Use of various combinations of free, complexed and total prostate-specificantigen levels as predictors of the pathologic stage of prostate cancer

Background: The efficacy of various combinations of total, free and complex ed prostate-specific antigen (PSA) levels were assessed to predict the path ologic stage of prostate cancer. Methods: Total PSA (tPSA), free PSA (fPSA) and complexed PSA (cPSA) levels were measured preoperatively in 52 patients with clinical localized prostat e cancer who had undergone radical prostatectomy. Pathologic stages were cl assified as: organ-confined (n = 27); capsular penetration (n = 14); semina l vesicle involvement (n = 8); involvement of the surgical margins (n = 10) ; and lymph node involvement (n = 3). Results: The fPSA/tPSA and fPSA/cPSA ratios significantly differed between patients with organ-confined disease and non-organ-confined disease (P = 0. 035, P = 0.033, respectively) and between those with favorable versus unfav orable pathology (P = 0.001, P = 0.014, respectively), but tPSA, cPSA, fPSA and the cPSA/tPSA ratio did not. Using a fPSA/tPSA cutoff level of 11%, th e prediction of organ-confined disease would increase from 52 to 67% and th e rate of predicting favorable pathology would increase from 42 to 62%. A f PSA/cPSA cutoff level of 12% would increase the rate of predicting organ-co nfined disease to 79% and the rate of predicting favorable pathology would increase to 69%. The positive predictive value of the fPSA/cPSA ratio was h igher than that of the fPSA/tPSA ratio, although the receiver operating cha racteristic curve of the fPSA/cPSA ratio was not different from that of the fPSA/tPSA ratio. Conclusion: Although there was no predictive difference found between fPSA/ tPSA and fPSA/cPSA ratio, both ratios may help predict the pathologic stage of prostate cancer.