UCSF

Objective

To evaluate the association of therapeutic hypothermia with magnetic resonance imaging (MRI) findings and 30-month neurodevelopment in term neonatal encephalopathy.

Study design

Cross-sectional analysis of 30-month neurodevelopment (IQR 19.0-31.4) in a prospective cohort of mild-to-severe neonatal encephalopathy imaged on day 4 (1993-2017 with institutional implementation of therapeutic hypothermia in 2007). MRI injury was classified as normal, watershed, or basal ganglia/thalamus. Abnormal motor outcome was defined as Bayley-II psychomotor developmental index <70, Bayley-III motor score <85 or functional motor deficit. Abnormal cognitive outcome was defined as Bayley-II mental developmental index <70 or Bayley-III cognitive score <85. Abnormal composite outcome was defined as abnormal motor and/or cognitive outcome, or death. The association of therapeutic hypothermia with MRI and outcomes was evaluated with multivariable logistic regression adjusted for propensity to receive therapeutic hypothermia.

Results

Follow-up was available in 317 (78%) surviving children, of whom 155 (49%) received therapeutic hypothermia. Adjusting for propensity, therapeutic hypothermia was independently associated with decreased odds of abnormal motor (OR 0.15, 95% CI 0.06-0.40, P < .001) and cognitive (OR 0.11, 95% CI 0.04-0.33, P < .001) outcomes. This association remained statistically significant after adjustment for injury pattern. The predictive accuracy of MRI pattern for abnormal composite outcome was unchanged between therapeutic hypothermia-treated (area under the receiver operating curve 0.76; 95% CI 0.61-0.91) and untreated (area under the receiver operating curve 0.74; 95% CI 0.67-0.81) infants. The negative predictive value of normal MRI was high in therapeutic hypothermia-treated and untreated infants (motor 96% vs 90%; cognitive 99% vs 95%).

Conclusions

Therapeutic hypothermia is associated with lower rates of brain injury and adverse 30-month outcomes after neonatal encephalopathy. The predictive accuracy of MRI in the first week of life is unchanged by therapeutic hypothermia. Normal MRI remains reassuring for normal 30-month outcome after therapeutic hypothermia.