Using a Mathematical Method to Predict Breast Cancer Recurrence

Patients diagnosed with breast cancer are normally treated with surgery, radiation therapy, chemotherapy and possibly hormone therapy, although many will not relapse even without chemotherapy.

Chair of the Department Applied and Computational Mathematics and Statistics Steven Buechler has developed a way to predict the recurrence of breast cancer using network theory and microarray data. A provisional patent on the method has been filed. The tool, taking advantage of research into the molecular biology of cancer cells, can guide individual treatment plans, perhaps avoiding chemotherapy when it provides little benefit.

Buechler’s work provides a new way of selecting genes for a test that predicts the recurrence rate of breast cancer. His approach is more stable and more accurate than previous methods. It uses the expression level of only four genes to predict relapse or define disease subtypes, making more targeted treatment possible. The method, which can be applied to all estrogen receptor positive tumors, has been validated in six independent datasets with different clinical traits and technical formats.

The method can also be applied to other types of cancer and may spare many patients from expensive, unnecessary chemotherapy and its associated side effects.