Efficacy of MRI-Directed Breast Ultrasound and Lesion Characteristics Affecting Visualization on Breast Ultrasound

To assess the efficacy of magnetic resonance imaging (MRI)-directed ultrasound (US) in further characterizing lesions detected on MRI and to evaluate the correlation between lesion characteristics and their visibility on US.

A retrospective review was performed of the records of patients with breast abnormalities detected on MRI between May 2018 and December 2020. All patients who were given a Breast Imaging-Reported and Data System Assessment Category 4 or 5 on breast MRI and sent to breast US for assessment of the MRI abnormality were included. Patients lost to follow-up or those who did not get their subsequent ultrasound at our institution were excluded. The following factors were recorded for all lesions described on the MRI reports: breast density, degree of background parenchymal enhancement, type of MRI finding (mass, areas of non-mass enhancement [NME], enhancing focus), largest dimension, T2 signal intensity, kinetics, shape, margins, enhancement pattern, and presence of ipsilateral breast malignancy. For the subset of the lesions that were identified on MRI-directed US, we recorded the following sonographic features: largest dimension, shape, and margin. The chi-square or Fisher exact tests of association were used to compare categorical characteristics.

Two hundred and thirty-three MRI-detected breast lesions were included in the study, which consisted of 171 masses, 7 enhancing foci, and 55 NME lesions. Of all lesions, 52% (89/171) of masses, 29% (16/55) of areas of NME, and 29% (2/7) of foci were seen on US. Moreover, 43 per 63 (68%) of malignant, 18 per 47 (38%) of high-risk and 46 per 123 (37%) of benign lesions were seen on MRI-directed US. Of the invasive malignant lesions, 77% (36/47) had a sonographic correlate. For masses, the average size on MRI was 1.1 cm, and MRI size had a statistically significant association with US visibility (P value 0.01). Similarly, spiculated margins on MRI for masses had a significant association with US visibility (P value 0.0006). Linear, segmental, and regional distributions of areas of NME were associated with decreased likelihood of sonographic visibility (P = 0.049).

MRI-directed US is a useful tool in the management of MRI-detected breast lesions. Our study indicates that masses, particularly those with spiculated margins on MRI, should be sent to US for further assessment. Additionally, masses are more likely to be seen under US as their size increases. MRI-directed US is less likely to visualize areas of NME and enhancing foci, which may be managed with MRI follow-up or MRI-guided biopsy depending on their level of suspicion.