The study (The Italian Trial for Breast MR in Multifocal/Multicentric Cancer, promoted by The Italian Society of Medical Radiology) included 90 women who had been diagnosed with breast cancer. Some of the women had cancer detected in both breasts. Each of the women had a mammogram and an MR examination to determine if there were additional cancers that had not been initially detected, said Francesco Sardanelli, MD, lead author of the study, chief of diagnostic imaging at the Istituto Policlinico San Donato and professor at the Postgraduation School of Radiagnostic at the University of Milan, Italy. Previous studies have shown that as many as 42% of women diagnosed with breast cancer in one breast have at least one other malignant lesion in that same breast, Dr. Sardanelli said. "We need to identify as many as possible malignant areas in the patient's breast, so we and she can determine what is the best choice of treatment," he added.
Mammography detected 124 malignant lesions, while MR detected 152, Dr. Sardanelli said. All of the women had mastectomies. Their breasts were then examined by a pathologist. The pathologist found 188 malignancies, he said. While MR did not find all of the malignancies, it did find more than mammography, said Dr. Sardanelli. In addition, "the malignancies that were missed by mammography were significantly larger and more aggressive (70% were invasive cancers) overall than the malignant areas that were missed by MR," said Dr. Sardanelli. MR imaging was even more effective for women with dense breasts, noted Dr. Sardanelli.
If MR and mammography were combined, there was no real difference in the detection rate, said Dr. Sardanelli. "It is improbable that a malignant lesion would be detected with mammography and not detected by MR," he said.
This is the first study to compare MR and mammography with whole breast pathology results, said Dr. Sardanelli. Because the whole breast was examined under the microscope, "we were able to determine exactly how many other malignancies -- even the smallest malignancies -- were actually in each breast," he said. This made the comparison of the tests very reliable, he said.
"We are aware that the clinical significance of using MR to detect malignant areas is a matter for debate. If combined with radiation therapy, breast-conserving surgery gives survival rates not significantly different from those obtained with mastectomy. Randomized studies comparing patients with or without pretreatment MRI would be able to clarify the problem, but they are really difficult to plan and perform. On the basis of our study, we suggest a pretreatment dynamic MRI examination at least in patients with nonfatty breasts," Dr. Sardanelli said. "Moreover, we need complete cooperation with surgeons. We should try to avoid performing mastectomies on patients who could have a conservative intervention with the same outcome. We could downsize the aggressive approach frequently used today on patients with multifocal breast cancer. If MRI demonstrates that multifocality is confined in a quadrant, we can suggest a breast-conserving surgical approach," he said.
The study appears in the October 2004 issue of the American Journal of Roentgenology.
Journal
American Journal of Roentgenology