Differential effects of hypothermia on neurovascular unit determine protective or toxic results: Toward optimized therapeutic hypothermia.

Therapeutic hypothermia (TH) benefits survivors of cardiac arrest and neonatal hypoxic-ischemic injury and may benefit stroke patients. Large TH clinical trials, however, have shown mixed results. Given the substantial pre-clinical literature supporting TH, we explored possible mechanisms for clinical trial variability. Using a standard rodent stroke model (n = 20 per group), we found smaller infarctions after 2 h pre- or post-reperfusion TH compared to 4 h. To explore the mechanism of this discrepancy, we used primary cell cultures of rodent neurons, astrocytes, or endothelial cells subjected to oxygen-glucose deprivation (OGD). Then, cells were randomly assigned to 33 degrees, 35 degrees or 37 degrees for varying durations after varying delay times. Both 33 and 35 degrees TH effectively preserved all cell types, although 33 degrees was superior. Longer cooling durations overcame moderate delays to cooling initiation. In contrast, TH interfered with astrocyte paracrine protection of neurons in a temperature-dependent manner. These findings suggest that longer TH is needed to overcome delays to TH onset, but shorter TH durations may be superior to longer, perhaps due to suppression of astrocytic paracrine support of neurons during injury. We propose a scheme for optimizing TH after cerebral injury to stimulate further studies of cardiac arrest and stroke.