Acute Electrophysiologic Effects of Intravenous Amiodarone Are Independent of a Sympatholytic Action in Humans

Abstract

Previous experiments in animals demonstrated a novel sympatholytic action of acute intravascular amiodarone (AM). It is not known if this action also occurs in humans. Twelve male volunteers performed handgrip for 10 min before and after 300 mg intravenous (IV) AM over 60 min. The effect of handgrip was determined from changes in blood pressure (BP), heart rate (HR), and cardiac noradrenaline (NA) spillover. Changes in cardiac spillover of dihydroxyphenylglycol (DHPG), the metabolite of NA, were measured during AM infusion. The electrophysiological effects of AM were determined from changes to the A-H intervals during right atrial stimulation (100 beats/min). Handgrip increased HR (63 ± 2 to 84 ± 5 beats/min and 65 ± 3 to 84 ± 4 beats/min), systolic BP (141 ± 4 to 179 ± 6 mm Hg and 140 ± 4 to 179 ± 7 mm Hg), and cardiac NA spillover (11.9 ± 4 to 44.3 ± 13 ng/min and 17.3 ± 4 to 55.5 ± 11 ng/min) before and after AM, respectively (P < 0.02 in all groups). There was good correlation between increases in cardiac NA spillover and HR (r² = 0.86) and systolic BP (r² = 0.87). AM increased the A-H interval (95.5 ± 18 to 107.8 ± 20 ms, P < 0.02). There was no difference in hemodynamic or NA response to handgrip before or after the AM infusion. There was also no change in DHPG cardiac spillover during AM infusion. Acute IV AM did not exert a sympatholytic action in humans, with no attenuation in hemodynamic or NA response to handgrip or increase in DHPG production, despite producing an electrophysiologic response.

Amiodarone is currently one of the most effective anti-arrhythmic agents in the prevention and treatment of ventricular tachyarrhythmias. ¹ This pharmacologically complex drug has a number of modes of action that may contribute to its anti-arrhythmic properties. These include acting as a potassium channel blocker, a sodium and calcium channel blocker, as well as a nonselective β- and α-adrenoreceptor antagonist. ²

In 1995 Du et al. ³ reported a novel action of amiodarone in perfused rat heart preparations. They showed that amiodarone had a sympatholytic action in that it appeared to modulate the sympathetic nervous system by promoting metabolism of noradrenaline to its inactive metabolite dihydroxyphenylglycol (DHPG). This resulted in decreased release of noradrenaline in response to nerve stimulation. It was subsequently demonstrated that N-desethylamiodarone, the major bioactive metabolite of amiodarone, also exerted this action and that the sympatholytic action occurred by inhibition of the vesicular monoamine transporter on the axoplasmic storage vesicles— the mechanism of action of reserpine. ^4,5

However, it has yet to be shown whether this anti-adrenergic action of amiodarone occurs in human subjects. This was examined by measuring the cardiac spillover of DHPG as well as the biochemical and hemodynamic responses to sympathetic activation before and after acute IV amiodarone administration. Electrophysiological effects of amiodarone were also simultaneously assessed.

METHOD

Twelve healthy adult male volunteers were recruited into the study. None had a history or clinical evidence of respiratory, cardiac, hepatic, or thyroid dysfunction. None were on any medication.

The experiment was approved by the Alfred Hospital Human Research Ethics Committee, after an external and independent expert had reviewed the study. The investigation conforms to the principles outlined in the Declaration of Helsinki. All subjects gave written informed consent for participation after detailed explanation of the aims and objectives, the experimental techniques, and possible risks of the procedures.

