The authors report a prospective study comparing magnetic resonance imaging (MRI) with mammography for screening in women having an estimated cumulative lifetime risk of breast cancer of at least 15% because of a familial or genetic predisposition. A clinical breast examination was done every 6 months and imaging studies (with independent readings) every 12 months. The study group included 1909 eligible women 25 to 70 years of age, 358 of whom were carriers of germline mutations. A BRCA1 mutation was by far the most common pathogenic mutation.
A total of 51 tumors were found during a median follow up of 2.9 years. There were 44 invasive cancers, 6 ductal carcinomas in situ, and 1 lymphoma. An additional woman had a lobular carcinoma in situ. The overall detection rate for all breast cancers was 9.5 per 1000 woman-years at risk. Twenty-two of the 32 cancers found by magnetic resonance imaging (MRI) were not found on mammograms. Eight of 13 cancers missed by MRI were visible on mammograms. Sensitivity figures for all breast cancers were 18% for clinical breast examination, 40% for mammography, and 71% for MRI. Twenty-two of 41 cancers found by screening were detected at the first imaging screening. Specificity figures for invasive breast cancer were 98% for clinical beast examination, 95% for mammography, and 90% for MRI. The overall discriminating ability of MRI significantly exceeded that of mammography. Invasive cancers measuring 10 mm or less in diameter were significantly more frequent in women under surveillance than in control women. More than half of control women but only approximately one fifth of screened women had positive axillary lymph nodes and/or micrometastases.
These findings indicate that MRI is more sensitive than mammography for detecting beast cancers in women at increased risk because of inherited susceptibility.