Stephen W. Ely

Blood Flow Regulation by Adenosine in Heart, Brain, and Skeletal Muscle

Although it has long been recognized that adenosine is a potent vasodilator, only during the past 20 years has serious consideration been given to its possible role in the local regulation of blood flow. Prior to that time other potential mediators of regional blood flow regulation, such as reduced PO2, increased PCO2, or decreased pH were often suggested, but definitive experimental proof for a physiological role for these agents was lacking. In more recent years, more evidence has appeared in support of a reduced pH, and additional substances such as potassium ions, inorganic phosphate, and increased osmolarity have been suggested with varying degrees of acceptance. In all likelihood, several factors, including a myogenic response, are involved in local blood flow regulation, but adenosine appears to be the primary factor. Since most of the studies on adenosine have been focused on its mediation of the control of coronary, cerebral, and skeletal muscle blood flow, in that order, discussion will be limited to these three tissues. This in no way means that adenosine does not act similarly in several other tissues. In the kidney, however, adenosine induces vasoconstriction, a subject considered in detail by Dr. Osswald in Chapter 25.

Measurements of coronary plasma and pericardial infusate adenosine concentrations during exercise in conscious dog: Relationship to myocardial oxygen consumption and coronary blood flow

A conscious dog model was employed to evaluate the relationship among myocardial oxygen consumption (MVO2), myocardial adenosine release and coronary blood flow (CNF) during graded treadmill exercise. Two methods were utilized simultaneously as indexes of changes in interstitial adenosine concentrations, (1) a pericardial infusate technique and (2) to measurement of arterial-coronary sinus plasma adenosine concentration differences. Graded exercise was associated with graded increases in MVO2, CBF, pericardial infusate adenosine concentration (PI Ado) and adenosine washout in coronary plasma. Regression analysis demonstrated significant linear relationships for MVO2 v. CBF (r = 0.78, P less than 0.01), MVO2 v. PI Ado (r = 0.52, P less than 0.01), and PI Ado v. CBF (r = 0.76, P less than 0.01). Coronary plasma adenosine arterio-sinus differences, sinus plasma concentrations, and net washout of adenosine also increased with graded exercise, however, a significant inter-animal variance was noted. These data suggest that the plasma adenosine assay is capable of detecting directional (qualitative) changes associated with changes in cardiac metabolism, however, it may not be reliable as a quantitative indicator of interstitial adenosine concentrations due to multiple factors which may influence the plasma adenosine pool. The pericardial infusate technique, which presumably represents a model of diffusion, is relatively invariant by comparison. The results demonstrate a significant correlation among MVO2, PI Ado and CBF, and thereby provide support to the hypothesis that adenosine is a major factor in the coupling of myocardial oxygen demand to oxygen supply under physiological conditions.

Inhibition of adenosine metabolism increases myocardial interstitial adenosine concentrations and coronary flow.

We employed an isolated guinea-pig heart model perfused at constant pressure (70 cmH2O) to test the hypothesis that inhibition of adenosine metabolism increases interstitial adenosine concentrations (as measured with epicardial discs) and coronary flow. Iodotubercidin (ITU, 1 microM) and EHNA (erythro-9-[2-hydroxy-3-nonyl] adenine, 5 microM) were used to inhibit adenosine kinase and deaminase, respectively during control conditions and during metabolic stimulation with 1 microM isoproterenol. The adenosine receptor blocker 8-phenyltheophylline (8-PT) was used during control conditions to assess whether the response seen was adenosine specific. ITU plus EHNA decreased heart rate (202 +/- 10 to 136 +/- 11 beats/min) and increased coronary flow (8.2 +/- 0.3 to 12.4 +/- 0.9 ml/min/g) without a change in MVO2, developed pressure or dP/dt. ITU plus EHNA increased adenosine concentrations in epicardial fluid (0.24 +/- 0.07 microM to 1.02 +/- 0.09 microM) and venous effluent (40 +/- 3 nM to 262 +/- 32 nM) during control conditions, and adenosine release increased from 389 +/- 96 pmols/min/g to 3480 +/- 365 pmols/min/g. 8-PT infusion reversed the effects on heart rate and coronary flow and resulted in a persistent elevation of epicardial fluid adenosine concentrations. During metabolic stimulation with 1 microM isoproterenol, ITU plus EHNA significantly limited the increase in heart rate and ventricular developed pressure and dP/dt while coronary flow increased to a significantly greater extent. Myocardial oxygen consumption was similar during metabolic stimulation between the two groups (vehicle vs. ITU plus EHNA). Epicardial fluid adenosine concentration in the vehicle-treated group increased from 0.17 +/- 0.3 microM to 0.34 +/- 0.02 microM at 15 min of isoproterenol stimulation whereas it increased from 1.10 +/- 0.02 microM to 2.90 +/- 0.46 microM in the ITU plus EHNA-treated group. Inhibition of adenosine metabolism during metabolic stimulation significantly increased venous adenosine concentrations and adenosine release and reduced inosine and hypoxanthine release proportionately. The release of adenosine+inosine+hypoxanthine was unchanged. Inhibition of adenosine metabolism provides evidence supporting the hypothesis that adenosine plays a role in regulating coronary vascular resistance as well as influencing heart rate and ventricular inotropy.

