European Heart Journal (2002) 23, 558–566
doi:10.1053/euhj.2001.2841, available online at http://www.idealibrary.com on
Efficacy of an additional MAZE procedure using
cooled-tip radiofrequency ablation in patients with
chronic atrial fibrillation and mitral valve disease
A randomized, prospective trial
T. Deneke1, K. Khargi2, P. H. Grewe1, A. Laczkovics2, S. von Dryander1, T. Lawo1,
K.-M. Mu ¨ller3and B. Lemke1
1Department of Cardiology,2Clinic for Cardiothoracic Surgery,3Institute of Pathology, University Hospital
Bergmannsheil, Bochum, Germany
Aims This study is the first prospective randomized trial
evaluating the efficacy of an antiarrhythmic surgical pro-
cedure in patients with chronic atrial fibrillation undergoing
mitral valve replacement.
Methods and Results Thirty consecutive patients with
chronic atrial fibrillation undergoing mitral valve replace-
ment were randomized for an additional modified MAZE-
operation using intra-operatively cooled-tip radiofrequency
ablation (group A) or mitral valve replacement alone
(group B). Biatrial contraction was studied and functional
capacity was evaluated in spiro-ergometry 6 months after
surgery. Thirty-day mortality was 0% in both groups. After
12 months, sinus rhythm was reinstituted significantly more
often in patients of group A (cumulative rate of sinus
rhythm 0·800) compared to patients in group B (0·267)
(P<0·01). 66·7% of patients in sinus rhythm of group A had
showed long-term success in only 17% of patients in group
A and 0% in group B. Maximal aerobic uptake at the
6-month spiro-ergometry revealed no significant difference
(9·3 vs 8·5 ml . min?1kg?1, P=0·530).
Conclusions A modified MAZE operation using cooled-
tip radiofrequency ablation can be safely combined with
mitral valve surgery and is highly effective in restoring sinus
rhythm. Biatrial contraction is found in 66·7% of patients
with sinus rhythm undergoing mitral valve replacement
plus the MAZE operation.
(Eur Heart J 2002; 23: 558–566, doi:10.1053/euhj.2001.2841)
? 2001 The European Society of Cardiology
Key Words: MAZE surgery, cooled-tip radiofrequency
ablation, sinus rhythm,
contraction, functional results
See page 514, doi:10.1053/euhj.2001.3016 for the Editorial
comment on this article
Atrial fibrillation as the most common arrhythmia has a
prevalence of 0·4% which increases to 9% in patients
aged 80 years and older[1,2]. In patients undergoing
mitral valve surgery the prevalence of atrial fibrillation is
reported to be as high as 80%. Therapeutic strategies
(pharmacological, catheter ablation, antiarrhythmic sur-
gery) for atrial fibrillation aim to diminish uncomfort-
able symptoms, such as tachycardias and palpitations,
by restoring sinus rhythm, improving haemodynam-
ics, reinstituting atrioventricular synchrony and reduc-
The MAZE operation, as an open heart surgical
procedure, was introduced by James L. Cox in 1987 as
a ‘cut-and-suture’ technique, a technique based on a
multiple wavelet theory, which proposes that different
depolarizing wave-fronts circle the atria. MAZE inci-
sions should reduce the atrial mass below the critical
reentry circuit size, so preventing atrial fibrillation[4–7,10].
MAZE surgery has been proven to be highly effective
in restoring sinus rhythm. In patients with concomitant
mitral valve disease, sinus rhythm can be restored in
Revision submitted 15 June 2001, accepted 27 June 2001, and
published online 2 October 2001.
Correspondence: Thomas Deneke, Department of Cardiology,
Bergmannsheil Bochum, University of Bochum, Bu ¨rkle-de-la-
Camp-Platz 1, 44789 Bochum, Germany.
? 2001 The European Society of Cardiology
80% to 90%. In patients with lone atrial fibrillation, 93%
of the patients may be converted to sinus rhythm
without additional antiarrhythmic drug therapy. Atrial
echocardiography, can be restored in 70% to 100% of
these patients[4–7,11,12]. Also, there seems to be a signifi-
cant improvement in health-related quality of life in
patients in whom the MAZE operation has been per-
formed for paroxysmal or chronic atrial fibrillation as
The major drawbacks of these studies are inhomo-
geneity of patients (patients with and without structural
heart disease, chronic and intermittent atrial fibril-
lation), diversity and modification of surgical methods
(cut and suture, cryoablation, radiofrequency ablation)
leading to data which is hard to reproduce.
