Hyperoxemic reperfusion myocardial infarct size
does not increase
CARLA B. SHNIER, BRIAN A. CASON, ANNE F. HORTON, AND ROBERT F. HICKEY Department of Anesthesia, University of California, San Francisco 94143; and Anesthesiology Service, Veterans Administration Medical Center, San Francisco, California 94121
SHNIER, CARLA B., BRIAN A. CASON, ANNE F. HORTON, AND ROBERT F. HICKEY. Hyperoxemic reperfusion does not increase myocardial infarct size. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H1307-H1312, 1991.-We tested the hypothesis that arterial hyperoxia during myocardial reperfusion increases reperfusion injury and infarct size. The anterolateral marginal coronary artery of 35 anesthetized rabbits was occluded for 45 min, then reperfused for 3 h with either normoxic [arterial PO, (PaO,) = 96.7 t 22.9 mmHg)] or hyperoxic (Pao, = 554.8 t 61.7 mmHg) blood. In the hyperoxic group only, Paoz was adjusted 10 s before the onset of reperfusion by raising inspired oxygen concentration to 100%. The area of infarction (AI) was defined by triphenyltetrazolium staining, and the area at risk (AR) by fluorescent microspheres. These areas were measured by planimetry. Heart rates and blood pressures did not differ between the two groups during occlusion or reperfusion. Infarct size (AJAR) was 49.1 * 16.5% in the normoxic group (n = 17) and 40.8 ? 16.1% in the hyperoxic group (n = 18). From these data, 90% confidence limits establish that the maximal true increase in AI/AR caused by hyperoxia would be 0%-l %. Hyperoxic reperfusion of ischemic myocardium compared with normoxic reperfusion does not significantly increase myocardial infarct size.
stantially increase blood oxygen content. High arterial PO, (Pao,) may nevertheless increase PO, at important sites of free radical injury, such as endothelial and myocardial cell membranes and leukocytes. Because hyperoxia can increase free radical production, and because previously anoxic tissue is particularly vulnerable to oxygen-mediated injury (12), we hypothesized that hyperoxemic reperfusion might increase reperfusion injury to ischemic myocardium. This issue is important because Pa 0, can be easily controlled in many clinical circumstances and could therefore be adjusted to maximize myocardial salvage during reperfusion of the heart after thrombolysis, angioplasty, or open heart surgery. Because the hypothesis of an adverse effect of Paoz on myocardial reperfusion injury has never been tested in the intact blood-perfused heart, we used a rabbit model of myocardial infarct size to answer the question, does hyperoxemia during myocardial reperfusion increase infarct size?
myocardial
Surgical preparation. With the approval of our hospital’s animal welfare committee, 35 adult New Zealand White rabbits (2.6-5.2 kg) were anesthetized with halothane in oxygen. The trachea was intubated via a tracheostomy, and ventilation was controlled to achieve a normal arterial pH and Pco~. Through a median sternotomy and pericardial incision, the heart was exposed. The anterolateral coronary artery was identified and encircled with a 2-O polyester suture at the midpoint of its epicardial course. The two ends of the suture were threaded loosely through a 6-cm piece of flexible plastic tubing forming a snare that would be tightened to occlude the coronary artery. Teflon catheters (Jelco) were inserted into the left atrium (20 gauge) to measure cardiac filling pressure, into the femoral artery (25 gauge) to measure systemic arterial pressure, and into the femoral vein (20 gauge) to provide intravenous access for fluid and heparin administration. A modified V5 lead electrocardiogram recorded heart rate. Ischemiaprotocol. To determine the effect of hyperoxic reperfusion on infarct size, we used a rabbit model of acute coronary artery occlusion-reperfusion. The left lateral coronary artery was occluded for 45 min, and then, by changing the fraction of inspired oxygen (FIN,), reperfused with blood having either a high Pao, or a
reperfusion
injury;
oxygen free radicals;
rabbit
REPERFUSION of ischemic myocardium limits myocardial infarct size (2, 5), decreasing morbidity and mortality from myocardial infarction (5, 32). However, recent work suggests that reperfusion itself can kill ischemic cells (termed reperfusion injury) (2, 4). Although the mechanisms of reperfusion injury are likely multiple, the most widely accepted mechanism is the generation of toxic oxygen free radicals during reperfusion. This mechanism was suggested by studies showing that reoxygenation of the ischemic or anoxic heart causes tissue damage (9,M) and is supported by studies demonstrating that reperfusion generates oxygen free radicals (3, 33) and other studies in which the administration of free radical scavengers during myocardial reperfusion reduces infarct size or diminishes postischemic functional impairment (1, 16, 26). Additionally, in some free radical-generating systems, higher Po2 leads to greater oxyradical production. In mitochondria, microsomes, and nuclei isolated from various animal sources, rate of free radical production increases during hyperoxia (11, 15, 30, 31). Because of the limited solubility of oxygen in water, increasing the PO, in arterial blood above the normal range does not subEARLY
0363-6135/91
$1.50 Copyright
METHODS
0 1991 the American
Physiological
Society
H1307
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
H1308
HYPEROXIC
REPERFUSION
normal Pao,. After 3 h of reperfusion we compared myocardial infarct size postmortem in the two Pao, groups. Before coronary ligation, we anticoagulated the blood with intravenous heparin sodium (400 U/kg) and lowered the FIN, to as close to 0.21 as possible without producing hypoxemia (Pao, = 95.5 t 13.3 mmHg). For each rabbit, the inspiratory halothane concentration between 0.6 and 1.0 vol/lOO ml that maintained anesthesia and normal hemodynamic values was established and held constant for the duration of the protocol. After a 30-min period of stabilization, we occluded the coronary artery by pulling the coronary ligature taut through the plastic tube and clamping the ligature in position. Hemodynamics, arterial blood gases, hemoglobin concentration, and hemoglobin oxygen saturation were measured before coronary ligation, after 30 min of occlusion, and after 30 min of reperfusion. Left atria1 and systemic arterial pressures were measured with Transpac II strain gauges (Abbott, Chicago, IL) and recorded on a Gilson polygraph (Middleton, WI). Hemoglobin concentration and oxygen saturation were measured with a Radiometer OSM 3 analyzer. Arterial blood gases were measured with a Radiometer ABL 2 analyzer (Copenhagen, Denmark). Reperfusion protocol. The 45 min of myocardial ischemia was followed by 3 h of myocardial reperfusion. Just before coronary ligature release, we assigned the rabbit randomly to one of two treatment groups. The control group underwent myocardial reperfusion with the normal Pao, set during the occlusion; the hyperoxemic group underwent myocardial reperfusion with a high Pao, attained by increasing the F1oz to 1.0 10 s before the moment of coronary ligature release. A rapid change in inspired oxygen concentration was initiated by flushing the ventilation circuit with pure oxygen at 50 l/min. This raised the inspiratory oxygen concentration in the hyperoxemic group to over 90% within 5 s. Arterial blood gases and hemoglobin concentration were measured at 1 and 5 min after ligature release and every 30 min thereafter. At 1 min of reperfusion, Pao, was above 450 mmHg in all hyperoxemic rabbits. Ventricular fibrillation during ischemia or reperfusion was treated by internal defibrillation (1 J/kg). After reperfusion, the rabbit was killed by anesthetic overdose, and myocardial infarct size was measured. Exclusion criteria. Rabbits were excluded from the study if they suffered from 1) ventricular fibrillation unresponsive to two consecutive electrical shocks of 4 J each, 2) mean blood pressure of ~40 mmHg during ischemia, 3) persistent hypoxemia of 0.5).
