Correlation of multi-parametric MRI of peripheral prostate cancer with clinically significant prostate cancer and Gleason score / 中国医学影像技术

ABSTRACT

Objective: To explore the correlation of multi-parametric MRI (mp-MRI) of peripheral prostate cancer based on prostate imaging reporting and data system version 2 (PI-RADS v2) scoring with clinically significant prostate cancer and Gleason score. Methods: Clinical and MRI data, including T2WI, diffusion-weighted imaging (DWI) and dynamic contrast-enhanced MRI (DCE-MRI) of 91 cases of peripheral prostate cancer (positive group), 70 of benign prostatic hyperplasia (BPH) and 21 cases of prostatitis (negative group) confirmed with prostate biopsy pathology were retrospectively analyzed. PI-RADS v2 scoring was performed blindly, then the correlations with clinically significant prostate cancer and Gleason score were analyzed. The sensitivity, specificity and accuracy of PI-RADS v2 scoring were analyzed using receiver operating characteristic (ROC) curve. Results: The area under the curve (AUC) for the detection of clinically significant prostate cancer of PI-RADS v2 scoring was 0.965 (95%CI [0.930, 1.000], P<0.01). Taken PI-RADS v2 scoring 4 as the cutoff value, Youden index was maximum (0.877),the sensitivity, specificity and accuracy of diagnosis of clinically significant prostate cancer was 98.61%, 89.09% and 92.86%, respectively. Taken PI-RADS v2 scoring 3 as the cut-off point, unnecessary prostate biopsy could be avoided in 36.26% (66/182) patients, but 2 patients with non-significant prostate cancer would be misdiagnosed. PI-RADS v2 scoring was highly positively correlated with clinically significant prostate cancer (r=0.853, P<0.01) and Gleason score (r=0.816, P<0.01). Conclusion: PI-RADS v2 scoring was positively correlated with clinically significant peripheral prostate cancer and Gleason score, which had high diagnostic accuracy in clinically significant prostate cancer and Preliminarily evaluation on tumor differentiation, therefore being helpful to avoiding unnecessary biopsy.