Performance of a Region of Interest-based Algorithm in Diagnosing International Society of Urological Pathology Grade Group ≥2 Prostate Cancer on the MRI-FIRST Database-CAD-FIRST Study.

BACKGROUND AND OBJECTIVE: Prostate multiparametric magnetic resonance imaging (MRI) shows high sensitivity for International Society of Urological Pathology grade group (GG) ≥2 cancers. Many artificial intelligence algorithms have shown promising results in diagnosing clinically significant prostate cancer on MRI. To assess a region-of-interest-based machine-learning algorithm aimed at characterising GG ≥2 prostate cancer on multiparametric MRI. METHODS: The lesions targeted at biopsy in the MRI-FIRST dataset were retrospectively delineated and assessed using a previously developed algorithm. The Prostate Imaging-Reporting and Data System version 2 (PI-RADSv2) score assigned prospectively before biopsy and the algorithm score calculated retrospectively in the regions of interest were compared for diagnosing GG ≥2 cancer, using the areas under the curve (AUCs), and sensitivities and specificities calculated with predefined thresholds (PIRADSv2 scores ≥3 and ≥4; algorithm scores yielding 90% sensitivity in the training database). Ten predefined biopsy strategies were assessed retrospectively. KEY FINDINGS AND LIMITATIONS: After excluding 19 patients, we analysed 232 patients imaged on 16 different scanners; 85 had GG ≥2 cancer at biopsy. At patient level, AUCs of the algorithm and PI-RADSv2 were 77% (95% confidence interval [CI]: 70-82) and 80% (CI: 74-85; p = 0.36), respectively. The algorithm's sensitivity and specificity were 86% (CI: 76-93) and 65% (CI: 54-73), respectively. PI-RADSv2 sensitivities and specificities were 95% (CI: 89-100) and 38% (CI: 26-47), and 89% (CI: 79-96) and 47% (CI: 35-57) for thresholds of ≥3 and ≥4, respectively. Using the PI-RADSv2 score to trigger a biopsy would have avoided 26-34% of biopsies while missing 5-11% of GG ≥2 cancers. Combining prostate-specific antigen density, the PI-RADSv2 and algorithm's scores would have avoided 44-47% of biopsies while missing 6-9% of GG ≥2 cancers. Limitations include the retrospective nature of the study and a lack of PI-RADS version 2.1 assessment. CONCLUSIONS AND CLINICAL IMPLICATIONS: The algorithm provided robust results in the multicentre multiscanner MRI-FIRST database and could help select patients for biopsy. PATIENT SUMMARY: An artificial intelligence-based algorithm aimed at diagnosing aggressive cancers on prostate magnetic resonance imaging showed results similar to expert human assessment in a prospectively acquired multicentre test database.

Managing Discordant Findings Between Multiparametric Magnetic Resonance Imaging and Transrectal Magnetic Resonance Imaging-directed Prostate Biopsy-The Key Role of Magnetic Resonance Imaging-directed Transperineal Biopsy.