of area at hyperoxic enclose in infarct one-tailed
AND
INFARCT
H1309
SIZE
limits and were not significantly different between groups at baseline, after 30 min of coronary occlusion, and after 30 min of myocardial reperfusion. As documented in Table 3, baseline Pao, also did not differ significantly between groups until reperfusion, when the Pao, of the experimental rabbits was increased by experimental design. The groups did not differ in rectal temperature, body weight, halothane concentration, or hemoglobin concentration. DISCUSSION
In this study, we found that hyperoxemia during myocardial reperfusion did not increase myocardial infarct size in rabbits subjected to temporary coronary artery occlusion. We hypothesized that hyperoxemic reperfusion of ischemic myocardium would increase infarct size because free radicals have been shown, in some circumstances, to be formed by PoZ-dependent processes (11, 30, 31). One of the fundamental assumptions of this hypothesis is that an increase in Pao, would increase the Po2 at sites where free radicals are generated. Because free radicals have extremely short half-lives, those sites of generation that are important in reperfusion injury are difficult to locate precisely. Previous work demonstrating the “noreflow” phenomenon, increases in microvascular permeability, and leukocyte-mediated tissue damage have implicated the blood and the endothelium as important sites in reperfusion injury (23, 29). It is certain that raising the FI o, to 1.0 in our experimental rabbits did increase the PO, in the arterial blood and likely, too, the Paz in sites close to the blood such as circulating leukocytes and endothelial cells (17). Another assumption made in this experiment was that 3 h of reperfusion are sufficient to allow the development of detectable differences between the experimental groups. This is a reasonable assumption, since free radicals are formed in large part early in reperfusion (33) and brain (22), and heart tissues (14) show increased damage produced by hyperoxia within this time frame. However, injury due to ischemia and reperfusion initiates an inflammatory process that resolves over days to weeks, and reperfused tissue that initially appears viable may nevertheless be “condemned.” A significant problem TABLE
3. Arterial
blood gases Baseline
30 Min Occlusion
30 Min Reperfusion
7.4520.04 7.44t0.06
7.44t0.08 7.46t0.05
7.43t0.06 7.43kO.05
33.9t3.8 34.4t5.2
33.9t4.8 31.8t4.5
35.Ok5.3 32.5t3.6
97.1t11.7 94.4t14.9
97.9t20.3 97.9t11.2
96.7t22.9 554.8261.7
14.4tl.4 5.0tl.9
14.2t2.1 14.9t1.64
14.1-+1.5 15.9t1.7*
-7
PH Reperfusion Group
~~ Heart rate, beats/min Left atria1 pressure, mmHg Mean arterial pressure, mmHg Rate pressure product X lo-‘, Values
are means
Normal High Normal High Normal High Normal High
Pao, Pao, Pao, Pao, Pao, Pao, Pao, Pao, G
~fr SD for 17 normal
Baseline
30 Min Occlusion
30 Min Reperfusion
278t39 290t30 4.9t2.5 3.9t2.0 61t9 66t9 171t35 192rt39
276t34 289t30 5.3t2.6 4.3t1.9 6lt9 64tll 168t33 185t43
262t25 278&35 4.8t2.4 4.3t2.0 60t7 62t9 158t27 173*39
Pao, and 18 high Pao, rabbits.
Normal Pao, group High Pao, group mmHg ho,, Normal Pao, group High Pao, group Pao2, mmHg Normal Pao, group High Pao, group Arterial 0, content, ml OJdl Normal Pao, group High Pao, group
Values are means t SD for I7 normal Pao, and 18 high Pao, rabbits. * P < 0.05 vs. normal Pao, group at 30 min reperfusion.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
HI310
HYPEROXIC
REPERFUSION
in acute reperfusion studies is that they may therefore overestimate the tissue-sparing effect of treatment, in that initial differences in tissue injury prove inconsequential when reassessed at a later time. In this experiment, however, we found no more tissue-sparing effect from normoxic reperfusion than from hyperoxic reperfusion. In fact, infarct size was slightly higher in the normoxic group (49.1%) than in the hyperoxic group (40.8%), although this difference did not reach statistical significance (P = 0.14 by two-tailed t test). A much larger study of -135 rabbits would be required to establish such a small difference with 80% certainty. The short time course of the study brings up another methodological concern. Because killing the animals at 3 h after coronary release precluded our measuring infarct size by histology, we used a histochemical staining technique. Triphenyltetrazolium chloride marks viable muscle by forming an insoluble red precipitate in the presence of dehydrogenase enzymes and coenzymes. Cells that are irreversibly damaged quickly lose enzyme activity and the ability to reduce tetrazolium salts, so they remain unstained. Because detection of dead tissue depends on absence of enzyme activity, early infarcts may be underestimated due to temporary retention of dehydrogenase enzyme activity in condemned cells (27). However, the triphenyltetrazolium chloride method of infarct size measurement is appropriate in our study for the following reasons. First, this staining technique was shown by ultrastructural analysis to be an accurate method for quantifying infarct size as early as 3 h after coronary occlusion (10). Second, reperfusion as performed in our study accelerates the washout of enzyme activity from cells that are irreversibly damaged and makes spurious staining less likely (28). Third, the hearts in our study were frozen before staining, which lessens artifactual tissue salvage in infarct size studies (7). Finally, if triphenyltetrazolium chloride underestimates the size of the infarct, there is no a priori reason to believe that it would act any differently in the two groups of rabbits. A final concern is that unequal anatomic areas at risk of infarction would bias our results. Though rabbits were randomly assigned to treatment groups after coronary ligation, the area at risk was, by chance, 24% smaller in the hyperoxic group than in the normoxic group. This chance inequality between the groups does not invalidate comparisons between the groups for two reasons. First, although a larger area at risk would favor a larger infarct, infarct size was expressed and compared as a percentage of the corresponding area at risk. Second, because collateral blood flow in the rabbit heart is essentially nil (21), cells within a smaller area at risk were as unlikely to be sustained by collateral flow as those within a larger area at risk. Although there is considerable evidence that reoxygenation of the ischemic or hypoxic tissue can cause irreversible cell injury (13), this is the first study of the effect of hyperoxic blood reperfusion on myocardial infarct size in vivo. The theoretical impetus for this study came from previous work in isolated heart (14), skeletal muscle (19), brain (22), and gut (25), demonstrating the influence of PO2 in reoxygenation-reperfusion injury.
