The Effects of Cardiac Sympathetic Nerve Stimulation on Perfusion of Stenotic Coronary Arteries in the Dog

Transcription

1 The Effects of Cardiac Sympathetic Nerve Stimulation on Perfusion of Stenotic Coronary Arteries in the Dog Gerd Heusch and Andreas Deussen From the Physiologisches Institul I der Univcrsitit Dusseldorf, Dusseldorf, West Germany SUMMARY. The interaction of sympathetic vasoconstriction with the coronary reserve distal to stenoses and the role of a-adrenergic mechanisms in the genesis of myocardial ischemia were studied in 42 anesthetized open-chest dogs. Left cardiac sympathetic nerve stimulation was performed after bilateral cervical vagotomy: with intact coronary arteries, with an intermediate stenosis, and with a severe stenosis on the circumflex coronary artery. Stenoses were produced by a wire snare and defined as intermediate by the reduction of the reactive hyperemia repayment following a 15-second occlusion from 460 ± 100 to 140 ± 30%; a severe stenosis was defined by only 25 ± 8% reactive hyperemia repayment. Cardiac sympathetic nerve stimulation decreased the end-diastolic coronary resistance of intact coronary arteries from 0.79 ± 0.05 to 0.53 ± 0.06 mm Hg X min X 100 g/ml (P < 0.01) and the end-diastolic poststenotic resistance of the moderately stenosed arteries from 0.65 ± 0.08 to 0.50 ± 0.07 mm Hg X min x 100 g/ml (P < 0.01). Cardiac sympathetic nerve stimulation increased the end-diastolic resistance distal to severe stenoses from 0.57 ± 0.04 to 0.97 ± 0.18 mm Hg X min X 100 g/ml (P < 0.01). This stimulation resulted in net lactate production of the circumflex-perfused myocardium; four dogs died by ventricular fibrillation. There was a hyperbolic correlation of the cardiac sympathetic nerve stimulation-induced change in resistance to the degree of coronary hyperemic reserve (r = 0.81). Phentolamine (2 mg/kg, iv) and rauwolscine (0.2 mg/kg, iv) prevented the increase in resistance distal to severe stenoses during cardiac sympathetic nerve stimulation, whereas prazosin (1.2 mg/ kg, iv) was ineffective. After j3-blockade with propranolol (2 mg/kg, iv), rauwolscine still prevented the increase in poststenotic resistance during cardiac sympathetic nerve stimulation. We conclude that there is a continuous unmasking of sympathetic vasoconstriction with increasing severity of a stenosis and, thus, decreasing coronary reserve. This vasoconstriction is mediated by postjunctional o 2 -receptors and can induce myocardial ischemia distal to severe coronary stenoses. (Circ Res 53: 8-15, 1983) THE direct a-vasoconstrictive effects of cardiac sympathetic nerves on coronary vessels can compete with their effects on myocardial performance and metabolism (Berne et al., 1965; Mohrman and Feigl, 1978; Murray and Vatner, 1979). In the presence of coronary stenoses, and, thus, limited, or at pharmacologically exhausted dilatory reserve, coronary vessels remain responsive to a-constrictive influences (Buffington and Feigl, 1981; Johannsen et al., 1982). Activation of cardiac sympathetic nerves in the presence of coronary stenoses may then decrease coronary blood flow and induce myocardial ischemia (Mudge et al., 1976). The purpose of this study was to investigate the effects of cardiac sympathetic nerve stimulation at different degrees of coronary stenoses and to study the interaction of sympathetic vasoconstriction with metabolic vasodilation over the whole range of coronary reserve. In addition, we wanted to investigate whether sympathetic vasoconstriction can overcome metabolic vasodilation, even in the absence of /Sblockade, and then induce ischemia in the myocardium supplied by severely stenotic coronary arteries. Finally, we attempted to differentiate a j - and a 2 - adrenergic receptors in the mediation of sympathetic vasoconstriction distal to coronary stenoses by use of phentolamine, prazosin, and rauwolscine. Methods General Preparation Forty-two mongrel dogs weighing kg were anesthetized with a-chloralose (50 mg/kg, iv) and urethane (250 mg/kg, iv). Dogs were ventilated with room air by a respirator (74052 Braun-Melsungen) at an end expiratory pressure of 5 cm H 2 O. Arterial P02, Pcoz, and ph were controlled every 30 minutes (BMS 2 Radiometer, Copenhagen) and maintained within normal limits (Feigl and D'Alecy, 1972). A left thoracotomy was performed through the 5th intercostal space and the heart suspended in a pericardial cradle. Coronary blood flow was measured on the left circumflex coronary artery with an electromagnetic flowmeter (Statham SP2202). Left ventricular pressure and pressure in the ascending aorta were recorded with tip manometers (Millar PC 350). Left ventricular dp/ dt and heart rate were derived from the left ventricular pressure signal. One small branch of the circumflex coronary artery was cannulated with a polyethylene catheter of 1 mm outer diameter for measurement of epicardial peripheral coronary pressure (Statham P 23 ID).

