Abstract
Objectives
We sought to identify clinical and MRI characteristics in men with PI-RADS category 3 index lesions that predict clinically significant prostate cancer (PCa) on MRI targeted biopsy.
Materials and Methods
Multi-center study of prospectively collected data for biopsy-naive men (n=247) who underwent MRI-targeted and systematic biopsies for PI-RADS 3 index lesions. The primary endpoint was diagnosis of clinically significant PCa (Grade Group ≥2). Multivariable logistic regression models assessed for factors associated with clinically significant PCa. The probability distributions of clinically significant PCa based on different levels of predictors of multivariable models were plotted in a heatmap.
Results
Men with clinically significant PCa had smaller prostate volume (39.20 vs 55.10 mL, p<0.001) and lower apparent diffusion coefficient (ADC) values (973 vs 1068 μm2/s, p=0.013), but higher prostate-specific antigen (PSA) density (0.21 vs 0.13 ng/mL2, p=0.027). On multivariable analyses, lower prostate volume (OR 0.95, 95%CI 0.92–0.97), lower ADC value (OR 0.99, 95%CI 0.99–1.00), and PSA density >0.15 ng/ml2 (OR 3.51, 95%CI 1.61–7.68) were independently associated with significant PCa.
Conclusion
Higher PSA density, lower prostate volume and ADC values are associated with clinically significant PCa in biopsy-naïve men with PI-RADS 3 lesions. We present regression-derived probabilities of detecting clinically significant PCa based on various clinical and imaging values that can be used in decision-making. Our findings demonstrate an opportunity for MRI refinement or biomarker discovery to improve risk stratification for PI-RADS 3 lesions.
Keywords: Prostate cancer, Magnetic Resonance Imaging
1. Introduction
Prostate multiparametric magnetic resonance imaging (MRI) is increasingly used to risk stratify in deciding whether a man should undergo prostate biopsy.[1] In an effort to optimize interpretation and reporting of multiparametric MRI, the Prostate Imaging - Reporting and Data System (PI-RADS) was developed as an international collaboration, with the objective to improve detection of clinically significant prostate cancer (Gleason group ≥2).[2,3] PI-RADS is a five-point scale designed to convey the perceived risk of clinically significant prostate cancer based on imaging features. Lesions felt to be equivocal for cancer are scored as PI-RADS 3. PI-RADS category 3 lesions are found in approximately 15%−29% of men undergoing multiparametric MRI.[4–7] Despite the increase in diagnostic accuracy of PI-RADS v2 compared to the earlier version, there was limited improvement in risk stratification of PI-RADS 3 lesions.[8,9] This leads to a clinical dilemma. While current guidelines recommend biopsy for men with PI-RADS 3 lesions,[10] the majority of these lesions (>70%) do not contain clinically significant cancer.[4,5,7,11] Therefore, PIRADS 3 lesions present a challenge to treating urologists who must decide whether to monitor with follow-up prostate-specific antigen (PSA) testing and imaging, or schedule immediate biopsy. Moreover, the increasing risk of biopsy-related infections warrant judicious patient selection for prostate biopsy.[12]
Studies that risk stratify men with PI-RADS category 3 lesions have either a relatively small sample size (< 150 men),[13,14] included PI-RADS v1,[4,15] or used 1.5 Tesla MRI magnets instead of the recommended 3 Tesla magnets.[11] To address this important clinical dilemma, we sought to identify clinical and MRI characteristics associated with detecting clinically significant cancer in biopsy-naïve men with PI-RADS 3 lesions through a contemporary multi-center study.
2. Methods:
2.1. Study population
We examined prospectively collected data for men who underwent MRI-targeted biopsy at Weill Cornell Medicine, UCLA and Stanford during 2015–2018. Our study was approved by the Institutional Review Boards of the respective institutions. Generally, men with an initial elevated prostate-specific antigen (PSA) underwent a repeat PSA test, and if elevated by age-specific criteria,[16] an MRI was obtained. The risks of missing significant cancer were reviewed and most men opted to proceed with a biopsy. Men with PI-RADS ≥ category 3 lesions are offered MRI-targeted biopsy. Only those with a category 3 index lesion on multiparametric MRI using PI-RADS v2 classification were included in this study (n=686). Subjects who had a prior biopsy were excluded from the study resulting in a sample of 247 biopsy-naïve men. Data on clinical staging (digital rectal examination) and family history were inconsistently collected. Therefore, this information was excluded from the analysis. Of note, in those who underwent a digital rectal exam, less than 3% had suspicions findings (≥T1c).
