Diffusion-weighted MR imaging for the evaluation of seminal vesicle invasion in prostate cancer: initial results

PURPOSE: To evaluate the accuracy of T2-weighted (T2WI) with diffusion-weighted imaging (DWI) as compared with T2WI alone for predicting seminal vesicle invasion (SVI) of prostate cancer. MATERIALS AND METHODS: A total of 30 patients with SVI and 136 patients without SVI who underwent prostate MR imaging were included in the study. MR images were analyzed retrospectively and independently by two readers for SVI on T2WI and T2WI with DWI using a 5-point scale. RESULTS: For predicting SVI, the specificity for T2WI with DWI and for T2WI was 97% and 87%, respectively, and accuracy was 96% and 87%, respectively, as determined by the experienced reader (P < 0.01). For the less experienced reader, the specificity for T2WI with DWI and for T2WI was 96% and 81%, respectively, and accuracy was 90% and 77%, respectively (P < 0.01). The diagnostic performance of the less experienced reader with T2WI with DWI led to significant improvement of the area under the receiver operating characteristic curve (Az) as compared with T2WI alone (Az = 0.815 versus 0.696; P < 0.01). Interreader agreement showed a substantial agreement (kappa = 0.613) for T2WI, and a substantial agreement (kappa = 0.737) for T2WI with DWI. CONCLUSION: For predicting SVI, T2WI with DWI showed a better diagnostic performance than T2WI alone. Additionally, the accuracy of the less experienced reader using T2WI with DWI showed a significant improvement as compared with T2WI alone.