Objective: The objective was to evaluate ultrasound and MRI in clinical appendicular and truncal fat necrosis.
Materials and methods: Thirty-three patients (14 men, 19 women, median age 55, range 29-95) were retrospectively evaluated. Histologically, three groups were seen: Group 1 (n = 18) consisted of patients with subcutaneous masses with septal and extrinsic oedema; in Group 2 (n = 11) necrosis occurred within lipomatous tumours and little oedema; and in Group 3 (n = 4) there were large complex masses consistent with Morel-Lavallée lesions. Two experienced radiologists reviewed MR (n = 30) and ultrasound (n = 32) images with consensus agreement. MRI was performed on a 1.5T system with T1-weighted, T2-weighted fat-suppressed and T1-weighted fat-suppressed post-intravenous gadolinium sequences obtained in two orthogonal planes. Ultrasound (linear 5- to 13.5-MHz probe) was performed in the longitudinal and short axis. Anatomical position, size, shape (oval, linear, ill-defined), internal architecture (lobules, septi or stranding), intrinsic signal characteristics, presence of surrounding pseudocapsule, extrinsic linear stranding and vascularity (gadolinium enhancement or power Doppler) were recorded.
Results: Anatomical locations were buttock/thigh (n = 17), leg (n = 6), upper limb (n = 5) and thoracic/abdominal wall (n = 5) with the majority of lesions (30 out of 33) oval/linear in shape. On ultrasound and MRI most lesions showed internal fat lobules, intervening septi and a surrounding pseudocapsule.
Conclusion: Fat necrosis can usually be identified as containing multiple fat lobules on ultrasound and MRI despite a varying degree of inflammatory change surrounding and within the mass.