Mean Diffusivity in Striatum Correlates With Acute Neuronal Death but Not Lesser Neuronal Injury in a Pilot Study of Neonatal Piglets With Encephalopathy

BACKGROUND: Diffusion MRI is routinely used to evaluate brain injury in neonatal encephalopathy. Although abnormal mean diffusivity (MD) is often attributed to cytotoxic edema, the specific contribution from neuronal pathology is unclear. PURPOSE: To determine whether MD from high-resolution diffusion tensor imaging (DTI) can detect variable degrees of neuronal degeneration and pathology in piglets with brain injury induced by excitotoxicity or global hypoxia-ischemia (HI) with or without overt infarction. STUDY TYPE: Prospective. ANIMAL MODEL: Excitotoxic brain injury was induced in six neonatal piglets by intrastriatal stereotaxic injection of the glutamate receptor agonist quinolinic acid (QA). Three piglets underwent global HI or a sham procedure. Piglets recovered for 20-96 hours before undergoing MRI (n = 9). FIELD STRENGTH/SEQUENCE: 3.0T MRI with DTI, T1 - and T2 -weighted imaging. ASSESSMENT: MD, fractional anisotropy (FA), and qualitative T2 injury were assessed in the putamen and caudate. The cell bodies of normal neurons, degenerating neurons (excitotoxic necrosis, ischemic necrosis, or necrosis-apoptosis cell death continuum), and injured neurons with equivocal degeneration were counted by histopathology. STATISTICAL TESTS: Spearman correlations were used to compare MD and FA to normal, degenerating, and injured neurons. T2 injury and neuron counts were evaluated by descriptive analysis. RESULTS: The QA insult generated titratable levels of neuronal pathology. In QA, HI, and sham piglets, lower MD correlated with higher ratios of degenerating-to-total neurons (P  0.99). Neuron counts were not related to FA (P > 0.30) or to qualitative injury from T2 -weighted MRI. DATA CONCLUSION: MD is more accurate than FA for detecting neuronal degeneration and loss during acute recovery from neonatal excitotoxic and HI brain injury. MD does not reliably detect nonfulminant, nascent, and potentially reversible neuronal injury. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 2.