European Heart Journal (1997) 18 {Supplement Q, C12-C18
Why does atrial fibrillation occur?
M. J. Janse
Department of Clinical and Experimental Cardiology, Academic Medical Center, University of Amsterdam, and the
Interuniversity Cardiology Institute, The Netherlands
A number of electrophysiological changes have been found
in isolated preparations from human atria that had been
fibrillating. Action potentials had a shorter duration and a
triangular configuration in contrast to action potentials
from normal atria that mostly showed a distinct plateau.
Refractory periods were also shorter and the normal rate
Introduction
In about 85% of patients with atrial fibrillation an
underlying cardiac abnormality or metabolic disorder
can be found, often associated with atrial enlargement1'1. Even in patients without such disorders, the
arrhythmia itself can lead to atrial dilatation. It is
therefore not unreasonable to assume that stretch may
be arrhythmogenic.
In the following brief description of pathophysiological mechanisms that may cause atrial fibrillation, I
shall first discuss the effects of acute and chronic stretch.
This will be followed by a description of cellular electrophysiological abnormalities and electrical remodelling.
Finally it will be attempted to relate these findings to
arrhythmia mechanisms.
adaptation of the refractory period disappeared, so that,
following a slowing of the heart rate, the refractory period
did not prolong. These changes largely seem to be the result
of prolonged episodes of rapid atrial activity and may be
called electrophysiological remodelling. In addition, a
marked dispersion refractoriness has been found which
might be due to different factors, such as fibrosis and local
denervation.
It is likely that atrial dilatation and fibrosis are important
factors in the occurrence and maintenance of atrial fibrillation. In an enlarged atrium, multiple re-entrant circuits
can co-exist. Fibrosis leads to inhomogeneities in both
conduction and refractoriness. Finally, the arrhythmia itself
causes persistent shortening of refractoriness. All of these
changes favour re-entry.
(Eur Heart J 1997; 18 (Suppl C): C12-C18)
Key Words: Action potential, refractory period, stretch,
fibrosis, dilatation, electrical remodelling, reentry.
The effects of acute and chronic
stretch
In recent years, the interaction between cardiac mechanics and electrophysiology has been the subject of many
experimental and clinical studies. The focused issue
of Cardiovascular Research of June 1996 is entirely
devoted to mechanoelectric feedback'21. In that issue,
Nazir and Lab review the literature dealing with mechanoelectrical feedback and atrial arrhythmias'31. Acute
stretch has different effects on the configuration of the
atrial action potential, depending on its timing. When
stretch is applied during the plateau phase of the action
potential, and the membrane potential is more positive
than the equilibrium potential of the so-called stretch
activated channels ( - 40 mV)'4], a repolarizing current is
induced which shortens the action potential. When
Correspondence: Michiel J. Janse, Department of Clinical and stretch is applied later, at times when membrane potenExperimental Cardiology, Academic Medical Center, University of tial is more negative than the equilibrium potential of
the stretch-activated channels, an inward current is
Amsterdam, Amsterdam, The Netherlands.
0195-668X/97/0C0012+07 $18.00/0
© 1997 The European Society of Cardiology
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Atrial fibrillation is often associated with atrial enlargement
and stretch is known to cause electrophysiological alterations. Acute stretch may, depending on the moment at
which it is applied, cause action potential shortening or
induce both early and delayed afterdepolarizations which,
when large enough, may initiate triggered premature action
potentials. The effects of acute stretch may be very different
from those of chronic stretch. In fact, in dogs with mitral
valve disease in which progressive atrial enlargement, leading to atrial fibrillation, developed over a period of years,
hardly any changes in transmembrane potential characteristics were found. In contrast, marked fibrosis developed
which could favour re-entry because of slow fragmented
conduction.
Why does atrial fibrillation occur?
A
03;
3
t
\V
.'A., \ 2
-50 -
<u
,
®,d5)
\
V
. \ ^ -
Et
©
250 ms
activated which lengthens the action potential and may
cause both early and delayed after-depolarizations.
These effects are schematically illustrated in Fig. 1.
