Serum PSA is currently the most widely used PCa screening test, however, PSA is organ- but not disease-specific, thus its value to predict PCa is limited and there is a need for additional disease markers. We consider SP as an ideal source to monitor pathological changes in the prostate since it more directly reflects local glandular production than other sources such as peripheral blood. Moreover, SP donation is completely non-invasive avoiding the discomfort and possible concerns associated with prostate massage necessary to obtain post-massage urine. Previously, we reported that the inactive free:total PSA ratios in SP from men with suspected PCa did not distinguish between those harboring malignant as opposed to benign disease . The present study was designed to assess changes in Dkk-3 levels in SP and its potential for discriminating benign from malignant prostatic conditions in patients who were to undergo TRUS-guided biopsy due to elevated serum PSA (> 4 ng/mL) and/or abnormal DRE.
With the IEMA developed in our laboratory, SP Dkk-3 levels were found significantly higher in patients with biopsy confirmed PCa (100.9 ± 12.3 vs. 69.2 ± 9.4 fmol/mg). Of note, this difference was more pronounced in the subgroup of patients with a clinical follow-up of > 3 years (100.0 ± 16.6 vs. 56.3 ± 8.4 fmol/mg) for whom, as a consequence, there was greater veracity for the negative biopsy finding to be correct (Table 1).
While Dkk-3 levels as a single marker distinguished between prostate cancer and non-cancer on the basis of TRUS biopsies in the study cohort with a similar diagnostic accuracy compared with serum PSA (SP Dkk-3: AUC = 0.633; serum PSA: AUC = 0.644), the diagnostic accuracy of SP Dkk-3 levels was enhanced in multivariate models including serum PSA (model A) and both, serum and SP PSA levels (model B). Neither SP Dkk-3 nor SP PSA levels correlated with biopsy Gleason scores. However, this might be in part due to the patient cohort characteristics with approximately 2/3 of all positive biopsies reported as Gleason 7 (either Gleason 3+4 or Gleason 4+3). Of interest SP Dkk-3 negatively correlated with SP PSA levels in the positive biopsy cohorts, which might be responsible for the higher accuracy of model B compared with model A.
The biological significance of the increase in SP Dkk-3 levels in patients with positive vs. negative PCa biopsy findings remains unclear. As expression of Dkk-3 in secretory epithelial cells of the prostate is reduced in both PCa [5, 13, 20] and BPH , one might have expected a similar effect on SP Dkk-3 levels in patients with both these conditions. However, it is conceivable that the elevated Dkk-3 levels in PCa SP are derived from the tumor neovasculature, where high levels of Dkk-3 are produced . Elevated serum PSA levels in PCa patients are supposed to be primarily an effect of tissue degradation and thus increased tissue permeability while in normal/BPH tissue an intact basement membrane prevents leakage . Likewise, Dkk-3 as a secreted protein of similar size (PSA protein backbone: 237 aminoacids; Dkk-3 protein backbone: 329 aminoacids) might be able to diffuse through the disorganized PCa tissue.
Due to the limited number of SP samples available, especially from patients with long-term follow-up, it was not useful to further stratify the biopsy confirmed PCa patients according to their risk in order to more closely investigate the SP Dkk-3 and SP PSA correlation within this cohort.
Given the limited diagnostic accuracy of single markers including PSA, much effort has been made to identify potential additional PCa markers from different sources including SP [22–24]. Our data demonstrate that SP is a valuable and valid source for biomarkers of prostatic disease. The most promising attempt to improve PCa diagnosis appears the establishment of a set of individual biomarkers. Thus, SP Dkk-3 should be considered in a panel of markers such as prostate cancer antigen 3 (PCA3) and/or hepsin  to be evaluated in combination in future validation studies.