THE observation that brief episodes of ischemia in the heart, occurring before a subsequent longer interruption of blood flow, provides protection against dysfunction and necrosis has been termed ischemic preconditioning . 1The protection is well described in a variety of animal models as well as in clinical settings, and it is not a trivial effect. In models of stunned myocardium in which dysfunction persists for hours or days after ischemia/reperfusion, preconditioning can virtually prevent contractile dysfunction. In models of infarction, the necrotic area within a region at risk can be reduced by 60–75%. Clinically, this can mean the difference between sustained inotropic support in the postoperative period or considerably greater functional capacity in patients after discharge. The study by Novalija et al. 2in this issue of ANESTHESIOLOGY continues the series of rather remarkable studies demonstrating that brief exposure to a volatile anesthetic, in this case sevoflurane, can mimic a brief ischemic insult and thereby precondition the myocardium, decreasing reperfusion damage and dysfunction.
The preconditioning protection observed with brief ischemia seems to be mediated by release of adenosine—it can be duplicated by adenosine administration, 3prevented by blockade of adenosine receptors 4and by inhibition of 5′-nucleotidase, 5which is responsible for generation of adenosine. Adenosine binds to its receptor (A1 and possibly A3), and via a G-protein–linked process, increases protein kinase C (PKC) activity. The resulting phosphorylation of the adenosine triphosphate (ATP)-sensitive K channel (KATP) results in the channel being less sensitive to inhibition by ATP. 6Physiologically, the KATPchannel opens when intracellular ATP stores are depleted, permitting K+to flow out of the cell, thus restoring the resting membrane potential and decreasing activity. This channel plays an important role in regulating the tone of vascular smooth muscle by causing hyperpolarization and relaxation when oxygen delivery results in decreased ATP production. In the heart, the KATPchannel is not normally active, but its sensitivity to inhibition by ATP is decreased with PKC activation. When KATPchannel activity is increased, the cardiac action potential shortens, accompanied by a mild negative inotropic action and remarkable protection against a subsequent sustained ischemic or hypoxic insult. Preconditioning can also be elicited by activation of a variety of ligand receptors (endothelin, δ-opiate, α-adrenergic) that increase PKC activity, as well as by drugs such as KATPchannel openers (e.g. , nicorandil or cromakalim). Of special relevance to anesthesiology, brief exposure to a volatile anesthetic can activate cardioprotection against a subsequent prolonged ischemia that is identical to ischemic preconditioning in that it can be inhibited by blocking KATPchannels or adenosine receptors. 7–12Similar effects of ischemia and isoflurane and sevoflurane in preconditioning have also been demonstrated in isolated human atrium. 13,14
Although initially attributed to KATPchannel effects on the sarcolemma, the remarkably profound protective effect exceeds the modest electrophysiologic changes. Furthermore, preconditioning actions are observed in the absence of alterations in electrophysiologic behavior. 15–17However, in addition to their location on the myocyte membrane, KATPchannels are located in the mitochondrial inner membrane, where they seem to regulate mitochondrial volume as well as the massive electrical and proton gradient that powers ATP synthesis. Preconditioning can be initiated by the opening of mitochondrial KATPchannels and prevented by their blockade. 18,19The model that emerges is one in which surface receptor activation turns on PKC activity, resulting in activation of mitochondrial KATPchannels to provide protection to myocytes. 17,20PKC activity is actually mediated by a large class of ubiquitous phosphorylating enzymes that have varying requirements for activity (Gproteins, phospholipids, diacylglycerol, and increased intracellular Ca2+). A recent study suggests that a particular isoform (PKC-δ) is the type that is translocated to the mitochondria to activate KATPchannels located there. 21Evidence is accumulating to document the functional role of KATPchannels in mitochondria, suggesting that channel activation leads to a decrease in the voltage gradient and a decrease in Ca2+accumulation. 20,22,23However, the exact pathway by which mitochondria and cells are protected remains to be defined.
Although demonstrating the cardioprotective effect of sevoflurane, the more newsworthy result in the article by Novalija et al. may be that this protection occurs not only in cardiac myocytes, but also extends to the endothelium of the coronary vasculature. A study dating back to the early 1990s demonstrated that a brief episode of ischemia also protects the functional integrity of the endothelium, demonstrating that the vasodilating capacity of the coronary vasculature was retained in hearts that were ischemically preconditioned. 24The ischemic preconditioning of the endothelium also seems to be mediated, at least in part, by adenosine receptors and KATPchannels. 25In addition, further studies have demonstrated that structural integrity of endothelial cells is better maintained after preconditioning. 26It is interesting that these structural studies of endothelial cells subjected to ischemia/reperfusion show marked mitochondrial swelling, an effect not observed in preconditioned endothelium. 26In addition, the structural evidence of protection seemed to last for up to 1 month. Further studies are required to demonstrate more precisely how volatile anesthetic preconditioning compares with ischemic preconditioning of both myocytes as well as the endothelium.
One of the major features of endothelial protection by brief ischemia or anesthetics is the ability to generate nitric oxide and mediate vasodilation. 2,25The presence of nitric oxide is not only important for regulating vascular tone, but also for its ability to prevent leukocyte adhesion and migration into reperfused tissues. Endothelial nitric oxide production can prevent recruitment of polymorphonuclear leukocytes (neutrophil) into ischemic regions. 27–29Because neutrophil accumulation and infiltration clearly contributes to the postischemic “no reflow” phenomenon, contractile dysfunction, and myocardial necrosis, prevention of their accumulation by maintained endothelial integrity is critically important. In addition, ischemia can induce expression of a variety of cell surface markers (P-selectin, intercellular adhesion molecule-1) and inflammatory mediators (tumor necrosis factor-α) that also contribute to neutrophil accumulation. 30If endothelial protection by anesthesia includes the prevention of the expression of the cell adhesion molecules such as P-selectin, the endothelial protection provided by the anesthetics during ischemia may have profound implications with regard to maintaining vascular integrity during the stressful period of reperfusion. Although there are conflicting data concerning the role of free radicals as well as the exact cellular biochemical pathways involved, the studies with regard to anesthetics suggest that there may be remarkable protection provided by these agents.
The good news for anesthesiologists is that volatile agents that we routinely use seem to provide a significant protective effect, not only on the myocardium, but on the vascular endothelium. If endothelium in other vascular beds shows similar degrees of protection, then the use of volatile anesthetics may provide important protection for a far wider variety of tissues. Over the next few years we can look forward to more detailed explanations of the pathways of anesthetic preconditioning, as well as the extent to which other tissues share the beneficial effects observed in the myocardium. Considerable effort will no doubt be expended to develop pharmacologic means to maximize protection, perhaps seeking other drugs that can provide the similar kind of protection provided by volatile anesthetics. In the mean time, we can be assured that at least certain anesthetic agents seem to precondition and protect, but much work remains to be performed to define fully the extent of protection.