BACKGROUND: Discordant findings between multiparametric magnetic resonance imaging (mpMRI) and transrectal image-guided biopsies of the prostate (TRUS-P) may result in inadequate risk stratification of localized prostate cancer. OBJECTIVE: To assess transperineal image-guided biopsies of the index target (TPER-IT) in terms of disease reclassification and treatment recommendations. DESIGN, SETTING, AND PARTICIPANTS: Cases referred for suspicion or treatment of localized prostate cancer were reviewed in a multidisciplinary setting, and discordance was characterized into three scenarios: type I-negative biopsies or International Society of Urological Pathology (ISUP) grade 1 cancer in Prostate Imaging Reporting and Data System (PI-RADS) ≥4 index target (IT); type II-negative biopsies or ISUP grade 1 cancer in anterior IT; and type III-<3 mm stretch of cancer in PI-RADS ≥3 IT. Discordant findings were characterized in 132/558 (23.7%) patients after TRUS-P. Of these patients, 102 received reassessment TPER-IT. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: The primary objective was to report changes in treatment recommendations after TPER-IT. Therefore, cores obtained by primary TRUS-P and TPER-IT were analyzed in terms of cancer detection, ISUP grade, and Cambridge Prognostic Group classification using descriptive statistics. RESULTS AND LIMITATIONS: TPER-IT biopsies that consisted of fewer cores than the initial TRUS-P (seven vs 14, p < 0.0001) resulted in more cancer tissue materials for analysis (56 vs 42.5 mm, p = 0.0003). As a result, 40% of patients initially considered for follow-up (12/30) and 49% for active surveillance (30/61) were reassigned after TPER-IT to surgery or intensity-modulated radiotherapy. CONCLUSIONS: Nonconcordance between pathology and imaging was observed in a significant proportion of patients receiving TRUS-P. TPER-IT better informed the presence and grade of cancer, resulting in a significant impact on treatment recommendations. A multidisciplinary review of mpMRI and TRUS-P findings and reassessment TPER-IT in type I-II discordances is recommended. PATIENT SUMMARY: In this report, patients with suspicious imaging of the prostate, but no or well-differentiated cancer on transrectal image-guided -biopsies, were offered transperineal image-guided biopsies for reassessment. We found that a large share of these had a more aggressive cancer than initially suspected. We conclude that discordant results warrant reassessment transperineal image-guided biopsies as these may impact disease risk classification and treatment recommendations.

Assessment of the Minimal Targeted Biopsy Core Number per MRI Lesion for Improving Prostate Cancer Grading Prediction.

BACKGROUND: To study the impact of MRI characteristics and of targeted biopsy (TB) core number on the final grade group (GG) prediction. MATERIALS AND METHODS: The cohort was 478 consecutive patients who underwent radical prostatectomy (RP) after positive mpMRI (multiparametric magnetic resonance imaging) followed by fusion TB. Endpoints were the upgrading and concordance rates between TB and RP specimens. RESULTS: Upgrading rate after TB was 40.6%. Patients with upgrading had lower PIRADS (Prostate Imaging-Reporting and Data System) scores (p < 0.001), smaller lesion size (p = 0.017), fewer TB cores (p < 0.001), and lower TB density (p = 0.015) compared with cases with grade concordance. There was a significant continuous improvement in upgrading rate when TB core number per lesion increased from 56.3% to 25.6% when <2 or ≥5 TB cores were taken, respectively (p = 0.002). The minimal TB number per lesion to reduce upgrading risk to approximately 30%was 4 in PIRADS 3, and 3 in PIRADS 4-5 cases. CONCLUSIONS: Grade group prediction by TB is significantly improved by higher PIRADS score, larger lesion size, and increased TB per lesion. At least four TB cores should be taken in PIRADS 3 score lesions, whereas three cores seem enough in PIRADS 4-5 cases to improve GG prediction and limit upgrading risk.

Impact of MRI and Targeted Biopsies on Eligibility and Disease Reclassification in MRI-positive Candidates for Active Surveillance on Systematic Biopsies.

OBJECTIVE: To assess the impact of concomitant targeted biopsies (TB) for predicting final disease reclassification in MRI-positive low-risk prostate cancer patients eligible for active surveillance (AS) on systematic biopsies (SB). MATERIALS AND METHODS: From a prospective database, we included all prebiopsy MRI-positive men fulfilling AS criteria at diagnosis (Toronto [n = 114], UCSF [n = 82], or PRIAS [n = 60] criteria) on SB. All patients underwent a combination of SB and software-based fusion TB, and an immediate radical prostatectomy. The primary endpoints were the pathologic upgrading and upstaging rates. RESULTS: Biopsy grade group was upgraded to grade group (GG) 2 and to GG≥3 on TB in 65.9%-76.7% and in 12.2-16.7%, respectively. The rate of GG ≥3 in radical prostatectomy specimens varied from 31.6% to 43.3% with no relation between strictest criteria and lower upgrading rates. The proportion of not organ-confined disease (35%-39%) was comparable among the AS cohorts. Negative TB was strongly associated with the absence of final GG ≥3. Tumor grade on TB was significantly correlated with the risk of final GG ≥3 in both Toronto and UCSF cohorts, not in the PRIAS cohort. In the PRIAS cohort, the only independent predictive factor for GG ≥3 disease was the maximal tumor length in any core (P = .034). CONCLUSION: In MRI-positive patients, the risk of disease reclassification was comparable whatever the SB-based AS criteria used. TB were predictive of final upgrading, with a varied impact according to the AS criteria. SB features remained relevant for reclassification prediction even in case of positive TB. The risk of upstaged disease remains important, approximately one third, and neither TB/SB parameters nor MRI findings could accurately predict it.