2 Heusch and DeHSsen/Sympathetic Effects on Stenotic Coronary Arteries A 5F Coumand catheter was advanced from the right jugular vein via the coronary sinus into a vein draining the myocardium supplied by the left circumflex coronary artery. This catheter was used for measurement of coronary venous pressure (Statham P23BB) and for withdrawal of coronary venous blood samples. The hemodynamic data were continuously recorded (R 612 Beckman). Oxygen content was determined in samples from the left femoral artery and the cannulated circumflex vein (Lex- O 2 -Con, Lexington Instruments). In the same samples, the plasma lactate concentration was measured by enzymatic analysis (monotest Lactat, Boehringer Mannheim). Oxygen and, with correction for hematocrit, lactate consumption of the circumflex-perfused myocardium were calculated by Fick's principle. In four dogs, we measured regional contractile function of the circumflex-perfused myocardium by ultrasound technique. One pair of ultrasonic crystals (2 mm in diameter) was implanted into the subendocardial third of the circumflex-perfused myocardium perpendicular to the long axis of the left ventricle. The distance between the two crystals was about 10 mm. The motion of the ultrasonic crystals was monitored by an ultrasonic imaging circuit (Fa. R. Oswald, Erkrath). Systolic segment shortening was related to end-diastolic length and is given in percent. Experimental Protocols The left inferior cardiac nerve was cut and the distal part stimulated at 20 Hz with 5-msec pulses of 1-7 V for 90 seconds. Preliminary studies with stimulation at 5 and 10 Hz have revealed qualitatively similar, but weaker, responses. We chose 20 Hz for a stimulation in the maximal physiological range (Horeyseck et al., 1976). Hemodynamic measurements and blood samples were taken at steady states after seconds of stimulation. The stimulation was first performed after bilateral cervical vagotomy with intact coronary arteries. Then a circumferential stenosis of the left circumflex artery was produced by a wire snare just proximal or distal to the flowprobe and proximal to the cannulated circumflex branch. An intermediate stenosis was defined by the reduction of the reactive hyperemia repayment (Coffman and Gregg, 1960) following a 15-second occlusion to about 100%. Then the degree of stenosis was increased and a severe stenosis defined by the absence of almost any reactive hyperemia. Cardiac sympathetic nerve stimulation was repeated with an intermediate and with a severe circumflex coronary stenosis. Then 30 of the dogs were divided into four groups undergoing different pharmacological interventions. The severe stenosis was maintained throughout the following experimental procedure in all these dogs. Ten dogs received phentolamine-methansulfonate (2 mg/kg, iv; Ciba), seven dogs prazosin-hydrochloride (1.2 mg/kg, iv; Pfizer), and six dogs rauwolstine-hydrochloride (0.2 mg/kg, iv; Roth) before a last cardiac sympathetic nerve stimulation was performed. In an additional seven dogs, j9-adrenergic blockade was produced by propranolol-hydrochloride (2 mg/kg, iv; Rhein Pharma) before administration of rauwolscine (0.2 mg/kg, iv); the cardiac sympathetic nerve stimulation was repeated after propranolol alone and after propranolol + rauwolscine. After each experiment, the perfusion area of the left circumflex coronary artery was stained, cut out, and weighed. Coronary blood flow was related to 100 g of myocardium. End-diastolic distal coronary resistances were calculated from the gradient between peripheral coronary pressure and coronary