Download Endocardial mapping of atrial fibrillation in the human right atrium

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Management of acute coronary syndrome wikipedia , lookup

Cardiac contractility modulation wikipedia , lookup

Mitral insufficiency wikipedia , lookup

Cardiac surgery wikipedia , lookup

Quantium Medical Cardiac Output wikipedia , lookup

Arrhythmogenic right ventricular dysplasia wikipedia , lookup

Electrocardiography wikipedia , lookup

Lutembacher's syndrome wikipedia , lookup

Dextro-Transposition of the great arteries wikipedia , lookup

Heart arrhythmia wikipedia , lookup

Atrial septal defect wikipedia , lookup

Ventricular fibrillation wikipedia , lookup

Atrial fibrillation wikipedia , lookup

Transcript
European Heart Journal (2000) 21, 550–564
doi:10.1053/euhj.1999.1851, available online at http://www.idealibrary.com on
Endocardial mapping of atrial fibrillation in the
human right atrium using a non-contact catheter
R. J. Schilling2, A. H. Kadish1, N. S. Peters2, J. Goldberger1 and D. Wyn Davies2
1
Northwestern Memorial Hospital, Chicago IL, U.S.A.; 2St. Mary’s Hospital and Imperial College School of
Medicine, London, U.K.
Background Endocardial mapping of atrial fibrillation in
humans is limited by its low resolution and by complexities
in the arrhythmia and atrial anatomy.
Methods and Results A catheter mounted non-contact
multielectrode was deployed in the right atrium of 11
patients with atrial fibrillation and used to reconstruct 3360
electrograms, superimposed onto a computer-simulated
model of the endocardium, using inverse solution mathematics. This allows construction of isopotential maps of
the right atrium. Patients had either sustained atrial fibrillation (n=3) for >6 months or developed atrial fibrillation
during the study (n=8). Spontaneous initiation of atrial
fibrillation was recorded in one patient and was demonstrated by the non-contact system to arise from two successive atrial ectopic beats from the site of a roving contact
catheter. Reconstruction of electrograms recorded during
atrial fibrillation was validated by comparison with contact
electrograms with cross-correlation. During established
atrial fibrillation, four patients predominantly had a single
right atrial wave front, two had two wave fronts and five
Introduction
Moe’s original hypothesis that atrial fibrillation was
caused by continuous reentry[1–3], was confirmed by later
experimental studies[4–6] so that it is now widely accepted
that atrial fibrillation is a reentrant arrhythmia[7]. Drug
treatment to prevent recurrent atrial fibrillation has
limited efficacy and chronic antiarrhythmic therapy can
be associated with increased mortality[8]. Catheter ablation strategies to restore and maintain sinus rhythm
after atrial fibrillation have been developed following the
success of surgical ablation procedures[9], but their
Revision submitted 5 July 1999, and accepted 21 July 1999.
Correspondence: D. Wyn Davies, Waller Department of Cardiology, St. Mary’s Hospital, Praed Street, London W2 1NY, U.K.
0195-668X/00/070550+15 $35.00/0
patients had three to five wave fronts for most of the time.
Periods of electrical silence were seen in the right atrium in
eight patients, after which, activity emerged from consistent
septal sites alone, suggesting a left atrial origin. During
intravenous administration of flecainide, atrial fibrillation
in two patients terminated spontaneously or following
pacing manoeuvres, while in the remaining patient sinus
rhythm was restored via atrial tachycardia.
Conclusion Non-contact mapping of the right atrium has
demonstrated modes of initiation and termination of atrial
fibrillation, characterized different patterns of right atrial
activation in atrial fibrillation and suggests that the left
atrium may sustain atrial fibrillation in some patients.
Simultaneous mapping of the right and left atrium is
required to further elucidate the mechanisms of human
atrial fibrillation.
(Eur Heart J 2000; 21: 550–564)
2000 The European Society of Cardiology
Key Words: Mapping, atrial, fibrillation.
success will be critically dependent on improved
understanding of atrial fibrillation mechanisms.
Atrial fibrillation has been mapped in animal
models[10–16] and humans[16–24] and has provided
insights on the arrhythmia mechanism, but studies have
been limited by the resolution of the mapping system
used[25], the extent of atrium mapped[20], or the need for
the procedure to be performed after thoracotomy[20–22],
which has an associated morbidity and mortality.
The aim of this study was to assess the feasibility of
using a non-contact mapping system to map atrial fibrillation in the human right atrium. Because the system
continuously acquires high-resolution data simultaneously from the entire right atrium and is deployed
percutaneously it may overcome some of the problems
associated with the atrial fibrillation mapping techniques
used to date.
2000 The European Society of Cardiology
Non-contact endocardial mapping of human right atrial fibrillation
Table 1
Patient
1
2
3
4
5
6
7
8
9
10
11
551
Patient demographics
Age
(years)
55
40
66
42
34
76
73
75
36
34
74
Drugs
RA lat/TA
(cm)
RA-AP
(cm)
Digoxin, warfarin
Disopyramide
Verapamil, warfarin
None
Amiodarone, warfarin
Digoxin, diltiazem, warfarin
Diltiazem, metoprolol
None
Amiodarone
None
Amiodarone
5·5
4·3
4
4
5·6
5·4
5·8
6·2
9·9
7·4
8·4
4·3
3·5
3·2
3·2
4·7
3·7
3·6
2·8
5·1
4·3
5·3
Underlying arrhythmia
AF
P Septal WPW
A flutter
Paroxysmal AF
AF
AF
A flutter
A flutter
A flutter
AVRT
A flutter
Underlying arrhythmia=the patients’ presenting clinical arrhythmia and the reason for undergoing
electrophysiological study; AF=atrial fibrillation; P Septal WPW=a postero-septal accessory
pathway presenting as Wolff–Parkinson–White syndrome; A flutter=atrial flutter; Drugs=the drug
therapy the patient was taking prior to the mapping study; RA lat/TA=the right atrial lateral wall
to tricuspid distance (cm); RA-AP=the right atrial anterior to posterior distance (cm).
Methods
Patients
Right atrial mapping was performed in 11 patients (8
male, mean age 55 years, range 34 to 76 years) (Table 1).