Performance of systematic, MRI-targeted biopsies alone or in combination for the prediction of unfavourable disease in MRI-positive low-risk prostate cancer patients eligible for active surveillance.

PURPOSE: To assess the upstaging/upgrading rates of low-risk prostate cancer (PCa) according to the biopsy scheme used (systematic (SB), targeted biopsies (TB), or both) in the setting of positive pre-biopsy MRI. PATIENTS AND METHODS: We included 143 consecutive men fulfilling the Toronto University active surveillance (AS) criteria who underwent a pre-biopsy positive MRI, a combination of SB and software-based fusion TB, and a radical prostatectomy, in two expert centres. The primary endpoints were the pathological upgrading and upstaging rates. Overall unfavourable disease (OUD) was defined by any pT3-4 and/or pN1 and/or ≥ GG 3. RESULTS: Using TB alone would have missed 21.7% of cancers including 16.7% of ≥ GG 3. The use of TB was significantly associated with a lower risk of ≥ Grade Group (GG) 3 disease (p < 0.006) in RP specimens. Combination of SB and TB lowered this risk by 39%, compared with TB alone. The biopsy scheme did not affect the upstaging rates which were substantial even in case of combination scheme (from 37 to 46%). OUD was detected in approximately 50% of cases. The presence of high grade on TB was the only independent predictive factor for both ≥ GG 2 (p = 0.015) and ≥ GG 3 (p = 0.023) in RP specimens. CONCLUSIONS: High grade on TB biopsies represented the major predictor of upgrading. Combination of SB and TB better defined the sub-group of patients having the lowest risk of reclassification, compared with TB or SB alone. The risk of non-organ-confined disease remained high, and could not be accurately predicted by MRI or systematic/targeted biopsy features.

Active surveillance eligibility of MRI-positive patients with grade group 2 prostate cancer: a pathological study.

PURPOSE: To assess the final pathology risk in MRI-positive grade group (GG) 2 prostate cancer (PCa) patients undergoing targeted (TB) and systematic (SB) biopsies, and thereby, the possibility of active surveillance (AS) in this population. PATIENTS AND METHODS: We included 242 consecutive men diagnosed with GG2 PCa by a combination of SB and software-based fusion TB undergoing a radical prostatectomy (RP). The primary endpoints were the pathological findings in RP specimens, including favourable disease which was defined by a pT2 and GG1-2 disease. RESULTS: The rate of upgrading was 33% including 3% of GG 4-5 disease. MRI lesion size (p = 0.038) and tumor length per core (p < 0.001) were significantly lower in case of favourable pathology. Only 34.2% of not organ-confined disease was reported when only SB were positive, compared with 45.7% and 57.1% when GG2 was detected on TB only and on TB plus SB, respectively (p = 0.035). The number of positive cores on SB was significantly higher in not organ-confined disease (4.3 versus 2.9; p = 0.005). The risk of not organ-confined disease was only 20.8% in men who had a PSAD ≤ 0.20 ng/ml/gr, 1-2 positive biopsies and a maximal tumor length ≤ 6 mm per core, compared with 52.3% in men who did not fulfil all these criteria (p = 0.003). CONCLUSIONS: This study identified clinical, imaging, and pathological factors that were significantly associated with the final pathology risk. In case of positive MRI followed by TB showing GG2, AS could be offered in patients having a PSAD ≤ 0.20, a tumor length ≤ 6 mm and 1-2 positive cores.

Added Value of Concomitant Systematic and Fusion Targeted Biopsies for Grade Group Prediction Based on Radical Prostatectomy Final Pathology on Positive Magnetic Resonance Imaging.

