Breast magnetic resonance imaging (MRI) is routinely used as a problem-solving tool, but its benefit for this indication remains unclear. The records of 3001 consecutive breast MR examinations between January 1, 2003 and June 6, 2007 were reviewed to identify all those performed for the clinical indication of problem solving. Details of clinical presentation, mammography and ultrasound (US) findings, follow-up recommendations, and pathology outcomes were recorded. Benign versus malignant outcomes were determined by biopsy or 12 months of follow-up imaging and linkage with the regional tumor registry. Problem solving was the clinical indication for 204 of 3001 (7%) of all examinations. Forty-two of 204 examinations (21%) had suspicious or highly suspicious MRI assessments with recommendation for biopsy and 62 of 204 (79%) examinations were assessed as negative, benign, or probably benign. Thirty-six biopsies were performed based on MRI findings and 14 cancers were diagnosed. Biopsy was indicated for 11 of 14 (79%) cancers based on suspicious mammographic or US findings identified prior to MRI. One incidental cancer was detected by MRI alone in a patient at high risk for breast cancer, and two cancers were detected in patients with suspicious nipple discharge and negative mammogram and US. A single false-negative MRI occurred in a patient whose evaluation for a palpable lump prompted biopsy. Problem-solving breast MRI rarely identifies otherwise occult cancer and can be falsely negative in patients with suspicious findings on mammogram and US. Until the benefits and risks of problem-solving MRI are clarified, it should be used judiciously.
© 2011 Wiley Periodicals, Inc.