Fabian Carol J, Kimler Bruce F
Department of Internal Medicine, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA.
J Steroid Biochem Mol Biol. 2007 Aug-Sep;106(1-5):31-9. doi: 10.1016/j.jsbmb.2007.05.022. Epub 2007 May 24.
Epidemiologic models used for cancer risk prediction, such as the Gail model, are validated for populations undergoing regular screening but often have suboptimal individual predictive accuracy. Risk biomarkers may be employed to improve predictive accuracy based on the Gail or other epidemiologic models and, to the extent that they are reversible, may be used to assess response in phase I-II prevention trials. Risk biomarkers used as intermediate response endpoints include high mammographic breast density, intra-epithelial neoplasia, and cytomorphology with associated molecular markers such as Ki-67. At the present time these biomarkers may not be used to predict or monitor individual response to standard prevention interventions but are used in early phase clinical trials as preliminary indicators of efficacy.
