Determiners of Fatal Reperfusion Brain Oedema

Summary

Brain oedema is an important aspect of infarction from cerebrovascular occlusion. In a cat stroke model where the middle cerebral artery (MCA) was reversibly or permanently occluded, we analyzed the incidence of fatal hemispheral oedema in 35 normo- (6 mM) and 35 hyperglycaemic (20 mM for 6 hours) animals, with (N = 45) and without (N = 25) restoration of blood flow with clip release at 4 and 8 hrs of occlusion. Fatal hemispheral oedema occurred in 23% of cats (16/70) while hyperglycaemia, for one, and restoration of blood flow, for another, each quadrupled its occurrence. Further, evidence of remote oedema in the form of posterior cingulate cortical pressure atrophy from transtentorial herniation was found in animals that were allowed to survive for 2 weeks and that exhibited infarcts that affected 12 to 95% of the MCA territory. Thus, hemispheral oedema in association with MCA occlusion developed sufficiently markedly as to cause transtentorial herniation in 47% of all cats (33/70).

We carried out biochemical analyses in 14 hyper- and 10 normoglycaemic cats after 4 hrs of MCA occlusion for ATP, phosphocreatine (PCr), lactate, glucose and glycogen. The biochemical findings then were correlated with the occurrence of reperfusion oedema following clip release after 4 hrs of occlusion point-by-point in the brains. Linear regression analyses of the brain metabolic and pathologic data revealed highly significant (p < 0.001) correlations of acute oedema with brain tissue ATP and PCr reductions1.5 < µM/ g, with lactic acid accumulation > 20 µM/g and with the extents of reduction in brain tissue glucose concentrations in the ischaemic territories.

Fatal reperfusion oedema developing following clip release after 4hrs of occlusion could be predicted with high reliability (r = 0.96, p < 0.001) from the above biochemical data obtained after 4 hrs of MCA occlusion using multivariate analyses based on partial least squares (PLS) regressions. This results suggests that MRI spectroscopy in vivo data may be usable to guide clinical ischaemic brain reperfusion procedures.