RESUMEN
BACKGROUND: The incremental value of dynamic contrast-enhanced (DCE) imaging in localizing radiorecurrent prostate cancer is uncertain. PURPOSE: To assess the added-value of DCE imaging to the combination T2 -weighted imaging (T2 W)+diffusion-weighted imaging (DWI) in detecting locally radiorecurrent prostate cancer (PCa), by radiologists with different levels of experience. STUDY TYPE: Analytic retrospective study. POPULATION: In all, 52 men with biological suspected PCa recurrence after radiotherapy were retrospectively included. FIELD STRENGTH/SEQUENCE: All men underwent prostatic MRI (1.5T or 3T), including T2 W, DWI, and DCE imagings, before biopsies. ASSESSMENT: Two junior (6 months' experience) and two senior readers (more than 3 years' experience) independently assigned a Likert score for each prostatic sextant on T2 W+DW+DCE imagings, then on T2 W+DW imagings, 4 weeks later. STATISTICAL TESTS: The reference standard was prostatic biopsies. For two levels of positivity of Likert score, 3/5 and 4/5, sensitivity, specificity, area under the receiver operating curve (AUC), and interreader agreement were compared. RESULTS: T2 W+DWI+DCE and T2 W+DWI imaging had similar AUC at lobe and sextant level (0.853-0.946 vs. 0.819-0.955, P from 0.071-0.534). Using a Likert score ≥4/5, T2 W+DWI+DCE significantly improved the sensitivity for junior readers at the patient, lobe, and sextant level (40-80% vs. 22-66%, P < 0.0001-0.041). Sensitivity was not significantly modified with DCE imaging for senior readers (54-95% vs. 50-91%, P from 0.074-1). Specificity was not modified for all readers (50-100% vs. 50%-100%, P from 0.134-1). DCE imaging improved interreader agreement for a Likert score ≥4/5 (kappa from 0.6-0.73 vs. 0.38-0.73). DATA CONCLUSION: The addition of DCE imaging did not significantly improve accuracy in recurrent PCa detection after radiotherapy, whatever the level of experience of the readers. However, the addition of DCE imaging slightly improved the sensitivity for less-experienced readers and increased their diagnostic confidence. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;48:1012-1023.