Under the Microscope: A Personalized Approach to Treating Breast Cancer

When facing breast cancer, the most pressing question on everyone’s mind is which therapies will work. The flip side of that question, though, is which ones won’t. This has become an increasingly important question as more people survive breast cancer, because minimizing patients’ exposure to unnecessary treatments and side effects is one important way to ensure survivors enjoy the healthiest and longest life possible.

With support from a V Foundation Translational Grant, Christina Curtis, Ph.D., assistant professor at the Stanford Cancer Institute, is working on a way to use gene expression levels to predict which breast cancer patients would benefit most from anthracyclines, which are among the oldest and most widely used types of chemotherapy drug. Although effective against breast cancer, anthracyclines don’t work for everyone and bring risks for serious cardiovascular side effects and secondary cancers.

“Determining who should receive anthracyclines is an old question that has remained unsolved,” said Curtis. “By using powerful computational approaches to assimilate the wealth of molecular data available from thousands of cancer patients, we are taking a new approach to address this question.”

Using big data

Curtis and her research team examined the expression levels of genes known as chromatin regulators in breast cancers to define a network describing their connectivity. Chromatin regulators play an important role in gene expression by opening and closing the nearly two meters of DNA packed tightly inside each cell in the human body. Multiple chromatin regulators work together in a protein complex, and their connectivity is crucial to understanding their function.

“Figuring out differences in how the nearly 400 chromatin regulatory genes were configured in breast cancer versus normal breast tissue points to ways that their behavior goes awry in disease,” said Curtis. “It’s important to consider these genes together because they have different functions and influence many other genes.”

With this information, the researchers looked at chromatin regulatory gene expression in human breast cancer cell lines from 10 different labs around the world and data from more than 1,000 women with early-stage breast cancer. Because the data they used was publicly available, the researchers didn’t have to conduct a new study to examine how chromatin regulators relate to anthracycline response.

“It was important to have data from a large number of patients because there is a high degree of patient-to-patient variability,” said Curtis. “This also allowed us to study distinct clinical subgroups of breast cancer instead of grouping them all together.”

Informing treatment decisions

The researchers discovered expression levels of a particular group of chromatin regulators dictated the response to anthracyclines but not other cancer drugs. They used the data to build statistical models that allow them to use a patient’s molecular profile to predict whether they will respond to anthracyclines.

“It is particularly encouraging that these findings held true across all major subgroups of breast cancer,” said Curtis. “This points to a common mechanism that might also be present in other types of cancer treated with anthracyclines.”

Although further validation is needed before this signature could be used to determine treatment plans, the researchers are working toward a diagnostic test that might be available in several years that uses the patient’s molecular profile to determine whether they should receive anthracycline treatment.

“We appreciate this grant, which has helped us move our research forward rapidly toward the ultimate goal of translating these findings to the clinic to benefit patients,” said Curtis.