Europace (2000) 2, 193–200
doi:10.1053/eupc.2000.0102, available online at http://www.idealibrary.com on
How to map and ablate atrial scar macroreentrant
tachycardia of the right atrium
F. G. Cosio, A. Pastor, A. Núñez and M. A. Montero*
Cardiology Service, Hospital Universitario de Getafe, Madrid, Spain
A special form of macroreentrant atrial tachycardia
(MRAT), due to reentrant activation around surgical scars,
can occur in patients after cardiac surgery. Scar MRAT
occurs usually after correction of congenital defects, such as
atrial or ventricular septal defects, and especially after
Mustard, Senning or Fontan procedures, but it can occur
also after myxoma, valvular or coronary bypass surgery.
The simplest form of scar MRAT is reentry around a
lateral right atrial surgical scar. A basic mapping array with
multiple simultaneous recordings from the anterior and
septal right atrium is very useful to make the electrophysi-
ological diagnosis. A line of double electrograms can be
mapped in the centre of the circuit and a fragmented
electrogram usually marks the pivoting point between the
inferior end of the scar and the inferior vena cava (IVC).
Extension of the scar toward the closest fixed obstacle,
usually the IVC, by means of radiofrequency ablation,
can interrupt the tachycardia and make it non-inducible.
Introduction
In recent years mapping and ablation studies have
defined macroreentry as the mechanism of the majority
of atrial tachycardias, including flutter[1–3]. Macro-
reentrant atrial tachycardia (MRAT) circuits have a
large diameter (several cm) and well defined central
obstacles and peripheral boundaries, and understanding
of atrial anatomy is essential to locate and characterize
the reentrant pathway. Mapping and ablation of MRAT
needs a specific approach, similar to that used to under-
stand flutter circuits. We will try to explain how acti-
vation mapping and entrainment based on anatomical
structure can lead to the diagnosis and effective treat-
ment of MRAT based on surgical scars. Atrial arrhyth-
mias after orthotopic cardiac transplantation pose
specific problems and will not be covered.
Manuscript accepted 14 April 2000.
*Dr Montero was supported by a Grant from Medtronic Ibérica,
S.A. Correspondence: Francisco G. Cosı́o, MD, Chief Cardiology
Service, Hospital Universitario de Getafe, Carretera de Toledo,
km 12.5, 28905 Getafe, Madrid, Spain.
1099–5129/00/030193+08 $35.00/0
Typical atrial flutter usually coexists with scar MRAT and
flutter isthmus ablation is probably indicated in all cases.
In patients having undergone baffle atrial surgery it can
be impossible to map the whole circuit and entrainment-
mapping is helpful to localize critical isthmuses in the
circuit. After the Fontan operation the right atrium can be
severely dilated and scarred, and multiple, complex reentry
circuits can be found. Left atrial MRAT based on large
areas of scar has been described, but there is still too little
experience with these to propose general rules for diagnosis
and management.
(Europace 2000; 2: 193–200)
 2000 The European Society of Cardiology
Key Words: Atrial macroreentrant tachycardia, cardiac
surgery, atrial reentry, radiofrequency ablation, atrial
mapping.
Atrial macroreentry vs ventricular
tachycardia
We think it may be useful to start with a warning note
because, in our opinion, direct application of concepts,
methods and terms successfully applied to diagnosis and
treatment of ventricular tachycardia can be a significant
impediment to understanding MRAT. Terms such as
diastolic activity or entry and exit to a slow conduction
zone are either inapplicable or applicable only in very
qualified terms in MRAT. In post-infarction ventricular
tachycardia there is an area of scar with surviving
myocardial fibres separated by fibrous tissue that main-
tains reentrant activation during diastole, i.e. outside the
QRS complex. Activation of this part of the circuit is
slow and can be recorded only by direct endocardial
or epicardial mapping and this has led to a division of
the cycle length into systole and diastole, based on the
electrocardiogram (ECG)[4,5]. The scar also contains the
more circumscribed anatomical isthmuses where radio-
frequency ablation can interrupt the reentrant circuit.
The concepts of entry and exit pointing to the slow
 2000 The European Society of Cardiology
194 F. G. Cosio et al.
conduction zone are relevant, as they may coincide with
anatomical isthmuses.
Scar MRAT is similar to typical flutter and differs
from ventricular tachycardia in that the reentry circuit
involves large parts of the right atrium[1,2] with a rela-
tively uniform conduction velocity, and a critical dis-
eased area concealing activation may not be present or
be small. Sequential endocardial mapping readily
records activation, covering the whole cycle length (CL)
with uniform voltage throughout the circuit, except for
specific areas in the centre, where the scar is located. The
concept of isthmus is mainly anatomical, i.e. a narrow
conducting structure between fixed obstacles, even
though conduction can be slow in the isthmic area. The
useful clues to understanding MRAT are the definition
of the fixed central obstacles (either anatomical or