Three patients had a history of chronic atrial fibrillation
in whom non-contact mapping studies were carried out
under the supervision of the local ethics committee in
order to assess the validity and feasibility of non-contact
mapping of human atrial fibrillation. No therapeutic
procedures were performed in these patients. In the
remaining patients non-contact mapping was performed
to characterize other arrhythmia substrates and atrial
fibrillation was induced during the course of the study.
One patient was undergoing an electrophysiological
study for a right posteroseptal accessory pathway and
the remaining patients had a history of paroxysmal atrial
fibrillation and flutter. All patients underwent preoperative echocardiography. Anticoagulation was
stopped 3 days prior to study in the patients on warfarin
therapy (Table 1) and was reinitiated immediately after
the procedure. Antiarrhythmic drugs were unchanged
after the procedure in those patients with chronic atrial
fibrillation and were stopped in patients with an accessory pathway or atrial flutter after successful catheter
ablation procedures subsequent to atrial fibrillation
mapping. A 12-lead electrocardiogram was recorded
during atrial fibrillation either immediately prior to the
mapping procedure in patients with chronic atrial fibrillation or during atrial fibrillation in the patients with
paroxysmal atrial fibrillation. Atrial fibrillation was
divided into ‘coarse’ or ‘fine’ according to the appearance of the diastolic interval on the surface ECG[26,27].
Mapping procedure
The study was approved by the local ethics committee
whose guidelines were followed (n=6) or the Insti-
tutional Review Board at Northwestern University
(n=5). All patients were studied in the post-absorptive
state, having given written informed consent. A quadripolar catheter was placed in the right ventricular apex
and a 7 French deflectable mapping/ablation 4 mm tip
catheter placed in the right atrium via right femoral
venous sheaths. Systemic arterial blood pressure was
continuously monitored via a 6 French femoral arterial
sheath. A 9 French sheath was placed in the left femoral
vein to introduce the non-contact mapping catheter.
Contact catheter data and surface ECGs were recorded
simultaneously on both a conventional electrophysiology system and the non-contact mapping system.
Non-contact mapping system
The non-contact mapping system (EnSite 3000Endocardial Solutions, St. Paul, MN, U.S.A.) has been
described in detail elsewhere[28]. Briefly, it consists of a
catheter-mounted multielectrode array, a custom built
amplifier system, and a Silicon Graphics workstation
that runs specially designed system software. The multielectrode array catheter consists of a 7·5 ml balloon
mounted on a 9F catheter (Fig. 1) around which is
woven a braid of 64 0·003 inch diameter wires. Each wire
has one 0·025 inch break in insulation producing a
non-contact unipolar electrode; these 64 electrodes constitute the multielectrode array. The raw far-field electrographic data are acquired and fed into a multichannel
recorder and amplifier system, sampling at 1·2 kHz, and
filtered with a band width of 0·1 to 300 Hz. Unipolar
signals are recorded using a ring electrode as a reference
which is located on the proximal shaft of the multielectrode array catheter at the level of the inferior vena
cava.
Before deployment of the multielectrode array,
patients were given 10 000 IU heparin, with later boluses
to maintain ACT between 300 and 400 s. The multielectrode array catheter was deployed in the right atrium
of all patients, over a 0·032 inch J-tipped guide wire,
Eur Heart J, Vol. 21, issue 7, April 2000
552
R. J. Schilling et al.
Figure 1 (a) A postero-anterior radiograph showing the multielectrode
array deployed around the balloon filled with 7·5 ml of contrast (B) in the
right atrium with the pigtail tip positioned near the superior vena cava. Also
seen is a mapping catheter positioned on the inferior vena cava-TV isthmus
(Map) a coronary sinus catheter (CS) and a halo catheter (H) which were
used to map and ablate this patient’s atrial flutter after mapping of atrial
fibrillation by creating a line of block across the TA-inferior vena cava
isthmus. (b) A postero-anterior radiograph showing the multielectrode array
(B) deployed in the right atrium and canted over towards the tricuspid
annulus by a guide wire passed through the central lumen and out through the
right ventricular outflow tract, so that the pigtail of the multielectrode array
is straightened and sited on the ventricular side of the TA. Also seen are
catheters sited at the high right atrium (HRA), right ventricular apex (RVA),
His bundle (His) and coronary sinus (CS). A mapping/ablation catheter has
been positioned on the TA-inferior vena cava isthmus (arrow).
Eur Heart J, Vol. 21, issue 7, April 2000
Non-contact endocardial mapping of human right atrial fibrillation
advanced to the superior vena cava (n=8) (Fig. 1(a)) or
the pulmonary outflow tract (n=3) having been positioned using a 7 French multipurpose catheter (Fig.
1(b)). This was done in patients with a postero-septal
accessory pathway or a large right atrium and atrial
flutter in order to maximize the system’s resolution of
that area by canting the balloon towards the atrioventricular junction or inferior vena cava tricuspid valve
isthmus. Previous data have demonstrated that the
accuracy of electrogram reconstruction improves with
decreasing distance between the multielectrode array
and the endocardium[28]. At the end of the procedure,
heparin was reversed with protamine and the sheaths
removed immediately the ACT was below 120 s.
Catheter location
The system locates any other catheter in the chamber
with respect to the multielectrode array by passing a
5·68 kHz, low current locator signal between the catheter being located and alternately between ring electrodes proximal and distal to the multielectrode array on
the non-contact catheter. This locator signal serves two
purposes. Firstly it can be used to construct a threedimensional computer model of the endocardium (virtual endocardium), providing the geometry matrix for
the inverse solution. Geometric points are sampled by
dragging the located catheter around the cardiac chamber at the beginning of a study, either during sinus
rhythm (sampled points being gated 6 ms before the
R-wave) or arrhythmia (no gating). Secondly the locator
signal is used to display and log the position of any
catheter on the virtual endocardium during a study (Fig.
2(a)). This is used to mark anatomical locations (identified using fluoroscopy and electrogram characteristics)
on the virtual endocardium.
