Background The present study was designed to explore the relation between the duration of ischemia and the rate and extent of myocardial functional recovery after reperfusion.
Methods and Results Isolated rat hearts were perfused with blood from a support animal for 15 minutes (flow rate, 2.5 mL/min; perfusion pressure, 60.1±1.3 mm Hg). Control left ventricular developed pressure (LVDP) was measured, and the hearts (six per group) were subjected to 10, 20, 30, 40, 50, 60, 70, or 80 minutes of global ischemia (37°C) and 60 minutes of reperfusion. Pacing (320 beats per minute) was instituted before and after ischemia. In all groups, transient arrhythmias occurred at the onset of reperfusion, to be followed by an early phase of recovery that peaked after 2 to 3 minutes of reperfusion. The relation between the extent of this initial recovery and the duration of preceding ischemia was described by a bell-shaped curve. Thus, the maximum initial mean recovery after 10, 20, 30, 40, 50, 60, 70, or 80 minutes of ischemia was 97%, 108%, 145%, 154%, 118%, 34%, 41%, and 24%, respectively, of preischemic LVDP. Possibly indicative of reperfusion-induced injury, LVDP then declined in all groups so that after 20 minutes of reperfusion, the mean recovery was 63%, 53%, 48%, 50%, 56%, 12%, 9%, and 5%, respectively. In the 10-, 20-, 30-, and 40-minute ischemia groups, there then was a secondary increase in LVDP, possibly indicating the start of recovery from stunning. After 60 minutes of reperfusion, the mean recovery of LVDP was 82%, 65%, 59%, 54%, 47%, 9%, 7%, and 4%, respectively; this second phase of recovery was inversely proportional to the duration of ischemia. To define the early phase of recovery that had been obscured by reperfusion-induced arrhythmias, we repeated the experiments with the inclusion of a cardioplegic infusion (St Thomas’ solution for 2 minutes before ischemia). This significantly reduced the incidence of ventricular fibrillation during early reperfusion. The extent of the initial postischemic recovery of LVDP was similar to that observed without cardioplegia; however, the mean secondary recovery was greater in all groups. Again, the relation of early transient (2 to 5 minutes) recovery to the duration of ischemia was represented by a bell-shaped curve, whereas the secondary recovery was inversely related.
Conclusions Although the results of the present study confirm the protective properties of cardioplegia, they also shed some light on the nature of reperfusion-induced injury and myocardial stunning and their complex relation to the severity of the preceding ischemia.
The mechanisms and pathological significance of myocardial stunning are attracting considerable attention.1 2 Myocardial stunning is considered to be a consequence of “reperfusion-induced injury” sustained during the early minutes of reperfusion after a brief (10- to 20-minute) period of severe global or regional ischemia. Stunning occurs in the absence of irreversible damage and may involve a component of ischemic as well as reperfusion injury. Conventionally, stunning is described as a postischemic contractile deficit from which the heart slowly, but nevertheless completely, recovers. However, one of the authors3 has drawn attention to the fact that the profile for functional recovery during stunning is more complex in that on reperfusion there is a rapid initial recovery of contractile function to a level that approaches or even exceeds the preischemic control value. However, this recovery is transient (as long as 5 minutes) and is followed by a rapid decline after which there may be a second but very slow recovery of function. Although the early transient recovery is seldom reported, it has been widely observed in several species both in vivo and in vitro. There are a number of possible explanations for this complex recovery profile. One is that the heart recovers quickly and completely from the reversible injury sustained during ischemia only to be further injured by transient reperfusion-induced injury that occurs during the early moments of reperfusion. Once this injury abates, the heart slowly recovers to its full functional capacity.
The primary objective of the present study in the isolated blood-perfused rat heart was to characterize the relations between the duration of ischemia and the extents of the early and late recoveries of function. In addition, our aim was to discover whether a cardioprotective intervention such as cardioplegia would modify either of these relations.
Animals and Cardioplegic Solution
Wistar rats (400 to 600 g) of either sex were used as support animals, and the isolated hearts were excised from 250- to 300-g male rats of the same strain. The animals were treated in compliance with the “Guidelines for the Care and Use of Laboratory Animals” published by the National Institutes of Health (NIH publication No. 85-23, revised 1985). St Thomas’ Hospital cardioplegic solution no. 2 (for composition, see Table 1⇓) was used in the study and was filtered through a 0.5-μm filter before infusion.
