, we understand that the true incidence of csPCa European Randomized Study of Screening for Prostate Cancer. Eur observed in surgical specimens may be overestimated by Urol 2019;76:43–51. prostate biopsies. In the era of MRI and guided biopsies it [2] Mottet N, van den Bergh RCN, Briers E, et al. EAU-EANM-ESTRO- seems important to analyse the efficacy of any tool for ESUR-SIOG guidelines on prostate cancer—2020 update. Part 1: screening, diagnosis, and local treatment with curative intent. Eur improving the detection of csPCa regarding PI-RADS cate- Urol 2021;79:243–62. gories. Results for the overall efficacy, net benefits, and clin- [3] US Preventive Services Task Force, Grossman DC, Curry SJ, et al. ical utility may result in confusion for clinicians. The overall Screening for prostate cancer: US Preventive Services Task Force analyses are important, but they do not guarantee the same recommendation statement. JAMA 2018;319:1901–13. [4] Moldovan PC, Van den Broeck T, Sylvester R, et al. What is the effectiveness across the PI-RADS categories [12,13,28]. 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;71:618–29. 5. Conclusions [5] Sathianathen NJ, Omer A, Harriss E, et al. Negative predictive value of multiparametric magnetic resonance imaging in the detection of clinically significant prostate cancer in the Prostate Imaging The efficacy of tools for the appropriate selection of candi- Reporting and Data System era: a systematic review and meta- dates for prostate biopsy varies regarding PI-RADS cate- analysis. Eur Urol 2020;78:402–14. gories. Proclarix performed better than PSAD and the [6] Mazzone E, Stabile A, Pellegrino F, et al. Positive predictive value of ERSPC MRI predictive model in the challenging scenario of Prostate Imaging Reporting and Data System version 2 for the detection of clinically significant prostate cancer: a systematic PI-RADS category 3. Proclarix was able to reach 100% detec- review and meta-analysis. Eur Urol Oncol. In press. http://doi.org/ tion of csPCa, avoiding almost a quarter of unnecessary 10.1016/j.eururo.2021.07.027. prostate biopsies and reducing iPCa overdetection from [7] Schoots IG. MRI in early prostate cancer detection: how to manage indeterminate or equivocal PI-RADS 3 lesions. Transl Androl Urol 16.6% to 11.2%. 2018;7:70–82. [8] Görtz M, Radtke JP, Hatiboglu G, et al. The value of prostate-specific antigen density for Prostate Imaging-Reporting and Data System 3 Author contributions: Juan Morote had full access to all the data in the lesions on multiparametric magnetic resonance imaging: a strategy study and takes responsibility for the integrity of the data and the accu- to avoid unnecessary prostate biopsies. Eur Urol Focus racy of the data analysis. 2021;7:325–31. Study concept and design: Morote. [9] Klocker H, Golding B, Weber S, et al. Development and validation of a novel multivariate risk score to guide biopsy decision for the Acquisition of data: Celma, Regis, Planas, Semidey, de Torres, Mast. diagnosis of clinically significant prostate cancer. BJUI Compass Analysis and interpretation of data: Morote, Campistol, Triquell. 2020;1:15–20. Drafting of the manuscript: Morote, Campistol, Triquell. [10] Steuber T, Heidegger I, Kafka M, et al. PROPOSe: a real-life Critical revision of the manuscript for important intellectual content: Trilla, prospective study of Proclarix, a novel blood-based test to support challenging biopsy decision-making in prostate cancer. Santamaria.Statistical analysis: Morote.Obtaining funding: Morote. Eur Urol Oncol. In press. http://doi.org/10.1016/j.euo.2020.12.003. Administrative, technical, or material support: None. [11] Boesen L, Nørgaard N, Løgager V, et al. Prebiopsy biparametric Supervision: Morote, Trilla. magnetic resonance imaging combined with prostate-specific Other: None. antigen density in detecting and ruling out Gleason 7–10 prostate cancer in biopsy-naïve men. Eur Urol Oncol 2019;2:311–9. [12] Morote J, Celma A, Planas J. Re: Lars Boesen, Nis Nørgaard, Vibeke Financial disclosures: Juan Morote certifies that all conflicts of interest, Løgager, et al. Prebiopsy biparametric magnetic resonance imaging combined with prostate-specific antigen density in detecting and including specific financial interests and relationships and affiliations rel- ruling out Gleason 7–10 prostate cancer in biopsy-naïve men. Eur evant to the subject matter or materials discussed in the manuscript (eg, Urol Oncol 2019;2:311–9. employment/affiliation, grants or funding, consultancies, honoraria, stock [13] Morote J, Celma A, Díaz F, et al. Prostatic-specific antigen density ownership or options, expert testimony, royalties, or patents