686
Point of View
Class III Antiarrhythmic Agents Have a Lot
of Potential but a Long Way to Go
Reduced Effectiveness and Dangers of Reverse
Use Dependence
L. M. Hondeghem, MD, PhD, and D. J. Snyders, MD
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
In a perspective,' it was pointed out that available
class Ic antiarrhythmic agents2 do not discriminate
well between normal and depressed tissue and that
consequently they are expected to have a lower
therapeutic index. The approach to developing more
potent antiarrhythmic agents to be tested in patients
who do not respond to class IA or IB agents is
problematic. Indeed, it is efficacy and safety rather
than potency that matters. In fact, simple basic
scientific principles suggest that a good sodium channel blocker should interfere markedly with abnormal
conduction in tissue responsible for arrhythmias,
while leaving the normal conduction intact. Unfortunately, recent data indicate that the concerns for the
safety of class lc agents were justified.3
At present, class III antiarrhythmic agents2 are
favored increasingly to treat patients with serious
tachycardias. Although these agents could be very
powerful antiarrhythmics, our investigation suggests
that the currently available drugs have electrophysiologic features that render them relatively less effective than an ideal class III agent could be and may
even render them proarrhythmic. We review briefly
the kinetic effects on action potential duration of
current class III agents. Desirable properties for
lengthening of action potential duration are derived,
and a simple model illustrating the time- and voltagedependent actions of a hypothetical compound is
developed. Finally, we apply current knowledge of
sodium and potassium channel blockade to illustrate
how therapy combining class I and class III effects
might prove to be especially effective.
Mechanisms Instead of Classification
Class III antiarrhythmic agents are defined as
antiarrhythmic drugs that act primarily by prolonging
The opinions expressed in this article are not necessarily those
of the editors or of the American Heart Association.
From the Departments of Medicine and Pharmacology, Vanderbilt University, Nashville, Tennessee.
L.M.H. is the Stahlman Professor of Cardiovascular Medicine.
Supported by the Stahlman endowment.
Address for correspondence: Luc Hondeghem, MD, PhD, Medical Center North CC-2209, Department of Medicine, Vanderbilt
University, Nashville, TN 37232.
the action potential duration.2 Agents in this category include amiodarone,4 sotalol,5 N-acetylprocainamide (NAPA),6 melperone,7 and several
other experimental agents. It is believed that these
agents act primarily by blocking cardiac potassium
channels, thereby reducing repolarizing currents and
prolonging action potential duration.8
Amiodarone is usually classified as a class III
antiarrhythmic agent, but it also has class I (blocks
sodium channels),9'10 class II (antiadrenergic
actions)," and class IV effects (blocks calcium
channels).12 Similarly, many of the class I agents,
noticeably the class 'A agents, not only block sodium
channels but also lengthen action potential duration
(e.g., quinidine, procainamide, and disopyramide).
Moreover, for drugs with multiple mechanisms of
action, it is difficult if not impossible to indicate the
primary classification unequivocally. Quinidine, for
example, causes both cycle length-dependent widening of the QRS-complex and lengthening of action
potential duration. However, whereas the sodium
channel block is most marked at fast heart rates and
nearly nonexistent at slow heart rates,13 the prolongation of the action potential duration is most
marked at slow heart rates.14'5 As a result, quinidine
could be labeled primarily as a class III agent at slow
heart rates or as a class I agent at fast heart rates. For
this reason, it is preferable to discuss mechanisms of
drug action, rather than drug classification. Thus, we
concentrate on drugs that lengthen action potential
duration (class III effects) rather than on class III
agents.
Although lengthening of action potential duration
can also be achieved by increasing inward currents,
we limit our considerations to block of outward
currents. Furthermore, although the heart has
numerous potassium channels, we concentrate our
discussion on the delayed outward rectifier current
and use it as an example of an important determinant
of action potential duration. However, the principles
presented could be applied equally well to the other
potassium or outward currents and inward currents
as well. Indeed, class III effects merely require a
relative reduction in outward currents or increase in
Hondeghem and Snyders Class III Antiarrhythmic Agents
100
v
100
NAPA 20 ^iM
T0
* Quinidine 1 pM
U_
* Sotalol 10 pM
-
T
o
0
E
Em0
C
0)
0
687
0
50
CL
0n
0
CL
a.