Shortening of the action potential favours reentry, because wavelength (i.e. the product of refractory
period, which normally equals action potential duration,
and conduction velocity) is shortened. It is noteworthy
in this respect that during atrial flutter in humans,
the ventricular contractions caused a shortening of the
flutter cycle length, presumably due to stretch-induced
shortening of the atrial refractory period, resulting in
shortening of the wavelength and a shorter time for the
circulating re-entrant wave to complete its circuit151.
The role of after-depolarizations, possibly induced by late systolic or early diastolic stretch, in
causing atrial fibrillation is very uncertain. It is generally
recognized that re-entry is responsible for fibrillation.
Thus, triggered action potentials caused by either early
or delayed after-depolarizations that are large enough to
reach threshold, may at best provide premature atrial
depolarizations that could induce re-entry, given the
circumstance that an appropriate electrophysiological
substrate for re-entry is present. Moreover, if atrial
dilatation is indeed an important factor in causing atrial
fibrillation, one should consider the effects of chronic
stretch rather than the effects of acute stretch.
Boyden et alS6^ published an important study in
which they measured the atrial cellular electrophysiological characteristics of dogs having naturally occurring
mitral valve disease with progressive enlargement of the
atria. Some animals were followed for 5 years before the
electrophysiological study was performed. Most dogs
developed atrial arrhythmias, including atrial fibrillation. Remarkably, the transmembrane potential characteristics of atrial cells of these animals were not
significantly different from those of control animals,
although some cells were found with resting membrane
potentials below — 60 mV and that were inexcitable. In
the atria, massive interstitial fibrosis and cellular hyper-
trophy were found. The authors concluded that the
morphological changes were much more important in
causing atrial fibrillation. The increased atrial size would
permit the coexistence of many re-entrant circuits, even
when wavelength remained unaltered. The increase in
connective tissue would alter anisotropic properties
and could lead to slow, inhomogeneous conduction,
unidirectional block and re-entry. Spachs et a/.t?1 demonstrated how in an isolated preparation from a 62-yearold patient with an enlarged and hypertrophied right
atrium, where virtually all muscle fibres were surrounded by collagenous septae, extra-cellular potentials
were fragmented, and micro re-entry occurred.
Cellular, electrophysiological
abnormalities
From the previous paragraph the conclusion may be
drawn that morphological changes are much more important than possible alterations in cellular electrophysiology in causing atrial fibrillation. However, a number
of studies, both in humans and in experimental animals,
have documented changes in cellular electrophysiology
in atrial fibrillation.
Early studies in which transmembrane potentials
were recorded from isolated preparations obtained from
fibrillating human atria, either from atria of normal size
or from dilated atria, showed that resting potentials
were significantly reduced compared with resting potentials from cells in non-fibrillating atria18'91. In later
studies, however, hypopolarized cells were infrequently
found (15% of cells from fibrillating atria versus 5%
in normal atria)[l0!. The functional significance of the
presence of partially depolarized cells is twofold: first,
conduction velocity will be decreased because of the
reduction of action potential upstroke velocity and
amplitude as a consequence of the low resting membrane
potential; second, since in partially depolarized cells the
recovery kinetics of both the fast and the slow inward
current are markedly delayed, the refractory period is
prolonged, and lags behind completion of repolarization. This so-called post-repolarization refractoriness
results in a spatial dispersion of refractory periods.
Both a reduced conduction velocity and an increase in
dispersion of refractoriness predispose to re-entry.
Attuel and colleagues1"1 were the first to report
on an intriguing abnormality observed in patients vulnerable to atrial arrhythmias, including fibrillation. In
these patients there was no, or hardly any, adaptation of
the atrial refractory period to changes in heart rate. At
the most rapid rates investigated (cycle lengths in the
order of 350—400 ms), the range of refractory period
durations was similar to that of normal patients (of
approximately 160-250 ms). However, upon slowing of
heart rate, no prolongation of the refractory period was
observed so that at cycle lengths between 800 and
1000 ms, the refractory periods of the atria of arrhythmic patients were much shorter than those of normal
Eur Heart J, Vol. 18, Suppl C 1997
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Figure 1 Schematic representation of the effects of
stretch. The solid lines depict two consecutive atrial action
potentials. Stretch applied at A shortens the action potential: stretch applied at B prolongs the action potential and
may result in an early after-depolarization. Stretch when
repolarization is complete (C) may produce a delayed
after-depolarization. (Reproduced from Mazir and Lab131
with permission, Mazir and Lab modified the scheme from
Hansen et al)29K)
C13
C14
M. J. Janse
280
RA
RA
•c
100 ms
15
n = 25
x = 158.4
sd = 9.1
10
200
140
100
200
Interval (ms)
300
Figure 2 Upper two tracings: Electrograms during a 4-s
period of atrial fibrillation recorded from an epicardial site
on the right ventricle (RV) and a site on the right atrium
(RA) in a patient during open-heart surgery. Lower
tracing: same tracing as the right atrium recording in A, in
which the vertical lines are the activation moments as
determined by an interactive computer program. Lowest
panel: 25 intervals between activation moments are expressed in a histogram. The mean atrial fibrillation interval at this particular site, that is, the index of local
refractoriness, was 158-4 ms. (Reproduced from Ramdat
Misier et a/.'121 with permission.)
