- Email: [email protected]
Cardiac electrophysiological effects of rilmenidine, a novel antihypertensive agent, in the conscious dog : Comparison with clonidine
Life Sciences, Vol. 54, No. 22, pp. PL 381-387, 1994 Copyright © 1994 Elsevier Science Lid Printed in the USA. All rights resereed 0024-3205/94 $6.00 + .00
Pergamon
0024-3205-(94) E0027-O PHARMACOLOGY LETTERS Accelerated Communication
C A R D I A C E L E C T R O P H Y S I O L O G I C A L E F F E C T S OF R I L M E N I D I N E , A NOVEL A N T I H Y P E R T E N S I V E A G E N T , IN T H E C O N S C I O U S DOG : COMPARISON WITH CLONIDINE M. Boucher, M. Dubar*, C. Chassaing, P. Vivet* and P. DuchSne-Marullaz INSERM U. 195, Faculty of Medicine, Clermont-Ferrand ; *IRI SERVIER, Courbevoie, France. (Submitted January 26, 1994; accepted February 15, 1994; received in final form March 15, 1994) Abstract : The cardiac electrophysiological effects of rilmenidine, a novel antihypertensive agent,
and clonidine were studied in the conscious dog. Sinus rate, corrected sinus recovery time (CSRT) and Wenckebach point (WP) were measured in seven intact dogs. Atrial rate and atrial effective refractory period (AERP) were measured in six atrioventricular (AV)-blocked dogs with ventricular pacing. In both groups, blood pressure was also monitored. Each dog received with at least a threeday interval rilmenidine as dihydrogen phosphate and clonidine as hydrochloride in four successive intravenous injections, 30 min apart. In intact dogs, rilmenidine was administered at 50, 50, 100 and 200 l.tg/kg and clonidine at 2.5, 2.5, 5 and 10 I.tg/kg. In AV-blocked dogs, doses of rilmenidine were 25, 25, 50 and 100 I.tg/kg, those of clonidine 5, 5, 10 and 20 I.tg/kg. Rilmenidine and clonidine decreased sinus rate and atrial rate from the first dose. In this regard, rilmenidine was respectively 24 and 23 times less potent than clonidine. A lengthening of CSRT was observed at all doses with rilmenidine and at the last three doses with clonidine (ratio : 17) and a lowering of WP at all doses with rilmenidine and clonidine (ratio : 22). A shortening of AERP was also seen with rilmenidine and clonidine from the second dose (ratio : 6). All these effects may at least partly be explained by a cholinergic activation mechanism. In intact dogs both drugs produced a lowering of mean blood pressure (ratio : 17), whereas in AV-blocked dogs, in which ventricular rate was kept constant by pacing, pressure effects were more complex, being the resultant of hypotensive and hypertensive effects, the latter due to alpha vascular stimulation. Taken together, these results indicate that in the conscious dog, rilmenidine and clonidine exert qualitatively identical electrophysiological effects, but with different potency ratios. Key Words: rilmenidine, clonidine, sinoatrial node, atrial myocardium, AV node
Introduction
Rilmenidine, or 2-(dicyclopropylmethyl) amino-2-oxazoline, is a novel antihypertensive agent, which binds more selectively to imidazoline-preferring receptors than to alpha2-adrenoceptors (1,2) and to alpha2- than to alphal-adrenoceptors (3). These properties may account for the low incidence of central side effects observed with rilmenidine during therapeutic use (4). The blood pressure lowering effects have been demonstrated in animals (2,5) as well as in humans (4,6), and have been shown to result from a decrease in vascular sympathetic tone of both central and peripheral origins (7,8). On the other hand, relatively little is known about the electrophysiological properties of this drug (9,10). Accordingly, the purpose of the present study was to investigate and quantify the effects of rilmenidine on sinoatrial node automaticity, atrial myocardium and atrioventricular (AV) node refractoriness, and blood pressure as a function of dose in conscious dogs, in comparison with clonidine, the well-known centrally acting antihypertensive alpha2-adrenoceptor agonist, already shown to decrease atrial effective refractory period (11). Correspondence to : Michel Boucher, Physiology, Faculty of Pharmacy, F-63001 ClermontFerrand Cedex 1, France.
PL-382
Eleetrophysiological Effects of Rilmenidine
Vol. 54, No. 22, 1994
Methods This study was performed on 13 mongrel dogs of either sex weighing between 14 and 23 kg. They were housed in individual cages in a large colony room, with food and water continuously available in their cages.The study conformed to the NIH Guidelines for Care and Use of Laboratory Animals.
Surgical preparation and intrumentation. In seven dogs (intact dogs) out of the 13, two wired stainless steel electrodes were implanted, under sodium pentobarbital anesthesia and aseptic conditions, 1.5 cm apart on the external surface of the right atrium near the sinoatrial node, and the leads tunneled out through the neck. Three of these dogs were in addition fitted with a catheter for long-term measurement of blood pressure; the catheter was inserted into the left omocervical artery and connected to a valve fixed on the neck. In the other six dogs (AV-blocked dogs), AV block was induced by crushing the His bundle with forceps introduced through the open right atrium during temporary occlusion of the venae cavae [modified Fredericq's technique (12)]. Two atrial surgical electrodes (in all six dogs) and an arterial catheter (in three dogs) were implanted as above. In addition, two ventricular surgical electrodes (in all six dogs) were implanted 1.5 cm apart on the external surface of the fight ventricle near the AV junction, and the leads tunneled out through the neck. All dogs were left to recover for at least eight to ten days prior to experimentation.
