Ischemically damaged brain can be accompanied by secondary degeneration of associated axonal connections e.g. Wallerian degeneration. Diffusion tensor imaging (DTI) is widely used to investigate axonal injury but the cellular correlates of many of the degenerative changes remain speculative. We investigated the relationship of DTI of directly damaged cerebral cortex and secondary axonal degeneration in the cerebral peduncle with cellular alterations in pan-axonal neurofilament staining, myelination, reactive astrocytes, activation of microglia/macrophages and neuronal cell death. DTI measures (axial, radial and mean diffusivity, and fractional anisotropy (FA)) were acquired at hyperacute (3 h), acute (1 and 2 d) and chronic (1 and 4 week) times after transient cerebral hypoxia with unilateral ischemia in neonatal rats. The tissue pathology underlying ischemic and degenerative responses had a complex relationship with DTI parameters. DTI changes at hyperacute and subacute times were smaller in magnitude and tended to be transient and/or delayed in cerebral peduncle compared to cerebral cortex. In cerebral peduncle by 1 d post-insult, there were reductions in neurofilament staining corresponding with decreases in parallel diffusivity which were more sensitive than mean diffusivity in detecting axonal changes. Ipsilesional reductions in FA within cerebral peduncle were robust in detecting both early and chronic degenerative responses. At one or four weeks post-insult, radial diffusivity was increased ipsilaterally in the cerebral peduncle corresponding to pathological evidence of a lack of ontogenic myelination in this region. The detailed differences in progression and magnitude of DTI and histological changes reported provide a reference for identifying the potential contribution of various cellular responses to FA, and, parallel, radial, and mean diffusivity.

Stroke is a major cause of disability in adults and children. Recently, we have developed an adult rat model of minor stroke containing a peri-infarct region with a modest T increase and mild ischemic damage. We hypothesized that a neonatal minor stroke with mild peri-ischemic changes could also be produced, but with potential ontogenic differences. Using our minor photothrombosis method, we produced a range of severities of ischemic lesions (mini, minor, moderate and severe) within magnetic resonance imaging (MRI) slices of adult and neonatal rats. In both age groups, the lesion region showed a marked increase in T and diffusion-weighted intensity and decrease in apparent diffusion coefficient (ADC), corresponding to a cortical infarct detected using fluorojade and hematoxylin and eosin staining. Perilesional regions showed modest increases in T and ADC in adults, but not neonates, and this corresponded to scattered cell death, but not necessarily extravasation of plasma protein, i.e. blood-brain barrier disruption. Mini and minor insults in neonates generally showed homogeneous and rather modest changes in T and ADC. MR perfusion maps demonstrated a penumbral area of greater hypoperfusion in adults compared with neonates. Together, the results indicate that, in neonatal cortex, a similar severity of photothrombosis occurs throughout the area of photoactivation, whereas, in adult brain, spontaneous clot lysis and/or partial thrombosis occurs adjacent to permanently occluded vessels. Thus, by comparing differing severities of photothrombotic ischemia in neonates and adults, ontogenic differences were detectable using MRI, with mature brain having a greater penumbral region. Mild ischemic injury and scattered cell death in both neonates and adults could be identified by a modest increase in T and decrease in ADC. A better understanding of the effects of development on ischemic responses and associated MRI changes will provide a basis for the improved diagnosis of mild or minor ischemic insults relevant to pediatric and adult stroke.

The extent of damage produced by a minor neonatal stroke followed by a diffuse HI two days later results in heterogeneous enhancement of T, ADC and histological injury near the lesion. Surgical procedures including mechanical head manipulation followed by HI also produced some enhanced heterogeneity of hypoxic-ischemic injury affirming the need for sham controls.