patients. These findings were largely supported by later
studies, in which action potentials from isolated atria
were recorded at different pacing rates'101. Again, a poor
adaptation of both action potential duration and refractory period to heart rate was found. Moreover, the
effective refractory period was usually shorter in atrial
tissue obtained from patients with atrial fibrillation than
from normal atrial preparations, except at short cycle
lengths where, because of post-repolarization refractoriness, refractory periods were longer. In addition, it was
found that the proportion of triangular action potentials
in the atrial fibrillation group was much greater than in
the control group (97% vs 23%). Finally, dispersion of
action potential duration was much greater in the atrial
fibrillation group than in the control group. Increased
dispersion in refractoriness was also found by direct
measurements during cardiac surgery in patients with
chronic atrial fibrillation'121. In that study, the average
interval between local activations during atrial fibrillation, the so-called atrial fibrillation interval, was used
Eur Heart J, Vol. 18, Suppl C 1997
160
180
AF interval (ms)
200
Figure 3 Relation between atrial fibrillation (AF) intervals and the refractory period, determined with the extrastimulus technique during regular pacing of the atria
at a basic cycle length of 400 ms, at four epicardial
sites. (Reproduced from Ramdat Misier et al)121 with
permission.)
as an index of local refractoriness (see Fig. 2). As shown
in Fig. 3, there was a good correlation between the
refractory period duration, determined with the extra
stimulus technique during regular pacing of the atria at
a cycle length of 400 ms, and the atrial fibrillation
interval, measured during atrial fibrillation at the same
right atrial epicardial sites. The advantage of using the
atrial fibrillation interval is that simultaneous recordings
during brief episodes of atrial fibrillation could be made
at 40 atrial sites, whereas classical determination of
refractory periods at 40 sites with the extra stimulus
technique would take an inordinate amount of time, and
would in fact be impossible during an operation. We
employed this technique in ten patients with idiopathic
paroxysmal atrial fibrillation and in a control group of
six patients. Reasons for operation were the 'corridor'
operation for drug refractory atrial fibrillation (eight
patients), surgical ablation of accessory pathways in
patients with atrial fibrillation (two patients) and surgical treatment of post-infarction ventricular tachycardia
refractory to medical therapy (six control patients).
After a routine median sternotomy, a multi-electrode
grid with up to 40 electrode terminals was placed over
the right atrium and atrial fibrillation was induced by
premature stimulation. The average atrial fibrillation
interval in patients with paroxysmal atrial fibrillation,
recorded at a total of 247 sites, was 152 ± 3 ms, compared with a value of 176 ± 81 ms recorded at 118 sites
in the control group (/ > <005). Dispersion in atrial
fibrillation intervals, defined as the variance of the
fibrillation intervals at all recording sites, was three
times larger in the atrial fibrillation group than in the
control group (Fig. 4). The question arises whether the
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
•J3
Why does atrial fibrillation occur?
180
10
15
20
Electrodes
25
30
short refractory period is the cause or the result of atrial
fibrillation. Recent studies suggest that the latter is the
case.
Electrical remodelling
In an experimental model, the hypothesis was tested that
atrial fibrillation itself causes the electrophysiological
changes in the atrium that favour both induction and
maintenance of atrial fibrillation'131.