Measurement. Electrocardiographic and blood pressure monitoring were carded out with a Cardiopan III T instrument (Massiot-Philips) and a Statham P23 Gb transducer connected to the arterial valve and linked up to the recorder via a pressure module. Corrected sinus recovery time (CSRT) and Wenckebach point (WP) were measured in intact dogs. Corrected sinus recovery time was measured according to the method described previously for humans by Mandel et al. (13). Corrected sinus recovery time, which corresponds to the postoverdrive pacing pause, was determined as the difference between the measured pause and the mean resting sinus cycle length. Atrial pacing was applied for 1 min using 2-ms rectangular pulses from a Hugo Sachs Electronik 6512 stimulator; the stimulation voltage was 1.5 times the threshold voltage and pacing frequency twice the spontaneous sinus rate. Wenckebach point was determined by carrying out atrial pacing (2ms rectangular pulses - threshold voltage x 1.5) and gradually increasing the pacing frequency from the spontaneous sinus rate onward until type I second degree AV block occurred. Atrial effective refractory period (AERP) was measured in AV-blocked dogs by the extrastimulus method involving single premature atrial stimuli. Atrial pacing was applied in 2-ms rectangular pulses from a Janssen programmable stimulator; the stimulation voltage was 1.5 times the threshold voltage and pacing frequency twice the spontaneous atrial rate observed before the first injection. Single premature atrial stimuli were brought closer to the preceding stimulus in 5-ms steps at every eight pacing stimuli. Throughout the experiment in the AV-blocked dogs, ventricular pacing (2-ms rectangular pulsesthreshold voltage x 1.25 - spontaneous ventricular rate x 1.25) was applied in order to avoid any interferences due to drug-induced ventricular bradycardia. During recording, the previously trained dogs were placed on a table and lightly restrained. One microcatheter was fitted in the cephalic vein before each test to allow painless drug administration. Protocol. In intact dogs, rilmenidine (dihydrogen phosphate) was administered at doses of 50, 50, 100 and 200 I.tg/kg and clonidine (hydrochloride) at doses of 2.5, 2.5, 5 and 10 ~tg/kg. In AVblocked dogs, the doses of rilmenidine were 25, 25, 50 and 100 I.tg/kg and those of clonidine 5, 5, 10 and 20 ~tg/kg. The doses of rilmenidine were chosen sufficiently high to produce net effects on all parameters from the first or the second dose onward, and those of clonidine to induce almost identical effects. The study design was strictly the same in all cases. Each dog received the four successive intravenous injections, lasting 30 s each, of each drug, 30 min apart. These experiments were carded out at least eight to ten days after surgery, by which time the dogs were thoroughly familiarized with the experimental conditions. At least 72 hours elapsed between each drug evaluation performed on the same dog in random order. Sinus rate (determined over 30 s), CSRT, WP, and mean blood pressure in intact dogs, and atrial rate (so called to differentiate it from sinus rate in intact dogs), AERP, and mean blood pressure in AV-blocked dogs were measured before the first injection and for 30 min after each injection. As previously shown in our laboratory, all these parameters remain highly stable throughout. Drugs. Rilmenidine (dihydrogen phosphate) was supplied by IRI Servier (France) and clonidine (hydrochloride) by Boehringer Ingelheim Laboratories (France). Drugs were dissolved in physiological saline and doses are expressed in terms of the salt.
Vol. 54, No. 22, 1 9 9 4
Eiectrophysiological Effects of Rilmenidine
PL-383
Statistical analysis. Results were expressed as means + SEM and also as mean maximal variations + SEM. This parameter was calculated as follows: The time by which maximal or minimal value had been attained was determined during the 30 min following each injection. The mean difference between individual values and their corresponding basal values was calculated, giving mean maximal variations + SEM. The effects of each drug on the different parameters were established using analysis of variance in complete blocks without repeated measures, followed, when the F value was significant, by multiple comparisons using Dunnett's test. Comparison of the effects of the two drugs on the same parameter was made using either effective dose values, i.e., ED10, ED40 and ED80, calculated from the corresponding regression lines when possible, or if not, areas under the curves obtained by plotting the chosen parameter against time. Results Effects on sinoatrial node automaticity. Sinus rate. In intact dogs, rilmenidine decreased sinus rate from the first dose onward (p < 0.001) (Fig.l). This bradycardic effect reached 20 + 2, 27 + 4, 35 + 6 and 42 + 5 beats/rain for cumulative doses of 50, 100, 200 and 400 lag/kg, respectively. The ED40, the dose producing a 40% decrease in sinus rate, was 151.0 + 29.6 ~tg/kg rilmenidine dihydrogen phosphate, i.e., 97.8 + 19.2 I.tg/kg rilmenidine base. Clonidine produced a decrease in sinus rate from the first dose onward (p < 0.001) (Fig.l), which reached 23 + 5, 28 + 4, 35 + 5 and 46 + 4 beats/min for cumulative doses of 2.5, 5, 10 and 20 lag/kg, respectively. The ED40 was 4.8 + 1.0 I.tg/kg clonidine hydrochloride, i.e., 4.1 + 0.9 p.gJkg clonidine base, thus indicating a potency ratio of about 24 in comparison with rilmenidine. Corrected sinus recovery time. In the same dogs, rilmenidine lengthened CSRT (p < 0.01) (Fig.l), which remained almost identical at the first three doses and then increased markedly at the last dose (maximal lengthenings of 371 + 117, 334 + 47, 377 + 115 and 768 + 272 msec, respectively). Clonidine induced a lengthening of CSRT from the second dose onward (p < 0.01) (Fig. 1), which reached 152 + 81,369 + 165, 469 + 149 and 820 + 298 msec for cumulative doses of 2.5, 5, 10 and 20 I.tg/kg, respectively. Calculated from the respective areas under the curves, the potency ratio was about 17. Atrial rate. In AV-blocked dogs, the ventricular rate of which was kept at a constant value by pacing, rilmenidine decreased atrial rate from the first dose onward (p < 0.05) (Fig.2). This bradycardic effect reached 24 + 13, 40 + 12, 64 + 5 and 82 + 7 beats/min for cumulative doses of 25, 50, 100 and 200 btg/kg, respectively. The ED80, the dose producing a 80% decrease in atrial rate, was 221.2 + 7.6 [tg/kg rilmenidine dihydrogen phosphate, i.e., 143.3 + 4.9 ~tg/kg rilmenidine base. Clonidine produced a decrease in atrial rate from the first dose onward (p <0.001)(Fig.2) (maximal decreases of 78 + 9, 89 + 6 and 91 + 8 beats/min for cumulative doses of 5, 10 and 20 I.tg/kg, respectively). During analysis of the results it was deemed preferable not to take into account the highest dose (40 I.tg/kg) since at this dose the atrial pacing carried out for the determination of the AERP frequently triggered (in four dogs out of six) atrial fibrillation of varying duration. The ED80 was 7.3 + 1.9 ~g/kg clonidine hydrochloride, i.e., 6.3 + 1.6 I.tg/kg clonidine base, thus indicating a potency ratio of about 23. Effects on atrial m y o c a r d i u m refractoriness. Atrial refractoriness was assessed by measuring AERP by the extrastimulus method in AV-blocked dogs. Rilmenidine shortened AERP from the second dose onward (p < 0.01) (Fig.2) (maximal shortenings of 3 + 2, 8 + 4, 15 + 3 and 21 + 6 msec for cumulative doses of 25, 50, 100 and 200 btg/kg, respectively). The ED10, dose decreasing AERP by 10%, was 91.3 + 2.6 I.tg/kg rilmenidine dihydrogen phosphate, i.e., 59.1 + 1.7 I.tg/kg rilmenidine base.