2.2. Magnetic resonance imaging and fusion biopsy
All subjects underwent contrast enhanced multiparametric MRI performed using 3 Tesla scanners without an endorectal coil. Studies were performed with T1-weighted and T2-weighted imaging, dynamic contrast-enhanced imaging, and diffusion-weighted imaging. Protocols were similar in all institutions and were consistent with the recommendations of PI-RADS v2.[2] All scans were read by fellowship trained radiologists at each respective center.
Subjects were classified according to their highest PI-RADS category (index). Only 5 subjects had two PI-RADS category 3 lesions. The lesion with the greatest diameter (index) in those men was included in the analysis. Thus, all men had organ-confined category 3 lesions as the highest scored lesions. The mean apparent diffusion coefficient (ADC) was calculated by drawing the region of interest in the most suspicious axial slice on the ADC map similarly in all institutions.[13] Prostate biopsy was performed using a robotic MRI/US fusion platform (Artemis™, Eigen, Grass Valley, California) under local anesthesia in the clinic setting. Following targeted biopsies, a 12–14 core systematic (template) biopsy was performed using a template generated by the biopsy device.[17] Each institution has previously reported on their experience in MRI/US fusion prostate biopsies. [18–20]
2.3. Study outcomes and statistical analysis
The primary outcome of interest was clinically significant prostate cancer defined as Grade group 2 or higher detected in a systematic or targeted core biopsy. The secondary outcome was detection of indolent prostate cancer (Grade group 1). Clinical (age, institution, PSA level prior to biopsy, prostate volume, and PSA density) and MRI characteristics (lesion diameter, side, location [apex, midgland or base], region [peripheral or transition zones], and ADC value) were compared between men with significant prostate cancer and those without by independent two-sample t-tests or Chi-square/Fisher’s Exact tests.
Multivariable logistic regression analyses assessed for clinical (age, PSA and center) and MRI parameters (prostate volume and ADC value) independently associated with clinically significant prostate cancer. Given that PSA density is derived from PSA level and prostate volume, two separate multivariable models including either the combination of PSA and prostate volume or high PSA density (>0.15 ng/ml2) alone were analyzed.[21] Diagnostic accuracy measures such as sensitivity, specificity and negative predictive value (NPV) were derived from the two multivariable models using Youden’s index to obtain the optimal probability of clinically significant prostate cancer thresholds from the Receiver Operating Characteristic curves. Finally, the probability distributions of clinically significant prostate cancer based on different levels of predictors of the two multivariable models were plotted in a heatmap. Analyses were two-sided with statistical significance evaluated at the 0.05 alpha level. Analyses were performed in R version 3.5.1 (Vienna, Austria).
3. Results:
A total of 55/247 (22%) of biopsy-naïve men with PI-RADS category 3 lesions on multiparametric MRI were diagnosed with clinically significant prostate cancer: 37 (67%) had Grade group 2, seven (13%) had Grade group 3, seven (13%) had Grade group 4 and four (7%) had Grade group 5. Additionally, 61/247 (25%) subjects had insignificant cancer (Grade group 1) on biopsy. Men with clinically significant prostate cancer had smaller prostate volumes (39.20 vs 55.10 mL, p < 0.001), lower ADC values (973 vs 1068 μm2/s, p=0.013), and higher PSA density (0.21 vs. 0.13 ng/ml2, p= 0.027) compared to those without (Table 1). The majority of men, 26/55 (47%), diagnosed with clinically significant prostate cancer were detected on both targeted and systematic cores, whereas only 11/55 (20%) men were detected on target biopsy only (Table 2). Men with clinically significant prostate cancer detected on target biopsy only had higher PSA levels (6.42 vs 4.40 ng/mL, p=0.01) and larger prostate volumes (46.31 vs 38.49 mL, p <0.001) compared to those that were detected on systematic biopsy only.