0X
a.
!50-
E1
1
j
Amiodarone
0
0
200
500
Cycle Length
1000
2000
[msec]
FIGURE 1. Plot of effects of class III agents that prolong
action potential duration with increasing cycle length. Data
redrawn from references 5 (sotalol), 6 (NAPA), and 15
(quinidine) with permission from the authors.
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
inward currents. Furthermore, the latter do not
necessarily require an increase of peak inward current. For example, slowing of inactivation of the
sodium or the calcium current resulting in more
inward current toward the end of the action potential
could also lengthen the action potential duration.16,17
Thus, clearly class III effects must not be equated
with potassium channel block.
Time- and Voltage-Dependence of Drug Effects
It is now well established that the interaction of
sodium and calcium channel blockers with their
receptor is time- and voltage-dependent and can be
accounted for in terms of the modulated receptor
hypothesis.18 More recently, our laboratory demonstrated that the interaction of quinidine with the
time-dependent outward current also is modulated
by time and voltage; quinidine primarily reduces the
outward current at negative membrane potentials,
and block becomes less pronounced during
depolarization.8 Thus, this use-dependent block of
potassium channels exhibits reverse use dependence;
block increases during diastole (phase 4) and
declines during the plateau, resulting in less block
with increasing use. This contrasts with its usedependent block of sodium channels that primarily
develops during depolarization (upstroke of the
action potential) and declines during diastole.
Reverse use dependence accounts nicely for the
observation that the prolongation of action potential
duration by quinidine is most marked at slow heart
rates14'5 (Figure 1). Most antiarrhythmic agents that
lengthen action potential duration (e.g., sotalol,5
NAPA,6 and melperone7) exhibit reverse use dependence similar to that of quinidine. Thus, they prolong
action potential duration at normal heart rates, but
the magnitude of the prolongation declines as heart
rate is increased (Figure 1).
Amiodarone19 when administered chronically
forms an exception: It lengthens action potential
duration at normal and at fast heart rates to a similar
200
500
--- Hypothetical
1000
2000
Cycle Length [msecl
FIGURE 2. Plot of action potential prolongation by amiodarone is nearly independent of cycle length. Data redrawn
from reference 19 with permission from the authors. The
dashed line represents a hypothetic agent with an "ideal"
prolongation of action potential duration as a function ofcycle
length. Arrow indicates a rate of 180 beats/min.
extent (Figure 2) (i.e., its action appears to be less
time and voltage dependent). Under voltage-clamp
conditions, amiodarone, contrary to quinidine, does
not preferentially bind at negative membrane potentials. If anything, block increases during depolarization instead.20
Optimal Electrophysiologic Characteristics for
Class III Effects
As a result of reverse use dependence, many drugs
with class III properties experimentally have the least
effect during tachycardias. Conversely, during bradycardia or after a long diastolic interval (e.g., compensatory pause after an ectopic), these drugs induce
maximum prolongation of action potential duration.
Excessive lengthening of the QT interval with these
agents has been associated with the development of
torsade de pointes.15 Thus, drugs exhibiting reverse
use dependence may become proarrhythmic after
long diastolic intervals and become less effective in
lengthening action potential duration at short cycle
lengths. Despite these shortcomings, agents with
predominantly class III properties appear to be moderately clinically effective, perhaps by reducing the
diastolic window during which a tachycardia can be
initiated.