PL-384
Electrophysiological Effects of Rilmenidlne
Vol. 54, No. 22, 1994
Clonidine produced a shortening of AERP from the second dose onward (p < 0.01) (Fig.2), which reached 7 + 4, 12 + 5 and 19 + 7 msec for cumulative doses of 5, 10 and 20 p.g/kg, respectively. The ED10 was 11.7 + 0.7 ~tg/kg clonidine hydrochloride, i.e., 10.1 + 0.6 ~tg/kg clonidine base, thus indicating a potency ratio of about 6. e--e RILMENIDINE 5 0 * 5 0 4 1 0 0 ~ ' 2 0 O p g / k g CLONIDINE 2.5.~2.545. l O p g / k g go~
T_
,:
•
•
*
,
E
70
a
~0
~=-~-~,,
*, ,: *• .: =:
,
u) 2100
|
1900
•
:
:
/ ~
~1700 1500
~13oo 1100 900
:
?00
[L
:
50
~
: ~
~
=:
z
20
30
30
80
.o
*
f O
*
:
,
,t
,
10
20
300
10
*
*
.
*
~
.t
•
*
*
f ~ ' . " f t : ! ~
20 300 10 TIME(mln)
20
300
10
FIG. 1 Time courses of sinus rate (SR), corrected sinus recovery time (CSRT), Wenckebach point (WP), and mean blood pressure (MBP) in conscious intact dogs after four successive intravenous injections (arrows) of 50, 50, 100 and 200 ~tg/kg rilmenidine dihydrogen phosphate (O) and of 2.5, 2.5, 5 and 10 I.tg/kg clonidine hydrochloride (O). Values are means for groups of seven dogs, except for mean blood pressure (only three dogs). Vertical lines show SEM. Significance : * p < 0.05, ** p < 0.01, *** p < 0.001 in comparison with corresponding control values. Effects
on
AV
node
refractoriness.
Atrioventricular nodal refractoriness was assessed by measuring WP during right atrial pacing in intact dogs. Rilmenidine lowered WP from the first dose onward (p < 0.01) (Fig.l), reflecting an increase in AV nodal refractoriness. This effect reached 12 +_ 4, 25 + 7, 40 +_ 8 and 58 + 10 beats/min for cumulative doses of 50, 100, 200 and 400 I.tg/kg, respectively. The ED40, the dose lowering WP by 40%, was 225.1 + 40.6 I.tg/kg rilmenidine dihydrogen phosphate, i.e., 145.8 + 26.3 I.tg/kg rilmenidine base.
Vol. 54, No. 22, 1 9 9 4
Electrophysiological Effects of Rilmenidine
PL-385
Clonidine lowered WP from the first dose onward (p < 0.01) (Fig. 1) (maximal lowerings of 16 _+6, 31 _+6, 43 _+6 and 58 -+ 6 beats/min for cumulative doses of 2.5, 5, 10 and 20 I.tg/kg, respectively). The ED40 was 7.6 -+ 1.5 I.tg/kg clonidine hydrochloride, i.e., 6.6 _+ 1.3 p.gJkg clonidine base, thus indicating a potency ratio of about 22. e~o RILMENIDINE 2 5 + 2 5 + 5 0 ÷ l O O I l g / k g ~ CLONIDINE 5 * 5 4 1 0 * 2 0 v g / k g 110 90 , , ; !
70
* *
:
: , : , * * , : *
50
30 10
140
,,: laO
:
I/d '< 1 2 0
^ 12°r
:
: *
*
::,
*:
I*:
-,oo
80[ ,o
.:
tO
: 1()
!
. 20
t
300
1()
.
:
t
2() 300 1() TIME(mln)
2()
.
t
300
.
::" 10
20
30
FIG.2 Time courses of atrial rate (AR), atrial effective refractory period (AERP), and mean blood pressure (MBP) in conscious ventricularly paced AV-blocked dogs after four successive intravenous injections (arrows) of 25, 25, 50 and 100 lag/kg rilmenidine dihydrogen phosphate ( e ) and of 5, 5, 10 and 20 p.g/kg clonidine hydrochloride (O). Values are means for groups of six dogs, except for mean blood pressure (only three dogs). Vertical lines show SEM. Significance: * p < 0.05, ** p < 0.01, , *** p < 0.001 in comparison with corresponding control values. Effects
on
mean
blood
pressure.
In the two experimental series, rilmenidine lowered mean blood pressure from the cumulative dose of 100 ~g/kg (p < 0.05) (Figs.1 and 2). In intact dogs, maximal lowerings were 5 + 2, 9 + 1, 16 + 2 and 21 + 4 m m H g for cumulative doses of 50, 100, 200 and 400 I.tg/kg, respectively, and in AVblocked dogs, they were 3 + 2, 8 + 3, 10 + 5 and 18 + 6 mmHg for cumulative doses of 25, 50, 100 and 200 I.tg/kg, respectively. Clonidine produced a lowering o f mean blood pressure from the first dose onward (p < 0.001) (Figs. 1 and 2), which remained almost identical at all doses (maximal lowerings of 15 + 3, 12 + 4, 19 + 5 and 17 + 4 m m H g in intact dogs and of 18 + 2, 17 + 4 and 15 + 2 m m H g in AV-blocked dogs). Calculated from the respective areas under the curves, potency ratios were about 17 in the two sets of experimental conditions, but this figure is very likely underestimated, inasmuch as an initial hypertensive effect due to alpha vascular stimulation, which buffers the hypotensive effect, was observed with clonidine and much less markedly with rilmenidine.
