| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1995;109:466-472
© 1995 Mosby, Inc.
CARDIOPULMONARY BYPASS, |
Detroit, Mich.
Presented in part at the Forty-ninth Annual Sessions of the Forum on Fundamental Surgical Problems, 1993 Clinical Congress, San Francisco, Calif., October 10-15.
Received for publication March 10, 1994. Accepted for publication July 22, 1994. Address for reprints: Norman A. Silverman, MD, Division Head, Cardiac and Thoracic Surgery, Henry Ford Hospital, 2799 W. Grand Blvd., Detroit, MI 48202.
Abstract
Mechanical function and coronary hemodynamics were assessed in 73 isolated rabbit hearts randomly subjected to 0, 10, 20, 30, or 45 minutes of 37 degrees C global ischemia and 45 minutes of reperfusion in either a modified Krebs buffer or homologous blood-perfused Langendorff mode (n = 7 to 9 hearts per group). Isovolumic developed pressure, resting coronary flow, and response to endothelium-dependent (bradykinin) and -independent (nitroglycerin) agonists were quantitated at defined preload and heart rate. Perfusate did not influence systolic performance, which was impaired after 30 minutes of ischemia and fell to 64% to 72% of preischemic values after 45 minutes of ischemia (p < 0.05). However, basal coronary flow was at least sixfold greater in crystalloid-perfused hearts. Moreover, coronary hyperemia (p < 0.05) persisted for Krebs-perfused hearts subjected to all but the longest ischemic interval. After equilibration, all postischemic blood-perfused hearts had basal flow unchanged from before ischemia. Bradykinin and nitroglycerin induced similar increases in coronary flow for each group before and after each ischemia interval. However, the magnitude of this increase was greater in blood-perfused hearts (p < 0.01) and was not attenuated by the ischemic times encompassed in this protocol. In contrast, endothelium-dependent and -independent coronary flow reserve was abolished after 20 minutes of ischemia or longer in Krebs-perfused hearts. These data suggest that the unphysiologic resting flow patterns of crystalloid-perfused isolated hearts obfuscate interpretation of the interaction between coronary flow reserve and ischemic injury. (J THORAC CARDIOVASC SURG 1995; 109: 466-72)
The isolated, perfused, small animal heart has been a standard experimental model for quantitating the consequences of ischemia and reperfusion. Although the focus of previous investigations has historically been on myocyte viability as assessed by postischemic ventricular performance and cardiac metabolism, the coronary arterial smooth muscle and the endothelium with its paracrine function are now recognized as separate and important considerations in formulating strategies to prevent ischemia-reperfusion injury. 
1-2 Crystalloid-perfused heart preparations have the advantages of lower cost, precision of formulation, simplicity, and avoidance of thrombus formation. However, clinical relevance will always be questionable because the roles of erythrocytes, platelets, and leukocytes in oxidative stress and reperfusion injury are excluded. Also, the lack of plasma proteins results in osmotic and oncotic inadequacies and the baseline high oxygen tension and coronary flow rate are unphysiologic. 3
The present study was designed to quantitate and compare perturbations in basal coronary flow and coronary hemodynamics stimulated by endothelium-dependent and endothelium-independent agonists after progressively severe, reversible ischemic insults in crystalloid- and blood-perfused isolated hearts. These observations suggest that the unphysiologic resting flow patterns attendant to asanguineous perfusates obfuscate interpretation of coronary flow reserve after ischemic injury.
MATERIALS AND METHODS
Langendorff perfusion.
The details of the blood-perfused ex vivo preparation have been previously described. 4 All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH Publication No. 80-23, revised 1978). Adult New Zealand White rabbits (2.5 to 3.8 kg) were anesthetized, heparinized, a tracheostomy performed, and the lungs ventilated with room air. The abdomen of the blood-donor rabbit was opened and the abdominal aorta was cannulated with an 18-gauge ball-end needle for blood collection. The thorax and abdomen of the heart-donor rabbit were opened. Both the abdominal aorta and ascending aorta were cannulated in situ and retrograde perfusion was initiated before the heart was harvested as blood was withdrawn and added to the venous reservoir. The pulmonary artery was cannulated for collection and measurement of coronary effluent. The heart was then harvested and mounted on the siliconized glass perfusion column. Normothermic retrograde perfusion was initiated at 50 mm Hg for 5 minutes and then increased to 80 mm Hg by varying the speed of the roller pump. The blood perfusate was oxygenated by bubbling the reservoir beaker with a mixture of 95% oxygen and 5% carbon dioxide to maintain oxygen tension between 120 and 200 torr and carbon dioxide tension between 35 and 45 torr. The pH was corrected to 7.35 to 7.45 with sodium bicarbonate. Crystalloid-perfused hearts were similarly cannulated and perfused in situ before being mounted to minimize ischemia. Hearts did not stop beating in the interval between harvesting and mounting. The modified Krebs buffer solution was bubbled with a 95% oxygen and 5% carbon dioxide mixture. The blood perfusion was a recirculating system whereas the Krebs perfusate was not recirculated.