PURPOSE: We assessed the added value of concomitant systematic biopsy for final grade group prediction in patients with positive magnetic resonance imaging who were undergoing targeted biopsy. MATERIALS AND METHODS: Included in study were 478 consecutive patients with prebiopsy positive multiparametric magnetic resonance imaging and a greater than 10-core systematic biopsy combined with fusion targeted biopsy who underwent radical prostatectomy. The primary end point was the grade group concordance between biopsy and radical prostatectomy pathology according to the biopsy technique. Clinical and biological factors associated with the performance of systematic biopsy were analyzed. RESULTS: Adding systematic biopsy to targeted biopsy modified the d'Amico risk classification toward more intermediate and high risk in 7.8% of cases, mainly from low to intermediate risk with low risk prostate cancer on targeted biopsy in 44.3%. This reclassification was significantly higher in patients with lower prostate specific antigen and with prostate specific antigen density less than 0.20 ng/ml/gm (11.7% vs 2.4%, p <0.001). The concordance rate between biopsy pathology and radical prostatectomy pathology significantly differed between targeted biopsy and targeted biopsy plus systematic biopsy (45.2% and 51.7%, respectively). The upgrading rate in radical prostatectomy specimens decreased by 22% when systematic biopsy was added to targeted biopsy. Patients in whom systematic biopsy did not modify grading were more likely to have pT3-4 and/or pN1 disease on final pathology (56.9% vs 38.3%, p=0.007). CONCLUSIONS: Grading concordance between biopsy pathology and radical prostatectomy pathology was improved by adding systematic biopsy in all patient subgroups. Patients with prostate specific antigen density less than 0.20 ng/ml/gm benefited the most from this combined biopsy strategy. Systematic biopsy reclassified a nonnegligible number of cases toward a higher risk category, mainly the low risk cases. Thus, systematic biopsy could modify treatment decision making.

Refining the risk-stratification of transrectal biopsy-detected prostate cancer by elastic fusion registration transperineal biopsies.

PURPOSE: To evaluate image-guided Transperineal Elastic-Registration biopsy (TPER-B) in the risk-stratification of low-intermediate risk prostate cancer detected by Transrectal-ultrasound biopsy (TRUS-B) when estimates of cancer grade and volume discorded with multiparametric Magnetic Resonance Imaging (MRI). METHODS: All patients referred for active surveillance or organ-conservative management were collegially reviewed for consistency between TRUS-B results and MRI. Image-guided TPER-B of the index target (IT) defined as the largest Prostate Imaging-Reporting Data System-v2 ≥ 3 abnormality was organized for discordant cases. Pathology reported Gleason grade, maximum cancer core length (MCCL) and total CCL (TCCL). RESULTS: Of 237 prostate cancer patients (1-4/2018), 30 were required TPER-B for risk-stratification. Eight cores were obtained [Median and IQR: 8 (6-9)] including six (IQR: 4-6) in the IT. TPER-B of the IT yielded longer MCCL [Mean and (95%CI): 6.9 (5.0-8.8) vs. 2.6 mm (1.9-3.3), p < 0.0001] and TCCL [19.7 (11.6-27.8) vs. 3.6 mm (2.6-4.5), p = 0.0002] than TRUS-B of the gland. On TPER-B cores, longer MCCL [Mean and (95%CI): 8.7 mm (6.7-10.7) vs. 4.1 mm (0.6-7.6), p = 0.002] were measured in Gleason score-7 cancers. TPER-B cores upgraded 13/30 (43.3%) patients. 14/30 (46.7%) met University College London-definition 1 and 18/30 (60.0%) definition 2, which correlate with clinically significant cancers > 0.5 mL and > 0.2 mL, respectively. 7/16 (43.8%) patients under active surveillance were re-allocated toward prostatectomy (n = 5) or radiation therapy (n = 2). In 14 patients not yet assigned, TPER-B risk-stratification spurred the selection (13/14, 92.9%) of treatments with curative intent. CONCLUSION: Image-guided TPER-B of the index target provided more cancer material for pathology. Subsequent re-evaluation of cancer volume and grade switched a majority of patients towards higher-risk groups and treatments with curative intent.

