Volatile Anesthetics Protect the Ischemic Rabbit Myocardium from Infarction

After we received approval for the study by the Institutional Animal Care and Use Committee, New Zealand White rabbits of either sex, weighing 1.8–2.5 kg, were anesthetized with intravenous or volatile agents according to the specific anesthesia protocol. After tracheotomy and intubation, the lungs were mechanically ventilated with 100% oxygen. The ventilation rate was 30–35 breaths/min and tidal volume was approximately 15 ml. The respiratory rate was adjusted to keep the arterial pH in the physiologic range of 7.35–7.45. A temperature probe was inserted into the rectum and normal body temperature was maintained with a heating pad. Catheters were placed in the left carotid artery and jugular vein to monitor blood pressure and to inject drugs, respectively. A left thoracotomy was performed in the fourth intercostal space and the pericardium was opened to expose the heart. A prominent branch of the left coronary artery was selected and surrounded with a 2–0 silk suture to form a snare. Pulling the snare reliably occluded the coronary artery with subsequent cyanosis and bulging of the myocardium distal to the occlusion. Release of the snare resulted in reperfusion and epicardial hyperemia. Infarction was produced by occluding the coronary branch for 30 min and reperfusing it for 3 h. Measurements of coronary collateral blood flow were not necessary because rabbits lack significant preformed collateral vessels. [13] 

For isolated heart experiments, rabbits were anesthetized with 30 mg/kg sodium pentobarbital, mechanically ventilated as noted previously, and subjected to a left thoracotomy. After passing a ligature under a coronary branch, we excised the heart and mounted it within 1 min by the aortic root on a Langendorff apparatus. The heart was perfused from a reservoir maintained 100 cm above it with Krebs-Henseleit buffer containing 118.5 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO⁴, 1.2 mM KH²PO⁴, 2.5 mM CaCl², 24.8 mM NaHCO³, and 10 mM glucose. The buffer was equilibrated with a 95% oxygen and 5% carbon dioxide mixture and maintained at 37 degrees Celsius. Buffer pH was in the 7.35–7.45 range. The partial pressure of oxygen of the solution ranged from 400–600 mmHg. A saline-filled balloon was passed through the left atrium into the left ventricle. The balloon was filled with saline such that peak ventricular pressure of the beating, nonworking heart equaled 100 mmHg. If that could not be achieved with a preload under 15 mmHg, the heart was discarded. After initial balloon adjustment, no changes were made for the rest of the experiment. Coronary flow was measured by timed collection of the effluent from the right heart. Infarction was produced by occluding the coronary arterial branch for 30 min with reperfusion for 2 h.

In the intravenous group, general anesthesia was induced with either sodium pentobarbital (30 mg/kg), propofol (40–45 mg/kg), or ketamine with xylazine (165 mg/kg and 26 mg/kg, respectively) and maintained with additional intermittent dosing of pentobarbital or ketamine with xylazine or by continuous infusion of propofol (180 micro gram [center dot] kg sup -1 [center dot] min sup -1). The inhalational groups underwent induction by breathing halothane (1 minimum alveolar concentration [MAC]= 0.77%), enflurane (1 MAC = 1.70%), or isoflurane (1 MAC = 1.15%) at concentrations equal to 3 +/- 0.5 MAC in humans. After tracheal intubation, anesthesia was maintained with 1.5 +/- 0.5 MAC for each inhalational agent. Each anesthetic agent was administered through an agent-specific, temperature-controlled, calibrated vaporizer. Amounts of both intravenous and volatile anesthetics could be titrated for each animal, if necessary, to maintain adequate general anesthesia during the procedure. No muscle relaxants were given to any animals so that the depth of general anesthesia could be observed. Adequacy of anesthesia was documented by absence of palpebral reflex and lack of limb movement with stimulation. During the experiments, arterial blood pressure, heart rate, electrocardiographic (ECG) activity, arterial oxygen saturation and pH, and temperature were monitored. Because the animal was ventilated with 100% oxygen, the blood was always fully saturated and oxygen partial pressure ranged from 180–300 mmHg. Infarct size after a 30-min coronary occlusion and 3-h reperfusion period was measured in rabbits anesthetized with each of the six agents, as described below.

After a 20-min equilibration period, a 2 MAC concentration of one of the three volatile anesthetics was included in the perfusate for 5 min, followed by a 10-min washout period. The heart then underwent a 30-min period of coronary occlusion to produce an infarct. Infarct size was measured after 2 h of reperfusion, as described below.

