OBJECTIVE: The purpose of this study was to prospectively evaluate the value of diffusion-weighted MRI (DWI) for the assessment of extracapsular extension (ECE) in patients with prostate cancer.
SUBJECTS AND METHODS: Between November 2010 and April 2012, 40 patients with intermediate- to high-risk prostate cancer were prospectively recruited. MR images were obtained at 3 T with a phased-array coil. Two independent readers scored the T2-weighted images alone and then in combination with DW images. ROIs were drawn on the apparent diffusion coefficient (ADC) maps, and histogram-derived values were calculated. Whole-mount histopathologic examination was the standard of reference. Reader performance was analyzed, and differences in patient characteristics and histogram-based ADC values, according to ECE status, were evaluated.
RESULTS: ECE was present in 23 of 40 (58%) patients and 23 of 43 (53%) tumors. The sensitivity for side-specific ECE detection significantly increased, from 0.22 to 0.44 for reader 1 and 0.33 to 0.82 for reader 2 (both p < 0.05) without a significant change in specificity for either reader with the addition of DWI and ADC mapping. The positive and negative predictive values for both readers also increased. The ADC parameters of median and 10th and 25th centiles showed a statistically significant difference between tumors with and those without ECE (p < 0.05).
CONCLUSION: The addition of DWI and ADC mapping to T2-weighted MRI improved the accuracy of preoperative detection of ECE. Median and 10th and 25th centile ADC values were significantly associated with the presence of ECE and may be useful in the pretreatment assessment of patients with prostate cancer.
Written by:
Lawrence EM, Gallagher FA, Barrett T, Warren AY, Priest AN, Goldman D, Sala E, Gnanapragasam VJ. Are you the author?
Department of Radiology, University of Cambridge, Addenbrooke's Hospital, Box 218, Level 5, Hills Rd, Cambridge CB2 0QQ UK.
Reference: AJR Am J Roentgenol. 2014 Sep;203(3):W280-6.
doi: 10.2214/AJR.13.11754
PubMed Abstract
PMID: 25148185