Mechanisms of Disease: biomarkers and molecular targets from microarray gene expression studies in prostate cancer.

Molecular biomarkers can serve as useful diagnostic markers, as prognostic markers for predicting clinical behavior, or as targets for new therapeutic strategies. Application of expression microarray technology, which allows the expression of all or most of the genes in the human genome to be analyzed simultaneously, has dramatically enhanced the discovery of prostate cancer biomarkers. The diagnostic markers identified include AMACR (alpha-methylacyl CoA racemase), a protein that has already been translated into clinical use as an aid in distinguishing prostate cancer from benign disease. Individual genes, such as the polycomb gene EZH2 whose expression indicates poor survival, have been identified. The power of microarray technology is that it has allowed the identification of gene signatures (each composed of multiple genes) that might provide improved prediction of clinical outcomes in human prostate cancer. The development of a new method for analyzing expression microarray data, called COPA, has led to the discovery of TMPRSS2-ERG gene fusion involvement in the development of prostate cancer, while expression analysis of castration-resistant prostate cancer has suggested the use of novel therapeutic approaches for advanced disease. Despite these successes, there are limitations in the application of microarray technology to prostate cancer; for example, unlike with other cancers, this approach has failed to provide a consistent unsupervised classification of the disease. Overcoming the reasons for these failures represents a major challenge for future research endeavors.