Table 1:
Clinical and MRI characteristics of men with prostate cancer with PI-RADS 3.
| Indolent prostate cancer or benign prostate (N=192) |
Clinically significant prostate cancer (N=55) |
p value | |
|---|---|---|---|
| Clinical characteristics | |||
| Mean age (Std), years | 63.1 (7.50) | 65.1 (6.81) | 0.08 |
| Mean PSA (Std), ng/ml | 6.25 (3.33) | 6.94 (7.32) | 0.50 |
| Mean Prostate volume (Std), ml | 55.1 (24.5) | 39.2 (15.6) | <0.001 |
| Mean PSA density (Std), ng/ml2 | 0.13 (0.10) | 0.21 (0.25) | 0.03 |
| Center | 0.17 | ||
| WCM | 60 (31.2%) | 16 (29.1%) | |
| Stanford | 73 (38.0%) | 15 (27.3%) | |
| UCLA | 59 (30.7%) | 24 (43.6%) | |
| MRI characteristics | |||
| Location | |||
| Apex | 50 (26.0%) | 16 (29.1%) | 0.78 |
| Mid | 119 (62.0%) | 33 (60.0%) | 0.91 |
| Base | 44 (22.9%) | 13 (23.6%) | 1.00 |
| Side | |||
| Right | 70 (36.6%) | 25 (45.5%) | 0.44 |
| Left | 108 (56.5%) | 28 (50.9%) | |
| Midline | 13 (6.81%) | 2 (3.64%) | |
| Region | |||
| Peripheral | 122 (63.5%) | 36 (65.5%) | 0.91 |
| Transition | 70 (36.5%) | 19 (34.5%) | |
| Mean (STD) Diameter, mm | 11.1 (5.85) | 11.1 (4.88) | 0.99 |
| Mean (Std) ADC value, μm2/s | 1068 (202) | 973 (214) | 0.01 |
Std: Standard deviation.
Table 2:
Clinically significant prostate cancer in PI-RADS category 3 lesion detected on systematic, target or both.
| Detected on systematic biopsy only (N=18) |
Detected on target biopsy only (N=11) |
Detected on systematic and target biopsies (N=26) |
p value | |
|---|---|---|---|---|
| Clinical characteristics | ||||
| Mean age, years | 62.68 (6.73) | 65.09 (7.13) | 66.63 (6.51) | 0.15 |
| Mean PSA (Std), ng/ml | 4.4 (2.36) | 6.42 (2.02) | 8.83 (10.00) | 0.01 |
| Mean Prostate volume (Std), ml | 38.49 (15.02) | 46.31 (20.87) | 36.74 (12.93) | <0.001 |
| Mean PSA density (Std), ng/ml2 | 0.12 (0.09) | 0.17 (0.10) | 0.28 (0.34) | <0.001 |
| Center, n (%) | ||||
| Cornell | 7 (38.9%) | 3 (27.3%) | 6 (23.1%) | 0.26 |
| Stanford | 2 (11.1%) | 3 (27.3%) | 10 (38.5%) | |
| UCLA | 9 (50.0%) | 5 (45.5%) | 10 (38.5%) | |
| MRI characteristics | ||||
| Location, n (%) | ||||
| Apex | 4 (22.2%) | 6 (54.6%) | 6 (23.1%) | 0.23 |
| Midgland | 9 (50.0%) | 6 (54.6%) | 18 (69.2%) | 0.61 |
| Base | 7 (38.9%) | 3 (27.3%) | 3 (11.5%) | 0.19 |
| Side, n (%) | ||||
| Right | 9 (50.0%) | 2 (18.2%) | 14 (53.9%) | 0.34 |
| Left | 9 (50.0%) | 8 (72.7%) | 11 (42.3%) | |
| Midline | - | 1 (9.1%) | 1 (3.9%) | |
| Region, n (%) | ||||
| Peripheral | 13 (72.2%) | 6 (54.6%) | 17 (65.4%) | 0.79 |
| Transition | 5 (27.8%) | 5 (45.5%) | 9 (34.6%) | |
| Mean (Std) Diameter, mm | 10.18(2.74) | 10.09(3.88) | 12.12(6.12) | 0.67 |
| Mean (Std) ADC value μm2/s) | 964.64(327.39) | 980.5(164.71) | 974.31(123.83) | 0.08 |
| Pathology | ||||
| Grade group, n (%) | ||||
| 2 | 13 (72.2) | 10 (90.9) | 14 (53.8) | |
| 3 | 3 (16.7) | - | 4 (15.4) | |
| 4 | 1 (5.6) | 1 (9.1) | 5 (19.2) | |
| 5 | 1 (5.6) | - | 3 (11.5) | |
Std: Standard deviation; Percentages may not add to 100% due to rounding.
