Classification of human liposarcoma and lipoma using ex vivo proton NMR spectroscopy. Academic Article uri icon

Overview

abstract

  • Prognostication in patients with liposarcoma is a complex and controversial subject based on recognition of lipoblasts, adipocyte nuclear atypia, and qualitative estimations of cellularity and cell size. We show here that for 30 patients with liposarcoma and 5 patients with lipoma, spectral differences on high-resolution, magic angle spinning proton nuclear magnetic resonance (hr-MAS 1H-NMR) spectroscopy relate to known biochemical changes and correlate with adipocyte tissue differentiation, histologic cell type, and cellularity. The NMR-visible level of triglyceride is shown to correlate with liposarcoma differentiation, since the triglyceride level in well-differentiated liposarcoma is 33-fold higher on average than for myxoid/round cell liposarcoma, which in turn is 6-fold higher than the dedifferentiated and/or pleomorphic subtypes. The NMR-visible phosphatidylcholine level serves as an estimate of total tissue cell membrane phospholipid mass and was found to correlate with liposarcoma subtype. Pleomorphic liposarcoma, the most aggressive and metastatic subtype, was found to have a threefold increase in NMR-visible phosphatidylcholine level compared with dedifferentiated liposarcoma. The level of NMR-visible phosphatidylcholine was twofold greater in well-differentiated liposarcoma compared with lipoma and was threefold larger for the hypercellular myxoid/round cell subtype compared with the pure myxoid histology. Thus, NMR-derived parameters of tissue lipid may be used for objective distinction of liposarcoma histologic subtype/grade and lipoma from liposarcoma. These biochemical parameters may ultimately improve prognostication in patients with liposarcoma.

publication date

  • February 1, 1999

Research

keywords

  • Lipoma
  • Liposarcoma
  • Magnetic Resonance Spectroscopy

Identity

Scopus Document Identifier

  • 0033004204

PubMed ID

  • 10080272

Additional Document Info

volume

  • 41

issue

  • 2