Prognostic significance of MRI findings in patients with myxoid-round cell liposarcoma

AJR Am J Roentgenol. 2004 Mar;182(3):725-31. doi: 10.2214/ajr.182.3.1820725.

Abstract

Objective: The aims of this study were to determine the prognostic significance of MRI findings in patients with myxoid-round cell liposarcomas and to clarify which MRI features best indicate tumors with adverse clinical behavior.

Materials and methods: The initial MRI studies of 36 pathologically confirmed myxoid-round cell liposarcomas were retrospectively reviewed, and observations from this review were correlated with the histopathologic features. MR images were evaluated by two radiologists with agreement by consensus, and both univariate and multivariate analyses were conducted to evaluate survival with a median clinical follow-up of 33 months (range, 9-276 months).

Results: Statistically significant MRI findings that favored a diagnosis of intermediate- or high-grade tumor were large tumor size (> 10 cm), deeply situated tumor, tumor possessing irregular contours, absence of lobulation, absence of thin septa, presence of thick septa, absence of tumor capsule, high-intensity signal pattern, pronounced enhancement, and globular or nodular enhancement. Of these MRI findings, thin septa (p < 0.05), a tumor capsule (p < 0.01), and pronounced enhancement (p < 0.01) were associated significantly, according to univariate analysis, with overall survival. Multivariate analysis indicated that pronounced enhancement was associated significantly with overall survival (p < 0.05).

Conclusion: Contrast-enhanced MRI findings can indicate a good or adverse prognosis in patients with myxoid-round cell liposarcomas.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Chi-Square Distribution
  • Female
  • Humans
  • Liposarcoma / pathology*
  • Liposarcoma, Myxoid / pathology*
  • Magnetic Resonance Imaging*
  • Male
  • Middle Aged
  • Predictive Value of Tests
  • Prognosis
  • Proportional Hazards Models
  • Retrospective Studies
  • Soft Tissue Neoplasms / pathology*
  • Statistics, Nonparametric