All hearts were immersed in a water-jacketed organ chamber to maintain a temperature of 37 degrees C. During stabilization after retrograde perfusion was established, the venae cavae were ligated, the sinus node crushed, the lungs removed, and all hearts were instrumented in the following manner. The left atrium was opened and a size 10 intraventricular latex balloon (Hugo Sachs Elektronik, March-Hugstetten, Germany) secured below the mitral anulus. A 1 mm diameter plastic catheter effected left ventricular venting. A saline filled Tygon microbore tube (Norton Performance Plastics, Wayne, N.J.) connected the balloon to a P23XL transducer (Viggo-Spectramed, Inc., Sacramento, Calif.) for measurement of left ventricular pressure, developed pressure, and the rate of positive and negative pressure development (±dp/dt). Pressure tracings were recorded on an eight-channel recorder (RS 3800, Gould, Inc., Cleveland, Ohio). Ventricular electrodes were attached and the hearts paced at 180 beats/min. For the purpose of drug testing, the bolus infusion was administered through a sidearm line to the main perfusion line.
Experimental protocol.
Hearts were perfused with either oxygenated Krebs buffer or homologous blood and randomly subjected to 10, 20, 30, or 45 minutes of 37° C global ischemia effected by stopping the perfusion pump. Control hearts were studied without antcedent ischemia. After a 45-minute equilibrium time with retrograde perfusion, baseline isovolumic systolic function and coronary flow were quantitated at the fixed paced rate and an end-diastolic pressure of 10 mm Hg. Endothelium-dependent coronary flow reserve was quantitated after bolus infusion of 1 ml 10-6 molar bradykinin. A 12-minute washout period allowed resting coronary flow to return to baseline before endothelium-independent coronary reserve was quantitated by bolus infusion of 500 µg nitroglycerin. Hearts were then removed from the perfusion column, excess fluid gently removed, and the hearts were placed in a drying oven for 48 hours. At that time, hearts were removed and weighed. Data were obtained from seven to nine hearts for each perfusate as controls and for each ischemic interval.
Systolic function.
The intraventricular balloon was filled with a volume of saline to produce an end-diastolic pressure of 10 mm Hg. Isovolumic developed pressure at this preload (millimeters of mercury) was calculated as the difference between peak left ventricular systolic pressure and this diastolic pressure.
Coronary flow.
Three-minute timed collection of coronary venous effluent was made at each sampling interval. Results were expressed as milliliters per minute per gram dry weight of the heart.
Statistical methods.
Data are presented as mean ± standard error of the mean. Paired observations were compared by Student's t test. Repeated observations were compared by analysis of variance with Tukey's method used to adjust for multiple tests. Differences were considered significant if p was 
0.05.
RESULTS
Systolic function (Table I).
Krebs-perfused hearts had no decrement in developed pressure after the 10- and 20-minute ischemic intervals. However, the 72% and 64% returns of preischemic developed pressure after 30 and 45 minutes of ischemia were depressed (p < 0.05 for each) compared with control values. In blood-perfused hearts, there was an incremental decrease in developed pressure after 20 minutes and longer ischemic intervals that reached significance after 30 and 45 minutes of ischemia with reperfusion (74% and 72%, respectively, p < 0.05 versus control for each). A similar pattern of maximum +dp/dt was noted, declining to 63% and 72% of control in Krebs- and blood-perfused hearts, respectively, after 45 minutes of ischemia (p < 0.05, each group). No difference in the two parameters of systolic function was noted in comparing Krebs- and blood-perfused hearts before and after each ischemic interval. Because heart rate was fixed by pacing, the mechanical determinants of coronary flow were comparable in the two groups throughout the experimental protocol.
|
Table II, Fig. 1).
|
|
Table II, Figs 2 to 4).
|
|
|
DISCUSSION
Conceptually, we have performed an experiment in an isolated small animal heart preparation to quantitate derangements in resting coronary flow and stimulated endothelium-dependent and endothelium-independent coronary flow reserve to incrementally more severe ischemic insults. We compared the relative susceptibility to unprotected, global normothermic ischemia of these coronary flow patterns with isovolumic systolic performance. Given an identical experimental protocol, implications from the data generated depend on whether the heart was perfused with homologous blood or an asanguineous solution. In a blood-perfused heart, systolic performance was compromised, but the response to bradykinin and nitroglycerin was unaffected even after the longest ischemic interval, suggesting resistance of the endothelium and coronary arterial smooth muscle to ischemia-reperfusion injury. In contrast, in Krebs-perfused hearts, the response to these agonists was blunted before a decrement in systolic function was noted. Does this imply that the coronary vasculature including the endothelium is more susceptible than the myocyte to deprivation of flow? Alternatively, does the high shear stress of the unphysiologic resting coronary flow in the crystalloid-perfused heart result in almost maximal relaxation of arterial smooth muscle mediated by endothelium-derived relaxing factor (EDRF), thus precluding an incremental effect when an exogenous agonist is infused?