In the adenosine receptor blockade group, 50 micro Meter 8-(p-sulfophenyl) theophylline (SPT) was included in the perfusate starting 5 min after the exposure period to halothane, and, therefore, 5 min before coronary occlusion and continuing for the first 15 min of ischemia. Hearts were subjected to a total of 30 min of ischemia followed by 2 h of reperfusion with SPT-free buffer.

In the PKC blockade experiments, an infusion of 5 micro Meter chelerythrine, a highly selective PKC blocker, was started 5 min after the exposure period to halothane and, therefore, 5 min before coronary occlusion, and continued throughout the 30 min of ischemia. Another group of rabbit hearts was treated with the same dose and schedule of chelerythrine but without previous exposure to halothane. Infarction in these groups was compared with that of a control group of untreated hearts that experienced only the 20-min period of equilibration before the coronary occlusion. A final group of hearts was ischemically preconditioned with a 5-min period of global ischemia achieved by arresting retrograde aortic perfusion followed by a 10-min period of reperfusion before the 30 min of regional ischemia.

Isolated rabbit hearts were used to measure the effects of halothane on intramyocardial adenosine production and release. Isolated hearts (n = 4) were perfused at a constant aortic pressure (100 cm water) with Krebs-Henseleit bicarbonate buffer. Adenosine levels in the coronary effluent collected from the right heart were determined by high-pressure liquid chromatography using a C-18 column and ultraviolet detection. Samples of coronary venous effluent were taken before and at the end of 5 min of exposure to halothane at 2 MAC.

At the end of each in situ study, hearts were quickly removed and mounted on a modified Langendorff apparatus and flushed with normal saline for 60 s. Isolated hearts were already on the apparatus at the end of the 2-h reperfusion period. In both experimental models the coronary branch was reoccluded and 1–10 micro meter fluorescent particles (Duke Scientific Corp., Palo Alto, CA) were infused into the perfusate to mark the risk zone as the nonfluorescent region. The hearts were removed from the Langendorff apparatus, weighed, and frozen. They were subsequently cut into 2-mm-thick transverse slices, which were incubated in 1% triphenyltetrazolium chloride buffer, pH 7.4, for 20 min. Triphenyltetrazolium chloride (T TC) reacts with reduced nicotinamide adenine dinucleotide in the presence of dehydrogenase enzymes and causes living tissue to stain a deep red. The region of infarcted tissue (T TC-negative) and the risk zone (area lacking fluorescence under ultraviolet light) were traced on plastic overlays. The areas of infarct and risk zone were determined by planimetry. Volumes of infarct and myocardium at risk were calculated by multiplying the planimetered areas by the slice thickness and summing values for all slices. Infarct size was expressed as a percentage of the risk zone.

Data were tabulated and computations made using Microsoft Systat software (Redmond, WA). All results are reported as means +/- SEM. Infarct volume (expressed as cubic centimeters) was plotted against risk zone volume (also cubic centimeters) for each group, and regression lines were analyzed for differences among groups. Because the relation between infarct and risk zone volumes has a non-zero intercept, a risk area of at least 0.5 cm³was required to minimize the likelihood that risk zone size alone could influence the results. [16]Intergroup comparisons were made with a one-way analysis of variance, and group differences were detected with Tukey's post-hoc test. Changes in hemodynamics were analyzed by analysis of variance with repeated measures. A probability value less than 0.05 for a comparison was considered to indicate a statistically significant difference. When multiple comparisons were made to determine the effect of different anesthetics on a hemodynamic parameter, a more conservative probability value calculated using the Dunn-Sidak correction was used to determine statistical significance. [17] 

(Figure 1and Table 1) show the infarction data for the three volatile anesthesia and three intravenous anesthesia protocols in the in situ model. Risk zones were not different in any of the groups. General anesthesia with any of the three intravenous anesthetics resulted in relatively large infarcts averaging about 40% of the risk region. There were no significant differences in infarct size within the three intravenous anesthetic groups. All three of the volatile anesthetics resulted in much smaller infarcts, averaging about 20% of the risk zone. There were also no significant differences among the infarct sizes within the inhalational groups. More importantly, infarction was significantly less in each of the inhalational groups than the propofol group, and halothane and isoflurane each produced smaller infarcts than did both pentobarbital and ketamine/xylazine.

Infarct volume is plotted against risk zone volume for the two classes of anesthetic agents in Figure 2. There is very little overlap between protected inhalational and unprotected intravenous groups, and the regression lines are statistically different. Protection is indicated by the lower slope for animals treated with the volatile anesthetics. Analysis of similar plots for each anesthetic (not shown) revealed no differences between any of the three volatile anesthetics or between any of the three intravenous anesthetic regimens.

