Comparison of Imaging and Pathologic Findings of Retroperitoneal Dedifferentiated Liposarcoma

2019 
Objective To investigate the imaging appearance of CT and MRI in retroperitoneal dedifferentiated liposarcoma (DDL) based on pathological findings. Methods Twelve patients with retroperitoneal DDL (13 lesions) who were surgically and pathologically confirmed were retrospectively collected in the Cancer Hospital of Chinese Academy of Medical Sciences. The correlation of CT and MRI features with histopathologic findings was analyzed. Results The CT and MRI images of retroperitoneal DDLs were large, heterogeneous soft-tissue masses, mostly lobulated (30.8%, 4/13) or multinodular (46.2%, 6/13), invading adjacent anatomic structures (46.2%, 6/13). The lesions contained different proportions of fatty and non-fatty components, and usually with clear boundaries. The CT images of dedifferentiated components showed non-fatty masses of soft tissue density or mixed density, among which ground-glass nodules may be related to mucinous components. Occasionally calcification or ossification was seen (45.5%, 5/11). The contrast-enhanced CT and MRI images of non-fatty components commonly showed intense heterogeneous enhancement (84.6%, 11/13), central cystic changes and necrosis (61.5%, 8/13), pathologically corresponding to multiple types of soft tissue sarcomas without significant specificity. The well-differentiated components were fatty masses with irregular fibrous septa or soft tissue nodules, which is pathologically corresponding to well differentiated liposarcoma. Lymph node or distant metastasis was rare. Conclusions The imaging manifestations of retroperitoneal DDLs are diverse and closely related to the proportion and distribution of different components. CT, MRI and contrast-enhanced imaging has a certain diagnostic value for retroperitoneal DDLs. Key words: Retroperitoneal neoplasms; Dedifferentiated liposarcoma; Tomography, X-ray Computed; Magnetic resonance imaging
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    0
    Citations
    NaN
    KQI
    []