functional) and the isthmus or isthmuses. The concepts
(and terms) diastolic potentials, or entrance and exit
points are not applicable.
Good tolerance of the arrhythmia makes sequential
mapping feasible in most cases, with the only limitation
being MRAT stability. Pacing studies to determine
post-entrainment return cycles at selected sites serve to
validate the activation map. Finally, interruption of
MRAT by radiofrequency ablation of a critical isthmus
proves that the mechanistic hypothesis was right.
The substrates of scar MRAT
Scar tachycardia is particularly troublesome after atrial
baffle procedures (Mustard and Senning) for redirecting
venous blood flow in transposition of the great arteries,
because the complex scar lines can result in various
MRAT circuits[6,7]. Fontan procedures are also com-
monly followed by scar MRAT, and in this case atrial
dilatation is an added factor contributing to abnormal
activation[8,9]. The presence of intra-atrial baffles after
Mustard or Senning procedures can distort anatomy
and prevent complete circuit mapping.
In adults, scar MRAT usually has a surgical incision
in the lateral right atrium as the central obstacle[1,2,10]. It
is most common after even simple surgery for atrial or
ventricular septal defects, but it has been described after
mitral or arrhythmia surgery[11]. Reentry around septal
patches has been described infrequently[1]. The circuit is
generally circumscribed to the anterolateral right atrial
wall and an isthmus can be found between the scar and
the inferior vena cava (IVC) or tricuspid valve. Local
slow conduction can be found in this isthmic region.
Mapping and entrainment can delineate the circuit, as in
flutter, to direct radiofrequency ablation. We have
treated nine cases of adult scar MRAT with lateral wall
right atrial circuits. Four had had surgery for atrial
septal defect, one for ventricular septal defect, three for
myxoma and one coronary bypass.
Scar MRAT often coexists with typical flutter. The
presence of the scar as a fixed barrier in the superior–
inferior direction probably facilitates formation of the
large obstacle, including the superior vena cava and
Europace, Vol. 2, July 2000
IVC[12] characteristic of typical flutter. In all but one
of our cases this coexistence was documented before
ablation, and, in the remainder, flutter appeared after
ablation of scar MRAT. Both flutter and scar MRAT
use the anterior wall as part of the circuit; however, the
IVC–tricuspid isthmus is not part of the scar MRAT
circuit (Fig. 1).
Large areas of electrical silence or low voltage electro-
grams suggesting atrial scarring have been described at
the centre of left atrial MRAT circuits in patients
without previous atrial surgery[13]. We have studied a
case of MRAT around a central area of low voltage
electrograms in the lateral right atrium in a patient
without previous cardiac surgery. The nature of these
abnormal electrogram zones of atrial myocardium in the
absence of surgical trauma remains to be determined.
Mapping technique
The ECG of scar MRAT can be suggestive of flutter
(continuous undulation) or focal tachycardia (discrete P
waves with isoelectric segments) and in many cases more
than one ECG pattern can be documented. Only
endocardial or epicardial mapping can accurately define
the mechanism. Mapping has to be anatomically
accurate, i.e. referred to a true atrial structure and
position, because activation paths and isthmuses are
defined in reference to anatomical landmarks. In this
respect it is important to discard topographic terms
commonly used for accessory pathways and AV
junctions, where inferior (caudal) is called posterior and
superior (cranial) is called anterior. Only an anatomi-
cally accurate nomenclature[14] referring to the erect
body can describe MRAT circuits in relation to atrial
structure. An added advantage in using this anatomi-
cally accurate nomenclature is that it matches positions
observed directly on the fluoroscopic views used in the
laboratory, as these show the chest upright.
Mapping atrial activation in the superior–inferior
direction is essential, because both MRAT and typical
flutter circuits have this main axis. In typical (counter-
clockwise) and reversed (clockwise) typical flutter the
terminal crest, acting as a functional line of block, is
responsible for this axis of activation. In scar MRAT the
superior–inferior axis can be due to the direction of the
surgical incisions, and/or the terminal crest posteriorly
and the tricuspid valve anteriorly. A single catheter with
multiple electrode pairs (24 electrodes, 2 mm intra-pair,
7–9 mm inter-pair separation), coming from the IVC to
the right atrial roof, then descending on the anterior
wall, can serve to obtain data from both the antero-
lateral wall and the posteroseptal wall in a superior–
inferior direction. Appropriate display of the recordings
on the screen can give an instant picture of the activation
sequence.
Changes from scar MRAT to typical flutter and
vice-versa are common during mapping, entrainment
and/or radiofrequency ablation and can be easily
recognized when recordings from the septal and anterior
 Scar macroreentrant tachycardia of the right atrium 195