Theoretically, the efficacy of agents with class III
properties could be increased. Indeed, common
sense dictates that agents that lengthen action potential duration optimally should elicit relatively little
action potential prolongation of the normal heart
beat, but maximally prolong action potential duration
of ectopics and tachycardias. Thus, as diagrammatically illustrated in Figure 2 (dashed line), the normal
cycle-length dependence of action potential duration
could remain intact for physiologic heart rates. However, when heart rates exceed a threshold (arrow in
Figure 2), the action potential would become increasingly longer causing the termination of the tachycar-
688
Circulation Vol 81, No 2, February 1990
dia. Equally important, an agent with such an electrophysiologic profile would not lengthen excessively
the action potential duration after long diastolic
intervals and, hence, would be less prone to precipitate torsade de pointes. As for sodium channel
blockers, this action potential prolongation could be
use dependent; little or no lengthening would occur
at normal heart rates, but during a tachycardia a
beat-by-beat prolongation would develop until the
tachycardia was terminated.
100
-
a)£
Open Block
S
0
,_
0
0)
c
0
50
.0
0CL
cl
Closed Block
-:z
o -
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Model
Use-dependent block of open and closed states has
been modeled for sodium and calcium channels. To
test whether it also can occur for potassium channels
such as the delayed outward rectifier and to illustrate
the kinetic differences between blockers of rested
and open channels, we implemented a modulated
receptor18 model for potassium channel block. This
model allows us to test whether it is at least theoretically possible to have an open channel blocker with
an electrophysiologic profile that would approach the
optimal drug effect presented above.
Briefly, potassium channels were modeled to exist
in two primary states: closed (C) at negative potentials and open (0) at plateau potentials as described
in the Beeler and Reuter model.16 In reality, there
may be multiple closed or open states,21 but for the
purpose of the present computations these two primary states should be adequate. Transition between
the two states was governed by voltage-dependent
rate constants. Both states may interact with drug to
form RD and OD states (see "Appendix") using
characteristic (kR, IR, ko, and lo) rate constants.
Drug-associated channels do not conduct potassium.
We incorporated these equations into the Beeler and
Reuter action potential model16 (although the results
would be qualitatively similar if another model were
used) of the cardiac action potential and computed
drug effects on action potential duration.
Under drug-free conditions, the action potential
duration shortened as heart rate was increased. On
addition of an agent that binds primarily to the rested
state, action potential duration was increased over
the range of normal cycle lengths (600-1,000 msec).
Moreover, with increasing heart rate, the prolongation of action potential duration declined, whereas
the prolongation increased for longer cycle lengths
(see Figure 3). This behavior is very similar to that of
quinidine, sotalol, and NAPA (Figure 1).
When computing the effects on action potential
duration for an open channel blocker, drug concentrations that gave relatively little lengthening of
action potential duration at normal heart rates gave
even less prolongation at slow heart rates. However,
as a tachycardia developed, action potential duration
prolonged markedly, and it was no longer possible to
sustain a tachycardia of cycle lengths less than 400
msec (Figure 3). In other words, such tachycardia
would become self-terminating.
200
500
1000
2000
Cycle Length [msecl
FIGURE 3. Plot of opposite rate dependency ofprolongation
of action potential duration for an agent blocking open or
closed channels. Open state blocking drug exhibits usedependent prolongation of action potential duration, whereas
the closed state blocker exhibits reverse use dependence.
Combination of Potassium and Sodium
Channel Block
In the above descriptions, we have concentrated on
drug effects on action potential duration. In reality,
maximal tachycardia rate of reentry arrhythmias is
determined by conduction velocity and effective
refractory period. Although action potential duration
is an important determinant of the effective refractory period, availability of sodium current is an
additional important parameter. Because recovery
from block for sodium channels occurs primarily at
negative potentials,'8 recovery mainly starts after
repolarization. Hence, its contribution to lengthening
of refractoriness is added to that resulting from
prolongation of action potential duration. Thus, prolongation of action potential duration and sodium
channel block that by themselves could not adequately prolong refractoriness to prevent or terminate a tachycardia (e.g., because of side effects)
might in combination succeed to accomplish this
therapeutic goal.
