Breast Cancer Genomics
Breast cancer continues to exact a major toll in the US: it is the most frequently diagnosed cancer in women and the second leading cause of cancer mortality. By now, many of the other grim breast cancer statistics are familiar: one million diagnoses a year worldwide (200,000 in the US); 40,000 deaths per year; one-fifth of all deaths in women aged 40-50; $60-$100 billion in direct and indirect costs every year. The most well-known and perhaps most sobering figure regarding the disease: one in eight women in the US will develop breast cancer during her lifetime.
What do we know about cancer that might enable us to fight it? We know that it is characterized by genomic instability-tumor cells divide like mad and in that process their genomes are rarely transmitted faithfully. Breast cancer is no exception: any ten women with the disease will have ten different tumors-their cancers will be of different sizes, some will be more aggressive than others, and those tumors will each express their own peculiar set of genes (albeit with some overlap). At Duke, genome scientists and clinicians have begun to make use of breast cancer' s heterogeneity to make predictions and guide treatment decisions. By examining the expression patterns of collections of genes that tend to be turned off or on together in an array of tumor samples and the clinical outcomes of patients who developed those tumors, researchers are now able to predict who is most likely to experience recurrent breast cancer and who is likely to remain cancer-free. This information has practical implications. If her risk is low, a patient may not want to endure the hardships of chemotherapy; if her risk is high, she may choose a more aggressive course of treatment.