Figure 1 Typical flutter and scar MRAT in a patient after surgery for atrial septal defect.
From top to bottom recordings are lead II and endocardial right atrial electrograms from the
high anterior wall (HA), the mid anterior wall (MA), the low anterior wall (LA), the
IVC–tricuspid isthmus (CTI) and the low posterior septum (LPS), the mid posterior septum
(MPS) and the high posterior septum (HPS). The schemas at the bottom show the reentrant
mechanism. On the left typical flutter shows counterclockwise right atrial activation, and the
electrogram sequence descending in the anterior wall and ascending in the septal wall. The
CTI bridges the interval between LA and LPS. On the right scar MRAT is centred on
the lateral wall. Note, anterior wall activation is very similar to flutter, but there is a change in
the CTI and septal activation sequence. Note that almost the whole cycle length is covered
by recorded electrograms in both instances. Values in ms.

right atrium are monitored. The first documented
arrhythmia in the laboratory can be flutter, and scar
MRAT may be documented only after flutter ablation
or entrainment, by a subtle change in the septal acti-
vation sequence, with little or no significant CL change
(Fig. 1). The anterior wall activation sequence can be
almost identical in scar MRAT and flutter, however
septal activation will often not be inferosuperior in scar
MRAT, and the electrogram from the IVC-tricuspid
valve isthmus will not be in sequence between the low
anterior and low septal electrograms, as during typical
flutter (Fig. 1).
A hallmark of scar MRAT is the recording of a
superior–inferior line of double electrograms over the
location of the scar (Fig. 2). Voltage of at least one of
the components can be quite low. Fragmented electro-
grams can be recorded sometimes during sinus rhythm
along this line. Spike separation decreases toward the
inferior pivoting point, generally close to the IVC, and a
continuous fragmented electrogram can often be re-
corded at this point, suggesting local slow conduction
(Fig. 2). If the superior vena cava is incorporated in the
central obstacle, wide separation of double potentials
can be recorded in the high lateral right atrium, close to
the superior vena cava.
Entrainment mapping
After mapping has been followed by a hypothetical
circuit description, entrainment is used to confirm or
rule-out participation of specific sites in the circuit core.
Entrainment should be used sparingly because scar
MRAT is often unstable and can be altered or inter-
rupted by pacing. Careful attention should be paid to
activation sequence after each pacing run, because
pacing can also change activation from flutter to scar
MRAT or vice-versa. Furthermore, the longest pacing
CL capable of entraining (15–25 ms below baseline
CL) should be used in order to avoid induction of
conduction delay within the circuit that could alter
interpretation of return cycles.
If one or two sites on the anterior wall show return
CL<20 ms longer than baseline CL this confirms that