New approach of ultra-focal brachytherapy for low- and intermediate-risk prostate cancer with custom-linked I-125 seeds: A feasibility study of optimal dose coverage.

PURPOSE: To present the feasibility study of optimal dose coverage in ultra-focal brachytherapy (UFB) with multiparametric MRI for low- and intermediate-risk prostate cancer. METHODS AND MATERIALS: UFB provisional dose plans for small target volumes (<7 cc) were calculated on a prostate training phantom to optimize the seeds number and strength. Clinical UFB consisted in a contour-based nonrigid registration (MRI/Ultrasound) to implant a fiducial marker at the location of the tumor focus. Dosimetry was performed with iodine-125 seeds and a prescribed dose of 160 Gy. On CT scans acquired at 1 month, dose coverage of 152 Gy to the ultra-focal gross tumor volume was evaluated. Registrations between magnetic resonance and CT scans were assessed on the first 8 patients with three software solutions: VariSeed, 3D Slicer, and Mirada, and quantitative evaluations of the registrations were performed. Impact of these registrations on the initial dose matrix was performed. RESULTS: Mean differences between simulated dose plans and extrapolated Bard nomogram for UFB volumes were 36.3% (26-56) for the total activity, 18.3% (10-30) for seed strength, and 22.5% (16-38) for number of seeds. Registration method implemented in Mirada performed significantly better than VariSeed and 3D Slicer (p = 0.0117 and p = 0.0357, respectively). For dose plan evaluation between Mirada and VariSeed, D100% (Gy) for ultra-focal gross tumor volume had a mean difference of 28.06 Gy, mean values being still above the objective of 152 Gy. D90% for the prostate had a mean difference of 1.17 Gy. For urethra and rectum, dose limits were far below the recommendations. CONCLUSIONS: This UFB study confirmed the possibility to treat with optimal dose coverage target volumes smaller than 7 cc.

IDEAL 2a Phase II Study of Ultrafocal Brachytherapy for Low- and Intermediate-risk Prostate Cancer.

PURPOSE: Focal therapy of prostate cancer requires precise positioning of therapeutic agents within well-characterized index tumors (ITs). We assessed the feasibility of low-dose-rate ultrafocal brachytherapy. METHODS AND MATERIALS: The present study was an institutional review board-approved European Clinical Trials Database-registered phase II protocol. Patients referred (October 2013 to August 2016) for active surveillance (prostate-specific antigen <10 ng/mL, cT1c-cT2a, Gleason score on referring biopsy specimens ≤6 (3+3), ≤3 positive biopsy cores, ≤50% of cancer) were preselected. Inclusion was confirmed when complementary image-guided biopsy findings informed a single Prostate Imaging Reporting and Data System (PI-RADS) ≥3, Gleason score ≤7a (3+4) lesion. A ultrasound-visible ancillary marker was positioned within the IT using a magnetic resonance imaging (MRI)/3-dimensional transrectal ultrasound (TRUS) elastic fusion-guided system (Koelis). Ultrafocal transperineal delivery of 125I seeds used classic 2-dimensional TRUS (Bard-FlexFocus) and dose optimization (Variseed Treatment Planning System). Following Simon's optimal design, 17 patients were required to assess the feasibility of delivering ≥95% of the prescribed dose (160 Gy) to the IT (primary objective). Adverse events (Common Terminology Criteria for Adverse Events) and quality of life (5-item International Index of Erectile Function, International Prostate Symptom Score) were recorded. One-year control biopsy specimens were obtained from the IT and untreated segments. RESULTS: Of the 44 preselected patients, 27 did not meet the inclusion criteria. Of the 17 ultrafocal brachytherapy-treated patients, 16 met the primary objective (per protocol success). The prescription dose was delivered to 14.5% ± 6.4% of the prostate volume, resulting in negligible urethral and rectal irradiation and toxicity. No recurrence was evidenced on the 1-year follow-up MRI studies or IT biopsy specimens. Seven nonclinically significant cancers and one Gleason score 7a (3+4) cancer (salvage prostatectomy) were observed in the untreated parenchyma. CONCLUSIONS: Recent technology has allowed for selective and effective brachytherapy of small MRI targets.