Isolated Blood-Perfused Rat Heart Preparation
The isolated blood-perfused rat heart preparation used in the present study (Fig 1⇓) has been described previously4 and was a modification of the original design of Gamble et al.5 A peristaltic pump (LKB Bromma 12000 Varioperspex) was interposed in the arterial portion of the circuit between the support animal and the isolated heart. The presence of the pump and a latex rubber balloon resulted in a compliant, bubble-free perfusion that could deliver blood to the isolated heart at a constant flow rate (2.5 mL/min). A separate port between the pump and the isolated heart enabled the aortic pressure to be monitored for calculation of coronary vascular resistance. The perfusion circuit was maintained at 37°C by connection in series to a circulator with an integral heater (Techne C-85A). All tubing was constructed from silicone rubber because of biocompatibility.6
Each support rat was anesthetized with pentobarbital (60 mg/kg IP) and received heparin (1000 U/kg IV). A femoral artery and vein were cannulated with 22-gauge Abbocath-T catheters for the supply and return, respectively, of blood. Blood was returned by gravity to the support animal through a 200-μm filter (Avon A100) with the flow regulated with an adjustable clamp. The unrestrained support rat was placed supine on a thermostatically controlled operating table set at 37°C. The animal was allowed to breathe spontaneously through a face mask supplied with air mixed with Carbogen (British Oxygen Company) (95% O2–5% CO2). The mixture was adjusted to maintain oxygen saturation at more than 99% and Po2 and Pco2 within the ranges of 150 to 250 and 45 to 60 mm Hg, respectively. Additional pentobarbital (6 mg IP) was administered as necessary, with care taken to maintain the mean blood pressure of the support animal in the range of 95 to 100 mm Hg; an increase in heart or respiratory rate was the indication for additional anesthesia. Monitoring was continuous and transfusion of blood from another rat of the same strain was limited to 2 mL throughout the experiment. Blood glucose was measured hourly and maintained between 9 and 12 mmol/L by injection of 0.1 mL 50% glucose as necessary into the circuit. Hourly, pH and base excess measurements were taken; these values remained in the ranges of 7.25 to 7.32 and −5.2 to +1.1 mmol/L, respectively, without the need for the addition of any buffer. After femoral cannulation, the extracorporeal flow was initiated and gradually increased to 2.5 mL/min until a steady state was achieved before isolated hearts were inserted into the perfusion circuit. Each support rat was used for two to four experiments, after which it was killed with a lethal dose of pentobarbital (300 mg IV).
Each donor rat was anesthetized with diethyl ether and a sublethal dose of pentobarbital (300 mg/kg IP) and received heparin (1000 U/kg IV). The heart was then rapidly excised, immersed in saline at 4°C, and immediately transferred to the perfusion circuit, with care taken to exclude all air from the aorta and cannula before cannulation. A microthermistor was inserted into the right ventricle through the pulmonary trunk. Pacing electrodes were attached to the right atrial appendage and the aorta, and the heart was paced at 320 beats per minute. Left ventricular function was measured with a 0.3-mL intraventricular balloon connected to a pressure transducer and 100-μL syringe. The heart was then sealed in a temperature-regulated chamber and maintained at 37°C. The balloon was inflated to achieve a left ventricular end-diastolic pressure (LVEDP) of 2 mm Hg, and continuous recordings were made on a Gould (model 2107) recorder. The left ventricular developed pressure (LVDP) was calculated by subtracting the LVEDP from the left ventricular systolic pressure (LVSP). The isolated heart was perfused aerobically for 15 minutes before ischemia was induced; this allowed sufficient time for gradual inflation of the balloon to avoid the mechanical induction of arrhythmias. Preischemic control recordings were obtained for LVSP, LVEDP, and aortic root pressure (an index of coronary vascular resistance). Ischemia was induced by clamping the blood infusion line and unclamping the parallel line (Fig 1⇑). In the second series of studies, cardioplegic solution (37°C) was infused through the aortic cannula at a pressure of 60 cm H2O for 2 minutes, and the solution was removed from the circuit as it left the heart. To avoid arrhythmias resulting from its inflation during early reperfusion, the intraventricular balloon was not deflated during the period of ischemia.
