Second-look US examination of MR-detected breast lesions.

PURPOSE

To review our institutional experience in using second-look ultrasound (SLUS) to identify breast lesions initially detected on MR imaging that were indeterminate or suspicious for malignancy.

METHODS

This Health Insurance Portability and Accountability Act compliant retrospective review included 83 women with 131 lesions initially identified as indeterminate or suspicious for malignancy on MR imaging from February 1, 2008 through July 31, 2009. An SLUS correlate was confirmed on the basis of concordant location, size, and morphologic features. The detection rate of SLUS was determined. Patients' demographics, lesion size, and MR imaging morphologic features (focus, mass, non-masslike) were reviewed to identify which factors led to successful detection on SLUS. Likelihood ratio χ(2) tests were used for statistical analysis.

RESULTS

SLUS correlates were found for 88 of 131 (67%) lesions initially detected on MR imaging; 27 of 88 (31%) were malignant. SLUS detected foci (67%) and masses (73%) more frequently than it did non-masslike lesions (54%). The detection rate of SLUS was independent of lesion size on MR imaging. Malignant lesions were not more likely than benign lesions to be detected on SLUS (61% versus 70%).

CONCLUSIONS

SLUS provides value in the clinical workup of breast lesions that are indeterminate or suspicious for malignancy. It identified two thirds of the MR-detected lesions evaluated and permitted performance of US-guided needle biopsy on 70 of 88lesions. The likelihood of finding MR-detected lesions on SLUS was significantly higher for foci and masses than for non-masslike lesions (P < 0.05).