Biomarker May Be an Early Predictor of Advanced Breast Cancer

CINCINNATI—Researchers have identified a molecule that may be more accurate than existing biological signposts used to predict which breast cancers will develop into advanced forms of the disease.

Detailed Oct. 24, 2007, in an early online edition of the International Journal of Cancer, the discovery could one day influence therapy decisions and prevent patients from unnecessarily undergoing aggressive cancer treatments.

When diagnosing breast cancer, pathologists currently look for elevated levels of three standard molecules known to make tumors grow in the breast. These molecules—

estrogen receptor (ER), progesterone receptor (PR) and HER2—are used as “biomarkers” for diagnosis and individually detect only a fraction of breast cancers.  

“The problem with these biomarkers is that many of them are present at some level in the normal breast,” says Georg Weber, MD, PhD, lead investigator of the new study and associate professor of pharmacy at the University of Cincinnati. “In addition, they are surface molecules that support growth so they are not necessarily a good predictor of tumor metastasis.”

Weber and his team have identified a molecule, osteopontin-c, that is absent from the normal breast and appears to more accurately predict breast cancer that will become metastatic and spread to distant organs from the original tumor site.

In normal levels, osteopontin is a protein used by the immune system to help cells move and migrate. There are three forms of osteopontin—a, b and c—which are formed by splicing, or “cutting and pasting,” ribonucleic acid (RNA) molecules to make variations of the original gene. Osteopontin-a is the normal form that helps with immune functions. Little is known about osteopontin-b, but if present its levels are very low. Osteopontin-c is the molecule Weber and his team discovered is a good biomarker of breast cancer.

In a two-year evaluation of 178 breast tumors, normal and abnormal tissue samples, they found that osteopontin-c was present in 78 percent of cancers and in 36 percent of the surrounding tissues. It was not detected at all in normal tissues.

In 56 breast cancers, 20 were positive and 36 were negative for estrogen receptor, 19 were positive and 37 were negative for progesterone receptor, and 26 were positive for HER2 with 30 negative.

“Osteopontin-c was present in a substantially higher number of breast cancers than the three biomarkers traditionally used to diagnose breast cancer,” says Weber. “We also found that the cancers containing osteopontin-c correlated with a higher tumor grade, meaning they were more likely to become aggressive cancer.”

“If we know that this molecule is not present in a patient with breast cancer, it’s more likely that we can treat them with conservative therapy rather than breast surgery, hormone therapy or chemotherapy because we know it’s less likely to metastasize,” he adds. “On the other hand, if we know that a patient has this molecule early in their diagnosis, we can treat it more aggressively because we know their cancer is likely to become invasive.”

The study was funded by grants from U.S. Department of Defense Breast Cancer Program and the UC cancer research program. The tissue procurement was supported by a grant from the National Institute of Health.

Collaborators include UC’s Mana Mirza and Elizabeth Shaughnessy, MD, John Hurley, Kristie Vanpatten and Gary Pestano of Ventana Medical Systems in Tucson, Ariz., as well as Bin He, MD, of the North Shore Medical Center in, Salem, Mass.