Medical Imaging News

January 24, 2007

Tool Helps Predict Chance of Breast Cancer Returning

Soon, radiation oncologists will have access to one more tool for use in treating and counseling their patients. A computer program developed by clinicians at Tufts-New England Medical Center in Boston promises to predict the risk of breast cancer returning in the same breast over a 10-year period in women who have had a lumpectomy.

The solution runs an algorithm that is based on the risk factors that have been reported in multiple trials, looking at which patients are at the highest risk for recurrence. Physicians are able to enter information about seven different factors and compute an estimate of that patient's risk of the cancer returning, both with and without radiation therapy. Those factors include the age of the patient at the time of treatment, the size and grade of the cancer, if lymphatic vessels are affected, and the use of chemotherapy or hormone therapy.

"There clearly is a benefit from radiation therapy for almost every patient, and it really is just the magnitude of the benefit that differs," said Mona Sanghani, MD, lead author of the study and an oncologist at Tufts-New England Medical Center. "This can be a useful tool for radiation oncologists when counseling patients about what they want to do. If the patient is very adverse to radiation therapy, it may help them in making a decision one way or the other."

Although it holds much promise, the predictive tool must be validated by independent clinical data before it can be employed by physicians.

"It was developed based on the literature, but the next phase is to test it against the breast-cancer database to determine the accuracy of it," Sanghani said, adding that the software is designed as another tool for physician use and not for patients to evaluate themselves. "It obviously can't replace clinical intuition and clinical judgment, but I think it will be an important supplement."

Sanghani believes that the final version will be available for use within a year. Her goal for the software is to make it available to physicians free of charge through the Tufts-New England Medical Center Web site.

"There are other algorithms like this based out of [Memorial] Sloan-Kettering [Cancer Center] and other places—but they look at different questions," she said. "This program focuses on a unique question that no algorithm has yet addressed."

The abstract, titled "Predicting the Risk of Local Recurrence in Patients with Breast Cancer: An Approach to a New Computer Based Predictive Tool," was presented at the 2006 Annual Meeting of the American Society for Therapeutic Radiology and Oncology.

Read more Medical Imaging News from our archives