Comparison of serum prostate specific membrane antigen, prostate specific antigen, and free prostate specific antigen levels in radical prostatectomy patients

Abstract

BACKGROUND. Higher preoperative prostate specific antigen (PSA) levels are associated with higher pathologic stage and grade in patients undergoing radical prostatectomy (RP). In earlier studies, serum prostate specific membrane antigen (PSMA) elevations were associated with clinical progression and hormone-refractory carcinoma. The goal of this study was to evaluate the serum markers PSMA, free PSA (FPSA), free:total PSA ratio (F:TPSA), and total PSA (PSA) in men undergoing RP. METHODS. Serum was obtained from 63 patients undergoing RP for clinically localized (T1c, T2) prostate carcinoma. Serum PSA and FPSA were determined by Hybritech Tandem-E(R) and Tandem-R(R), respectively, and PSMA was determined by Western blot analysis. Serum values for these markers were compared with the pathologic stage, surgical margin status, Gleason sum, prostate size (as calculated via reconstruction and transrectal ultrasound), tumor size based on pathologic assessment of the whole mount, and World Health Organization (WHO) grade of the prostatectomy specimen. Markers were also compared against demographic information and the patients' age and race. RESULTS. There was a weak correlation between serum PSA and positive surgical margins, higher Gleason sum, and WHO grade (P < 0.05). Receiver operating characteristic curve (ROC) analysis comparing sensitivity and specificity of the markers to positive and negative margins as well as seminal vesicle invasion demonstrated PSA and FPSA predictive ability for seminal vesicle invasion. The area under the curve for PSA and FPSA in this case was 0.7318 and 0.7432, respectively. There was also a weak correlation between the FPSA level and margins, with a low ROC area under the curve of 0.6789. The FPSA cannot distinguish the more advanced stage of disease. There was no significant correlation between F:TPSA and PSMA with regard to the study variables in predicting organ confinement. High PSMA levels only correlated with higher stage and were maximal in pT4a classified disease. CONCLUSIONS. Higher PSA and FPSA levels are likely to be associated with more locally advanced disease. Total PSA was the best marker. However, the cutoff values necessary for significant accuracy between PSA and FPSA are not of clinical usefulness due to the lack of specificity and sensitivity of the markers at those cutoffs. F:TPSA and PSMA levels as currently measured are of limited value in discriminating more aggressive disease in patients with clinically localized CaP.