Predictive Value Of PSA-derived Indexes Represented By FPSAD On Prostate Puncture Results

OBJECTIVE: In this study,the clinical data of 416 patients with prostate cancer(PCa)or benign prostatic hyperplasia(BPH)diagnosed by prostate puncture biopsy were analyzed.To investigate the value of PSA-derived indexes represented by free prostate specific antigen density(f PSAD)for predicting prostate puncture results,and to establish a prediction model to further improve its clinical application.METHODS: Clinical data(PSA,f PSA,age,prostate volume,height,weight,and pathological findings)of patients who underwent transrectal ultrasound-guided prostate puncture from January 2016 to June 2022 were collected from Gansu Provincial People’s Hospital,and the included PSA-derived indexes were calculated and divided into PCa and BPH groups based on pathological findings.Data analysis was performed by SPSS 21.0 and Med Calc 20.0 software.Comparisons between the two groups were performed using nonparametric tests,and correlations between the indicators and PCa tests were observed by Spearman correlation analysis.The main observed values(AUC value,Jorden index,best cut-off value,sensitivity,specificity,positive predictive value,negative predictive value,diagnostic compliance rate,etc.)were derived from the subject operating characteristic curve(ROC)analysis based on the correlation results,and each index was analyzed at different PSA,PV intervals,and logistic regression analysis was applied to establish prediction models.RESULTS: General information of the patient: age 69.88±8.15 years,prostate volume59.32±34.46 ml,divided into PCa group(206 patients,49.52%)and BPH group(210 patients,50.48%).PSAD(0.774),f PSAD(0.764),A-PSAD(0.759),PSAMR(0.750),(F/T)/PSAD(0.702),PSA(0.688),PSAM(0.686),%(F/T)/PV(0.613),and F/PSAD(0.609)had statistically significant AUC values(P < 0.05),and f/t PSA did not have statistically significant AUC values(P > 0.05).In the PSA 4 – 10 ng/ml interval %(F/T)/PV(AUC = 0.758,Sen = 74.07%,Spe =76.06%)was significantly better and statistically significant than the other indicators,and the predictive model had a significantly improved AUC of 0.789 predictive power;in the PSA10 –20 ng/ml and >20 ng/ml intervals,PSAD was significantly better than other indicators,but the AUC of A-PSAD,f PSAD and PSAMR were above 0.7 and statistically significant.In the PSA10 – 20 ng/ml interval the prediction model had a significantly improved AUC of 0.804,while in the PSA>20 ng/ml interval only the A-PSAD failed to significantly improve the prediction ability.f PSAD(AUC = 0.723,Sen = 67.03%,Spe = 77.78%)in the PV < 40 ml interval(AUC = 0.786,Sen = 78.49%,Spe = 72.04%)was significantly better than the other indicators and statistically significant,with an AUC of 0.734 for the prediction model;A-PSAD(AUC = 0.786,Sen = 78.49%,Spe = 72.04%)was significantly better than the other indicators and statistically significant,with an AUC of 0.796 for the prediction model in the PV 40-80 ml interval;A-PSAD(AUC = 0.816,Sen = 77.27%,Spe = 80.95%)was significantly better and statistically significant in the PV greater than 80 ml interval,with a predictive model AUC value of 0.864.Conclusions: 1.The predictive power of the included PSA-derived indicators f PSAD,PSAMR,A-PSAD,and %(F/T)/PV was generally similar to PSAD,but superior to PSAD in individual intervals.(F/T)/PSAD,F/PSAD,and PSAM were not observed to be superior in this study,although there was some correlation between prediction and prostate puncture results.2.f PSAD was superior to PSAD in the interval of PSA 4 – 10 ng/ml and PV less than 40 ml,and had the strongest predictive power in the interval of PV less than 40 ml;3.%(F/T)/PV had the strongest predictive power in the interval of PSA 4 – 10 ng/ml compared to the included indexes.4.APSAD is similar to PSAD in different PSA intervals,but it is better than PSAD in different PV intervals,and has the strongest predictive power in PV 40 – 80 ml and >80 ml compared with the included indicators;5.The predictive power of PSAD remained the strongest in the interval of PSA 10 – 20 ng/ml and >20 ng/ml;6.The predictive modeling of the combined controversial PSA-derived indicators significantly improved the predictive power of PCa.