Prediction Of Biparameter MRI PI-RADS Combined With Clinical Indicators For Clinically Significant Prostate Cancer

Objective:The incidence of clinically significant prostate cancer（csPCa）among men in my country ranks among the highest in Asia,which is a threat to the health of most elderly men.Clinical indicators such as total prostate specific antigen（tPSA）,free prostate specific antigen（fPSA）,the ratio of free prostate antigen to total prostate antigen（fPSA/tPSA）,prostate specific antigen density（PSAD）and other clinical indicators hav e been widely used in clinics,and has certain diagnostic value to exist.Based on the prostate imaging reporting and data system version 2.1（PI-RADS v2.1）,this article explores the biparametric MRI（bp-MRI）PI-RADS combined with clinical indicators to predict clinically significant prostate cancer,in order to provide a basis for clinical puncture,or to provide a prediction of the malignant degree of the disease for patients with contraindications of puncture biopsy.Methods:A total of 74 patients with elevated PSA or abnormal digital rectal examination who attended the Department of Urology of Jincheng General Hospital from March 2019 to December 2020.Under the guidance of rectal ultrasound examination,these were pathologically confirmed as prostate cancer or benign prostatic hyperplasia.Detailed and complete records of the general information of patients,including name,age,prostate specific serum index test data and prostate cancer patients’pathological Gleason score and so on.All patients underwent prostate biparametric MRI examination,and the prostate volume was measured and recorded in detail.The image PI-RADS score was judged by two imaging doctors independently and blindly.When there was a dispute,the two groups discussed and gave the same results.The patients were grouped according to whether the pathological Gleason score was greater than or equal to 7（non-csPCa group and csPCa group）,the Mann-Whitney U test was used to compare whether the differences in clinical and imaging indicators between the two groups were statistically significant;binary Logistic regression analysis and discriminant analysis were used to predict clinically significant prostate cancer.Results:There was no statistically significant difference in age between the two groups（P>0.05）.The median prostate volume in the non-csPCa group was 53.97 cm³,which was obviously greater than that in the csPCa group（41.39cm³）,and the difference was statistically meaning（P<0.05）.There was no statistical discrepancy in fPSA/tPSA between the two groups（P>0.05）,and there were significant differences in tPSA,fPSA,PSAD and PI-RADS（P<0.05）.Spearman related analysis evidenced that PI-RADS score was positively correlated with tPSA,fPSA and PSAD（r>0.01,P<0.01）,and binary Logistic regression analysis showed that PI-RADS score was the main predictor of csPCa（P<0.05）.The discriminant analysis showed that the overall prediction accuracy was 85.1%,among which the accuracy of non-csPCa group was 91.3%,the accuracy of csPCa group was 75.0%,and the Kappa correlation coefficient was 0.677,P<0.001.The establishment of discriminant function indicated that the correlation coefficients of age,f/t,PI-RADS and PSAD were higher.The data of ungrouped cases were divided into two groups by using the typical discriminant function model and the mean values of each index in non-csPCa group and csPCa group.The calculated values were?Y^<1>=-0.671,?Y^<2>=1.106,and then Y₀=0.002.When Y>Y₀,it belongs to the csPCa group;when Y<Y₀,it belongs to the non-csPCa group.Conclusion:Biparametric MRI PI-RADS combined with clinical indicators can predict clinically significant prostate cancer,and the main predictor of csPCa is PI-RADS score.The Y₀calculated by the discriminant analysis statistical method is used to judge the data of the unknown group of prostate disease patients and predict whether it is csPCa,in order to reduce unnecessary biopsy,and play a vital role in the diagnosis and treatment of csPCa.