Myxoid liposarcoma: magnetic resonance imaging appearances with clinical and histological correlation

Murali Sundaram, Gregg Baran, Gloria Merenda, Douglas J. McDonald

Research output: Contribution to journalArticlepeer-review

68 Scopus citations


Myxoid liposarcoma is the most common type of liposarcoma. The magnetic resonance imaging (MRI) features of this tumor were evaluated and correlated with its clinical and histological features in seven patients to determine under what circumstances the tumor should be considered in differential diagnosis and why its signal intensity differs from those of lipoma and lipoma-like (lipoblastic) liposarcoma. In all patients the tumor presented in a lower extremity (5 thigh, 2 calf) as a painless, slowly growing mass which had been present for several months to several years. MRI examination revealed the tumors to be encapsulated, noninfiltrating, and usually septated. On T1-weighted sequences five of seven lesions (71%) showed lacy, amorphous, or linear foci of high signal within a low signal mass. These foci are believed to represent fat within the tumor and distinguish it from several other benign and malignant masses. If an indolent mass in a lower extremity demonstrates a predomiantly low signal with a few amorphous or linear high signal foci on T1-weighted sequences, one should consider the possibility of myxoid liposarcoma even if it appears benign by all other criteria. Histologic evaluation showed that myxoid liposarcomas contain less than 10% mature fat, which accounts for their low signal on T1-weighted sequences in contrast to the high signal of lipomas and lipoblastic liposarcomas.

Original languageEnglish
Pages (from-to)359-362
Number of pages4
JournalSkeletal Radiology
Issue number5
StatePublished - Jul 1 1990


  • Long T-high signal foci
  • Magnetic resonance-imaging
  • Myxoid liposarcoma


Dive into the research topics of 'Myxoid liposarcoma: magnetic resonance imaging appearances with clinical and histological correlation'. Together they form a unique fingerprint.

Cite this