New biomarkers for prostate cancer are needed. We determined whether a novel serum marker, total PSP94 can be used to accomplish these goals. We conducted a case-control study of 1,212 men with no previous history of prostate cancer and who underwent a prostate biopsy from 1998 to 2000 because of an increased PSA or an abnormal DRE. Serum PSP94 levels were assessed using a sandwich enzyme-linked immunosorbent assay technique. Cases were patients with prostate cancer, and controls were patients who had no evidence of cancer. Multivariate logistic regression analysis was used to determine whether or not PSP94 levels improved the predictive value for prostate cancer. Of the 1,212 men 596 (49.2%) had cancer detected. The median PSP94 level was significantly lower among cases (2.60 ng/ml) than among controls (3.40 ng/ml, p <0.0001). The adjusted odds ratios for the presence of prostate cancer for patients with the lowest quartile of PSP94, compared to patients in the highest quartile was 2.70 (95% CI 1.8 - 4.0, p <0.0001). Among a subgroup of 649 men in whom PSA had a low predictive value (PSA less than 20 ng/ml, normal DRE and less than 70 years), 260 (40.1%) were found to have cancer. In this subgroup total PSP94 levels helped discriminate between patients with high grade disease (Gleason score 8 or more, median 1.90 ng/ml), moderate grade disease (Gleason score 7, median 2.34 ng/ml) and low grade disease (Gleason score 6 or less, median 2.60 ng/ml, p = 0.007). PSA and the FTPSA were not able to distinguish between patients with different grades in this group. Patients with low total PSP94 levels had a high probability for having prostate cancer detected at biopsy. The total PSP94 level was able to help identify patients with high grade disease among a subset of patients in whom PSA and FTPSA are least informative.
Read full abstract