Reconstruction of electrograms
The electrical activity detected by the electrodes on the
surface of the multielectrode array is generated primarily
by the potential field on the endocardial surface. The
electrograms detected by non-contact electrodes are of
lower amplitude and frequency than the source on the
endocardium. A technique to enhance and resolve these
far-field potentials has been devised from an inverse
solution to Laplace’s equation using a boundary element
method. The inverse solution considers how a signal
detected at a remote point would have appeared at
source; the boundary element method is used to apply the
inverse solution, allowing resolution of a matrix of such
signals originating at a known boundary (the blood–
endocardial boundary). Using these techniques the system is able, from the multielectrode array potentials, to
reconstruct >3000 unipolar electrograms simultaneously
and superimpose them onto the virtual endocardium
(Fig. 2(a)) to produce isopotential or isochronal maps
with a colour range representing voltage or timing of
onset. The colour range on the isopotential maps can be
narrowed in order to produce a form of activation map
with the threshold for activation being set at an arbitrary
level by the operator. Electrograms may also be selected
553
from any site of interest on the virtual endocardium and
displayed individually (Fig. 2(b)).
Mapping protocol
The mean cycle length of the atrial fibrillation in the
right and left atrium were determined from bipolar
electrograms recorded simultaneously by conventional
multipolar catheters in fluoroscopically stable positions
in the high right atrium and coronary sinus.
If atrial fibrillation was chronic no attempt was made
to terminate it, but atrial fibrillation induced and persisting during the procedure was then terminated using
intravenous flecainide at a dose of 2 mg . kg 1, with a
maximum dose of 150 mg infused over 10 min. The
infusion was terminated if the patient reverted to sinus
rhythm. Recording of endocardial activation was
continued throughout this infusion and non-contact
mapping data was analysed off-line for all patients.
For the purposes of analysing termination of atrial
fibrillation after intravenous flecainide, data were
recorded continuously during the infusion and 10 s
samples of data were analysed every minute. The
number of wave fronts present was plotted against
time.
Validation of electrogram reconstruction
A previously described method for computerized crosscorrelation comparison was used[28] to compare reconstructed with contact electrograms, taken from the same
endocardial point identified by the catheter location
system. Signals for both contact and reconstructed electrograms were filtered with a bandwidth of 4 to 300 Hz.
Definitions
Atrial fibrillation
Atrial fibrillation was defined as an irregular tachycardia
with beat-to-beat change in contact intracardiac atrial
electrogram timing and morphology (either right atrium
or coronary sinus) and with an irregular ventricular
response[15–17].
Number of wave fronts
A wave front was defined as a discreet and moving front
of endocardial depolarization identified on unipolar
isopotential maps. They were identified as regions of
negative polarity on isopotential maps and when two
wave fronts collided and fused, they were then counted as
one. Conversely, when a wave front bifurcated, two wave
fronts were present from then on[14]. A wave front was
defined as the leading edge of a progressive and sequential wave of depolarization across the endocardium.
Wave fronts were detected by increasing the sensitivity
threshold of the isopotential map without displaying
noise. The initiation of a wave front was confirmed
by differentiating between endocardial activation and
background noise. This was achieved by examining
Eur Heart J, Vol. 21, issue 7, April 2000
554
R. J. Schilling et al.
reconstructed electrograms from the location of this wave
front initiation. Further confirmation of the validity of
endocardial activation, as indicated on the isopotential
map, was performed by electrogram examination each
time the wave front changed direction.
Reentry circuits
A reentry circuit was defined as a complete circuit of
re-excitation with a wave front completing at least two
rotations through the same circuit.
Eur Heart J, Vol. 21, issue 7, April 2000
Focal activity
Focal activity or breakthrough of electrical activity was
defined as a wave front emerging from a point on the
endocardium where the surrounding endocardium was
electrically silent so that activation was not propagating
from a detectable wave front.
Atrial fibrillation type
The pattern of atrial fibrillation in the right atrium
was divided into three categories using a similar
Non-contact endocardial mapping of human right atrial fibrillation
classification to that of Konings et al.[18] and defined as
follows:
(1) Type I atrial fibrillation—only one wave front
present in the right atrium for >50% of the time.
(2) Type II atrial fibrillation—two independent wave
fronts present in the right atrium for >50% of the time.
(3) Type III atrial fibrillation—three or more wave
fronts present in the right atrium for >50% of the time.
Our definitions differed from that of Konings et al. in
that data were acquired from the entire right atrium and
we stipulated that a set number of wave fronts should be
present for >50% of the time period analysed, rather
than for the entire period. Thus, in this study, larger
areas of the right atrium were analysed and for much
longer periods than had been possible in Konings’ study,
so that there was a variation in the number of wave
fronts present in most of the cases analysed. After atrial
fibrillation had been established for at least 30 s and
prior to any drug or pacing intervention, atrial fibrillation was then categorized as type I, II or III atrial
fibrillation according to the number of wave fronts
present on activation maps for the majority of the time
period analysed.
Statistics
Data are presented as means and standard deviations.
Means of continuous, normally distributed data
were compared with Student’s t-test. Other data were
compared using Chi-squared analysis.
Results
Isopotential maps of atrial fibrillation were recorded for
558597 s (meanSD) (range 9 to 1453 s) on the
non-contact mapping system.
Mean atrial fibrillation cycle length
Mean right and left atrial fibrillation cycle lengths were
determined from 322 electrogram intervals (mean 29
555
intervals per patient, range 16–60 per patient). The right
atrial fibrillation cycle length was 16936·5 ms
(meanstandard deviation), (range 58–285 ms) and left
atrial fibrillation cycle length was 169·240·9 ms (range
43–292 ms) which were not significantly different
(P=0·8).
Validation of atrial fibrillation electrogram
reconstruction
Comparison of non-contact reconstructed electrograms
and contact electrograms was performed for all patients
(Fig. 3). Approximately 3600 electrogram complexes
were compared with a mean cross-correlation of
0·740·19.
Initiation of atrial fibrillation
Initiation of atrial fibrillation in four patients without
chronic atrial fibrillation was caused by catheter manipulation and in four atrial fibrillation occurred during
attempted entrainment of atrial flutter. This happened
before introduction of the multielectrode array in one
patient. Initiation of atrial fibrillation was recorded by
the non-contact mapping system in one patient (Fig. 2(a)
and (b)). Figure 2(a) shows a series of activation maps
which demonstrate that the mapping/ablation catheter,
the position of which was determined on the virtual
endocardium using the locator signal, had initiated two
atrial ectopic beats, resulting in a macroreentry circuit.
