Gene fusion subtypes in prostate carcinoma have clinical implications

Aug 3 2007

Building on an earlier study that used Oncomine to identify ETS-fusion proteins in prostate cancer (Science, Oct 28 2005), Tomlins et al. report the identification of four additional classes of ETS rearrangements in prostate cancer, each of which activates ETV1 by a different genetic mechanism (Nature, Aug 2 2007).

Tomlins et al. queried the Oncomine database to determine the tissue specificity of three of the new fusion gene partners, demonstrating that two new partners are prostate cancer specific while the third partner is a generally over-expressed 'house-keeping' gene. In addition, Oncomine Concepts Map was applied to test for similarities between the set of genes regulated by ETV1 with sets of genes related to other biological concepts. This analysis revealed a clear association between ETV1 activation and cellular invasion, a hallmark of cancer.

Oncomine, a dynamic repository that combines rich data, sophisticated analysis, and responsive user interface, empowers oncology research by bringing data analysis tools to a large cancer gene expression database, which includes data from 20,000+ microarray experiments curated from 300+ independent studies. Oncomine Concepts Map, an extension of Oncomine, is a compilation of nearly 8,000 Oncomine cancer gene signatures together with 11,500 gene, protein, drug, and pathway signatures collected from the literature and other public sources.

"This study marks an important advance in our understanding of the molecular basis of prostate cancer. Oncomine again proved invaluable in examining the expression of genes in cancer, and in this case aided in the characterization of new sub-types of prostate cancer," said Dan Rhodes, Ph.D., co-founder and Chief Science Officer of Compendia Bioscience. "The authors went on to show that one of the new subtypes is unlikely to respond to conventional anti-androgen therapy, and another may be adversely affected by standard therapies. This strongly suggests that using this type of molecular information to distinguish patient populations will have important clinical implications."