Quantitative Characterization of Radiation Dose Dependent Changes in Normal- Appearing White Matter of Cerebral Tumor Patients Using Diffusion Tensor Imaging

Reviewer: John P. Plastaras, MD, PhD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 8, 2006

Share article


Presenter: V. Nagesh
Presenter's Affiliation: University of Michigan, Ann Arbor, MI
Type of Session: Plenary

Background

  • Radiation effects on normal brain tissue can include neuroinflammation, demyelination, disruption of the blood-brain barrier, cerebral edema, and white matter necrosis.
  • Hypothesis: Radiation causes demyelination and structural degradation in normal appearing white matter (NAWM).

Materials and Methods

  • 20 patients with malignant glioma, low grade astrocytoma, and benign conditions
  • Radiation: median dose, 67 Gy
  • Imaging:
    • Diffusion tensor imaging using MRI was used before radiation, during radiation (at 1 week, 3 weeks, and at 6 weeks), and after radiation (1 month, 3 months, and 6 months after completion of RT)
    • Principle: MRI can image the diffusion of protons in water. In a neuron, water preferentially diffuses along the length of the axon (parallel) rather than in the radial direction (perpendicular), because it is blocked by the myelin sheath.
    • Demyelination would increase the amount of perpendicular diffusion compared to parallel diffusion because the myelin sheath would no longer block the water diffusion.
  • The genu of the corpus callosum was selected as a site of NAWM because all the fibers run in the same direction.
  • The "fractional anisotropy" (FA) is the ratio of parallel to perpendicular diffusion. A high FA indicates an intact, myelinated neuron. A low FA is a sign of demyelination.

Results

  • 1 month following radiation, there was evidence of demyelination (FA decreased 14%).
  • After radiation, overall diffusivity increased in both the perpendicular and parallel directions, but the perpendicular direction increased more, indicating demyelination.
  • There was a radiation dose-dependent response in both overall diffusivity (+ 4.4 x 10 -6 mm 2 s -1 /Gy and FA (-0.034/Gy).
  • There was a significant correleation between dose and FA during radiation (from 3 weeks to 19 weeks), but at 32 weeks, this was no longer significant.
  • Parallel diffusion changes (indicating axonal damage) were not significant early on, but became significant at 32 weeks. This indicates that axon damage is a late radiation change.

Author's Conclusions

  • Diffusion tensor MR imaging of corpus callosum genu can measure white matter demyelination and axonal damage in patients undergoing brain radiotherapy.
  • There were dose-dependent acute and subacute decreases in NAWM myelination from brain irradiation.
  • Axonal injury occurs later and is also dose-dependent.

Clinical/Scientific Implications

  • The early and late sequelae of brain radiotherapy are poorly understood. The technique described by the authors using diffusion tensor MR imaging of white matter provides a tool to study these changes.
  • The relationship between demyelination / axonal damage and clinically important side-effects, such as somnolence syndrome, neurocognitive deficits, and brain necrosis, are still unknown. Such imaging techniques may be used for predication of these late effects.