Study Of PSAD And Mp-MRI In Multifocal And Clinically Significant Prostate Cancer

Objective:Multi-parameter magnetic resonance imaging(mp-MRI)and prostate-specific antigen(PSA)have a critical role in the diagnosis,staging and management of Pca.Prostate-specific antigen density(PSAD)excludes the effect of prostate epithelial cell count on tPSA concentration.There are no studies of PI-RADS V2.1,PSAD,and Apparent diffusion coefficient(ADC)value for the prediction of multifocal and clinically significant prostate cancer(MF-csPca).Therefore,this study investigated the diagnostic efficacy and predictive value of various patient observations including PSA value,PSAD value and PI-RADS V2.1 score for MF-csPca,the correlation with Gleason grading,as well as the diagnostic efficacy of different combined models in MF-csPca to further improve the accuracy of preoperative diagnosis of multifocal csPca.Methods:Patients with suspected Pca and mp-MRI for abnormally elevated PSA from January 2018 to August 2020 at our Hospital were collected,the enrolled cases were grouped according to pathological diagnosis and Gleason grading results:Group A:including MF-csPca group,Group B:including prostatitis,benign prostatic hyperplasia(BPH)and clinically insignificant cancer(ciPca)group.Clinically significant prostate cancer(csPca)was defined as Pca with a Gleason score≥ 3+4,or volume≥ 0.5 cm3,or with extraperitoneal invasion,and clinically insignificant prostate cancer(ciPca)was defined as Pca with a Gleason score<3+4,or volume<0.5 cm3,and without extraperitoneal invasion.SPSS 21.0 and MedCalc software were used for statistical analysis,firstly,the univariate analysis was performed,and receiver operating characteristic(ROC)curve were plotted for indicators of statistically significant differences between the two groups in the univariate analysis of variance and the area under the curve(AUC)of ROC was derived.Multi-factor analysis was then performed using binary logistic regression to derive strong predictors for indicators with statistically significant differences between the two groups.Then,Spearman correlation analysis was used to assess the correlation between predictors and Gleason classification.The combined prediction models were developed separately according to binary logistic regression,the ROC curves were plotted to calculated the diagnostic efficacy,the accuracy,sensitivity and specificity value,negative predictive value(NPV)and positive predictive value(PPV)of each combined prediction model separately and compare the differences in AUC value between the different combined models and their individual diagnoses of MF-csPca(P<0.05).Results:A total of 202 patients were included in the study,Group A:MF-csPca group(86 patients),Group B(116 patients in total):including prostatitis(31 patients),BPH(77 patients)and ciPca(8 patients),where prostatitis was all chronic non-bacterial prostatitis.The above patients were divided into 2 groups.32(15.9%)cases had a PI-RADS 2 score,75(37.1%)cases had a PI-RADS 3 score,40(19.8%)cases had a PI-RADS 4 score and 55(27.2%)cases had a PI-RADS 5 score among all patients.Univariate analysis showed that:prostate volume,tPSA,fPSA f/tPSA,PSAD,ADC value and PI-RADS V2.1 were statistically significant between group A and group B.Further multivariate analysis showed that PI-RADS V2,1,ADC value and PSAD were Influential predictors of MF-csPca.Correlation analysis with Gleason’s classification showed PI-RADS V2.1 score were moderately positively correlated with Gleason’s classification,rs=0.531,P<0.001,PSAD was moderately positively correlated with Gleason’s classification,rs=0.475,P<0.001,ADC value were moderately negatively correlated with Gleason classification,rs=-0.499,P<0.001.ROC curve analysis PI-RADS V2.1,ADC value,and PSAD diagnostic MF-csPca AUC value was 0.833,0.857,and 0.839,respectively.Three combined models were developed based on the above results:the combined model 1(PI-RADS V2.1 PSAD),combined model 2(ADC value+PSAD value)and combined model 3(PI-RADS score+ADC value+PSAD value)were established,with diagnostic MF-csPca AUC value were 0.876,0.887,and 0.896,respectively.Conclusions:The PI-RADS V2.1 score,PSAD value and ADC value were all Influential predictors of MF-csPca,and were significantly more effective than prostate volume,fPSA value,f/tPSA value and slightly better than tPSA value in the diagnosis of MF-csPca,and correlated with pathological Gleason grading,with the PI-RADS V2.1 score having the highest correlation.It is suggested that PI-RADS V2.1 score,PSAD value and ADC are not only helpful for the diagnosis of MF-csPca,but also a guide for the preoperative assessment of the invasiveness of Pca.The diagnostic efficacy of the combined model for MF-csPca was higher than that of a Single indicator,the combined diagnostic efficacy of the model three(PI-RADS V2.1 score+ADC value+PSAD value)for MF-csPca was the highest,suggesting that the combination of PSAD and ADC value,in addition to PI-RADS V2.1 score by mp-MRI,is needed to further improve the accuracy of the preoperative diagnosis of MF-csPca and reduce the need for unnecessary biopsies.