Anesth Analg 2000;91:1112-1117
© 2000 International Anesthesia Research Society
CARDIOVASCULAR ANESTHESIA
Dexmedetomidine and Hemodynamic Responses to Simulated Hemorrhage in Experimental Heart Failure
D. W. Blake, MB, BS, PhD, FANZCA
Department of Pharmacology, Anesthesia Research Unit, University of Melbourne, Victoria, Australia
Address correspondence and reprint requests to D. W. Blake, MB, BS, PhD, FANZCA, Department of Pharmacology, Anesthesia Research Unit, University of Melbourne, Parkville, Victoria, 3010 Australia. Address e-mail to d.blake{at}pharmacology.unimelb.edu.au
 |
Abstract
|
|---|
 2-Adrenoreceptor agonists may counteract the increased basal sympathetic nervous activity in patients with congestive heart failure (CHF), but they may also compromise reflex responses to hypovolemia. We have tested responses to simulated hemorrhage (central hypovolemia) after IV dexmedetomidine in normal animals and in experimental chronic CHF. Rabbits (n = 14) were treated with IV doxorubicin (or control saline) for 8 weeks inducing biventricular dilatation and myocardial damage. Cardiac output (CO) was measured continuously with a transit-time Doppler implanted on the ascending aorta. Progressive inflation of a cuff around the inferior vena cava (simulated hemorrhage) was used to reduce cardiac index at a constant rate. Arterial baroreceptor-mediated vasoconstrictor and heart rate responses were tested with repeated cuff inflations. Although resting CO was reduced in CHF, the blood pressure and heart rate changes with dexmedetomidine were not exaggerated. The slope of the vasoconstrictor response to graded hypovolemia was attenuated by dexmedetomidine with an earlier onset of decompensation. There was no added effect of CHF on the response until the dose of dexmedetomidine was sufficient to reduce resting CO in addition to arterial blood pressure and heart rate.
Implications: As an adjunct to anesthesia, dexmedetomidine may be useful in reducing basal sympathetic nervous activity. This study in experimental animals suggests this may be achieved without compromising protective responses to decreased blood volume.
|
Introduction
|
|---|
Adrenoreceptor agonists, such as dexmedetomidine, reduce resting sympathetic nervous system tone and sympathoadrenal responses to tracheal intubation or surgery (1). Although blood pressure and heart rate (HR) are reduced and more frequent support of blood pressure is needed, there may be a benefit from increased hemodynamic stability in high-risk patient groups with cardiac disease, such as in patients undergoing vascular or cardiac surgery (2–5). Improved hemodynamics might be expected in a particular subset of these patients, with chronic congestive heart failure (CHF), because of an increased resting sympathetic tone associated with increased afterload and HR (6). However, 2-adrenoreceptor agonists also modify baroreceptor-mediated reflexes already attenuated in CHF (6–8), and therefore could compromise responses to hypovolemia during surgery. The aim of this study was to test this hypothesis in rabbits, an animal model of CHF with many of the features of chronic CHF seen in humans (9–11), with doxorubicin-induced cardiomyopathy. Baroreceptor-mediated responses to hypovolemia were tested by repeated inflations of a vena caval cuff to reduce venous return and simulate hemorrhage (12,13).
|
Methods
|
|---|
Experiments were performed in 14 female New Zealand White rabbits weighing 1.9–3.5 kg (mean 3.1 kg). Experiments were approved by the Animal Ethics and Experimentation Committees of the Royal Melbourne Hospital and the University of Melbourne.
Surgical Procedures
Under general anesthesia, an inflatable cuff was placed around the inferior vena cava (IVC cuff) via a right thoracotomy (12). Two to three weeks later, an ultrasonic transit-time flow probe (Type 6STM; Transonics Systems, Ithaca, NY) was placed extrapericardially around the ascending aorta via a left thoracotomy. The first study was performed 2–3 weeks after the last surgery, when the rabbits had recovered and were gaining weight. Before each experiment, catheters were inserted under local anesthesia in a central ear artery and a marginal ear vein for arterial pressure measurement and for drug infusion. Tubing from the IVC cuff and the flowmeter plug, buried subcutaneously, was also retrieved using local anesthesia.