After the 15-minute period of aerobic perfusion, hearts (six per group) were randomized into 16 groups that were rendered ischemic at 37°C for 10, 20, 30, 40, 50, 60, 70, or 80 minutes with or without a preischemic cardioplegic infusion. Pacing was stopped at the onset of ischemia and restarted at the time of reperfusion. During the 60 minutes of reperfusion, measurements of postischemic recovery were recorded at defined intervals. To minimize any variability arising from the use of different support animals, the hearts in the cardioplegia versus noncardioplegia series that were subjected to the same ischemic period were, whenever possible, perfused from the same support animal. Similarly, the order in which isolated hearts were perfused was reversed in experiments with separate support animals to minimize the possible effect of factors released from one heart on the function of a subsequent heart.
Hearts were rendered ischemic by diverting the blood away from the heart through a parallel line that returned the blood directly to the venous reservoir. Continuous extracorporeal circulation was chosen in preference to stopping and starting the infusion pump since it would avoid (1) stasis of blood in the circuit and (2) sudden changes in the blood pressure of the support rat with consequent neurohormonal changes that might affect the function of the isolated heart.
During administration of the cardioplegic solution, the perfusion cannula and isolated heart were raised above the level of the heart chamber. In this way the cardioplegic solution could be collected, avoiding its transfusion into the support rat. The volume of cardioplegic solution infused into each heart was measured. The heart was then lowered into its chamber, which was resealed. Administration of the cardioplegic solution at 37°C minimized any cooling effect that resulted from exposure of the heart to ambient temperature for 2 minutes (the transient drop in temperature measured by the right ventricular thermistor was less than 1°C and lasted for less than 4 minutes). Reperfusion was achieved by clamping the parallel line and redirecting the blood to the isolated heart. Atrial pacing was restarted at the time of reperfusion. Measurements of LVEDP and LVSP were made at 1-minute intervals during the first 5 minutes of reperfusion and at 5-minute intervals thereafter. The LVEDP was adjusted as necessary (the change in the volume of the intraventricular balloon was less than 0.005 mL over 60 minutes) so as to match the preischemic value. Measurements of aortic root pressure were taken at 5-minute intervals during the 60 minutes of reperfusion.
To confirm the adequacy of perfusion and oxygenation in the blood-perfused preparation, arterial and venous oxygen contents were measured in a subgroup of six hearts subjected to 40 minutes of ischemia without cardioplegia. Arterial samples for these measurements were taken from a tap situated between the femoral artery and the pump. (The tap was located at this point so as not to decrease the blood flow to the isolated heart.) To avoid contact with air, venous samples were obtained through a cannula introduced into the right ventricle and secured with a ligature around the pulmonary trunk. The volume of each sample was restricted to 0.25 mL, and samples were obtained before the induction of ischemia and at 1, 5, and 20 minutes after the onset of reperfusion.
Of the 122 isolated hearts initially entered into the study, 26 were excluded on the basis of the following predefined criteria: (1) hearts developing sustained ventricular arrhythmias that would preclude the accurate measurement of preischemic function (7 were excluded at this stage); (2) preischemic aortic root pressure exceeding 90 mm Hg (pressures above this value were associated with poor preischemic function, probably as a consequence of particulate or gaseous embolism of the coronary bed; 4 hearts were excluded on these grounds); (3) cardiovascular instability of the support animal, secondary to hypovolemia due to momentary reductions in femoral venous return, during any part of the perfusion period (since this had a profound effect on the isolated heart, usually resulting in a temporary increase in LVDP) or death of the support animal (8 hearts were excluded on these grounds); and (4) technical failures during cannulation of the heart (eg, poor insertion of the balloon) or inadequate administration of cardioplegic solution (7 hearts were excluded for these reasons). Excluded hearts were immediately replaced.
Values are given as mean±SEM. A two-way ANOVA was performed in different groups of data, and association between variables was determined by linear regression analysis. A value of P<.05 was considered statistically significant.
Preischemic Control Values
There were no significant differences between any of the groups for mean LVDP or aortic root pressure. The value for preischemic mean LVDP for all groups combined was 100.7±0.3 mm Hg (measured at an LVEDP of 2 mm Hg). The combined mean preischemic aortic root pressure for all groups was 60.1±1.3 mm Hg.