Europace, Vol. 2, July 2000

196 F. G. Cosio et al.

Figure 2 Double electrograms recorded from the lateral right atrium during scar
MRAT in the same patient as in Fig. 1. Five panels show, from top to bottom, lead II
and endocardial right atrial electrograms from the high anterior wall (HA), the mid
anterior wall (MA), the low anterior wall (LA) as fixed references. The last electrogram
is recorded at five levels of the lateral wall from close to the superior vena cava (L1) to
close to the IVC (L5). The right atrial schema at the bottom shows the location of the
sequential recording sites and the direction of reentry. Note double electrograms along
the supposed superior–inferior surgical scar. Electrogram separation decreases toward
the IVC, where a prolonged, fragmented electrogram is recorded. Note that scar MRAT
cycle length spontaneously became longer here than in the initial recording shown in
Fig. 1. Values in ms.

MRAT involves the anterior right atrium (Fig. 3). If
return CLs are >20 ms longer than the baseline CL at
the low septum or IVC–tricuspid isthmus, this rules out
typical flutter (Fig. 4). Pacing some areas of the circuit
involved with scar can be difficult, but logical approxi-
mations can be made if return CLs pacing the posterior
right atrial wall are closer to the basic CL than pacing
the septal wall.
Conversely, entrainment can help rule out participa-
tion of surgical scars in a flutter circuit, when conduc-
tion slowing around a scar alters the local activation
sequence, suggesting a false pivoting point of activation
around the scar (Fig. 5). In these cases, the local return
cycle will be long, ruling out the suggested participation
in the reentry circuit.
Some authors have defined anatomical isthmuses in
the scar MRAT circuit as the sites where pacing results
in concealed entrainment (paced atrial deflections are
identical to baseline, indicating the absence of fusion).
However, this technique can be very difficult to use
because of low voltage P waves and QRS-T overlap.
Nevertheless, it could be a useful adjunct particularly
when atrial baffles prevent complete activation
mapping[8,11].
Ablation of scar MRAT
Defining a critical isthmus for radiofrequency ablation
Actual noting of activation times and drawing the
hypothetical circuit on an anatomical scheme of
the atrium helps its relationship with fixed anatomical
obstacles to be understood (tricuspid valve, superior
vena cava, IVC). In our experience most scar MRATs
have an inferior isthmus coinciding with the low pivot-
ing point of activation, where radiofrequency ablation
can effectively interrupt the circuit. At this site double
electrograms may fuse into one wide fragmented electro-
gram suggesting local slow conduction. In only one case
was activation difficult to trace during significant seg-
ments of CL (30%) around this pivoting point, suggest-
ing concealed activation through extensive scarring,
similar to post-infarction ventricular tachycardia. When
the isthmus is very small, MRAT can be interrupted by

Europace, Vol. 2, July 2000

 Scar macroreentrant tachycardia of the right atrium 197

Figure 3 Entrainment mapping of scar MRAT in the patient in Figs 1 and 2. Lead II and
electrograms from HA, MA and LA are shown as in Fig. 2. The bottom electrogram is from
the pacing site. The schema on the left indicates the pacing sites and the location of the scar.
First post-pacing cycle length is equal to following cycle lengths both at the right atrial ‘roof’
(R) and the mid anterior wall (MA), confirming the participation of these sites in the circuit
core. For further explanation see text. S=stimulus artifact. Values in ms.

Figure 4 Entrainment mapping of scar MRAT in the patient in Figs 1–3. Lead II and
electrograms from HA, MA and LA are shown as in Figs 2 and 3. The bottom electrogram
is from the pacing site. The schema on the bottom indicates the pacing sites. First
post-pacing cycle length is definitely longer (>40 ms) both at the IVC–tricuspid isthmus
(CTI) and the low septum (LPS), ruling out participation of these sites in the circuit core.
For further explanation see text. S=stimulus artifact. Values in ms.