venous pressure and the circumflex coronary blood flow. Statistics All data are presented as mean ± SEM. The reactions to cardiac sympathetic nerve stimulation were analyzed by paired Ntest. The control conditions of the four experimental subgroups before drug administration were compared by a one-way analysis of variance. In addition, a nonlinear regression between the coronary reactive hyperemia repayment and the reaction of end-diastolic distal coronary resistance to cardiac sympathetic nerve stimulation was performed. This regression analysis was based on a least squares fitting method using a BMDP 3 R program. We chose a hyperbolic expression since there is a mathematical hyperbolic relationship between resistance and flow. Results With intact coronary arteries, cardiac sympathetic nerve stimulation induced a marked increase in myocardial performance and metabolism (Table 1). Systolic segment shortening of the circumflex-perfused myocardium increased from 11.5±2.6 to 14.1 ± 3.4% (n = 4, P < 0.05). In compensation for the increased myocardial metabolic needs, coronary vascular resistance decreased (Fig. 1). With an intermediate circumflex coronary stenosis, cardiac sympathetic nerve stimulation induced a similar increase in myocardial performance, as under conditions of intact coronary arteries (Table 1). Systolic segment shortening of the circumflex-perfused myocardium increased from 11.4 ± 2.7 to 14.1 ± 4.2% (n = 4, P < 0.05). Circumflex coronary blood flow still increased, and poststenotic coronary resistance still decreased, although to a lesser extent than in intact coronary arteries (Fig. 1). ENDDIASTOLIC DISTAL CORONARY RESISTANCE NO I I IT* INTERMEDIATE T*1T r CONTROL CSNS FIGURE 1. Effects of cardiac sympathetic nerve stimulation (CSNS) on the end-diastolic resistances of intact and stenotic coronary arteries. The decrease in the resistance distal to an intermediate stenosis dunng CSNS is less marked than that of intact coronary arteries. The resistance of the vascular bed distal to a severe stenosis is increased during CSNS.

3 10 Circulation Research/Vo/. 53, No. 1, July 1983 Circumflex coronary blood flow (ml/min pet 100 g) Mean aortic pressure Mean peripheral coronary pressure Peak left ventricular pressure Left ventricular end-diastolic pressure Left ventricular dp/dt,^ (mm Hg/sec) Heart rate (beats/min) Circumflex coronary venous pressure TABLE 1 Effects of Cardiac Sympathetic Nerve Stimulation (CSNS) on Perfusion of Stenotic Coronary Arteries Myocardial oxygen consumption Oxygen content in a circumflex vein (ml/100 ml) Myocardial lactate consumption (jimol/min per 100 g) No stenosis (n = 42) (460 ± 100% RH) Intermediate stenosis (n = 42) (140 ± 30% RH) Severe stenosis (n = 38) (25 ± 8% RH) Control CSNS Control CSNS Control CSNS 78 ± ± 9* 95 ± ±8* 94 ± ± 8* 103 ±5 116 ±7* 3 5 ± ± 0.4* 2100 ± ± 300* 166 ±5 179 ±5* 7.8 ± ± ± ±1.2* 5.6 ± ± ± ±36* Data are mean ± SEM. * P < 0.01 CSNS vs. control.rh = reactive hyperemia repayment. 65 ± 3 95 ± 8 78 ± 7 8 ± 9* 99 ± 6* 73 ±5* 50 ± 3 50 ±3 95 ±8 101 ±8 * 55 ±4 55 ± ± ±5* 100 ± ± 3' 3.4 ± ±0.3' 4 6 ± ± ± ±290* 1810 ± ± 210* 170 ±4 183 ±4* 171 ± ± 3* 7.7 ± ± ± ± ± ±12* 4.8 ± ± ± ± ± ± 0.9* 59 ± ±38* 39 ± ±12* With a severe circumflex coronary stenosis, total left ventricular performance was still increased during cardiac sympathetic nerve stimulation (Table 1). Systolic segment shortening of the circumflex-perfused myocardium, however, was reduced from 8.4 ± 2.5 to 5.1 ± 0.8% («= 4, P < 0.05). Circumflex coronary blood flow decreased in spite of an increase in perfusion pressure (Fig. 2). Oxygen consumption of the circumflex-perfused myocardium did not increase anymore in spite of an increased oxygen extraction. Myocardial lactate consumption was reversed to net lactate production during sympathetic stimulation (Table 1). Poststenotic coronary resistance increased (Fig. 1). Four out of 42 dogs died by ventricular fibrillation during this sympathetic stimulation. There was a continuous shift from a decrease in coronary resistance to an increase in coronary resistance during cardiac sympathetic nerve stimulation with increasing severity of the stenosis and, thus, decreasing coronary reserve, as evidenced by the reduced reactive hyperemia response to a 15- second occlusion (Fig. 3). At a coronary reserve of less than 79%, an increase in resistance was the predominant effect of cardiac sympathetic nerve stimulation. The four experimental subgroups receiving either phentolamine, prazosin, rauwolscine, or propranolol + rauwolscine did not differ significantly in any of the measured parameters before drug administration. Phentolamine decreased the resistance of the poststenotic vascular bed by 25%, and prevented the increase in resistance during sympathetic stimulation as well as the net lactate production (Fig. 4; Table 2). Prazosin decreased aortic and left ventricular pressures to approximately the same degree as did phen-

4 Heusch and Deussen/Sympathetic Effects on Stenotic Coronary Arteries 11 CSNS h \ LEFT VENTRICULAR PRESSURE ImmHg] LEFT VENTRICULAR DIASTOLIC PRESSURE 5- ImmHg ) LEFT VENTRICULAR iooo dp/dt (mmhgai FIGURE 2. Effects of cardiac sympathetic nerve stimulation (CSNS) on perfusion of a severely stenotic coronary artery. Coronary blood flow is decreased in spite of an increase in perfusion pressure. The arrow indicates a 15-second occlusion without a following reactive hyperemia AORTIC PRESSURE ImmHg] POSTSTENOTIC CORONARY PRESSURE ImmHg I CIRCUMFLEX VENOUS PRESSURE ImmHg] CIRCUMFLEX CORONARY BLOOD FLOW 50- [ml/mm] 0_ tolamine. However, it increased the resistance of the poststenotic vascular bed by 28%. Prazosin failed to prevent the increase in poststenotic resistance during sympathetic stimulation, whereas the reactions of left ventricular performance were similar to those after phentolamine (Fig. 4; Table 2). NO A INTERMEDIATE CRITICAL 2.0 E Rauwolscine was effective to the same extent as phentolamine. It decreased the resistance of the poststenotic vascular bed by 18% and prevented the increase in resistance during cardiac sympathetic nerve stimulation as well as the net lactate production (Fig. 4; Table 2). After propranolol, the increase in poststenotic vascular resistance during cardiac sympathetic nerve stimulation was even more marked than without propranolol (Fig. 5). The additional administration of rauwolscine did not change the control resistance 1.5»1.0 u z z0.5 ENDDIASTOLIC DISTAL CORONARY RESISTANCE PHENTOLAMINE PRAZOSIN RAUWOLSCINE 10-r 08- T V, 06- < z o 04-- o BOO -r _+ I 1 REACTIVE HYPEREMIA REPAYMENT [ I ] FIGURE 3. The interaction of the coronary reactive hyperemia repayment percent after a 15-s occlusion and the response of end-diastolic distal coronary resistance to cardiac sympathetic nerve stimulation. With decreasing coronary reserve, there is a continuous unmasking of sympathetic vasoconstnction. "*(> <: 0.05 CONTROL CSNS FIGURE 4. Effects of cardiac sympathetic nerve stimulation (CSNS) on coronary resistance distal to a severe stenosis. Phentolamine and rauwolscine prevent an increase in resistance, whereas after prazosin the increase in-resistance during CSNS is still present.