Experimental Protocol

The experiment was conducted in a dedicated cardiac catheterization laboratory. A 12-lead electrocardiogram was recorded. Brachial artery blood pressure was monitored noninvasively (Spacelabs, Redmond, WA, U.S.A.). ³[H]NA tracer was infused through a 21G cannula inserted into a forearm vein. Coronary sinus catheterization with a Webster coronary sinus thermodilution catheter proceeded through an 8F venous sheath inserted in the medial cubital vein. Catheter placement into the coronary sinus was under fluoroscopic guidance. Two 2F quadpolar electrodes were introduced through a 7F sheath in the right femoral vein. The electrodes were placed to record high right atrium and bundle of His activity. A 3F artery catheter was then placed in the right femoral artery.

Baseline blood pressure, heart rate, and coronary sinus blood flow were recorded and venous and artery samples obtained. Volunteers then performed handgrip for 10 min on a Jamar hand dynamometer (Jackson, MI, U.S.A.), maintaining a grip of 20% of their maximum. Coronary sinus blood flow, venous and arterial blood samples, blood pressure, and heart rate measurements were obtained after 5 and 10 min of handgrip. The volunteer was then allowed 30 min to rest and recover to baseline blood pressure and heart rate.

Prior to starting the amiodarone infusion, coronary sinus blood flow, venous and arterial blood samples, blood pressure and heart rate recordings, and A-H and H-V intervals were obtained. The heart was then paced at 100 beats/min for 5 min and A-H and H-V intervals again recorded. Amiodarone infusion was then commenced at 300 mg/h. Volunteers rested quietly during this time. Coronary sinus blood flow, venous and artery samples, blood pressure, and heart rate recordings were obtained at regular intervals (10, 20, 30, 40, and 50 min) during the amiodarone infusion. At 40 min, A-H and H-V intervals were measured before and after atrial pacing (100 beats/min). At 50 min handgrip was recommenced for 10 min at the same kilogram force as in the first part of the experiment. Coronary sinus blood flow, venous and arterial blood samples, blood pressure, and heart rate measurements were obtained at 5 and 10 min of handgrip. At the end of the experiment all catheters were removed, and the volunteer was allowed home after a 2-h observation period.

All blood samples were placed immediately on ice. Samples were then centrifuged for 10 min at 3,000 rpm at 4°C and the plasma removed and stored at −80°C until analyzed by high-performance liquid chromatography.

Measurement of Intervals

Intervals were recorded as previously described. ⁶ The A-H interval was measured from the initial rapid deflection of the A electrogram to the initial deflection of the His spike on the electrogram (normal range 60–120 ms). The H-V interval was measured from the initial His deflection to the earliest onset of ventricular activation (normal range 35–55 ms).

Analysis of Plasma Catecholamines

Catecholamines (noradrenaline, DHPG) were extracted from 1 ml plasma with alumina adsorption, separated by liquid chromatography according to previously described methods, and the amounts quantified by electrochemical detection. ^7,8 Fractions of the eluant leaving the electrochemical cell were collected into scintillation vials for the measurement of ³[H]-labeled catecholamines by liquid scintillation spectroscopy.

Noradrenaline and Dihydroxyphenylglycol Spillover Measurements

Cardiac noradrenaline spillover was estimated according to the established formula ^9–12:EQUATION

where NA_v and NA_a are the coronary venous plasma and arterial plasma noradrenaline concentrations, respectively.

The fractional cardiac extraction of ³[H]NA was calculated by:EQUATION

where 3H-NA_a and 3H-NA_v are the concentrations of 3H-NA in arterial plasma and coronary venous plasma, respectively.

Plasma flow is from the formula:EQUATION

where CSBF is the coronary sinus blood flow and Hct the hematocrit.

Cardiac spillover of DHPG was estimated by the formula:EQUATION

where DHPG_a is the concentration of endogenous or tritiated labeled DHPG in arterial plasma, and DHPG_v is that in coronary venous plasma.

Statistics

All data are presented as mean ± SEM. SPSS (Chicago, IL, U.S.A.) statistical software package was used. Statistical comparisons were made with a t test for paired data. Where multiple groups were compared an analysis of variance was used. A value of P < 0.05 was considered statistically significant.

