Multivariate analysis of plasma hormones in patients with metastatic prostate cancer receiving combined LHRH-analog and antiandrogen therapy.

The following hormones, the plasma protein SHBG, and the tumor markers prostatic acid phosphatase (PAP) and prostate-specific antigen (PSA) were assayed at 18 time-points over 1 month during a double-blind randomized study in 36 stage D2 cancer patients receiving either "buserelin+placebo" or "buserelin+nilutamide (Anandron)": LH, FSH, estradiol, testosterone, dihydrotestosterone, androstenedione, 5-androstene-3,17 beta-diol, dehydroepiandrosterone and its sulfate, cortisol, 17 alpha-hydroxyprogesterone, pregnenolone, 17 alpha-hydroxypregnenolone, and 3 alpha-androstanediol glucuronide. Multivariate analysis of the treatment values (over 10,000 assays) by two complementary methods, correspondence factorial analysis (CFA) and the minimum spanning tree (MST) method, identified those variables within the pathways of androgen metabolism that were correlated over time and, in a comparison of the two treatment groups, identified the enzyme targets of nilutamide action in humans. Whereas nilutamide tended to decrease androstenedione slightly, it affected no other variables including cortisol, except for pregnenolone, 17 alpha-OH-pregnenolone, and 17 alpha-OH-progesterone, which were increased. These increases are indicative of weak inhibition of C17,20 lyase by nilutamide, but, according to the multivariate analysis, are insufficient to account for the more marked and rapid fall in PAP and PSA noted on addition of nilutamide to buserelin that must therefore be explained by another mechanism such as androgen receptor blockade by nilutamide.