This then divided into three wave fronts in the septum,
which then continued to propagate in an unpredictable
manner resulting in persistent atrial fibrillation.
Mapping of established atrial fibrillation
Atrial fibrillation, which had been established for at least
1 min, was recorded in all patients. The results of atrial
fibrillation mapping and termination are given in Table 2.
Figure 2 (a) A series of unipolar activation maps showing initiation of atrial fibrillation. The frames have not been
evenly spaced in time in order to show the activation patterns clearly, but the timing of each frame with reference to the
electrograms and the surface ECG is shown in (b) as a yellow number. A focal atrial wave front emerges from the point
of contact of the catheter (indicated by the green locator line) (frame 1), splits around the TV (frame 2) and fuses in the
superior right atrium (frame 3). A period of electrical silence in the right atrium (frame 4) followed by a second focal
activation wave front (frame 5), splits and is blocked in the isthmus (frame 6). The more superior limb rotates around
the TV and forms a macroreentry circuit which competes two circuits (frames 7 to 11) before degenerating to atrial
fibrillation. (b) Wave forms showing the period during which the activation maps were recorded. The yellow numbers
indicate the timing of the corresponding frame shown in (a) and a time scale is shown at the bottom of the figure in blue
(ms). ECG II=surface ECG lead II; Catheter=bipolar electrograms recorded from the roving contact catheter which
was the source of focal activity; a to f are reconstructed unipolar electrograms recorded from positions displayed on the
activation maps. Note that after the first focal activation, electrograms from a and f are of roughly equal timing, but
after the second beat, electrograms from a to f show double potentials. The first one is a far-field ventricular activation,
which is immediately followed by atrial electrograms in the same sequence and similar morphology. In the subsequent
beat when a macroreentry circuit is established the activation timing from a to f is again similar in morphology and
timing as a result of activation progress around the tricuspid annulus. This is then followed by more disorganized activity
corresponding with the initiation of atrial fibrillation (not shown in figures).
Eur Heart J, Vol. 21, issue 7, April 2000
556
R. J. Schilling et al.
Table 2
Patient
1
2*
3*
4*
5
6
7*
8
9
10
11
Results
AF on ECG
AF
type
Focal activation
Termination by flecainide
Coarse
Fine
Fine
Fine
Coarse
Fine
Coarse
Fine
Coarse
Coarse
Fine
I
III
III
III
II
III
I
II
I
I
III
No
No
Sup and CS sept
No
Sup sept
Sup mid and CS sept
Sup and CS sept
Sub sept
Sup mid and CS sept
Sup sept
Sup mid and CS sept
No
To A flutter (pace terminated)
To A flutter (blocked in isthmus)
Atrial ectopics
No
No
No
No
No
No
No
AF on ECG=the appearance of the baseline atrial activity on ECG, recorded during arrhythmia either immediately prior or during the
procedure (see text for details); AF type=the type of atrial fibrillation as identified on non-contact mapping of the right atrium (see text);
Focal activation=evidence of a wave front emerging from a focus surrounded by electrically silent tissue; Termination by
flecainide=whether termination using flecainide was attempted and the mode of termination (see text).
*Indicates patients in whom AF was initiated by catheter manipulation.
By these criteria four patients had predominantly type
I, two patients had predominantly type II and five
patients had predominantly type III atrial fibrillation in
the right atrium. The patients with type I atrial fibrillation had coarse P wave activity on the surface ECG
(Fig. 4(a)). Of patients with type II atrial fibrillation, one
had coarse and one had fine P wave activity on the ECG.
All of the patients with type III atrial fibrillation had fine
P wave activity (Fig. 4(b)) on the surface ECG. A coarse
pattern of atrial fibrillation on the surface ECG was
significantly associated with type I atrial fibrillation
(P=0·01).
Periods of right atrial electrical silence lasting >50 ms
were observed in the right atrium of eight patients (3
type I, 2 type II, 3 type III atrial fibrillation) despite
evidence of continuing atrial fibrillation on the surface
ECG and coronary sinus catheter electrograms. In all
patients, the re-initiation of right atrial activation was
via consistent locations on the right atrial septum,
possibly representing breakthrough from the left atrium.
In three patients the focus was in the superior septum in
the region of Bachmann’s bundle. In two patients, two
foci were observed in the superior septum (Bachmann’s
bundle) and near the coronary sinus os. In three
patients, three foci were observed in the superior
septum/Bachmann’s bundle, mid septum and near the
coronary sinus os.
Example of type I atrial fibrillation
An example of 1·5 cycles of type I atrial fibrillation is
shown in Fig. 5 and the patient’s 12-lead ECG is seen in
Fig. 4(a). The reentry wave front rotates around a line of
block situated in the posterior septum and posterolateral wall near the crista terminalis. This line of block
did not vary during atrial fibrillation, but slight variations in the path of the wave front were seen in other
regions of the right atrium.
Eur Heart J, Vol. 21, issue 7, April 2000
Example of type II atrial fibrillation
An example and description of type II atrial fibrillation
is shown in Fig. 6.
Example of type III atrial fibrillation
An example of type III atrial fibrillation is shown and
described in Fig. 7.
Termination of atrial fibrillation
Termination of atrial fibrillation following flecainide
infusion was observed and recorded in three patients, all
with type III atrial fibrillation. Mean flecainide dose was
140 mg.
Termination of atrial fibrillation example 1
Flecainide infusion in a patient with type III atrial
fibrillation resulted in progressively fewer wave fronts
with longer periods of right atrial electrical silence. This
then evolved into a single reentry wave front, which
encountered a line of block in the superior posteroseptal wall. The right atrium was then reactivated several times from the septum, presumably as breakthrough
from the left atrium. The length of these periods of
electrical silence and number of wave fronts present over
the period of flecainide infusion are plotted in Fig. 8.
Following one period of electrical silence, right atrial
tachycardia emerged as a uniform wave front from a
focus on the lateral right atrial wall in the region of the
crista terminalis. A second wave front emerged from a
similarly located focus, which was then followed by a
longer pause. This was then followed by slow regular
activity emerging from the same right atrial point, as
recorded during sinus rhythm, indicating the resumption
of sinus rhythm.
Non-contact endocardial mapping of human right atrial fibrillation
557
Figure 3 Three examples of the surface ECG with contact (C) and reconstructed
(R) electrograms recorded from the same endocardial location in patients with type
I (top), II (middle) and III (bottom) atrial fibrillation.