We developed the first randomized, prospective trial
so far to evaluate the effectiveness of intra-operative
radiofrequency cooled tip ablation in a modified MAZE
procedure during mitral valve surgery. In a randomized
fashion we evaluated the efficacy and functional out-
come of patients undergoing a modified MAZE pro-
cedure, in addition to mitral valve replacement, in
patients with chronic atrial fibrillation and mitral
valve disease. Is there a rationale for the addition of
antiarrhythmic surgery in patients undergoing open
The primary end-point of this study was sinus rhythm at
postoperative follow-up. Secondary end-points were
defined as clinical outcome, survival, atrial transport
function and functional capacity at follow-up.
Between February 1998 and October 1999, 30 consecu-
tive patients (of all 49 patients with mitral valve replace-
ment alone) in whom mitral valve replacement and
chronic atrial fibrillation were indicated (permanent
atrial fibrillation for 1 year or at least two non-successful
medical or electrical cardioversions 6 months before
surgery) were included in the Clinic for Cardio-Thoracic
Surgery at the University Hospital Bergmannsheil
Bochum, Germany. Of the 30 patients, nine were male
(30%) and 21 female (70%) with a mean age of 68 years
(range 49 to 77 years). The primary indication for mitral
valve replacement was mitral valve regurgitation in 18
patients, mitral valve stenosis in five patients and com-
bined mitral valve disease in seven patients. The mean
duration of permanent atrial fibrillation before surgery
was 3·1 years (range between 1 to 30 years). The
underlying heart disease was degenerative in 14, docu-
mented mitral valve prolapse in one patient, rheumatic
heart disease in five, and unknown in 10.
After informed consent was obtained in all patients,
they were consecutively randomized to undergo either
mitral valve surgery and antiarrhythmic surgery (modi-
fied MAZE operation) (group A) or mitral valve surgery
without antiarrhythmic surgical intervention (group B).
All patients received prosthetic mitral valve replacement
and in patients randomized to group A an additional
intra-operative cooled-tip radiofrequency ablation as
previously described by Sie et al. (manuscript in prep-
aration). The MAZE procedure includes biatrial appen-
dectomy, right atrial radiofrequency ablation of the
isthmus, intercaval and appendage–tricuspid anulus
lines, left atrial isolation of each pulmonary vein ostium
and interconnection lines, and a line connecting the
mitral valve anulus and left atrial appendage by radio-
frequency ablation (modification by Khargi et al., in
press) (see Fig. 1). Compared to the original Cox
MAZE-III procedure, the right atrial suture lines have
been changed to ablation lines (except the incision to
enter the right atrial cavity) and the left atrial lines have
been modified as described above (see Fig. 1)[4–6]. As a
major modification, ablation lines in the right and left
atrium were performed in addition to cut and suture
lines to open both atria[14,15].
The duration of the surgical procedure in patients of
group A (270 min vs 205 min), bypass duration (188 min
vs 127 min) and aortic clamp duration (103 min vs
85 min) were longer compared to patients who did not
undergo the MAZE procedure.
was performed using
Department of Cardiology and Angiology of the
University Hospital Bergmannsheil Bochum, Germany.
Follow-up dates of data acquisition were postoperative
day 1, day 12 (before discharge), and at 3, 6, 9 and 12
months after the operation. Antiarrhythmic drug
therapy using sotalol (at least 80 mg twice a day) was
constituted for 6 months and was then switched to
metoprolol (at least 95 mg metoprololsuccinate). All
patients received anticoagulant therapy constituting an
INR value of 2·5 to 3·5. When patients were in atrial
fibrillation at follow-up, electrical cardioversion (direct
current shock up to twice 360 Joule and two different
defibrillator pedal positions) was performed up to twice
during the follow-up period. At each follow-up date a
medical and clinical history and an electrocardiogram
(ECG) was performed. Holter ECG analysis was per-
formed at the 6 and 12 month follow-up. After 12 days
and at 6 and 12 months an additional echocardio-
graphic study (including transmitral and transtricuspidal
bicycle ergometry and a ramp protocol with a workload
rise of 10 Watts every minute) was performed at the
6 month follow-up (see Table 1).
survivors were followed-upclosely at the
Efficacy of additional MAZE surgery 559
Eur Heart J, Vol. 23, issue 7, April 2002
All data was collected in between February 1998 and
standard deviation (median). Student’s unpaired t-test
(two-tailed) was used for comparison between the two
groups. Differences were considered significant at a
P-value <0·05. A highly significant difference was
postulated at a P-value of <0·01. The survival rate and
maintenance rate of sinus rhythm were calculated
according to the Kaplan–Meier method and groups were
variables were expressedas mean?
compared using the Log rank-test (significant difference
postulated at a P-value <0·05).
