HUMAN GLANDULAR KALLIKREIN 2: A POTENTIAL SERUM MARKER FOR PREDICTING THE ORGAN CONFINED VERSUS NONORGAN CONFINED GROWTH OF PROSTATE CANCER

Abstract

We measured serum levels of human glandular kallikrein 2 (hK2) in patients with prostate cancer treated with radical retropubic prostatectomy for clinically localized prostate cancer to determine whether preoperative hK2 levels discriminate stage pT2a/b from pathological stage T3a or greater cancer. This finding would help to predict preoperatively the organ confined versus non-organ confined growth of prostate cancer. A total of 68 consecutive men underwent radical retropubic prostatectomy for clinically localized prostate cancer. Serum was obtained 1 day preoperatively before prostatic manipulation. hK2, and total and free prostate specific antigen (PSA) were measured using immunofluorometric assays. Mean, median and range of hK2, total and free PSA, and the ratio of free-to-total PSA (percent free PSA) were calculated. Each analyte or combination of analytes was evaluated to determine whether it significantly contributed to enhance the discrimination of organ confined from non-organ confined cancer. We calculated the statistical significance of observed differences using the Mann-Whitney U and Kruskal-Wallis tests. Sensitivity and specificity calculations were performed for hK2, total PSA and the algorithm, (hK2) x (total PSA/free PSA) in addition to receiver operating characteristics curves and the respective areas under the curves. Multivariate logistic regression analysis was done for hK2, and total and free PSA RESULTS: Disease was organ and non-organ (extraprostatic extension) confined in 38 and 30 men, respectively. In organ confined cancer mean hK2 was significantly lower than in non-organ confined cancer (0.09 ng./ml., range less than 0.03 to 0.23 versus 0.30, range 0.04 to 0.94, p <0.0001). In addition, there was significantly higher free and total but not percent free PSA in non-organ than in organ confined cases. There were also statistically significant differences in hK2, free PSA and total PSA at each pathological disease stage (p <0.001, <0.01 and <0.05, respectively). Sensitivity for detecting organ confined disease was 37% at 100% specificity (correct identification of all non-organ confined cancer) using hK2 measurements compared with a sensitivity of 14% for total PSA. At a specificity of 95%, sensitivity was 40% for hK2 versus 23% for total PSA, which was a statistically significant gain in sensitivity (p <0.05). Receiver operating characteristics curves demonstrated that hK2 had the largest area under the curve, followed by the algorithm, (hK2) x (total PSA/free PSA), and total PSA (0.76, 0.75 and 0.72, respectively). However, none of area under the curve differences was statistically significant. Compared with total and free PSA hK2 testing improved the preoperative evaluation of patients who underwent radical retropubic prostatectomy due to the superior discrimination of organ from non-organ confined cancer.