PL-386
Electrophysiologieal Effects of Rilmenidine
Vol. 54, No. 22, 1994
In the conscious dog, rilmenidine at cumulative intravenous doses of between 25 and 400 p.g/kg decreased sinus heart rate. This result is consistent with previous studies showing heart rate slowing after intravenous rilmenidine in animals (5,7). In this regard, our study shows rilmenidine to be 2324 times less potent than clonidine. As evidenced by literature data, the clonidine-induced bradycardia results from a decrease in cardiac adrenergic activity, both through a central and a peripheral presynaptic mechanism (14,15), and from an enhancement of cardiac cholinergic activity (14,16). As regards rilmenidine, the same mechanisms have been postulated to account for its bradycardic effects (7). In intact dogs, rilmenidine lowered mean blood pressure with a potency 17 times less than clonidine. In ventricularly paced AV-blocked dogs this potency ratio was more difficult to assess since pressure effects were more complex, being the resultant of hypotensive effects and hypertensive ones due to alpha vascular stimulation. In this regard, rilmenidine was less effective than clonidine in producing initial hypertensive effect, very likely due to its lower affinity for alphal- and alpha2-adrenoceptors (3). This hypotensive effect, already well-documented in different species and in various experimental conditions (2,4-6), results from a decrease in cardiac sympathetic activity from both central and peripheral origins (7,8). Rilmenidine simultaneously lengthened CSRT at all doses used, i.e., between 50 and 400 I.tg/kg, with a potency ratio of 1 : 17 in comparison with clonidine. Two studies in patients with and without essential hypertension failed to show any effect of rilmenidine on CSRT (9,10), but, and this is very likely the reason for this discrepancy, the doses used were markedly lower than ours (50 btg base/kg per os and 1 mg dihydrogen phosphate/patient intravenously, respectively) and thus produced no effect on heart rate either. Rilmenidine was also shown to lower WP at doses of between 50 and 400 t.tg/kg with a potency ratio of 1 : 22 in comparison with clonidine. As for CSRT and very likely for the same reasons, no significant effect was seen either on WP in patients with essential hypertension given per os 50 I.tg base/kg rilmenidine (9) or on AH interval in patients without hypertension given intravenously 1 mg rilmenidine (10). The same above-mentioned mechanisms as for bradycardia very likely account for these effects on CSRT and WP. Rilmenidine decreased AERP very transiently and by at most about 20 msec, but in this regard, it was only six times less potent than clonidine. This effect probably results from an increase in cardiac cholinergic activity, the AERP shortening effect of which is well established (17). Given the reported effect of the stimulation of the cardiac adrenergic activity on atrial refractoriness (18), an increase in AERP would have been expected in response to a fall in cardiac adrenergic tone produced by these two drugs. In fact, this latter effect may explain why the AERP shortening is relatively slight and short-lasting. As regards direct stimulation of cardiac postsynaptic alphal- and/or alpha2adrenoceptors, this possibility must be ruled out, since it would result in a lengthening of AERP (19). An identical effect has also been observed with 1 mg rilmenidine given intravenously to patients without hypertension (10) and with clonidine in chloralosed AV-blocked dogs (11), but at doses of 25 and 50 I.tg base/kg administered per os to patients with essential hypertension, rilmenidine caused no significant shortening of AERP and even produced a very slight lengthening of AERP at the lowest dose, close to that generally used in therapy (9). In conclusion, the results presented here show that in the conscious dog rilmenidine and clonidine exert, on sinoatrial node automaticity, atrial myocardium and AV node refractoriness, and blood pressure, qualitatively identical effects, but with different potency ratios.
Acknowledgmen|~ We thank Dr. O. Crambes of IRI Servier for his advice and Mme D. Hosmalin for the preparation of the manuscript.
References 1.G.BRICCA, M. D O N T E N W I L L , A. MOLINES, J. FELDMAN, E. TIBIRICA, A. BELCOURT and P. BOUSQUET, Eur.J.Pharmacol. 163 373-377 (1989). 2.R.E. GOMEZ, P. ERNSBERGER, G. FEINLAND and D.J. REIS, Eur.J.Pharmacol. 195 181191 (1991). 3.P.A. VAN ZWlETEN, Am.J.Cardiol. 61 6D-14D (1988).
Vol. 54, No. 22, 1 9 9 4
Electrophysiological Effects of Rilmenidine
PL-387
4.B. BEAU, F. MAHIEUX, M. PARAIRE, S. LAURIN, B. BRISGAND and P. VITOU, Am.J.Cardiol. 6.1.195D-102D (1988). 5.E. KOENIG-BERARD, C. TIERNEY, B. BEAU, B. DELBARRE, F. LHOSTE and C. LABRID, Am.J.Cardiol. 61 22D-31D (1988). 6.C. FIORENTINI, C. GUILLET and M. GUAZZI, Arch.Mal.Coeur 82 (V) 39-46 (1989). 7.M. LAUBIE, J.C. POIGNANT, J. SCUVEE-MOREAU, H. DABIRE, A. DRESSE and H. SCHMITT, J.Pharmacol. 16 259-278 (1985). 8.J.L. MONTASTRUC, I. MACQUIN-MAVIER, M.A. TRAN, C. DAMASE-MICHEL, E. KOENIG-BERARD and P. VALET, Am.J.Med. 87 (3C) 14S-17S (1989). 9.F. ZANNAD, E. ALLOT, J. FLORENTIN, J.P. SAULNIER and J.M. GILGENKRANTZ, Am.J.Cardiol. 61 67D-71D (1988). 10.J.L. TONET, C. GUILLET, G. JONDEAU, F. POULAIN, P. VIVET, R. FRANK and Y. GROSGOGEAT, Eur.J.Clin.Pharmacol. 41 537-540 (1991 ). l l.P. DUCHENE-MARULLAZ, J. LAVARENNE, P. LAPALUS, M. BOUCHER et C. CHASSAING, C.R.Soc.Biol. 166 601-604 (1972). 12.M. BOUCHER and P. DUCHENE-MARULLAZ, J.Pharmacol.Methods 13 95-107 (1985). 13.W.J. MANDEL, H. HAKAYAMA, R. DANZIG and H.S. MARCUS, Circulation 4__4459-66 (1971). 14.J.R. BOISSIER, J.F. GIUDICELLI, J. FICHELLE, H. SCHMITT and H. SCHMITT, Eur.J. Pharmacol. 2 333-339 (1968). 15.P. DUCHENE-MARULLAZ, J. LAVARENNE, P. LAPALUS, M. BOUCHER and Y. MONGHEAL, Eur.J.Pharmacol. 28 76-80 (1974). 16.W. HOEFKE and W. KOBINGER, Arzneimittelforschung 16 1038-1050 (1966). 17.E.N. PRYSTOWSKY, G.V. NACCARELLI, W.M. JACKMAN, R.L. RINKENBERGER, J.J. HEGER and D.P. ZIPES, Am.J.Cardiol. 51 96-100 (1983). 18.F.A. KRALIOS and C.K. MILLAR, Am.J.Physiol. 240 590-596 (1981). 19.W.C. GOVIER, N.C. MOSAL, P. WHITTINGTON and A.H. BROOM, J. Pharmacol.Exp. Ther. 154 255-263 (1966).