AND
INFARCT
SIZE
The discrepancy between our results, which showed no increase in injury from hyperoxic reperfusion, and these cited above, may be explained by differences in both study design and the criteria used to measure injury. For example, Hearse et al. (14) showed that after a period of oxygen deprivation, reoxygenation of isolated crystalloid-perfused rat hearts caused structural damage and a dose-related massive increase in cardiac enzyme leakage. The relationship between reperfusion Paz and structural damage was not quantified. This study established reoxygenation as one of the potential causative events in reperfusion injury by showing that the extent of cardiac enzyme leakage varies directly with the Pop of the hemoglobin-free perfusate. By contrast, we found no increase in injury resulting from hyperoxic reperfusion. Three differences in study design may account for the discrepancy between Hearse’s findings and ours. First, we used a whole animal blood-perfused heart, whereas Hearse et al. (14) used an isolated crystalloid-perfused heart to study reoxygenation damage. Although reperfusion PO+ varied over a similar range, because the oxygen-carrying capacity of blood is much greater than that of crystalloid, the reperfusates in the two studies provided dissimilar amounts of oxygen. Hearts in our experiments were reperfused at a Pao, of 96.7 mmHg vs. a Paoz of 554.8 mmHg (14.6 vs. 16.0 ml oxygen/100 ml blood). We thus compared the damage of restoring normal to slightly supernormal amounts of oxygen to the ischemic myocardium. By contrast, Hearse’s reperfusate contained no hemoglobin, and so increasing the Po2 from 71 to 677 mmHg restored only subnormal amounts of oxygen (0.2-2.5 ml/l00 ml crystalloid) to the anoxic myocardium. Second, it should be noted that, unlike in Hearse’s study, the measured tissue injury in our study was due to a combination of ischemia, reperfusion, and reoxygenation. The conditions of ischemia and reperfusion were the same in both of our experimental groups, allowing comparison of the effects of different Pozs at reoxygenation. It was possible, however, that injury due to ischemia and injury due to the stress of reperfusion were sufficiently large and variable that they obscured any smaller differences in injury due to the Po2 of the reperfusate. Ischemia and the physical stress of reperfusion were not considerations in Hearse’s continuously perfused hearts and so may account for the disparity in experimental findings between his study and ours. Third, differences in experimental outcome may also be attributable to the presence of formed elements in blood. Blood contains platelets and white cells that can contribute to the apparent reperfusion injury by plugging the microvasculature and preventing complete reflow. Although crystalloid-perfused hearts are not leukocytefree, they do contain fewer white cells than blood-perfused hearts (18). Because white cells are a significant source of free radicals, it may be that blood-reperfused hearts suffer greater obligatory reperfusion injury than crystalloid-perfused hearts, masking any increase in reperfusion injury due to hyperoxia. More recent work has used a variety of blood-perfused tissues to show that, in contrast to our findings, the partial pressure of oxygen in arterial blood is important
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
HYPEROXIC
REPERFUSION
in reperfusion injury (19, 22, 25). In a canine model of temporary skeletal muscle ischemia, Korthius et al. (19) found only small increases in microvascular permeability as a result of anoxic reperfusion but noted large increases in microvascular permeability during normoxic reperfusion. These findings are supported by similar experiments in gastric mucosa (25) where, after ischemia, reperfusion with hypoxic blood caused less microvascular injury than did normoxic blood. In both of these studies the investigators measured microvascular damage in nonworking tissue to quantify reperfusion injury, while we measured infarct size in the beating rabbit heart. Conceivably these two approaches differ in their sensitivity to changes in reperfusion Paz. Additionally, our results differ from these two previous investigations because our questions were different. We did not ask if oxygen contributed to reperfusion injury, but if oxygen in excess of that normally provided to maintain life and function would increase reperfusion injury. This question is more relevant to the treatment of human myocardial reperfusion where maintaining a perfusing Paz below that achievable breathing room air is not acceptable. Defining the role of Po2 in reperfusion injury is complicated by the fact that as Po2 in the reperfusate changes so does oxygen content. Otani et al. (24) examined the effect of oxygen content on reperfusion injury apart from the effect of Po2 by reperfusing isolated dog hearts with different degrees of hemodilution (Hb = 8, 4, or 0 g/dl). They found greater recovery of ventricular function and metabolism in the extreme hemodilution group (Hb = 4 g/dl, PO, = 100 m .mHg) th .an in the moderate hemodilution group (Hb = 8 g/dl, Po2 = 100 mmHg) or the hemoglobin-free plasma group (Po2 = 300 mmHg). This result was consistent with increased oxygen delivery leading to increased reperfusion injury by oxygen free radical formation. An alternate explanation for the benefit of extreme hemodilution is the associated decrease in leukocytes and platelets in the reperfusate or the lowered vascular resistance leading to increased blood flow. It is likely that the reason why hemoglobin-free hearts fared the worst was that reperfusion with oxygenated hemoglobin-free plasma, even at a Po2 of 300 mmHg, provided insufficient oxygen to meet the demands of this damaged myocardium. The advantage of our experimental design compared was twofold. with those of these previous investigators First ? we P rov ided a range of oxygen content th .at would sustain in vivo working heart muscle indefinitely; the hemoglobin-free perfusates used by Hearse et al. and Otani et al. would not. Similarly, the blood perfusates used in the skeletal muscle and gastric mucosa models were severely hypoxemic in the test groups exhibiting less reperfusion injury. Second, our study quantified reoxygenation injury by one of the most important predictors of cardiac morbidity and mortality, infarct size. of the In comparison with infarct size, the significance rate of myocardial enzyme leakage and of microvascular permeability is less certain. Like us, Mickel et al. (22) compared the effect of normoxic to hyperoxic reperfusion in an intact animal. This group demonstrated an increase in lipid peroxidation and an increase in mortality in gerbils breathing
AND
INFARCT
H1311
SIZE
100% oxygen compared with those breathing air after transient global brain ischemia. Cerebral infarct size was not measured. The reason our results do not agree with those of Mickel et al. may be because of differences in species (rabbit vs. gerbil) and in tissue studied (heart vs. brain). Additionally, though there is a strong suggestion that an increase in cerebral reperfusion injury due to arterial hyperoxia caused the increase in mortality, covariant causes such as changes in blood pressure and in ventilation cannot be ruled out. Based on our study, we speculate that the administration of supplemental oxygen to otherwise normoxic patients undergoing myocardial reperfusion after acute coronary artery occlusion will not increase infarct size. However, the extrapolation of these experimental results to humans should be made with caution. First, there may be significant species differences in free radical-producing systems, and these differences may determine sensitivity to therapeutic interventions. For example, xanthine oxidase, an important producer of superoxide radical in the dog and rat, is absent from the myocardium of rabbits and humans (6,8). Therefore, inhibitors of xanthine oxidase may be expected to reduce reperfusion injury in dogs and rats, but not in rabbits or humans. The relevant free radical-generating systems operating in human myocardial reperfusion injury are, at present, unknown. However, rabbit hearts do resemble human hearts at least insofar as neither generates free radicals by the well-studied xanthine oxidase system. Second, other experimental conditions such as the severity of ischemia and rapidity of reperfusion may also affect the extent of reperfusion injury (23). The current study modeled only one situation, absolute ischemia followed by sudden maximal reperfusion. This differs from the situation in older humans where, because of coronary collateral vessels, coronary occlusion may produce submaximal ischemia in much of the anatomic area at risk and reperfusion through diseased vessels may be more gradual. Lesser degrees of ischemia and gradual reintroduction of oxygen may lessen the amount of reoxygenation-reperfusion injury (14). Because we found no increase in infarct size with hyperoxic reperfusion under severe ischemia-reperfusion conditions, there is unlikely to be one in less extreme clinical situations. In summary, although oxygen free radicals are implicated in myocardial reperfusion injury, we found that hyperoxic reperfusion did not increase myocardial infarct size in the rabbit. This work was supported in part by a grant to B. A. Cason from the Veterans Administration Research Advisory Group. C. B. Shnier was supported by a Research Fellowship Award of the Heart and Stroke Foundation of Canada. Address for reprint requests: B. A. Cason, Anesthesiology Service (129), Veterans Administration Medical Center, 4150 Clement St., San Francisco, CA 94121. Received
17 April
1990; accepted
in final
form
5 December
1990.