The cardiothoracic surgical procedure was completed in
all patients. Patients in group A (n=15) and patients in
group B (n=15) did not differ in regard to age, pre-
operative duration of atrial fibrillation, left atrial size in
echocardiography, and pre-operative left ventricular
ejection fraction, as assessed during cardiac ventriculo-
graphy (see Table 2). The 12-month follow-up was
completed in all but two surviving patients from group
left atrium (surgical access via inter-atrial groove): Dots=endocardial RF-ablation lines;
LAA=excised left atrial appendage; LSPV=left superior pulmonary vein ostium;
LIPV=left inferior pulmonary vein ostium; RSPV=right superior pulmonary vein
ostium; RIPV=right inferior pulmonary vein ostium.
Schematic view (looking from behind) of the modified MAZE pattern of the
Table 1 Data acquisition during follow-up period (*up to twice)
Post-op follow-up1 day 12 days3 months6 months9 months 12 months
560 T. Deneke et al.
Eur Heart J, Vol. 23, issue 7, April 2002
B, who were unable to attend the hospital but had
documented (on an outclinic ECG) atrial fibrillation
(one patient unable to travel to our hospital, one patient
with severe psychosis). Mean postoperative follow-up
duration was 22?7 (21) months in group A and 21?6
(20) months (P=0·698).
At postoperative day 1, eight patients from group A and
four patients from group B were in sinus rhythm. The
first electric cardioversion was performed before dis-
charge at day 12. In group A, six of the remaining seven
patients in atrial fibrillation underwent electric direct
current shock cardioversion (successful in three), one
patient spontaneously cardioverted. In group B, two
patients were electrically cardioverted (not successful).
At the 3-month follow-up, nine patients out of group A
and four patients out of group B were in sinus rhythm.
In group A, one patient spontaneously converted from
atrial fibrillation to sinus rhythm, whereas two patients
after previously successful cardioversion converted back
to atrial fibrillation. No patient from group B spon-
taneously converted but two patients were electrically
cardioverted without success. At the 6-month follow-up,
one patient in group A spontaneously converted to sinus
rhythm; this increased to 10 patients. In Group B, 11
patients were in atrial fibrillation and three were cardio-
verted with two successful conversions to sinus rhythm.
After 9 months, nine patients in group A were in sinus
rhythm. In group B, one patient previously successfully
cardioverted, converted back to atrial fibrillation. This
resulted in 10 patients in atrial fibrillation. At the
12-month follow-up in group A, nine patients were in
sinus rhythm and two in atrial fibrillation, whereas in
group B, 11 were in atrial fibrillation (one patient
previously successfully cardioverted, converted back to
atrial fibrillation) and three in sinus rhythm (see Fig. 2).
In the first 6 months postoperation, six patients in
group A were electrically cardioverted with an immedi-
ate success rate of 50%, and long-term success of 17%.
Three patients spontaneously converted to sinus rhythm.
In group B, seven patients were electrically cardioverted
with a immediate success rate of 33%, but long-term
success was 0%, with all converted patients reverting
back to atrial fibrillation. No spontaneous conversion to
sinus rhythm occurred after post-operative day 1 (see
All patients in permanent sinus rhythm in group B
converted to sinus rhythm spontaneously on the first
post-operative day. Twelve months post-surgery, 81·8%
of surviving patients in group A and 21·4% of surviving
patients in group B were in sinus rhythm.