Table 2⇓ shows the mean preischemic and postischemic arterial and venous oxygen content, oxygen saturation, and partial pressure values in the selected group that was rendered ischemic for 40 minutes without cardioplegic protection. Venous blood values were within the normal range, suggesting that blood and oxygen delivery were adequate for the needs of the isolated heart.
Arrhythmias were analyzed according to the Lambeth Conventions.7 Fig 2A⇓ shows the relation between the duration of preceding ischemia and the number of hearts fibrillating during the first, second, third, and fourth minutes of reperfusion in the first series of studies, in which cardioplegia was omitted before ischemia. Fig 2B⇓ shows the corresponding results for the second series of studies, in which cardioplegia was used.
Increasing durations of ischemia were associated with an increasing vulnerability to ventricular fibrillation (VF) at the onset of reperfusion. Thus, after 10, 20, 30, 40, 50, 60, 70, and 80 minutes of ischemia, 17%, 0%, 50%, 100%, 100%, 100%, 100%, and 100%, respectively, of the hearts were fibrillating during the first minute of reperfusion. However, the fibrillation was not sustained, and in some groups most hearts had reverted to sinus rhythm by the second or third minute of reperfusion. By 5 minutes of reperfusion, all hearts in all groups had reverted to sinus rhythm. As can be seen from Figs 2A⇑ and 3⇓, the relation between the severity of reperfusion-induced VF (as expressed by incidence and duration) and the duration of the preceding ischemia conforms with the well-established bell-shaped curve.8 As is discussed, the high incidence of VF early during reperfusion in some groups (notably those with 50 and 60 minutes of ischemia) made it difficult to define the initial phase of the early recovery profile for contractile function.
When cardioplegia was given before the period of ischemia, there was a dramatic reduction in the incidence and duration of reperfusion-induced VF in all groups (Fig 2B⇑). Few hearts developed VF (except in the group reperfused after 60 minutes of ischemia), and all hearts in all groups reverted to sinus rhythm by the third minute of reperfusion.
Postischemic Functional Recovery Profiles (Noncardioplegia Series)
Fig 4⇓ shows the profile for the recovery of LVDP during 60 minutes of reperfusion after 10, 20, 30, 40, 50, 60, 70, and 80 minutes of ischemia.
Transient Early Recovery (0 to 20 Minutes of Reperfusion)
After each duration of ischemia (even after 80 minutes), there was an initial recovery of function (in most groups reaching its maximum after 2 to 3 minutes of reperfusion). However, in all groups, LVDP then declined over the ensuing 20 minutes.
Due to the high incidence of arrhythmias during the first 4 minutes of reperfusion, it was not possible to quantify accurately the early peak in all groups (especially in those after 40, 50, and 60 minutes of ischemia). However, despite this complication, it is clear that the extent of the early recovery was not related in a simple manner to the duration of the preceding ischemia. Thus, in the 10- and 30-minute ischemia groups (in which arrhythmias did not prevent the detection of maximal early recovery), the surprising observation was made that the maximum early LVDP was greater after 30 minutes of ischemia (145.0±9.3) than after 10 minutes (96.5±9.2, P<.05). However, despite the occurrence of arrhythmias, there was evidence to suggest that the relation between the duration of ischemia and the early recovery of LVDP (measured after 5 minutes) was described by a bell-shaped curve (Fig 5A⇓).
As can be seen from Fig 4⇑, by 20 minutes of reperfusion, LVDP had declined to its lowest point in all groups (63.4±8.4%, 52.5±3.4%, 47.5±2.9%, 49.5±4.2%, 56.1±7.4%, 12.3±9.6%, 9.0±4.9%, and 5.0±2.2% in the 10-, 20-, 30-, 40-, 50-, 60-, 70-, and 80-minute ischemia groups, respectively). In the 20-, 30-, 40-, and 50-minute ischemia groups, this value was significantly lower (P<.0.5) than the peak observed during the early phase of recovery.
Venous oxygen content and saturation measured at 1, 5, and 20 minutes of reperfusion (Table 2⇑) were within normal ranges and did not correlate with contractile function in the group under investigation (n=6; 40 minutes of ischemia without cardioplegia). The observed decline in function could not, therefore, be explained in terms of inadequate oxygen supply.