Multivariable logistic regression, including PSA and prostate volume, revealed smaller prostate volume (odds ratio [OR] 0.95, 95% CI 0.92–0.97) and lower ADC value (OR 0.99, 95% CI 0.99–1.00) were independently associated with clinically significant prostate cancer (Table 3, Model 1). Multivariable regression, including PSA density, revealed high PSA density > 0.15 ng/ml2 (OR 3.51, 95% CI 1.61–7.68) was also independently associated with clinically significant prostate cancer (Table 3, Model 2). The sensitivity, specificity and NPV of the first model were 79%, 80% and 91% and that for the second model were 60%, 80% and 85%. Figure A illustrates the probability of clinically significant prostate cancer (based on the multivariable models) over various levels of the predictors (age, PSA volume, PSA and ADC) while Figure B does the same for the second model including PSA density. Older men with smaller prostate volume and lower ADC are more likely to have clinically significant prostate cancer (Figure A) and older men with smaller ADC values and elevated PSA density are more likely to have clinically significant prostate cancer (Figure B).
Table 3:
Multivariable logistic regression model predicting clinically significant prostate cancer in men with PI-RADS 3 lesions. Model 1 includes PSA level and prostate volume and model 2 includes PSA density.
| Model 1 | Model 2 | |||
|---|---|---|---|---|
| OR (95% CI) | p value | OR (95% CI) | p value | |
| Age (per year) | 1.05 (0.99–1.11) | 0.079 | 1.02 (0.97–1.08) | 0.402 |
| PSA (per 1 ng/ml) | 1.02 (0.95–1.10) | 0.614 | - | - |
| Prostate volume (per ml) | 0.95 (0.92–0.97) | < 0.0001 | - | - |
| PSA density (> 0.15 ng/ml2) | - | - | 3.51 (1.61–7.68) | 0.002 |
| ADC value (per 1 μm2/s) | 0.99 (0.99–1.00) | 0.029 | 0.99 (0.99 – 1.00) | 0.007 |
| Center | ||||
| Cornell | Ref | Ref | Ref | Ref |
| Stanford | 0.63 (0.19–2.07) | 0.443 | 0.67 (0.21–2.11) | 0.494 |
| UCLA | 1.21 (0.45–3.28) | 0.705 | 1.18 (0.45–3.08) | 0.738 |
Figure:
Probability of clinically significant prostate cancer based on various levels (intervals) of the predictors: age, apparent diffusion coefficient (ADC), prostate volume and prostate-specific antigen (PSA) (panel A) and age, ADC, PSA density (panel B). The white color represents the optimal threshold for clinically significant prostate cancer probability obtained using Youden’s index and shades of red indicate the distribution of clinically significant prostate cancer probability among those predicted to be diagnosed with clinically significant prostate cancer while shades of purple represent the same for those predicted to not have clinically significant prostate cancer.
4. Discussion:
In this multicenter study, one in five biopsy-naïve men with PI-RADS v2 category 3 index lesion harbored clinically significant prostate cancer, in line with the 15%−31% reported in the literature.[4,5,7,22] We found low prostate volume, high PSA density and low ADC values to be associated with clinically significant prostate cancer in men with PI-RADS category 3 index lesions. Moreover, we used multivariable models to create a heatmap that may be used as a scoring algorithm to aid in deciding whether to biopsy men with PI-RADS 3 lesions.
Several studies attempt to define characteristics associated with clinically significant prostate cancer in equivocal lesions. Sheridan et al. did not find PSA density to be associated with significant disease in equivocal lesions on multiparametric MRI. The authors attributed their findings to the small sample size (111 lesions).[23] However, a recent meta-analysis suggested that PSA density ≥0.15 ng/ml2 may represent an index to decide whether to biopsy or not PI-RADS 3 lesions.[24] Others have reported that older men and men with lower prostate volume have an increased likelihood of detecting clinically significant prostate cancer.[11,23] Felker et al. and Hermie et al. found lower ADC values (800–1000 μm2/s) to be associated with clinically significant disease.[13,14] Our findings confirm that detecting clinically significant disease is more likely in men with lower prostate volume, higher PSA density (> 0.15 ng/ml2) and lesions with low ADC values.