(Table 2) presents hemodynamic data for all six anesthetic groups at baseline, at the beginning of ischemia, at the beginning of reperfusion, and at the end of the study. Baseline blood pressures were lower, in general, in the volatile anesthetic groups as compared with the pentobarbital and ketamine/xylazine groups. Mean arterial pressure and rate pressure product (the product of heart rate and mean arterial pressure)(RPP) declined during the experiment in the halothane, enflurane, and isoflurane groups. In addition, in the ketamine/xylazine group, arterial pressure decreased significantly at the beginning of reperfusion when compared with baseline and remained depressed throughout the rest of the study. Rate pressure product also declined in rabbits anesthetized with pentobarbital.

Blood pressure and RPP were generally higher in the propofol and pentobarbital groups than in the three groups of animals anesthetized with inhalational agents. It should be noted, however, that RPP in the ketamine/xylazine group was comparable to that measured in the animals treated with volatile anesthetics.

To determine whether blood pressure was influencing infarction, normalized infarct size was plotted against mean arterial pressure at the four different time points in the halothane group, which had the largest number of animals and widest range of mean arterial pressures. The data were analyzed by linear regression. The r values for the regression analyses were 0.004 at baseline, 0.295 during early ischemia, 0.334 in early reperfusion, and 0.085 during late reperfusion. At no time was a significant correlation between the two parameters found. The plot of infarct size versus mean blood pressure during ischemia is shown in Figure 3. Thus arterial pressure did not appear to significantly influence infarct size in these animals.

In this series of experiments, hearts were transiently exposed for only 5 min to volatile anesthetics included in the perfusate. As seen in Figure 4and Table 3, a 5-min exposure to either halothane, enflurane, or isoflurane provided equal and significant degrees of protection. There were no significant differences in the size of risk zones in any of the groups. Whereas 37.7 +/- 4.8% of the risk zone infarcted in untreated hearts, infarction was significantly less (P <0.0005) after exposure to halothane (6.6 +/- 2.3%), enflurane (9.3 +/- 3.8%), or isoflurane (5.2 +/- 1.6%). These small infarcts were comparable to those observed in hearts preconditioned with a 5-min ischemic period (9.8 +/- 3.1%). The degree of protection did not differ among the three volatile anesthetic and the ischemically preconditioned groups. These data are consistent with the results found in the in situ experiments.

Chelerythrine, a highly selective PKC inhibitor, blocked the protective effects of halothane in the isolated heart, as shown in Table 4and Figure 5. This halothane group is a different group of hearts from that shown in Figure 4and was studied concurrently with the other groups shown in Figure 5. Halothane again caused a marked reduction in infarct size, to only 9.7 +/- 1.8% of the risk zone. The addition of chelerythrine to halothane-exposed hearts resulted in 34.8 +/- 3.5% infarction (P < 0.005 vs. halothane), which was not different from that in untreated controls (31.9 +/- 2.6% infarction) or hearts exposed to chelerythrine alone (23.8 +/- 2.7% infarction). Halothane was also no longer protective in the presence of the adenosine A¹receptor blocker SPT (26.2 +/- 5.2% infarction, p = NS vs. control, P < 0.05 vs. halothane). Previous studies in isolated hearts have shown that SPT infused before ischemia in non-preconditioned hearts has little effect on infarct size. [18]In this series, average risk zone volume was inexplicably smaller in hearts treated with the combination of halothane and SPT. To ensure that this smaller risk zone had no independent effect on the infarct:risk zone ratio, [16]a plot of infarct versus risk zone volumes similar to that shown in Figure 2was constructed (not shown). The data points for the halothane-SPT hearts fell along the same regression line as those of the other unprotected hearts, implying that different risk zone sizes in these hearts had no significant effect on the extent of infarction.

Heart rate, left ventricular developed pressure, and coronary flow data are displayed in Table 5and Table 6. There were no differences among groups at baseline. Although developed pressure and coronary flow were lower during ischemia and at the end of the reperfusion period, the three volatile anesthetic agents produced nearly identical changes. Similarly, in the adenosine receptor and PKC blockade experiments, parallel changes in hemodynamics were observed during the experiments.

In this protocol we wanted to determine whether halothane might be acting to increase adenosine release by the ischemic heart. The mean adenosine level in the effluent from four Krebs buffer-perfused isolated hearts was low and averaged 0.093 +/- 0.048 micro Meter. At the end of a 5-min period of perfusion with buffer containing 2 MAC halothane, adenosine levels actually decreased to 0.029 +/- 0.010 micro Meter (P = 0.27 vs. basal level). If halothane had protected these hearts by increasing the level of interstitial adenosine, we would have expected an increased adenosine concentration by at least ten times. Thus there was no indication that halothane protected the heart by increasing adenosine levels.

