In the UK, there are ∼20 000 deaths within 30 days of surgery every year, 9000 of which have a cardiac cause.93 The number of major cardiac complications is likely to be in the region of 150 000 per annum. As 60% of patients who die within 30 days of surgery suffer from coronary artery disease94 it is reasonable to assume that the majority of cardiac complications of anaesthesia and surgery results from myocardial ischaemia leading to myocardial injury (Fig. 1).
Myocardial responses to ischaemia
Acute myocardial ischaemia
The acute occlusion or progressive constriction of a coronary artery causes reduction or abolition of systolic shortening and thickening of the ischaemic wall.81 Ischaemic segments also demonstrate paradoxical wall motion (termed post‐systolic shortening or post‐systolic thickening). These functional changes relate directly to the severity of the reduction in coronary blood flow. As post‐systolic shortening and thickening occur after aortic valve closure, they do not contribute to ejection and result in an internal shift of blood in the ventricle, and may impair relaxation. Acute or progressive ischaemia of the left ventricle also cause an increase in chamber stiffness in the ischaemic and in remote non‐ischaemic segments.81 This generalized increase in myocardial chamber stiffness contributes to an elevation of the left ventricular end‐diastolic pressure, especially in the presence of volume loading.81 The increase in end‐diastolic pressure contributes to a vicious circle as it further impairs coronary blood flow by increasing diastolic wall tension. The mechanisms of the increase in ventricular chamber stiffness of remote, well‐perfused myocardium have not been elucidated. However, ventricular stiffening does not depend on loading conditions and is likely to result from the release of mediators.
Mechanisms of myocardial ischaemia
The time‐course of the effects of ischaemia on cardiac tissue is well known. There is a marked reduction of contractile function resulting from decreased ATP production a few seconds after the onset of ischaemia. Leakage of potassium ions is responsible for the alterations of ST‐segments. Within minutes, an intracellular acidosis develops associated with an increase in myoplasmic Ca2+ and the beginning of cell swelling. Later, cellular lesions become irreversible. The ultrastructure of the cells becomes altered and macromolecules (CK‐MB, troponins) are released. An increased concentration of cytosolic and mitochondrial Ca2+ plays a central role in the damage to the cells and their membranes104 (Fig. 2).
Myocardial ischaemia occurs in the presence of fixed or dynamic coronary artery stenoses and, in the case of the right ventricle, in response to afterload mismatch. The main causes of ischaemia with fixed coronary stenoses include tachycardia, excessive left ventricular filling, and hypoxaemia. Tachycardia, systolic hypertension and β‐adrenergic stimulation increase the oxygen requirements and may decrease oxygen delivery. Causes of ischaemia in the presence of dynamic stenoses include those described above and, in addition, activation of sympathetic and parasympathetic systems. Moreover, several endothelium‐derived mediators may enhance vasoconstriction.
In the normal heart, the role of the autonomic nervous system is overshadowed by local metabolic coronary vasoregulation. Normally, activation of the sympathetic nervous system at the level of α1‐ and β1‐adrenoceptors increases blood pressure, heart rate, and contractility. As a result, myocardial oxygen consumption increases and local vasoregulation decreases coronary vascular resistance. The direct effect of α1‐adrenoceptor stimulation, vasoconstriction, is minimal under normal circumstances but may be exaggerated in the presence of coronary artery disease, extreme exercise, or haemorrhagic shock. Activation of the parasympathetic system causes bradycardia and hypotension, thereby reducing myocardial oxygen requirements. Local regulation increases coronary vascular resistance and the direct coronary vasodilatation caused by acetylcholine is masked.