In chronically instrumented conscious goats, in
which 27 electrodes were sutured onto the epicardium of
both atria, a special device was implanted that could
detect the presence or absence of atrial fibrillation and
could also induce atrial fibrillation by delivering a 1-s
burst of biphasic stimuli (strength four times diastolic
threshold; interval 20 ms). Initially, electrically induced
atrial fibrillation lasted only a few seconds before it
terminated spontaneously. However, when the arrhythmia was repetitively re-induced, the episodes of atrial
fibrillation gradually became longer, until chronic fibrillation (i.e. lasting longer than 24 h) occurred within
periods varying among individual animals from several
days to 2 weeks (Fig. 5). Repetitive induction of atrial
fibrillation did not alter conduction velocity, but gave
rise to a marked shortening of the refractory period
from control values of 151 ± 12 ms to 93 ± 20 ms after
24 h. In addition, the normal rate of adaptation of the
refractory period was abolished, or even reversed, resulting in a shortening of the refractory period at slower
heart rates.
These changes could also be obtained by regular
pacing at a rapid rate. When the atria were paced at a
cycle length of 180 ms, the ventricles responded in a 2:1
fashion. After 24 h, the atrial refractory period had
shortened from 140 to 105 ms. When the atrial pacing
cycle was lengthened to 360 ms, with the ventricles now
being activated in a 1:1 manner so that ventricular rate
and haemodynamic conditions remained constant, it
took 24 h for the atrial refractory period to return to its
original value. These results indicate that, following
cardioversion of atrial fibrillation after 1 or 2 days,
conditions remain favourable for re-induction of atrial
fibrillation for at least 12 h. This electrical remodelling
may be responsible for the well known fact that cardioversion has a much higher success rate when atrial
fibrillation is of recent onset'141.
On the ventricular level, persisting T wave
changes following a single paroxysm of ventricular
tachycardia had been described in 1935[151. Rosenbaum
et a/.'161 used the term cardiac 'memory' to describe the T
wave inversion that developed after about 24 h of pacing
and which could persist for several weeks after pacing
was stopped. Katz'171 has suggested that such longlasting T wave abnormalities could be initiated by
stretch, which would induce changes in gene expression
that in their turn would lead to the formation of
abnormal potassium channels. It is possible that such a
process might also occur on an atrial level. Future
research concerning the changes in gene expression
resulting in alterations of the ion channels involved in
atrial repolarization, such as the transient outward current or the delayed rectifier, might unravel the mechanisms of the atrial fibrillation-induced and persisting
shortening of the action potential and the refractory
period. Once the mechanism is known, ways may be
found to counteract the changes in gene expression.
It must be acknowledged that, besides the shortening of the refractory period, other factors must also
play a role in the development of chronic atrial fibrillation. The refractory period reaches a steady state
within a few days following induction of atrial fibrillation, whereas it often takes a few additional weeks
for atrial fibrillation to become persistent'131. Possibly,
atrial enlargement and the development of fibrosis are
required as well.
The wavelength and inducibility of
atrial fibrillation
It has been recognized for a long time that during
re-entrant rhythms, the conduction time of the reentrant impulse travelling around an area of block must
be long enough to allow fibres proximal to the zone of
block to recover their excitability. The wavelength
for circus movement re-entry has been defined as the
distance travelled by the depolarization wave during the
refractory period: wavelength=conduction velocity x
refractory period. When the wavelength is short, because
of depressed conduction, the refractory period is shortened, or both, small areas of conduction block may
already be sufficient for the establishment of re-entrant
circuits. Since conduction block is more likely to occur
in small areas than in a large segment of atrial myocardium, it is to be expected that inducibility of atrial
fibrillation depends on wavelength. If wavelength during
fibrillation is long, fewer wavelets can circulate through
Eur Heart J, Vol. 18, Suppl C 1997
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Figure 4 Atrial fibrillation intervals recorded simultaneously at 37 sites (electrodes) in a control patient and
at 32 sites in a patient with paroxysmal atrial fibrillation.
D indicates controls and • atrial fibrillation. (Reproduced
from Ramdat Misier et a/.1'2' with permission.)
C15
C16
M. J. Janse
Burst
pacing_AF
Duration of
fibrillation
> Sinus rhythm
Control
20 s
After
2 weeks
>24h
2s
the atria and fibrillation may be self-terminating. If
wavelength is short, a greater number of wavelets will be
present and fibrillation will tend to be stable and longlasting. Wavelength is therefore also important for
maintenance of fibrillation.
