• M Z Zuo, W L Zhao, C G Wei, C Y Zhang, R Wen, Y F Gu, M J Li, Y Y Zhang, J F Wu, X Li, and J K Shen.
• Department of Radiology, Second Affiliated Hospital of Soochow University, Suzhou 215004, China.
• Zhonghua Yi Xue Za Zhi. 2017 Dec 19; 97 (47): 3693-3698.

AbstractObjective: To investigate the preliminary applicability of Prostate Imaging Reporting and Data System version 2 (PI-RADS v2) score in the condition of 3.0T multi-parametric magnetic resonance imaging (Mp-MRI) combined with clinical classic indicators for the diagnosis of prostate cancer (PCa). Methods: The clinical and MRI materials of 247 patients of suspicious prostate disease treated in Second Affiliated Hospital of Soochow University from June 2015 to November 2016 were analyzed retrospectively, including 110 cases with PCa and 137 cases without cancer.All cases underwent the high-resolution axial T(2)-weighted imaging (T(2)WI), diffusion weighted imaging (DWI) and dynamic contrast enhancement-magnetic resonance imaging (DCE-MRI) and were confirmed pathologically by puncture biopsies.The Mp-MRI materials of all cases were scored according to PI-RADS v2.The prostate volume and prostate specific antigen (PSA) density (PSAD) value were calculated according to the formulas.The univariate and multivariate analysis were performed for the observed indicators (age, prostate volume, PSA, PSAD and PI-RADS v2 score) to determine the independent predictors for PCa.Then, a Logistic regression model (combined prediction model) was established by the independent predictors for combined diagnosis of PCa.The receiver operating characteristic curve (ROC) curve analysis was performed to get the sensitivity and specificity of each independent predictor and the model to diagnose PCa.The differences of AUC values of each independent predictor and the model were compared with each other to evaluate the diagnostic performance for PCa. Results: The differences in the age, prostate volume, PSA, PSAD and the PI-RADS v2 score between patients with PCa and non-cancer group were all statistically significant (t=2.870, Z=-4.230, -7.787, -9.477, -10.826, all P<0.05). The PSAD and PI-RADS v2 score were independent predictors for PCa (OR=3.331, 10.546, both P<0.05). The Logistic regression combined prediction model by PI-RADS v2 score and PSAD to forecast PCa was Logit(P)=-5.097+ 2.309×PSAD+ 1.214×PI-RADS v2 score.The area under the curve (AUC) of ROC in the combined model (0.911) was higher than that in the PI-RADS v2 score (0.886) and PSAD (0.851) and the differences were all statistically significant (Z=2.416, 2.716, both P<0.05); but the difference in the AUC value between PI-RADS v2 score and PSAD was not statistically significant (Z=1.191, P=0.234). The diagnostic sensitivity of PSAD, PI-RADS v2 score and the model were: 0.891, 0.782 and 0.855, respectively; the specificity were 0.449, 0.912 and 0.847, respectively on their positive thresholds (0.15 μg·L(-1)·ml(-1,) 4 and -0.82). Conclusion: PI-RADS v2 score combined with PSAD in diagnosing PCa is superior to the single application of them and it can lead to high diagnostic sensitivity and specificity for PCa.

10 keywords...

hide…