Materials

Materials included 8F coronary sinus sheath (Cordis, Miami, FL, U.S.A.), 3F artery catheter (Cook Australia, Brisbane, Queensland), 7F Triple Lumen guide with super arrow flex sheath (Arrow International, Reading, PA, U.S.A.), 2F Quadpolar electrode catheter (Arrow International), Webster CCS 7/8U 90A coronary sinus thermodilution catheter (Webster Laboratories), Cardiolab Electrophysiology Software (Prucka Engineering, Milwaukee, WI, U.S.A.), EP-2 Clinical Stimulator (Digital Cardiovascular Instruments), Hemodynamic monitoring system, Model 90651 (Spacelabs), flow monitoring system Model NT9500 (Astro-Med), Hand Dynamometer 5030J1 (Jamar), tritium-labeled noradrenaline (0.8 m curie/min, levo–[7–³H] noradrenaline specific activity 12–20 Curie/mmol, New England Nuclear, PerkinElmer, Boston, MA, U.S.A.), and Cordarone X injection (amiodarone hydrochloride, Sanofi-Winthrop, New South Wales, Australia)1.

RESULTS

Subject Data

The average age of volunteers was 40.2 years (range 21–71), height was 1.77 ± 0.02 m, weight was 82.95 ± 3.4 kg, and body mass index 26.4 ± 1.2 kg/m². The total dose of amiodarone administered was 300 mg per patient, giving an average of 3.68 ± 0.2 mg/kg body weight. Resting heart rate and systolic and diastolic blood pressures were 63 ± 2 beats/min, 141 ± 4 mm Hg, and 74 ± 2 mm Hg, respectively.

Electrophysiological Effects of Amiodarone

With infusion of amiodarone there was a prolongation of both the A-H and H-V values. All the values presented here are at an atrial stimulated rate of 100 beats/min. The A-H interval rose from 95.5 ± 20 ms to 107.8 ± 20 ms, and the H-V interval increase was from 54.8 ± 3 ms to 62.6 ± 2 ms. This represented a significant increase of 15 and 16% (P < 0.02) in the A-H and H-V intervals, respectively, following 40 min of amiodarone infusion. There was a similar increase noted in the A-H and H-V intervals at resting heart rates (data not shown).

Catecholamine and Hemodynamic Changes During IV Amiodarone Infusion

There was no change in the noradrenaline or DHPG plasma arterial or coronary sinus venous DHPG concentrations during the amiodarone infusion—prior to handgrip—measured at 10-min intervals for 50 min. There was a small nonsignificant rise in the DHPG cardiac spillover during this time (Fig. 1), but no change in the noradrenaline cardiac spillover. Heart rate and systolic and diastolic blood pressure were measured every 10 min during the amiodarone infusion. There were no significant changes to any of these variables during this period (with average values of 67 ± 3 beats/min, 140 ± 4 mm Hg, 74 ± 3 mm Hg, respectively, average during the infusion, P = NS).

FIGURE 1.: Change in cardiac dihydroxyphenylglycol (DHPG) spillover during amiodarone infusion.

Hemodynamic Response to Handgrip Before and After Amiodarone Infusion

Following 10 min of handgrip, there was a significant increase in heart rate, systolic blood pressure, and diastolic blood pressure both before and after amiodarone infusion. Before amiodarone, the heart rate rose 35 ± 10% (P < 0.01), the systolic blood pressure rose 27 ± 3% (P < 0.01), and the diastolic blood pressure rose 26 ± 6% (P < 0.01). After amiodarone there was a 31 ± 7% change in heart rate (P < 0.01) and 28 ± 3% and 29 ± 3% changes in systolic and diastolic blood pressure, respectively (P < 0.01 for both). However, there was no difference in the increase in hemodynamic parameters between each group (Table 1).