Our results demonstrate several important aspects when encountering men with equivocal lesions. Clinicians are encouraged to measure and record ADC values. There is variation in institutional computation and assessment of ADC values, which are protocol-dependent, and a feedback loop should be established to determine significant thresholds for detection of clinically significant prostate cancer.[25] Of note, there was no significant variation in ADC values across the centers in this study. Contrary to the findings of others, lesion diameter was not associated with significant cancer in PI-RADS 3 lesions in our series.[26] Moreover, men must be counseled that PI-RADS category 3 lesions may harbor unfavorable intermediate or high-risk prostate cancer: 7% of men in our series had Grade group ≥3. Lastly, one third of men with clinically significant prostate cancer were only detected on systematic cores, of which 30% were high grade (Gleason score ≥3), underscoring the importance of obtaining a systematic biopsy.
Based on our results and the work of others,[24] we now base our recommendations whether to biopsy men with PI-RADS 3 lesions primarily on PSA density. Certainly, proceeding with a prostate biopsy is a shared physician-patient decision, and the recommendations provided should be individualized based on clinical information. As such, deferring a biopsy in a 81 year-old man with low PSA density may not have the same implications as in a 51 year-old. Similarly, men with prior negative biopsies may a have a lower need to proceed with another biopsy compared to biopsy naïve men with a family history of prostate cancer.
Our findings must be interpreted in the context of the study design. The inherent interobserver variability that is associated with PI-RADS scoring and limitations of MRI to demonstrate clinically significant prostate cancer must be noted.[27,28] All MRI scans and prostate biopsies were performed in academic centers, limiting the generalizability of our findings to settings with limited prostate MRI expertise. However, the scans in each center were read by more than one radiologist per the PI-RADS assessment replicating clinical practice, which allowed us to evaluate standard clinical interpretation. Additionally, other factors that may significantly improve risk stratification for PI-RADS 3 lesions such as family history and prostate exam findings were not captured consistently across sites. Moreover, we only used available MRI data such as lesion diameter and ADC value. It is possible that artificial intelligence algorithms applied to prostate MR images may improve risk stratification in men with equivocal lesions.[29] Further prospective investigations are needed to explore the association of lesion characteristics on MRI and clinically significant prostate cancer, as there is opportunity in future PI-RADS updates to improve the performance of multiparametric MRI and in the assessment and reporting of findings in equivocal lesions.[30]
5. Conclusion
PSA density, prostate volume and ADC values are associated with clinically significant prostate cancer in men with PIRADS 3 lesions. We provide a visual depiction using a heatmap of how the risk of detecting clinically significant prostate cancer changes at different levels of predictors. This heatmap can be factored into decision making about the need for biopsy when counseling patients with equivocal lesions. Nonetheless, for cutoffs to be implemented into decision-making, further studies including different cohorts of men are needed. Our findings demonstrate an opportunity for biomarker discovery or PI-RADS refinement to improve clinically significant prostate cancer stratification for PI-RADS category 3 lesions in biopsy-naïve men.
One fifth of biopsy naïve men harbor clinically significant prostate cancer
PSA density, prostate volume and ADC values are associated with significant cancer
Both systematic and target biopsies should be obtained in men with PIRADS 3 lesions
Acknowledgments
Funding: The Frederick J. and Theresa Dow Wallace Fund of the New York Community Trust
6. References:
- [1].Bjurlin MA, Carroll PR, Eggener S, Fulgham PF, Margolis DJ, Pinto PA, et al. Update of the AUA policy statement on the use of multiparametric magnetic resonance imaging in the diagnosis, staging and management of prostate cancer. J Urol 2019:101097JU0000000000000617. doi: 10.1097/JU.0000000000000617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Weinreb JC, Barentsz JO, Choyke PL, Cornud F, Haider MA, Macura KJ, et al. PI-RADS Prostate Imaging - Reporting and Data System: 2015, Version 2. Eur Urol 2016;69:16–40. doi: 10.1016/j.eururo.2015.08.052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Epstein JI, Egevad L, Amin MB, Delahunt B, Srigley JR, Humphrey PA, et al. The 2014 international society of urological pathology (ISUP) consensus conference on gleason grading of prostatic carcinoma: definition of grading patterns and proposal for a new grading system. Am J Surg Pathol 2016;40:244–52. doi: 10.1097/PAS.0000000000000530. [DOI] [PubMed] [Google Scholar]
- [4].Borkowetz A, Hadaschik B, Platzek I, Toma M, Torsev G, Renner T, et al. Prospective comparison of transperineal magnetic resonance imaging/ultrasonography fusion biopsy and transrectal systematic biopsy in biopsy-naïve patients. BJU Int 2018;121:53–60. doi: 10.1111/bju.14017. [DOI] [PubMed] [Google Scholar]
- [5].Hansen NL, 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;122:40–9. doi: 10.1111/bju.14049. [DOI] [PubMed] [Google Scholar]
- [6].Kasivisvanathan V, Rannikko AS, Borghi M, Panebianco V, Mynderse LA, Vaarala MH, et al. MRI-Targeted or Standard Biopsy for Prostate-Cancer Diagnosis. N Engl J Med 2018;378:1767–77. doi: 10.1056/NEJMoa1801993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Pokorny MR, de Rooij M, Duncan E, Schröder FH, Parkinson R, Barentsz JO, et al. Prospective study of diagnostic accuracy comparing prostate cancer detection by transrectal ultrasound-guided biopsy versus magnetic resonance (MR) imaging with subsequent MR-guided biopsy in men without previous prostate biopsies. Eur Urol 2014;66:22–9. doi: 10.1016/j.eururo.2014.03.002. [DOI] [PubMed] [Google Scholar]
- [8].Kasel-Seibert M, Lehmann T, Aschenbach R, Guettler FV, Abubrig M, Grimm M-O, et al. Assessment of PI-RADS v2 for the Detection of Prostate Cancer. Eur J Radiol 2016;85:726–31. doi: 10.1016/j.ejrad.2016.01.011. [DOI] [PubMed] [Google Scholar]
- [9].Rosenkrantz AB, Oto A, Turkbey B, Westphalen AC. Prostate Imaging Reporting and Data System (PI-RADS), Version 2: A Critical Look. AJR Am J Roentgenol 2016;206:1179–83. doi: 10.2214/AJR.15.15765. [DOI] [PubMed] [Google Scholar]
- [10].EAU Guidelines: Prostate Cancer ∣ Uroweb n.d. https://uroweb.org/guideline/prostate-cancer/?type=summary-of-changes (accessed June 21, 2019).
- [11].Dell’Oglio P, Stabile A, Mazzone E, Soligo M, Brunetti L, Damascelli A, et al. Which patients with PI-RADs 3 lesion at multiparametric MRI can avoid prostate biopsy. European Urology Supplements 2018;17:e701. doi: 10.1016/S1569-9056(18)31322-8. [DOI] [Google Scholar]
- [12].Loeb S, Vellekoop A, Ahmed HU, Catto J, Emberton M, Nam R, et al. Systematic review of complications of prostate biopsy. Eur Urol 2013;64:876–92. doi: 10.1016/j.eururo.2013.05.049. [DOI] [PubMed] [Google Scholar]
- [13].Felker ER, Raman SS, Margolis DJ, Lu DSK, Shaheen N, Natarajan S, et al. Risk Stratification Among Men With Prostate Imaging Reporting and Data System version 2 Category 3 Transition Zone Lesions: Is Biopsy Always Necessary? AJR Am J Roentgenol 2017;209:1272–7. doi: 10.2214/AJR.17.18008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Hermie I, Van Besien J, De Visschere P, Lumen N, Decaestecker K. Which clinical and radiological characteristics can predict clinically significant prostate cancer in PI-RADS 3 lesions? A retrospective study in a high-volume academic center. Eur J Radiol 2019;114:92–8. doi: 10.1016/j.ejrad.2019.02.031. [DOI] [PubMed] [Google Scholar]
- [15].Hansen NL, Koo BC, Warren AY, Kastner C, Barrett T. Sub-differentiating equivocal PI-RADS-3 lesions in multiparametric magnetic resonance imaging of the prostate to improve cancer detection. Eur J Radiol 2017;95:307–13. doi: 10.1016/j.ejrad.2017.08.017. [DOI] [PubMed] [Google Scholar]
