Cardiac Vagal Reflex Modulates Intestinal Vascular Capacitance and Ventricular Preload in Anesthetized Dogs With Acute Myocardial Infarction
Background The purpose of the present study was to examine the effects of the cardiac vagal reflex on intestinal vascular capacitance and cardiac filling pressure during experimental acute myocardial infarction (AMI).
Methods and Results AMI was induced in anesthetized dogs through injection of microspheres into the left main coronary artery. Intestinal blood volume was measured with blood-pool scintigraphy. Portal venous pressure was varied through graded inflation of a portal venous constrictor to determine the intestinal vascular pressure-volume relation. Induction of AMI decreased intestinal blood volume to 88±3% of the control value (P<.01) and shifted the pressure-volume relation toward the pressure axis. This change was associated with increased left ventricular (LV) end-diastolic pressure (LVEDP) (from 6±1 to 17±2 mm Hg, P<.01) and LV segment length (to 112±4% of the control value, P<.01). During AMI, blockade of the cardiac vagal reflex by intrapericardial application of 2% lidocaine further decreased intestinal blood volume (to 83±3% of the control value, P<.05, versus AMI without lidocaine), increased LVEDP (to 22±2 mm Hg, P<.05, versus AMI without lidocaine), and tended to increase LV segment length (to 115±5%, P<.10). Lidocaine had no effect in dogs with AMI that had been vagotomized.
Conclusions These results suggest that the cardiac vagal reflex modulates the decrease in the intestinal vascular capacitance induced by AMI and modulates ventricular preload through pooling of blood in the intestinal circulation.
The cardiac vagal reflex can be activated during AMI, especially when the inferior wall of the LV is involved.1 2 The result is a vasodepressor effect due to decreased sympathetic and increased parasympathetic efferent activity. Furthermore, it has been well established that AMI is associated with activation of vasopressor neurohumoral agents.3 Thus, cardiovascular responses during AMI are determined by or are the result of integrated vasopressor and vasodepressor effects. In dogs with AMI, Thames and Abboud4 have shown an interaction between the cardiac vagal reflex and arterial baroreflexes in the control of renal sympathetic nerve activity. A previous study5 also showed that the cardiac vagal reflex attenuates the arterial baroreflex-induced increase in cardiac sympathetic activity during AMI. Moreover, Rutlen and Underwood6 have shown that transient ischemia induced by selective occlusion of the left circumflex coronary artery increases total vascular capacitance. The previous studies did not assess the contribution of the cardiac vagal reflex to the changes in splanchnic vascular capacitance or its role in the modulation of ventricular preload during AMI.
The richly innervated splanchnic vascular bed contributes to a major portion of the blood volume mobilized by reflex mechanisms.7 8 Because the splanchnic circulation contains ≈25% to ≈30% of the total blood volume, changes in splanchnic vascular capacitance may play an important role in redistribution of blood between the peripheral circulation and the heart. The present study was designed to determine the effects of the cardiac vagal reflex on intestinal vascular capacitance and LV filling pressure in a canine model of the experimental AMI. AMI was induced by injection of microspheres into the left main coronary artery. The intestinal vascular P-V relation was determined with the newly described method of radionuclide plethysmography, which uses blood-pool scintigraphy and alteration of portal venous pressure.9
Twelve mongrel dogs of either sex (14 to 22 kg) were anesthetized initially with sodium thiopental (25 mg/kg IV). Dogs were placed in the supine position and ventilated with a mixture of oxygen (30%) and nitrous oxide (70%) with use of a constant-volume respirator (model 607, Harvard Apparatus Co) and a closed rebreathing system. Fentanyl was given at an initial dose of 50 mg/kg for 5 minutes, and supplemental doses (20 to 50 mg/kg per hour) were given as necessary to maintain anesthesia. Pao2, Paco2, and pH were measured periodically during the experiment so values could be maintained within physiological ranges. Body temperature was maintained by a warming blanket. Splenectomy was performed through a midline abdominal incision. A 4×4-cm sheet of radiographic apron material was sutured immediately under the left ventral abdominal wall so that the count rate in the anterior wall of the abdomen could be determined.9 10 11 A pneumatic occluder was placed around the portal vein near the liver. A small fluid-filled catheter was advanced into the portal vein from a small arcade vein. The end of the catheter was positioned just upstream from the occluder to measure the pressure on the intestinal side of the occluder. After instrumentation, the abdomen was closed.
