Authors:
R Johnston, L-M Wong, A Warren, N Shah, D Neal
Journal name: 
ANZ J Surg
Citation info: 
83(4):234-238
Abstract: 
BACKGROUND: To review the role of 1.5T (standard) magnetic resonance imaging (MRI) in predicting pathological T stage of prostate cancer in patients undergoing radical prostatectomy. METHODS: All patients undergoing radical prostatectomy between 2005 and 2010, who had a preoperative MRI, were included in the study. All MRI examinations were performed with a 1.5-T magnet without an endorectal coil. All the MRIs and prostate specimen histology slides were reviewed by genitourinary radiologists and pathologists at a centralized multidisciplinary team meeting. We calculated the sensitivity and specificity according to D'Amico risk classification. Likelihood ratios were used to test the discriminative ability of MRI. RESULTS: A total of 568 patients were identified as eligible. Median age was 62 years (35-74). Average prostate-specific antigen was 8.70 ng/mL (0.5-63). Risk categories included low, intermediate and high risk and contained 198 (34.9%), 303 (53.3%) and 67 (11.8%) patients, respectively. Comparisons between MRI findings and final histology gave a sensitivity of 20.0% and specificity of 80.2%. The positive likelihood ratio was 1.25. Although there was a trend (sr rho = 0.79) towards improved sensitivity as the clinical stage increased, this did not reach statistical significance (P = 0.68). CONCLUSION: Standard MRI does not improve preoperative local staging. Therefore, we recommend that standard MRI has no role in the local staging of prostate cancer. The use of higher field strength magnets (e.g. 3.0 T) and/or endorectal coil or the addition of other techniques such as dynamic contrast-enhanced MRI, diffusion-weighted MRI and MR spectroscopy imaging should therefore be preferentially used in routine clinical practice.
DOI: 
http://doi.org/10.1111/ans.12094
Research group: 
Neal Group
E-pub date: 
01 Apr 2013