Endothelins act on vascular smooth muscle and are extremely powerful vasoconstrictors. However, they exert complex physiological effects. Through activation of endothelin B (ETB)‐receptors, endothelial nitric oxide synthase is activated and nitric oxide is released causing cGMP‐mediated vasodilatation. Activation of ETB‐receptors also increases the activity of cyclo‐oxygenase leading to the release of prostaglandin I2 (PGI2). The latter causes vasodilatation and minimizes smooth muscle cell proliferation. When the endothelium is normal, there is a delicate balance between vasoconstriction and vasodilatation in response to endothelins. The vasodilatory role of the endothelium becomes more apparent when it is damaged by atheroma, hypercholersterolaemia, hypertensive heart disease, and after reperfusion. An imbalance of mediators can then develop and facilitate vasoconstriction. There are several important mechanisms involved in this vasoconstriction. Norepinephrine causes α1‐adrenoceptor‐mediated vasoconstriction. When the endothelium is damaged, acetylcholine causes muscarinic receptor mediated vasoconstriction,79 instead of endothelium‐dependent vasodilatation. Thus, with endothelial damage both sympathetic and parasympathetic stimulation may cause coronary vasoconstriction. As sympathetic overactivity is part of the perioperative stress response, exaggerated vasoconstriction may be expected to contribute to perioperative myocardial ischaemia and cardiac damage. Similarly, the effects of endothelins are altered when the endothelium is damaged and activation of nitric oxide‐synthase and cyclo‐oxygenase does not occur. Consequently, only effects of endothelins in these circumstances are those on endothelin A (ETA)‐ and ETB‐receptors in vascular smooth muscle resulting in vasoconstriction and smooth muscle proliferation.
Myocardial stunning (flow‐contractility mismatch)
The term myocardial stunning was coined by Braunwald and Kloner14 in 1982 to describe a reduction in function after a brief period of ischaemia followed by reperfusion. The impairment of function could last for several hours or days at a time when coronary blood flow was normal and there was no obvious cellular damage. In clinical practice, interventions such as transluminal coronary angioplasty, coronary artery bypass graft surgery, and thrombolysis after myocardial infarction are human models of ischaemia‐reperfusion phenomena. During the perioperative period, a high proportion of adult patients suffer from episodes of myocardial ischaemia, most of which are silent but can be prolonged. Importantly, silent myocardial ischaemia is supposed to be a common cause of myocardial stunning and is a predictor of adverse cardiac outcome.
Mechanisms of myocardial stunning
Myocardial ischaemia followed by reperfusion causes reversible or irreversible damage depending on its duration (Figures 1 and 2). In stunning, ischaemic damage is, in principle, reversible.67 Three main mechanisms are involved in the establishment of stunned myocardium: formation of free oxygen radicals, accumulation of intracellular Ca2+, and degradation of contractile proteins. Many studies have shown that during ischaemia, but more importantly during reperfusion, considerable production of free oxygen radicals occurs.99 Free radicals do not have a single target, but adversely affect many components of the cell including sarcolemmal and subcellular membranes of organelles. The role of free radicals in stunning is confirmed by the improved post‐ischaemic functional recovery in the presence of superoxide dismutase.138 Production of free radicals involves xanthine oxidase, oxidation of catecholamines, uncoupling of mitochondrial respiration, and activation of neutrophils.
During ischaemia and the early phase of reperfusion, there is an increase in the concentration of intracellular Ca2+. Although disappearing in the late phase of reperfusion, Ca2+ overload can decrease the sensitivity of contractile proteins to Ca2+, thus diminishing the developed force.9 Ca2+ overload my result from altered characteristics of the Na+/Ca2+ antiport and from altered Ca2+ fluxes at the level of the sarcoplasmic reticulum. Such alterations in Ca2+ handling may be attributable to ischaemia‐induced intracellular acidosis. During the early phase of reperfusion, the H+/Na+ antiport is maximally stimulated. While the acidosis is progressively corrected, there is an increase in intracellular Na+ leading to a further increase in Ca2+.
Furthermore there is some evidence that translocation of heat‐shock proteins (Hsp‐27, αB‐crystalline) with covalent binding to myofibrils, together with degradation of contractile proteins such as troponin I, as evidenced in a transgenic mouse model overexpressing troponin I fragments,88 may be, at least in part, involved in the pathogenesis of myocardial stunning. Also, an increase in coronary vascular resistance and a reduction in vasodilator response were previously reported during reperfusion and may represent some sort of vascular counterparts of stunning in endothelial cells (‘microvascular stunning’).10 However, not all studies confirm this phenomenon.32
The concept of myocardial hibernation was put forward by Rahimtoola in 1985.110111 In the hibernating myocardium, ventricular function is diminished as a consequence of insufficient coronary blood flow (Fig. 1). However, this reduction is not necessarily permanent: an improved balance of supply and demand may augment myocardial function.778 The issue of myocardial hibernation is clinically important because the risk of adverse cardiac outcome in cardiac and non‐cardiac surgery increases with a reduction of the ejection fraction. If coronary revascularization increases the ejection fraction the risk of adverse cardiac outcome is likely to be reduced. The likelihood of improved function after coronary reperfusion can be predicted by the result of a dobutamine stress echocardiogram. If dobutamine worsens ventricular function (reversible ischaemia) coronary revascularization is likely to improve cardiac function.1
Mechanisms of myocardial hibernation
In hibernating myocardium, cardiac metabolism is downregulated. It has been proposed that abolition of contractility of hibernating cardiac tissue is attributable to chronic stunning caused by multiple episodes of severe ischaemia followed by repetitive reperfusion. Other experimental models suggest that hibernation occurs as a result of chronic low‐flow states. In either case, hibernating myocardium should be salvageable by restitution of an adequate coronary blood flow.
