Background: Since free prostate-specific antigen (PSA) and gamma-seminoprotein (gamma-SM) recognize similar epitope(s) of PSA, the significance of serum-free PSA and gamma-SM in the early detection of prostate cancer was compared.
Methods: A prospective clinical trial was conducted on 701 male volunteers, age 50 years or older. Free PSA (Tandem-R free PSA, Hybritech) and gamma-SM (gamma-SM, Chugai) levels were determined, and biopsies were performed if the PSA (Tandem-R, Hybritech) level was > 4 ng/ml, or if digital rectal examination (DRE) was suspicious.
Results: One hundred and eighty-seven men (27%) had either a PSA level > 4 ng/ml or a suspicious DRE. Of 116 biopsies performed, cancer was detected in 13 (1.9%, 13/701). Receiver-operating characteristic analysis of free PSA to PSA ratio (free PSA ratio, %) and gamma-SM to PSA ratio (gamma-SM ratio), to differentiate normal biopsy findings from cancer, showed that the optimal values were 12% and 0.38, respectively. Positive predictive value for cancer was 24% (12 cancers/50 biopsies) for PSA alone, 42% (8/19) for the combination of PSA and DRE, 45.5% (10/22) for the combination of PSA and gamma-SM ratio, and 50% (10/20) for the combination of PSA and free PSA ratio. Regression analysis showed that gamma-SM highly correlated with free PSA, but that the analytical detection limit of gamma-SM was 1 ng/ml, significantly higher than that of free PSA.
Conclusions: Free PSA determination might effectively eliminate unnecessary biopsies in subjects with PSA > 4 ng/ml, and gamma-SM might provide a complementary index to free PSA, but its validity should be further studied in other settings, such as after radical prostatectomy or during endocrine treatment.