[tousled1]

It’s been known for some time that breast cancer is not just one disease, but a grouping of several different subtypes, including those that overexpress the protein HER2 and those that are fueled from hormones. Building on that knowledge, recently developed technology now allows for a broad population-based molecular analysis of the disease, which promises to give oncologists the ability to choose the best treatment for each individual patient based on a tumor’s genomic expression.

Genomic expression data from breast cancer has already identified four distinct subtypes (luminal A and luminal B, which are both estrogen receptor-positive but differ in prognosis; HER2-positive; and basal type, which are negative for both hormone and HER2). Each subtype responds differently to each type of drug.

In a presentation on the genomic approaches to identifying a specific breast cancer subset and choosing an appropriate treatment, Joe Gray, PhD, from the Lawrence Berkeley National Laboratory, presented data from a multi-institutional effort to identify specific markers of each subset.

Dr. Gray and colleagues were able to expose a number of cell culture models to an even larger number of drug agents which could slow down or stop cell growth. By correlating the genomic characteristics of each specific cell model to how they react to a variety of these drug agents, scientists were able to map out what molecular variables were associated with a response to each agent.

To translate this research into the clinic, markers are selected, such as HER2. A test was created and confirmed by Dr. Gray’s group and compared with the more standard test using FISH (fluorescent in vitro hybridization). Both tests used data from a study looking at Taxol with or without Tykerb (lapatinib), a HER2-targeted agent, in newly diagnosed metastatic breast cancer.

When the two tests were compared, both Dr. Gray’s genomic test as well as a predictive HER2-associated 6-gene response signature developed by his group, were as sensitive as HER2-FISH in predicting response to Tykerb. —Zach Moore, PhD