We examined cerebral lipid peroxidation, estimated by a thiobarbituric acid test, in rat brain regions after 30 minutes of severe forebrain ischemia and at recirculation periods of up to 72 hours. The lipid peroxide levels remained unaltered in all brain regions during ischemia and during the first hour of recirculation but were selectively increased between 8 and 72 hours of recirculation in the ischemia-sensitive regions of the hippocampus, striatum, and cortex. The most pronounced increases (30-37%) were seen at 48 hours of recirculation. In contrast, lipid peroxide levels were unchanged in infarcted brain regions 24 hours after intracarotid injection of microspheres, indicating that reoxygenation of the ischemic brain is a prerequisite for lipid peroxidation. We assessed the lipid peroxidation capacity of cerebral homogenates obtained from rats subjected to ischemia and recirculation by measuring the production of lipid peroxides after aerobic incubation. The homogenates from rats exposed to 30 minutes of ischemia or to 1 hour of recirculation were not more susceptible to peroxidation. However, the production of lipid peroxides was selectively increased in the hippocampus, striatum, and cortex at 8-48 hours of recirculation, suggesting a loss of efficacy of the antioxidant systems. These results, showing a delayed and long-lasting increase in lipid peroxidation that occurs in ischemia-sensitive brain regions and parallels the development of neuronal necrosis, support the hypothesis that free radical processes participate in postischemic neuronal damage. (Stroke 1989;20:918-924) I t has been suggested that cellular damage in cerebral ischemia is at least partly due to oxidative damage, notably that caused by free radical formation and lipid peroxidation.1 However, definitive proof of its occurrence is still controversial. Several laboratories have provided evidence that lipid peroxidation occurs in vivo either during or after brain ischemia and reperfusion, 2 -8 but other data have failed to support the hypothesis.9 -12 The controversy derives from the different methods that have been used for the detection and analysis of lipid peroxidation or from the use of different models of brain ischemia.In our study, we investigated lipid peroxidation using the four-vessel occlusion model of transient ischemia, 13 which produces severe forebrain ischemia and leads to delayed neuronal necrosis in 17 Regional blood flow was therefore measured during and after vascular occlusion to record precisely the ischemic conditions. We assessed the extent of lipid peroxidation by measuring the level of thiobarbituric acidreactive substances (TBARS) in various cerebral structures at different times after the induction of ischemia. Simultaneously, we investigated the capacity of cerebral lipid peroxidation by measuring the accumulation of TBARS in brain homogenates incubated under aerobic conditions. Materials and MethodsWe performed the study with the four-vessel occlusion model on 98 male Wistar rats (Iffa Credo, France), wei...