Objectives. To investigate whether different molecular forms of prostate-specific antigen (PSA) ratios could provide prognostic information about the stage and grade of prostate cancer.

Methods. Serum specimens were examined from 53 patients who underwent radical prostatectomy for clinically localized prostate cancer and from 94 patients diagnosed as having no evidence of malignancy (total PSA between 4.1 and 20.0 ng/mL). The free/total PSA (fPSA/tPSA) ratio and complexed/total PSA (cPSA/tPSA) ratio in 18 patients with biochemical recurrence were compared with those of patients either without recurrence or with no evidence of malignancy.

Results. The fPSA/tPSA and cPSA/tPSA ratios differed significantly (P <0.05) between patients with organ-confined and those with non-organ-confined disease, but the tPSA, cPSA, and fPSA levels did not (P >0.05). The median values of the fPSA/tPSA ratio in patients with recurrence (7.0%) were significantly lower than in the patients without recurrence (8.9%) or those without evidence of malignancy (15.2%) (P = 0.02 and P <0.01, respectively). The median values of the cPSA/tPSA ratio in patients with recurrence (97.4%) were significantly higher than in patients without recurrence (92.9%) or those without evidence of malignancy (86.7%) (P <0.01 and P <0.01, respectively). At the time of recurrence, 6 (33%) of 18 patients expressed higher fPSA/tPSA ratios (15% or greater) and lower cPSA/tPSA ratios (less than 95%). Five (83%) of these 6 patients appeared to have aggressive tumors according to pathologic stage.

Conclusions. The fPSA/tPSA and cPSA/tPSA ratios differed significantly among the three groups. Thus, a subset of tumors appears to be capable of producing high fPSA/tPSA and low cPSA/tPSA ratios at the time of recurrence, and some of these have an aggressive phenotype. Until this latter phenomenon can be adequately explained, use of these ratios for prognostic purposes should be approached cautiously.