Termination of atrial fibrillation example 2
Flecainide infusion resulted in progressively fewer wave
fronts, until a single wave front remained rotating in a
macroreentry circuit around the right atrium, consistent
with typical atrial flutter. This terminated several times
and random reentry[29,30] was re-established by local
reactivation from the septum. After a further two rotations, this wave front then blocked in the region of the
TA-inferior vena cava isthmus, resulting in termination
of tachycardia and resumption of sinus rhythm.
Termination of atrial fibrillation example 3
In the remaining patient, an infusion of flecainide
resulted in organization and slowing of the wave front so
that a consistent atrial flutter reentry circuit was formed.
This then required overdrive pacing to terminate
tachycardia.
Procedural complications
No patient suffered complications as a result of the
mapping procedures. Postoperative investigations
remained unchanged and echocardiography revealed
no evidence of intracardiac thrombus, trauma or
pericardial effusion.
Discussion
We describe the validation and first results of noncontact mapping of human atrial fibrillation using a
previously validated system[28]. These data are the first
to provide maps of the entire right atrial endocardium
including the septum. Non-contact mapping of atrial
fibrillation in the human right atrium has confirmed the
presence of a wide variation in atrial fibrillation patterns
in different patients, which correspond with the three
broad classifications of types I, II and type III, as
described by Konings et al.[18]. The system has also
recorded and demonstrated right atrial activation
during termination of atrial fibrillation by intravenous
flecainide.
Drug therapy in atrial fibrillation is disappointing; the
success rate of chemical cardioversion of acute atrial
fibrillation is only 70–86%[31–33], whereas chronic
antiarrhythmic therapy has a high recurrence rate
and an associated mortality[8]. Developments of
potentially curative procedures have been based on
data from multi-site mapping studies reported by
Allessie et al.[4,5] which supported the multiple wavelet
hypothesis proposed by Moe et al. in 1959[1–3]. Cox
et al. used such mapping data, from patients undergoing
surgical ablation of Wolff–Parkinson–White syndrome,
to develop the surgical Maze procedure[9]. The
high success rate (100%) when using the most recent
modification in a highly selected group, has been tempered by a recurrence rate for atrial fibrillation or atrial
flutter of 7% and an associated morbidity and mortality[9]. Attempts to reproduce the Maze procedure
using percutaneous catheter techniques have so far had
low success rates, high recurrence rates and required
Eur Heart J, Vol. 21, issue 7, April 2000
558
R. J. Schilling et al.
(a)
I
aVR
V1
V4
II
aVL
V2
V5
III
aVF
V3
V6
RHYTHM STRIP: II
–1
–1
25 mm.s ; cm.mV
F
40
08523
LOC 00000–0000
(b)
I
aVR
V1
V4
II
aVL
V2
V5
III
aVF
V3
V6
VI
II
V5
Figure 4 (a) A surface ECG recorded prior to mapping of the patient with type I atrial fibrillation, showing
a coarse pattern of atrial activity. (b) A surface ECG recorded prior to mapping of a patient with type III
atrial fibrillation showing a fine pattern of atrial activity.
long procedure times[34–36]. Increased understanding of
the mechanisms underlying atrial fibrillation by detailed
mapping techniques is likely to improve these results.
Many mapping studies of atrial fibrillation have been
published to date[4–6,10–14]. Human studies have recorded
data from limited endocardial contact sites[19,23] or have
used epicardial electrode arrays to record high resolution data over limited areas[11,18,20,37] or low resolution
data over large areas[16,24], often in patients with
Eur Heart J, Vol. 21, issue 7, April 2000
normal atria who had atrial fibrillation induced during a
surgical procedure.
Validation of non-contact mapping of atrial
fibrillation
This mapping system has previously been validated in
the human left ventricle during sinus rhythm[29] and
Non-contact endocardial mapping of human right atrial fibrillation
559
Figure 5 (a) A series of unipolar activation maps, at 30 ms intervals, of type I atrial fibrillation showing
a macroreentry wave front rotating around a line of block indicated by the blue line. The wave front
passes superiorly up the septum (frame 1) and passes over the TV and line of block (frame 2). It then
passes inferiorly along the anterolateral aspect of the line of block (frames 3 to 5) before turning around
the inferior vena cava (frame 5) and passing up the septum (frame 6) to complete another circuit (frames
6 to 9). (b) Unipolar reconstructed electrograms taken from the points labelled on the activation maps a
to g showing sequential activation around the right atrium, also seen is the surface ECG lead II. The
yellow numbers indicate the points at which the corresponding sequential activation map frame has been
displayed. Note the double potentials seen on electrogram c which are closely spaced corresponding to the
electrogram’s proximity to the end of a line of block. By comparison, electrogram g has widely spaced
double potentials; the second potential (of lower amplitude and frequency) corresponds with activation of
electrogram d which is taken from the opposite side to the line of block from electrogram g.
ventricular tachycardia[38], and in the canine right
atrium during atrial fibrillation and atrial flutter[39]. The
low amplitude and fractionated electrograms recorded
during atrial fibrillation may not be as accurately
reconstructed as those recorded during sinus rhythm or
ventricular tachycardia. While satisfactory, the crosscorrelation data presented in this study are similar, but
not as good as that reported for cross-correlation of
Eur Heart J, Vol. 21, issue 7, April 2000
560
R. J. Schilling et al.
Figure 6 A series of unipolar activation maps of type II atrial fibrillation shown at 30 ms intervals.
After a period of electrical silence (frame 1), two simultaneous sources of right atrial activation are
seen from the superior and inferior septum (frame 2). The inferior wave front progresses inferiorly
towards the border of the right atrium and inferior vena cava. The superior wave front splits in the
region of the posterior septum, with one wave front turning at the septum and progressing superiorly
and the other progressing across the lateral right atrium (frames 3 and 4) before blocking in the
inferior vena cava-TC isthmus (frame 5). The septal wave front follows a complex path, finally
progressing interiorly (frame 5), turning (frame 6), passing around the superior vena cava (frames 6
to 7), moving inferiorly (frame 7) and blocking at the lateral right atrium (frame 8). The inferior
wave front progresses superiorly and blocks at the superior septum (frames 4 to 9). Electrical silence
(frame 10) is followed by reactivation of right atrium from the superior septum (frame 11).
contact and reconstructed electrograms in the fibrillating
canine right atrium[30] (0·8), and the human left ventricle
during sinus rhythm[28] (0·83) and ventricular tachycardia[38] (0·86). Possible reasons for this difference are the
increased size of the human right atrium compared with
the canine right atrium (resulting in an increased distance between the multielectrode array and the endocardium), the increased complexity of the atrial fibrillation
electrogram and the decreased amplitude of the atrial
electrogram.