The present study shows that the halogenated, inhalational anesthetic agents halothane, enflurane, and isoflurane profoundly protect the heart from infarction as compared with intravenous general anesthesia. In the in situ hearts, concurrent administration of a volatile anesthetic before and during ischemia reduced the extent of infarction by about 50%. This protection appeared to be independent of mean arterial pressure over a wide range. In addition, there were no differences in the much larger degree of infarction among the three intravenous drugs despite the diversity of the agents, which included a barbiturate, a hypnotic, and a dissociative drug with a veterinary analgesic. We must ask whether the inhalational anesthetics protect the heart or whether general intravenous anesthetics exacerbate infarction. In a previous study, we occluded a coronary artery branch for 30 min and induced myocardial infarction in conscious rabbits in the absence of anesthesia. [19]The experiment was performed by surgically implanting an inflatable occluder on the coronary artery 1 week before the experiment. In that study, infarct size averaged 38% of the risk region and was not different from the 39% infarction seen in the open-chest, pentobarbital-anesthetized group in the present study. Based on these data, it would appear that pentobarbital anesthesia does not increase the extent of infarction. Therefore, by extrapolation, we conclude that the inhalational anesthetics must act to protect the hearts.

Unfortunately, the volatile anesthetics caused blood pressure to be lower than that seen with the intravenous anesthetics. We addressed the possibility that altered hemodynamics might have been the critical determinant of infarct size in these experiments. First, we plotted infarct size against blood pressure at four different time points in in situ preparations anesthetized with halothane (Figure 3). There was no correlation between the extent of infarction and blood pressure at any time point. Next we calculated rate pressure products for each experiment. As indicated in Table 2, RPP values in the propofol and pentobarbital groups were, as expected, higher than in the three groups anesthetized with halothane isoflurane, and enflurane. However, RPP in the third intravenous group anesthetized with ketamine/xylazine was significantly lower than that in the pentobarbital rabbits, tended to be lower in the propofol group and was comparable to that observed in the three groups with volatile anesthetics. Yet infarct size in the rabbits anesthetized with ketamine/xylazine was similar to that measured in the other two groups treated with injectable agents and much greater than that in the groups anesthetized with halothane, isoflurane, and enflurane. Thus a lower RPP cannot be the critical factor accounting for a lower infarct size in these latter three groups. Finally, because of the critical importance of the possible effect of changing hemodynamics on infarct size, we elected also to study the three volatile anesthetic agents in an isolated heart preparation in which peripheral hemodynamic effects are eliminated. In this series, a 5-min exposure to halothane, enflurane, or isoflurane at 2 MAC before ischemia reduced the amount of infarction in the ischemic heart by an amount similar to that seen in the in situ animals. Therefore, it would appear that altered hemodynamics in rabbits anesthetized with inhalational agents are not critical factors in the salvage of ischemic myocardium in these experiments. It should be noted that, except when agents were added to the perfusate, isolated hearts are anesthetic free, indicating a direct protective effect of the previous exposure to the volatile anesthetics.

The protection seen in the isolated hearts with the volatile anesthetics is similar in degree and time course to the phenomenon known as ischemic preconditioning (Figure 4). [18]In ischemic preconditioning, exposing the heart to a transient period of ischemia followed by reperfusion puts the heart into a protected state for about 1 h. [20,21]A similar phenomenon was seen here, with protection persisting even when the anesthetic was allowed to wash out. These results suggest that volatile anesthetics “precondition” ischemic rabbit myocardium.

Ischemic preconditioning was originally described by Murry et al. in 1986. [22]As a result of this remarkable phenomenon, exposure of the heart to a brief period of ischemia causes myocardium to adapt rapidly to become highly resistant to infarction from a subsequent ischemic insult. This protection has been shown to involve stimulation of adenosine receptors, [18]and recent evidence indicates that protection is a direct consequence of activation of PKC. [23]The degree of protection afforded by the inhalational anesthetics seen here was comparable to the protection afforded by ischemic preconditioning in the rabbit model in the current study. To test whether volatile anesthetics might be protecting the heart by a similar mechanism, we observed the effect of an adenosine receptor blocker, SPT, on that protection. As is seen in ischemic preconditioning, [18]SPT abolished the protection, suggesting that both preconditioning and volatile anesthesia may be working through a similar mechanism. It should be noted that blocking adenosine receptors with 100 micro Meter SPT had no effect on infarct size in non-preconditioned isolated hearts, [18]and it is assumed that the same holds for the 50-micro Meter concentration used in the present protocol.