In conscious dogs, in which multiple electrodes
for recording and stimulation had been attached to both
atria, both refractory periods and conduction velocity
were measured. To change wavelength, a variety of
drugs (acetylcholine, propafenone, lidocaine, ouabain,
quinidine, sotalol) were administered, and the refractory
period, conduction velocity, and their product were
correlated with the induction of atrial arrhythmias during premature stimulation'181. In all dogs (n=19), atrial
arrhythmias (n = 549) including atrial fibrillation
(n = 208) could be induced by single premature stimuli.
Although at shorter refractory periods, a relatively high
incidence of atrial fibrillation was observed, prolongation of the refractory period did not always prevent
atrial fibrillation. In fact, the predictive power of refractory period duration alone, or conduction velocity
alone, for induction of arrhythmias was poor. In contrast, wavelength correlated very well with inducibility
of atrial arrhythmias.
It is therefore reasonable to assume that drugs
that would increase wavelength, by prolonging the refractory period, increasing conduction velocity, or both,
would be anti-fibrillatory. Indeed, mapping experiments
in dogs with atrial fibrillation in which the arrhythmia
was terminated by flecainide or propafenone'19'201
showed that, because of the use-dependent increase
in refractoriness, the size of the re-entrant circuits
Eur Heart J, Vol. 18, Suppi C 1997
increased and the number of re-entrant wavelets decreased, until block in the remaining circuit(s) occurred
and sinus rhythm was restored.
Dispersion of refractoriness
Although one factor predisposing to re-entry, the shortening of the refractory period, may be the result of
prolonged episodes of rapid atrial activity, another
factor, dispersion of refractory periods, is unlikely to be
due to rapid atrial excitation. Different factors may be
involved in this dispersion. Vagal stimulation shortens
the atrial refractory period in a non-uniform way'21'221,
possibly because 'fibres immediately adjacent to vagal
post-ganglionic endings are exposed to relatively high
concentrations of the cholinergic mediator and are profoundly affected, while fibres more remote from sites of
acetylcholine liberation are influenced to a much less
degree''2'1. Both the shortening of the refractory period
and the increase in dispersion are arrhythmogenic
because 'an early ectopic impulse generated during a
period of vagal stimulation is bound to be propagated
along an irregular wavefront as the impulse encounters
areas in varying states of excitability. The likelihood of
fibrillation must be enhanced by such irregularity''211.
These mechanisms may operate in the syndrome of
vagally mediated paroxysmal atrial fibrillation occurring
in relatively young patients without structural heart
disease described by Coumel et a/.'231. It is difficult to
find an explanation for the much rarer adrenergically
mediated paroxysmal atrial fibrillation'241, since stellate
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Figure 5 Prolongation of the duration of episodes of electrically induced atrial fibrillation (AF) after maintaining AF for
24 h and 2 weeks respectively. The three tracings show a single
atrial electrogram recorded from the same goat during induction
of AF by a 1-s burst of stimuli (50 Hz, 4 x threshold). In the
upper tracing the goat has been in sinus rhythm all the time and
AF self-terminated within 5 s. The second tracing was recorded
after the goat had been connected to the fibrillation pacemaker
for 24 h showing a clear prolongation of the duration of AF to
20 s. The third tracing was recorded after 2 weeks of electrically
maintained AF. After induction of AF this episode became
sustained and did not terminate. (Reproduced from Wijffels
et a/.'131 with permission.)
Why does atrial fibrillation occur?
ganglion stimulation was found to have no effect on
atrial refractoriness'221. Whereas vagal stimulation may
result in functional inhomogeneities in normal hearts
and cause atrial fibrillation, it has been suggested that
fibrosis provides the pathological basis for electrophysiological inhomogeneities in atrial fibrillation in rheumatic
heart disease. Extracellular electrograms from atrial
strips from patients with rheumatic mitral stenosis were
found to be fragmented ('toothbrush appearance') and
this was attributed to fibrosis'251. Similar fragmented
electrograms have been found in patients with atrial
fibrillation; the fragmentation increased during premature stimulation, and the conduction delay was
greater in patients with atrial fibrillation than in control
patients1261.
