Identification and meta‐analysis of a small gene expression signature for the diagnosis of estrogen receptor status in invasive ductal breast cancer

In breast cancer, the determination of estrogen receptor (ER) expression is crucial for the decision on therapeutic strategies. Current ER expression analysis is based on immunohistochemical (IHC) staining of ER on formalin fixed tissue sections. However, low levels of ER expression frequently escape detection because of varying sensitivities of routine histopathological laboratories. Moreover, in estimating ER by IHC the receptor protein only is tested instead of the complex underlying ER pathway, which reflects its biological activity. To overcome this limitation, we have used the microarray technology to study 56 samples of invasive ductal carcinoma. We infer a robust and reliable signature of 10 genes, which is associated with ER expression and presumably therapeutically relevant biological processes. In a meta-analysis, the signature was tested on 3 further independent microarray gene expression data sets, covering different laboratories, array platforms, and clinics. The classification based on the signature showed a very low misclassification rate. In summary, the expression of few genes is sufficient to determine ER status. Future decisions on antiestrogen based therapy in breast cancer could be based on this signature rather than on immunostaining alone.