| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of Anesthesiology, Neurobiology, and Surgery, Duke University Medical Center, Durham, North Carolina
Address correspondence and reprint requests to David S. Warner, MD, Department of Anesthesiology, Box 3094, Duke University Medical Center, Durham, North Carolina 27710. Address email to warne002{at}mc.duke.edu
In this issue, Kawaguchi et al. (1) make two important contributions to our understanding of the effects of anesthetics on ischemic brain. First, they confirm their previous finding that isoflurane is only transiently protective against a severe focal ischemic insult (i.e., protection is evident at 2 days postischemia but completely dissipated at 2 wk postischemia (2). Second, they provide evidence that the decay of isoflurane neuroprotection is likely mediated by apoptosis. Thus, although isoflurane is capable of inhibiting immediate necrotic cell death, it fails to provide defense against delayed apoptotic (programmed) cell death with the end result being no protection. To understand the significance of these findings, it is necessary to review the evolution of research into therapeutics for ischemic brain injury.
Pharmacologic treatment of ischemic brain injury has been under investigation for four decades (3). Despite this, we still have little solid evidence that such therapy is effective. Within our own discipline, study has focused on the role of anesthetics. Early studies, performed predominantly by anesthesiologists and neurosurgeons, exploited the fact that anesthetics reduce cerebral metabolic rate. The principal goal was to reduce the incidence of perioperative neurologic morbidity. However, as basic scientists began to unravel the molecular pathobiology of cerebral ischemia, many nonanesthetic drugs became therapeutic candidates. This was associated with increasing interest in the neurology community to pharmacologically treat stroke. As a result, there was a paradigm shift. Early work focused on neuroprotection, i.e., prevention of injury from anticipated intraoperative ischemic insults (e.g., carotid endarterectomy, cerebral aneurysms clipping, or cardiopulmonary bypass), for which pretreatment with sedative pharmacologic agents was feasible. But as the field shifted focus to treatment of spontaneous stroke and the inherent delayed entry of such patients into medical care, advances in experimental therapeutic strategies shifted to neuroresuscitation, i.e., identification of nonsedative pharmacologic interventions that would be effective even when given several hours after ischemia onset.
Parallel to this paradigm shift was the development and refinement of numerous (mostly rodent) models of ischemic brain injury (4–7). The severity of the ischemic insult is typically titrated so that all animals sustain major injury. This allows efficient screening of pharmacologic interventions because all animals can be subject of study regarding efficacy of the compound in reducing histologic and neurologic damage and also because these severe insults model the duration of spontaneous stroke. Anesthesiologists rapidly adopted these model systems to conduct perioperative neuroprotection research. As a result, the experimental question shifted. Instead of asking if anesthetics can reduce the incidence of ischemic injury (8), the question became "Can anesthetics reduce the severity of injury when it occurs?" This is an important difference because unlike the spontaneous stroke population, in which a large fraction of patients have neurologic deficits, the typical intraoperative ischemic insult is likely less severe because few patients actually sustain neurologic injury. This suggests that the duration of ischemic insult used for the study of spontaneous stroke may be inappropriate for the study of intraoperative ischemic insults. The same argument can be extended to out-of-hospital versus intraoperative cardiac arrest, i.e., two circumstances that are likely to have very different durations of ischemic insult.
Current theory is that anesthetics protect by the same mechanism that they provide anesthesia, i.e., antagonism of excitatory neurotransmission and potentiation of inhibitory neurotransmission. Conveniently, these two actions simultaneously reduce brain energy requirement. However, these two mechanisms are only relevant during the ischemic insult. For example, while halothane produces robust protection during focal ischemia (9), it fails to provide any protection when given after ischemia (10). Similarly, intraoperative thiopental, although shown to provide at least short-term protection in cardiac valve surgery patients (11), had no benefit when given to humans postcardiac arrest (12). Thus anesthetics would be expected to do no more than increase the time to energy failure after onset of ischemia. This, presumably, would reduce the time available for triggering irreversible ischemic injury and by extension should increase the available time for operative procedures that might by necessity cause temporary arrest of either global or regional blood flow. Oddly, over the past 20 yr there has been essentially no effort to define this potential for anesthetics (or other drugs) despite advances in animal modeling of ischemic insults that make such inquiry feasible.
The findings of Kawaguchi et al. (1) are not unique to isoflurane. Short durations of postischemic hypothermia (3–4 h) provided protection evident during the first few days after severe forebrain ischemia but by 2 mo postischemia, hypothermic animals had the same magnitude of injury as did normothermic controls (13). This confusion was resolved with the observation that postischemic moderate hypothermia provides sustained protection in rodents, but only if hypothermia is maintained >12 h (14). This was recently confirmed in comatose humans resuscitated from out-of-hospital ventricular fibrillation (15). Despite this advance for hypothermia, pharmacologic therapy has lagged behind. Like isoflurane, almost all purported neuroprotective/neuroresuscitative compounds have failed to provide protection against severe ischemic insults when outcome was assessed after long postischemic intervals (16–18) [but see (19)]. Of note, there have been no long-term outcome studies in animals or humans that have demonstrated sustained benefit from barbiturates, etomidate, or propofol, regardless of duration of the ischemic insult.
