Prognostic significance of the MDM2/HMGA2 ratio and histological tumor grade in dedifferentiated liposarcoma.

2021 
Dedifferentiated liposarcoma (DDLPS) is a relatively common soft tissue sarcoma that results from the progression of well-differentiated liposarcoma (WDLPS). This study aimed to investigate the progression process and to clarify the pathological and genetic factors related to poor prognosis in DDLPS. In 32 DDLPS cases and five WDLPS cases, genetic factors were analyzed by custom comparative genomic hybridization (CGH) array, which was designed to densely cover gene regions known to be frequently amplified in WD/DDLPS. The analyses comparing primary and metastatic lesions and those comparing histologically different areas in the same tumor revealed intra-tumoral genetic heterogeneity and progression. According to a prognostic analysis comparing the good-prognosis and the poor-prognosis groups, we selected MDM2 and HMGA2 as candidate genes associated with poor and good prognosis, respectively. The ratios of the amplification or gain levels of MDM2 and HMGA2 expressed in log ratios (log[MDM2/HMGA2] = log[MDM2]-log[HMGA2]) were significantly associated with prognosis. An amplification or gain level of MDM2 that was more than twice that of HMGA2 (MDM2/HMGA2 > 2, log[MDM2/HMGA2] > 1) was strongly related to poor OS (P < .001) and poor distant metastasis-free survival (DMFS) (P < .001). In the pathological analysis of 44 cases of DDLPS, histological tumor grade, cellular atypia, and MDM2 immunoreactivity were related to overall survival (OS), while HMGA2 immunoreactivity tended to be associated with OS. Cellular atypia was also associated with DMFS. In conclusion, histological grade and MDM2 expression were determined to be prognostically important, and the MDM2/HMGA2 amplification or gain ratio was found to have significant prognostic value by the custom CGH array analysis.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    22
    References
    2
    Citations
    NaN
    KQI
    []