Summary
In summary, it is likely that atrial dilatation and the
development of fibrosis are important factors for the
occurrence and maintenance of atrial fibrillation. In an
enlarged atrium, multiple re-entrant circuits can coexist.
Fibrosis leads to imhomogeneities in conduction and
refractoriness. Finally, the arrhythmia itself causes persistent shortening of refractoriness. All of these changes
favour re-entry.
References
[1] Murgatroyd F, Camm AJ. Atrial arrhythmias. Lancet 1993;
341: 1317-22.
[2] Lab M, Taggart P, Sachs F, eds. Focussed issue on mechanoelectric feedback. Cardiovasc Res 1996; 32: 1-188.
[3] Nazir SA, Lab MJ. Mechanoelectric feedback and atrial
arrhythmias. Cardiovasc Res 1996; 32: 52-61.
[4] Bustamente JO, Ruknudin A, Sachs F. Stretch-activated
channels in heart cells: relevance to cardiac hypertrophy.
J Cardiovasc Pharmacol 1991; 17 (Suppl 2): SI 10-3.
[5] Ravelli F, Disertori M, Cozzi F, Antolini R, Allessie M.
Ventricular beats induce variations in cycle length of rapid
(type II) atrial flutter in humans. Evidence for leading circle
reentry. Circulation 1994; 89: 2107-16.
[6] Boyden PA, Tilley LP, Pham TD, Liu SK, Fenoglio JJ Jr, Wit
AL. Effects of left atrial enlargement on atrial transmembrane
potentials and structure in dogs with mitral valve fibrosis. Am
J Cardiol 1982; 49: 1896-908.
[7] Spachs MS, Dolber PC, Heidlage JF. Influence of the passive
anisotropic properties on directional differences in propagation following modification of the sodium conductance in
human atrial muscle. A model of reentry based on anisotropic
discontinuous propagation. Circ Res 1988; 62: 811-32.
[8] Hordof AJ, Edie R, Malm JR, Hoffman BF, Rosen MR.
Electrophysiologic properties and response to pharmacologic
aspects of fibers of diseased human atria. Circulation 1976; 54:
774-9.
[9] Ten Eick RA, Singer DH. Electrophysiological properties of
diseased human atrium. 1. Low diastolic potential and altered
cellular response to potassium. Circ Res 1979; 44: 545-57.
[10] Le Heuzey JY, Boutjdir M, Gagey S, Lavergne T, Guize L.
Cellular aspects of atrial vulnerability. In: Attuel P, Coumel P,
Janse MJ, eds. The atrium in health and disease. Mt Kisco,
NY: Futura Publishing Company, 1989: 81-94.
[11] Attuel P, Childers R, Cauchemez B, Poveda J, Mugica J,
Coumel P. Failure in rate adaptation of the atrial refractory
period: its relationship to vulnerability. Int J Cardiol 1982; 2:
179-97.
[12] Ramdat Misier AR, Opthof T, van Hemel NM, Defauw
JJAM, de Bakker JMT, Janse MJ. Increased dispersion of
'refractoriness' in patients with idiopathic paroxysmal atrial
fibrillation. J Am Coll Cardiol 1992; 19: 1531-5.
[13] Wijffels MCEF, Kirchhof CJHJ, Dorland R, Allessie MA.
Atrial fibrillation begets atrial fibrillation. A study in awake
chronically instrumented goats. Circulation 1995; 92: 1954-68.
[14] Crijns HJGM, van Wijk LM, van Gilst WH, Kingma HJ, van
Gelder JC, Lie K.I. Acute conversion of atrial fibrillation to
sinus rhythm: clinical efficacy of flecainide acetate. Comparison of two regimens. Eur Heart J 1988; 9: 634-8.
[15] Graybiel A, White PD. Inversion of the T wave in lead I or II
of the electrocardiogram in young individuals with neurocirculatory asthenia, with thyrotoxicosis, in relation to certain
infections, and following paroxysmal ventricular tachycardia.
Am Heart J 1935; 10: 345-54.
[16] Rosenbaum MB, Blanco HH, Elizari WV, Lazzari JO,
Davidenko JM. Electronic modulation of the T wave and
cardiac memory. Am J Cardiol 1982; 50: 213-22.
