Therapeutic Hypothermia for Neonatal Hypoxic –Ischemic Encephalopathy: Do Lesions on MRI Predict Outcome?

Abstract & Commentary

By Toshiki Takenouchi, MD, and Steven Weinstein, MD. Dr. Takenouchi is Clinical Fellow in Neonatal Neurology, and Dr. Weinstein is Director of the Pediatric Comprehensive Epilepsy Program, Weill Cornell Medical School, New York Presbyterian Hospital. Neither Dr. Takenouchi nor Dr. Weinstein report any financial relationships relevant to this field of study.

Synopsis: MRI analysis of a subset of infants from the Total Body Hypothermia for Neonatal Encephalopathy (TOBY) trial showed that total body hypothermia decreases brain tissue injury in infants with hypoxic-ischemic encephalopathy (HIE).

Source: Rutherford M, Assessment of brain tissue injury after moderate hypothermia in neonates with hypoxic-ischaemic encephalopathy: A nested substudy of a randomised controlled trial. Lancet Neurol, online November 6, 2009. DOI:10.1016/S1474-4422(09)70295-70299.

Toby was a European, multicenter, randomized, controlled clinical trial conducted between 2002 and 2006 to evaluate the efficacy of total body hypothermia to treat neonatal hypoxic-ischemic encephalopathy (HIE) in infants 3 36 weeks gestation. Infants with moderate to severe encephalopathy were randomized, at < 6 hours of age, to either total body hypothermia at 33.5°C for 72 hours (n=163) or normothermic intensive care (n=162). At 18 months, there was no significant reduction in the combined rate of death or severe disability, but there was better neurologic outcome in survivors.

In the current subset analysis, the authors analyzed the MRIs of 131 infants recruited from the original TOBY study (n=325). The amplitude of integrated EEG, an objective measure of encephalopathy, was similar in each group. MRI was obtained in a subset of early survivors, representing 40% of the original subjects, but selection criteria for imaging was not provided. The median time of postnatal scans was eight days, ranging from two to 30 days. 37 of 59 (63%) infants scanned at less than eight days of age had major abnormalities on MRI that were predictive of poor outcome, compared to 34 of 72 (47%) of infants scanned after eight days (p=0.08).

Two independent radiologists, blinded to the outcomes, reviewed T1- and T2-weighted MR images and counted lesions in the basal ganglia, thalami, posterior limb of the internal capsule, subcortical white matter, and cortex as well as identifying intracranial hemorrhages and sinus thrombosis. Therapeutic hypothermia was associated with a reduction in lesions in the basal ganglia or thalamus (OR=0.36, 95% CI 0.15-0.84; p=0.02), subcortical white matter (OR=0.30, 0.12-0.77; p=0.01), and posterior limb of the internal capsule (OR=0.38, 0.17-0o85; p=0.02). Compared with the control group, infants treated with hypothermia had fewer scans that were predictive of later neurological abnormalities (OR=0.41, 0.20-0.92; p=0.03) and were more likely to have normal scans (OR=2.81, 1.13-6.93; p=0.03). The predictive accuracy of moderate or severe lesions in the basal ganglia and thalami, severe white-matter lesions, or an abnormal posterior limb of the internal capsule, for death or severe disability at 18 months of age was 0.84 (95% CI 0.74-0.94) in the cooled group and 0.81 (0.71-0.91) in the non-cooled group.


Moderate to severe neonatal HIE is a serious complication of pregnancy and delivery, since it is associated with a high incidence of neurological impairment. Since 2005, several large, international, multicenter, randomized control trials have shown marginal benefits of therapeutic hypothermia for term infants with moderate to severe HIE, especially for short-term neurologic outcome.1 Although therapeutic hypothermia has become widely available and is considered a standard of care for neonatal HIE, the optimal temperature and duration of treatment remain to be determined.

Brain MRI has become a standard to objectively assess anatomical changes from the ischemic insult and is highly predictive of outcome in this population. However, the predictive value of MRI in infants undergoing therapeutic hypothermia needs to be validated. The current study retrospectively analyzed a large number of MRIs in infants treated with hypothermia using an objective scoring system and demonstrated that hypothermia decreases cerebral lesions seen on MRI, consistent with a similar, but smaller report by Rutherford in 2005.2

There are technical limitations of this study that warrant caution in interpretation. First, the authors had to combine MRI data sets from scanners of different manufacturers, with different magnetic field strengths, and different acquisition parameters. Second, the investigators only utilized T1- and T2-weighted images and did not include diffusion weighted imaging (DWI). DWI has been considered the most sensitive method to identify acute ischemic injury with MRI and has been widely utilized in the neonatal population. Third, each MRI was obtained at the discretion of the treating physician and the acquisition of MRI ranged from two to 30 days with the median of eight days. Therefore, it is not possible to determine the best timing, acquisition sequences, or parameters of scanning from this study.


1. Azzopardi DV, et al. Moderate hypothermia to treat perinatal asphyxial encephalopathy. N Engl J Med 2009;361:1349-1358.

2. Rutherford MA, et al. Mild hypothermia and the distribution of cerebral lesions in neonates with hypoxic-ischemic encephalopathy. Pediatrics 2005;116:1001-1006.