Am Fam Physician. 2008;77(4):514-516
Background: When breast cancer is diagnosed unilaterally, the other breast should be evaluated for occult malignancy. Using mammography in addition to clinical examination has been shown to increase the cancer detection rate by up to 3 percent over clinical evaluation alone, and this approach is widely recommended. However, false-negative results are common and up to 10 percent of women with unilateral breast cancer eventually have cancer detected in the contralateral breast.
The role of magnetic resonance imaging (MRI) in detecting breast cancer is still uncertain, but there has been increasing evidence that it may be useful in assessing certain high-risk populations. One study on the use of MRI for screening in women with a familial or genetic predisposition to breast cancer found a 1.2 percent increase in cancer detection compared with traditional mammography. In women recently diagnosed with breast cancer, preliminary studies have suggested that MRI may also have clinical value, but there have been insufficient data to establish the negative predictive value of such testing.
The Study: The study prospectively gathered data through 25 U.S. MRI centers, ranging from academic institutions to private and community practices. To be included, patients had to have been diagnosed with unilateral breast cancer within the previous 60 days and had to have normal clinical and mammographic findings in the contralateral breast within the previous 90 days. Exclusion criteria included pregnancy, having chemotherapy or hormonal therapy for breast cancer within the preceding six months, or having a breast MRI within the preceding year.
All participants had T1-weighted MRI of the contralateral breast, with and without contrast, using a 1.5-tesla or larger magnet and a dedicated breast-surface coil. Studies were interpreted according to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) classification (see accompanying table). A BI-RADS score of 0, 4, or 5 was considered positive for a suspicious lesion, and biopsy was recommended. Patient's clinical status was followed for one year after the MRI study was performed.
| Score | Interpretation |
|---|---|
| 0 | Needs additional imaging evaluation |
| 1 | Negative |
| 2 | Benign finding—annual mammography advised |
| 3 | Probably benign—short-term follow-up advised |
| 4 | Suspicious abnormality—biopsy should be considered |
| 5 | Highly suggestive of malignancy—appropriate action should be taken |
Results: Of the 1,007 potentially eligible women, 969 were included in the analysis. The most common initial cancer diagnosis was infiltrating ductal carcinoma (58 percent), followed by ductal carcinoma in situ (20 percent), infiltrating lobular carcinoma (10 percent), and various less common tumors. Approximately 43 percent of participants were premenopausal at the time of diagnosis, and 80 percent did not have a first-degree relative with breast cancer.
Of 969 women, 135 (13.9 percent) were found to have a suspicious contralateral breast lesion on MRI. There were 121 biopsies performed, with 30 (24.8 percent) testing positive for cancer. During the follow-up period, three lesions reported as benign on MRI were subsequently found to be malignant. Estimated sensitivity of MRI for detecting potentially malignant lesions in the contralateral breast was 91 percent, and specificity was 88 percent. The positive predictive value was 21 percent and the negative predictive value was 99 percent.
There were no significant differences in sensitivity of MRI to detect suspicious lesions based on breast density, menopausal status, or the histologic features of the index cancer. Specificity was significantly greater for postmenopausal women compared with pre- and perimenopausal women (91 versus 84 percent), as were positive predictive value (31 versus 11 percent) and likelihood of malignancy of a biopsy specimen (35 versus 14 percent).
Conclusion: This study demonstrated that MRI had a 3.1 percent diagnostic yield for detecting occult contralateral breast cancer among women recently diagnosed with unilateral breast cancer. The negative predictive value of MRI for contralateral breast cancer detection was 99 percent. These findings may have clinical implications for the management of unilateral breast cancer.
editor's note: Have we reached the point where MRI should be recommended for breast cancer screening in the general population? Probably not. The cost of MRI remains prohibitive for this purpose, and studies examining its effectiveness (from a cost and a clinical perspective) in this situation are lacking.1,2 MRI may be more effective when used judiciously in specific clinical situations.
This study does offer compelling data for using MRI to screen the contralateral breast after a unilateral breast cancer is diagnosed. It may also modestly increase detection rates in women at high risk of familial breast cancer. Like any test, MRI is not perfect, as shown by the three malignant lesions missed in this study. Additionally, clinical examination and traditional mammography can still detect lesions overlooked by MRI.3 Rather than replacing clinical examination and traditional mammography, MRI will probably work best in addition to these modalities to maximize breast cancer detection in select high-risk populations.—k.t.m.