Purpose: We hypothesize that serum human glandular kallikrein-2 (hK2) levels predict the presence of prostate cancer among men prescreened by prostate-specific antigen (PSA).
Patients and methods: We conducted a cross-sectional study of 324 men who had no history of prostate cancer and who were referred for prostate biopsy. PSA and hK2 levels were measured using specific nonisotopic immunometric techniques. Cases were patients who were diagnosed with adenocarcinoma of the prostate from biopsy, and controls were patients who had no evidence of cancer from biopsy. The odds ratio for detection of prostate cancer was determined for hK2 measurements, controlling for age, total-PSA level, digital rectal examination, and symptoms of urinary obstruction.
Results: Of 324 men, 159 (49.1%) had cancer. Mean hK2 levels and hK2:free-PSA ratios were significantly higher in cases than in controls (1.18 v 0.53 ng/mL, respectively, for hK2, P =.0001; 1.17 v 0.62 for hK2:free-PSA ratio, P =.0001). The crude odds ratio for prostate cancer detection for patients in the highest quartile of hK2 level was 5.83 (95% confidence interval [CI], 2.8 to 12.1; P =.0001) compared with patients in the lowest quartile. The adjusted odds ratio was 6.72 (95% CI, 2.9 to 15.6; P =.0001). Similarly, the crude and adjusted odds ratios for prostate cancer detection using the hK2:free-PSA ratio were 7.36 (95% CI, 3.6 to 15.1; P =.0001) and 8.06 (95% CI, 3. 7 to 17.4; P =.0001), respectively. These odds ratios were higher than that observed for prostate cancer detection by total-PSA level (2.73; P =.03).
Conclusion: Among men prescreened with PSA for prostate cancer, patients with high hK2 measurements have a five- to eight-fold increase in risk for prostate cancer, adjusting for PSA level and other established risk factors. hK2 measurements may be a useful adjunct to PSA in improving patient selection for prostate biopsy.