5 12 Circulation Research/Vol. 53, No. 1, July 1983 TABLE 1ADL.C 2i. Effects of Cardiac Sympathetic Nerve Stimulation (CSNS) on Perfusion of Severely Stenotic Coronary Arteries after Phentolamine (2 mg/kg, iv), Prazosin (12 mg/kg, iv) or Rauwolsdne (0.2 mg/kg, iv) Circumflex coronary blood 48 ± 6 61 ± 9* flow Mean aortic pressure 64 ± 5 69 ± 4* (mmhg) Mean peripheral coronary 36 ± 8 44 ± 9f pressure Peak left ventricular pressure 76 ± 4 84 ± 6* Left ventricular end-diastolic pressure Left ventricular dp/dw (mm Hg/sec) Heart rate (beats/min) Circumflex coronary venous pressure (mmhg) Myocardial oxygen consumption Oxygen content in a circumflex vein (ml/100 ml) Myocardial lactate 15 ± 3 34 ± 12f consumption (junol/min per 100 g) Data are mean ± SEM. P < 0.01 CSNS vs. control; t P < 0.05 Phentolamine (n = 10) Prazosin (n = 7) Rauwolscine (n = 6) Control CSNS Control CSNS Control CSNS 4.2 ± ± 0.6f 1800 ± ± 280' 176 ±3 187 ±5* 7.3 ± ± ± ±2.3f 3.8 ± ±0.6 of the poststenotic vascular bed, but prevented the increase in resistance during sympathetic stimulation and a net lactate production (Fig. 5; Table 3). Discussion The major results of this study are: 1. With decreasing coronary reserve, there is a continuous shift from a decrease in coronary resistance to an increase in coronary resistance in response to cardiac sympathetic nerve stimulation. 2. The sympathetic vasoconstriction in severely stenotic coronary arteries is mediated by postjunctional a2-receptors. 3. The vasoconstrictive effects of sympathetic stimulation can overcome the metabolic dilatory mechanisms, even in the absence of /3-blockade, and induce ischemia in the myocardium supplied by severely stenotic coronary arteries. 30 ±4 56 ±4 38 ±4 73 ±5 25±5f 60 ±4' 42±4 80 ±5* 52 ± ± 5 44 ± 5 76 ± 5 28 ±14 16 ± 10f 10 ±4 64 ± 10* 60 ±4' 46 ±5 80 ±5 4.2 ± ± ± ± 0.6t 1780 ± ±240* 2100 ± ± 190* 175 ±6 189 ±9* 175 ±8 185 ± 9* 6.5 ± ± ± ± ± ± ± ±1.8t 3.6 ± ±0.7f 4.8 ± ± ±6f The Interaction of Sympathetic Vasoconstriction with Coronary Reserve Sympathetic vasoconstriction has already been demonstrated at distinct points within the range of coronary reserve. In the upper range, there is an increase in reactive hyperemia repayment after a- blockade (Schwartz and Stone, 1977), as well as an enhanced vasodilation during exercise after a-blockade (Murray and Vatner, 1979; Heyndrickx et al., 1982). With limited coronary reserve in the presence of an intermediate stenosis in terms of our definition (Buffington and Feigl, 1981), or after maximal coronary dilation by adenosine (Johannsen et al., 1982), a-sympathetic vasoconstriction is still present. We extended these studies over the whole range of coronary reserve within one experimental model and demonstrated a continuous unmasking of the constrictive effects of sympathetic stimulation with

6 Heusch and Deussen/Sympathetic Effects on Stenotic Coronary Arteries 13 ENDDIASTOLIC DISTAL CORONARY RESISTANCE frnmufl-fnin roogl L * J r p<005 1.O M PROPRANOLOL CONTROL PROPRANOLOL RAUWOLSCINE CSNS FIGURE 5. Effects of cardiac sympathetic nerve stimulation (CSNS) on coronary resistance distal to a severe stenosis. The marked increase in resistance during CSNS under fl-blockade with propranolol can be prevented by rauwolscine. decreasing coronary hyperemic reserve. Below a reactive hyperemia repayment of 79%, the decrease in resistance during sympathetic stimulation was reversed to an increase in resistance. Since the coronary circulation may form vascular waterfalls (Downey and Kirk, 1975; Munch and Downey, 1980), the coronary venous pressure may be an inaccurate estimate of outflow pressure for the calculation of coronary resistances (Bellamy, 1980). Thus, the observed increase in the resistance of the poststenotic vascular bed distal to a severe stenosis during sympathetic stimulation may be due to enhanced extravascular compression, and coronary vascular resistance may be virtually constant. However, since left ventricular end-diastolic pressure was unchanged during this sympathetic stimulation, an enhanced extravascular compression seems not to be responsible for the increase in the end-diastolic resistance of the poststenotic vascular bed. The prevention of the increase in poststenotic resistance by phentolamine and rauwolscine, which do not prevent the