A midline sternotomy was performed on each animal. After the pericardium was opened, an ultrasonic flow probe (model T101, Transonic Systems Inc.) was placed around the ascending aorta. Ultrasonic crystals were implanted in the subendocardium of the LV free wall to measure segment length, and pacing electrodes were sutured to the right atrium to control heart rate. A soft polyvinyl catheter was sutured close to the origin of the left circumflex coronary artery for topical application of lidocaine to block the cardiac vagal reflex. After administration of lidocaine, fluid accumulation was prevented by withdrawal of excess liquid through a second polyvinyl catheter that was sutured on the dorsal wall of the heart. After instrumentation, the pericardium was reapproximated, and care was taken to not compromise pericardial volume.
LV pressure was measured with an 8F micromanometer-tipped catheter with a reference lumen (model Pc480, Millar Instruments). Aortic pressure was recorded with a fluid-filled catheter introduced via the left brachial artery. A midline cervical incision was made to expose both vagi. Pressure, flow, and LV segment-length signals were conditioned (model VR-16, Electronics for Medicine/Honeywell) and recorded with a computer (IBM model PC/AT). The data were subsequently analyzed with a computer using specially designed software (CVSOFT, ODESSA Computer Systems Ltd.).
Induction of AMI
Experimental AMI was induced in eight animals. From the femoral artery, a 5F left coronary artery catheter was advanced into the left main coronary artery under fluoroscopy. A small amount of contrast medium was injected into the left main coronary artery to determine that the contrast was distributed to both left anterior descending and left circumflex coronary arteries. Polystyrene microspheres (50 μm in diameter) were dispersed in dextran to a concentration of 4 mg/mL. AMI was induced by bolus injections of the microsphere suspension (2.5 mL every 5 minutes). The bolus volume was reduced to 1 mL each when LVEDP reached ≈15 mm Hg. When LVEDP reached ≈20 mm Hg, no additional microspheres were injected. The induction of AMI required ≈1 hour, during which 4 to 6 mg of microspheres/kg body wt was injected. This model of AMI has been shown to produce features similar to those observed clinically such as increases in LVEDP and serum enzyme activities, ECG signs of AMI, multiple small transmural infarcts, and a reduction in myocardial blood flow.12 13
Relative changes in the intestinal blood volume were assessed with the use of equilibrium blood-pool scintigraphy.9 10 11 14 In brief, erythrocytes were labeled in vivo with 99mTc. Static scintigrams were recorded for 60 seconds with a gamma camera (model DYNA-MO 4, Picker) equipped with a parallel-hole, high-sensitivity collimator and interfaced to a nuclear medicine computer system (model DPS-3300, ADAC Laboratories). To minimize the amount of circulating free 99mTc, scintigrams were not recorded until at least 30 minutes after erythrocyte labeling. The camera was positioned above the abdomen to include the entire field of the small intestine. The intestinal region of interest was defined as the sum of two regions on each side of the abdominal midline. Because the position of the camera in relation to the animal was unchanged during successive recordings, the regions of interest were drawn only on the first scintigram. Of course, radioactivity emanating from these regions of interest reflected the amount of blood contained in the vessels of the ventral abdominal wall as well as the vessels of splanchnic circulation. The contribution of the abdominal wall was determined by obtaining count rates from the region of interest drawn over the central portion of the leaded sheet. The net intestinal count rate was calculated by subtracting the count rate emanating from the ventral abdominal wall (ie, the region of interest drawn over the leaded sheet) from the total count rate in the intestinal region of interest (all count rates were expressed per pixel). Count rates were corrected for physical decay and biological decay according to the method described previously.11 12 Biological decay refers to changes in the count rate due to loss of the isotope from the erythrocyte or change in hematocrit. To correct for biological decay, reference blood samples (100 μL) were drawn every 3 minutes for determination of specific radioactivity. Count rates of reference blood samples were plotted against time, and the resulting curve was smoothed with a five-point running-average filter. The specific activity at the time of each scintigram was obtained from these filtered data through interpolation. Thus, the net intestinal count rate was divided by the specific radioactivity of blood at the time of each scintigram. Relative changes in the intestinal blood volume were determined by quantifying changes in the intestinal count rate as percentages of the count rate of the first scintigram.
ntestinal vascular P-V relation was determined according to a previously described method.9 10 11 Portal venous pressure was increased from baseline (≈7 mm Hg) to ≈14 and 20 mm Hg through graded inflation of the portal venous occluder. At each of these three portal venous pressure levels, a 60-second abdominal scintigram was recorded. After each scintigram, the portal venous pressure was returned to the baseline level through deflation of the occluder. In this manner, three sets of data (portal venous pressure and intestinal vascular volume) were collected, which allowed us to determine an intestinal vascular P-V relation, a process that required 4 to 5 minutes.
