Prognostic Factor Analysis for Breast Cancer Using Gene Expression Profiles

Abstract

The survival of patients with breast cancer is highly sporadic, from a few months to more than 15 years. Recently, the large-scale gene expression profiling of tumors has been used as a promising means of predicting prognosis factors. In this study, we used large-scale gene expression datasets of tumors to identify prognostic factors in breast cancer. We conducted survival log-rank tests and used an unsupervised clustering method to find individual genes associated with worse survival rates. For each gene, log-rank test was performed between two groups of patients who had high and low gene expression levels. After the genes which were associated with poor survival were separated into high- or low-expressed gene sets, we clustered genes to find co-expressed or reciprocally expressed gene sets through maximal clique and bi-clique algorithms. As a result, four prognostic gene sets were constructed among 24,924 genes. Based on the ratio of high and low expression levels, our four scores predicted the worse survival rates in three independent data sets. In addition, we found that cancer patient with poor prognosis, i.e., triple-negative cancer types, HER2-enriched, and high-grade patients, had higher scores than those with other types of breast cancer. In conclusion, based on a gene expression analysis, we suggest that our well-defined method of the prediction of survival outcome may be useful for developing prognostic factors in breast cancer.