Predictive value of neonatal MRI showing no or minor degrees of brain injury after hypothermia

Pediatr Neurol. 2014 May;50(5):447-51. doi: 10.1016/j.pediatrneurol.2014.01.013. Epub 2014 Jan 7.

Abstract

Background: Magnetic resonance imaging is a surrogate biomarker for major neurodevelopmental disabilities in survivors of perinatal hypoxic-ischemic encephalopathy because injury to the basal ganglia/thalami is highly predictive of major neuromotor and cognitive problems. Major disabilities and the appearance of neonatal magnetic resonance imaging are improved with therapeutic hypothermia. We evaluated neurodevelopmental outcomes when conventional magnetic resonance imaging showed minimal or no brain injury.

Methods: Institutional review board-approved series of 62 infants (≥36 weeks; ≥1800 g; 34 boys/28 girls) cooled for hypoxic-ischemic encephalopathy between 2005 and 2011 who underwent neonatal magnetic resonance imaging and Bayley Scales of Infant and Toddler Development-III at 22 ± 7 months of age. Magnetic resonance imaging at 5-14 (mean 8) days was scored as normal (score = 0), showing focal gray or white matter injury only (score = 1), or basal ganglia/thalamic and/or watershed lesions with or without more extensive hemispheric injury (score = 2). Sensitivity, specificity, and positive and negative predictive values for magnetic resonance scores 0 and 1 and statistical interaction between magnetic resonance imaging score and age at magnetic resonance imaging were determined.

Results: Magnetic resonance score = 0 was seen in 35/62 patients; 26/35 (74%) were typically developing, seven (20%) had moderate and two (6%) had severe delay. Magnetic resonance score = 1 was seen in 17/62 (27%) patients; 5/17 (29%) were normal, 11/17 (65%) had moderate delay, and 1/17 (6%) had severe neurodevelopmental delay. Of the 52 patients with magnetic resonance scores of 0 and 1, 40% were abnormal. The negative predictive value of a normal magnetic resonance imaging was 74%. For score 1, sensitivity was 95% (confidence interval 63%-83%), specificity 84% (confidence interval 70%-90%), positive predictive value 84% (confidence interval 71%-93%), and negative predictive value 74% (confidence interval 62%-82%).

Conclusions: Caution is warranted when prognosticating about neurodevelopmental status in early childhood after hypoxic ischemic encephalopathy with cooling, and longer follow-up studies are needed to determine the prognostic significance of a neonatal magnetic resonance imaging showing no or minor degrees of brain injury.

Keywords: MRI; hypothermia; hypoxic ischemic encephalopathy; infant; newborn; prognosis.

MeSH terms

  • Basal Ganglia / pathology
  • Brain / growth & development
  • Brain / pathology*
  • Child Development
  • Developmental Disabilities / diagnosis
  • Developmental Disabilities / etiology
  • Developmental Disabilities / pathology
  • Female
  • Follow-Up Studies
  • Humans
  • Hypoxia-Ischemia, Brain / complications
  • Hypoxia-Ischemia, Brain / diagnosis*
  • Hypoxia-Ischemia, Brain / pathology*
  • Hypoxia-Ischemia, Brain / therapy
  • Infant
  • Infant, Newborn
  • Magnetic Resonance Imaging*
  • Male
  • Nerve Fibers, Myelinated / pathology
  • Nerve Fibers, Unmyelinated / pathology
  • Neuropsychological Tests
  • Prognosis
  • Sensitivity and Specificity
  • Severity of Illness Index
  • Thalamus / pathology