Back to 2016 Annual Meeting Posters
Breast MRI: a retrospective review of the utility and pitfalls in newly diagnosed breast cancer
*Laura C Lamb1,2, Kristen Zarfos2, *Bethany Carr2
1University of Connecticut, Avon, CT;2The Hospital of Central Connecticut, New Britain, CT
Objective: To determine whether a preoperative breast MRI can prevent women from returning to the operating room for additional surgical excisions.
Design: Single surgeon, single institution, retrospective analysis of women with newly diagnosed breast cancer.
Setting: Breast cancer accounts for 25% of female cancers. Breast cancer accounts for 450,000 deaths annually worldwide.
Patients: 96 patients had breast MRIs from December 2013 through December 2014.
Interventions: Each patient with newly diagnosed breast cancer underwent a preoperative bilateral breast MRI.
Main Outcome Measures: 1.Whether preoperative MRIs would change the planned procedure 2. Minimizing unnecessary biopsies3. Minimizing re-operations
Results: MRI’s correlation to pathology, the gold standard was 64% overall, while the correlation for ultrasound was 43% and 44% for mammography. As breast density increased, MRI’s correlation to pathology increased. For women with breast density of 25-50% there was a 27% correlation while for women with 51-75% density there was a 92% correlation. True positives were higher with the use of MRI with 42% of breast biopsies on MRI being positive for carcinoma. 38% patients had a change in operative plan based on their MRI findings.
Conclusions: MRI is a sensitive imaging modality which is most closely correlated to pathology. This led to 38% of patients having a change in their operative plan. Breast biopsies were more likely to result in findings of carcinoma when aided by MRI as well. We found that 42% of our breast biopsies of lesions found on MRI were positive as opposed to 25% nationally. The >50% breast density population in particular seems to benefit from the increased sensitivity of MRI as compared to ultrasound and mammography.
Back to 2016 Annual Meeting Posters