Doxorubicin Treatment
Doxorubicin (Adriamycin, Farmitalia Carlo Erba, 1 mg/ml in 0.9% saline) was injected into a marginal ear vein at a dose of 1 mg/kg twice weekly for 8 weeks in seven rabbits. Control rabbits were injected with saline alone. The hearts of five of the doxorubicin-treated rabbits were fixed immediately postmortem and examined histologically. All demonstrated biventricular dilatation and gross disorganization of the myocardium with generalized fibrous tissue infiltration and changes suggesting apoptosis.
Hemodynamic Measurements
Arterial pressure was measured via a catheter inserted on each experiment day at the base of a central ear artery. The transducer (Model P23 XLTM; Viggo-Spectramed, Oxnard, CA) was placed at heart level and calibrated with a mercury manometer. The flow probe was connected to a Transonics Systems flowmeter (T206) to measure ascending aortic flow, and cardiac output (CO) was derived using the known aortic diameter. HR was measured by a tachometer triggered by the flow pulse. Signals were amplified and recorded on a Grass polygraph (Model 7) and sent to a personal computer after analog-to-digital conversion. This provided 10 s mean values for the primary variables mean arterial pressure (MAP, mm Hg), CO ml/min and HR (beats/min), and for the derived variables cardiac index (CI = CO/kg body weight) and systemic vascular conductance index (SVCI = CI/MAP x 100, described below as units). Rabbits were restrained in an experiment box from 60 min prior to the onset of recording.
Acute Central Hypovolemia (Simulated Hemorrhage)
The IVC cuff was progressively inflated with saline using a micrometer-driven syringe to externally compress the IVC, so that CI decreased at a constant rate of 8% of its baseline level per minute (12,13). When CI had decreased to 33% of its baseline level or MAP to 40 mm Hg, whichever occurred first, CI was maintained constant for a further 5 min prior to deflation of the IVC cuff. The controlled variable, CI, was integrated each 2 s and displayed on the Grass polygraph. The SVCI response to IVC cuff inflation is reported in preference to vascular resistance, because previous studies found a linear relation between CI and SVCI, in contrast to vascular resistance. The vascular response to IVC cuff inflation was also found to be similar to the response to actual acute blood loss in the rabbit (12).
Drugs
Dexmedetomidine (Abbott Laboratories, UK) was diluted in normal saline to produce a 1–2 mL bolus for IV injection. Doses of 1.0, 3.0, and 10 µg/kg IV were given over 1 min, followed by an infusion of 5% of the bolus dose per minute. Simulated hemorrhage was initiated 12 min after the bolus dose. A dexmedetomidine dose of 1.0 µg/kg was found to be the threshold for MAP and HR changes in preliminary experiments.
Experiment Protocol
Four treatments were tested in each rabbit in both the control and CHF groups: IV saline and IV dexmedetomidine 1.0, 3.0, and 10 µg/kg. The order of treatments was according to a 4 x 4 latin square design. Progressive IVC cuff inflation to simulate hemorrhage (duration 10–13 min) was performed prior to each IV infusion and then repeated 12 min after the start of each treatment. At least 90 min was allowed for full recovery between IVC cuff inflations (12). Rabbits received two treatments (four IVC cuff inflations) on each experiment day, with at least two days for recovery between experiment days. Responses obtained prior to each treatment were used to test for time-related effects and residual drug effects.
Analysis of Results
Phases I and II of the hemodynamic response to central hypovolemia were separated by the point at which SVCI reached a minimum prior to an abrupt increase. The levels of the hemodynamic variables were recorded as 60 s averages at 5 points in each simulated hemorrhage: 1) baseline before saline or drug; 2) 10 min after drug or saline bolus and start of infusion; 3) during IVC cuff inflation in the final 60 s of Phase I; 4) the initial 60 s of Phase II; and 5) after 5 min of Phase II with CI maintained constant. The effects of dexmedetomidine and of control treatment versus doxorubicin-induced cardiomyopathy on variables at the five points were evaluated by two-way analysis of variance. The patterns of MAP, HR, and SVCI change between these points and across Phases I and II of the response to simulated hemorrhage were compared by repeated-measures analysis of variance with the Greenhouse-Geisser correction for serial autocorrelation (14). Trends were tested by the main effect of treatment (simulated hemorrhage) and by the treatment x drug dose and treatment x CHF (doxorubicin treatment) interactions. The statistical analyses were performed using Systat 5.0 (SPSS, Chicago, IL).