In 1986, Reimer and colleagues112 reported that a brief period of ischaemia decreased the rate of ATP depletion during a further period of ischaemia. Murry and colleagues89 reported that brief periods of ischaemia made the heart more resistant to infarction during a subsequent acute coronary occlusion, reducing the infarct size by 70–80%. This phenomenon, termed myocardial preconditioning is the most powerful means of achieving cardiac protection. Protection conferred by ischaemic and (most) pharmacological preconditioning protocols exhibits two windows: an acute memory phase that develops within minutes of the ischaemic stimulus and lasts only between 1 and 3 h (classic or early preconditioning), and a longer, more delayed, phase (late preconditioning) starting after 12–24 h and persisting for 2–4 days.156 Delayed protection by preconditioning is thought to be primarily attributable to alterations in gene expression. Myocardial preconditioning has been documented in all animal species in which it has been studied and in human cardiac tissue. Whether preconditioning facilitates the recovery of function in stunned myocardium is still debated.19101 However, delayed preconditioning, in contrast to early preconditioning, always confers protection against stunning.
Preconditioning can result from successive episodes of angina or silent myocardial ischaemia. Indeed, infarct size is known to be smaller, if it has been preceded by angina.6991100 Preconditioning also reduces the risk of ischaemia‐induced ventricular tachycardia and ventricular fibrillation.145 In an almost experimental situation, preconditioning can occur during coronary angioplasty where several temporary occlusions are applied and the effects of a longer period of occlusion are minimized. During coronary angioplasty, sequential occlusions cause less angina, smaller ST‐segment changes and lesser lactate production than the first 90 second occlusion.28 Also, delayed preconditioning can be observed after prolonged nitroglycerin administration in patients undergoing angioplasty 24 h later. Although ischaemic preconditioning, or agents that mimic preconditioning, elicit cardiac protection in patients undergoing cardiac surgery,51 the clinical use of preconditioning in cardiac surgery is controversial as some studies show benefits48148 while others do not,5972102 or even demonstrate deleterious effects.
Mechanisms of myocardial preconditioning
Ischaemic stimuli cause the release of stress mediators such as adenosine, bradykinin, norepinephrine, and opioids. The mechanisms of preconditioning involve several types of triggers and mediators156 (Fig. 3). Amongst them, adenosine A1‐ and A3‐receptors, bradykinin2‐receptors, δ1‐opioid receptors and α1‐adrenoceptors play an important role. Via G‐proteins, phospholipase C (PLC) and protein kinase C (PKC), these receptors act on mitochondrial and sarcolemmal KATP channels and Ca2+ channels.152153 As many mediators are involved in preconditioning, and many substances are capable of preventing it, a ‘summation hypothesis’ has been proposed by Downey and colleagues.3136 In order for ischaemic preconditioning or pharmacological preconditioning to occur, it is necessary to reach an activation threshold. This threshold represents the sum of the activity of several mediators. The role of lactate as a triggering mechanism is controversial, even though lactate can open KATP channels and increase the expression of heat‐shock proteins.465 Ischaemic preconditioning is caused by temporary occlusions, generally lasting 5 min, followed by reperfusion. A comparison of various numbers of cycles of ischaemia‐reperfusion shows that protection is best with 1–4 cycles. With six or more cycles protection is reduced und ultimately lost.47 While one cycle may be enough to trigger the release of protective substances, many more cycles may be unable to enhance further release of triggers. A high number of short ischaemic periods may start to cause cumulative ischaemic damage, thus decreasing the efficacy of preconditioning.47
Preconditioning reduces cardiomyocyte necrosis in vivo and in vitro. Apoptosis is known to increase with ischaemia. In ischaemia‐reperfusion, both ischaemia and reperfusion contribute equally to apoptosis.76 By contrast, necrosis occurs primarily during reoxygenation. Preconditioning effectively reduces necrosis and apoptosis. Opening of KATP channels together with modulation of Ca2+ homeostasis may explain why inhalation anaesthetics inhibit apoptosis in cardiomyocytes.3151
Role of adenosine
The role of adenosine in preconditioning has been well documented. Adenosine A1‐receptor activation plays an important role. These receptors are coupled with KATP channels68 via Gi‐proteins. Activation of adenosine receptors decreases the production of reactive oxygen species and attenuates myocardial stunning.92 The role of adenosine receptors in preconditioning is confirmed by the observation that adenosine A1‐receptor antagonists can block KATP channels, thereby preventing ischaemic preconditioning. In addition, preconditioning can be mimicked by adenosine A1‐receptor agonists.