REFERENCES 1. AMBROSIO, G., L. C. BECKER, G. M. HUTCHINS, H. F. WEISMAN, AND M. L. WEISFELDT. Reduction in experimental infarct size by recombinant human superoxide dismutase: insights into the pathophysiology of reperfusion injury. Circulation 74: 1424-1433, 1986.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
H1312
HYPEROXIC
REPERFUSION
2. BECKER, L. C., AND G. AMBROSIO. Myocardial consequences of reperfusion. Prog. Cardiouasc. Dis. 30: 23-44, 1987. 3. BOLLI, R., B. S. PATEL, M. 0. JEROUDI, E. K. LAI, AND P. B. MCCAY. Demonstration of free radical generation in “stunned” myocardium of intact dogs with the use of the spin trap a-phenyl N-tert-butyl nitrone. J. CLin. Inuest. 82: 476-485, 1988. 4. BRAUNWALD, E., AND R. A. KLONER. Myocardial reperfusion: a double-edged sword? J. Clin. Invest. 76: 1713-1719, 1985. 5. CAMPBELL, C. A., K. PRZYKLENK, AND R. A. KLONER. Infarct size reduction: a review of the clinical trials. J. CZin. Pharmacol. 26: 317-329,1986. 6. DOWNEY, J. M., T. MIURA, L. J. EDD, D. E. CHAMBERS, T. MELLER, D. J. HEARSE, AND D. M. YELLON. Xanthine oxidase is not a source of free radicals in the ischemic rabbit heart. J. Mol. Cell CardioZ. 19: 1053-1060, 1987. 7. DOWNEY, J. M., C. SHIRATO, T. MIURA, AND T. TOYOFUDU. Tetrazolium is unreliable as an index of drug-induced salvage (Abstract). J. Mol. Cell. Cardiol. 20, Suppl. V: S-70, 1988. 8. EDDY, L. J., J. R. STEWART, H. P. JONES, T. D. ENGERSON, J. M. MCCORD, AND J. M. DOWNEY. Free radical-producing enzyme, xanthine oxidase, is undetectable in human hearts. Am. J. Physiol. 253 (Heart Circ. Physiol. 22): H709-H711, 1987. 9. FEUVRAY, D., AND J. D. LEIRIS, Ultrastructural modifications induced by reoxygenation in the anoxic isolated rat heart perfused without exogenous substrate. J. Mol. Cell. Cardiol. 7: 307-314,1975. 10. FISHBEIN, M. C., S. MEERBAUM, J. RIT, U. LANDO, K. KANMATSUSE, J. C. MERCIER, E. CORDAY, AND W. GANZ. Early phase acute myocardial infarct size quantification: validation of the triphenyl tetrazolium chloride tissue enzyme staining technique. Am. Heart J. 101: 593-600,198l. 11. FREEMAN, B. A., AND J. D. CRAPO. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J. BioL. Chem. 256: 10986-10992,198l. 12. GUARNIERI, C., F. FLAMIGNI, AND C. M. CALDARERA. Role of oxygen in the cellular damage induced by re-oxygenation of hypoxic heart. J. Mol. Cell. Cardiol. 12: 797-808, 1980. 13. HEARSE, D. J. Reperfusion of the ischemic myocardium. J. Mol. Cell Cardiol. 9: 605-616, 1977. 14. HEARSE, D. J., S. M. HUMPHREY, AND G. R. BULLOCK. The oxygen paradox and the calcium paradox: two facets of the same problem? J. Mol. Cell. Cardiol. 10: 641-668, 1978. 15. JAMIESON, D., B. CHANCE, E. CADENAS, AND A. BOVERIS. The relation of free radical production to hyperoxia. Annu. Reu. Physiol. 48: 703-719, 1986. 16. JOLLY, S. R., W. J. KANE, M. B. BAILIE, G. D. ABRAMS, AND B. R. LUCCHESI. Canine myocardial reperfusion injury: its reduction by the combined administration of superoxide dismutase and catalase. Circ. Res. 54: 277-285, 1984. in oxidative stress: 17. JONES, D. P. The role of oxygen concentration hypoxic and hyperoxic models. In: Oxidatiue Stress, edited by H. Sies. London: Academic, 1985, p. 151-195. AND 18. KELLER, A. M., R. M. CLANCY, M. L. BARR, C. C. MARBOE, P. J. CANNON. Acute reoxygenation injury in the isolated rat heart: role of resident cardiac mast cells. Circ. Res. 63: 1044-1052, 1988. 19. KORTHIUS, R. J., J. K. SMITH, AND D. L. CARDEN. Hypoxic renerfusion attenuates postischemic microvascular injury. Am. J.
AND
INFARCT
SIZE
Physiol. 256 (Heart Circ. Physiol. 25): H315-H319, 1989. 20. LIE, J. T., P. C. PAIROLERO, K. E. HOLLEY, AND J. L. TITUS. Macroscopic enzyme-mapping verification of large, homogeneous, experimental myocardial infarcts of predictable size and location in dogs. J. Thorac. Cardiovasc. Surg. 69: 599-605, 1975. 21. MAXWELL, M. P., D. J. HEARSE, AND D. M. YELLON. Species variation in the coronary collateral circulation during regional myocardial ischaemia: a critical determinant of the rate of evolution and extent of myocardial infarction. Cardiouasc. Res. 21: 737746, 1987. 22. MICKEL, H. S., Y. N. VAISHNAV, 0. KEMPSKI, D. V. LUBITZ, J. F. WEISS, AND G. FEUERSTEIN. Breathing 100% oxygen after global brain ischemia in Mongolian gerbils results in increased lipid peroxidation and increased mortality. Stroke 18: 426-430, 1987. 23. OPIE, L. H. Reperfusion injury and its pharmacologic modification. Circulation 80: 1049-1062, 1989. 24. OTANI, H., K. OMOTO, K. TANAKA, T. SATO, M. UMEMOTO, A. TATSUMI, Y. SAITO, T. OSAKO, M. FUKUNAKA, K. KASAHARA, A. MASUDA, A. NONOYAMA, AND T. KAGAWA. Reperfusion injury induced by augmented oxygen uptake in the initial reperfusion period: possible efficacy of extreme hemodilution. J. Mol. CeZZ. Cardiol. 17: 457-465, 1985. 25. PERRY, M. A., AND S. S. WADWHA. Gradual reintroduction of oxygen reduces reperfusion injury in cat stomach. Am. J. Physiol. 254 (Gastrointest. Liver Physiol. 17): G366-G372, 1988. 26. PRZYKLENK, K., AND R. A. KLONER. Superoxide dismutase plus catalase improve contractile function in the canine model of the Circ. Res. 58: 148-156, 1986. “stunned myocardium.” 27. REIMER, K. A., AND R. B. JENNINGS. Can we really quantitate myocardial cell injury? In: Therapeutic Approaches to Myocardial Infarct Size Limitation, edited by D. J. Hearse and D. M. Yellon. New York: Raven, 1984, p. 163-184. 28. SCHAPER, J., AND W. SCHAPER. Reperfusion of ischemic myocardium: ultrastructural and histochemical aspects. J. Am. Coil. Cardial. 4: 1037-1046, 1983. 29. SIMPSON, P. J., R. F. TODD III, J. C. FANTONE, J. K. MICKELSON, J. D. GRIFFIN, AND B. R. LUCCHESI. Reduction of experimental canine myocardial reperfusion injury by a monoclonal antibody (anti-Mol, anti-CDllb) that inhibits leukocyte adhesion. J. Clin. Invest. 81: 624-629, 1988. J. F., B. A. FREEMAN, AND J. D. CRAPO. Hyperoxia 30. TURRENS, increases H20, release by lung mitochondria and microsomes. Arch. Biochem. Biophys. 217: 411-421, 1982. 31. TURRENS, J. F., B. A. FREEMAN, J. G. LEVITT, AND J. D. CRAPO. The effect of hyperoxia on superoxide production by lung submitochondrial particles. Arch. Biochem. Biophys. 217: 401-410, 1982. 32. YUSUF, S., R. COLLINS, R. PETO, C. FURBERG, M. J. STAMPFER, S. Z. GOLDHABER, AND C. H. HENNEKENS. Intravenous and intracoronary fibrinolytic therapy in acute myocardial infarction: overview of results on mortality reinfarction and side-effects from 33 randomized controlled trials. Eur. Heart J. 6: 556-585, 1985. 33. ZWEIER, J. L., J. T. FLAHERTY, AND M. L. WEISFELDT. Direct measurement of free radical generation following reperfusion of ischemic myocardium. Proc. Natl. Acad. Sci. USA 84: 1404-1407, 1987.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
does not increase
CARLA B. SHNIER, BRIAN A. CASON, ANNE F. HORTON, AND ROBERT F. HICKEY Department of Anesthesia, University of California, San Francisco 94143; and Anesthesiology Service, Veterans Administration Medical Center, San Francisco, California 94121
SHNIER, CARLA B., BRIAN A. CASON, ANNE F. HORTON, AND ROBERT F. HICKEY. Hyperoxemic reperfusion does not increase myocardial infarct size. Am. J. Physiol. 260 (Heart Circ. Physiol. 29): H1307-H1312, 1991.-We tested the hypothesis that arterial hyperoxia during myocardial reperfusion increases reperfusion injury and infarct size. The anterolateral marginal coronary artery of 35 anesthetized rabbits was occluded for 45 min, then reperfused for 3 h with either normoxic [arterial PO, (PaO,) = 96.7 t 22.9 mmHg)] or hyperoxic (Pao, = 554.8 t 61.7 mmHg) blood. In the hyperoxic group only, Paoz was adjusted 10 s before the onset of reperfusion by raising inspired oxygen concentration to 100%. The area of infarction (AI) was defined by triphenyltetrazolium staining, and the area at risk (AR) by fluorescent microspheres. These areas were measured by planimetry. Heart rates and blood pressures did not differ between the two groups during occlusion or reperfusion. Infarct size (AJAR) was 49.1 * 16.5% in the normoxic group (n = 17) and 40.8 ? 16.1% in the hyperoxic group (n = 18). From these data, 90% confidence limits establish that the maximal true increase in AI/AR caused by hyperoxia would be 0%-l %. Hyperoxic reperfusion of ischemic myocardium compared with normoxic reperfusion does not significantly increase myocardial infarct size.