Cumulative frequencies of sinus rhythm were 0·733 in
group A and 0·267 in group B after 6 months and after
12 months, cumulative frequencies of sinus rhythm were
0·800 in group A compared to 0·267 in group B with a
significant difference (P=0·005) (see Fig. 3).
rhythm in all but three patients (two in group A and one
radiofrequency MAZE procedure) and in group B (mitral valve replacement without
modified MAZE procedure)
Patient characteristics in group A (mitral valve replacement plus modified
Female vs male
9 vs 6
49 to 75
1 to 10
12 vs 3
64 to 77
1 to 30
AF pre-op (years)mean (median)
LA pre-op (mm)
Mitral valve insufficiency (n)
Mitral valve stenosis (n)
AF=atrial fibrillation; LA=left atrial; LVEF=left ventricular ejection fraction.
(SR) in group A (mitral valve replacement plus modified
radiofrequency MAZE procedure) and group B (mitral
valve replacement without modified MAZE procedure)
Cardioversion and conversion to sinus rhythm
Group A Group B
DC-shock=direct current shock electric cardioversion using up to
twice 360 Joule and two different defibrillator pedal positions.
Efficacy of additional MAZE surgery561
Eur Heart J, Vol. 23, issue 7, April 2002
in group B). In these patients short runs of atrial
tachycardias were found in less than 10% of the
Holtered time interval.
Echocardiographic study was performed in all patients
available for follow-up. The mean pressure gradient of
the prosthetic mitral valve was 5?2 (4) mmHg in group
A and 4?1 (4) mmHg in group B without mitral valve
regurgitation in either group (P=0·812).
Doppler-echocardiography revealed left atrial con-
traction (transmitral A-wave) at post-operative day 12 in
5/8 (62·5%) patients in sinus rhythm (group A) and in
4/4 (100%) patients in sinus rhythm in group B. At 6
months follow-up, 7/10 (70%) patients in sinus rhythm
(group A) and 4/4 (100%) patients in sinus rhythm
(group B) and after 12 months 6/9 (66·7%) patients in
sinus rhythm (group A) and 3/3 (100%) patients in sinus
GROUP A (MAZE+)
conversion (CV=direct current shock cardioversion, spon=spontaneous cardioversion) and patient death (*=renal
bleeding, mediastinitis, **=sudden cardiac death, ***, ****=respiratory insufficiency).
Flow chart characterization of patient cardiac rhythm (SR=sinus rhythm, AF=atrial fibrillation),
300 200 100
Group A: 0·800
Group B: 0·267
P < 0·01
replacement and modified radiofrequency MAZE procedure (group A)
and without the MAZE procedure (group B).
Cumulative frequencies of sinus rhythm after mitral valve
562 T. Deneke et al.
Eur Heart J, Vol. 23, issue 7, April 2002
rhythm (group B) demonstrated left atrial contraction.
Right atrial contraction (transtricuspid A-wave) was
detected in all patients in sinus rhythm in groups A and
B (100%) during the complete follow-up period (see
Clinical parameters of functional capacity were evalu-
ated at each follow-up but were consistent over the
follow-up period from postoperative month 3 onwards.
Mean NYHA classification was considered 2·5 (2·5)
in group A and 2·6 (3·0) in group B (P=0·531) (see
Spiro-ergometry was performed at the 6-month
follow-up in 11/12 (91·7%) patients in group A and 11/15
(73·3%) patients in group B. In group A, one patient
with severe gonarthrosis and in group B, four patients
(one severe coxarthrosis, one bronchial carcinoma under
chemotherapy unable to attend hospital, one patient
with tracheostoma, one patient with severe psychosis
unable to attend hospital) were unable to perform the
bicycle stress test.
Spiro-ergometry was performed using bicycle ergometry
under standardized conditions with a workload increase
of 10 Watts per minute. Maximum workload achieved
by patients in group A was 73?29 (76) Watts vs 43?16
(38) Watts in group B (P=0·008).
Mean restingheart rates
beats . min?1(beats . min?1) in group A and 68?13
(71) beats . min?1in group B, increasing to 96?16 (94)
beats . min?1and 99?17 (96) beats . min?1under
maximum workload (P=0·247, P=0·666).
Mean aerobic capacity or maximal oxygen uptake
(VO2max) was measured as 9·3?3·2 ml . min?1kg?1in
group A and 8·5?3·0 ml . min?1kg?1in group B
(P=0·530). Anaerobic threshold could be determined in
all but two patients from each group and corresponded
were 74?10 (73)
to an oxygen uptake of 7·2?1·8 ml . min?1kg?1and
6·4?2·3 ml . min?1kg?1(P=0·551) (see Table 4).