effects of sympathetic stimulation on myocardial performance, further excludes a significant role of extravascular compression in the observed resistance increase. TABLE 3 Effects of Cardiac Sympathetic Nerve Stimulation (CSNS) on Perfusion of Severely Stenotic Coronary Arteries after Propranolol (2 mg/kg, iv) and Propranolol (2 mg/kg, iv) + Rauwolscine (0.2 mg/kg, iv) Circumflex coronary blood flow Mean aortic pressure (mmhg) Mean peripheral coronary pressure Peak left ventricular pressure Left ventricular end-diastolic pressure Left ventricular dp/dt, (mm Hg/sec) Heart rate (beats/min) Circumflex coronary venous pressure (mmhg) Myocardial oxygen consumption Oxygen content in a circumflex vein (ml/100 ml) Myocardial lactate consumption (jimol/min per 100 g) Data are mean ± SEM. P < 0.01 CSNS vs. control; t P < Propranolol (n = 7) Propranolol + Rauwolscine (n = 7) Control CSNS Control CSNS 44 ±5 85 ±6 52 ±4 94 ±6 33 ±5* 91 ±6* 59 ±4* 102 ± 7 38 ±6 71 ±7 40 ±4 90 ±6 5.8 ± ± ± ± ± 60* 1140 ± ±9 117 ± ± ± ± ± ± ± 0.7f 4.8 ± ± ± ± ±7 9 ± 8t 30 ± ±9t 76 ± 7 44 ±4* 96 ±6* 6.1 ± ± ± ± ± l.of 50 ± ± 12

7 14 Another source of error for the calculation of poststenotic coronary resistance may be an underestimation of myocardial blood flow by varying amounts of collateral blood flow from the nonoccluded artery, since we measured only arterial inflow. Thus, the decrease in arterial inflow distal to a severe stenosis might be replaced by enhanced collateral flow during sympathetic stimulation, and poststenotic coronary vascular resistance may be constant. However, if we regard the difference between aortic and poststenotic coronary pressure as the driving pressure gradient for collateral blood flow, this gradient was rather reduced during sympathetic stimulation. In addition, the decrease in systolic segment shortening and the net lactate production of the circumflex-perfused myocardium during sympathetic stimulation indicate a real myocardial perfusion impairment. Thus, the reactions of calculated end-diastolic coronary resistance during cardiac sympathetic nerve stimulation are predominantly due to an interaction of metabolic dilatory mechanisms and sympathetic vasoconstriction in the poststenotic coronary vascular bed. ai- vs. a 2 -Receptors in the Mediation of Sympathetic Vasoconstriction Phentolamine decreased the resistance of the poststenotic vascular bed and prevented the increase in resistance during cardiac sympathetic nerve stimulation. These beneficial effects of phentolamine could be due to blockade of postjunctional a-receptors and the prevention of a-receptor-mediated vasoconstriction, as well as to blockade of prejunctional a-receptors and a metabolic vasodilation resulting from increased norepinephrine release (Gorman and Sparks, 1982; Saeed et al., 1982). To differentiate these mechanisms, we employed more specific drugs. Prazosin is supposed to be a relatively specific antagonist of postjunctional ai-receptors (Hoffman and Lefkowitz, 1980), but failed to decrease the resistance of the poststenotic vascular bed and to prevent the increase in resistance during sympathetic stimulation. The same dosage of 1.2 mg/kg prazosin intravenously has been employed in various vascular preparations without evidence of a partial agonism (Holtz et al., 1982; Saeed et al., 1982). Thus the increase in the resting resistance of the poststenotic vascular bed cannot be attributed to a partial a-agonism and is rather explained by the decrease in poststenotic perfusion pressure and thus by less distention of the dilated poststenotic vascular bed. The increase in poststenotic resistance during sympathetic stimulation was obviously resistant to a\ -antagonism. In contrast to prazosin, rauwolscine, which is a very potent a 2 -receptor antagonist (McGrath, 1982), was effective to the same degree as phentolamine, both in decreasing the resting resistance of the post- Circulation Research/Vol. 53, No. 1, July 1983 stenotic vascular bed and in preventing the increase in resistance during sympathetic stimulation. The decrease in the resting resistance could be abolished by previous j9-blockade. Thus this decrease may be due to enhanced norepinephrine release and a resulting metabolic dilation, which is no longer present after /3-blockade. A recent study by Gorman and