filed, behavior according to multiparametric magnetic resonance imaging result. Urol Oncol 2020;38:410–7. received, or pending), are the following: None. [14] Alberts AR, Roobol MJ, Verbeek JFM, et al. Prediction of high-grade prostate cancer following multiparametric magnetic resonance imaging: improving the Rotterdam European Randomized Study Funding/Support and role of the sponsor: This study was financed in part of Screening for Prostate Cancer risk calculators. Eur Urol by Instituto de Salud Carlos III (ES) (PI20/01666). Proteomedix provided 2019;75:310–8. Proclarix evaluations. The sponsors played no direct role in the study. [15] Macagno A, Athanasiou A, Wittig A, et al. Analytical performance of thrombospondin-1 and cathepsin D immunoassays part of a novel CE-IVD marked test as an aid in the diagnosis of prostate cancer. Acknowledgments: The likelihood of csPCa according to the ERSPC MRI PLoS One 2020;15:e0233442. [16] Epstein JI, Egevad L, Amin MB, et al. The 2014 International Society risk calculator was assessed using the Prostate Cancer Research Founda- of Urological Pathology (ISUP) consensus conference on Gleason tion (Reeuwijk, The Netherlands) web application at www.prostatecan- grading of prostatic carcinoma: definition of grading patterns and cer-riskcalculator.com. proposal for a new grading system. Am J Surg Pathol 2016;40:244–52. [17] Vickers AJ, Russo G, Lilja H, et al. How should molecular markers and magnetic resonance imaging be used in the early detection of Appendix A. Supplementary data prostate cancer. Eur Urol Oncol. In press. http://doi.org/10.1016/j. euo.2021.01.010. [18] Fan YH, Pan PH, Cheng WM, et al. The Prostate Health Index aids Supplementary data to this article can be found online at multi-parametric MRI in diagnosing significant prostate cancer. Sci https://doi.org/10.1016/j.euros.2021.12.009. Rep 2021;11:1286. 44 EUROPEAN UROLOGY OPEN SCIENCE 37 (2022) 38–44 [19] Porpiglia F, Russo F, Manfredi M, et al. The roles of multiparametric [25] Verbeek JFM, Bangma CH, Kweldam CF, et al. Reducing unnecessary MRI, PCA3, and PHI: which is the best predictor of prostate cancer biopsies while detecting clinically significant prostate cancer after a negative biopsy? Results of a prospective study. J Urol including cribriform growth with the ERSPC Rotterdam risk 2014;192:60–6. calculator and 4Kscore. Urol Oncol 2019;37:138–44. [20] Roobol MJ, Vedder MM, Nieboer D, et al. Comparison of two [26] Grönberg H, Eklund M, Picker W, et al. Prostate cancer diagnostics prostate cancer risk calculators that include the Prostate Health using a combination of the Stockholm3 blood test and Index. Eur Urol Focus 2015;1:185–90. multiparametric magnetic resonance imaging. Eur Urol [21] Gnanapragasam VJ, Burling K, George A, et al. The Prostate Health 2018;74:722–8. Index adds predictive value to multi-parametric MRI in detecting [27] Maggi M, Del Giudice F, Falagario UG, et al. SelectMDx and significant prostate cancers in a repeat biopsy population. Sci Rep multiparametric magnetic resonance imaging of the prostate for 2016;6:35364. men undergoing primary prostate biopsy: a prospective assessment [22] Punnen S, Nahar B, Soodana-Prakash N, et al. Optimizing patient’s in a multi-institutional study. Cancers 2021;13:2047. selection for prostate biopsy: a single institution experience with [28] Hendriks RJ, van der Leest MMG, Israël B, et al. Clinical use of the multi-parametric MRI and the 4Kscore test for the detection of SelectMDx urinary-biomarker test with or without mpMRI in aggressive prostate cancer. PLoS One 2018;13:e0201384. prostate cancer diagnosis: a prospective, multicenter study in [23] Falagario UG, Martini A, Wajswol E, et al. Avoiding unnecessary biopsy-naïve men. Prostate Cancer Prostat Dis 2021;24:1110–9. magnetic resonance imaging (MRI) and biopsies: negative and [29] Morote J, Díaz F, Celma A, et al. Behaviour of SelectMDx and positive predictive value of MRI according to prostate-specific prostate-specific antigen density in the challenging scenario of antigen density, 4Kscore and risk calculators. Eur Urol Oncol Prostate Imaging-Reporting and Data System category 3 lesions. 2020;3:700–4. Eur Urol. In press. https://doi.org/10.1016/j.eururo.2021.09.019. [24] Marzouk K, Ehdaie B, Vertosick E, Zappala S, Vickers A. Developing [30] Govers TM, Hessels D, Vlaeminck-Guillem V, et al. Cost- an effective strategy to improve the detection of significant prostate effectiveness of SelectMDx for prostate cancer in four European cancer by combining the 4Kscore and multiparametric MRI. Urol countries: a comparative modeling study. Prostate Cancer Prostat Oncol 2019;37:672–7. Dis 2018;22:101–9.