- [16].Vickers AJ, Eastham JA, Scardino PT, Lilja H. The memorial sloan kettering cancer center recommendations for prostate cancer screening. Urology 2016;91:12–8. doi: 10.1016/j.urology.2015.12.054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Sonn GA, Natarajan S, Margolis DJA, MacAiran M, Lieu P, Huang J, et al. Targeted biopsy in the detection of prostate cancer using an office based magnetic resonance ultrasound fusion device. J Urol 2013;189:86–91. doi: 10.1016/j.juro.2012.08.095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Filson CP, Natarajan S, Margolis DJA, Huang J, Lieu P, Dorey FJ, et al. Prostate cancer detection with magnetic resonance-ultrasound fusion biopsy: The role of systematic and targeted biopsies. Cancer 2016;122:884–92. doi: 10.1002/cncr.29874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Kasabwala K, Patel N, Cricco-Lizza E, Shimpi AA, Weng S, Buchmann RM, et al. The Learning Curve for Magnetic Resonance Imaging/Ultrasound Fusion-guided Prostate Biopsy. European Urology Oncology 2019;2:135–40. doi: 10.1016/j.euo.2018.07.005. [DOI] [PubMed] [Google Scholar]
- [20].Sonn GA, Fan RE, Ghanouni P, Wang NN, Brooks JD, Loening AM, et al. Prostate magnetic resonance imaging interpretation varies substantially across radiologists. Eur Urol Focus 2017. doi: 10.1016/j.euf.2017.11.010. [DOI] [PubMed] [Google Scholar]
- [21].Epstein JI, Walsh PC, Carmichael M, Brendler CB. Pathologic and clinical findings to predict tumor extent of nonpalpable (stage T1c) prostate cancer. JAMA 1994;271:368–74. [PubMed] [Google Scholar]
- [22].Venderink W, van Luijtelaar A, Bomers JGR, van der Leest M, Hulsbergen-van de Kaa C, Barentsz JO, et al. Results of Targeted Biopsy in Men with Magnetic Resonance Imaging Lesions Classified Equivocal, Likely or Highly Likely to Be Clinically Significant Prostate Cancer. Eur Urol 2017. doi: 10.1016/j.eururo.2017.02.021. [DOI] [PubMed] [Google Scholar]
- [23].Sheridan AD, Nath SK, Syed JS, Aneja S, Sprenkle PC, Weinreb JC, et al. Risk of clinically significant prostate cancer associated with prostate imaging reporting and data system category 3 (equivocal) lesions identified on multiparametric prostate MRI. AJR Am J Roentgenol 2018;210:347–57. doi: 10.2214/AJR.17.18516. [DOI] [PubMed] [Google Scholar]
- [24].Maggi M, Panebianco V, Mosca A, Salciccia S, Gentilucci A, Di Pierro G, et al. Prostate Imaging Reporting and Data System 3 Category Cases at Multiparametric Magnetic Resonance for Prostate Cancer: A Systematic Review and Meta-analysis. Eur Urol Focus 2019. doi: 10.1016/j.euf.2019.06.014. [DOI] [PubMed] [Google Scholar]
- [25].deSouza NM, Winfield JM, Waterton JC, Weller A, Papoutsaki MV, Doran SJ, et al. Implementing diffusion-weighted MRI for body imaging in prospective multicentre trials: current considerations and future perspectives. Eur Radiol 2018;28:1118–31. doi: 10.1007/s00330-017-4972-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [26].Scialpi M, Martorana E, Aisa MC, Rondoni V, D’Andrea A, Bianchi G. Score 3 prostate lesions: a gray zone for PI-RADS v2. Turk J Urol 2017;43:237–40. doi: 10.5152/tud.2017.01058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Muller BG, Shih JH, Sankineni S, Marko J, Rais-Bahrami S, George AK, et al. Prostate Cancer: Interobserver Agreement and Accuracy with the Revised Prostate Imaging Reporting and Data System at Multiparametric MR Imaging. Radiology 2015;277:741–50. doi: 10.1148/radiol.2015142818. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Johnson DC, Raman SS, Mirak SA, Kwan L, Bajgiran AM, Hsu W, et al. Detection of individual prostate cancer foci via multiparametric magnetic resonance imaging. Eur Urol 2019;75:712–20. doi: 10.1016/j.eururo.2018.11.031. [DOI] [PubMed] [Google Scholar]
- [29].Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL. Artificial intelligence in radiology. Nat Rev Cancer 2018;18:500–10. doi: 10.1038/s41568-018-0016-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Rosenkrantz AB, Babb JS, Taneja SS, Ream JM. Proposed Adjustments to PI-RADS Version 2 Decision Rules: Impact on Prostate Cancer Detection. Radiology 2017;283:119–29. doi: 10.1148/radiol.2016161124. [DOI] [PubMed] [Google Scholar]