Update on molecular signatures in breast cancer

Routine clinical and pathological criteria cannot precisely determine the individual prognosis of breast cancer patients, leading to the "over-treatment" of many of them. Three intermediate- or high-throughput, gene-expression-based tools have been developed to make up for this lack of precision. After validation studies on independent series, they are now commercially available. Trials are in progress to assess their real clinical benefits in terms of the determination of prognosis and an enhanced-therapeutic-decision process. These first-generation molecular signatures rely mostly on estrogen receptors, HER2, and proliferation. Would a prognostic algorithm, based on a combination of a well-calibrated determination by immunohistochemistry of these markers, give as much information? This question needs to be addressed.