Hormonal role of adenosine in maintaining patency of the ductus arteriosus in fetal lambs.

The hypothesis that endogenously released adenosine plays an important role in maintaining patency of the fetal lamb ductus arteriosus was tested. The design of the study was (1) to determine the effect, if any, of exogenous adenosine on blood flow through the ductus arteriosus and (2) to evaluate the relationship among the partial pressure of oxygen in arterial blood, circulating endogenous plasma adenosine concentration, and the rate of blood flow through the ductus. When exogenous adenosine (5 mumoles) was administered during oxygen-induced ductal constriction, ductal blood flow increased from 101 +/- 6 ml/min to 153 +/- 4 ml/min (p less than 0.01). When fetal blood adenosine concentrations were measured during nonventilation and ventilation with 100% oxygen, endogenous adenosine concentrations fell to less than one-half of the preventilation levels, i.e., from 1.12 +/- 0.17 to 0.49 +/- 0.03 microM (p less than 0.01). Finally, when fetal lambs were ventilated with increasing concentrations of oxygen (0%, 10%, 20%, 60%, and 100%) and measurements obtained simultaneously at each level, there was a significant monoexponential relationship among the rise in PO2, the fall in plasma adenosine concentration, and the decrease in ductal blood flow. These data suggest that: (1) adenosine is a potent vasodilator of the lamb ductus arteriosus during oxygen-induced vasoconstriction; (2) fetal endogenous plasma adenosine levels fall significantly when PO2 is increased; and (3) the fall in adenosine concentrations parallels a decrease in ductal blood flow. The findings suggest that the endogenous vasodilator adenosine plays an important role in maintaining ductal patency in utero.

Time-dependent effects of theophylline on myocardial reactive hyperaemias in the anaesthetized dog.

1 The effects of a loading dose of theophylline (5 mg kg−1 i.v.) on the hyperaemias resulting from short‐term (15 and 30 s) interruptions in coronary blood flow and intracoronary adenosine were studied at given intervals over a 2 h period in the anaesthetized dog. 2 These hyperaemic responses were affected differently by theophylline and each effect was time‐dependent. The reactive hyperaemic response progressively decreased after drug delivery, reaching 46% of control at 2 h. In contrast, after a maximal attenuation to 23% of control 5 min after theophylline, the hyperaemia resulting from intracoronary adenosine progressively increased over the same period, reaching 64% of control 2 h after the loading dose. 3 Two‐compartment model results based on plasma theophylline measurements and the time course of theophylline accumulation in pericardial infusates, suggested that complete drug distribution throughout the heart may require at least 20 min following a single intravenous dose. 4 If it is assumed that theophylline blocks coronary vascular adenosine receptors, these pharmacokinetics are consistent with the time‐dependent pattern of response attenuation we observed for the adenosine‐induced hyperaemias, but they cannot entirely explain the pattern of response attenuation observed for the occlusion‐induced hyperaemias. The continued increase in attenuation of this response after complete drug distribution suggests an additional pharmacodynamic action of theophylline. 5 We conclude that a single therapeutic dose of theophylline results in distinct time‐dependent pharmacological effects with respect to the ability of the coronary vasculature to dilate in response to temporary interruptions in oxygen supply and in response to exogenously administered adenosine. These effects deserve consideration in both experimental studies in which adenosine antagonists are used to assess adenosine action in vivo, and in clinical practice where theophylline pharmacotherapy for pulmonary disorders is commonplace.

A Critique on the Use of Adenosine Deaminase to Test the Adenosine Hypothesis: Disregarded Implicit Assumptions

The idea that intracoronary injection of adenosine deaminase should destroy interstitial fluid adenosine and abolish regulation of coronary blood flow has sparked a number of controversial studies. In some studies adenosine deaminase infusion failed to block vascular responses attributed to endogenous adenosine, but in others it depressed them. Such results are to be expected because the amount of adenosine deaminase calculated from direct measurements of enzyme kinetics (in the adenosine concentration range of 10−8-1(10−3 M) necessary to destroy adenosine in the myocardial interstitial fluid is considerably greater than that used by investigators who have used this approach.

Response of lamb ductus arteriosus to nitroglycerin and nitroprusside.

