Human glandular kallikrein as a tool to improve discrimination of poorly differentiated and non-organ-confined prostate cancer compared with prostate-specific antigen.

摘要

OBJECTIVES: Human glandular kallikrein (hK2) possesses 80% structure identity with prostate-specific antigen (PSA) and is secreted by identical prostate epithelial cells. Although increasing with pathologic stage, PSA is not clinically sufficient to predict histologic grade and pathologic stage of prostate cancer (PCa) in individual cases. To address this issue, serum hK2 in various PCa grades was investigated. METHODS: Sera from 122 consecutive patients with PCa, graded as well-differentiated (G1, n = 35); moderately differentiated (G2, n = 61), and poorly differentiated (G3, n = 26) PCa, was studied. In patients who underwent radical prostatectomy (n = 42), 24 had organ-confined (pT2a-b) and 18 extracapsular (pT3a or greater) disease. hK2 was measured by an indirect immunofluorometric assay with a functional sensitivity of 0.03 ng/mL. Total PSA (tPSA), free PSA (fPSA), and PSA bound to alpha(1)-antichymotrypsin (PSA-ACT) were also measured. Multivariate logistic regression analysis was used for evaluation of the best combinations of tumor markers. RESULTS: Median hK2 and tPSA increased twofold from G1 to G2 tumors (hK2 0.07 versus 0.14 ng/mL, P <0.002; tPSA 6.1 versus 12.1 ng/mL, P <0.0002). Between G2 and G3 tumors, hK2 increased threefold (0.14 versus 0.43 ng/mL, P <0.02), and tPSA showed no significant increase (12.1 versus 26.5 ng/mL, P <0.18). The f/t PSA ratio decreased between G1 and G2 cancers (0.15 vs. 010, P <0.001); no difference was found between G2 and G3 tumors (0.10 versus 0.11, P = 0.93). However, the hK2/fPSA ratio distinguished between G1 and G3 tumors and G2 and G3 tumors (0.085 [G1] and 0.11 [G2] versus 0.22 [G3], P <0.0002 and P <0.002, respectively). Using multivariate regression analysis, the fPSA/(tPSA x hK2) ratio differentiated G2 and G3 tumors (P <0.01). In the tPSA range of 3 to 15 ng/mL, hK2, the hK2/fPSA ratio, and the fPSA/(tPSA x hK2) ratio differentiated between the G1/G2 and G3 tumors, and tPSA, the f/t PSA ratio, and PSA-ACT did not. In radical prostatectomy cases, hK2 (0.06 versus 0. 156, P <0.005) and the fPSA/(tPSA x hK2) ratio (2.104 versus 0.828, P <0.005) discriminated between pT2a-b and pT3a or greater PCa. CONCLUSIONS: hK2 significantly improved the identification of poorly differentiated (G3) tumors compared with PSA. By multivariate logistic regression analysis, the hK2/fPSA and fPSA/(tPSA x hK2) ratios further improved the detection of PCa grade. This improvement was also seen with the intermediate range of tPSA. hK2 was also helpful in the prediction of organ-confined disease. Thus, hK2 may be a useful tool for more accurate prediction of tumor grade or stage and allow better clinical decision-making.