DO anesthetics protect the brain from ischemic injury? The answer to a seemingly simple question has eluded researchers for more than a quarter of a century. The contribution by Sakai et al. 1in this issue of Anesthesiology suggests that we may finally have an answer to at least part of the long-standing controversy—at least in rats.
Working with laboratory rats, Hiroaki Sakai and coworkers in David Warner's laboratory at Duke University show conclusively that isoflurane is neuroprotective during focal cerebral ischemia and that, in distinction to several other influential studies,2,3the protection from isoflurane is long-lasting, evident for a month after the experimental stroke.
To understand the significance of the article by Sakai et al. , some history of investigations regarding anesthetic neuroprotection is in order. Disputes regarding whether clinical anesthetics confer neuroprotection in experimental models of brain ischemia date to the late 1960s. John Michenfelder at the Mayo Clinic argued that because even very-high-dose barbiturates do not reduce brain metabolism more than does brain ischemia, barbiturates should not be protective after cardiac arrest or global ischemia.4Correspondingly, clinical trials in human cardiac arrest were negative,5but barbiturates did seem to be beneficial in experimental focal ischemia.6As time passed, numerous studies pro and con appeared, examining a wide range of different models and anesthetic agents. The controversy continued through the 1980s, when it became clear that even small changes in brain temperature during or after ischemia had a major impact on outcome, with hypothermia being protective and mild hyperthermia being deleterious. A new wave of experimentation with volatile anesthetics ensued, based on the knowledge that hypothermia was a confounding variable in previous neuroprotection studies with these agents. Of note, similar issues vexed studies on human neuroprotection: barbiturate neuroprotection after coronary artery bypass graft surgery was reported by Nussmeier et al. ,7only to be refuted by Zaidan et al. 8when postbypass hypothermia was prevented. However, in experimental ischemia, agents like isoflurane remained viable neuroprotectants even when intraischemic and postischemic brain temperature was carefully controlled.9But doubts persisted, and neuroprotection advocates10continued to clash with critics.11Soon the pendulum swung decidedly in the negative direction, when the durability, rather than the potency, of anesthetic protection came into question. Seminal studies by Kawaguchi and others found that although isoflurane decreased the degree of injury present several days or a week after the ischemia, animals given isoflurane fared just as poorly as controls when examined several weeks to a month afterward.2,12That is, isoflurane protection fades. That anesthetic neuroprotection is transient was further supported by work of Elsersy et al. 13and Bayona et al. 14These investigations revealed that in rodent forebrain ischemia and focal ischemia models, respectively, anesthetic protection was not sustained.
Clearly, the results of these previous investigations differ from those of Sakai et al. , and an explanation of this discrepancy requires a close examination of the experimental model that was used. Sakai's control group for the stroke treatment notably involved awake rats—an experimentally difficult preparation involving intensive care and observation of study animals. All previous studies performed with modern standards of blood pressure control, adequate respiratory monitoring, and preservation of normothermia used nitrous oxide-fentanyl–anesthetized rats in the control group, possibly obscuring the benefits of the inhalation anesthetic.
Another important feature of the Sakai study is that the experimental ischemia model involved a 50- or 80-min temporary, rather than permanent, occlusion of the middle cerebral artery. This injury, although severe, was not as severe as the permanent occlusion group used by other investigators including Kawaguchi and others. As suggested by David Warner, perhaps these other studies were asking too much of any potentially neuroprotective agent to protect against permanent middle cerebral artery occlusion.15In fact, the work of Christian Werner's group indicates that volatile agent neuroprotection can be sustained provided the injury is of mild to moderate severity.16It seems clear that given the right circumstances, anesthetics can achieve enduring neuroprotection.
The current studies raise several questions. One of the more interesting is whether neuroprotection is intrinsic to the state of anesthesia or is dependent on some particular quality of isoflurane distinct from its capacity to produce unconsciousness and prevent movement from a noxious stimulus. As noted, the use of an awake control group experiencing ischemia was probably critical to the outcome. Further studies with different anesthetic agents and with mechanism-based examinations of neuroprotective actions should be able to reveal the answer to this question.
An important question is whether the findings of Sakai et al. will renew interest in testing anesthetics in human neuroprotection trials, an effort that had clearly waned after the failure of countless clinical studies of stroke neuroprotection. An examination of preclinical data demonstrating the neuroprotective efficacy of mild hypothermia may provide some perspective. In a variety of models of ischemia, and in a variety of species, mild hypothermia was shown to be profoundly neuroprotective. Moreover, this protection was demonstrated months after the ischemic injury.17Based on these indisputable results, clinical trials of mild hypothermia in head-injured patients were initiated; the failure of hypothermia to improve outcome18in these patients was met with disappointment. Similarly, in patients undergoing intracranial aneurysm clipping, hypothermia did not provide any benefit.19If an intervention that has, arguably, shown effective neuroprotection in preclinical studies fails in clinical trials, what then are the prospects for demonstrating the neuroprotective efficacy of anesthetic agents given the disparate findings of anesthetic protection in preclinical investigations? Clearly, much more work is needed before isoflurane can be evaluated as a neuroprotectant, especially considering the large and expensive clinical trial that necessarily lies ahead. In today's environment of outcomes research driving evidence-based clinical practice, whether it is even possible to obtain good evidence for anesthetic neuroprotection in humans is a matter for debate. But the study by Sakai et al. provides strong rationale for such clinical research.
Is the story of anesthetic neuroprotection nearing its final chapter? Although the contribution of Sakai et al. is a major step in neuroprotection in the laboratory, history teaches us that much more remains to be written in the clinic.
* Department Anesthesia and Perioperative Care, University of California at San Francisco, San Francisco, California. bicklerp@anesthesia.ucsf.edu. † Department of Anesthesiology, University of California at San Diego, San Diego, California.