Pergamon
0024-3205-(94) E0027-O PHARMACOLOGY LETTERS Accelerated Communication
C A R D I A C E L E C T R O P H Y S I O L O G I C A L E F F E C T S OF R I L M E N I D I N E , A NOVEL A N T I H Y P E R T E N S I V E A G E N T , IN T H E C O N S C I O U S DOG : COMPARISON WITH CLONIDINE M. Boucher, M. Dubar*, C. Chassaing, P. Vivet* and P. DuchSne-Marullaz INSERM U. 195, Faculty of Medicine, Clermont-Ferrand ; *IRI SERVIER, Courbevoie, France. (Submitted January 26, 1994; accepted February 15, 1994; received in final form March 15, 1994) Abstract : The cardiac electrophysiological effects of rilmenidine, a novel antihypertensive agent,
and clonidine were studied in the conscious dog. Sinus rate, corrected sinus recovery time (CSRT) and Wenckebach point (WP) were measured in seven intact dogs. Atrial rate and atrial effective refractory period (AERP) were measured in six atrioventricular (AV)-blocked dogs with ventricular pacing. In both groups, blood pressure was also monitored. Each dog received with at least a threeday interval rilmenidine as dihydrogen phosphate and clonidine as hydrochloride in four successive intravenous injections, 30 min apart. In intact dogs, rilmenidine was administered at 50, 50, 100 and 200 l.tg/kg and clonidine at 2.5, 2.5, 5 and 10 I.tg/kg. In AV-blocked dogs, doses of rilmenidine were 25, 25, 50 and 100 I.tg/kg, those of clonidine 5, 5, 10 and 20 I.tg/kg. Rilmenidine and clonidine decreased sinus rate and atrial rate from the first dose. In this regard, rilmenidine was respectively 24 and 23 times less potent than clonidine. A lengthening of CSRT was observed at all doses with rilmenidine and at the last three doses with clonidine (ratio : 17) and a lowering of WP at all doses with rilmenidine and clonidine (ratio : 22). A shortening of AERP was also seen with rilmenidine and clonidine from the second dose (ratio : 6). All these effects may at least partly be explained by a cholinergic activation mechanism. In intact dogs both drugs produced a lowering of mean blood pressure (ratio : 17), whereas in AV-blocked dogs, in which ventricular rate was kept constant by pacing, pressure effects were more complex, being the resultant of hypotensive and hypertensive effects, the latter due to alpha vascular stimulation. Taken together, these results indicate that in the conscious dog, rilmenidine and clonidine exert qualitatively identical electrophysiological effects, but with different potency ratios. Key Words: rilmenidine, clonidine, sinoatrial node, atrial myocardium, AV node
Introduction
Rilmenidine, or 2-(dicyclopropylmethyl) amino-2-oxazoline, is a novel antihypertensive agent, which binds more selectively to imidazoline-preferring receptors than to alpha2-adrenoceptors (1,2) and to alpha2- than to alphal-adrenoceptors (3). These properties may account for the low incidence of central side effects observed with rilmenidine during therapeutic use (4). The blood pressure lowering effects have been demonstrated in animals (2,5) as well as in humans (4,6), and have been shown to result from a decrease in vascular sympathetic tone of both central and peripheral origins (7,8). On the other hand, relatively little is known about the electrophysiological properties of this drug (9,10). Accordingly, the purpose of the present study was to investigate and quantify the effects of rilmenidine on sinoatrial node automaticity, atrial myocardium and atrioventricular (AV) node refractoriness, and blood pressure as a function of dose in conscious dogs, in comparison with clonidine, the well-known centrally acting antihypertensive alpha2-adrenoceptor agonist, already shown to decrease atrial effective refractory period (11). Correspondence to : Michel Boucher, Physiology, Faculty of Pharmacy, F-63001 ClermontFerrand Cedex 1, France.
PL-382
Eleetrophysiological Effects of Rilmenidine
Vol. 54, No. 22, 1994
Methods This study was performed on 13 mongrel dogs of either sex weighing between 14 and 23 kg. They were housed in individual cages in a large colony room, with food and water continuously available in their cages.The study conformed to the NIH Guidelines for Care and Use of Laboratory Animals.
Surgical preparation and intrumentation. In seven dogs (intact dogs) out of the 13, two wired stainless steel electrodes were implanted, under sodium pentobarbital anesthesia and aseptic conditions, 1.5 cm apart on the external surface of the right atrium near the sinoatrial node, and the leads tunneled out through the neck. Three of these dogs were in addition fitted with a catheter for long-term measurement of blood pressure; the catheter was inserted into the left omocervical artery and connected to a valve fixed on the neck. In the other six dogs (AV-blocked dogs), AV block was induced by crushing the His bundle with forceps introduced through the open right atrium during temporary occlusion of the venae cavae [modified Fredericq's technique (12)]. Two atrial surgical electrodes (in all six dogs) and an arterial catheter (in three dogs) were implanted as above. In addition, two ventricular surgical electrodes (in all six dogs) were implanted 1.5 cm apart on the external surface of the fight ventricle near the AV junction, and the leads tunneled out through the neck. All dogs were left to recover for at least eight to ten days prior to experimentation.