Secondary Recovery Profiles (20 to 60 Minutes of Reperfusion)
Beyond 20 minutes of reperfusion, there was no further significant decline in LVDP in any group. In the 50-, 60-, 70-, and 80-minute ischemia groups, steady-state function was maintained for the remainder of the experiment. However, in the 10-, 20-, 30-, and 40-minute ischemia groups, a slow secondary phase of recovery was apparent. Thus, mean LVDP in these groups recovered from their 20-minute values of 63.4±8.4%, 52.5±3.4%, 47.5±2.9%, and 49.5±4.2% to 82.1±9.4%, 65.0±4.2%, 59.3±1.5%, and 53.7±2.5%, respectively, by 60 minutes of reperfusion (this increase in function did not reach statistical significance).
Relations Between Duration of Ischemia and Early and Late Recoveries of Function
As is evident from Fig 6A⇓ (peak recorded early recovery) and Fig 5A⇑ (recovery after 5 minutes of reperfusion), there is a bell-shaped relation between the duration of ischemia and the extent of early recovery. It is also clear that the greatest early recovery was seen in the 30- and 40-minute ischemia groups. However, the occurrence of transient reperfusion arrhythmias prevented absolute definition of the peak recovery.
In contrast to the early transient recovery of LVDP, the recovery after 60 minutes of reperfusion (Fig 7A⇓) was described by a simple inverse relation (r=.97) to the duration of preceding ischemia.
Postischemic Functional Recovery Profiles (Cardioplegia Series)
In general, the results of this series of experiments reflected those obtained in the experiments carried out without the use of cardioplegia.
Transient Early Recovery (0 to 20 Minutes of Reperfusion)
The dramatic reduction in the incidence and duration of reperfusion arrhythmias consequent to the use of cardioplegia (Fig 2B⇑) allowed the early recovery phase after each duration of ischemia to be clearly defined (Fig 8⇓). The peak recovery of LVDP took place at between 2 and 5 minutes of reperfusion and in each instance was followed by a major decline in LVDP, which fell to a minimum between 10 and 25 minutes of reperfusion.
As with the first series of experiments, the relation between the early transient recovery of LVDP and the duration of the preceding ischemia was described by a bell-shaped curve. This was seen regardless of whether recovery was expressed in terms of the LVDP after 5 minutes of reperfusion (Fig 5B⇑) or the peak LVDP measured during the first 5 minutes of reperfusion (Fig 6B⇑).
Secondary Recovery Profiles (20 to 60 Minutes of Reperfusion)
In the 70- and 80-minute ischemia groups, a steady state of depressed function was observed, whereas a secondary recovery was apparent in the 10-, 20-, 30-, 40-, 50-, and 60-minute ischemia groups (Fig 8⇑). Thus, LVDP in these latter groups recovered from the 20-minute values of 76.3±9.3%, 61.6±6.6%, 37.5±8.3%, 52.4±7.7%, 48.0±4.4%, and 45.7±8.8% to 96.9±6.0%, 79.2±4.1%, 62.0±3.8%, 58.5±8.2%, 57.6±2.5%, and 55.8±5.5%, respectively, by 60 minutes of reperfusion. In the 20- and 30-minute ischemia groups, the 60-minute value was significantly greater (P<.05) than the 20-minute value.
As with the preceding series of experiments, there was an inverse relation (r=.96) between the duration of ischemia and the postischemic LVDP measured after 60 minutes of reperfusion (Fig 7B⇑).
Protection by Cardioplegia
Comparison of the results from the first and second series of experiments provides some information on the protective properties of cardioplegia. First (not previously reported), the use of a preischemic infusion of cardioplegia provides a profound protection against postischemic arrhythmias. This protection (Fig 2B⇑ compared with Fig 2A⇑) was afforded regardless of the duration of ischemia.
With regard to the postischemic recovery of LVDP, cardioplegia unexpectedly failed to improve the early transient peak of recovery but increased the mean recovery of LVDP during the second phase of recovery in all groups studied so that the mean recovery in the cardioplegia groups was always greater than that in the time-matched noncardioplegic group (Fig 7A⇑ and 7B⇑). However, these differences were statistically significant only in the 60-, 70-, and 80-minute ischemia groups (55.8±5.5%, 37.4±7.2%, and 22.3±6.3% versus 9.1±6.1%, 7.5±3.9%, and 4.0±1.7%, respectively; P<.05 for each).