[17] Katz AM. T wave 'memory': possible causal relationship to
stretch-induced changes in cardiac ion channels? J Cardiovasc
Electrophysiol 1992; 3: 150-9.
[18] Rensma PL, Allessie MA, Lammers WJEP, Bonke FIM,
Schalij MJ. Length of excitation wave and susceptibility to
re-entrant arrhythmias in normal conscious dogs. Circ Res
1988; 62: 394-410.
[19] Wang Z, Page P, Nattel S. Mechanism of flecainide's antiarrhythmic action in experimental atrial fibrillation. Circ Res
1992; 71: 271-87.
[20] Villemaire C, Talajic M, Nattel S. Comparative mechanism of
antiarrhythmic drug action in experimental atrial fibrillation:
importance of use dependent effects on refractoriness. Circulation 1993; 88: 1030-44.
[21] Alessi R, Nusynowitz M, Abildskov JA, Moe GK. Nonuniform distribution of vagal effects on the atrial refractory
period. Am J Physiol 1958; 194: 406-10.
[22] Zipes DP, Mihalick MJ, Robbins GT. Effects of selective
vagal and stellate ganglion stimulation on atrial refractoriness.
Cardiovasc Res 1974; 8: 647-55.
[23] Coumel P, Attuel P, Lavelle JP, Flammang D, Leclerq JF,
Slama R. Syndrome d'arythmie auriculaire d'origine vagale.
Arch Mai Coeur 1978; 71: 645-51.
Eur Heart J, Vol. 18, Suppl C 1997
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Atrial fibrillation is more common with increasing age. Spach and Dolber'271 showed that with advancing age, extensive collagenous septa develop in the atria,
leading to progressive electrical uncoupling of the sideto-side connections of parallel-oriented atrial fibres. This
led to 'zigzag' conduction in the transverse direction and
to fragmented extracellular electrograms. Fibrosis may
result not only in slow, zigzag conduction, but also in an
increase in dispersion of refractoriness since, in wellcoupled cells, the current flow during repolarization
will tend to decrease dispersion by prolonging action
potentials with a short duration and shortening action
potentials with a long duration. Dispersion of atrial
refractoriness does increase with age and is increased in
patients with atrial fibrillation'10'121. Another factor that
could contribute to dispersion of the atrial refractory
period is the presence of partially depolarized cells'8'91,
because in these cells the recovery of excitability lags
behind completion of repolarization (post-repolarization
refractoriness). Finally, in atria with regional fibrosis,
sympathetic and parasympathetic fibres may be interrupted. This may cause regional supersensitivity to
circulating catecholamines and acetylcholine and thus
create inhomogeneity in refractoriness'281.
C17
C18 M. J. Janse
[24] Coumel P. Neurogenic and humoral influences of the auto- [27]
nomic nervous system in the determination of paroxysmal
atrial fibrillation. In: Attuel P, Coumel P, Janse MJ, eds. The
atrium in health and disease. Mount Kisco, NY: Futura
Publishing Company, 1989: 213-32.
[25] Van Dam RT, Durrer D. Excitability and electrical activity of [28]
human myocardial strips from the left atrial appendage in
cases of mitral stenosis. Circ Res 1961; 9: 509-514.
[26] Cosio FG, Pacacias J, Vidal JM, Cocina EG, Gomez-Sanchez [29]
A, Tamargo L. Electrophysiological studies in atrial
fibrillation. Slow conduction of premature impulses: a possible
manifestation of the background for reentry. Am J Cardiol
1983; 51: 122-30.
Spach MS, Dolber PC. Relating extracellular potentials and
their derivatives to anisotropic propagation at a miscroscopic
level in human cardiac muscle: evidence for electrical uncoupling of side-to-side connections with increasing age. Circ
Res 1986; 58: 356-71.
Inoue H, Zipes DP. Results of sympathetic denervation in the
canine heart: supersensitivity that may be arrhythmogenic.
Circulation 1987; 75: 877-87.
Hansen DE, Craig CS, Hondeghem LM. Stretch-induced
arrhythmias in the isolated canine ventricle: evidence for the
importance of mechano-electric feedback. Circulation 1990;
81: 1094-105.
Downloaded from http://eurheartj.oxfordjournals.org/ by guest on September 30, 2016
Eur Heart J, Vol. 18, Suppl C 1997