In six animals, while the heart rate was maintained constant with the use of electrical pacing, LVEDP, LV segment length, aortic blood flow and pressure, and the intestinal vascular P-V relation were recorded in three stages: (1) the control state, at least 30 minutes after completion of the instrumentation; (2) after induction of AMI, at least 15 minutes after hemodynamic stabilization; and (3) 3 to 5 minutes after intrapericardial application of 2% lidocaine (1 mL/min for 2 minutes). The intrapericardial fluid was then withdrawn via the second catheter. In two of these animals, after induction of AMI, vehicle (saline) was applied intrapericardially before lidocaine application. In three of these animals, blood samples were drawn from the pulmonary artery after intrapericardial application of lidocaine. Plasma concentrations of lidocaine were measured with a high-performance liquid chromatographic method15 ; none was detected.
Vagotomy (two animals) and intrapericardial application of lidocaine (two animals) were performed under control conditions. In two additional animals subjected to AMI, vagotomy was performed before intrapericardial application of lidocaine. The change in the intestinal vascular P-V relation in response to vagotomy and subsequent intrapericardial application of lidocaine was determined.
In the control state, after induction of AMI, and after intrapericardial application of lidocaine, changes in hemodynamic values, the slope of the P-V relation, and IBV7 were compared with the use of one-way ANOVA with a factorial-measure design. The slope of the P-V relation was calculated with linear regression. Significance of changes in the P-V relation after the induction of AMI or after intrapericardial application of lidocaine was determined with two-way ANOVA with a repeated-measure design. The data were expressed as mean±SEM. Values of P<.05 were considered to be significant.
Changes in hemodynamics and intestinal blood volume in response to induction of AMI and intrapericardial application of lidocaine are shown in the Table.⇓ AMI was associated with a decrease in cardiac output and mean aortic pressure but no statistically significant change in heart rate. As IBV7 decreased, LVEDP and LV segment length increased. Blockade of the cardiac vagal reflex through intrapericardial application of lidocaine resulted in an additional increase in LVEDP and a further decrease in IBV7. Blockade of the cardiac vagal reflex produced slight but not statistically significant changes in LV segment length (P<.10), mean aortic pressure, and cardiac output.
The effects of AMI and intrapericardial application of lidocaine on the intestinal vascular P-V relation are shown in Figs 1⇓ and 2⇓. AMI shifted the relation toward the pressure axis, indicating a decrease in intestinal vascular capacitance. After AMI, blockade of the cardiac vagal reflex through intrapericardial application of lidocaine resulted in a further leftward shift of the relation, whereas intrapericardial application of the saline vehicle had no effect (not shown). After AMI, cervical vagotomy also resulted in a further leftward shift of the P-V relation; subsequently, application of lidocaine produced no further shift (Fig 2). Induction of AMI and blockade of the cardiac vagal reflex through intrapericardial lidocaine or vagotomy had no significant effect on the slope of the P-V relation (Figs 1 and 2 and Table). To examine the effects of the cardiac vagal reflex on the intestinal vascular P-V relation under control conditions, we performed cervical vagotomy and intrapericardial application of lidocaine in additional control animals. These interventions did not affect the intestinal vascular P-V relation in animals without AMI (Figs 3⇓ and 4⇓).
The results of the present study show that after experimental induction of AMI, the cardiac vagal reflex attenuates intestinal venoconstriction and, probably through this mechanism, modulates LV filling pressure. After AMI, intestinal blood volume decreased (ie, the vascular P-V relation shifted toward the pressure axis) and LVEDP and LV end-diastolic segment length increased. Subsequent blockade of the cardiac vagal reflex with intrapericardial lidocaine further decreased intestinal blood volume, increased LVEDP, and tended to increase LV segment length. Alternatively, cervical vagotomy had similar effects on intestinal blood volume, after which lidocaine had no further effect. These findings suggest that after AMI, vagal afferent activity inhibits venoconstriction.