Several distinct findings emerged from the present study. First, the profile of the postischemic recovery of contractile function was not a simple progression but rather a biphasic phenomenon characterized by an early transient recovery that usually peaked within 5 minutes and subsequently declined. Second, the extent of the early recovery was not linearly related to the duration of the preceding ischemic period; instead, the relation was described as a bell-shaped curve in which early recovery increased with increasing durations of ischemia (10 to 40 minutes of ischemia) and then subsequently declined. Third, over a relatively narrow range of ischemic durations (20 to 50 minutes), the extent of the transient recovery might exceed the preischemic control value. Fourth, LVDP declined from its early peak over 15 to 20 minutes and generally reached its lowest point by 20 minutes of reperfusion. Fifth, after moderate ischemic insults (up to 40 minutes in the noncardioplegic group and up to 60 minutes in the cardioplegia group), there was a progressive secondary recovery of LVDP, which slowly continued throughout the remaining 40 minutes of the reperfusion period under study. Sixth, in the hearts subjected to more severe ischemia, the depressed function remained at steady state for the remainder of the reperfusion period. Last, in contrast to the early recovery profile, the recovery of LVDP after 60 minutes of reperfusion was inversely related to the duration of the preceding ischemia—the longer the ischemia, the poorer was the recovery of function.
In addition, results of the present study have shown that, as expected, cardioplegia slowed ischemic injury so that reperfusion resulted in a superior secondary recovery of contractile function, especially after longer durations of ischemia (60 to 80 minutes, when striking protection was observed). Cardioplegia did not appear to influence the extent of the early transient recovery. However, cardioplegia was able to exert a profound protective effect against reperfusion-induced arrhythmias during this phase.
Our findings illustrate the potential dangers of using short durations (less than 30 minutes) of reperfusion when assessing the functional consequences of ischemia and the ability of interventions to protect against injury. In addition, they allow some speculation on the existence and significance of “reperfusion-induced injury,” particularly in the context of myocardial stunning.
Does Reversible Reperfusion-Induced Injury Exist?
The very existence of both irreversible (ie, causing necrosis) and reversible (ie, contributing to stunning) reperfusion-induced injury is controversial; the arguments have been reviewed elsewhere.2 One of the most convincing pieces of evidence for the existence of reversible reperfusion-induced injury comes from one of the many studies by Bolli et al9 in which mercaptopropionyl glycine was administered to dog hearts subjected to 15 minutes of regional ischemia and 4 hours of reperfusion. A major attenuation of stunning was observed. The crucial feature of the study was that the drug was given only 1 minute before reperfusion. Mercaptopropionyl glycine is a potent antioxidant that is believed to be able to attenuate the known burst of free radical production during the early minutes of reperfusion. Bolli et al9 also showed that antioxidant interventions were protective if given 1 minute before reperfusion but were relatively ineffective if given 1 minute after reperfusion. Taken together, these (and other) findings lend support to the hypothesis that a transient burst of free radical production during the early moments of reperfusion imposes an early and transient oxidant stress on the heart.
It is further hypothesized3 that this stress (possibly by promoting the oxidation of sulfhydryl groups at or near the active center of key proteins and enzymes and/or altering the redox state of glutathione) has an adverse effect on the regulation of proteins responsible for controlling ions involved in excitation and contraction. The consequent ionic perturbations then result in disturbances of cardiac contraction (stunning) and rhythm (reperfusion-induced arrhythmias). Finally, it has been proposed3 that the recovery from this oxidative stress may be rapid (simple reduction of oxidized thiol groups) or slow (the resynthesis and replacement of damaged proteins, a process that may require hours or days to accomplish).
The results of the present study (both with and without cardioplegia) strongly suggest that after periods of global ischemia (as long as 50 to 60 minutes), the rat heart has the inherent ability to recover 100% of its preischemic function. This would indicate that ischemia per se did not inflict lethal or long-lasting damage. Thus, in the noncardioplegic series of experiments, the mean peak recovery of LVDP at 2 minutes of reperfusion after 30 minutes of ischemia was 145.0±9.3%. Similarly, in the cardioplegia series with 50 minutes of ischemia, LVDP initially recovered to 137.5±24.1% of its preischemic value. Because all hearts were paced, this transient hyperactivity cannot be attributed to any rate-dependent phenomenon. Thus, as depicted in Fig 9⇓, it appears that during this early and rapid phase of recovery, factors come into play to attenuate and eventually reverse the high level of functional recovery. As a consequence, LVDP declines and, depending on the severity of the factors, may either never recover (indicative of infarction) or recover very slowly (possibly indicative of a process requiring the repair or replacement of proteins).