The combination of a drug that exhibits usedependent block of sodium channels with fast diastolic recovery from block (class IB) and one that
exhibits normal use-dependent (not reversed) prolongation of action potential duration (not yet available) may markedly extend refractoriness in a usedependent fashion. However, conduction of the
normal heart beat and its duration would be left
relatively undisturbed. This condition is somewhat
approximated by amiodarone: It has class IB (fast
recovery from block during diastole) sodium channel-blocking properties10 and can sustain its action
potential prolongation effects at fast heart rates.19
Moreover, the fact that prolongation of the action
potential duration is not increased much by bradycardia probably accounts for the fact that torsade de
pointes occurs relatively rarely with this agent.22 We
anticipate that should its class III effects be aug-
Hondeghem and Snyders Class III Antiarrhythmic Agents
mented by fast heart rates, it would be even more
effective against tachycardias.
689
axl
R
O 0
Summary
With regard to currently available class III agents,
although their class III effect may reduce the likelihood of tachycardia initiation, their reverse usedependent prolongation of action potential duration
reduces their effectiveness during tachycardias and
may even render them proarrhythmic, especially
after long diastolic intervals. In contrast, agents that
exhibit normal use-dependent prolongation of refractoriness hold great promise: While having relatively
less effects on the normal heart beat, they could
induce self-termination of a tachycardia. Prolongation of refractoriness can be achieved by lengthening
of action potential duration and delaying recovery of
excitability. Combination of these drug actions may
yield important clinical applications.
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Acknowledgments
The authors thank Drs. Balser, Bennett, Echt, Murray, and Roden for their helpful suggestions and Mrs.
Hondeghem and Ms. Barrett for secretarial assistance.
Appendix
The Beeler-Reuter model16 was used to calculate
cardiac ventricular action potentials, which were
integrated using an improved Euler scheme. The
action potentials were initiated by a 2-msec "stimulation" pulse of 20 gA/cm2 at the frequency of the
various trains. For all trains, the initial priming values
for membrane potential, kinetic parameters, and
[Ca2+]i were set to their end-diastolic values for a
steady-state 1-Hz train of action potentials, either in
control or for the various drug modeling runs. Therefore, the calculations simulate the effect of an abrupt
change in cardiac rate from the 1-Hz steady state in
control or drug. Drug effects were modeled on the
delayed rectifier (called L1x in the Beeler-Reuter
model). Rested state block was modeled as occurring
through a lipophilic pathway23:
ko[D] 4| 10
OD
with the following voltage-dependent rate constants:
ko=ko(0)texp(6 zFE/2RT)
lo=lo(0)texp(- 8 zFE/2RT)
ko(O)=50/M/msec
lo(0)=0.0001/msec
6=0.5
[D] =2 ptM
R, T, F, and z have their usual meaning, and 8
represents the fractional electrical distance (the fraction of the membrane field seen by the charged drug
at the receptor site on the open channel; the arbitrary
value 6=0.5 assumes that the binding site is halfway
into the electrical field). Action potential duration
was determined as the time interval between
upstroke and repolarization to -75 mV, which for
this model approximates 90% repolarization.
References
1. Hondeghem LM: Antiarrhythmic agents: Modulated receptor
applications. Circulation 1987;75:514-520
[D]=3 ,uM
2. Vaughan Williams EM: Classification of antiarrhythmic drugs,
in Sandoe E, Flensted-Jensen E, Olesen KH (eds): Symposium
on Cardiac Arrhythmias. Elsinore, Denmark, Astra, 1970, pp
449-472
3. Suppression Trial (CAST) Investigators: Preliminary report:
Effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. The Cardiac Arrhythmia. NEnglJMed 1989;321:406-412
4. Singh BN, Vaughan Williams EM: The effect of amiodarone,
a new anti-anginal drug, on cardiac muscle. Br J Pharnacol
1970;39:657-667
