MR imaging of extracapsular silicone from breast implants: Diagnostic pitfalls

OBJECTIVE. We sought to identify pitfalls in recognition of extracapsular silicone on MR imaging. MATERIALS AND METHODS. Three experienced observers reviewed MR images from 3,59 women with current (n = 320), prior (n = 15), or both current and prior (n = 24) silicone gel implants. Axial and sagittal fast spin-echo T2-weighted images with water suppression, axial inversion-recovery T2-weighted images with water suppression, and axial T2-weighted images with silicone suppression were obtained in a dedicated phased array breast coil on a 1.5-T magnet. Images were reviewed again when only one observer saw extracapsular silicone, and reasons for disagreement were recorded. RESULTS. Rupture was identified in 265 women (77%) with current silicone implants and 378 (55%) of 687 implants. Observers agreed in describing extracapsular silicone in 85 (12%) of 687 breasts with current silicone gel implants, of which 81 (95%) showed definite evidence of rupture on MR imaging. One observer reported extracapsular silicone in another 79 breasts. Confusion over contour deformity due to weakening versus breach of the capsule accounted for 33 (42%) of 79 disagreements. Another 20 (25%) of the 79 disagreements were attributed to poor conspicuity of extracapsular silicone on fast spin-echo T2-weighted images combined with intermittent observer failure to review inversion-recovery images. Subtlety of findings (n = 17, 22%) and technical issues (n = 9, 11%) with failed water suppression of pleural effusion or cysts and ghosting artifacts accounted for remaining disagreements. CONCLUSION. Extracapsular rupture is usually manifest as local spread of silicone in the breast and is not well-depicted on fast spin-echo T2-weighted images. Water-suppressed inversion-recovery T2-weighted images are often needed to identify extracapsular silicone. Distinction of the bulge in the fibrous capsule from herniation through the capsule remains problematic.