Objective: Reliable diagnosis of liposarcoma by magnetic resonance imaging (MRI) is essential for surgical planning. The purpose of this study was to correlate radiologists' certainty of the diagnosis of liposarcoma on musculoskeletal MRI with pathology results.
Materials and Methods: Between January 2001 and February 2009, 105 patients who radiologically suggested liposarcoma on their MRI reports were retrospectively reviewed. Among them, pathologically confirmed 64 patients (benign, 42; malignant, 22) were included. Two musculoskeletal radiologists reviewed MR reports and classified these into 3 groups according to the degree of certainty of liposarcoma (CL) by consensus: C1, high; C2, undetermined; and C3, low. Classified data were compared with pathology results.
Results: The number of cases in each group was C1, 32; C2, 12; and C3, 20. Group C1 included 17 liposarcomas (53%), 7 lipomas, 4 other sarcomas, 2 hibernomas, 1 abscess, and 1 epidermal cyst. There were 11 lipomas (92%) and one liposarcoma in C2. In C3, all patients had lipomas. Fifteen (47%) of 32 variable benign or malignant tumors were incorrectly diagnosed as liposarcomas.
Conclusion: On the diagnosis of liposarcoma, the radiologists' high degree of CL showed high agreement. However, nonadipose tumors were sometimes misdiagnosed as liposarcomas with a high CL. Therefore, we should consider other soft tissue tumors and benign lipomas for the differential diagnosis of liposarcoma.
From the *Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, Seoul; †Department of Radiology, Kangwon National University College of Medicine, Chuncheon, Korea; ‡Departments of Orthopedic Surgery, and §Pathology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea.
Received for publication January 23, 2011; accepted April 12, 2011.
Reprints: Hye Won Chung, MD, PhD, Department of Radiology, University of Ulsan College of Medicine, Asan Medical Center, 86, Asanbyeongwon-gil, Songpa-gu, Seoul 138-736, Korea (e-mail: firstname.lastname@example.org).
The authors did not receive funding for this work, and they have no conflicts of interest.