Role of bradykinin
Bradykinin is an inflammatory stress mediator and a vasodilator. In some experimental models, an infusion of bradykinin has been shown to reduce ischaemic injury,1536 while bradykinin receptor antagonists negated the protection conferred by ischaemic preconditioning.1536
Role of opioids
Role of adrenergic receptors
Both α‐ and β‐adrenoceptors are involved in preconditioning. While preconditioning is induced primarily by β1‐adrenoceptors,34 β2‐adrenoceptor may play a role via activation of L‐type calcium channels. Brief episodes of ischaemia cause the release of norepinephrine in the myocardium, while exogenous α1‐adrenoceptor agonists may cause pharmacological preconditioning.132
Role of free oxygen radicals and nitric oxide
Free radicals cause myocardial damage during ischaemia‐reperfusion. However, treatment with small amounts of free radicals before an ischaemic insult can reduce infarct size in vitro,6 an effect that is abolished by free radical scavengers. Nitric oxide‐cGMP signalling is also important.253392152 Inhalation anaesthetics may modulate the activity of various isoenzymes of nitric oxide synthase (nNOS, eNOS, iNOS) as they are heterogeneously distributed in the myocardium.
A preischaemic increase in Ca2+ represents a second messenger in the development of ischaemic preconditioning,85 even though Ca2+ overload is a major contributor to cell damage. Short‐time administration of increased Ca2+ concentrations to myocardial tissue is an effective preconditioning stimulus, which can be inhibited by administration of Ca2+ channel blockers.141
Protein kinase C (PKC)
PKC transfers γ‐phosphoryls from ATP to hydroxyl groups of serine/threonine residues in proteins. This phosphorylation controls the function of many cellular effectors. PKC plays an important role in ischaemic and pharmacological preconditioning.50113134139 PKC activators can induce, while PKC inhibitors prevent preconditioning.150 Inhala tion anaesthetics may directly activate PKC. Importantly, preconditioning‐associated isoform translocation of PKC to subcellular targets is highly dependent on species, age, and on the type of preconditioning stimulus.
Role of ATP‐dependent potassium channels (KATP channels)
Ultimately, KATP channels hold the central role in ischaemic and pharmacological preconditioning153 (Figures 2 and 3). Sarcolemmal KATP channels were described by Noma in 1983.96 These channels open when ATP levels fall, allowing potassium efflux so causing membrane hyperpolarization and reducing the action potential duration. These changes decrease the open probability of voltage‐gated Ca2+ channels. The resulting reduction in Ca2+ concentration preserves ATP levels and reduces coronary vascular tone.118 The increase in extracellular potassium also facilitates coronary vasodilatation and increases blood flow to the ischaemic region.5 KATP channels mediate the response to hypoxia and the hyperaemic response to brief coronary occlusions. However, reduction of action potential duration does not correlate with the reduction in infarct size. Surface KATP channels were initially thought to mediate preconditioning. More recent evidence indicates that mitochondrial KATP channels play a pivotal role in mediating cardiac preconditioning.35153 Opening of mitochondrial KATP channels may optimize mitochondrial energy production, decrease mitochondrial Ca2+ overload, and prevent opening of mitochondrial permeability transition pores (Fig. 4).304383157 Numerous studies have confirmed the important role of these channels. The role of KATP channels in human preconditioning is evidenced by the observation that ischaemic preconditioning does not occur in patients taking sulfonylureas, as these agents block KATP channels.1321
Other beneficial effects associated with cardiac preconditioning
Glycogen depletion and lactate accumulation during ischaemic preconditioning periods play a role in myocardial protection. Indeed, transient exposure to lactate improves contractile recovery in rat heart.29 Entrapment of neutrophils and platelets in the coronary vasculature occurs in ischaemia. The protective effects of preconditioning also extend to the endothelium of the coronary vasculature and to the adhesion properties of platelets and leucocytes to these vessels. Ischaemic preconditioning reduces ICAM‐1 production and thus neutrophil entrapment.116 In turn, reduced neutrophil and platelet entrapment is associated with enhanced post‐ischaemic function.4445