Europace, Vol. 2, July 2000

198 F. G. Cosio et al.

Figure 5 Entrainment of a false isthmic site in a patient with common flutter and a history
of mitral valve surgery. From top to bottom recordings are lead II and endocardial right atrial
electrograms from the high anterior wall (HA), the mid anterior wall (MA), the low anterior
wall (LA), the low lateral wall (LL), the low posterior septum (LPS), the mid posterior
septum (MPS) and the high posterior septum (HPS). On the right, after pacing, note that the
LL electrogram appears to bridge the gap between the LA and the LPS; however, the first
post-pacing pause at this site is 90 ms longer than baseline cycle length, ruling out
participation of LL in the circuit core. For further explanation see text. S=stimulus artifact.
Values in ms.

catheter pressure (Fig. 5), making it necessary to
deliver radiofrequency ablation during sinus rhythm
with anatomical guidance. Once the isthmus is defined it
is necessary to draw a line of applications to link the scar
with a fixed anatomical obstacle. In our cases this was
usually the IVC, but the tricuspid valve has been the
obstacle chosen by others[10].
Ablation endpoints
Tachycardia interruption and non-inducibility are nec-
essarily the end-points. There are no standard methods
to ensure complete conduction block on potentially
critical areas of the circuit. How aggressive induction
protocols should be is open to question. We use up to
three extrastimuli over CL 600-500-400-350 and burst
pacing to loss of 1:1 capture with stimulus strength twice
threshold. If induction was very reproducible at baseline
we tended to use a less aggressive protocol. Interestingly,
atrial fibrillation appears hard to induce in these
patients, perhaps because the scar plus the ablation lines
help prevent reentry breaking into multiple fronts.
An added end-point should be, in our opinion, com-
plete IVC–tricuspid isthmus block. All our patients but
one had typical flutter documented before and/or during
the radiofrequency ablation procedure, and the one that
did not had typical flutter clinically a few months later
and had to undergo a second procedure. Criteria
for IVC–tricuspid isthmus block have been clearly
defined[15,16]. Mapping and radiofrequency ablation of
both scar MRAT and flutter can become a very long
procedure, depending on operator and team experience.
We postponed part of the procedure to a second day if
3–4 h was exceeded, depending on patient tolerance.
Post ablation prognosis
The post radiofrequency ablation prognosis of adult
scar MRAT is good in our experience. We have had
no scar MRAT recurrences in eight of our cases in whom
scar MRAT was rendered non-inducible. We have, how-
ever, observed typical flutter recurrences. This might
suggest that myocardial thickness is less extensive in the
MRAT isthmus than in the IVC–tricuspid isthmus. This
favourable prognosis cannot be expected in cases with
complex congenital defects, after Mustard, Senning or
Fontan procedures[7,8]. There is very little experience
with ablation in patients with large scars of unknown
origin, such as described in left atrial MRAT[13].
The myocardial substrate is unknown in these cases