Measurement. Electrocardiographic and blood pressure monitoring were carded out with a Cardiopan III T instrument (Massiot-Philips) and a Statham P23 Gb transducer connected to the arterial valve and linked up to the recorder via a pressure module. Corrected sinus recovery time (CSRT) and Wenckebach point (WP) were measured in intact dogs. Corrected sinus recovery time was measured according to the method described previously for humans by Mandel et al. (13). Corrected sinus recovery time, which corresponds to the postoverdrive pacing pause, was determined as the difference between the measured pause and the mean resting sinus cycle length. Atrial pacing was applied for 1 min using 2-ms rectangular pulses from a Hugo Sachs Electronik 6512 stimulator; the stimulation voltage was 1.5 times the threshold voltage and pacing frequency twice the spontaneous sinus rate. Wenckebach point was determined by carrying out atrial pacing (2ms rectangular pulses - threshold voltage x 1.5) and gradually increasing the pacing frequency from the spontaneous sinus rate onward until type I second degree AV block occurred. Atrial effective refractory period (AERP) was measured in AV-blocked dogs by the extrastimulus method involving single premature atrial stimuli. Atrial pacing was applied in 2-ms rectangular pulses from a Janssen programmable stimulator; the stimulation voltage was 1.5 times the threshold voltage and pacing frequency twice the spontaneous atrial rate observed before the first injection. Single premature atrial stimuli were brought closer to the preceding stimulus in 5-ms steps at every eight pacing stimuli. Throughout the experiment in the AV-blocked dogs, ventricular pacing (2-ms rectangular pulsesthreshold voltage x 1.25 - spontaneous ventricular rate x 1.25) was applied in order to avoid any interferences due to drug-induced ventricular bradycardia. During recording, the previously trained dogs were placed on a table and lightly restrained. One microcatheter was fitted in the cephalic vein before each test to allow painless drug administration. Protocol. In intact dogs, rilmenidine (dihydrogen phosphate) was administered at doses of 50, 50, 100 and 200 I.tg/kg and clonidine (hydrochloride) at doses of 2.5, 2.5, 5 and 10 ~tg/kg. In AVblocked dogs, the doses of rilmenidine were 25, 25, 50 and 100 I.tg/kg and those of clonidine 5, 5, 10 and 20 ~tg/kg. The doses of rilmenidine were chosen sufficiently high to produce net effects on all parameters from the first or the second dose onward, and those of clonidine to induce almost identical effects. The study design was strictly the same in all cases. Each dog received the four successive intravenous injections, lasting 30 s each, of each drug, 30 min apart. These experiments were carded out at least eight to ten days after surgery, by which time the dogs were thoroughly familiarized with the experimental conditions. At least 72 hours elapsed between each drug evaluation performed on the same dog in random order. Sinus rate (determined over 30 s), CSRT, WP, and mean blood pressure in intact dogs, and atrial rate (so called to differentiate it from sinus rate in intact dogs), AERP, and mean blood pressure in AV-blocked dogs were measured before the first injection and for 30 min after each injection. As previously shown in our laboratory, all these parameters remain highly stable throughout. Drugs. Rilmenidine (dihydrogen phosphate) was supplied by IRI Servier (France) and clonidine (hydrochloride) by Boehringer Ingelheim Laboratories (France). Drugs were dissolved in physiological saline and doses are expressed in terms of the salt.
Vol. 54, No. 22, 1 9 9 4
Eiectrophysiological Effects of Rilmenidine
PL-383
Statistical analysis. Results were expressed as means + SEM and also as mean maximal variations + SEM. This parameter was calculated as follows: The time by which maximal or minimal value had been attained was determined during the 30 min following each injection. The mean difference between individual values and their corresponding basal values was calculated, giving mean maximal variations + SEM. The effects of each drug on the different parameters were established using analysis of variance in complete blocks without repeated measures, followed, when the F value was significant, by multiple comparisons using Dunnett's test. Comparison of the effects of the two drugs on the same parameter was made using either effective dose values, i.e., ED10, ED40 and ED80, calculated from the corresponding regression lines when possible, or if not, areas under the curves obtained by plotting the chosen parameter against time. Results Effects on sinoatrial node automaticity. Sinus rate. In intact dogs, rilmenidine decreased sinus rate from the first dose onward (p < 0.001) (Fig.l). This bradycardic effect reached 20 + 2, 27 + 4, 35 + 6 and 42 + 5 beats/rain for cumulative doses of 50, 100, 200 and 400 lag/kg, respectively. The ED40, the dose producing a 40% decrease in sinus rate, was 151.0 + 29.6 ~tg/kg rilmenidine dihydrogen phosphate, i.e., 97.8 + 19.2 I.tg/kg rilmenidine base. Clonidine produced a decrease in sinus rate from the first dose onward (p < 0.001) (Fig.l), which reached 23 + 5, 28 + 4, 35 + 5 and 46 + 4 beats/min for cumulative doses of 2.5, 5, 10 and 20 lag/kg, respectively. The ED40 was 4.8 + 1.0 I.tg/kg clonidine hydrochloride, i.e., 4.1 + 0.9 p.gJkg clonidine base, thus indicating a potency ratio of about 24 in comparison with rilmenidine. Corrected sinus recovery time. In the same dogs, rilmenidine lengthened CSRT (p < 0.01) (Fig.l), which remained almost identical at the first three doses and then increased markedly at the last dose (maximal lengthenings of 371 + 117, 334 + 47, 377 + 115 and 768 + 272 msec, respectively). Clonidine induced a lengthening of CSRT from the second dose onward (p < 0.01) (Fig. 1), which reached 152 + 81,369 + 165, 469 + 149 and 820 + 298 msec for cumulative doses of 2.5, 5, 10 and 20 I.tg/kg, respectively. Calculated from the respective areas under the curves, the potency ratio was about 17. Atrial rate. In AV-blocked dogs, the ventricular rate of which was kept at a constant value by pacing, rilmenidine decreased atrial rate from the first dose onward (p < 0.05) (Fig.2). This bradycardic effect reached 24 + 13, 40 + 12, 64 + 5 and 82 + 7 beats/min for cumulative doses of 25, 50, 100 and 200 btg/kg, respectively. The ED80, the dose producing a 80% decrease in atrial rate, was 221.2 + 7.6 [tg/kg rilmenidine dihydrogen phosphate, i.e., 143.3 + 4.9 ~tg/kg rilmenidine base. Clonidine produced a decrease in atrial rate from the first dose onward (p <0.001)(Fig.2) (maximal decreases of 78 + 9, 89 + 6 and 91 + 8 beats/min for cumulative doses of 5, 10 and 20 I.tg/kg, respectively). During analysis of the results it was deemed preferable not to take into account the highest dose (40 I.tg/kg) since at this dose the atrial pacing carried out for the determination of the AERP frequently triggered (in four dogs out of six) atrial fibrillation of varying duration. The ED80 was 7.3 + 1.9 ~g/kg clonidine hydrochloride, i.e., 6.3 + 1.6 I.tg/kg clonidine base, thus indicating a potency ratio of about 23. Effects on atrial m y o c a r d i u m refractoriness. Atrial refractoriness was assessed by measuring AERP by the extrastimulus method in AV-blocked dogs. Rilmenidine shortened AERP from the second dose onward (p < 0.01) (Fig.2) (maximal shortenings of 3 + 2, 8 + 4, 15 + 3 and 21 + 6 msec for cumulative doses of 25, 50, 100 and 200 btg/kg, respectively). The ED10, dose decreasing AERP by 10%, was 91.3 + 2.6 I.tg/kg rilmenidine dihydrogen phosphate, i.e., 59.1 + 1.7 I.tg/kg rilmenidine base.