We have proposed that in the preconditioned state a coupling is established between adenosine receptors and PKC that does not normally exist. [14,15]This change makes the heart's own adenosine very protective. Indeed, the necessity of PKC activation for cardioprotection from halothane was confirmed by our experiments. Chelerythrine, a specific PKC inhibitor, completely abolished the protective effect of halothane in the isolated heart.

The physiologic trigger of the preconditioned state for the heart seems to be transient stimulation of PKC through adenosine receptors. [18,24]That is not the only possible trigger, however. Indeed, transient stimulation of any of the known PKC-coupled receptors in the heart has been found to put the myocardium into a preconditioned state. These include the alpha¹-adrenergic, [25,26]angiotensin AT¹, [27]bradykinin B², [28]and endothelin ET-1 [29]receptors. Furthermore, transient exposure to direct activators of PKC, including diacylglycerol analogs [23,30]and phorbol esters, [29,31]can also put the heart into a preconditioned state. Protection in hearts pharmacologically preconditioned with agents other than adenosine can still be blocked by introducing an adenosine receptor blocker during ischemia. [25,31] 

How then do the volatile inhalational anesthetics cause the heart to become preconditioned? The isolated heart experiments provided no indication that halothane acted to promote adenosine release in the non-ischemic heart because adenosine levels were not increased after exposure to halothane. Therefore, halothane probably acted more directly to stimulate PKC. Indeed, one report has shown that halothane is a direct activator of pertussis toxin-sensitive G proteins, [32]also noted to be involved in the protection of ischemic preconditioning [33]and similar to those binding to adenosine A¹receptors. [34]However, Schmidt et al. [35]found an inhibitory effect of halothane on Gⁱproteins in human myocardium. If halothane had either inhibited or stimulated a Gⁱprotein subunit in this model, we might have expected a consistent corresponding change in heart rate after halothane exposure, but we did not see this in our data. If coupling of the adenosine A¹receptor with a Gⁱprotein subunit were enhanced by volatile anesthetics, then the effect might well be indistinguishable in the cardiomyocyte from increased adenosine receptor stimulation.

We are not the first to try to determine whether anesthetics modulate the vulnerability of the heart to ischemia. As already noted, several studies using end points other than infarction support a protective role for the inhalational anesthetics. [1–7,9–12]Two studies actually measured infarct size. [36,37]One used a method of tetrazolium staining similar to ours to measure cell death after 30 min of ischemia in rabbit hearts in situ. [36]These investigators reported that halothane anesthesia actually resulted in a higher percentage of necrosis (64%) compared with alpha-chloralose (54%) and pentobarbitone anesthesia (44%). However, in that study, risk zones in each animal were not measured. Risk area was arbitrarily obtained from a set of five entirely different rabbits that underwent 90 min of myocardial ischemia. As we have noted, occlusion of a coronary branch results in wide variance in the amount of tissue rendered ischemic, [16]and the extent of infarction cannot be determined accurately unless the infarct size is normalized to the size of the ischemic zone in each animal. Therefore, the validity of that report is questionable.

In the second study, the authors used the same method that we did to measure risk zones and cell death in in situ rabbit hearts. [37]Isoflurane anesthesia resulted in 30% less infarction of the risk zone than did ketamine-xylazine and 25% less than did pentobarbitone anesthesia, but the data did not reach statistical significance. A similar observation was made in the present study, in which the decrease in normalized infarct size in the enflurane group also approached but did not reach statistical significance (P < 0.06) when compared with that in animals anesthetized with pentobarbital. Another interesting finding of the study by Haessler et al. [37]was that ischemic preconditioning protected the animals anesthetized with isoflurane less than the animals in which pentobarbital was used, perhaps an expected finding if the former were already partially preconditioned by the anesthetic. We cannot explain why the anesthetic effect was so much less dramatic in their study compared with our present results.

Thus all anesthetic agents do not appear to have equivalent effects on the ischemic heart, and studies such as this one might ultimately influence agent selection. The potential value of preconditioning the heart before procedures in which the heart will be subjected to a period of ischemia is well recognized. Until now, however, none of the currently approved agents could pharmacologically precondition the heart unless they were given as an intracoronary infusion. But if inhalation of a halogenated anesthetic for a brief period can confer the protection of preconditioning, then a relatively simple method exists to protect the ischemic heart in various clinical situations.