Europace, Vol. 2, July 2000

 Scar macroreentrant tachycardia of the right atrium 199

Figure 6 Interruption of scar MRAT by catheter pressure during mapping. From top to
bottom recordings are lead II and endocardial right atrial electrograms from the high anterior
wall (HA), the mid anterior wall (MA), the low anterior wall (LA), the low lateral wall (LL),
the low posterior septum (LPS), the mid posterior septum (MPS) and the high posterior septum
(HPS). The first six cycles show scar MRAT. Note descending anterior wall activation and
simultaneous activation of MPS and HPS with a descending sequence from MPS to LPS. The
LL electrogram, at the low end of the scar, shows a double deflection of changing morphology.
Interruption of scar MRAT coincides with the disappearance of one of the components of the
LL electrogram. The first post-MRAT beat is of sinus origin (note descending activation of
anterior and septal walls) and a fragmented electrogram is recorded at LL. For further
explanation see text. Values in ms.
and elimination of one circuit might not rule out the [2] Cosio FG, Arribas F, López-Gil M, Palacios J. Atrial flutter
possibility of other circuits developing later in other mapping and ablation. I. Atrial flutter mapping. PACE 1996;
19: 841–53.
parts of the atria. [3] Cosio FG, Arribas F, López-Gil M, González D. Atrial flutter
mapping and ablation II. Radiofrequency ablation of atrial
flutter circuits. PACE 1996; 19: 965–75.
Conclusions [4] De Bakker JMT, van Capelle FJL, Janse MJ, et al. Slow
conduction in the infarcted human heart. ‘‘Zigzag’’ course of
Scar MRAT in adults generally occurs after surgery for activation. Circulation 1993; 88: 915–26.
atrial or ventricular septal defect or myxoma and, less [5] Stevenson WG, Weiss JN, Wiener I, Nademanee K. Slow
commonly, after valvular or coronary bypass surgery. It conduction in the infarct scar: relevance to the occurrence,
is generally possible to define the circuit in relation to detection, and ablation of ventricular reentry circuits resulting
from myocardial infarction. Am Heart J 1989; 117: 452–67.
atrial anatomy by endocardial mapping and entrain- [6] Cronin CS, Nitta T, Mitsuno M, et al. Characterization and
ment of selected sites. Ablation of an isthmus, generally surgical ablation of acute atrial flutter following the Mustard
between the inferior end of the scar and the IVC is procedure. A canine model. Circulation 1993; 88(Suppl 2):
usually successful in interrupting the tachycardia and 461–71.
making it non-inducible. Recurrences appear to be rare. [7] Van Hare GF, Lesh MD, Ross BA, Perry JC, Dorostkar.
Mapping and radiofrequency ablation of intraatrial reentrant
When scar MRAT is due to atrial baffle procedures or tachycardia after the Senning or Mustard procedure for
atrial dilatation after the Fontan procedure, MRAT can transposition of the great arteries. Am J Cardiol 1996; 77:
be a more difficult challenge because of circuit complex- 985–91.
ity and the difficulty of mapping in the presence of [8] Triedman JK, Saul JP, Weindling SN, Walsh EP. Radio-
distorted anatomy. frequency ablation of intra-atrial reentrant tachycardia after
surgical palliation of congenital heart disease. Circulation
1995; 91: 707–14.
[9] Deal BJ, Mavroudis C, Backer CL, Johnsrude CL, Rocchini
References AP. Impact of arrhythmia circuit cryoablation during Fontan
conversion for refractory atrial tachycardia. Am J Cardiol
[1] Lesh MD, Van Hare GF, Epstein LM, et al. Radiofrequency 1999; 83: 563–9.
catheter ablation of atrial arrhythmias: results and mech- [10] Kalman JM, Van Hare GF, Olgin JE, Saxon LA, Stark SI,
anisms. Circulation 1994; 89: 1074–89. Lesh MD. Ablation of ‘incisional’ reentrant atrial tachycardia

Europace, Vol. 2, July 2000

200 F. G. Cosio et al.
complicating surgery or congenital heart disease. Circulation
1996; 93: 502–12.
[11] Baker BM, Lindsay B, Bromberg BI, Frazier DW, Cain ME,
Smith JM. Catheter ablation of clinical intraatrial reentrant
tachycardias resulting from previous atrial surgery: localizing
and transecting the critical isthmus. J Am Coll Cardiol 1996;
28: 411–17.
[12] Rosenblueth A, Garcı́a Ramos J. Estudios sobre el flútter y la
fibrilación. II. La influencia de los obstáculos artificiales en el
flútter auricular experimental. Arch Inst Cardiol Mex 1947;
17: 1–19.
[13] Jas P, Haı̈ssaguerre M, Shah DC, Takahashi A, Lavergne T,
Clementy J. A new electrophysiological substrate for
spontaneous left atrial flutters (Abstr). PACE 1998; 21: 794.
Europace, Vol. 2, July 2000
[14] Working Group Report. Living anatomy of the atrioventricu-
lar junctions. A guide to electrophysiologic mapping. A con-
sensus Statement from the Cardiac Nomenclature Study
Group, Working Group of Arrhythmias, European Society of
Cardiology, and the Task Force on Cardiac Nomenclature
from NASPE. Eur Heart J 1999; 20: 1068–75.
[15] Poty H, Saoudi N, Abdel Aziz A. Radiofrequency ablation
of type 1 atrial flutter: Prediction of late success by electro-
physiological criteria. Circulation 1995; 92: 1389–92.
[16] Shah DC, Takahashi A, Jaı̈s P, Hocini M, Clementy J,
Haı̈ssaguerre M. Local electrogram-based criteria of cavotri-
cuspid isthmus block. J Cardiovasc Electrophysiol 1999; 10:
662–9.