In certain forms of congenital heart disease, patency of the ductus arteriosus is critical for survival. Since the administration of prostaglandin is associated with adverse side effects, this study was undertaken to evaluate the effects of nitroglycerin and nitroprusside on ductal blood flow during oxygen-induced ductal closure. Fifteen near-term fetal lambs were instrumented acutely. Ductal blood flow and pre- and post-ductal pressures were monitored continuously. After obtaining control data, intravenous bolus injections of nitroglycerin (250 micrograms), nitroprusside (250 micrograms), or prostaglandin E1 (5 micrograms) were administered during ventilation with either 100% nitrogen or 100% oxygen. All three agents significantly increased ductal blood flow during nitrogen ventilation (PO2 = 15 +/- 1 mm Hg). When the lambs were ventilated with 100% oxygen, the arterial PO2 increased to 107 +/- 14 mm Hg, and this was associated with a marked decrease in ductal blood flow from 275 +/- 44 to 83 +/- 11 ml/min (P less than 0.05). When nitroglycerin was administered during oxygen-stimulated ductal closure, ductal blood flow increased 184%, from 79 +/- 18 to 225 +/- 18 ml/min (P less than 0.05); nitroprusside increased ductal blood flow 126%, from 86 +/- 20 to 195 +/- 25 ml/min (P less than 0.05); prostaglandin E1 increased ductal blood flow 110%, from 84 +/- 18 to 178 +/- 17 ml/min (P less than 0.05). These data demonstrate that both nitroglycerin and nitroprusside are potent vasodilators of the ductus arteriosus and, like prostaglandin E1, can markedly attenuate the oxygen-induced ductal vasoconstriction. These results imply that nitroglycerin and nitroprusside may be useful clinically in maintaining ductal patency.

Changes in work rate to oxygen consumption ratio during hypoxia and ischemia in immature and mature rabbit hearts

This study was designed to evaluate the relative response of myocardial efficiency to reduced oxygen supply (hypoxia and ischemia) in immature and mature isolated rabbit hearts. Hearts were subjected to either 15 min of hypoxia (60% or 30% O2) or reductions in coronary flow to 75%, 50%, 25%, and 15% of basal flow followed by 12 min of total global ischemia and 15 min of reperfusion. In order to examine changes in cardiac efficiency, we utilized the ratio of isovolumic contractile function (rate-pressure product) to myocardial oxygen consumption (RPP/MVO2). Under basal conditions, immature hearts displayed lower aortic pressure. RPP, coronary resistance and RPP/MVO2. Moderate hypoxia (60% O2) resulted in similar reductions in RPP and MVO2 in both age groups, with RPP/MVO2 remaining unchanged. During severe hypoxia, RPP/MVO2 increased significantly in mature hearts but not in immature hearts (P < 0.05). Underperfusion produced greater reductions in RPP and heart rate, whereas reperfusion after ischemia resulted in greater recovery of RPP, dP/dt and MVO2 in immature compared to mature hearts. When oxygen supply was limited by reductions in coronary perfusion. RPP/MVO2 tended to increase in mature hearts, whereas the ratio declined significantly in immature hearts. These data demonstrate that, in this model, a reduction in oxygen supply by hypoxia or hypoperfusion decreases efficiency in immature hearts, but increases efficiency in mature hearts under the same conditions.

Myocardial reactive hyperemia in the newborn

Adenosine may be an important metabolic regulator of coronary blood flow during active hyperemia in the newborn. In this study, the adenosine uptake blocker dipyridamole (DPY) and receptor antagonist theophylline (THEO) were used to assess the role of adenosine in the reactive hyperemic response of the neonatal heart. Eighteen anesthetized, open-chest lambs were instrumented with aortic and coronary sinus catheters as well as an extracorporeal shunt to the circumflex coronary artery incorporating a 2.0-mm electromagnetic flow transducer. Ten-second occlusions of the circumflex coronary artery catheter were performed, and the resulting reactive hyperemia was used to determine peak hyperemic blood flow (PHF), duration of hyperemia, and the blood flow repayment. These values were determined prior to treatment and 30 min following administration of a saline vehicle, DPY (0.2 mg/kg) or THEO (5 mg/kg). DPY resulted in increases in PHF from 220 +/- 12 to 247 +/- 14 ml/min/100 g heart tissue (P less than 0.05), durations of hyperemia from 29 +/- 2 to 38 +/- 2 sec (P less than 0.01), and blood flow repayments from 65 +/- 2 to 102 +/- 4 ml/100 g (P less than 0.001). THEO resulted in decreases in PHF from 224 +/- 15 to 198 +/- 12 ml/min/100 g (P less than 0.05), durations of hyperemia from 28 +/- 2 to 22 +/- 2 sec (P less than 0.05), and blood flow repayments from 64 +/- 2 to 46 +/- 2 ml/100 g (P less than 0.01). The data indicate that DPY enhances reactive hyperemia while antagonism of adenosine with THEO attenuates it. These results suggest that adenosine plays a role in mediating reactive hyperemia in the newborn heart.