Precision Matters in MR Imaging-targeted Prostate Biopsies: Evidence from a Prospective Study of Cognitive and Elastic Fusion Registration Transrectal Biopsies.

Purpose To measure the precision in placement of a biopsy needle in a magnetic resonance (MR) imaging-detected target with transrectal ultrasonography (US), to document the clinical relevance of precision, and to report on the precision of cognitive and software-based registrations. Materials and Methods This prospective study was approved by the institutional review board and performed between June 2013 and September 2013. Patients provided informed verbal consent. Two cores each were obtained with cognitive and fusion techniques in 88 patients with a Prostate Imaging Reporting and Data System version 1 score of at least 3. Precision was measured with Euclidian geometry by using the Digital Imaging and Communications in Medicine archives of the biopsy as the distance from the core to the center (dCC) and the distance from the core to the surface of the target modeled as a sphere. To address clustering of data from multiple cores in the same patients, analyses of precision focused on the best shot for a patient or a technique. The Welch unequal variance t test and Yates corrected χ2 test were used as appropriate. Results Mean precision was 2.5 mm (95% confidence interval: 1.8 mm, 3.3 mm). Positive cores were closer to the center than were negative cores (dCC: 1.7 mm vs 3.1 mm, respectively; P = .025). More cancers were detected with on-target than off-target cores (33 of 71 cores [46.5%] vs three of 17 cores [17.6%]; P = .03). Cores obtained with the fusion technique achieved a higher precision than did cores obtained with the cognitive technique (dCC: 2.8 mm vs 7.1 mm, respectively; P < .0001). Targeted cores demonstrated cancer in 44 patients. Fewer cancers were detected with the cognitive technique than with the fusion technique (31 of 44 patients [70.5%] vs 40 of 44 patients [90.9%]; P = .03). Conclusion A deformable MR imaging/transrectal US image registration system achieved a higher precision and depicted cancer in more patients than did the cognitive freehand technique. © RSNA, 2018.

Are Magnetic Resonance Imaging-Transrectal Ultrasound Guided Targeted Biopsies Noninferior to Transrectal Ultrasound Guided Systematic Biopsies for the Detection of Prostate Cancer?

PURPOSE: In men with suspicion of prostate cancer the standard of cancer detection is transrectal ultrasound guided 10 to 12-core systematic biopsy. The targeted biopsy only strategy using magnetic resonance imaging-transrectal ultrasound image registration is gaining in popularity. We assessed the noninferiority of targeted vs systematic biopsy. MATERIALS AND METHODS: Between June and October 2014 a total of 108 biopsy naïve patients with prostate specific antigen between 4 and 20 ng/ml, normal rectal examination and a single suspicious image on magnetic resonance imaging were included in study at 7 centers. Patients underwent systematic biopsy by a first operator blinded to magnetic resonance imaging, immediately followed by 3 targeted biopsies within the suspicious image by a second operator. The primary end point was the cancer detection rate. The noninferiority margin was set at -5%. The secondary end points were the detection rate of clinically significant prostate cancer (maximum cancer core length 5 mm or greater for Gleason 6 or any Gleason 7 or greater disease) and procedure duration. RESULTS: Systematic and targeted biopsies detected cancer in 66 (61.1%) and 61 patients (56.5%), respectively. The mean difference was -4.5% with a 95% CI lower bound of -11.8%. A total of 13 patients with protocol violations were excluded from the per protocol analysis, which showed a mean difference of -5.2% with a 95% CI lower bound of -13.1%. Clinically significant prostate cancer was detected in 50 (46.2%) and 52 patients (48.1%) with systematic and targeted biopsies, respectively (p = 0.69). The mean ± SD duration of image fusion plus targeted biopsy was 16.7 ± 7 minutes vs 7.4 ± 3 for systematic biopsy (p <0.001). CONCLUSIONS: Targeted biopsy seemed to be inferior to systematic biopsy for overall cancer detection. Detection of clinically significant prostate cancer did not differ between targeted and systematic biopsies.

Prostate Imaging Reporting and Data System and Likert Scoring System: Multiparametric MR Imaging Validation Study to Screen Patients for Initial Biopsy.