stantially increase blood oxygen content. High arterial PO, (Pao,) may nevertheless increase PO, at important sites of free radical injury, such as endothelial and myocardial cell membranes and leukocytes. Because hyperoxia can increase free radical production, and because previously anoxic tissue is particularly vulnerable to oxygen-mediated injury (12), we hypothesized that hyperoxemic reperfusion might increase reperfusion injury to ischemic myocardium. This issue is important because Pa 0, can be easily controlled in many clinical circumstances and could therefore be adjusted to maximize myocardial salvage during reperfusion of the heart after thrombolysis, angioplasty, or open heart surgery. Because the hypothesis of an adverse effect of Paoz on myocardial reperfusion injury has never been tested in the intact blood-perfused heart, we used a rabbit model of myocardial infarct size to answer the question, does hyperoxemia during myocardial reperfusion increase infarct size?
myocardial
Surgical preparation. With the approval of our hospital’s animal welfare committee, 35 adult New Zealand White rabbits (2.6-5.2 kg) were anesthetized with halothane in oxygen. The trachea was intubated via a tracheostomy, and ventilation was controlled to achieve a normal arterial pH and Pco~. Through a median sternotomy and pericardial incision, the heart was exposed. The anterolateral coronary artery was identified and encircled with a 2-O polyester suture at the midpoint of its epicardial course. The two ends of the suture were threaded loosely through a 6-cm piece of flexible plastic tubing forming a snare that would be tightened to occlude the coronary artery. Teflon catheters (Jelco) were inserted into the left atrium (20 gauge) to measure cardiac filling pressure, into the femoral artery (25 gauge) to measure systemic arterial pressure, and into the femoral vein (20 gauge) to provide intravenous access for fluid and heparin administration. A modified V5 lead electrocardiogram recorded heart rate. Ischemiaprotocol. To determine the effect of hyperoxic reperfusion on infarct size, we used a rabbit model of acute coronary artery occlusion-reperfusion. The left lateral coronary artery was occluded for 45 min, and then, by changing the fraction of inspired oxygen (FIN,), reperfused with blood having either a high Pao, or a
reperfusion
injury;
oxygen free radicals;
rabbit
REPERFUSION of ischemic myocardium limits myocardial infarct size (2, 5), decreasing morbidity and mortality from myocardial infarction (5, 32). However, recent work suggests that reperfusion itself can kill ischemic cells (termed reperfusion injury) (2, 4). Although the mechanisms of reperfusion injury are likely multiple, the most widely accepted mechanism is the generation of toxic oxygen free radicals during reperfusion. This mechanism was suggested by studies showing that reoxygenation of the ischemic or anoxic heart causes tissue damage (9,M) and is supported by studies demonstrating that reperfusion generates oxygen free radicals (3, 33) and other studies in which the administration of free radical scavengers during myocardial reperfusion reduces infarct size or diminishes postischemic functional impairment (1, 16, 26). Additionally, in some free radical-generating systems, higher Po2 leads to greater oxyradical production. In mitochondria, microsomes, and nuclei isolated from various animal sources, rate of free radical production increases during hyperoxia (11, 15, 30, 31). Because of the limited solubility of oxygen in water, increasing the PO, in arterial blood above the normal range does not subEARLY
0363-6135/91
$1.50 Copyright
METHODS
0 1991 the American
Physiological
Society
H1307
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
H1308
HYPEROXIC
REPERFUSION
normal Pao,. After 3 h of reperfusion we compared myocardial infarct size postmortem in the two Pao, groups. Before coronary ligation, we anticoagulated the blood with intravenous heparin sodium (400 U/kg) and lowered the FIN, to as close to 0.21 as possible without producing hypoxemia (Pao, = 95.5 t 13.3 mmHg). For each rabbit, the inspiratory halothane concentration between 0.6 and 1.0 vol/lOO ml that maintained anesthesia and normal hemodynamic values was established and held constant for the duration of the protocol. After a 30-min period of stabilization, we occluded the coronary artery by pulling the coronary ligature taut through the plastic tube and clamping the ligature in position. Hemodynamics, arterial blood gases, hemoglobin concentration, and hemoglobin oxygen saturation were measured before coronary ligation, after 30 min of occlusion, and after 30 min of reperfusion. Left atria1 and systemic arterial pressures were measured with Transpac II strain gauges (Abbott, Chicago, IL) and recorded on a Gilson polygraph (Middleton, WI). Hemoglobin concentration and oxygen saturation were measured with a Radiometer OSM 3 analyzer. Arterial blood gases were measured with a Radiometer ABL 2 analyzer (Copenhagen, Denmark). Reperfusion protocol. The 45 min of myocardial ischemia was followed by 3 h of myocardial reperfusion. Just before coronary ligature release, we assigned the rabbit randomly to one of two treatment groups. The control group underwent myocardial reperfusion with the normal Pao, set during the occlusion; the hyperoxemic group underwent myocardial reperfusion with a high Pao, attained by increasing the F1oz to 1.0 10 s before the moment of coronary ligature release. A rapid change in inspired oxygen concentration was initiated by flushing the ventilation circuit with pure oxygen at 50 l/min. This raised the inspiratory oxygen concentration in the hyperoxemic group to over 90% within 5 s. Arterial blood gases and hemoglobin concentration were measured at 1 and 5 min after ligature release and every 30 min thereafter. At 1 min of reperfusion, Pao, was above 450 mmHg in all hyperoxemic rabbits. Ventricular fibrillation during ischemia or reperfusion was treated by internal defibrillation (1 J/kg). After reperfusion, the rabbit was killed by anesthetic overdose, and myocardial infarct size was measured. Exclusion criteria. Rabbits were excluded from the study if they suffered from 1) ventricular fibrillation unresponsive to two consecutive electrical shocks of 4 J each, 2) mean blood pressure of ~40 mmHg during ischemia, 3) persistent hypoxemia of 0.5).