Initiation of atrial fibrillation
There are little data pertaining to the mechanisms of
spontaneous initiation of atrial fibrillation[40]. One study
has published data describing spontaneous and adenosine triphosphate induced conversion of atrial flutter to
atrial fibrillation and atrial fibrillation to atrial flutter in
a canine sterile pericarditis model[13]. We have presented
data on only one patient with a history of paroxysmal
atrial fibrillation in whom atrial fibrillation was initiated
by catheter-induced right atrial ectopic beats. These
Eur Heart J, Vol. 21, issue 7, April 2000
resulted in the establishment of a reentry circuit which
encountered varying lines of block causing multiple
wave fronts and atrial fibrillation akin to previous
descriptions[13].
Mapping of atrial fibrillation
Atrial reentry
The range of activation patterns presented here are
compatible with previous studies[14,18–20]. We demonstrate three broad classifications of activation pattern
of atrial fibrillation ranging from a single reentry wave
front in the right atrium (type I), two simultaneous
wave fronts (type II) and three or more simultaneous
wave fronts (type III). This fits with the similar classification of Konings et al.[18] but it is of interest that the
majority of patients described in Konings’ study had a
predominantly type I pattern, whereas most of our small
number of patients had predominantly type III patterns.
This may be explained by bias resulting from the
small numbers included in our study, or by the global
Non-contact endocardial mapping of human right atrial fibrillation
561
Figure 7 (a) A series of activation maps, shown at 20 ms intervals, during type III atrial fibrillation.
Activity emerges from the coronary sinus septum and splits into three separate lateral, central and
medial wave fronts (frames 1 and 2). The centre of these wave fronts is slow and of low amplitude and
splits further into two wave fronts travelling superiorly and inferiorly (frame 3). The more inferior
wave front collides in the isthmus with the medial wave front that has rotated around the TV (frames
3 and 4) while the more superior wave front collides with a line of block established by earlier
activation by the medial wave front (frame 4). The lateral and medial wave fronts fuse in the
anterolateral right atrium (frames 2 and 3) and the resulting wave front divides in the superior right
atrium with one front rotating around the TV (frames 3 and 4) and the other rotating in the superior
septum before fusing with the central superior wave front (frames 4 and 5). Subsequent activation
maps show the continuing complex patterns of activation seen in type III atrial fibrillation. (b)
Surface ECG lead II and unipolar reconstructed electrograms from points a to e on the activation
maps are shown demonstrating the complexity of electrograms seen when compared to type I atrial
fibrillation (Fig. 4(b)). The points in time at which each activation map is made are shown by the
yellow numbers corresponding to each frame at the top of the figure.
Eur Heart J, Vol. 21, issue 7, April 2000
562
R. J. Schilling et al.
3
2
Pt 4
SR
1
0
Number of wave fronts
4
3
Pt 3 2
SR
1
AFL
0
4
3
Pt 2 2
SR
1
AFL
0
30
60
90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570
Time (s)
Figure 8 The number of wave fronts plotted against time (s) over the period of flecainide infusion for the
three patients in whom flecainide terminated atrial fibrillation. The patient number corresponds with the
numbering of Table 2. Onset of atrial flutter is indicated by an arrow with the legend AFL and restoration of
sinus rhythm is indicated by an arrow and the legend SR.
nature of our mapping technique which revealed the
presence of other wave fronts which may not have been
apparent to Konings’ group when mapping a small
atrial area.
The observation of type I right atrial activity during
atrial fibrillation is consistent with previously published
data from studies which performed simultaneous left
and right atrial mapping[16,18]. These describe wave
fronts penetrating the septum from a single reentry wave
front in the right atrium and reactivating the left
atrium[16]. A single reentry circuit during atrial fibrillation, which may be maintaining atrial fibrillation, is a
potential target for catheter ablation and may therefore
greatly simplify catheter ablation of atrial fibrillation.
However, the relevance of these single reentry circuits
to the maintenance of atrial fibrillation has not been
examined in this study.
Eur Heart J, Vol. 21, issue 7, April 2000
Focal atrial activation
Focal atrial activation during atrial fibrillation has been
described by several mapping studies[14,18,20,22,37,41,42].
The presence of micro-reentry presenting as focal activation can only be excluded by studies using highdensity mapping. Of these high density mapping studies,
the majority have used epicardial arrays applied during
surgery and have concluded that the focal activation
seen was a result of epicardial breakthrough of an atrial
fibrillation wave front propagating in a free running
atrial trabeculum[18,20,37]. One study has used simultaneous endo- and epicardial mapping in canines to
confirm the existence of this phenomenon[42].
The data presented in this paper demonstrate activity
emerging solely from the septum mainly in the region of
Bachmann’s bundle and the septum near the coronary
sinus where the orientation of atrial fibres might
Non-contact endocardial mapping of human right atrial fibrillation
encourage inter-atrial conduction. We surmise from
these data that this activity is the result of breakthrough
of activation from the left atrium. These conclusions are
supported by previous studies which have used simultaneous left and right atrial mapping in canines[14] and
humans[22] to identify breakthrough of activation from
one atrium to the other, in some cases resulting in
reactivation of unstable reentry circuits.
Termination of atrial fibrillation
In this study, intravenous flecainide administration during atrial fibrillation resulted in a progressive reduction
in the number of wave fronts within the right atrium,
culminating in a single reentry wave front. Three modes
of arrhythmia termination were then observed, firstly
the right atrium wave front continued to circulate,
resulting in an organized reentrant arrhythmia requiring
termination by pacing. Secondly the reentry wave front
encountered a line of block at a ‘vulnerable’ site in the
circuit (i.e. the inferior vena cava-TV isthmus); the right
atrium was then reactivated by wave fronts emerging
from the septum until sinus rhythm was restored in the
left atrium. Thirdly the reentry wave front blocked and
was followed by an atrial tachycardia, emerging from a
focus in the right atrium remote from the septum which
stopped before sinus rhythm was restored.
