Adrienne Rebecca Newburg MD, Presenter: Nothing to Disclose

Chloe Muy-Chou Chhor MD, Abstract Co-Author: Nothing to Disclose

Jiyon Lee MD, Abstract Co-Author: Nothing to Disclose

Samantha Lynn Heller MD, PhD, Abstract Co-Author: Nothing to Disclose

Hildegard B. Toth MD, Abstract Co-Author: Nothing to Disclose

Linda Moy MD, Abstract Co-Author: Nothing to Disclose

Prior studies showed the likelihood of identifying an ultrasound (US) correlate for an MRI-detected abnormality depends on lesion type. NME was less likely to be seen on US compared to a mass or focus. Targeted second-look US may result in prolonged work-up time, added expense, and false reassurance in the setting of a negative US. Our study was performed to determine the utility of targeted US and to determine how often a MRI-US discordant lesion was found.

An IRB-approved retrospective review was performed for breast MRI examinations performed from 2005-2008. Data regarding patient demographics, MRI findings and subsequent sonographic and pathologic results were recorded. Of 2,222 breast MRI exams, 70 (3.2%) NME lesions were identified for which targeted US was recommended. An additional 85 NME lesions went directly to an MRI biopsy because the interpreting radiologist felt it unlikely that an US correlate would be seen. The rate of subsequent malignancy was analyzed.

Targeted US was performed in 59 of 70 (84%) women. In the remaining 11 (16%) cases, targeted US was not performed because 5 women underwent mastectomy or had metastatic disease. MRI-guided biopsy was pursued directly in 2 women and 4 women did not undergo further imaging at our institution. Mean age was 46.7 years, range was 25 to 99 years. In 14 (24%) of 59 sonograms, an US correlate was seen. An US-guided biopsy was performed in 7 (50%) of 14 cases. None yielded cancer. One of 7 (14%) yielded papillomas which were subsequently excised. One of 7 (14%) yielded atypia. Three benign biopsies were discordant with the MRI findings. At subsequent MRI biopsy, one lesion was an invasive ductal carcinoma (IDC). Forty-five of 59 (76.3%) cases had no US correlate; 15 proceeded to MRI-guided biopsy. Two (13%) yielded cancer, 1 IDC and 1 DCIS. An additional 2 (13.3%) cases demonstrated atypia/ADH. In the 11 remaining cases, pathology was benign. The cancer yield for the 85 NME lesions that went directly to MRI biopsy was 12% (12/85); 2 were IDC and 10 were DCIS.

The yield for detecting an US correlate for an MRI-detected NME is low (24%) with no detection of malignancy.

Confident MRI-US correlation for an MRI-detected abnormality can be challenging. It may be advisable to forego targeted US and proceed directly to MRI-guided core biopsy.

Newburg, A, Chhor, C, Lee, J, Heller, S, Toth, H, Moy, L, Utility of Targeted Ultrasound in the Evaluation of Breast MRI-detected Non-mass Enhancement (NME).  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13021233.html