Sparks (1982) has just demonstrated a significant constrictive resting tone in severely stenotic coronary arteries which could be abolished by phentolamine and phenoxybenzamine, but could not be attributed to postjunctional a-adrenergic effects. The genesis of this vasoconstrictive resting tone in severely stenotic coronary arteries remains unclear. The increase in poststenotic coronary resistance during sympathetic stimulation, however, could be prevented by rauwolscine, without and with previous /3-blockade. According to the criteria of Mc- Grath (1982), the ineffectiveness of prazosin and the effectiveness of rauwolscine, even in the presence of /3-blockade, provide proof that the increase in poststenotic resistance during sympathetic stimulation is mediated by postjunctional a 2 -receptors. Sympathetic Vasoconstriction and Myocardial Ischemia Whether a-sympathetic vasoconstriction of coronary arteries can overwhelm metabolic vasodilation and induce myocardial ischemia is still controversial. Studies by Mudge and co-workers (19761, 1979) in patients with coronary artery disease indicate a significant coronary vasoconstriction and precipitation of angina pectoris in response to the cold pressor test. In contrast, experiments in conscious pigs with intravenous administration of norepinephrine showed a competition of a-vasoconstriction with metabolic vasodilation distal to a severe stenosis, but no sufficient vasoconstriction to overcome the metabolic dilation and induce ischemia (Gewirtz et al., 1982). Experiments in anesthetized dogs with moderate coronary stenoses revealed an a-constrictive limitation of oxygen delivery up to cardiac failure, but no net lactate production during inrracoronary infusion of norepinephrine (Buffington and Feigl, 1981). Inrracoronary infusion of norepinephrine distal to severe coronary stenoses in anesthetized dogs still resulted in an increase of coronary blood flow (Gorman and Sparks, 1982). Our findings of a less potent vasodilation in moderately stenosed coronary arteries than in intact vessels, but without evidence of ischemia during cardiac sympathetic nerve stimulation, correspond well to the experiments of Buffington and Feigl (1981). Our results of a marked vasoconstriction in severely stenotic coronary arteries, together with net lactate production and ventricular fibrillation in four out of 42 dogs, however, are in contrast to the findings of Gewirtz et al. (1982) and Gorman and Sparks (1982). Besides possible differences between the effects of

8 Heusch and Deussen/Sympathetic Effects on Stenotic Coronary Arteries exogenous and endogenous norepinephrine, this contrast could be explained by only tiny differences in the severity of the stenoses and in the poststenotic hyperemic reserve. Whereas, in our preparation, the coronary hyperemic reserve, as indicated by the reactive hyperemia repayment following a 15 -second occlusion, was almost exhausted, there was still a substantial reserve in the preparation of Gorman and Sparks (1982) and at least some reserve in the subepicardium in the preparation of Gewirtz et al. (1982). Only tiny differences in the degree of hyperemic reserve may then be relevant, since the relationship between the response of coronary resistance to sympathetic stimulation and the coronary reserve is very steep in this range of stenoses (Fig. 3). This explanation emphasizes how a slight progress in atherosclerotic narrowing of a coronary artery can suddenly induce clinical symptoms up to myocardial ischemia and ventricular fibrillation. Since we cannot exclude a small remaining vasodilatory reserve in the subepicardium even after elimination of reactive hyperemia in our preparation, a transmural steal phenomenon as observed by Gallagher et al. (1982) might act in addition to the increase in total poststenotic resistance and even further compromise subendocardial perfusion during sympathetic stimulation. Under our experimental conditions, a-adrenergic vasoconstriction in severely stenotic coronary arteries is not only a contributing but a necessary factor for the genesis of myocardial ischemia during sympathetic stimulation, since after administration of phentolamine or rauwolscine a marked increase in myocardial performance and metabolism never induced ischemia. In conclusion, our