Previous studies have used either low resolution endocardial catheter mapping in humans[21] or high resolution
epicardial mapping of canines[10,12,13] and humans[17] to
demonstrate the mechanism of termination of a number
of different models of atrial fibrillation using a variety of
drugs. It has been shown that flecainide increases the size
and reduces the number of reentry circuits[10]. In addition, flecainide may convert atrial fibrillation to a regular tachycardia before terminating the arrhythmia[10,12,17]
and termination of atrial fibrillation may follow the
establishment of a macroreentry circuit[10]. Mechanisms
of termination of atrial fibrillation, similar to the cases
described here, have been demonstrated previously,
showing that failure of re-excitation of the right atrium
from the left atrium was important, and that termination
may be preceded by wave fronts emerging from a
focus[12]. A considerable interrelation between atrial fibrillation and atrial flutter and spontaneous conversion
between the two has also been shown previously[13],
which fits with our observations of the development of a
reentry circuit identical to right atrial flutter preceding
the termination of atrial fibrillation.
Limitations of study
This study does not examine activation within the left
atrium. It is well known that the conduction properties
of the left atrium and right atrium are inhomogeneous[44,45] and that the fibrillation rates may differ between the two[44]. We are therefore unable to draw any
563
conclusions about the activation patterns of the left
atrium and while it is likely that the local activation
emanating from the right atrial septum is likely to be left
atrial in origin, our data could not prove this.
Because these results describe initial studies to confirm
the feasibility of using a novel mapping technology to
study atrial fibrillation, the number of patients included
are limited.
Conclusion
Non-contact mapping of human right atrial fibrillation
has been validated. Identification of types I, II and III
atrial fibrillation using non-contact mapping of the
entire human right atrium confirms previous observations made by mapping limited areas of the right
atrium and identifies a wide variability in pathophysiology of atrial fibrillation. The ability to rapidly produce
high resolution simultaneous global maps of atrial
fibrillation offers the potential opportunity to customize
a catheter ablation procedure, according to the pattern of atrial fibrillation seen in an individual patient.
Further studies of the left atrium may confirm this
system’s potential for guiding catheter ablation of atrial
fibrillation.
This work was supported by the British Heart Foundation.
References
[1] Moe GK, Abildskov JA. Atrial fibrillation as a self sustaining
arrhythmia independent of focal discharge. Am Heart J 1959;
58: 59–70.
[2] Moe GK. On the multiple wavelet hypothesis of atrial fibrillation. Arch Int Pharmacodyn Ther 1962; 140: 183–8.
[3] Moe GK, Rheinboldt WC, Abilskov JA. A computer model
of atrial fibrillation. Am Heart J 1964; 64: 200–20.
[4] Allessie MA, Lammers WJEP, Smeets JRLM, Bonke FIM,
Hollen J. Total mapping of atrial excitation during
acetylcholine-induced atrial flutter and fibrillation in the isolated canine heart. In: Kulbertus HE, Olsson SB, Schlepper
M, eds. Atrial fibrillation. Molndal, Sweden: Lindgren and
Soner; 1982: 44–62.
[5] Allessie MA, Lammers WJEP, Bonke FIM, Hollen J. Experimental evaluation of Moe’s multiple wavelet hypothesis of
atrial fibrillation. In: Zipes DP, Jalife J, eds. Cardiac Arrhythmias. New York: Grune & Stratton, 1985: 265–76.
[6] Allessie MA, Kirchhof CJHJ, Bonke FIM. The role of the
sinus node in supra ventricular arrhythmias. In: Mazgalev T,
Dreifus LS, Michelson EL, eds. Electrophysiology of the
Sinoatrial and Atrioventricular Nodes. New York: Alan R
Liss, 1988: 53–66.
[7] Waldo AL. Mechanisms of atrial fibrillation, atrial flutter, and
ectopic tachycardia: a brief review. Circulation 1987; 75:
III-37–III-40.
[8] Flaker GC, Blackshear JL, McBride R, Kronmal RA,
Halperin JL, Hart RG. Antiarrhythmic drug therapy and
cardiac mortality in atrial fibrillation: the Stroke Prevention
in Atrial Fibrillation Investigators. J Am Coll Cardiol 1992;
20: 527–32.
[9] Cox JL, Schuessler RB, Lappas DG, Boineau JP. An 8Y year
clinical experience with surgery for atrial fibrillation. Ann
Surg 1996; 224: 267–75.
Eur Heart J, Vol. 21, issue 7, April 2000
564
R. J. Schilling et al.
[10] Wang Z, Page P, Nattel S. Mechanism of flecainide’s antiarrhythmic action in experimental atrial fibrillation. Circ Res
1992; 71: 271–87.
[11] Kirchhof C, Chorro F, Scheffer GJ et al. Regional entrainment of atrial fibrillation studied by high-resolution mapping
in open-chest dogs. Circulation 1993; 88: 736–49.
[12] Wang J, Bourne GW, Wang Z, Villemaire C, Talajic M,
Nattel S. Comparative mechanisms of antiarrhythmic drug
action in experimental atrial fibrillation importance of usedependent effects on refractoriness. Circulation 1993; 88:
1030–44.
[13] Ortiz J, Niwano S, Abe H, Rudy Y, Johnson NJ, Waldo AL.
Mapping the conversion of atrial flutter to atrial fibrillation
and atrial fibrillation to atrial flutter: insight into mechanisms.
Circ Res 1994; 74: 882–94.
[14] Kumagai K, Khrestian C, Waldo AL. Simultaneous multisite
mapping studies during induced atrial fibrillation in the sterile
pericarditis model: insights into the mechanism of its maintenance. Circulation 1997; 95: 511–21.
[15] Allessie MA, Wijffels MCEF, Kirchhof CJHJ. Experimental
models of arrhythmias: Toys or truth? Eur Heart J 1994; 15
(Suppl A): 2–8.
[16] Cox JL, Canavan TE, Schuessler RB et al. The surgical
treatment of atrial fibrillation II. Intraoperative electrophysiologic mapping and description of the electrophysiologic basis
of atrial flutter and atrial fibrillation. J Thorac Cardiovasc
Surg 1991; 101: 406–26.