PURPOSE: To compare the diagnostic performance of the magnetic resonance (MR) imaging-based Prostate Imaging Reporting and Data System (PI-RADS) and a Likert scale in the detection of prostate cancer in a cohort of patients undergoing initial prostate biopsy. MATERIALS AND METHODS: This institutional review board-approved two-center prospective study included 118 patients with normal digital rectal examination (DRE) results but elevated prostate-specific antigen (PSA) levels (4-20 ng/mL) who were referred for initial prostate biopsies and had one suspicious (Likert scale score, ≥3) focus at prebiopsy 1.5-T multiparametric MR imaging performed with T2-weighted, diffusion-weighted [DW], and dynamic contrast material-enhanced imaging. Targeted core biopsies and random systematic core biopsies were performed. The elementary unit for analysis was the core. Relationships were assessed by using the Mann-Whitney U test. Yates corrected and Pearson χ(2) tests were used to evaluate categoric variables. A training set was randomly drawn to construct the receiver operating characteristic curves for the summed PI-RADS scores and for the Likert scale scores. The thresholds to recommend biopsy were obtained from the Youden J statistics and were tested in the remaining validation set in terms of predictive characteristics. Interobserver variability was analyzed by using weighed κ statistics in a random set of 50 patients. RESULTS: Higher T2-weighted, DW, and dynamic contrast-enhanced imaging PI-RADS scores were observed in areas that yielded cancer-positive cores. The percentage of positive cores increased with the sum of scores aggregated in five classes as follows: For summed PI-RADS scores of 3-5, the percentage of positive cores was 2.3%; for scores of 6-8, it was 5.8%; for scores of 9 or 10, it was 24.7%; for scores of 11 or 12, it was 51.8%; and for scores of 13-15, it was 72.1% (P for trend, <.0001). For the threshold of summed PI-RADS scores of 9 or greater, sensitivity was 86.6%, specificity was 82.4%, the positive predictive value was 52.4%, the negative predictive value was 96.5%, and accuracy was 83.2%. The respective data for Likert scale scores of 3 or greater were 93.8%, 73.6%, 44.3%, 98.1%, and 73.3%. Good interobserver agreement was observed for the Likert scale (κ = 0.80) and the summed PI-RADS (κ = 0.73) scoring systems. CONCLUSION: PI-RADS provided the site-specific stratified risk of cancer-positive cores in biopsy-naive men with normal DRE results and elevated PSA levels. There was no significant difference between summed PI-RADS scores of 9 or greater and Likert scale scores of 3 or greater in the detection of cancer in the peripheral zone.

Validation of the European Society of Urogenital Radiology scoring system for prostate cancer diagnosis on multiparametric magnetic resonance imaging in a cohort of repeat biopsy patients.

BACKGROUND: Wide variations in acquisition protocols and the lack of robust diagnostic criteria make magnetic resonance imaging (MRI) detection of prostate cancer (PCa) one of the most challenging fields in radiology and urology. OBJECTIVE: To validate the recently proposed European Society of Urogenital Radiology (ESUR) scoring system for multiparametric MRI (mpMRI) of the prostate. DESIGN, SETTING, AND PARTICIPANTS: An institutional review board-approved multicentric prospective study; 129 consecutive patients (1514 cores) referred for mpMRI after at least one set of negative biopsies. INTERVENTION: Transfer of mpMRI-suspicious areas on three-dimensional (3D) transrectal ultrasound images by 3D elastic surface registration; random systematic and targeted cores followed by core-by-core analysis of pathology and mpMRI characteristics of the core locations. The ESUR scores were assigned after the procedure on annotated Digital Imaging and Communications in Medicine archives. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Relationships between ESUR scores and biopsy results were assessed by the Mann-Whitney U test. The Yates correction and Pearson χ(2) tests evaluated the association between categorical variables. A teaching set was randomly drawn to construct the receiver operating characteristic curve of the ESUR score sum (ESUR-S). The threshold to recommend biopsy was obtained from the Youden J statistics and tested in the remaining validation set in terms of sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. RESULTS AND LIMITATIONS: Higher T2-weighted, dynamic weighted imaging and dynamic contrast-enhanced ESUR scores were observed in areas yielding cancer-positive cores. The proportion of positive cores increased with the ESUR-S aggregated in five increments (ESUR-S 3-5: 2.9%; ESUR-S 6-8: 11.1%; ESUR-S 9-10: 38.2%; ESUR-S 11-12: 63.4%; and ESUR-S 13-15: 83.3%; p<0.0001). A threshold of ESUR-S ≥ 9 exhibited the following characteristics: sensitivity: 73.5%; specificity: 81.5%; positive predictive value: 38.2%; negative predictive value: 95.2%; and accuracy: 80.4%. Although the study was not designed to compare repeat biopsy strategies, more targeted cores than random systematic cores were found to be positive for cancer (36.3% compared with 4.9%, p<0.00001). CONCLUSIONS: In the challenging situation of repeat biopsies, the ESUR scoring system was shown to provide clinically relevant stratification of the risk of showing PCa in a given location.