results demonstrate a continuous unmasking of sympathetic vasoconstrictive influences with decreasing coronary reserve. This sympathetic vasoconstriction is mediated by postjunctional a 2 -receptors and can induce myocardial ischemia distal to severe coronary stenoses during cardiac sympathetic nerve stimulation. Our results may be relevant to clinical studies, since a-vasoconstriction might play a significant role in the pathogenesis of myocardial ischemia (Mudge et al., 1976, 1979). Prof. Dr. V. Thdmer kindly reviewed this manuscript. Dipl. Phys. H. Muller performed the statistical analysis of the data. We thank B. Patzerfor excellent technical assistance and her help in preparing the manuscript. Part of these results were presented at the 33rd Annual Fall Meeting of the American Physiological Society in San Diego, Address for reprints: Dr. Gerd Heusch, Physiologie I, Universitat Diisseldorf, Moorenstr. 5, 4000 Dusseldorf, West Germany. Received December 6, 1982; accepted for publication April 29, References Bellamy RF (1980) Calculation of coronary vascular resistance. Cardiovasc Res 14: Berne RM, De Geest H, Levy MN (1965) Influence of the cardiac nerves on coronary resistance. Am J Physiol 208: Buffington CW, Feigl EO (1981) Adrenergic coronary vasoconstriction in the presence of coronary stenosis in the dog. Circ Res 48: Coffman JD, Gregg DE (1960) Reactive hyperemia characteristics of the myocardium. Am J Physiol 199: Downey JM, Kirk ES (1975) Inhibition of coronary blood flow by a vascular waterfall mechanism. Circ Res 36: Feigl EO, D'Alecy LG (1972) Normal arterial blood ph, oxygen, and carbon dioxide tensions in unanesthetized dogs. J Appl Physiol 32: Gallagher KP, Osakada G, Matsuzaki M, Kemper WS, Ross J (1982) Myocardial blood flow and function with critical coronary stenosis in exercising dogs. Am J Physiol 243: H698-H707 Gewirtz H, Most AS, Williams DO (1982) The effect of generalized alpha-receptor stimulation on regional myocardial blood flow distal to a severe coronary artery stenosis. Circulation 65: Gorman MW, Sparks HV (1982) Progressive coronary vasoconstriction during relative ischemia in canine myocardium. Circ Res 51: Heyndrickx GR, Muylaert P, Pannier JL (1982) a-adrenergic control of oxygen delivery to myocardium during exercise in conscious dogs. Am J Physiol 242: H805-H809 Hoffman BB, Lefkowitz RJ (1980) Alpha-adrenergic receptor subtypes. N Engl J Med 302: Holtz J, Saeed M, Sommer O, Bassenge E (1982) Norepinephrine constricts the canine coronary bed via postsynaptic a 2 -adrenoceptors. Eur J Pharmacol 82: Horeyseck G, Janig W, Kirchner F, Thamer V (1976) Activation and inhibition of muscle and cutaneous postganglionic neurones to hindlimb during hypothalamically induced vasoconstriction and atropine-sensitive vasodilation. Pfluegers Arch 361: Johannsen UJ, Mark AL, Marcus ML (1982) Responsiveness to cardiac sympathetic nerve stimulation during maximal coronary dilation produced by adenosine. Circ Res 50: McGrath JC (1982) Evidence for more than one type of postjunctional a-adrenoceptor. Biochem Pharmacol 31: Mohrman DE, Feigl EO (1978) Competition between sympathetic vasoconstriction and metabolic vasodilation in the canine coronary circulation. Circ Res 42: Mudge GH, Grossman W, Mills RM, Lesch M, Braunwald E (1976) Reflex increase in coronary vascular resistance in patients with ischemic heart disease. N Engl J Med 295: Mudge GH, Goldberg S, Gunther S, Mann T, Grossman W (1979) Comparison of metabolic and vasoconstrictor stimuli on coronary vascular resistance in man. Circulation 59: Munch DF, Downey JM (1980) Prediction of regional myocardial blood flow in dogs. Am J Physiol 239: H308-H315 Murray PA, Vatner SF (1979) a-adrenoceptor attenuation of the coronary vascular response to severe exercise in the conscious dog. Circ Res 45: Saeed M, Sommer O, Holtz J, Bassenge E (1982) a-adrenoceptor blockade by phentolamine causes j3-adrenergic vasodilation by increased catecholamine release due to presynaptic a-blockade. J Cardiovasc Pharmacol 4: Schwartz PJ, Stone HL (1977) Tonic influence of the sympathetic nervous system on myocardial reactive hyperemia and on coronary blood flow distribution in dogs. Circ Res 41: INDEX TERMS: Sympathetic vasoconstriction Coronary stenosis Myocardial ischemia a i - and a 2 -Receptors 15