5. Strauss HC, Bigger JT, Hoffman BF: Electrophysiological and
beta-receptor blocking effects of MJ 1999 on dog and rabbit
cardiac tissue. Circ Res 1970;26:661-678
6. Dangman KH, Miura DS: Effects of therapeutic concentrations of procainamide on transmembrane action potentials of
normal and infarct zone Purkinje fibers and ventricular muscle
cells. J Cardiovasc Pharmacol 1989;13:846-852
7. Refsum H, Amlie JP, Platou ES, Owren T, Landmark K:
Electrophysiological effects of melperone in the dog heart in
situ-A new antiarrhythmic drug. Cardiovasc Res 1981;
15:131-136
8. Roden DM, Bennett PB, Snyders DJ, Balser JR, Hondeghem
LM: Quinidine delays IK activation in guinea pig ventricular
myocytes. Circ Res 1988;62:1055-1058
9. Mason JW, Hondeghem LM, Katzung BG: Block of inactivated sodium channels and of depolarization-induced automaticity in guinea pig papillary muscle by amiodarone. Circ Res
Open (activated) channel block was modeled as
occurring through a hydrophilic pathway, where the
drug diffused into the channel23:
10. Follmer C, Aomine, Yeh J, Singer D: Amiodarone-induced
block of sodium current in isolated cardiac cells. J Pharmacol
Exp Ther 1987;243:187-194
axl
0
R
P3x
kR[D] IR
RD
with
kR=50/M/msec
1R=0.00002/msec
1984;55:277-287
690
Circulation Vol 81, No 2, February 1990
11. Polster P, Broeckhuysen J: The adrenergic antagonism of
amiodarone. Biochem Pharmacol 1976;25:133-140
12. Yabek S, Kato R, Singh B: Acute electrophysiological effects
of amiodarone and desethylamiodarone in isolated cardiac
muscle. J Cardiovasc Pharmacol 1986;8:197-207
13. Hondeghem LM, Katzung BG: Test of a model of antiarrhythmic drug action: Effects of quinidine and lidocaine on myocardial conduction. Circulation 1980;61:1217-1224
14. West TC, Amory DW: Single fiber recording of the effects of
quinidine at atrial and pacemaker sites in the isolated right
atrium of the rabbit. J Phannacol Exp Ther 1960;130:183-193
15. Roden DM, Hoffman BF: Action potential prolongation and
induction of abnormal automaticity by low quinidine concentrations in canine purkinje fibers. Circ Res 1985;56:857-867
16. Beeler GW, Reuter H: Reconstruction of the action potential
of ventricular myocardial fibers. J Physiol 1977;268:177-210
17. Platou ES, Refsum H, Hotvedt R: Class III antiarrhythmic
action linked with positive inotropy: Antiarrhythmic, electrophysiological, and hemodynamic effects of the sea-anemone
polypeptide ATX II in the dog heart in situ. J Cardiovasc
Pharmacol 1986;8:459-465
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
(Circulation 1990;81:686-690)
18. Hondeghem LM, Katzung BG: Time- and voltage-dependent
interactions of antiarrhythmic drugs with cardiac sodium
channels. Biochim Biophys Acta 1977;472:377-398
19. Anderson KP, Walker R, Dustman T, Lux RL, Ershler PR,
Kates RE, Urie PM: Rate-related electrophysiologic effects of
long-term administration of amiodarone on canine ventricular
myocardium in vivo. Circulation 1989;79:948-958
20. Balser JR, Hondeghem LM, Roden DM: Amiodarone reduces
time dependent IK activation. Circulation 1987;75(Suppl
IV):IV-151
21. Bennett PB, McKinney L, Kass RS, Begenisich T: Delayed
rectification in the calf cardiac purkinje fiber: Evidence for
multiple state kinetics. Biophys J 1985;48:553-567
22. Roden DM: Arrhythmogenic potential of class III antiarrhythmic agents: Comparison with Class I agents, in Singh BN (ed):
Control of Cardiac Arrhythmias by Lengthening Repolarization.
Futura Book Publishing Co, Mount Kisco, NY, 1988, pp
559-576
23. Hille B: Local anesthetics: Hydrophylic and hydrophobic
pathways for the drug-receptor reaction. J Gen Physiol 1977;
69:497-515
Class III antiarrhythmic agents have a lot of potential but a long way to go. Reduced
effectiveness and dangers of reverse use dependence.
L M Hondeghem and D J Snyders
Downloaded from http://circ.ahajournals.org/ by guest on July 12, 2017
Circulation. 1990;81:686-690
doi: 10.1161/01.CIR.81.2.686
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1990 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/81/2/686
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/