Activation of the cardiac vagal reflex has been demonstrated in humans and animals with AMI.2 16 Embolization of the canine left main coronary artery reduces sympathetic nerve activity and inhibits arterial baroreflex-induced vasoconstriction.17 18 Furthermore, Thoren18 showed that the activity of feline cardiac vagal afferents from ischemic myocardium increases. Activation of vagal afferents in ischemic myocardium may be due to locally increased concentrations of substances such as prostacyclin20 as previous studies have shown that the intracoronary administration of prostacyclin increases vagal afferent activity.10 21 22
With an external reservoir technique, Rutlen and Underwood12 reported that selective occlusion of the left circumflex coronary artery increased total intravascular volume, which was attenuated with vagotomy or epicardial lidocaine administration. This study involved only transient (2.5-minute) ischemic episodes, and the hemodynamic results were modest (ie, mean arterial pressure decreased by 5 mm Hg, and left atrial pressure increased by 1 mm Hg). In contrast, in the present study, global ischemia was produced through embolization of the left main coronary artery and was characterized by features similar to those observed clinically (eg, a significant decrease in cardiac output and an increase in LVEDP).12 This experimental model of AMI resulted in a decrease, rather than an increase, in intestinal blood volume. The present finding is consistent with the concept that peripheral vasoconstriction is one of clinical hallmarks of global ischemia. Furthermore, we observed that after induction of AMI, blockade of the cardiac vagal reflex through intrapericardial application of lidocaine or vagotomy accentuated the reduction in intestinal blood volume. Thus, AMI may be associated with integrated vasopressor and vasodepressor control of vascular capacitance. Because the venodilating effect of the cardiac vagal reflex appears to be overridden by the concomitant intestinal venoconstriction, the net AMI effect on intestinal capacitance vessels is a decrease in intestinal blood volume.
Induction of AMI and subsequent interruption of the cardiac vagal reflex shifted the intestinal vascular P-V relation toward the pressure axis, indicating a decrease in intestinal vascular capacitance. Changes in splanchnic vascular capacitance are important mechanisms in the regulation of blood volume and cardiac filling. We showed that the decreases in intestinal blood volume during AMI and after blockade of the cardiac vagal reflex were associated with an increase in LVEDP and an apparent increase in LV dimensions, independent of changes in heart rate. Our findings are consistent with those of a previous study that demonstrated that intestinal blood volume decreased and LVEDP increased during experimental AMI.11 Furthermore, stimulation of the cardiac vagal reflex through infusion of veratridine, arachidonic acid, or prostacyclin into the circumflex coronary artery was reported to decrease LVEDP and LV dimension.23 These changes in ventricular filling pressure are likely to result, at least in part, from a redistribution of blood to the intestinal circulation.
Intestinal vascular P-V relations were approximately linear over the range of portal venous pressure between 7 and 20 mm Hg. Factors (eg, collagen) that tend to reduce vascular compliance at a higher venous pressures may not play a dominant role over the range studied. The present study is consistent with previous ones that have shown a linear intestinal vascular P-V relation.9 10 11 Because increments in portal venous pressure were associated with unknown volume increments in the different vascular compartments between the large arteries and the portal vein, we cannot be certain that the slope of the P-V relation accurately represents the slope of the small vein-venule vascular compartment.
Intrapericardial application of lidocaine has been shown to effectively block the cardiac vagal reflex.24 It is unlikely that lidocaine had any systemic effect because in our study, no drug could be detected in pulmonary arterial blood immediately after the intrapericardial application of lidocaine. Furthermore, the effects of lidocaine on the intestinal vascular P-V relation and LVEDP during AMI were abolished with vagotomy. This suggests that lidocaine-induced changes were mediated through cardiac vagal afferents.
For several reasons, the results of this investigation in anesthetized, open-chest dogs that sustained microsphere embolization of their left main coronary arteries should be interpreted cautiously. The potential role of this reflex in the intact, normal state remains to be determined. The fact that vagotomy or intrapericardial application of lidocaine failed to affect intestinal vascular P-V relation in control animals suggests that under normal conditions, the cardiac vagal reflex has little or no role in the control of intestinal vascular tone, but this tentative conclusion must be confirmed in other models. The effects of the cardiac vagal reflex on LV preload must also be addressed in more intact models and in patients. In our model, interruption of the cardiac vagal reflex increased LVEDP and apparently increased left ventricular dimensions. Whether the magnitude of these changes is the same in an unanesthetized closed-chest model in which the heart is larger remains to be determined.
An increase in ventricular filling pressure is a common finding in patients with AMI. Our data and those of others suggest that the decrease in intestinal vascular capacitance tends to increase filling pressure, which, if sufficiently high, can result in pulmonary congestion. The cardiac vagal reflex may have a beneficial modulating effect by limiting the decrease in venous capacitance and thereby limiting the increase in ventricular filling pressure.
