Establishment And Validation Of Biochemical Recurrence Risk Prediction Model After Radical Prostatectomy Based On Meta-analysis

Objective:This study intends to use a Meta-analysis approach to establish a prediction model for biochemical recurrence(BCR)risk assessment after radical prostatectomy(RP)in prostate cancer(PCa)patients that is easily generalizable in clinical practice,thereby providing assistance for early intervention in high-risk individuals with BCR,with the aim of promoting rehabilitation,reducing family and socioeconomic burden,and improving survival outcomes for PCa patients.Methods:(1)Risk factors for the development of BCR after radical prostatectomy in patients with PCa were reviewed by means of a Meta-analysis based on predefined criteria for the literature search and quality assessment,and the Risk ratio(RR)corresponding to the risk factors involved were combined(2)The risk assessment prediction model for BCR incidence was established by combining the RR values obtained for each risk factor and the respective 95%confidence intervals(95%CI)according to the method of Meta-analysis to calculate the regression coefficients of the corresponding risk factors and calculate the risk factor scores.A total of 170 PCa patients who underwent radical prostatectomy at the Department of Urology,Jiangxi Provincial People’s Hospital between June 2012 and July 2017 were used as the validation cohort to evaluate the predictive model by comparing the sensitivity,specificity and the area under the ROC curve(AUC)of the model for the prediction of high-risk patients with BCR when different cut point values were taken,and the best predictive cut point value was selected to risk stratify PCa patients.Results:In this meta-analysis,a total of 25 original studies were included and screened for10 risk factors for BCR development in prostate cancer patients after RP that differed significantly,including age(≤65years),BMI(≥25 kg/m~2),preoperative PSA(10 ng/ml≤PSA<20 ng/ml,PSA≥20ng/ml),proportion of positive needle passes(>50%),postoperative pathologic Gleason score(=3+4,=4+3,≥8),postoperative pathologic stage(≤p T3a,≥p T3b)For seminal vesicle invasion,positive surgical margins,lymph node metastasis,peripheral nerve invasion,the pooled RR and its 95%Cl were 1.06,1.90,1.08,2.29,1.82,1.54,2.09,3.11,1.50,2.27,1.93,2.21,1.69,and 1.64,respectively.For inclusion of the above risk factors in the construction of the risk assessment model,the corresponding scores for each risk factor were as follows:age(>65 years to confer 0 points,≤65 years to confer 0.5 points),BMI(<25 kg/m~2 to confer0 points,≥25 kg/m~2 to confer 6.5 points),preoperative PSA(<10 ng/ml to confer 0points,10 ng/ml≤PSA<20 ng/ml to confer 1 point,≥20 ng/ml to confer 8.5 points),proportion of positive needle passes(≤50%to confer 0 points,>50%to confer 6 points),and Postoperative Gleason score(=6 confer 0 points,=3+4 confer 4.5 points,=4+3confer 7.5 points,≥8 confer 11.5 points),postoperative pathological stage(≤p T3a confer 4 points,≥p T3b confer 8 points),seminal vesicle invasion(negative confer 0points,positive confer 6.5 points),surgical margins(negative confer 0 points,positive confer 8 points),lymph node metastasis(negative confer 0 points,positive confer 5.5points),peripheral nerve invasion(negative confer 0 points,positive confer 5 points).In this study,the developed prediction model for the onset of BCR was validated in a retrospective cohort study,and it was found that the optimal predictive cut point value of the model was selected at 28.0 points,that is,when the total score of patients involved was≥28.0 points,the sensitivity of the model prediction was 0.787,the specificity was 0.835,the AUC and 95%CI were 0.797(0.714,0.880),and it had good predictive efficacy.Conclusions:This study developed a prediction model for BCR incidence risk assessment using Meta-analysis,which was validated in a retrospective cohort population with a good prediction efficacy,and this model is simple to use and easily generalizable,which may facilitate early detection of patients at high risk of BCR and timely individualized and comprehensive measures of salvage therapy to improve patient survival.