|
Results
|
|---|
Dexmedetomidine or Saline Infusion Before Simulated Hypovolemia
Saline infusion had no effect on baseline circulatory variables (P always >0.5). There were transient increases in MAP and decreased HR in the 5 min after IV bolus of dexmedetomidine (3 and 10 µg/kg), but baseline circulatory variables remained stable over the infusion period, 10–25 min after the initial IV bolus. The doxorubicin-treated rabbits had similar resting MAP and HR to controls (Fig. 1). Resting CI and SVCI were decreased in rabbits with CHF, compared with controls (P = 0.006, P = 0.02) after IV saline and dexmedetomidine (1 and 3 µg/kg; Fig. 1). There was a dose-related decrease in MAP and HR with dexmedetomidine infusions (25% and 14%, respectively, after 10 µg kg-1 dexmedetomidine, P < 0.001), but no difference in response between CHF and control rabbits (P > 0.2). Dexmedetomidine (10 µg kg-1) also decreased resting CI (18%, P = 0.02), although there was no additional effect with CHF (P = 0.14 for interaction). Dexmedetomidine (1 and 3 µg/kg) increased SVCI (P = 0.04) in control, but not CHF rabbits, although with 10 µg/kg the effect was abolished by the reduction in CI.
View larger version (28K):
[in this window]
[in a new window]
|
Figure 1. Effects of IV saline and dexmedetomidine (1.0 [D1], 3.0 [D3], and 10.0 [D10] µg/kg) on mean arterial pressure (MAP), cardiac index (CI), systemic vascular conductance index (SVCI), and heart rate. Values in control rabbits (left panel, n = 7) and doxorubicin-treated rabbits with congestive heart failure (right panel, CHF, n = 7) averaged over 2 min prior to testing the response to simulated hemorrhage. Error bars indicate standard deviation.  Significant interaction (P < 0.05) between the effects of dexmedetomidine and CHF.
|
|
Acute Simulated Hemorrhage
Because resting CI was similar after saline, dexmedetomidine 1.0 µg/kg and dexmedetomidine 3.0 µg/kg, the responses to progressive reduction in CI (simulated hemorrhage) after these treatments are plotted together in Fig. 2. Responses after dexmedetomidine 10 µg/kg are compared with saline infusion in Fig. 3. In both control and CHF groups, two phases were observed in response to simulated hemorrhage as previously described. In Phase I, MAP was maintained close to control by decreased SVCI and increased HR (12). Phase II was marked by an abrupt decrease in MAP and increase in SVCI, then a partial recovery over 5 min. Across all drug doses, the effect of CHF was to decrease both resting CI and CI at the end of Phase I (P = 0.002, FDF1,45 = 10.6, right shift of the curve), the average decrease in CI during Phase II (36%) being similar to controls (P = 0.5). This indicates that the CHF rabbits tolerated a similar simulated decrease in preload or simulated hemorrhage volume before decompensation and the onset of Phase II. However, dexmedetomidine shortened Phase I; thus, after 10 µg/kg the average change in CI was 22%, compared with 44% after saline infusion (P = 0.000, FDF3,45 = 10.3). The shift to the right of the curve with CHF opposed shortening of the duration of Phase I by dexmedetomidine (P = 0.03, FDF3,45 = 3.2 for drug x CHF at the end Phase I).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 2. Mean arterial pressure (MAP, mm Hg), systemic vascular conductance index (SVCI units), and heart rate (HR) are plotted for three levels of cardiac index (CI, ml min-1 kg-1): baseline, Phases I and II of the response to simulated hemorrhage. Left column: responses in control rabbits (n = 7). Right column: responses in rabbits with congestive heart failure after treatment for 8 weeks with doxorubicin (CHF, n = 7). §Significant right shift of curves with CHF (P = 0.002). Interactions between dexmedetomidine and CHF are not significant. Responses during control IV saline infusion are shown as dashed lines ( ); solid lines indicate IV dexmedetomidine infusions with 1.0 (•) and 3.0 ( ) µg/kg initial IV bolus. Significant shift of curve with dexmedetomidine P < 0.001. CI points in each panel: Left, baseline prior to simulated hemorrhage. Center, last minute of Phase I, prior to abrupt decrease in MAP. Right, average for first minute of Phase II. Error bars indicate SD (n = 7 rabbits).
