Magnetic resonance imaging for follow-up monitoring of decompressive craniectomy in experimental "malignant" hemispheric stroke

Acute ischemia in the complete territory of the carotid or the middle cerebral artery may lead to cerebral edema with raised intracranial pressure and progression to coma and death. Although clinical data suggest benefit for patients undergoing decompressive surgery for massive space occupying hemispheric stroke, little data about the effects of this procedure on morbidity and outcome is available. The experimental data support an early surgical approach. For early and probably most effective treatment of severe, space-occupying cerebral ischemia, the "malignant" character of the brain edema has to be recognized early after onset of vessel occlusion. Hereby magnetic resonance imaging (MRI) may allow to determine the clinical significance of brain edema early after onset, simultaneously allowing to monitor the evolution of ischemia. We performed serial SE-MRI in rats with acute hemispheric infarctions treated by decompressive craniectomy. Focal cerebral ischemia was induced in 36 rats using an endovascular occlusion technique. Decompressive craniectomy was performed 4 and 24 hours after vessel occlusion in groups of 12 animals each. Twelve animals were not treated by decompressive craniectomy (control group). Four, 24, 48, 72 and 168 hours after MCAO all animals were examined with conventional T1- and T2-weighted SE-MRI. Shift of the midline structures and compression of the ventricles were scored. Changes in weight and neurological performance were measured daily. The infarction volume was calculated by triphenyltetrazolium chloride staining 168 hours after MCAO. While mortality in the untreated group was 33,3%, none of the animals treated by a decompressive craniectomy died (mortality 0%). Neurological behaviour, weight loss and infarction volume were significantly better in the animals treated by early decompressive craniectomy (p < 0.05). Four hours after MCAO all untreated animals showed a massive shift of the midline structures and a massive compression of the ventricles; only 7 of 12 animals treated early by craniectomy showed mild mass effects. Correlation of the histological brain damage with T2-weighted MRI 4 hours after MCAO was poor (r = 0,41); later than 24 hours there was a good correlation (r > 0,7). Our results suggest that decompressive craniectomy in malignant cerebral ischemia reduces mortality and significantly improves outcome. If performed early after vessel occlusion, it also significantly reduces infarction size. In the acute phase of hemispheric infarction conventional SE-MRI is not sensitive in estimation of infarction size. Later than 24 hours, conventinal SE-MRI proved to be useful in monitoring brain edema and infarction size in this rat model of malignant hemispheric stroke.