Selective vulnerability of hippocampal neurons in acceleration-induced experimental head injury.

Traumatically induced subtotal hippocampal neuronal loss traditionally has been considered a consequence of intracranial hypertension and impaired cerebral perfusion. We have examined the frequency and distribution of hippocampal lesions in an acceleration model of brain injury in 54 anesthetized nonhuman primates undergoing physiologic monitoring and subjected postinjury to comprehensive neuropathologic examination. Hippocampal lesions occurred in 32/54 animals (59%). These lesions always involved the CA-1 hippocampal subfield and were bilateral in 24 animals. Hippocampal involvement was not associated with marked elevation of intracranial pressure or depression of cerebral perfusion pressure. These lesions occurred in the absence of involvement of other brain regions considered selectively vulnerable to hypoxic insults. Hippocampal damage occurred in 46% of animals with mild injury characterized by brief periods of unconsciousness and no residual neurologic deficit. Ninety-four percent of animals with severe injuries and prolonged posttraumatic coma had hippocampal involvement. Traumatically induced selective neuronal necrosis of the hippocampus is a specific lesion not explained by the conventional mechanistic theories of head injury. An alternative hypothesis, such as excitotoxicity involving glutamate or other neurotransmitters, may account for the lesions demonstrated in this study.