Imaging features of dermatofibrosarcoma protuberans

Abstract

Aims: The study highlights diffusion-weighted imaging (DWI) and dynamic enhancement features of DFSP and characterizes unenhanced and enhanced computed tomography (CT) and magnetic resonance imaging (MRI) scans. Settings and Design: Image findings and clinical histories of 23 patients with DFSP were reviewed. Nine patients underwent CT before and after intravenous administration of contrast material. MRI was performed for 17 patients. CT and MRI findings were analyzed using location, size, edge, shape, infiltration sign, density and signal enhancement mode, and degree. Results: Patients showed 26 superficial and one deep lesion. Ten superficial lesions bulged onto the skin surface. Fourteen lesions were well-defined and 13 ill-defined. All lesions were nodular, with nine being multilobular. Thirteen showed infiltration to adjacent skin, fat, and fascia. Seven lesions on CT were iso- or hypo-dense to muscle without calcification. Contrast-enhanced CT showed inhomogeneous moderate and progressive enhancement in the arterial phase. Small tortuous vessels were seen in the arterial phase in one case. Sixteen tumors displayed signals that were similar to muscle by T1WI. Ten lesions were either hyper-intense to muscle or iso-intense to fat; the deep DFSP was hypo-intense by T2WI. All lesions were hyper-intense homogeneously or heterogeneously under fat-suppressed T2WI. Twelve superficial lesions showed high-intermediate signal, and one deep lesion showed low-intermediate signal with DWI. Seven cases showed low signal diffusion coefficient (ADC) images. Dynamic enhancement and signal intensity-time (SI-T) curves of four tumors showed rapid SI increases followed by steady or slightly rising SI. All lesions showed inhomogeneous, progressive enhancement in the arterial phase. Conclusions: This report is the first on dynamic curves and highlights DWI and T2WI features of DFSP. DFSP can be correctly diagnosed by combining a patient's clinical manifestations with imaging characteristics.