Discrimination of Benign From Malignant Prostatic Disease by Selective Measurements of Single Chain, Intact Free Prostate Specific Antigen

Abstract

Free prostate specific antigen (PSA) in serum consists of heterogeneous molecular subforms. Recently we developed an immunoassay for selective measurement of a subfraction of free PSA called intact PSA, which has been shown to be closely associated with prostate cancer. We assessed the ability of serum intact PSA to discriminate between benign and malignant prostatic disease. In serum of 178 men with benign disease and 255 men with prostate cancer we measured total PSA and free PSA using a commercially available immunoassay. Intact PSA levels were analyzed by a newly developed assay specific for noncleaved, that is single chain forms of free PSA. Internally cleaved "nicked" PSA was calculated by subtracting intact from free PSA. We also calculated ratios of intact PSA-to-free PSA (intact-to-free PSA) and nicked PSA-to-total PSA (nicked-to-total PSA). We compared means, medians and ranges of all analytes and ratios in patients with and without cancer for the entire total PSA range and in a subset with total PSA ranging from 2 to 10 ng./ml. Furthermore, various combinations of PSA forms were tested for their predictive ability. For statistical comparison we used the Mann-Whitney U test and ROC analysis. The ratio intact-to-free PSA was significantly higher in cancer (median 48.5%) compared to noncancer cases (median 41.8%, p <0.0001). Conversely, the ratio nicked-to-total PSA was significantly higher in men without compared to those with prostate cancer (median 11.0% and 6.0%, respectively, p <0.0001). Highest discriminative ability was observed for a combination of intact, total and free PSA (log [intact, free, total], AUC = 0.773) followed by nicked-to-total PSA (AUC 0.755). In the subgroup of patients with total PSA levels from 2 to 10 ng./ml. only the AUC of log intact, free, total (AUC 0.706, p = 0.0017) and nicked-to-total PSA (AUC 0.704,p = 0.0019) were significantly larger compared to the AUC of total PSA (AUC 0.602). By contrast to measuring crude free PSA concentration, selective determination of specific free PSA subforms, intact PSA and nicked PSA proved to be useful to discriminate men with benign from malignant prostatic disease. These markers may serve to generate specific serum profiles of PSA for improved specificity and early detection of prostate cancer. To translate the encouraging statistical advantage shown in this study into a clinically applicable tool warrants further investigation.