Late preconditioning is mediated by inducible nitric oxide synthase129146 and can be elicited by nitric oxide donors.46130 It can also be triggered by heat stress, lipopolysaccharides (LPS), or monophosphoryl lipid A; these are known to trigger delayed endogenous protective mechanisms against myocardial ischaemia‐reperfusion injury appearing after 24 h and lasting for several days.11 Alterations of gene expression of protective and anti‐protective proteins along with KATP channel opening have been proposed as the main mechanisms for this delayed protection. The endocannabinoid system is involved in the protection conferred by LPS, in relation with nitric oxide production.74 Two endocannabinoids act through interaction with G‐protein coupled membrane receptors, namely CB1‐ and CB2‐receptors, which are present throughout the body.103 Endocannabinoids have been implicated in the inflammatory response. In isolated rat hearts, endocannabinoids acting through CB2‐receptors and nitric oxide, play a role in the protection conferred by heat stress against myocardial ischaemia.52 Hemin is an activator of the potent antioxidant enzyme heme oxygenase‐1 and may play a role in delayed protection.149 Enhanced expression of heme oxygenase‐1 has been observed during the recovery phase of porcine myocardial stunning.125 Indeed, in experimental studies, a significant attenuation of stunning, as evidenced by enhanced recovery of wall thickening was observed in animals pretreated for 1 week with hemin.143 Whether inhalation anaesthetics are capable of elicting late preconditioning is not yet clear.
Ischaemic preconditioning can be generated by short episodes of myocardial, limb, or visceral ischaemia. In the heart itself, preconditioning can develop in areas remote from the preconditioning ischaemic stimulus (for review see156). This remote preconditioning may involve the release of adenosine, bradykinin, norepinephrine, and activation of KATP channels. Systemic effects of localized ischaemic preconditioning have been reported. This raises the issue that regional ischaemia of non‐vital tissues might protect remote vital organs.37 In volunteers, transient remote ischaemia of one limb induced remote ischaemic preconditioning of the opposite limb as evidenced by preservation of endothelial function (estimated as extent of vasodilator response).66 Cytokines and other metabolic mediators are likely to play a role. The autonomic system may also be involved as well as modulation of platelet, endothelium, and leucocyte function.
Protective effects of anaesthetics against ischaemia
Anaesthetics and myocardial stunning
Most inhalation anaesthetics and high dose opioids confer protection, increasing the rate of myocardial recovery after reperfusion.58115 In 1988, Warltier and colleagues142 demonstrated that administration of halothane and isoflurane before ischaemia improved the speed of recovery of function after a brief (15 min) period of ischaemia. After 5 h, the functional recovery in the presence of inhalation anaesthesia was 100% vs 50% only in the controls.142 Since then, these observations have been confirmed repeatedly.22134 However, some studies of isolated heart preparations showed no protection by inhalation anaesthetics.97117 More recently, sevoflurane and desflurane were shown to confer cardiac protection.108109119133135 It is likely that protection by inhalation anaesthetics is attributable to pharmacological preconditioning of the heart, as in many studies the inhalational anaesthetic was given before ischaemia and reperfusion. Indeed, in some studies the administration of the inhalation anaesthetic was discontinued before ischaemia‐reperfusion and resulted in reduced infarct size.1824 Nonetheless, protective effects of inhalation anesthetics were also reported if inhalation anaesthetics were administered exclusively during the reperfusion phase.