|
|
Dexmedetomidine decreased MAP and HR throughout simulated hemorrhage (P = 0.000, FDF3,45 = 2.8; P = 0.000, FDF3,45 = 11.6). There was no significant interaction of the MAP or HR responses with CHF, nor any overall effect of CHF on MAP or HR (P > 0.3). SVCI was increased throughout the simulated hemorrhage response by dexmedetomidine (P = 0.009, FDF3,43 = 4.4), but this effect was opposed by CHF (P = 0.01, FDF1,43 = 7.1). The gradient of the SVCI response in Phase I was decreased by dexmedetomidine (P = 0.000, FDF3,45 = 10.3), but not changed by CHF (P = 0.9).
Changes during Phase II of the simulated hemorrhage response are shown in Fig. 4. It was possible to maintain CI for a further 5 min at the value required to initiate Phase II (CI reached at the right-hand point in Figs. 2 and 3), without obvious stress to the rabbits. In control rabbits receiving saline, there was a recovery in MAP to 64% of control associated with a decrease in SVCI to values similar to that prior to simulated hemorrhage. The recovery in MAP was limited by both CHF and dexmedetomidine (P = 0.04, FDF1,45 = 4.2 and P = 0.000, FDF3,45 = 8.0). The SVCI after 5 min did not differ significantly between controls and CHF (P = 0.5). However, the increase in SVCI (failure of vasoconstriction) at the onset of Phase II (Fig. 4, left point) was greater in controls (P = 0.005, FDF1,43 = 8.8). Change in SVCI during Phase II was reduced by both CHF (P = 0.001, FDF1,43 = 12.5) and by dexmedetomidine (P = 0.000, FDF3,4 = 11.0). Most of the effect of dexmedetomidine was from the 10 µg/kg dose, after which there was a further decrease in MAP during Phase II.
|
Discussion
|
|---|
Doxorubicin-induced cardiomyopathy is a dose-dependent process with cumulative myocyte damage. During its development, there is an overall increase in adrenergic nervous system activity (9), as described with CHF from other causes that may be due in part to baroreceptor dysfunction and loss of feedback inhibition (7,15). The CHF group in this study had a decreased CI and increased resting vasoconstrictor tone (decreased SVCI). Decreases in MAP and HR with dexmedetomidine were not exacerbated by CHF, although the increase in SVCI, seen with the smaller doses in control rabbits, was attenuated. This is consistent with the central 2-adrenoreceptor action of dexmedetomidine to decrease sympathetic nervous system outflow, although the largest dose caused a significant reduction in CI, probably via a peripheral effect.
Reducing vasoconstrictor tone is an important treatment approach in CHF, depending on its etiology, but this approach may exacerbate impaired baroreceptor function and limit vasoconstrictor responses to maintain blood pressure during hypovolemia. Interface with baroreceptor function would also be more likely with an 2-adrenergic agonist acting on central neural pathways than with a direct-acting peripheral vasodilator. Although altered baroreceptor function in CHF is well described for baroreceptor-mediated HR changes (6), little is known about effects on vascular tone. Wang et al.(7) demonstrated an increase in threshold pressure and a decrease in peak discharge with single-unit carotid sinus baroreceptor activity in dogs with pacing-induced heart failure. However, this decrease in receptor sensitivity to arterial pressure may not be associated with a reduction in the sympathetic nervous system response to unloading of the baroreceptors as achieved by central hypovolemia. This is suggested by the absence of any significant change to the vasoconstrictor response in the CHF group in these experiments, except that the response is obtained over a lower range of CI.
In normal rabbits, dexmedetomidine resulted in a dose-related attenuation of the vasoconstrictor response to hypovolemia and earlier decompensation. In CHF, although dexmedetomidine decreased resting CI, the early vasoconstrictor response to hypovolemia was unaltered. Dexmedetomidine (at smaller doses) improved the resting circulation in CHF, without significant impairment of responses to central hypovolemia. However, when the dose of dexmedetomidine was increased to significantly reduce resting CI, vasoconstriction with further reduction in CI was abolished. MAP and HR responses were then similar in normal and heart failure rabbits. Recovery of vasoconstriction after decompensation was impaired by either dexmedetomidine or CHF, but the effects were not additive. This study therefore suggests a cardiovascular benefit from dexmedetomidine in a model of CHF without significant loss of the protective sympathetic nervous system response to hypovolemia.
|
Acknowledgments
|
|---|
This study was supported by a research grant from the Australian and New Zealand College of Anesthetists.
