Neuroprotection with delayed initiation of prolonged hypothermia after in vitro transient global brain ischemia

Prolonged therapeutic hypothermia (32-34 degrees C for 12-24 h) improves the functional outcome of comatose cardiac arrest survivors. It is generally believed that rapidly achieving target temperature optimizes neuroprotection. However, this hypothesis has not been tested systematically. In this study, we compared the neuroprotective effect of prolonged hypothermia initiated between 0 and 8 h after reoxygenation using an in vitro model of simulated global brain ischemia. Organotypic hippocampal slices were prepared from 5-day-old Wistar rat pups and cultured for 1 week prior to analysis. Ischemia was simulated by normothermic oxygen-glucose deprivation (OGD). Hypothermia (33 degrees C) was initiated 0-8 h after reoxygenation and maintained until 24 h post-injury. CA1 regional cell death was quantified by propidium iodide (PI) fluorescence. Release of 14-3-3 beta protein was evaluated as a potential surrogate maker for neuroprotection. Hypothermia initiated 0, 1, 2, or 4 h after 30 min OGD reduced 24 h CA1 regional PI fluorescence by 47 +/- 34%, 85 +/- 4%, 88 +/- 3%, and 88 +/- 5% (P < 0.05 for all versus normothermic reoxygenation). Direct comparison of hypothermia initiated 4 or 8 h after reoxygenation revealed equivalent neuroprotection following 15 and 30 min OGD, but neither was protective after 60 min OGD. Hypothermia initiated 4 or 8 h after 30 min OGD reduced 14-3-3 beta release by 73 +/- 11% and 92 +/- 4%, respectively (P < 0.01 for both versus normothermic reoxygenation). In this model, the neuroprotective effect of prolonged post-ischemic hypothermia is both optimal and equivalent when initiated between 1 and 8 h after reoxygenation. These results suggest the need for further in vivo studies to define the therapeutic window within which prolonged hypothermia is optimally neuroprotective after cardiac arrest.