of area at hyperoxic enclose in infarct one-tailed
AND
INFARCT
H1309
SIZE
limits and were not significantly different between groups at baseline, after 30 min of coronary occlusion, and after 30 min of myocardial reperfusion. As documented in Table 3, baseline Pao, also did not differ significantly between groups until reperfusion, when the Pao, of the experimental rabbits was increased by experimental design. The groups did not differ in rectal temperature, body weight, halothane concentration, or hemoglobin concentration. DISCUSSION
In this study, we found that hyperoxemia during myocardial reperfusion did not increase myocardial infarct size in rabbits subjected to temporary coronary artery occlusion. We hypothesized that hyperoxemic reperfusion of ischemic myocardium would increase infarct size because free radicals have been shown, in some circumstances, to be formed by PoZ-dependent processes (11, 30, 31). One of the fundamental assumptions of this hypothesis is that an increase in Pao, would increase the Po2 at sites where free radicals are generated. Because free radicals have extremely short half-lives, those sites of generation that are important in reperfusion injury are difficult to locate precisely. Previous work demonstrating the “noreflow” phenomenon, increases in microvascular permeability, and leukocyte-mediated tissue damage have implicated the blood and the endothelium as important sites in reperfusion injury (23, 29). It is certain that raising the FI o, to 1.0 in our experimental rabbits did increase the PO, in the arterial blood and likely, too, the Paz in sites close to the blood such as circulating leukocytes and endothelial cells (17). Another assumption made in this experiment was that 3 h of reperfusion are sufficient to allow the development of detectable differences between the experimental groups. This is a reasonable assumption, since free radicals are formed in large part early in reperfusion (33) and brain (22), and heart tissues (14) show increased damage produced by hyperoxia within this time frame. However, injury due to ischemia and reperfusion initiates an inflammatory process that resolves over days to weeks, and reperfused tissue that initially appears viable may nevertheless be “condemned.” A significant problem TABLE
3. Arterial
blood gases Baseline
30 Min Occlusion
30 Min Reperfusion
7.4520.04 7.44t0.06
7.44t0.08 7.46t0.05
7.43t0.06 7.43kO.05
33.9t3.8 34.4t5.2
33.9t4.8 31.8t4.5
35.Ok5.3 32.5t3.6
97.1t11.7 94.4t14.9
97.9t20.3 97.9t11.2
96.7t22.9 554.8261.7
14.4tl.4 5.0tl.9
14.2t2.1 14.9t1.64
14.1-+1.5 15.9t1.7*
-7
PH Reperfusion Group
~~ Heart rate, beats/min Left atria1 pressure, mmHg Mean arterial pressure, mmHg Rate pressure product X lo-‘, Values
are means
Normal High Normal High Normal High Normal High
Pao, Pao, Pao, Pao, Pao, Pao, Pao, Pao, G
~fr SD for 17 normal
Baseline
30 Min Occlusion
30 Min Reperfusion
278t39 290t30 4.9t2.5 3.9t2.0 61t9 66t9 171t35 192rt39
276t34 289t30 5.3t2.6 4.3t1.9 6lt9 64tll 168t33 185t43
262t25 278&35 4.8t2.4 4.3t2.0 60t7 62t9 158t27 173*39
Pao, and 18 high Pao, rabbits.
Normal Pao, group High Pao, group mmHg ho,, Normal Pao, group High Pao, group Pao2, mmHg Normal Pao, group High Pao, group Arterial 0, content, ml OJdl Normal Pao, group High Pao, group
Values are means t SD for I7 normal Pao, and 18 high Pao, rabbits. * P < 0.05 vs. normal Pao, group at 30 min reperfusion.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
HI310
HYPEROXIC
REPERFUSION
in acute reperfusion studies is that they may therefore overestimate the tissue-sparing effect of treatment, in that initial differences in tissue injury prove inconsequential when reassessed at a later time. In this experiment, however, we found no more tissue-sparing effect from normoxic reperfusion than from hyperoxic reperfusion. In fact, infarct size was slightly higher in the normoxic group (49.1%) than in the hyperoxic group (40.8%), although this difference did not reach statistical significance (P = 0.14 by two-tailed t test). A much larger study of -135 rabbits would be required to establish such a small difference with 80% certainty. The short time course of the study brings up another methodological concern. Because killing the animals at 3 h after coronary release precluded our measuring infarct size by histology, we used a histochemical staining technique. Triphenyltetrazolium chloride marks viable muscle by forming an insoluble red precipitate in the presence of dehydrogenase enzymes and coenzymes. Cells that are irreversibly damaged quickly lose enzyme activity and the ability to reduce tetrazolium salts, so they remain unstained. Because detection of dead tissue depends on absence of enzyme activity, early infarcts may be underestimated due to temporary retention of dehydrogenase enzyme activity in condemned cells (27). However, the triphenyltetrazolium chloride method of infarct size measurement is appropriate in our study for the following reasons. First, this staining technique was shown by ultrastructural analysis to be an accurate method for quantifying infarct size as early as 3 h after coronary occlusion (10). Second, reperfusion as performed in our study accelerates the washout of enzyme activity from cells that are irreversibly damaged and makes spurious staining less likely (28). Third, the hearts in our study were frozen before staining, which lessens artifactual tissue salvage in infarct size studies (7). Finally, if triphenyltetrazolium chloride underestimates the size of the infarct, there is no a priori reason to believe that it would act any differently in the two groups of rabbits. A final concern is that unequal anatomic areas at risk of infarction would bias our results. Though rabbits were randomly assigned to treatment groups after coronary ligation, the area at risk was, by chance, 24% smaller in the hyperoxic group than in the normoxic group. This chance inequality between the groups does not invalidate comparisons between the groups for two reasons. First, although a larger area at risk would favor a larger infarct, infarct size was expressed and compared as a percentage of the corresponding area at risk. Second, because collateral blood flow in the rabbit heart is essentially nil (21), cells within a smaller area at risk were as unlikely to be sustained by collateral flow as those within a larger area at risk. Although there is considerable evidence that reoxygenation of the ischemic or hypoxic tissue can cause irreversible cell injury (13), this is the first study of the effect of hyperoxic blood reperfusion on myocardial infarct size in vivo. The theoretical impetus for this study came from previous work in isolated heart (14), skeletal muscle (19), brain (22), and gut (25), demonstrating the influence of PO2 in reoxygenation-reperfusion injury.