TABLE 1: Hemodynamic changes with handgrip pre- and post amiodarone infusion

Correlation Between Hemodynamic Changes and Cardiac Noradrenaline Spillover During Handgrip

After 10 min of handgrip, heart rate and systolic and diastolic blood pressure increased. There was also an increase in cardiac noradrenaline spillover. The increase in cardiac noradrenaline spillover correlated with both increases in heart rate (r² = 0.87) and in systolic blood pressure (r² = 0.87).

Catecholamine and Coronary Sinus Blood Flow Response to Handgrip Before and After Amiodarone Infusion

Coronary sinus blood flow, cardiac noradrenaline spillover, and plasma arterial noradrenaline concentrations all increased significantly after 10 min of handgrip (146 ± 18 ml/min to 281 ± 60 ml/min, 11.9 ± 4 ng/min to 44.3 ± 13 ng/min, and 234 ± 35 pg/ml to 569 ± 145 pg/ml, respectively). Following amiodarone infusion, there was no dampening of the autonomic and hemodynamic response, with increase of coronary sinus blood flow from 166 ± 62 ml/min to 347 ± 103 ml/min, an increase of cardiac noradrenaline spillover from 17.3 ± 4 ng/min to 55.5 ± 10 ng/min, and an increase of plasma arterial noradrenaline concentrations from 299 ± 72 pg/ml to 735 ± 121 pg/ml (Table 2). This represents increases from baseline of 99 ± 26%, 193 ± 95%, and 153 ± 58% before amiodarone and 102 ± 24%, 256 ± 48%, and 166 ± 62% after amiodarone infusion.

TABLE 2: Catecholamine response to handgrip pre- and post amiodarone infusion

DISCUSSION

Previous studies in experimental animals have revealed that acute amiodarone treatment exerts a sympatholytic effect, similar to reserpine, and manifested by rapid loss of noradrenaline from neuronal storage vesicles with subsequent deamination to its metabolite DHPG. ^3–5 The current study has failed to find evidence of such a mechanism in humans and has demonstrated a dissociation between electrophysiological and anti-sympathetic effects of amiodarone.

Amiodarone is an effective ^13–19 though pharmacologically complex anti-arrhythmic agent. Acute effects include blockade of both rested and inactivated inward sodium channels ²⁰ and of the L-type inward calcium current (i_Ca). ² Outward currents including the delayed rectifier potassium current, both rapid and slowly activating components (I_K, I_Kr, I_Ks), the muscarinic acetylcholine receptor-operated potassium current (I_K,ACh), and the sodium-activated potassium currents (I_K,Na) are all suppressed by amiodarone at therapeutic levels. ²¹ There is also evidence of noncompetitive antagonism of both the α and β adrenoreceptors. ²² Bacq et al. ²³ showed that amiodarone inhibited the release of transmitter from isolated blood-perfused cat spleens following nerve stimulation. They suggested that this was not due to an increase in uptake or selective stimulation of presynaptic α receptor but most probably represented a neurone-blocking action of the drug similar to guanethidine or bretylium. They dismissed changes of metabolism or catecholamine depletion as a mechanism for reduced transmitter overflow. However, Du et al. ³ demonstrated a “reserpine-like” sympatholytic action of amiodarone. There was an increase in DHPG efflux following perfusion of rat hearts with amiodarone; in addition there was a decrease in noradrenaline release following sympathetic nerve stimulation. In humans, there is evidence of an anti-adrenergic effect of chronic amiodarone usage. In a cross-sectional study, Kaye et al. ²⁴ detected a significantly lower cardiac noradrenaline spillover rate in heart failure patients on amiodarone therapy compared with those without the drug. However, there have not been any data, to our knowledge, on the acute effects of IV amiodarone on the human sympathetic system. In addition, it is not known whether such sympatholytic properties (or other anti-sympathetic actions) might contribute to the electrophysiologic effects of amiodarone.

