Breast ImagingE2196. MR-Guided Breast Biopsies: Our Experience in Patients With Diagnosed Breast Malignancy or Lesions
Amiraian D, Patel N, McDonough M. Mayo Clinic, Jacksonville, FL
Address correspondence to D. Amiraian (email@example.com)
Objective: We retrospectively evaluated our 10-year experience with MR-guided breast biopsies, specifically in patients with previously or currently diagnosed breast malignancy or high-risk breast lesions. We sought to compare our institution’s performance with MR-guided breast biopsies to that reported in the literature, and to further discern whether MR-guided breast biopsies are truly beneficial for this population of patients.
Materials and Methods: A database search was performed to identify all patients who underwent MR-guided breast biopsies at our institution from 2003 to 2013. Two radiology residents retrospectively reviewed electronic medical records for biopsy results, breast imaging, and relevant history. Data were compiled and subsequently analyzed.
Results: Over a 10-year period, 451 consecutive MR-guided breast biopsies were performed on female patients with previously or currently diagnosed breast malignancy or high-risk pathology findings. MR-guided breast biopsies yielded a malignancy rate of 21%. An additional 18% of MR-guided breast biopsies revealed a high-risk lesion (atypical ductal hyperplasia, lobular neoplasia, papilloma, and radial scar). These are lesions that would be recommended for excision at our institution. When including all malignancies and high-risk lesions going on to excision, overall positive predictive value (PPV) for cases requiring surgery was 35%. Of the 80 high-risk lesions based on biopsy, 13 (16%) were upstaged to malignancy at excision. When these were included, the ultimate malignancy rate (or PPV3) increased to 24%. PPV3 for the biopsied lesions differed based on their location in relation to the previously diagnosed breast pathology findings. This was highest (32%) if the biopsied lesion was in the same breast, versus 18% if the biopsied lesion was in the contralateral breast. PPV3 was 24% if there was previously diagnosed pathology in both breasts. It should be noted that our PPVs are based solely on biopsies performed using MR guidance, even though a significant number of lesions seen on MRI and recommended for biopsy were performed under other modalities.
Conclusion: Currently, there is limited literature and a wide range of reported data regarding the utility of MR-guided breast biopsies. MRI is extremely sensitive in detecting breast malignancy that may otherwise be occult on other breast imaging modalities. However, MRI lacks specificity in some cases due to overlapping characteristics of benign and malignant lesions, often necessitating MR-guided breast biopsies. Our results add to the growing body of literature regarding the role of MR-guided breast biopsies particularly in patients with previously or currently diagnosed breast pathology. Such patients make up a substantial number of those undergoing diagnostic or screening breast MRI. Our PPVs fall within the range of those previously reported; however, our PPVs would likely be even higher since we only included patients who had MR-guided breast biopsies rather than all biopsies by other modalities. This study includes one of the larger sample sizes to date of women undergoing MR-guided breast biopsies. Our findings from this large cohort support and encourage the use of MR-guided breast biopsies in this patient population to provide additional information for surgical management.