Due to postoperative sinus bradycardia one patient
from group A (6·7%) received a permanent DDD
pacemaker and onepatient
postoperatively received a VVI pacemaker due to
in groupB (6·7%)
Survival and complications
Overall survival at 12 months was 83%, in group A 73%
vs 93% in group B (P=0·131). One patient from group A
died after 40 days due to renal bleeding under standard
anticoagulation as performed after prosthetic mitral
valve implantation (INR 2·5 to 3·5). One patient died
after 45 days from mediastinitis. One sudden cardiac
death occurred after 4 months (group A) and one death
due to respiratory insufficiency following severe lung
fibrosis after 7 months (group A). One patient from
group B died of respiratory insufficiency due to severe
chronic obstructive bronchial disease 10 months after
the surgical procedure (see Fig. 2).
No systemic thrombembolic complications occurred
in any of the patients of group A or B during follow-up.
To our knowledge this is the first prospective, random-
ized trial evaluating the efficacy and clinical effect of an
antiarrhythmic surgical procedure on atrial fibrillation.
The aim of any atrial fibrillation surgery is to diminish
uncomfortable symptoms of the arrhythmia by restoring
sinus rhythm, to reinstitute atrioventricular synchrony,
to regain atrial transport function to improve haemo-
dynamics, and to reduce the risk of thromboembolic
Functional data on group A and B at 6-month follow-up (*statistically
Max heart rate (beats . min?1)
Mean heart rate increase (beats . min?1)
VO2max (ml . min?1kg?1)
Anaerobic threshold (VO2)
100% (10/10) 100% (4/4)
Efficacy of additional MAZE surgery563
Eur Heart J, Vol. 23, issue 7, April 2002
During the duration of this study, 30 consecutive
patients with indications for mitral valve surgery alone
and chronic atrial fibrillation were included, whereas 19
patients were indicated for mitral valve surgery but had
documented sinus rhythm. Sixty-one percent of our
pre-mitral valve surgery patients had chronic atrial
fibrillation, which is comparable to data published by
Melo et al. (80% atrial fibrillation).
The efficacy of the MAZE procedure[11,12]or a modi-
fied MAZE procedure using cryoablationor using
radiofrequency left atrial ablationin patients with
mitral valve disease undergoing mitral valve surgery has
already been documented. In our study, the additional
MAZE operation was performed using intra-operative
cooled-tip radiofrequency ablation and ablation lines
different from the proposed Cox-MAZE III procedure.
Whereas the right atrial ablation was performed using
the scheme used by Sie et al. the left atrial procedure was
modified to encircle each single pulmonary vein and to
interconnect these lines. The addition of this procedure
produced a significantly higher rate of restored sinus
rhythm compared to mitral valve replacement alone. At
the 1-year follow-up, the rate of sinus rhythm was found
to be 80% vs 26·7% in the control group. Reports about
rates of sinus rhythm ranking, is in between 60% and up
to 90% in patients with mitral valve heart disease, using
either the classical MAZE procedure or the modified
versions[3,12,17,18]. This indicates that the modified pro-
cedure used in our study is as effective as other modified
MAZE procedures. In patients with atrial fibrillation as
a primary indication for open heart surgery, restoration
of sinus rhythm was found to be as high as 100% under
antiarrhythmic therapy[19,20]. These data are achieved
in patients with only 33% underlying heart disease, in
contrast to patients with mitral valve disease.
antiarrhythmic surgery alone
Spontaneous conversion to sinus rhythm occurred only
during the first 6 months after surgery (three patients).
Electric cardioversion using direct current shock was
effective in restoring sinus rhythm in 50% of patients
who underwent electro-shock, but long-term success was
seen in only 17% of all patients. From this experience we
decided to wait 6 months post surgery for a spontaneous
conversion to sinus rhythm. In patients without spon-
taneous sinus rhythm, direct current shock cardiover-
sion should be performed after 6 months at least once.