Intravenous anaesthetics appear to confer less protection.2327144 Fentanyl and propofol appear to be equivalent.114 In the isolated heart, as opposed to the intact instrumented heart, propofol has been shown to reduce infarct size and cellular damage.707182 Propofol‐induced protection was not abolished by block of KATP channels. While the effects of opioids on infarct size have been well demonstrated, there are only a few studies of their effects on myocardial stunning.114144 In the isolated heart, high concentrations of fentanyl have been shown to offer significant protection.58 Protection was mediated by δ‐opioid receptors, adenosine A1‐receptors, PKC, and KATP channels.5657 The role of δ‐opioid receptors is supported by the abolition of their protective effect by naloxone.20122131 As wash‐out of opioids does not prevent their effect, they must act as preconditioning agents.275 It is further possible that opioids act beneficially via a reduction in adhesion and migration of neutrophils.128140
Anaesthetics and cardiac preconditioning
Many anaesthetic agents have been shown to reduce infarct size in experimental models. Not all anaesthetics have the same efficacy. There is greater reduction of infarct size by halothane, enflurane, and isoflurane in comparison with pentobarbital, ketamine‐xylazine, or propofol anaesthesia in rabbits. Dogs anaesthetized with barbiturates exhibit larger infarcts than their conscious counterparts.53 This may be explained by the observation that barbiturates competitively antagonize adenosine A1‐receptors, a pivotal signalling pathway in cardiac preconditioning. Indeed, adenosine receptor antagonists decrease anaesthesia‐induced preconditioning.24113 The same is true of PKC antagonists.134 Albeit not proven, it may be speculated that pharmacological preconditioning by inhalation anaesthetics may be of smaller magnitude than ischaemic preconditioning.18
Pharmacological preconditioning by inhalation anaesthetics appears to be primarily mediated by stimulation of adenosine receptors113 and activation of KATP channels.62113 Protective effects by inhalation anaesthetics also occur in the presence of cardioplegic protection. Ischaemic and anaesthetic‐induced preconditioning are not additive suggesting the same end‐effector,12 identified in many studies as KATP channels. Nonetheless, sevoflurane can potentiate late ischaemic preconditioning in an in vivo rabbit model.87 The signalling components involved in cardiac protection by inhalational anaesthetics, are similar to ischaemic preconditioning, but show distinct differences.26 Both sarcolemmal and mitochondrial KATP channels may mediate anaesthesia‐induced preconditioning as demonstrated in desflurane‐mediated preconditioning.135 Yet, mitochondrial KATP channels may play the more important role. Halothane partially blocks sarcolemmal KATP channels,113 while isoflurane does not. Anaesthesia‐induced preconditioning is clearly species dependent. Halothane preconditions in rabbit but not rat or human. Isoflurane preconditions in rabbit and human, but not rat (for review see155156).
Anaesthetics may also modulate the effects of ischaemic preconditioning. Several anaesthetic agents have direct effects on KATP channels (barbiturates) or have prominent physiological effects that are induced by KATP channels (isoflurane, halothane).17 Accordingly, ischaemic preconditioning is abolished by glibenclamide under ketamine‐xylazine anaesthesia but not pentobarbital anaesthesia. In a comparison of the effects of ischaemic preconditioning under pentobarbital, isoflurane, and ketamine‐xylazine anaesthesia, infarct size was not different in the absence of preconditioning, but the magnitude of infarct size limitation by ischaemic preconditioning was different depending upon the basal anaesthesia.38 In the presence of halothane anaesthesia, nicorandil given before ischaemia did not demonstrate protective effects, whereas ischaemic preconditioning did reduce infarct size. Yet, a KATP channel blocker prevented the combined effect of ischaemia and nicorandil.90 The complexity of modulatory effects of anaesthetics on cardiac preconditioning has been substantiated in a cellular model of simulated ischaemia. Modulatory effects of anaesthetics were demonstrated by the inhibition of diazoxide‐induced mitochondrial KATP channel opening by R‐ketamine, thiopental and pentobarbital. Conversely, urethane, 2,2,2‐trichloroethanol (a main metabolite of α‐chloralose) and fentanyl potentiated the channel‐opening effect of diazoxide. This potentiation could be blocked by chelerythrine, a specific PKC inhibitor. By contrast, S‐ketamine, propofol, xylazine, midazolam and etomidate do not affect mitochondrial KATP channel activity.152 These observations illustrate the complex interference of anaesthetics with ischaemic preconditioning and stress the concept of anaesthetics acting as modulators of cardiac preconditioning.
