Malignancy Risk High With Indeterminate Breast Lesions

From Medscape Medical News
Norra MacReady

May 7, 2010 (San Diego, California) — Indeterminate breast lesions in high-risk women have a relatively high probability of being malignant and should be treated aggressively, investigators reported here at the American Roentgen Ray Society 2010 Annual Meeting.

Of 59 indeterminate lesions identified on magnetic resonance (MR) mammography in 55 women, 13 (22%) proved to be malignant on follow-up MR, mammography, or ultrasound performed 6 months after the initial MR study, said lead author Martin Korzeniowski, MD, from McMaster University in Hamilton, Ontario.

Biopsy or surgical intervention was performed whenever appropriate. The patients all had breast cancer or were deemed to be high risk, either because of personal or family medical history or genetic predisposition.

These findings suggest that "malignant lesions in women with a high risk of breast cancer may present atypically, with an indeterminate morphology or kinetic pattern, and may require more aggressive workup," said Dr. Korzeniowski.

The patients were drawn retrospectively from a database of 727 consecutive magnetic resonance imaging (MRI) scans performed at McMaster University between January 2007 and December 2008. Lesions were classified according to the Breast Imaging Reporting and Data System (BIRADS), developed by the American College of Radiology. Each lesion received a score ranging from 0 (incomplete examination) to 6 (known, biopsy-proven malignancy).

In this study, lesions were considered indeterminate if they could not be definitively classified as suspicious for malignancy but had suspicious abnormalities with a reasonable probability of being malignant (BIRADS 4), or if there was more than a 95% chance that they were malignant (BIRADS 5).

The 59 lesions that met those criteria were followed up within 6 months of the original MRI examination. Of the 13 malignancies, 9 were infiltrating ductal carcinomas, 2 were ductal carcinomas in situ, and 2 were metastatic lymph nodes. The remaining 46 lesions were benign.

The "substantial" cancer yield in this study suggests that follow-up examinations for indeterminate lesions should be performed sooner than is current practice, Dr. Korzeniowski told meeting attendees.

"The rate of cancer these authors found is much higher than other studies have suggested," Constance Lehman, MD, professor of radiology at the University of Washington and director of imaging at the Seattle Cancer Care Alliance, said in an interview with Medscape Radiology. At the University of Washington and other centers, less than 2% of patients with indeterminate lesions turn out to have malignant disease, she said.

"My guess is that their methods of determining if a lesion is indeterminate are different" than those used at other institutions, said Dr. Lehman, who was not involved in this research. Usually, indeterminate lesions must meet specific criteria involving morphology, margins, and enhancement pattern, and are assigned a BIRADS score of 3. Those lesions are considered "probably benign" because they have more than a 98% chance of being benign, she told Medscape Radiology.

Still, Dr. Lehman said, "this is a very interesting study. It is really important for us to continue to evaluate breast MRIs in high-risk women. This research will stimulate more debate."

Dr. Korzeniowski and Dr. Lehman have disclosed no relevant financial relationships.

American Roentgen Ray Society (ARRS) 2010 Annual Meeting: Abstract 018. Presented May 3, 2010.