PL-384
Electrophysiological Effects of Rilmenidlne
Vol. 54, No. 22, 1994
Clonidine produced a shortening of AERP from the second dose onward (p < 0.01) (Fig.2), which reached 7 + 4, 12 + 5 and 19 + 7 msec for cumulative doses of 5, 10 and 20 p.g/kg, respectively. The ED10 was 11.7 + 0.7 ~tg/kg clonidine hydrochloride, i.e., 10.1 + 0.6 ~tg/kg clonidine base, thus indicating a potency ratio of about 6. e--e RILMENIDINE 5 0 * 5 0 4 1 0 0 ~ ' 2 0 O p g / k g CLONIDINE 2.5.~2.545. l O p g / k g go~
T_
,:
•
•
*
,
E
70
a
~0
~=-~-~,,
*, ,: *• .: =:
,
u) 2100
|
1900
•
:
:
/ ~
~1700 1500
~13oo 1100 900
:
?00
[L
:
50
~
: ~
~
=:
z
20
30
30
80
.o
*
f O
*
:
,
,t
,
10
20
300
10
*
*
.
*
~
.t
•
*
*
f ~ ' . " f t : ! ~
20 300 10 TIME(mln)
20
300
10
FIG. 1 Time courses of sinus rate (SR), corrected sinus recovery time (CSRT), Wenckebach point (WP), and mean blood pressure (MBP) in conscious intact dogs after four successive intravenous injections (arrows) of 50, 50, 100 and 200 ~tg/kg rilmenidine dihydrogen phosphate (O) and of 2.5, 2.5, 5 and 10 I.tg/kg clonidine hydrochloride (O). Values are means for groups of seven dogs, except for mean blood pressure (only three dogs). Vertical lines show SEM. Significance : * p < 0.05, ** p < 0.01, *** p < 0.001 in comparison with corresponding control values. Effects
on
AV
node
refractoriness.
Atrioventricular nodal refractoriness was assessed by measuring WP during right atrial pacing in intact dogs. Rilmenidine lowered WP from the first dose onward (p < 0.01) (Fig.l), reflecting an increase in AV nodal refractoriness. This effect reached 12 +_ 4, 25 + 7, 40 +_ 8 and 58 + 10 beats/min for cumulative doses of 50, 100, 200 and 400 I.tg/kg, respectively. The ED40, the dose lowering WP by 40%, was 225.1 + 40.6 I.tg/kg rilmenidine dihydrogen phosphate, i.e., 145.8 + 26.3 I.tg/kg rilmenidine base.
Vol. 54, No. 22, 1 9 9 4
Electrophysiological Effects of Rilmenidine
PL-385
Clonidine lowered WP from the first dose onward (p < 0.01) (Fig. 1) (maximal lowerings of 16 _+6, 31 _+6, 43 _+6 and 58 -+ 6 beats/min for cumulative doses of 2.5, 5, 10 and 20 I.tg/kg, respectively). The ED40 was 7.6 -+ 1.5 I.tg/kg clonidine hydrochloride, i.e., 6.6 _+ 1.3 p.gJkg clonidine base, thus indicating a potency ratio of about 22. e~o RILMENIDINE 2 5 + 2 5 + 5 0 ÷ l O O I l g / k g ~ CLONIDINE 5 * 5 4 1 0 * 2 0 v g / k g 110 90 , , ; !
70
* *
:
: , : , * * , : *
50
30 10
140
,,: laO
:
I/d '< 1 2 0
^ 12°r
:
: *
*
::,
*:
I*:
-,oo
80[ ,o
.:
tO
: 1()
!
. 20
t
300
1()
.
:
t
2() 300 1() TIME(mln)
2()
.
t
300
.
::" 10
20
30
FIG.2 Time courses of atrial rate (AR), atrial effective refractory period (AERP), and mean blood pressure (MBP) in conscious ventricularly paced AV-blocked dogs after four successive intravenous injections (arrows) of 25, 25, 50 and 100 lag/kg rilmenidine dihydrogen phosphate ( e ) and of 5, 5, 10 and 20 p.g/kg clonidine hydrochloride (O). Values are means for groups of six dogs, except for mean blood pressure (only three dogs). Vertical lines show SEM. Significance: * p < 0.05, ** p < 0.01, , *** p < 0.001 in comparison with corresponding control values. Effects
on
mean
blood
pressure.
In the two experimental series, rilmenidine lowered mean blood pressure from the cumulative dose of 100 ~g/kg (p < 0.05) (Figs.1 and 2). In intact dogs, maximal lowerings were 5 + 2, 9 + 1, 16 + 2 and 21 + 4 m m H g for cumulative doses of 50, 100, 200 and 400 I.tg/kg, respectively, and in AVblocked dogs, they were 3 + 2, 8 + 3, 10 + 5 and 18 + 6 mmHg for cumulative doses of 25, 50, 100 and 200 I.tg/kg, respectively. Clonidine produced a lowering o f mean blood pressure from the first dose onward (p < 0.001) (Figs. 1 and 2), which remained almost identical at all doses (maximal lowerings of 15 + 3, 12 + 4, 19 + 5 and 17 + 4 m m H g in intact dogs and of 18 + 2, 17 + 4 and 15 + 2 m m H g in AV-blocked dogs). Calculated from the respective areas under the curves, potency ratios were about 17 in the two sets of experimental conditions, but this figure is very likely underestimated, inasmuch as an initial hypertensive effect due to alpha vascular stimulation, which buffers the hypotensive effect, was observed with clonidine and much less markedly with rilmenidine.