In our series of patients, we found a spontaneous
conversion to sinus rhythm after mitral valve surgery
alone in 26·7% of the patients. This is equivalent to data
published by Handa et al. or Melo et al. who stated that
the percentage of patients with sinus rhythm after mitral
valve surgery was between 8% and 27%[17,21]. All
patients regaining sinus rhythm did so directly after the
surgical procedure. No further cardioversions during
follow-up were successful in establishing long-term sinus
These results are comparable to the data discussed by
Sueda et al. and Kawaguchi et al. In these retrospective
reports, the outcome of patients after mitral valve
surgery, plus modifications of the MAZE procedure,
was compared to a carefully selected control group of
patients without MAZE surgery. In both studies, results
were similar with a rate of sinus rhythm in the MAZE
group of 86% to 88% compared to sinus rhythm in only
26·7% to 14% in patients undergoing only mitral valve
surgery. No higher rates of mortality were seen due to
the additional antiarrhythmic procedure. Sueda et al.
have restricted their MAZE approach to the left atrium
due to their experience in activation mapping in patients
with mitral valve disease and chronic atrial fibrillation,
suggesting that the origin of the arrhythmia was in the
left atrium. We performed biatrial modified MAZE
surgery using radiofrequency ablation, as shown to be
effective by the excellent results of Cox et al., in patients
with chronic and intermittent atrial fibrillation due to
different underlying heart diseases. Our method is a first
step in extending its use to patients with atrial fibril-
lation due to different causes and so with possibly
different underlying mechanisms. In a second step, the
procedure should be minimized to the least invasive but
most effective procedure[19,22,23].
We found no difference in resting or maximum
heart rate between the two groups of patients, even
though maximum heart rate was quite low in both
groups (96 beats . min?1in the MAZE group and
99 beats . min?1in group B). This low heart rate may be
explained by the use of high-dosed beta-blockers in both
groups. Cox modified his initial procedure (MAZE I
procedure) — due to the high incidence of sinus node
dysfunction under exercise — finally to the MAZE III
procedure with only a 25% rate of postoperative sick
sinus syndrome. Modifications using cryoablation in
some parts of the atria (preserving all patterns of blood
supply to the sinus node region) further reduced the risk
of iatrogenic sinus node dysfunction and the need for
permanent pacemaker implantation to 3·3%[12,19,20]. In
the modified strategy used in this study the sinus node
region and its blood supply is left in its original state and
so far there has been no sinus node dysfunction, except
for one patient with postoperative sinus bradycardia
receiving a permanent pacemaker (6·7%).
As seen in our study, mitral valve replacement can
safely be combined with a modified MAZE operation
using a radiofrequency cooled-tip. There was a 30-day
mortality of 0%. One death due to complicated medias-
tinitis, which occurred 6 weeks after the procedure,
might be related to the longer than usual operation time
(mean of 65 min longer in group A patients). The over-
all survival rate (83%) at 1 year seems lower than the
survival rate reported by Izumoto et al. (95·1%). The
survival rate in group A (mitral valve replacement plus
MAZE) at 1 year was found to be 73% with no
significant difference to group B. This mortality rate
includes one death due to renal bleeding (effect of
anticoagulation and not MAZE operation related) and
one respiratory insufficiency (not MAZE related) with
severely impaired pulmonary function pre-operation.
After 6 months we switched postoperative sotalol
564T. Deneke et al.
Eur Heart J, Vol. 23, issue 7, April 2002
due to one sudden cardiac death possibly related to
Patients included in our study had a mean 3·1 year
history of chronic permanent atrial fibrillation. Also
mean left atrial size was found to be large with a mean of
almost 60 mm in the MAZE group, which included
a giant left atria. These patient characteristics are two
factors predicting poor outcome after antiarrhythmic
Another goal of antiarrhythmic surgery is to restore
the atrial transport function and by this means, on the
one hand, reduce the risk of thrombembolism and on
the other improve haemodynamics. In our study, biatrial
contraction, as detected via the Doppler-flow profile
(biphasic transmitral and transtricuspidal flow pattern)
was evident in 67% of patients in sinus rhythm 12
months after completed MAZE surgery. Right atrial
contraction was found in all patients with restored sinus
rhythm (100%). These findings are consistent with find-
ings by Isobe et al. and Kosakai et al. (66·7 to 71% left
atrial transport function) but are lower than data col-
lected by Cox et al. in patients with lone atrial fibril-
lation undergoing a MAZE III procedure (94% left
atrial contraction). This difference may be explained by
different patient selection: all our patients had con-
comitant structural heart disease whereas only 33% of
the patients operated on by Cox et al. underwent an
additional cardiac surgical procedure. Concomitant
structural heart disease may involve structural altera-
tions of the atria by left atrial dilation and potentially
left atrial myocardial fibrosis and lead to impaired
left atrial function even when atrial contraction is
restored. On the other hand, transthoracic Doppler
studies may not be the most effective method to detect
biatrial contraction, and transoesophageal Doppler
echocardiography may reveal biatrial systole in a higher
percentage of patients.
