BACKGROUND: Data on utilizing complexed prostate specific antigen (cPSA) offering increased diagnostic performance over other available clinical parameters in diagnosis of prostate cancer is still controversial. Our objective was to determine diagnostic performance of cPSA compared to total prostate specific antigen (tPSA) and corresponding ratios for possible routine application.
METHODS: In a prospective study including overall 315 consecutive men, 177 patients with suspicious digital rectal examination, and/or tPSA value > 2.5 ng/ml underwent prostate biopsy. Serum samples for tPSA, cPSA and free PSA were analyzed using automated chemiluminometric technology.
RESULTS: Area under the curve (AUC) for cPSA, although greater, was not statistically different compared to that of tPSA (p = 0.253). AUCs of f/c, f/t and c/t ratios were all found significantly inferior. At clinically relevant 2.37 ng/ml threshold, cPSA performed with 85% sensitivity and significantly higher specificity of 63.1%, compared to same sensitivity and specificity of 57.2% at a 3.00 ng/ml cut off for tPSA.
CONCLUSIONS: Utilizing automated assay systems at predetermined cut off value for cPSA we would be able to save 27.1% of the biopsies while missing 13.4% of the cancers. Therefore, results of this study indicate higher discriminatory power of cPSA in diagnosis of prostate cancer for clinically relevant 2.5-4 ng/ml tPSA range.