With respect to the role of apoptosis identified by Kawaguchi et al. (1), an important lesson was learned from a study that looked at the relationship between focal ischemia duration and infarct evolution. Du et al. (20) subjected rats to either 30 or 90 min of middle cerebral artery occlusion. Infarct size was assessed at 1, 3, or 14 days postischemia. The 90 min ischemia group had large infarcts at 1 day that remained large over the subsequent 2 wk. In contrast, the 30 min ischemia group had no infarct at 1 day, but over the ensuing 2 wk, infarct size progressively increased to be the same size as the 90 min group. Treatment of the 30 min rats with cycloheximide, a protein synthesis inhibitor, prevented infarct formation implicating inhibition of apoptosis. This is consistent with the proposal that there are unique ischemia duration thresholds for the initiation of apoptotic versus necrotic responses to ischemia.
The data of Kawaguchi et al. (1), who studied 70 min of ischemia, can be interpreted in two different ways. One possibility is that isoflurane directly inhibits necrotic responses to focal ischemia but fails to inhibit apoptosis that is concurrently stimulated, with the net result being no protection. Alternatively, isoflurane may attenuate the severity of the ischemic insult, such that the brain responds as if it was exposed to a duration of ischemia sufficient to stimulate apoptosis but insufficient to stimulate necrosis, as was shown by Du et al. (20). Accordingly, Kawaguchi et al. (1,2) have suggested that isoflurane may act as a bridge, abating necrotic responses to ischemia, until an anti-apoptotic drug can be administered. Although this holds experimental promise, there are no clinically available anti-apoptotic drugs. In contrast, anesthetics are available. If the insult is brief, presence of anesthetics alone might be sufficient to reduce the incidence of permanent neurologic morbidity.
The classic scenario for an intraoperative ischemic insult is temporary vessel occlusion during cerebral aneurysm clipping, although this could also be extended to carotid occlusion during endarterectomy, aeric emboli during cardiopulmonary bypass, or even brief periods of profound systemic hypotension. For such procedures, vascular occlusion times are typically much shorter than 70 min. Several studies have examined the limits of the duration of temporary occlusion before infarction occurs during cerebral aneurysm surgery (21,22). Lavine et al. (22) retrospectively reviewed such patients and found infarction to be rare with an occlusion duration <10 min but frequent with durations of more than 10 min. There was a suggestion, however, that the high frequency of infarction with >10 min occlusion was substantially reduced in those patients receiving IV "neuroprotective agents."
Thus, a clinical model is available for study of the relevance of anesthetics in altering the duration of ischemia required for infarct occurrence. Perhaps a randomized study eventually should be undertaken in humans. What the laboratory can now offer is the following: The threshold durations of ischemia required to initiate apoptotic versus necrotic responses should be defined in the absence of anesthetics. Then, systematic appraisal of the effect of different anesthetics and different doses of anesthetics can be made with respect to defining whether anesthetics can "buy surgical time" within the context of an ischemia duration that is clinically relevant. In other words, we cannot eliminate the possibility that anesthetics provide sustained neuroprotection against intraoperative ischemia on the basis of laboratory studies that study clinically irrelevant durations of ischemia.
Until such work is completed, the work of Kawaguchi et al. (1) has an important clinical implication. We currently have no evidence that anesthetics alone can provide meaningful long-term neuroprotection against cerebral ischemia. Although the use of anesthetics for neuroprotection poses little additional hazard, beyond that which is already encountered by their requisite use to provide anesthesia, reliance upon anesthetics to increase permanent tolerance of human brain to an ischemic insult is unjustified on the basis of existing scientific evidence. Study of clinically relevant durations of ischemic insults may provide a different conclusion.
References
This article has been cited by other articles:
|
|
 |
|
  I. Nasu, N. Yokoo, S. Takaoka, K. Takata, T. Hoshikawa, M. Okada, and Y. Miura The dose-dependent effects of isoflurane on outcome from severe forebrain ischemia in the rat. Anesth. Analg., August 1, 2006; 103(2): 413 - 8, table of contents. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bekker, R. Shah, D. Quartermain, Y.-S. Li, and T. Blanck Isoflurane preserves spatial working memory in adult mice after moderate hypoxia. Anesth. Analg., April 1, 2006; 102(4): 1134 - 1138. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