The atrioventricular (AV) node is innervated by both sympathetic and parasympathetic neurones. Catecholamines increase impulse conduction through the AV node. The fibers in the upper and midregion of the AV node have low maximum diastolic potentials and action potentials with a slow upstroke and a low amplitude. Catecholamines increase the rate of rise of phase 0 depolarization and increase the upstroke velocity of the action potentials. Cells in the lower region of the AV node have higher maximum diastolic potentials and action potentials with higher upstroke velocities and amplitudes. These fibers are not affected by catecholamines. ²⁵ The A-H interval represents the conduction time through the upper fibers. This interval is markedly influenced by changes in autonomic tone. Wallace and Sarnoff ²⁶ and Rae ²⁷ demonstrated that cardiac sympathetic nerve stimulation resulted in shortening of the A-H interval, consistent with the observations here. Parasympathetic stimulation results in lengthening of the A-H interval. The H-V interval is a measure of infranodal conduction, involving the lower AV node fibers, conduction through the bundle of His, the bundle branches, and the terminal Purkinje system. There does not appear to be any change in this interval with sympathetic stimulation; i.e., intraventricular conduction is unaffected by the autonomic system. ²⁷

The acute electrophysiological effects of IV amiodarone are an increase in the A-H interval with a modest increase in the H-V interval. ^21,28–31 There is no significant effect on the corrected QT intervals or QRS duration. The increase in the A-H interval with amiodarone may either be due to its class III, class IV action, or to an anti-adrenergic effect. The change in H-V interval increase would be consistent with class I and class III effects of amiodarone. However, because we see no effect on change in heart rate or change in blood pressure (with handgrip) but do see an electrophysiological effect, this suggests that amiodarone's effect on A-V conduction is not mediated by sympathetic actions.

The electrophysiological effects of amiodarone observed in the current study were a significant prolongation of both the A-H and H-V intervals. The increase in A-H interval, of approximately 15%, was similar to that observed in previous acute studies with amiodarone, despite differences in dosing regimens used. ^6,31,32 This increase was seen in both native (unpaced) and paced studies. In contrast to previous studies we also found an increase in the H-V interval. This may have resulted from a more prolonged infusion rather than the 1–5 min used previously. ^21,28–31 Despite these electrophysiologic effects, there was no evidence of noradrenaline leakage from neuronal vesicles during the period of the infusion (as evidenced by DHPG spillover) nor by impairment in sympathetic responsiveness, the latter being determined by both noradrenaline kinetics and hemodynamic responses (which themselves were closely correlated).

Previous studies have indicated that the solvent for IV amiodarone (polysorbate 80) does have electrophysiological properties on canine myocardium ³³; thus this compound may potentiate and contribute to the electrophysiological effects of amiodarone. However, polysorbate 80 does not have sympatholytic actions when infused alone in rats as demonstrated in our earlier studies. ³

Limitations of this study include the infusion rate of amiodarone used and the study population. Had we administered amiodarone in a more rapid method (i.e., over 5 min rather than 60 min) we might have achieved higher cardiac concentrations of the drug and elicited a sympatholytic action; however, we chose a regimen that is used practically in our institution (and which did effect electrophysiologic changes in the current study). It is also possible that in unwell patients with increased sympathetic activity, sympatholytic changes may have been demonstrated more readily than in normal volunteers in whom subtle changes may have been missed, although the latter were subjected to periods of enhanced sympathetic activity (by handgrip).

In summary, acute IV administration of amiodarone to normal volunteers results in significant prolongation of both A-H and H-V intervals, both in the rested and stimulated state. These electrophysiological effects are not accompanied by biochemical evidence of a sympatholytic action nor by hemodynamic changes suggestive of an effect on the sympathetic nervous system, either at rest or with sympathetic activation.

ACKNOWLEDGMENTS

Many thanks to Dr. Archer Broughton, Leonie Johnstone, and Flora Socratous for their help during these studies and with the plasma assays.