[Does plasmatic dilution influence the validity of PSA-tests?]

The aim of this study is to improve the validity of PSA-based tests by considerations of the dilutions caused by the individual plasmatic blood volume (or other relevant body fluids).

Retrospective analyses of patients who underwent adenomectomy (n = 32) or radical prostatectomy (pT1 - 4 N0 M0; n = 60) were carried out. We calculated the PSA density (PSAD) and two new PSA-based parameters (mPSA (Vp), mPSAD (Vp)) that converted t-PSA concentration and PSAD with regard to individual Vp (according to Sprenger's modified Retzlaff formula). Comparative statistics of receiver operating characteristics (ROC-) curves, AUC, sensitivity, specificity, positive and negative predictive values for t-PSA, mPSA (Vp), PSAD and mPSAD (Vp) were performed.

PSA was positively correlated with local tumour stage. With regard to the whole range of t-PSA (n = 92; 0.1 - 88.4 microg/L) the diagnostic selection between prostate carcinoma (CaP) and benign disease (BPH) was significantly improved by PSAD (AUC = 0.803) and mPSAD (Vp) (AUC = 0.806) (p < or = 0.003) compared to t-PSA und mPSA (Vp) (AUC = 0.531). Within the range of 4.0 - 10.0 microg/L PSA, the areas under the ROC curves were much better for t-PSA (AUC = 0.663), mPSA (Vp) (0.694) and PSAD (0.931) in gereral; mPSAD (Vp) provided the best AUC (0.947). However, although considering Vp does demonstrate better AUCs, this tendency does not reach the level of significance (yet).

The most conclusive way to improve PSA test validity is to adjust PSA to different "volumes". Therefore, elaborated devices for a better preoperative investigation of the whole volume of the prostate gland and its distinct partial volumes (such as carcinoma or benign tissues), as well as applied knowledge on the distribution and kinetics of PSA in body fluids, might substantially help to optimise the detection of curable patients with unknown carcinoma of the prostate.