In a tightly controlled and well-designed study, Li et al. (2011) evaluated whether hypothermia and magnesium, given alone and when combined, reduces hippocampal CA1 sector damage when administered after global ischemia in rats. Postischemic hypothermia is widely known to reduce ischemic injury, and the authors' previous work shows that adding magnesium augments hypothermic neuroprotection in global and focal ischemia models. Thus, in this study, the authors had expected magnesium to improve CA1 cell survival when added with very mild hypothermia (35°C). However, use of 24 hours of hypothermia, a 2-day treatment with magnesium, and the combination treatment all failed to lessen injury significantly when initiated 4 hours after ischemia. A second experiment expanded upon these findings by also comparing 33°C and 35°C treatments and initiating them at 2 hours postischemia. The earlier interventions resulted in modest protection, but there was no significant benefit from adding magnesium with hypothermia. Unfortunately, the question of whether magnesium could provide additive or synergistic protection may have been missed, as the treatments provided quite limited protection (∼20%) on their own. Furthermore, a recent meta-analysis of combination treatments in focal ischemia models emphasizes the importance of having a “solid therapeutic foundation” to start with (O'Collins et al., in press). In other words, had one of the treatments provided significant protection against global ischemia then the other is more likely to add to that, at least up to a point (ceiling effect).
The modest level of protection, especially for hypothermia, stands in contrast to many earlier studies (MacLellan et al., 2009) and, as the authors argue, it was likely due to using a severe ischemic insult. The latter is due to tight physiological controls that differ from earlier work (e.g., choice of anesthetic, slight changes in temperature and blood pressure targets). It should be kept in mind, however, that greater durations of cooling, such as 48 hours, might have provided significantly better protection even at the late intervention delays after severe global ischemia (Colbourne et al., 1999). Regardless of the explanation(s), the Li et al. study nicely demonstrates that neuroprotection can vary considerably with slight changes in model parameters. Investigators ought to be concerned, especially when efficacy with the gold-standard treatment, prolonged hypothermia, can be lost after relatively minor model changes.
