Protein predicts prognosis for breast cancer patients

Protein predicts prognosis for breast cancer patients

January 14, 2000

Low levels of a protein that helps link cells together appears to be the single most effective factor for predicting which patients with early breast cancer will need chemotherapy following surgery, report researchers from the University of Chicago Medical Center in the January 15, 2000 issue of Cancer Research.

The researchers, studying women with node-negative breast cancer--the small, early tumors sought by mammography--found that by adding E-cadherin to other biomarkers that they have studied they are able to distinguish those patients with a 90 percent long-term survival from those whose survival is only 44 percent.

"We found that, for women with no affected lymph nodes, E-cadherin is the strongest prognostic factor for a poor long-term outcome," said first author Ruth Heimann, MD, PhD, assistant professor of radiation and cellular oncology at the University of Chicago. "Our anaylsis suggests that biomarkers like E-cadherin are even more important than tumor size, tumor grade, the presence of estrogen receptors or the age of the patient."

It is not the original breast cancer that kills women but the tumor's spread to other sites. Surgeons can remove the initial tumor, radiation therapy is effective in treating disease that has reached nearby lymph nodes, and chemotherapy can decrease the risk of distant metastases. But chemotherapy is not always effective and has significant side effects. Only 20 to 30 percent of women with node-negative breast cancer will develop metastatic disease. If physicians could predict who was at risk they could increase the intensity of chemotherapy for those women and reduce or even eliminate the therapy and its side effects for those not at risk.

Unfortunately, the standard tools for assessing risk of metastasis--such as the size of the tumor, its grade, the presence of estrogen receptors and the proportion of dividing cells--are "insufficient," note the authors, "for clinical decision making."

So Heimann and her colleagues have spent years looking for better "biomarkers," proteins found at varying levels in early breast cancers, which can provide clues about whether the cancer is likely to have spread.

The focal point for this search is a rare medical archive, a database of tissue samples and long-term follow-up information from 2,136 breast cancer patients treated with mastectomy, but not radiation or chemotherapy, at the University of Chicago between 1927 and 1987. The researchers found sufficient archival material to test the importance of several biomarkers, including E-cadherin, from 168 women with node-negative breast cancer.

A previous study (Cancer Research, July 1, 1998) found that two other markers could predict a good outcome. More than 90 percent of patients with either high nm23, a protein that prevents the cancer from spreading, or low angiogenesis, minimal growth of new blood vessels to supply the budding tumor, were alive an average of 14 years after treatment, compared to only 70 percent of those with low nm23 or high vessel density.

High or intermediate levels of E-cadherin add little assurance of good prognosis, but abnormally low levels of this protein indicate a substantially increased risk of metastasis.

The authors argue that the loss of E-cadherin is a later but crucially important step in the long progression from normal to highly malignant tissue. The growth of new blood vessels tends to occur quite early for most cancers; even small tumors found via mammogram frequently have a high microvessel count (MVC). However, patients with high MVC combined with high protective levels of nm23 also do quite well, 90 percent surviving 14 years.

Loss of nm23 is more significant and appears to occur slightly later. Fourteen-year disease-free survival falls to 69 percent for patients who have a high microvessel count combined with low nm23, but still have high or moderate E-cadherin levels.

The loss of E-cadherin is an additional and still later step toward malignancy. Only 44 percent of breast cancer patients with a high microvessel count, low nm23 and low E-cadherin levels are alive and disease free 14 years after surgery.

The mechanism of how E-cadherin prevents tumor spread is unclear. The protein is involved in cell-cell adhesion. It also acts to suppress the invasive character of cancer cells. As cancerous cells become more invasive they produce less E-cadherin.

"The additional information we can now derive from tumor biomarkers such as E-cadherin, nm23 and microvessel counts should begin to allow us to tailor therapy according to individual tumor types," added Samuel Hellman, MD, A.N. Pritzker Distinguished Service Professor of radiation and cellular oncology at the University of Chicago and co-author of the paper.

"Two-thirds of node-negative patients will never develop metastases--even with out chemotherapy. As we find more and more biomarkers that can predict which patients those will be, we can eliminate the unnecessary morbidity caused by treating patients who are well and perhaps increase the intensity of therapy in patients who are at high risk for metastatic disease."

Fusheng Lan and Russell McBride, both from the University of Chicago, also contributed to the study.

Low levels of E-cadherin, a protein that links cells together, appears to be the single most effective factor for predicting which patients with early breast cancer will need chemotherapy following surgery. E-cadherin levels, combined with two other biomarkers, can distinguish patients with 90% long-term survival from those with 44% survival.