Association of Diffusion Tensor Imaging Measures at Term with Absent Fidgety Movements in Infants with Hypoxic-ischemic Encephalopathy

Background: MRI is standard of care in the evaluation of brain injury after neonatal hypoxic-ischemic encephalopathy (HIE). However, clinicians still struggle to interpret MRI findings and counsel families on possible future neurodevelopmental impairment. The Precht General Movements Assessment (GMA) has been shown to be specific and sensitive for prediction of motor impairment; an absent fidgety GMA at 3-4 months post-term has high specificity (82–99%) and sensitivity (95–100%) for development of cerebral palsy (Al-Mayahi et al 2018). MRI with diffusion tensor imaging (DTI) allows clinicians to evaluate white matter injury by using anisotropic properties of water diffusion to evaluate white matter microstructure. Tract-based spatial statistics (TBSS) can be used to evaluate fractional anisotropy (FA) as a biomarker for white matter microstructural injury in an automated, whole-brain, voxel-wise approach (Porter et al 2010). This study aims to evaluate the association of FA on DTI with absent fidgety movements on GMA at 3-4 months in infants with HIE using TBSS. This may allow earlier prediction of cerebral palsy, thereby enabling earlier therapeutic intervention.

Methods: Infants treated with hypothermia for moderate or severe HIE between 2017-2021 were consented for participation and scanned using a 3T MRI scanner with DTI after rewarming (day of life 5 +/- 1.8). Infants found to have a genetic syndrome were excluded. All surviving infants had a GMA performed and scored by a certified examiner blinded to MRI/DTI results at 3-4 months of age. DTI data was preprocessed with the FMRIB Software Library, version 6.0, and infants with excessive motion artifact were excluded. Tissue FA was then determined in all infants, and using TBSS, was compared between infants with normal fidgety GMA and absent fidgety GMA.

Results: Eighteen infants with moderate to severe encephalopathy which underwent therapeutic hypothermia met study criteria; three surviving infants with absent fidgety movements were compared with fifteen infants with normal fidgety GMA. Infants with absent fidgety GMA had significantly decreased FA of several white matter tracts compared with infants with HIE that had a normal GMA. Tracts which significantly differed included the corpus callosum, bilateral corticospinal tracts, bilateral acoustic and optic radiations, and bilateral thalamus to sensory cortex tracts.

Conclusion: Decreased FA on initial MRI may help us understand the regions of the brain critical to fidgety movements at 3-4 months post-term while also allowing us to predict absent fidgety GMA and by proxy, cerebral palsy in infants with moderate and severe HIE prior to hospital discharge. This information may enable physicians to more accurately counsel families earlier in an infant’s clinical course and could potentially allow them to pursue therapeutic interventions prior to 3-4 months of age.