Premium
Prognostic value of Prostate Imaging and Data Reporting System (PI‐RADS) v. 2 assessment categories 4 and 5 compared to histopathological outcomes after radical prostatectomy
Author(s) -
Lim Christopher S.,
McInnes Matthew D.F.,
Lim Robert S.,
Breau Rodney H.,
Flood Trevor A.,
Krishna Satheesh,
Morash Christopher,
Shabana Wael M.,
Schieda Nicola
Publication year - 2017
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.25539
Subject(s) - medicine , prostatectomy , receiver operating characteristic , confidence interval , magnetic resonance imaging , prostate cancer , prostate , stage (stratigraphy) , nuclear medicine , urology , radiology , cancer , biology , paleontology
Purpose To assess Prostate Imaging and Data Reporting System (PI‐RADS) v. 2 score 4/5 lesions compared to Gleason score (GS) and stage after radical prostatectomy (RP) and to validate the proposed 15‐mm size threshold that differentiates category 4 versus 5 lesions. Materials and Methods With Institutional Review Board (IRB) approval, 140 men underwent 3T magnetic resonance imaging (MRI) and RP between 2012–2015. Two blinded radiologists: 1) assigned PI‐RADS v. 2 scores, 2) measured tumor size on axial T 2 ‐weighted‐MRI, and 3) assessed for extraprostatic extension (EPE). Interobserver agreement was calculated and consensus diagnoses achieved through reference standard (MRI‐RP maps). PI‐RADS v. 2 scores and tumor size were compared to GS and stage using chi‐square, analysis of variance (ANOVA), and receiver operating characteristic (ROC) curve analysis. Results In all, 80.7% (113/140) of tumors were category 4 ( n = 45) or 5 ( n = 68) lesions (κ = 0.45). Overall tumor size was 18.2 ± 7.7 mm and category 5 lesions were larger (22.6 ± 6.8 versus 11.5 ± 1.9 mm, P < 0.001). High‐risk (GS ≥8) tumors were larger than low‐ and intermediate‐risk tumors ( P = 0.016) and were more frequently, but not significantly so, category 5 lesions (78.9% [15/19] vs. 22.1% [4/10], P = 0.18). 67.3% (76/113) of patients had EPE. Category 5 lesions were strongly associated with EPE ( P < 0.0001). Area under the ROC curve for diagnosis of EPE by size was 0.74 (confidence interval 0.64–0.83), with size ≥15 mm yielding a sensitivity/specificity of 72.4/64.9%. Size improved sensitivity for diagnosis of EPE compared to subjective assessment (sensitivity/specificity ranging from 46.1–48.7%/70.3–86.5%, κ = 0.29) ( P = 0.028). Conclusion PI‐RADS v. 2 category 5 lesions are associated with higher Gleason scores and EPE. A 15‐mm size threshold is reasonably accurate for diagnosis of EPE with increased sensitivity compared to subjective assessment. Level of Evidence : 3 Technical Efficacy: Stage 2 J. MAGN. RESON. IMAGING 2017;46:257–266