This work was supported by grants-in-aid from the Alberta Heart and Stroke Foundation (Calgary) to Dr Manyari and to Dr Tyberg. Dr Tyberg is a medical scientist of the Alberta Heritage Foundation for Medical Research (Edmonton). We thank Perry Anderson, Gerry Groves, and Cheryl Meek for their skillful technical assistance and Dr I. Belenkie for his helpful criticism of the manuscript.
Selected Abbreviations and Acronyms
|AMI||=||acute myocardial infarction|
|LV||=||left ventricular (ventricle)|
|LVEDP||=||left ventricular end-diastolic pressure|
|IBV7||=||intestinal blood volume at a portal venous pressure of 7 mm Hg|
- Received October 11, 1995.
- Revision received January 11, 1996.
- Accepted January 22, 1996.
- Copyright © 1996 by American Heart Association
Thames MD, Klopfenstein HS, Abboud FM, Mark AL, Walker JL. Preferential distribution of inhibitory cardiac receptors with vagal afferents to the inferoposterior wall of the left ventricle activated during coronary occlusion in the dog. Circ Res.. 1978;43:512-519.
Webb SW, Adgey AA, Pantridge JF. Autonomic disturbance at onset of acute myocardial infarction. Br Med J.. 1972;3:89-92.
Dzau VJ. Contributions of neuroendocrine and local autocrine-paracrine mechanisms to the pathophysiology and pharmacology of congestive heart failure. Am J Physiol Endo Metab.. 1982;62:E76-E82.
Thames MD, Abboud FM. Reflex inhibition of renal sympathetic nerve activity during myocardial ischemia mediated by left ventricular receptors with vagal afferents in dogs. J Clin Invest.. 1979;63:395-402.
Felder RB, Thames MD. Interaction between cardiac receptors and sinoaortic baroreceptors in the control of efferent cardiac sympathetic nerve activity during myocardial ischemia in dogs. Circ Res.. 1979;45:728-736.
Rutlen DL, Underwood RS. Reflex influence of selective coronary artery occlusion on the total capacitance vasculature in the dog. J Clin Invest.. 1984;73:241-250.
Haase EB, Shoukas AA. Carotid sinus baroreceptor reflex control of venular pressure-diameter relations in rat intestine. Am J Physiol.. 1991;260:H752-H758.
Ozono K, Bosnjak JZ, Kampine JK. Reflex control of mesenteric vein diameter and pressure in situ in rabbits. Am J Physiol.. 1989;256:66-H1072.
Scott-Douglas NW, Manyari DE, Smiseth OA, Robinson VJB, Wang SY, Smith ER, Tyberg JV. Measurement of intestinal vascular capacitance in dogs: an application of blood-pool scintigraphy. J Appl Physiol.. 1995;78:232-238.
Wang Y, Drakonakis C, Alderman JL, Rutlen DL. Changes in regional vascular capacitance during prostacyclin-induced cardiac vagal reflex in pigs. Am J Physiol.. 1994;267:H535-H539.
Wang SY, Manyari DE, Scott-Douglas NW, Smiseth OA, Smith ER, Tyberg JV. Splanchnic venous pressure-volume relation during experimental acute ischemic heart failure: differential effects of hydralazine, enalaprilat, and nitroglycerin. Circulation.. 1995;91:1205-1212.
Rutlen DL, Wackers FJ, Zaret BL. Radionuclide assessment of peripheral intravascular capacity: a technique to measure intravascular volume changes in the capacitance circulation in man. Circulation.. 1981;64:146-152.
Baker DG, Coleridge HM, Coleridge JCG. Vagal afferent ‘C’ fibers from the ventricle. In: Hainsworth R, Kidd C, Linden RJ, eds. Cardiac Receptors. Cambridge, UK: Cambridge University Press; 1979:117-137.
Kezdi P, Kordenat RK, Misra SN. Reflex inhibitory effects of vagal afferents in experimental myocardial infarction. Am J Cardiol.. 1974;33:858-860.
Toubes DB, Brody MJ. Inhibition of reflex vasoconstriction after experimental coronary embolization in the dog. Circ Res.. 1970;26:211-224.
Thoren P. Activation of left ventricular receptors with non-medullated vagal afferents during occlusion of a coronary artery in the cat. Am J Physiol.. 1976;37:1046-1051.
Nganele DM, Hintze TH. Cardiac chemical reflex control of preload in conscious dogs. Am J Physiol.. 1990;258:H1055-H1063.
Evans RG, Ludbrook J, Potocnic SJ. Intracisternal naloxone and cardiac nerve blockade prevent vasodilatation during simulated hemorrhage in awake rabbits. J Physiol.. 1989;409:114-119.