Comparability of prostate trials

The present overview of advanced prostate cancer required the identification of randomized clinical trials studying the question of maximal androgen blockade versus the classic castration therapy. The heterogeneity of the trials concerned the type of castration (surgical or chemical) and the type of antiandrogen (flutamide, Anandron, or cyproterone acetate) added to castration. This paper reviews the different types of heterogeneity that might exist among trials that are involved in the overview: study design, randomization procedure, treatment evaluation, statistical evaluation, and data maturity. In order to overcome these various types of heterogeneity and to compare like with like, the treatment comparison should be stratified a posteriori by question (i.e., type of castration or type of anti-androgen studied) and by study. In this way, one may draw valid conclusions. Of course, those trials with a larger number of patients and a longer follow-up will contribute more to the overview's results.