OBJECTIVES: To evaluate the diagnostic performance of preoperative multiparametric MRI with extracapsular extension (ECE) risk-scoring in the assessment of prostate cancer tumour stage (T-stage) and prediction of ECE at final pathology. MATERIALS AND METHODS: Eighty-seven patients with clinically localised prostate cancer scheduled for radical prostatectomy were prospectively enrolled. Multiparametric MRI was performed prior to prostatectomy, and evaluated according to the ESUR MR prostate guidelines by two different readers. An MRI clinical T-stage (cTMRI), an ECE risk score, and suspicion of ECE based on tumour characteristics and personal opinion were assigned. Histopathological prostatectomy results were standard reference. RESULTS: Histopathology and cTMRI showed a spearman rho correlation of 0.658 (p < 0.001) and a weighted kappa = 0.585 [CI 0.44;0.73](reader A). ECE was present in 31/87 (36 %) patients. ECE risk-scoring showed an AUC of 0.65-0.86 on ROC-curve for both readers, with sensitivity and specificity of 81 % and 78 % at best cutoff level (reader A), respectively. When tumour characteristics were influenced by personal opinion, the sensitivity and specificity for prediction of ECE changed to 61 %-74 % and 77 %-88 % for the readers, respectively. CONCLUSIONS: Multiparametric MRI with ECE risk-scoring is an accurate diagnostic technique in determining prostate cancer clinical tumour stage and ECE at final pathology. KEY POINTS: • Multiparametric MRI is an accurate diagnostic technique for preoperative prostate cancer staging • ECE risk scoring predicts extracapsular tumour extension at final pathology • ECE risk scoring shows an AUC of 0.86 on the ROC-curve • ECE risk scoring shows a moderate inter-reader agreement (K = 0.45) • Multiparametric MRI provides essential knowledge for optimal clinical management.
European Radiology, 2015, Vol 25, Issue 6, p. 1776-85