Deformable anatomic templates improve analysis of gliomas with minimal mass effect in eloquent areas

Neurosurgery. 2013 Sep;73(3):534-42. doi: 10.1227/01.neu.0000431479.87160.e2.

Abstract

Background: Despite improvements in advanced magnetic resonance imaging and intraoperative mapping, cases remain in which it is difficult to determine whether viable eloquent structures are involved by a glioma. A novel software program, deformable anatomic templates (DAT), rapidly embeds the normal location of eloquent cortex and functional tracts in the magnetic resonance images of glioma-bearing brain.

Objective: To investigate the feasibility of the DAT technique in patients with gliomas related to eloquent brain.

Methods: Forty cases of gliomas (grade II-IV) with minimal mass effect were referred for a prospective preoperative and postoperative DAT analysis. The DAT results were compared with the patient's functional magnetic resonance imaging, diffusion tensor imaging, operative stimulation, and new postoperative clinical deficits.

Results: Fifteen of the 40 glioma patients had overlap between tumor and eloquent structures. Immediate postoperative neurological deficits were seen in 9 cases in which the DAT showed the eloquent area both within the tumor and within or at the edge of the resection cavity. In 6 cases with no deficits, DAT placed the eloquent area in the tumor but outside the resection cavity.

Conclusion: This is proof of concept that DAT can improve the analysis of diffuse gliomas of any grade by efficiently alerting the surgeon to the possibility of eloquent area invasion. The technique is especially helpful in diffuse glioma because these tumors tend to infiltrate rather than displace eloquent structures. DAT is limited by tract displacement in gliomas that produces moderate to severe mass effect.

MeSH terms

  • Adult
  • Aged
  • Brain Mapping
  • Brain Neoplasms / pathology*
  • Female
  • Glioma / pathology*
  • Humans
  • Magnetic Resonance Imaging
  • Male
  • Middle Aged
  • Retrospective Studies
  • Software*