A possible alternative explanation for the biphasic recovery profile might involve transient stimulation (from catecholamine or calcium influx) in early reperfusion. This might stimulate the heart into hyperactivity for a few minutes, and the effect would be superimposed onto an otherwise continuous slow recovery from ischemic injury. Our results do not support this hypothesis in that early functional recovery did not correlate linearly with ischemic duration. We have measured catecholamine stimulation during early reperfusion in separate experiments (C. Lawson, unpublished data) in which we measured epinephrine and norepinephrine levels during extended periods of extracorporeal circulation and demonstrated constant plasma levels as long as adequate anesthesia was maintained. We used constant flow (see “Experimental Design”) during ischemia and reperfusion to avoid possible neurohormonal changes and to eliminate hyperemia during early reperfusion.
The possible role of intracellular calcium in transient stimulation is discussed further.
When Does Reperfusion-Induced Injury Occur?
Results from the present study suggest that any injurious effect was initiated very soon after reperfusion, possibly during the first 1 or 2 minutes of reperfusion. This would account for the reversal of the very early recovery of function seen in most hearts. Our results, showing a progressive decline of LVDP over a 20-minute period, suggest that the injurious factor continued to operate, albeit at a declining level, over the ensuing 5 to 10 minutes of reperfusion. By 20 minutes of reperfusion, steady-state contractile conditions were present in all study groups, indicative of the near-disappearance of any short-term injurious effect.
What Is the Underlying Mechanism?
As we discussed, the study of Bolli et al9 suggested that if interventions were to be protective against stunning, they must be present during the first few minutes of reperfusion; if they were administered late in the reperfusion process, they would be ineffectual. Our study,10 taken together with subsequent studies by Bolli and colleagues11 12 and others,13 14 have shown conclusively in a number of species, both in vivo and in vitro, that during reperfusion there is a burst of radical production that peaks during the first 10 minutes of reperfusion and then slowly subsides. This time course is remarkably similar to that observed for the early but transient recovery and subsequent decline of LVDP in the present study (Figs 4⇑ and 8⇑). Our results therefore lend further support for an association between oxidant stress and the impairment of contractile function during reperfusion.
There is an additional facet of our results that might lend even more support to the above hypothesis. Analysis of the relation between early postischemic functional recovery and the duration of preceding ischemia demonstrated a clear and unexpected bell-shaped relation. Furthermore, in the groups that recovered significant late function (10, 20, 30, 40, and 50 minutes of ischemia without cardioplegia), the extent of the subsequent failure was greater in hearts with intermediate periods of ischemia (20, 30, and 40 minutes) than in hearts subjected to 10 or 50 minutes of ischemia. This observation is entirely consistent with a hypothesis advanced by one of the authors3 that the initial effect of oxidant stress should be to promote intracellular calcium overload (a hypothesis that gains some support from the literature15 ), which should have a positive inotropic effect, possibly even to the extent of leading to contractile activity greater than preischemic control values (exactly as observed in the present study). However, the cell possesses powerful mechanisms (eg, the Na+–Ca2+ exchange) for the normalization of intracellular calcium levels that should lead to early elimination of calcium overload and consequent reduction in any inotropic stimulus. Certainly, there are no published results that suggest that levels of intracellular calcium remain elevated in the stunned myocardium.
Another published experimental observation lends more support to the above hypothesis and allows for an explanation of the bell-shaped relations between duration of ischemia and early postischemic recovery. Zweier et al16 reported that the relation between duration of ischemia and severity of free radical production was also bell shaped. Thus, in the rat heart, they showed that peak radical production occurred after 30 minutes of ischemia; with less than 20 minutes or more than 60 minutes of ischemia, there was little radical production on reperfusion. This would certainly be consistent with our observation of a peak early recovery and major decline in hearts subjected to durations of ischemia of between 20 and 50 minutes.