When conflicting interpretations exist, confirmation of alternative explanations is sought in the literature. Unfortunately, the various methods for quantitating coronary artery endothelial cell and smooth muscle dysfunction after ischemia may further obfuscate rather than clarify the issue. 5-8 For instance, "endothelial function" can be tested in vivo by infusion of agonists to stimulate endothelial muscarinic, serotonergic, purinergic, thrombin, or vasopressinergic receptor-mediated EDRF release. EDRF causes relaxation of the underlying smooth muscle by a mechanism that involves cyclic guanosine monophosphate-dependent protein phosphorylation. What is usually measured is a change in coronary flow or resistance resulting from this vasodilation. Differences in dosage, agonist, and species make strict comparison difficult. Alternatively, ex vivo methods use harvested rings of epicardial vessels. Comparative arterial vascular responsiveness, that is, force of contraction, is quantitated by adding agonists to the bath with the endothelium intact or denuded and/or in the resting state or preconstricted. The lack of "gold standards" applicable to all laboratories is evident. Moreover, these ex vivo tests of vascular responsiveness are performed in epicardial conductance vessels, whereas the effects of an in vivo infusion of vasoactive agent are more likely at the coronary arteriolar level.
Several studies, mostly involving isolated canine epicardial coronary arterial rings, suggest that brief ischemia results in vascular endothelial dysfunction. Gross and colleagues 9 showed attenuated endothelium-dependent responses after repetitive 5-minute occlusions but unaffected response to endothelium-independent agonists. After 15 minutes of ischemia, Headrick, Angello, and Berne 10 showed injury to the large coronary conduit vessels and Dauber and coworkers 11 also showed concomitant microvascular damage. The time course of recovery of this "stunned" endothelial function has been assessed by Kim and colleagues. 12 After 15 minutes of ischemia, the impaired vasodilatory response to acetylcholine required 90 minutes to return to baseline. The response to nitroprusside was unaffected by their protocol. In vitro endothelium-dependent vasodilatory responses were blunted after 10 minutes of ischemia, abolished after 20 minutes, and required 120 minutes of reperfusion to recover. In contrast, the in vivo study of McFalls and associates 13 found no diminution in endothelium-dependent coronary flow reserve of swine after 10 minutes of ischemia.
More easily comparable to the present study is the work of Hashimoto and coworkers 14 In crystalloid-perfused isolated rabbit hearts subjected to 30 minutes of ischemia and 45 minutes of reperfusion, serotonin increased coronary flow 30.5% before ischemia and only 13.4% after ischemia, whereas the smooth muscle response to adenosine was unaffected. Although postischemic systolic dysfunction was comparable to our 30-minute ischemic groups, the difference in endothelium-dependent and -independent flow reserve could be attributable to different agonists and lower basal flow rates. Similarly, Tsao and Lefer 15 subjected isolated, crystalloid-perfused rat hearts to 30 minutes of global ischemia and found impaired relaxation to an alternative endothelium-dependent vasodilator, acetylcholine, after 20 minutes of reperfusion. No change in responsiveness to nitroglycerin was noted. However, unlike the present study, flow patterns to these agonists were quantitated after preconstriction of the coronary vasculature by U4619.
A detailed review of endothelial and arterial smooth muscle function after 1 hour of ischemia or more is not applicable to the presented data. However, conflicting results are reported that parallel the information available for "reversible ischemic" insults, that is, divergence of endothelium-dependent and -independent responses and actual existence of an injury. Representative of these controversies are the observations by Pearson, Schaff, and Vanhoutte, 16,17 showing persistence of altered serotonergic and purinergic receptor activity 1 month after a 1-hour ischemic interval. Dignan and associates 18 found no vascular injury after 2 hours of ischemia, and Viehman and coworkers 19 had to extend ischemia to 3 hours to demonstrate impaired vasoreactivity.
This study has several limitations. Admittedly, this is a descriptive and not a mechanistic protocol and therefore the questions posed in the first paragraph of the discussion are not answered. The relative contributions of altered viscosity, oxygen tension, formed blood cell elements, and wall shear stress were not analyzed. The crystalloid-perfused studies were purposely performed in a "standard" manner to compare with a more physiologic blood perfusate, and no attempt was made to control for the aforementioned variables. Second, the concentrations of agonist actually presented to the endothelial receptor may have been lower in the crystalloid-perfused hearts because of the increased resting coronary flow. However, preliminary dose-response curves (data not shown) in both crystalloid and blood perfusates showed maximal response to the concentration of agonist infused by bolus injection. Therefore, we support the quantitative validity of these observations and propose that despite the simplicity of a crystalloid-perfused heart model, it is inappropriate for assessment of the interaction between coronary flow reserve and ischemic injury.
Footnotes
From the Divisions of Cardiac and Thoracic Surgerya and Hypertension Research,b Henry Ford Hospital, Detroit, Mich. 
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