AND
INFARCT
SIZE
The discrepancy between our results, which showed no increase in injury from hyperoxic reperfusion, and these cited above, may be explained by differences in both study design and the criteria used to measure injury. For example, Hearse et al. (14) showed that after a period of oxygen deprivation, reoxygenation of isolated crystalloid-perfused rat hearts caused structural damage and a dose-related massive increase in cardiac enzyme leakage. The relationship between reperfusion Paz and structural damage was not quantified. This study established reoxygenation as one of the potential causative events in reperfusion injury by showing that the extent of cardiac enzyme leakage varies directly with the Pop of the hemoglobin-free perfusate. By contrast, we found no increase in injury resulting from hyperoxic reperfusion. Three differences in study design may account for the discrepancy between Hearse’s findings and ours. First, we used a whole animal blood-perfused heart, whereas Hearse et al. (14) used an isolated crystalloid-perfused heart to study reoxygenation damage. Although reperfusion PO+ varied over a similar range, because the oxygen-carrying capacity of blood is much greater than that of crystalloid, the reperfusates in the two studies provided dissimilar amounts of oxygen. Hearts in our experiments were reperfused at a Pao, of 96.7 mmHg vs. a Paoz of 554.8 mmHg (14.6 vs. 16.0 ml oxygen/100 ml blood). We thus compared the damage of restoring normal to slightly supernormal amounts of oxygen to the ischemic myocardium. By contrast, Hearse’s reperfusate contained no hemoglobin, and so increasing the Po2 from 71 to 677 mmHg restored only subnormal amounts of oxygen (0.2-2.5 ml/l00 ml crystalloid) to the anoxic myocardium. Second, it should be noted that, unlike in Hearse’s study, the measured tissue injury in our study was due to a combination of ischemia, reperfusion, and reoxygenation. The conditions of ischemia and reperfusion were the same in both of our experimental groups, allowing comparison of the effects of different Pozs at reoxygenation. It was possible, however, that injury due to ischemia and injury due to the stress of reperfusion were sufficiently large and variable that they obscured any smaller differences in injury due to the Po2 of the reperfusate. Ischemia and the physical stress of reperfusion were not considerations in Hearse’s continuously perfused hearts and so may account for the disparity in experimental findings between his study and ours. Third, differences in experimental outcome may also be attributable to the presence of formed elements in blood. Blood contains platelets and white cells that can contribute to the apparent reperfusion injury by plugging the microvasculature and preventing complete reflow. Although crystalloid-perfused hearts are not leukocytefree, they do contain fewer white cells than blood-perfused hearts (18). Because white cells are a significant source of free radicals, it may be that blood-reperfused hearts suffer greater obligatory reperfusion injury than crystalloid-perfused hearts, masking any increase in reperfusion injury due to hyperoxia. More recent work has used a variety of blood-perfused tissues to show that, in contrast to our findings, the partial pressure of oxygen in arterial blood is important
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
HYPEROXIC
REPERFUSION
in reperfusion injury (19, 22, 25). In a canine model of temporary skeletal muscle ischemia, Korthius et al. (19) found only small increases in microvascular permeability as a result of anoxic reperfusion but noted large increases in microvascular permeability during normoxic reperfusion. These findings are supported by similar experiments in gastric mucosa (25) where, after ischemia, reperfusion with hypoxic blood caused less microvascular injury than did normoxic blood. In both of these studies the investigators measured microvascular damage in nonworking tissue to quantify reperfusion injury, while we measured infarct size in the beating rabbit heart. Conceivably these two approaches differ in their sensitivity to changes in reperfusion Paz. Additionally, our results differ from these two previous investigations because our questions were different. We did not ask if oxygen contributed to reperfusion injury, but if oxygen in excess of that normally provided to maintain life and function would increase reperfusion injury. This question is more relevant to the treatment of human myocardial reperfusion where maintaining a perfusing Paz below that achievable breathing room air is not acceptable. Defining the role of Po2 in reperfusion injury is complicated by the fact that as Po2 in the reperfusate changes so does oxygen content. Otani et al. (24) examined the effect of oxygen content on reperfusion injury apart from the effect of Po2 by reperfusing isolated dog hearts with different degrees of hemodilution (Hb = 8, 4, or 0 g/dl). They found greater recovery of ventricular function and metabolism in the extreme hemodilution group (Hb = 4 g/dl, PO, = 100 m .mHg) th .an in the moderate hemodilution group (Hb = 8 g/dl, Po2 = 100 mmHg) or the hemoglobin-free plasma group (Po2 = 300 mmHg). This result was consistent with increased oxygen delivery leading to increased reperfusion injury by oxygen free radical formation. An alternate explanation for the benefit of extreme hemodilution is the associated decrease in leukocytes and platelets in the reperfusate or the lowered vascular resistance leading to increased blood flow. It is likely that the reason why hemoglobin-free hearts fared the worst was that reperfusion with oxygenated hemoglobin-free plasma, even at a Po2 of 300 mmHg, provided insufficient oxygen to meet the demands of this damaged myocardium. The advantage of our experimental design compared was twofold. with those of these previous investigators First ? we P rov ided a range of oxygen content th .at would sustain in vivo working heart muscle indefinitely; the hemoglobin-free perfusates used by Hearse et al. and Otani et al. would not. Similarly, the blood perfusates used in the skeletal muscle and gastric mucosa models were severely hypoxemic in the test groups exhibiting less reperfusion injury. Second, our study quantified reoxygenation injury by one of the most important predictors of cardiac morbidity and mortality, infarct size. of the In comparison with infarct size, the significance rate of myocardial enzyme leakage and of microvascular permeability is less certain. Like us, Mickel et al. (22) compared the effect of normoxic to hyperoxic reperfusion in an intact animal. This group demonstrated an increase in lipid peroxidation and an increase in mortality in gerbils breathing
AND
INFARCT
H1311
SIZE
100% oxygen compared with those breathing air after transient global brain ischemia. Cerebral infarct size was not measured. The reason our results do not agree with those of Mickel et al. may be because of differences in species (rabbit vs. gerbil) and in tissue studied (heart vs. brain). Additionally, though there is a strong suggestion that an increase in cerebral reperfusion injury due to arterial hyperoxia caused the increase in mortality, covariant causes such as changes in blood pressure and in ventilation cannot be ruled out. Based on our study, we speculate that the administration of supplemental oxygen to otherwise normoxic patients undergoing myocardial reperfusion after acute coronary artery occlusion will not increase infarct size. However, the extrapolation of these experimental results to humans should be made with caution. First, there may be significant species differences in free radical-producing systems, and these differences may determine sensitivity to therapeutic interventions. For example, xanthine oxidase, an important producer of superoxide radical in the dog and rat, is absent from the myocardium of rabbits and humans (6,8). Therefore, inhibitors of xanthine oxidase may be expected to reduce reperfusion injury in dogs and rats, but not in rabbits or humans. The relevant free radical-generating systems operating in human myocardial reperfusion injury are, at present, unknown. However, rabbit hearts do resemble human hearts at least insofar as neither generates free radicals by the well-studied xanthine oxidase system. Second, other experimental conditions such as the severity of ischemia and rapidity of reperfusion may also affect the extent of reperfusion injury (23). The current study modeled only one situation, absolute ischemia followed by sudden maximal reperfusion. This differs from the situation in older humans where, because of coronary collateral vessels, coronary occlusion may produce submaximal ischemia in much of the anatomic area at risk and reperfusion through diseased vessels may be more gradual. Lesser degrees of ischemia and gradual reintroduction of oxygen may lessen the amount of reoxygenation-reperfusion injury (14). Because we found no increase in infarct size with hyperoxic reperfusion under severe ischemia-reperfusion conditions, there is unlikely to be one in less extreme clinical situations. In summary, although oxygen free radicals are implicated in myocardial reperfusion injury, we found that hyperoxic reperfusion did not increase myocardial infarct size in the rabbit. This work was supported in part by a grant to B. A. Cason from the Veterans Administration Research Advisory Group. C. B. Shnier was supported by a Research Fellowship Award of the Heart and Stroke Foundation of Canada. Address for reprint requests: B. A. Cason, Anesthesiology Service (129), Veterans Administration Medical Center, 4150 Clement St., San Francisco, CA 94121. Received
17 April
1990; accepted
in final
form
5 December
1990.
