Diffuse axonal injury in children: Clinical correlation with hemorrhagic lesions

Karen A. Tong, Stephen Ashwal, Barbara A. Holshouser, Joshua P. Nickerson, Christopher J. Wall, Lori A. Shutter, Renatta J. Osterdock, E. M. Haacke, Daniel Kido

Research output: Contribution to journalArticlepeer-review

295 Scopus citations

Abstract

An inception cohort of 40 children and adolescents with traumatic brain injury and suspected diffuse axonal injury were studied using a new high-resolution magnetic resonance imaging susceptibility-weighted technique that is very sensitive for hemorrhage. A blinded comparison was performed between the extent of parenchymal hemorrhage and initial clinical variables as well as outcomes measured at 6 to 12 months after injury. Children with lower Glasgow Coma Scale scores (≤8, n = 30) or prolonged coma (>4 days, n = 20) had a greater average number (p = 0.007) and volume (p = 0.008) of hemorrhagic lesions. Children with normal outcomes or mild disability (n = 30) at 6 to 12 months had, on average, fewer hemorrhagic lesions (p = 0.003) and lower volume (p = 0.003) of lesions than those who were moderately or severely disabled or in a vegetative state. Significant differences also were observed when comparing regional injury to clinical variables. Because susceptibility-weighted imaging is much more sensitive than conventional T2*-weighted gradient-echo sequences in detecting hemorrhagic diffuse axonal injury, more accurate and objective assessment of injury can be obtained early after insult, and may provide better prognostic information regarding duration of coma as well as long-term outcome.

Original languageEnglish (US)
Pages (from-to)36-50
Number of pages15
JournalAnnals of Neurology
Volume56
Issue number1
DOIs
StatePublished - Jul 2004
Externally publishedYes

ASJC Scopus subject areas

  • Neurology
  • Clinical Neurology

Fingerprint

Dive into the research topics of 'Diffuse axonal injury in children: Clinical correlation with hemorrhagic lesions'. Together they form a unique fingerprint.

Cite this