The author thanks Ms. Linda Duncan (Department of Pharmacology, University of Melbourne) for her expert technical assistance and Professor P. Bhatal (Department of Pathology, Royal Melbourne Hospital) for the histological examination of myocardium from doxorubicin-treated rabbits. Dexmedetomidine was provided by Abbott Laboratories, UK.
|
References
|
|---|
-
Hayashi Y, Maze M. Alpha 2 adrenoceptor agonists and anesthesia [review]. Br J Anaesth 1993; 71: 108–18.[Free Full Text]
-
Flacke JW, Bloor BC, Flacke WE, et al. Reduced narcotic requirement by clonidine with improved hemodynamic and adrenergic stability in patients undergoing coronary bypass surgery. Anesthesiology 1987; 67: 11–9.[Web of Science][Medline]
-
Talke P, Li J, Jain U, et al. Effects of perioperative dexmedetomidine infusion in patients undergoing vascular surgery: the study of Perioperative Ischemia Research Group. Anesthesiology 1995; 82: 620–33.[Web of Science][Medline]
-
Myles PS, Hunt JO, Holdgaard HO, et al. Clonidine and cardiac surgery: haemodynamic and metabolic effects, myocardial ischaemia and recovery. Anesth Intensive Care 1999; 27: 137–47.[Web of Science][Medline]
-
Jalonen J, Hynynen M, Kuitunen A, et al. Dexmedetomidine as an anesthetic adjunct in coronary artery bypass grafting. Anesthesiology 1997; 86: 331–45.[Web of Science][Medline]
-
Sopher SM, Smith ML, Eckberg DL, et al. Autonomic pathophysiology in heart failure: carotid baroreceptor-cardiac reflexes. Am J Physiol 1990; 259: H689–96.[Abstract/Free Full Text]
-
Wang W, Chen JS, Zucker IH. Carotid sinus baroreceptor sensitivity in experimental heart failure. Circulation 1990; 81: 1959–66.[Abstract/Free Full Text]
-
Evans RG, Haynes JM, Ludbrook J. Effects of 5-HT-receptor and alpha 2-adrenoceptor ligands on the haemodynamic response to acute central hypovolaemia in conscious rabbits. Br J Pharmacol 1993; 109: 37–47.[Web of Science][Medline]
-
Arnolda L, McGrath BP, Cocks M, Johnston CI. Vasoconstrictor role for vasopressin in experimental heart failure in the rabbit. J Clin Invest 1986; 78: 674–9.
-
Noshiro T, Way D, McGrath BP. Angiotensin converting enzyme inhibition improves baroreflex-induced noradrenaline spillover responses in rabbits with heart failure. J Auton Nerv Syst 1997; 66: 87–93.[Web of Science][Medline]
-
Langton D, Jover B, McGrath BP, Ludbrook J. Cardiovascular responses to graded treadmill exercise during the development of doxorubicin induced heart failure in rabbits. Cardiovasc Res 1990; 24: 959–68.[Abstract/Free Full Text]
-
Ludbrook J, Potocnic SJ, Woods RL. Simulation of acute haemorrhage in unanesthetised rabbits. Clin Exp Pharm Physiol 1988; 15: 575–84.[Web of Science][Medline]
-
Blake DW, Van Leeuwen AF, Petring OU, et al. Haemodynamic response to simulated haemorrhage in the rabbit: interaction of i.v. anesthesia and hypoxia. Br J Anaesth 1995; 75: 610–5.[Abstract/Free Full Text]
-
Ludbrook J. Repeated measurements and multiple comparisons in cardiovascular research. Cardiovasc Res 1994; 28: 303–11.[Free Full Text]
-
Niebauer M, Zucker IH. Static and dynamic responses of carotid sinus baroreceptors in dogs with chronic volume overload. J Physiol 1985; 369: 295–310.[Abstract/Free Full Text]
Accepted for publication July 11, 2000.
|
Top
Abstract
Introduction
). Error bars indicate SD (n = 7).