PL-386
Electrophysiologieal Effects of Rilmenidine
Vol. 54, No. 22, 1994
In the conscious dog, rilmenidine at cumulative intravenous doses of between 25 and 400 p.g/kg decreased sinus heart rate. This result is consistent with previous studies showing heart rate slowing after intravenous rilmenidine in animals (5,7). In this regard, our study shows rilmenidine to be 2324 times less potent than clonidine. As evidenced by literature data, the clonidine-induced bradycardia results from a decrease in cardiac adrenergic activity, both through a central and a peripheral presynaptic mechanism (14,15), and from an enhancement of cardiac cholinergic activity (14,16). As regards rilmenidine, the same mechanisms have been postulated to account for its bradycardic effects (7). In intact dogs, rilmenidine lowered mean blood pressure with a potency 17 times less than clonidine. In ventricularly paced AV-blocked dogs this potency ratio was more difficult to assess since pressure effects were more complex, being the resultant of hypotensive effects and hypertensive ones due to alpha vascular stimulation. In this regard, rilmenidine was less effective than clonidine in producing initial hypertensive effect, very likely due to its lower affinity for alphal- and alpha2-adrenoceptors (3). This hypotensive effect, already well-documented in different species and in various experimental conditions (2,4-6), results from a decrease in cardiac sympathetic activity from both central and peripheral origins (7,8). Rilmenidine simultaneously lengthened CSRT at all doses used, i.e., between 50 and 400 I.tg/kg, with a potency ratio of 1 : 17 in comparison with clonidine. Two studies in patients with and without essential hypertension failed to show any effect of rilmenidine on CSRT (9,10), but, and this is very likely the reason for this discrepancy, the doses used were markedly lower than ours (50 btg base/kg per os and 1 mg dihydrogen phosphate/patient intravenously, respectively) and thus produced no effect on heart rate either. Rilmenidine was also shown to lower WP at doses of between 50 and 400 t.tg/kg with a potency ratio of 1 : 22 in comparison with clonidine. As for CSRT and very likely for the same reasons, no significant effect was seen either on WP in patients with essential hypertension given per os 50 I.tg base/kg rilmenidine (9) or on AH interval in patients without hypertension given intravenously 1 mg rilmenidine (10). The same above-mentioned mechanisms as for bradycardia very likely account for these effects on CSRT and WP. Rilmenidine decreased AERP very transiently and by at most about 20 msec, but in this regard, it was only six times less potent than clonidine. This effect probably results from an increase in cardiac cholinergic activity, the AERP shortening effect of which is well established (17). Given the reported effect of the stimulation of the cardiac adrenergic activity on atrial refractoriness (18), an increase in AERP would have been expected in response to a fall in cardiac adrenergic tone produced by these two drugs. In fact, this latter effect may explain why the AERP shortening is relatively slight and short-lasting. As regards direct stimulation of cardiac postsynaptic alphal- and/or alpha2adrenoceptors, this possibility must be ruled out, since it would result in a lengthening of AERP (19). An identical effect has also been observed with 1 mg rilmenidine given intravenously to patients without hypertension (10) and with clonidine in chloralosed AV-blocked dogs (11), but at doses of 25 and 50 I.tg base/kg administered per os to patients with essential hypertension, rilmenidine caused no significant shortening of AERP and even produced a very slight lengthening of AERP at the lowest dose, close to that generally used in therapy (9). In conclusion, the results presented here show that in the conscious dog rilmenidine and clonidine exert, on sinoatrial node automaticity, atrial myocardium and AV node refractoriness, and blood pressure, qualitatively identical effects, but with different potency ratios.
Acknowledgmen|~ We thank Dr. O. Crambes of IRI Servier for his advice and Mme D. Hosmalin for the preparation of the manuscript.
References 1.G.BRICCA, M. D O N T E N W I L L , A. MOLINES, J. FELDMAN, E. TIBIRICA, A. BELCOURT and P. BOUSQUET, Eur.J.Pharmacol. 163 373-377 (1989). 2.R.E. GOMEZ, P. ERNSBERGER, G. FEINLAND and D.J. REIS, Eur.J.Pharmacol. 195 181191 (1991). 3.P.A. VAN ZWlETEN, Am.J.Cardiol. 61 6D-14D (1988).
Vol. 54, No. 22, 1 9 9 4
Electrophysiological Effects of Rilmenidine
PL-387
4.B. BEAU, F. MAHIEUX, M. PARAIRE, S. LAURIN, B. BRISGAND and P. VITOU, Am.J.Cardiol. 6.1.195D-102D (1988). 5.E. KOENIG-BERARD, C. TIERNEY, B. BEAU, B. DELBARRE, F. LHOSTE and C. LABRID, Am.J.Cardiol. 61 22D-31D (1988). 6.C. FIORENTINI, C. GUILLET and M. GUAZZI, Arch.Mal.Coeur 82 (V) 39-46 (1989). 7.M. LAUBIE, J.C. POIGNANT, J. SCUVEE-MOREAU, H. DABIRE, A. DRESSE and H. SCHMITT, J.Pharmacol. 16 259-278 (1985). 8.J.L. MONTASTRUC, I. MACQUIN-MAVIER, M.A. TRAN, C. DAMASE-MICHEL, E. KOENIG-BERARD and P. VALET, Am.J.Med. 87 (3C) 14S-17S (1989). 9.F. ZANNAD, E. ALLOT, J. FLORENTIN, J.P. SAULNIER and J.M. GILGENKRANTZ, Am.J.Cardiol. 61 67D-71D (1988). 10.J.L. TONET, C. GUILLET, G. JONDEAU, F. POULAIN, P. VIVET, R. FRANK and Y. GROSGOGEAT, Eur.J.Clin.Pharmacol. 41 537-540 (1991 ). l l.P. DUCHENE-MARULLAZ, J. LAVARENNE, P. LAPALUS, M. BOUCHER et C. CHASSAING, C.R.Soc.Biol. 166 601-604 (1972). 12.M. BOUCHER and P. DUCHENE-MARULLAZ, J.Pharmacol.Methods 13 95-107 (1985). 13.W.J. MANDEL, H. HAKAYAMA, R. DANZIG and H.S. MARCUS, Circulation 4__4459-66 (1971). 14.J.R. BOISSIER, J.F. GIUDICELLI, J. FICHELLE, H. SCHMITT and H. SCHMITT, Eur.J. Pharmacol. 2 333-339 (1968). 15.P. DUCHENE-MARULLAZ, J. LAVARENNE, P. LAPALUS, M. BOUCHER and Y. MONGHEAL, Eur.J.Pharmacol. 28 76-80 (1974). 16.W. HOEFKE and W. KOBINGER, Arzneimittelforschung 16 1038-1050 (1966). 17.E.N. PRYSTOWSKY, G.V. NACCARELLI, W.M. JACKMAN, R.L. RINKENBERGER, J.J. HEGER and D.P. ZIPES, Am.J.Cardiol. 51 96-100 (1983). 18.F.A. KRALIOS and C.K. MILLAR, Am.J.Physiol. 240 590-596 (1981). 19.W.C. GOVIER, N.C. MOSAL, P. WHITTINGTON and A.H. BROOM, J. Pharmacol.Exp. Ther. 154 255-263 (1966).