Animal models of pediatric abusive head trauma

Abstract

Background: Abusive head trauma (AHT), previously known as the shaken baby syndrome, is a severe and potentially fatal form of traumatic brain injury in infant children who have been shaken, and sometimes also sustained an additional head impact. The clinical and autopsy findings in AHT are not pathognomonic and, due to frequent obfuscation by perpetrators, the circumstances surrounding the alleged abuse are often unclear. The concept has evolved that the finding of the combination of subdural hemorrhage, brain injury, and retinal hemorrhages ("the triad") is the result of shaking of an infant ("shaken baby syndrome") and has led to the ongoing controversy whether shaking alone is able to generate sufficient force to produce these lesions.

Objective: In an attempt to investigate whether shaking can engender this lesion triad, animal models have been developed in laboratory rodents and domestic animal species. This review assesses the utility of these animal models to reliably reproduce human AHT pathology and evaluate the effects of shaking on the immature brain.

Results: Due largely to irreconcilable anatomic species differences between these animal brains and human infants, and a lack of resemblance of the experimental head shaking induced by mechanical devices to real-world human neurotrauma, no animal model has been able to reliably reproduce the full range of neuropathologic AHT changes.

Conclusion: Some animal models can simulate specific brain and ophthalmic lesions found in human AHT cases and provide useful information on their pathogenesis. Moreover, one animal model demonstrated that shaking of a freely mobile head, without an additional head impact, could be lethal, and produce significant brain pathology.

Keywords: Abusive head trauma; Animal models; Neuropathologic changes; Pathogenetic mechanisms.

Fig. 1

Marked neuronal APP immunopositivity in the cervical spinal cord (Fig. 1) and retinal ganglion cell layer (Fig. 2). Injured ganglion cell axons (Fig. 2) are also APP-immunopositive (arrows). (Original magnification ×10 (Fig. 1) and ×40 (Fig. 2))

Fig. 2

Marked neuronal APP immunopositivity in the cervical spinal cord (Fig. 1) and retinal ganglion cell layer (Fig. 2). Injured ganglion cell axons (Fig. 2) are also APP-immunopositive (arrows). (Original magnification ×10 (Fig. 1) and ×40 (Fig. 2))

Fig. 3

Numerous APP-immunopositive damaged axons in the hemispheric white matter (higher power in the inset). (Original magnification ×10 (×40 inset))

Fig. 4

Widespread BBB breakdown in the cerebral cortex (A), cerebellar folia (B), brainstem (C), and cervical spinal cord (D) using immunohistochemical labeling of extravasated albumin as a surrogate marker of increased vascular permeability. (Original magnification ×4)