Jaime Lynn Geisel MD, Abstract Co-Author: Nothing to Disclose

Shelley Lynn Goodstine MD, Presenter: Nothing to Disclose

Madhavi Raghu MD, Abstract Co-Author: Nothing to Disclose

Jeffrey C. Weinreb MD, Abstract Co-Author: Research Support, Bayer AG Consultant, General Electric Company Consultant, Bayer AG Consultant, Bracco Group

Daniel Cornfeld MD, Abstract Co-Author: Stockholder, General Electric Company

M. Lisa Varipapa BA, Abstract Co-Author: Nothing to Disclose

Liane Elizabeth Philpotts MD, Abstract Co-Author: Nothing to Disclose

To analyze the outcome of breast MRI cases for which biopsy was recommended in an attempt to identify findings or features that have a lower likelihood of malignancy, in order to reduce false positives.

A retrospective review of the breast imaging section database from January 2007- December 2008 was performed to identify all breast MRI exams in which a biopsy was recommended (BI-RADS 4 or 5). These cases were reviewed to assess the size of the lesion and the MR characteristics. The method of biopsy (ultrasound, MR, or other) was determined. The pathological results were correlated with the MR findings. True positive cases (cancers) were reviewed and compared to false positive cases to assess for any defining features.

Of 275 breast MR biopsy recommendations, 42 were ultimately canceled (15%) due to non-reproduction of the lesion. 233 biopsies were performed - 154 by MR (66%), 69 by US (30%), 4 by stereotactic (2%), and 6 (2%) by other. Masses accounted for 62% (145/233) and non-mass-like enhancement 38% (87/233) of biopsied lesions. Malignancy was found in 29 cases (12%), with high-risk findings in 48 (21%) and benign in 156 (67%). Similar rates of malignancy were found in masses 12.4% (17/137) and NMLE 12.5% (12/96). For lesions with an US correlate, 20% were malignant, versus no US correlate where only 9% were malignant. Almost half the cancers (14/29, 48%) were identified and sampled with US, masses (71%) more commonly than NMLE (17%). 78 lesions were diagnosed by MR core after a negative US and 7 were malignant (2 masses and 5 NMLE). Cancer size ranged from 3-20mm (mean 9mm), with 6 cases (21%) measuring 5mm or less.

Malignancies are found in MR lesions at fairly similar rates relative to mass or NMLE. Lesions with an US correlate have a higher chance of malignancy but lack of an US finding does not exclude cancer. Small size should not preclude biopsy as many cancers found were 5mm or less. There appears to be a great deal of overlap between benign and malignant findings such that reducing the false positives is difficult.

There appears to be a great deal of overlap between benign and malignant findings such that reducing the false positives is difficult.

Geisel, J, Goodstine, S, Raghu, M, Weinreb, J, Cornfeld, D, Varipapa, M, Philpotts, L, Analysis of Biopsy Recommendations from Breast MRI Exams: Attempts to Reduce False Positives.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8009033.html