Melo et al. proposed a so-called Santa Cruz Score to
report results of surgery on atrial fibrillation. This score
is based on two findings after surgery: (1) the rhythm
and (2) haemodynamic activity of the atria. A successful
operation is defined as a Santa Cruz Score of 4, meaning
sinus rhythm and biatrial contraction. In our study in
the MAZE group, 54·5% of patients (6/11) scored a 4,
27·3% a 2 (3/11) meaning absence of atrial fibrillation
but only right atrial transport and 18·2% of patients
(2/11) scored a 0, indicating atrial fibrillation 12 months
postoperatively. These data are consistent with the data
published by Melo et al. after mitral valve surgery and
concomitant modified MAZE surgery.
A secondary objective of our study was to determine
the effect of the MAZE surgery on clinical and func-
tional results. Even though there was a significantly
higher maximum workload during spiro-ergometry after
6 months in group A patients we found no significant
differences concerning NYHA class, maximum oxygen
uptake and measures at the anaerobic threshold reflect-
ing degrees of cardiac disability. There seems to be
a tendency towards a higher functional capacity in
those patients who underwent additional antiarrhythmic
surgery but the study was not designed to elaborate
In our study most patients had an aerobic threshold
related to a moderate to severe functional impairment,
resulting in a depressed cardiac ability to perform
It has been recently shown that the MAZE operation
significantly improves health-related quality of life in
patients undergoing surgery for paroxysmal and chronic
atrial fibrillation. In our study a successful MAZE
surgery produced a tendency towards improved func-
tional capacity when compared to patients after mitral
valve replacement without antiarrhythmic surgery.
Our study is limited by the small number of patients
analysed which has decreased over follow-up duration.
The trend of our data is consistent with data obtained
in larger patient collectives. Further development in
ablative techniques (epicardial approach, radiofre-
quency ablation) and surgical skill will improve outcome
and make the MAZE procedure a more widely used
operative cooled-tip radiofrequency ablation catheters.
Postoperative bradycardia with the need for permanent
pacemaker implantation is rare.
2. An additional MAZE procedure using cooled-tip
radiofrequency ablation significantly more often restores
sinus rhythm in patients with chronic atrial fibrillation
compared to patients undergoing mitral valve surgery
alone. Sinus rhythm is usually restored within the
first 6 months after surgery and direct current shock
cardioversion has only a slight effect on rates of sinus
3. Complete success, determined by a biatrial transport
function in Doppler-echocardiographic evaluation, is
found in about 67% of patients with sinus rhythm after
a combined procedure.
4. The MAZE procedure, in combination with mitral
valve replacement, produced a tendency towards better
functional performance on exercise tests or clinical
evaluation in our small number of patients.
Therefore long-term follow-up studies, including
patients undergoing the MAZE operation using radio-
frequency ablative techniques, have to be performed
before this therapy can be universally recommended.
functional capacities of
 Ostrander Jr LD, Brandt RL, Kjeldsberg MO et al.
Electrocardiographic findings among the adult population of
a total natural community, Tecumseh, Michigan. Circulation
1965; 31: 888–98.
 Kannel WB, Abbott RD, Savage DD et al. Epidemiologic
features of chronic atrial fibrillation: The Framingham study.
N Engl J Med 1982; 307: 1018–22.
Efficacy of additional MAZE surgery565
Eur Heart J, Vol. 23, issue 7, April 2002
 Melo JQ, Neves J, Adraga ˜o P et al. When and how to report Download full-text
results of surgery on atrial fibrillation. Eur J Cardiothorac
Surg 1997; 12: 739–44.
 Cox JL, Schuessler RB, Agostino HJ et al. The surgical
treatment of atrial fibrillation. J Thorac Cardiovasc Surg
1991; 101: 569–83.
 Cox JL, Boineau JP, Schuessler RB et al. Operations for atrial
fibrillation. Clin Cardiol 1991; 14: 827–34.
 Cox JL, Boineau JP, Schuessler RB et al. Successful surgical
treatment of atrial fibrillation. Review and clinical update.
JAMA 1991; 266: 1976–80.