[17] Brugada J, Gursov S, Brugada P, Atie J, Guiradon G, Andries
E, Cibenzoline transforms random reentry into ordered reentry in the atria. Eur Heart J 1993; 14: 267–72.
[18] Konings KTS, Kirchhof CJHJ, Smeets JRLM, Wellens HJJ,
Penn OC, Allessie MA. High density mapping of electrically
induced atrial fibrillation in humans. Circulation 1994; 89:
1665–80.
[19] Daoud EG, Pariseau B, Niebauer M et al. Response of type I
atrial fibrillation to atrial pacing in humans. Circulation 1996;
94: 1036–40.
[20] Holm M, Johansson R, Brandt J, Luhrs C, Olsson SB.
Epicardial right atrial free wall mapping in atrial fibrillation:
documentation of repetitive activation with a focal spread — a
hitherto unrecognised phenomenon in man. Eur Heart J 1997;
18: 290–310.
[21] Sih HJ, Ropella KM, Swiryn S, Gerstenfeld EP, Sahakian
AV. Observations from intraatrial recordings on the termination of electrically induced atrial fibrillation in humans.
PACE 1994; 17: 1231–42.
[22] Harada A, Sasaki K, Fukushima T et al. Atrial
activation during chronic atrial fibrillation in patients with
isolated mitral valve disease. Ann Thorac Surg 1996; 61:
104–12.
[23] Jais P, Haissaguerre M, Shah DC, Chouairi S, Clementy J.
Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. PACE 1996; 19: 1998–2003.
[24] Chang BC, Schuessler RB, Stone CM et al. Computerized
activation sequence mapping of the human atrial septum. Ann
Thorac Surg 1990; 49: 231–41.
[25] Josephson ME, Seides SF. Clinical cardiac electrophysiology.
Technique and interpretations, 1st edn. Philadelphia, PA: Lea
& Febiger, 1979; 198–9.
[26] Hewlett AW, Wilson FN. Coarse auricular fibrillation in man.
Arch Int Med 1915; 15: 786–93.
[27] Nelson RM, Jenson CB, Davis RW. Differential atrial
arrhythmias in cardiac surgical patients. J Thorac Cardiovasc
Surg 1969; 58: 581–7.
Eur Heart J, Vol. 21, issue 7, April 2000
[28] Schilling RJ, Peters NS, Davies DW. A non-contact catheter
for simultaneous endocardial mapping in the human left
ventricle: comparison of contact and reconstructed electrograms during sinus rhythm. Circulation 1998; 98: 887–98.
[29] Hoffman BF, Rosen MR. Cellular mechanisms for cardiac
arrhythmias. Circ Res 1981; 49: 1–15.
[30] Allessie MA, Kirchhof CJCHJ, Konings KTS. Unravelling
the mysteries of atrial fibrillation. Eur Heart J 1996; 17 (Suppl
C): 2–9.
[31] Crijins HJGM, van Wiljk LM, van Gilst WH, Kingma JH,
van Gelder IC, Lie KI. Acute conversion of atrial fibrillation
to sinus rhythm: Clinical efficacy of flecainide acetate: Comparison of two regimens. Eur Heart J 1988; 9: 634–8.
[32] Borgeat A, Goy JJ, Maendly R, Kaufmann U, Grbic M,
Sigwart U. Flecainide versus quinidine for conversion of atrial
fibrillation to sinus rhythm. Am J Cardiol 1986; 58: 496–8.
[33] Bianconi L, Boccadamo R, Pappalardo A, Gentili C, Pistolese
M, Effectiveness of intravenous propafenone for conversion of
atrial fibrillation and flutter of recent onset. Am J Cardiol
1989; 64: 335–8.
[34] Avitall B, Hare J, Mughal K et al. Ablation of atrial fibrillation in a dog model (Abstr). J Am Coll Cardiol 1994; 23:
276A.
[35] Kempler P, Littmann L, Chuang CH et al. Radiofrequency
ablation of the right atrium: acute and chronic effects (Abstr).
PACE 1994; 17: 797.
[36] Elvan A, Pride HP, Zipes DP. Replication of the ‘maze’
procedure by radiofrequency catheter ablation reduces the
ability to induce atrial fibrillation (Abstr). PACE 1994; 17:
274.
[37] Holm M, Johansson R, Olsson SB, Brandt J, Luhrs C. A new
method for analysis of atrial activation during chronic atrial
fibrillation in man. IEEE Transactions on biomedical
engineering 1996; 43: 198–210.
[38] Schilling RJ, Peters NS, Davies DW. Feasibility of noncontact catheter for endocardial mapping of human ventricular tachycardia. Circulation 1999; 99: 2543–52.
[39] Kadish A, Hauck J, Pederson B, Beatty G, Gornick C.
Mapping of atrial activation with a noncontact, multielectrode catheter in dogs. Circulation 1999; 99: 1906–13.
[40] Ferguson TB, Schuessler RB, Hand DE, Boineau JP, Cox JL.
Lessons learned from computerized mapping of the atrium.
Surgery for atrial fibrillation and atrial flutter. J Electrocardiol
1993; 26: 210–19.
[41] Haissaguerre M, Marcus FJ, Fischer B, Clementy J. Radiofrequency catheter ablation in unusual mechanisms of atrial
fibrillation: Report of three cases. J Cardiovasc Electrophysiol
1994; 5: 743–51.
[42] Schuessler RB, Kawamoto T, Hand DE et al. Simultaneous epicardial and endocardial activation sequence
mapping in the isolated canine right atrium. Circulation 1993;
88: 250–63.
[43] Allessie MA, Kirchhof CJHJ, Scheffer GJ, Chorro FJ,
Brugada J. Regional control of atrial fibrillation by rapid
atrial pacing in conscious dogs. Circulation 1991; 84: 1689–
97.
[44] Alessi R, Nusynwitz M, Abildskov JA, Moe GK. Nonuniform distribution of vagal effects on the atrial refractory
period. Am J Physiol 1958; 194: 406–10.
[45] Han J, Millet D, Chizzonitti B, Moe GK. Temporal dispersion
of recovery of excitability on atrium and ventricle as a
function of heart rate. Am Heart J 1966; 71: 481–7.