Prospective comparison of T2w-MRI and dynamic-contrast-enhanced MRI, 3D-MR spectroscopic imaging or diffusion-weighted MRI in repeat TRUS-guided biopsies.

OBJECTIVES: To compare T2-weighted MRI and functional MRI techniques in guiding repeat prostate biopsies. METHODS: Sixty-eight patients with a history of negative biopsies, negative digital rectal examination and elevated PSA were imaged before repeat biopsies. Dichotomous criteria were used with visual validation of T2-weighted MRI, dynamic contrast-enhanced MRI and literature-derived cut-offs for 3D-spectroscopy MRI (choline-creatine-to-citrate ratio >0.86) and diffusion-weighted imaging (ADC × 10(3) mm(2)/s < 1.24). For each segment and MRI technique, results were rendered as being suspicious/non-suspicious for malignancy. Sextant biopsies, transition zone biopsies and at least two additional biopsies of suspicious areas were taken. RESULTS: In the peripheral zones, 105/408 segments and in the transition zones 19/136 segments were suspicious according to at least one MRI technique. A total of 28/68 (41.2%) patients were found to have cancer. Diffusion-weighted imaging exhibited the highest positive predictive value (0.52) compared with T2-weighted MRI (0.29), dynamic contrast-enhanced MRI (0.33) and 3D-spectroscopy MRI (0.25). Logistic regression showed the probability of cancer in a segment increasing 12-fold when T2-weighted and diffusion-weighted imaging MRI were both suspicious (63.4%) compared with both being non-suspicious (5.2%). CONCLUSION: The proposed system of analysis and reporting could prove clinically relevant in the decision whether to repeat targeted biopsies.

Role of parametric imaging in contrast-enhanced sonography of hepatic focal nodular hyperplasia.

PURPOSE: To assess the value of parametric imaging during contrast-enhanced sonographic examination in the diagnosis of focal nodular hyperplasia (FNH) of the liver. MATERIALS AND METHODS: Thirty-one patients with solitary FNH underwent contrast-enhanced sonographic examination between January 2003 and June 2004 using SonoVue and a Sequoia scanner equipped with Cadence Contrast Pulse Sequencing software. Contrast enhancement from a time sequence of perfusion frames was estimated using QontraXt software, which provides quantification of perfusion parameters. From the time-intensity curves, we obtained the following parameters: peak enhancement value, Tr (time recovery corresponding to time needed to reach 63% of the peak value), beta parameter corresponding to the exponential factor, and slope corresponding to the slope of the tangent to the arterial phase of enhancement. RESULTS: Among the 4 parameters studied, the slope of the arterial phase of enhancement was the most sensitive to image the centrifugal arterial flow originating from the central portion of the lesion, whereas peak enhancement value and Tr were the most sensitive to image full enhancement of the lesion. A blinded review revealed equivalent sensitivity in the diagnosis of FNH between the interpretation of the original videoclips and that of the parametric images. CONCLUSIONS: The results of this study show that parametric imaging can be used in place of original videoclips for clinical reporting of FNH; furthermore, it could help less-experienced sonologists diagnose FNH.