 Izumoto H, Kawazoe K, Eishi K, Kamata J. Medium-term
results after modified Cox/Maze procedure combined with
other cardiac surgery. Eur J Cardio Surg 2000; 17: 25–9.
 Jais P, Haissaguerre M, Shah DC et al. A Focal Source
of Atrial Fibrillation Treated by Discrete Radiofrequency
Ablation. Circulation 1997; 95: 572–6.
 Leich J, Klein G, Yee R, Guiraudon G. Sinus node-
atrioventricular node isolation: Long-term results with the
corridor operation for atrial fibrillation. J Am Coll Cardiol
1991; 17: 970–5.
 Moe GK. On the multiple wavelet hypothesis of atrial
fibrillation. Arch Int Pharmacodyn Ther 1962; 140: 183–8.
 Izumoto H, Kawazone K, Kitahara H, Kamata J. Operative
results after the Cox/maze procedure combined with a mitral
valve operation. Ann Thorac Surg 1998; 66: 800–4.
 Isobe F, Kawashima Y. The outcome and indications of the
Cox maze III procedure for chronic atrial fibrillation with
mitral valve disease. J Thorac Cardiovasc Surg 1998; 116:
 Lo ¨nnerholm S, Blomstro ¨m P, Nilsson L, Oxelbark S, Jideus
L, Blomstro ¨m-Lundqvist C. Effects of the Maze Operation on
health-related quality of life in patients with atrial fibrillation.
Circulation 2000; 101: 2607–11.
 Beukema WP, Sie H, Misier A, Haalebos M, Schipper C.
Radiofrequency ablation of atrial fibrillation in patients
undergoing concomitant cardiac surgery. First experience
(Abstr). PACE 1997; 20.
 Lemke B, Khargi K, Deneke T et al. A prospective random-
ized trial to evaluate the effectiveness of the Cox-MAZE-III-
Operation using cooled tip radiofrequency ablation in patients
with chronic atrial fibrillation and mitral valve disease (Abstr).
PACE 2000; 303.
 Kosakai Y, Kawaguchi AT, Isobe F et al. Cox maze pro-
cedure for chronic atrial fibrillation associated with mitral
valve disease. J Thorac Cardiovasc Surg 1994; 108: 1049–55.
 Melo J, Adraga ˜o P, Neves J et al. Surgery of atrial fibrillation
using radiofrequency catheter ablation: Assessment of results
at one year. Eur J Cardiothorac Surg 1999; 15: 851–5.
 Kim YJ, Sohn DW, Park DG et al. Restoration of atrial
mechanical function after maze operation in patients with
structural heart disease. Am Heart J 1998; 136: 1070–4.
 Cox JL, Schuessler RB, Lappas DG, Boineau JP. An 8½-year
clinical experience with surgery for atrial fibrillation. Ann
Surg 1996; 224: 167–73.
 Cox JL, Boineau JP, Schuessler RB, Jaquiss RDB, Lappas
DG. Modification of the maze procedure for atrial flutter
and atrial fibrillation. J Thorac Cardiovasc Surg 1995; 110:
 Handa N, Schaff HV, Morris JJ et al. Outcome of valve repair
and the Cox maze procedure for mitral regurgitation and
associated atrial fibrillation. J Thorac and Cardiovasc Surg
1999; 118: 628–35.
 Sueda T, Nagata H, Orihashi K et al. Efficacy of a simple left
atrial procedure for chronic atrial fibrillation in mitral valve
operations. Ann Thorac Surg 1997; 63: 1070–5.
 Kawaguchi AT, Kosakai Y, Sasako Y, Eishi K, Nakano K,
Kawashima Y. Risks and benefits of combined maze pro-
cedure for atrial fibrillation associated with organic heart
disease. J Am Coll Cardiol 1996; 28: 985–90.
 Kamata J, Kawazoe K, Izumoto H et al. Predictors of sinus
rhythm restoration after Cox maze procedure concomitant
with other cardiac operations. Ann Thorac Surg 1997; 64:
 Weber KT, Janicki JS, McElroy PA. Cardiopulmonary exer-
cise testing. In: Weber KT, Janicki JS, eds. Cardiopulmonary
Exercise Testing. Philadelphia: W. B. Saunders Co., 1986: 153.
566 T. Deneke et al.
Eur Heart J, Vol. 23, issue 7, April 2002