REFERENCES 1. AMBROSIO, G., L. C. BECKER, G. M. HUTCHINS, H. F. WEISMAN, AND M. L. WEISFELDT. Reduction in experimental infarct size by recombinant human superoxide dismutase: insights into the pathophysiology of reperfusion injury. Circulation 74: 1424-1433, 1986.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
H1312
HYPEROXIC
REPERFUSION
2. BECKER, L. C., AND G. AMBROSIO. Myocardial consequences of reperfusion. Prog. Cardiouasc. Dis. 30: 23-44, 1987. 3. BOLLI, R., B. S. PATEL, M. 0. JEROUDI, E. K. LAI, AND P. B. MCCAY. Demonstration of free radical generation in “stunned” myocardium of intact dogs with the use of the spin trap a-phenyl N-tert-butyl nitrone. J. CLin. Inuest. 82: 476-485, 1988. 4. BRAUNWALD, E., AND R. A. KLONER. Myocardial reperfusion: a double-edged sword? J. Clin. Invest. 76: 1713-1719, 1985. 5. CAMPBELL, C. A., K. PRZYKLENK, AND R. A. KLONER. Infarct size reduction: a review of the clinical trials. J. CZin. Pharmacol. 26: 317-329,1986. 6. DOWNEY, J. M., T. MIURA, L. J. EDD, D. E. CHAMBERS, T. MELLER, D. J. HEARSE, AND D. M. YELLON. Xanthine oxidase is not a source of free radicals in the ischemic rabbit heart. J. Mol. Cell CardioZ. 19: 1053-1060, 1987. 7. DOWNEY, J. M., C. SHIRATO, T. MIURA, AND T. TOYOFUDU. Tetrazolium is unreliable as an index of drug-induced salvage (Abstract). J. Mol. Cell. Cardiol. 20, Suppl. V: S-70, 1988. 8. EDDY, L. J., J. R. STEWART, H. P. JONES, T. D. ENGERSON, J. M. MCCORD, AND J. M. DOWNEY. Free radical-producing enzyme, xanthine oxidase, is undetectable in human hearts. Am. J. Physiol. 253 (Heart Circ. Physiol. 22): H709-H711, 1987. 9. FEUVRAY, D., AND J. D. LEIRIS, Ultrastructural modifications induced by reoxygenation in the anoxic isolated rat heart perfused without exogenous substrate. J. Mol. Cell. Cardiol. 7: 307-314,1975. 10. FISHBEIN, M. C., S. MEERBAUM, J. RIT, U. LANDO, K. KANMATSUSE, J. C. MERCIER, E. CORDAY, AND W. GANZ. Early phase acute myocardial infarct size quantification: validation of the triphenyl tetrazolium chloride tissue enzyme staining technique. Am. Heart J. 101: 593-600,198l. 11. FREEMAN, B. A., AND J. D. CRAPO. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J. BioL. Chem. 256: 10986-10992,198l. 12. GUARNIERI, C., F. FLAMIGNI, AND C. M. CALDARERA. Role of oxygen in the cellular damage induced by re-oxygenation of hypoxic heart. J. Mol. Cell. Cardiol. 12: 797-808, 1980. 13. HEARSE, D. J. Reperfusion of the ischemic myocardium. J. Mol. Cell Cardiol. 9: 605-616, 1977. 14. HEARSE, D. J., S. M. HUMPHREY, AND G. R. BULLOCK. The oxygen paradox and the calcium paradox: two facets of the same problem? J. Mol. Cell. Cardiol. 10: 641-668, 1978. 15. JAMIESON, D., B. CHANCE, E. CADENAS, AND A. BOVERIS. The relation of free radical production to hyperoxia. Annu. Reu. Physiol. 48: 703-719, 1986. 16. JOLLY, S. R., W. J. KANE, M. B. BAILIE, G. D. ABRAMS, AND B. R. LUCCHESI. Canine myocardial reperfusion injury: its reduction by the combined administration of superoxide dismutase and catalase. Circ. Res. 54: 277-285, 1984. in oxidative stress: 17. JONES, D. P. The role of oxygen concentration hypoxic and hyperoxic models. In: Oxidatiue Stress, edited by H. Sies. London: Academic, 1985, p. 151-195. AND 18. KELLER, A. M., R. M. CLANCY, M. L. BARR, C. C. MARBOE, P. J. CANNON. Acute reoxygenation injury in the isolated rat heart: role of resident cardiac mast cells. Circ. Res. 63: 1044-1052, 1988. 19. KORTHIUS, R. J., J. K. SMITH, AND D. L. CARDEN. Hypoxic renerfusion attenuates postischemic microvascular injury. Am. J.
AND
INFARCT
SIZE
Physiol. 256 (Heart Circ. Physiol. 25): H315-H319, 1989. 20. LIE, J. T., P. C. PAIROLERO, K. E. HOLLEY, AND J. L. TITUS. Macroscopic enzyme-mapping verification of large, homogeneous, experimental myocardial infarcts of predictable size and location in dogs. J. Thorac. Cardiovasc. Surg. 69: 599-605, 1975. 21. MAXWELL, M. P., D. J. HEARSE, AND D. M. YELLON. Species variation in the coronary collateral circulation during regional myocardial ischaemia: a critical determinant of the rate of evolution and extent of myocardial infarction. Cardiouasc. Res. 21: 737746, 1987. 22. MICKEL, H. S., Y. N. VAISHNAV, 0. KEMPSKI, D. V. LUBITZ, J. F. WEISS, AND G. FEUERSTEIN. Breathing 100% oxygen after global brain ischemia in Mongolian gerbils results in increased lipid peroxidation and increased mortality. Stroke 18: 426-430, 1987. 23. OPIE, L. H. Reperfusion injury and its pharmacologic modification. Circulation 80: 1049-1062, 1989. 24. OTANI, H., K. OMOTO, K. TANAKA, T. SATO, M. UMEMOTO, A. TATSUMI, Y. SAITO, T. OSAKO, M. FUKUNAKA, K. KASAHARA, A. MASUDA, A. NONOYAMA, AND T. KAGAWA. Reperfusion injury induced by augmented oxygen uptake in the initial reperfusion period: possible efficacy of extreme hemodilution. J. Mol. CeZZ. Cardiol. 17: 457-465, 1985. 25. PERRY, M. A., AND S. S. WADWHA. Gradual reintroduction of oxygen reduces reperfusion injury in cat stomach. Am. J. Physiol. 254 (Gastrointest. Liver Physiol. 17): G366-G372, 1988. 26. PRZYKLENK, K., AND R. A. KLONER. Superoxide dismutase plus catalase improve contractile function in the canine model of the Circ. Res. 58: 148-156, 1986. “stunned myocardium.” 27. REIMER, K. A., AND R. B. JENNINGS. Can we really quantitate myocardial cell injury? In: Therapeutic Approaches to Myocardial Infarct Size Limitation, edited by D. J. Hearse and D. M. Yellon. New York: Raven, 1984, p. 163-184. 28. SCHAPER, J., AND W. SCHAPER. Reperfusion of ischemic myocardium: ultrastructural and histochemical aspects. J. Am. Coil. Cardial. 4: 1037-1046, 1983. 29. SIMPSON, P. J., R. F. TODD III, J. C. FANTONE, J. K. MICKELSON, J. D. GRIFFIN, AND B. R. LUCCHESI. Reduction of experimental canine myocardial reperfusion injury by a monoclonal antibody (anti-Mol, anti-CDllb) that inhibits leukocyte adhesion. J. Clin. Invest. 81: 624-629, 1988. J. F., B. A. FREEMAN, AND J. D. CRAPO. Hyperoxia 30. TURRENS, increases H20, release by lung mitochondria and microsomes. Arch. Biochem. Biophys. 217: 411-421, 1982. 31. TURRENS, J. F., B. A. FREEMAN, J. G. LEVITT, AND J. D. CRAPO. The effect of hyperoxia on superoxide production by lung submitochondrial particles. Arch. Biochem. Biophys. 217: 401-410, 1982. 32. YUSUF, S., R. COLLINS, R. PETO, C. FURBERG, M. J. STAMPFER, S. Z. GOLDHABER, AND C. H. HENNEKENS. Intravenous and intracoronary fibrinolytic therapy in acute myocardial infarction: overview of results on mortality reinfarction and side-effects from 33 randomized controlled trials. Eur. Heart J. 6: 556-585, 1985. 33. ZWEIER, J. L., J. T. FLAHERTY, AND M. L. WEISFELDT. Direct measurement of free radical generation following reperfusion of ischemic myocardium. Proc. Natl. Acad. Sci. USA 84: 1404-1407, 1987.
Downloaded from www.physiology.org/journal/ajpheart by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on September 11, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