Inherent in the above hypothesis is the (untested) prediction that antioxidants given in the present model at the time of reperfusion would tend to eliminate both the oxidant stress component of the complex and the interacting series of events occurring during early reperfusion so as to eliminate the bell-shaped relation and restore a linear inverse relation between early postischemic recovery and the duration of preceding ischemia.
Extrapolation from the present study is clearly limited by its observation in the rat heart. However, we believe that the results provide support for the concept that reperfusion, in addition to ischemia, can exert unfavorable effects on cardiac contractile function and rhythm. Our results suggest that regardless of the mechanism, these deleterious events most probably occur during the very early moments of reperfusion. Furthermore, our results are fully consistent with the hypothesis that oxidant stress contributes significantly to this form of reperfusion injury. Finally, our finding (often observed but rarely reported) that the profile for functional recovery is often biphasic suggests that several mechanisms may need to be targeted if postischemic contractile function is to be optimized.
This work was supported in part by grants from STRUTH and the trustees of the Joint Research Board, St Bartholomew’s Hospital, London, UK.
Reprint requests to A. Manché, MPhil, FRCS(CTh), Cardiovascular Research, The Rayne Institute, St Thomas’ Hospital, London SE1 7EH, UK.
- Received October 20, 1994.
- Revision received January 18, 1995.
- Accepted January 22, 1995.
- Copyright © 1995 by American Heart Association
Bolli R. Mechanism of myocardial ‘stunning.’ Circulation. 1990;82:723-738.
Hearse DJ, Bolli R. Reperfusion induced injury: manifestations, mechanisms, and clinical relevance. Cardiovasc Res. 1992;26:101-108.
Qiu Y, Manché A, Hearse DJ. Contractile and vascular consequences of blood versus crystalloid cardioplegia in the isolated blood-perfused rat heart. Eur J Cardiothorac Surg. 1993;7:137-145.
Gamble WJ, Conn PA, Edalji-Kumar, Plenge R, Monroe RG. Myocardial oxygen consumption of blood-perfused, isolated, supported rat heart. Am J Physiol. 1970;219:604-612.
Walker MJA, Curtis MJ, Hearse DJ, Campbell RWF, Janse MJ, Yellon DM, Cobbe SM, Coker SJ, Harness JB, Harron DWG, Higgins AJ, Julian DG, Lab MJ, Manning AS, Northover BJ, Parratt JR, Riemersma RA, Riva E, Russell DC, Sheridan DJ, Winslow E, Woodward B. The Lambeth Conventions: guidelines for the study of arrhythmias in ischaemia, infarction, and reperfusion. Cardiovasc Res. 1988;22:447-455.
Bolli R, Jeroudi MO, Patel BS, Arouma OI, Halliwell B, Lai EK, McCay PB. Marked reduction of free radical generation and contractile dysfunction by antioxidant therapy begun at the time of reperfusion. Circ Res. 1989;65:607-622.
Garlick PB, Davies MJ, Hearse DJ, Slater TF. Direct detection of free radicals in the reperfused rat heart using electron spin resonance spectroscopy. Circ Res. 1987;61:757-760.
Bolli R, Patel BS, Jeroudi MO, Lai EK, McCay PB. Demonstration of free radical generation in stunned myocardium of intact dogs with the use of the spin trap α-phenyl N-tert-butyl nitrone. J Clin Invest. 1988;82:476-485.
Blasig IE, Ebert B, Hennig C, Pali T, Tosaki A. Inverse relationship between ESR spin trapping of oxyradicals and degree of functional recovery during myocardial reperfusion in isolated working rat heart. Cardiovasc Res. 1990;24:263-270.
Zweier JL, Flaherty JT, Weisfeldt ML. Direct measurement of free radical generation following reperfusion of ischemic myocardium. Proc Natl Acad Sci U S A. 1987;84:1404-1407.
Marban E, Koretsune Y, Corretti M, Chacko VP, Kusoka H. Calcium and its role in myocardial cell injury during ischemia and reperfusion. Circulation. 1989;80(suppl IV):IV-17-IV-22.
Zweier JL, Flaherty JT, Weisfeldt ML. Determination of time-window of ischemic duration which results in myocardial free radical generation upon reperfusion. Circulation. 1988;78(suppl II):II-373. Abstract.