To date, there are few data on the possibility of inhalation anaesthetics conferring late preconditioning, whereas delayed protection by opioids is well established.121 Kehl and colleagues60 examined the effect of isoflurane administration 24 h before a 60 min coronary occlusion followed by 3 h reperfusion in a canine model. While isoflurane exerted early protection, there was no late protection. By contrast, delayed preconditioning was observed in a rabbit model.137
Finally, improved collateral blood flow may also play a role in the beneficial effects elicited by anaesthetics. Indeed, sevoflurane increases collateral flow, an effect not reversed by glibenclamide.64 In addition, halothane, isoflurane, and sevoflurane reduce the number of neutrophils4573 and platelets44 sequestered in the coronary vasculature after ischaemia. This may contribute to the observed beneficial effects. Inhalation anaesthetics further suppress the postischaemic expression of CD11b45 and thus decrease neutrophil adhesion to the endothelium.86 However, sevoflurane does not reduce the expression of glycoprotein IIb/IIIa, a platelet adhesion molecule involved in the platelet–endothelium interaction.44
In the absence of ischaemia, isoflurane causes opening of KATP channels, an effect blocked by sulfonylureas.61 This results in a reduction in infarct size in experimental animals.63 Isoflurane also decreases infarct size in an in vitro model of human myocardium.113 Sarcolemmal and mitochondrial KATP channels appear to be involved.61106113133136 Isoflurane increases the open probability of the sarcoplasmic KATP channel for a given ATP concentration.39 In a cellular model, isoflurane significantly enhanced the diazoxide‐mediated activation of mitochondrial KATP channels. This effect was completely blocked by chelerythrine (a PKC inhibitor). Pretreatment with inhalation anaesthetics potentiated the diazoxide‐mediated protection against ischaemia. Cardioprotection was unaffected by the sarcoplasmic KATP channel blocker HMR‐1098, but sensitive to modulation of nitric oxide and adenosine‐Gi signalling pathways.153 Administration of isoflurane before aortic cross‐clamping in patients undergoing coronary artery bypass surgery causes cardiac index to be higher after cardiopulmonary bypass with less changes in ST‐segments than in the control group. However, there were no differences in terms of arrhythmias.42 Thus, isoflurane may offer some additional protection to cardioplegia. These findings are consistent with the observation of lower (albeit not statistically significant) perioperative levels of CK‐MB and troponin reported by Belhomme8 when isoflurane is used. Moreover, isoflurane was found to increase 5′‐nucleotidase activity in atrial tissue indicating increased PKC activity.8
Sevoflurane reduces infarct size in dogs via opening of KATP channels.133 Preservation of myocardial blood flow through collateral circulation, observed with sevoflurane, is independent of KATP channels. In sepsis, ultrastructural changes in the myocardium have been documented and sevoflurane protected cardiac output in septic (caecal ligation and perforation) rats.124 Recently, the first clinical double‐blinded multicentre study has shown sevoflurane to protect myocardium and kidney in patients undergoing coronary artery bypass grafting.54 This study also visualized for the first time PKC translocation (predominantly isoforms δ and ϵ) to subcellular targets such as the sarcolemma, mitochondria, intercalated disks, and nuclei in response to sevoflurane. Moreover, the observed renoprotective effect of sevoflurane raises the intriguing possibility that systemically administered sevoflurane may confer multiorgan protection in high‐risk patients.
In isolated human atrial trabeculae, desflurane improved the recovery of isometric contraction after a 30 min period of anoxia. The preconditioning effect of desflurane was abolished by glibenclamide, 5‐hydroxydecanoate (5‐HD), DPX (an adenosine receptor blocker), phentolamine, and propranolol.41 These observations suggest that preconditioning by desflurane is mediated by mitochondrial KATP channels,41 adenosine A1‐receptors, and α‐ and β‐adrenoceptors. In contrast, selective block of sarcolemmal KATP channels did not reduce desflurane‐induced preconditioning, while it abolished anoxia‐induced preconditioning. Desflurane increases sympathetic activity in volunteers and releases catecholamines from myocardial stores in rat and human myocardium.40 Preconditioning by desflurane may thus be, at least partly, elicited by stimulation of the α/β‐adrenoceptor pathways.77
Pharmacological interventions by nonanaesthetic agents currently used for the prevention of perioperative ischaemia
Several classes of drugs have been proposed in order to reduce the risk of ischaemic complications of anaesthesia and surgery. However, based on current clinical data, only beta‐blockers, α2‐adrenoceptor agonists, and possibly statins may have the potential to affect perioperative cardiovascular outcome.
