Imaging of irradiated brain tumours. Value of magnetic resonance imaging

J Neuroradiol. 1989;16(2):81-132.
[Article in English, French]

Abstract

Until recently, computerized tomography (CT) was the most sensitive and reliable imaging method to follow up patients with an irradiated brain tumour, but it is difficult or impossible with CT to differentiate between radionecrosis, residual tumour or recurrent tumour. Magnetic resonance imaging (MRI) has become the most sensitive examination. It ensures optimal focusing of radiography, thereby increasing its effectiveness and reducing its complications. In one-third of the cases studied, MRI has shown pathological signals that were invisible at CT, making for a better adjustment of treatment and a more accurate diagnosis. As regards specificity, MRI does not seem to provide criteria that would enable a radiation lesion to be distinguished from a tumoral recurrence. However, we found that certain signs may contribute to the aetiological diagnosis: a perilesional high-intensity signal extending to the grey matter and/or the corpus callosum is in favour of a recurrent tumour; a high-intensity signal on T1- and T2-weighted sequences and a disproportionally moderate mass syndrome are in favour of a radionecrosis; a post-irradiation leucoencephalopathy sparing the grey matter and the corpus callosum is in favour of a remission. Injecting a gadolinium complex always gives a better distinction between oedema, tumour and necrosis, it may also improve MRI sensitivity and sensitivity in some cases, and it reduces the time taken by the examination. MRI is now the reference morphological examination; its specificity can be further increased by positron emission tomography and assays of polyamines in red blood cells.

Publication types

  • Review

MeSH terms

  • Brain Neoplasms / diagnosis*
  • Brain Neoplasms / etiology
  • Brain Neoplasms / pathology
  • Humans
  • Magnetic Resonance Imaging
  • Neoplasms, Radiation-Induced / diagnosis*
  • Neoplasms, Radiation-Induced / pathology