Objectives: "I he study aimed to evaluate the diagnostic yield of estimation of eomplexed prostatespecific antigen (cPSA) in random sample of patients with symptoms suggestive of prostate cancer (PC) and had serum total PSA (tPSA) of 2.5-20 ng/ml and no previous histologic proof of Pc.
Patients and Methods: Prior to prostatic examination, venous blood samples were collccted for ELlSA estimation PSA serum leycls. Then, patients underwent trans-reclal ultrasonography (TRUS) to assess the volume of the prostat(O, the adenoma and the peripheral zone (PZ) of the prostate. All patients underwent IO-('ore TRUS-guided biopsy taking from the prostatic PZ~
Results: Thc study included 223 malc patients fulfilled the study inclusion eriteria with mean prostatie volume of 60.1±14.8 em" and mean PZ volumc of 41.8±9.4 em3 • Histopathologiclll examination of core "iopsies defined 23 cases with PC (PC group) with a frcqucncy of cancer detection of 10.3% while the other 200 patients were free of PC (Control group). Mean estimated serum tPSA and cPSA levels wert: signillcantly higher in PC group compared to control grollP, while serum free PSA (fPSA) levels were non-significantly lower in PC group compared to control group. Mean PSA densities were significantly higher in PC group compared to control group. There was a positive signi licant correlation between the presence of PC and age of patients and with levels and dcnsitics of tPSA and ePSA. ROC curve ana!,sis defincd serum cPSA level as highly specific predictor for the presence of PC with AUC=0.987 that showed a significant difference in favor of cPSA Icvel in comparison to tPSA and fPSA. Logestie regrcssion analysis defined cPSA level, PZ cPSAD and PZ PSAD as the signifieant predictors of PC and ROC curv,e analysis of the three parameters defined PZ cPSAD and cPSA serum level as thc highest specific prc~Ectors of PC.
Conclusion: Estimation of serum cPSA could define cases of PC with high specificity during screening of patients had serum tPSA levels ranged between 2.5 and 20 ng/mland used incombination with detcmlination of PZ cPSAD are highly significant spccific predictors of Pc.