While nitroglycerin is used successfully in the treatment of myocardial ischaemia, there is no evidence that its prophylactic administration before anaesthesia and surgery decreases the risk of perioperative cardiac complications.127
Calcium channel blockers
Though effective in the management of ischaemic heart disease, Ca2+ channel blockers have never been shown to offer any protection against perioperative cardiac complications of anaesthesia and surgery.126127 This absence of protection seems surprising in view of the strong antioxidant effect of certain calcium channel blockers.147
These compounds facilitate the release of the coronary vasodilator adenosine by the ischaemic myocardium, thereby improving collateral blood flow toward the compromised area. Though promising results were obtained,80 development of the only agent tested in clinical trials, acadesine, was stopped.
There is renewed interest in the use of clonidine, dexmedetomidine, and (temporarily, as it is not currently being further developed for clinical use) mivazerol; these drugs reduce the level of sympathetic activity and make the circulation more stable. Clonidine decreases the risk of perioperative myocardial ischaemia95 and a recent meta‐analysis has also shown a reduction in the risk of adverse outcome.127 Mivazerol has been tested in a large multicentre trial and was shown to decrease the incidence of cardiac complications in vascular surgical patients but not in non‐vascular surgical patients,98 yet its development was stopped.
This drug is both a nitrate and a KATP channel opener. It is effective in the management of ischaemic heart disease and its associated dysrhythmias. It may prove useful in the perioperative prevention of cardiac complications.55 In clinical practice, nicorandil, a KATP channel opener and nitrate, is widely used in the treatment of angina. Nicorandil induces myocardial preconditioning. In isolated human heart muscle nicorandil conferred cardioprotection (improved recovery of function in a hypoxia‐reoxygenation model). This effect was abolished by ischaemic preconditioning.16 In a rabbit model early treatment with nicorandil (pre‐ischaemia) decreased infarct size, while nicorandil administration after ischaemia was ineffective. Ischaemic preconditioning reduced infarct size and the combination of ischaemic preconditioning and nicorandil showed efficacy intermediate between ischaemic preconditioning and before administration of nicorandil.49 The effect of nicorandil was blocked by 5‐hydroxydecanoate, a KATP channel blocker. Thus, nicorandil appears to protect by opening KATP channels, and to interact with ischaemic preconditioning. By contrast, nicorandil appears to offer additional protection when administered with isoflurane, in terms of functional recovery of the stunned myocardium.105
In a case‐controlled study, Poldermans and colleagues107 evaluated the effects of statins on perioperative mortality in patients undergoing major vascular surgery. In statin‐treated patients, the risk for perioperative mortality was ∼20% of that observed in non‐statin‐treated patients. The authors concluded that perioperative statin use may reduce perioperative mortality in high‐risk vascular patients.
These drugs, at present, occupy centre stage for cardiac prophylaxis because several studies have shown a reduction in the incidence of cardiac complications of anaesthesia and surgery in patients deliberately given beta‐blockers prophylactically.154 The beneficial effects of perioperative β‐blocker administration are discussed in detail in another article in this issue.
Myocardial ischaemic injury is a potential perioperative threat. Ischaemia induces a palette of myocardial states with distinct pathophysiological backgrounds ranging from the paradoxically beneficial effects of preconditioning on one side to the complete loss of cellular integrity and to cell death on the other side. Preconditioning mimicking agents such as inhalation anaesthetics and opioids induce a pronounced protective cardiac phenotype and thus may decrease, along with β‐blockers, α2‐adrenoceptor agonists, and anti‐inflammatory/preconditioning‐mimicking statins, the deleterious effects of myocardial ischaemia in perioperative medicine.
This work was supported by the Swiss National Science Foundation Grant 3200‐063417.00 and Grant 3200B0‐103980/1, the Swiss Heart Foundation, a Grant from the Swiss University Conference, and a Grant from the Hartmann–Müller Foundation, Zürich, Switzerland.
1Institute of Anaesthesiology, University Hospital Zürich, Switzerland. 2Institute of Pharmacology and Toxicology, University of Zürich, Switzerland. 3Nuffield Department of Anaesthetics, The John Radcliffe Hospital, Oxford, UK