Download Morphological features of the P-waves at surface electrocardiogram

CLINICAL RESEARCH
Europace (2014) 16, 578–586
doi:10.1093/europace/eut248
Electrocardiology and risk stratification
Morphological features of the P-waves at surface
electrocardiogram as surrogate to mechanical
function of the left atrium following a successful
modified maze procedure
Jonathan Buber 1,2*†, David Luria 1,2†, Leonid Sternik 1,2†, Rafael Kuperstein 1,2,
Avishay Grupper 1,2, Ilan Goldenberg 1,2, Ehud Raanani 1,2, Micha S. Feinberg 1,2,
Eyal Nof 1,2, Michael Eldar 1,2, and Michael Glikson 1,2
1
Leviev Heart Center, Chaim Sheba Medical Center, 5265601 Tel Hashomer, Israel; and 2The Sackler Faculty of Medicine, Tel Aviv University, 69978 Ramat Aviv, Israel
Received 25 April 2013; accepted after revision 18 July 2013; online publish-ahead-of-print 29 August 2013
Aims
Absent left atrium (LA) mechanical contraction may occur following the modified Cox-maze operation, and was found to
impose a potential risk for the occurrence of thrombo-embolic stroke. It is unknown whether certain morphological
P-wave characteristics can surrogate absent LA mechanical activity. The aim of this study was to evaluate the morphological features of the P-waves on the surface electrocardiogram (ECG) of patients who underwent the maze operation
and to relate them to the contractile profile of the LA.
.....................................................................................................................................................................................
Methods
Electrocardiogram tracings of 150 consecutive patients that were in sustained sinus rhythm following the maze operation
and results
were evaluated. P-waves were scrutinized for morphology, duration, axis, and amplitude. Clinical, surgery-related, and
echocardiographic data were collected and analysed. Forty-seven patients (31%) had no evidence of LA contraction
at 3 months after surgery (baseline assessment) and on follow-up echocardiography. Multivariate analysis showed that
a positive-only P-wave deflection at lead V1 (P ¼ 0.03), a negative-only deflection at aVL, and a P-wave amplitude of
≤0.05 mV at the septal-anterior leads (P , 0.001 for both) were associated with absent LA mechanical contraction.
In a secondary analysis, a risk score involving the above three parameters was developed for the prediction of stroke
occurrence. Patients at the high-risk score group had a 30% survival freedom of stroke compared with 70% for patients
at intermediate risk (P , 0.001).
.....................................................................................................................................................................................
Conclusion
Absent LA mechanical contraction following the modified maze operation may be accompanied by a distinguished
pattern of the P-waves on the surface ECG.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Electrocardiogram † P-waves † Left atrial activity
Introduction
The left atrial (LA) contractile function contributes to the overall left
ventricular performance via its so called ‘booster’ effect during diastolic filling, a contribution that may be of high haemodynamic importance for patients with impaired diastolic function of the left
ventricle. Several previous studies have shown that the LA may not
regain its contractile function following the performance of the
maze procedure for patients with atrial fibrillation (AF) undergoing
cardiac surgery.1,2 As recently reported by our group, lack of LA
mechanical contraction (LAMC), along with very large LA size following a ‘modified’ Cox maze III procedure as diagnosed during a routine
echocardiography examination, may pose significant risk for subsequent cerebral thrombo-embolic events.3
* Corresponding author. Department of Cardiology, Boston Children’s Hospital, 300 Longwood Avenue, Boston MA 02115, USA. Tel: +1 617 355 6000; fax: +1 617 739 8632,
E-mail: [email protected]
†
These authors contributed equally to original concept and to authorship of this investigation.
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2013. For permissions please email: [email protected].
Morphological features of the P-waves at surface ECG as surrogate to mechanical function of LA
What’s new?
† Left atrial standstill following the modified maze operation is
not uncommon, and has been shown to be associated with occurrence of thrombo-embolic strokes. Left atrial contractile
pattern can be evaluated by using standard echocardiographic
techniques.
† The P-waves on the surface electrocardiograms (ECGs) represent the pattern of atrial electrical activation. We report
that lack of mechanical atrial contraction is associated with
specific P-waves morphological features on the surface ECG.
† These features include a positive-only deflection at V1, a
negative-only deflection at aVL, and P-wave amplitude of
≤0.05 mV at the septal-anterior leads.
† When combined together into a risk score, these parameters
were also found to be associated with the occurrence of
strokes of thrombo-embolic origin at follow-up.
† These findings contribute to the understanding of the
complex relationships between the mechanical and the electrical activation of the left atrium following the modified
maze operation.
The electrical activation patterns of both the right and LA are
readily represented by the P-waves on the surface electrocardiogram
(ECG), and may become modified following the maze operation,
ablation procedures, or cardioversion of AF.4 – 7 However, a clear relation between the P-wave morphology and LAMC has not been previously established. The aim of this study was to determine whether
lack of LAMC following the maze procedure may be characterized on
the surface ECG by certain P-wave morphological features.
Methods
Study population and data collection
Data were collected from 236 consecutive patients who underwent the
modified maze procedure in our institution, using a combination of radiofrequency and cryoablation lesions8 from February 2004 to November
2009. The lesion set utilized at the time of the operation was consistent
in all study patients, and involved incision in the interatrial groove, isolation of the right pulmonary veins via a unilateral ablation line, encirclement of the left pulmonary veins and subsequent line connection
between both islands of pulmonary veins, ablation lines from the ablation
line isolating the left pulmonary vein to the base of the LA appendage amputation site and to the posterior MV annulus, and an ablation line drawn
on the right-sided aspect of the interatrial septum up to the caudal aspect
of the opening of the coronary sinus extended to the inferior vena. The
procedures were performed using a combination of radiofrequency
and cryoablation lesions. In 95% of the patients, the LA appendage was
removed and the stump ligated.
Clinic-based outpatient follow-up included a complete physical examination, review of hospitalization records after the index surgery (if
existed), 12-lead ECG, and a 24 h Holter monitoring. A detailed echocardiogram was obtained immediately following surgery, at 3 and 12 months
after the surgery and annually thereafter. Electrocardiogram tracings
were obtained immediately after surgery, every 3 months during the
first post-operative year, and annually thereafter. Twenty four or 48 h
579
Holter studies were obtained at 3- and 12-month period after the operation. All available clinical data from visits not performed in our institute,
including ECG and Holter exams, were obtained.
Patients were included in this current analysis if: (i) sinus rhythm
without conduction disorders that required pacemaker implantation
was documented beginning in the first post-surgery clinic visit and
throughout follow-up; (ii) at least one detailed pre-operative and two
follow-up echocardiographic examinations, as well as three postoperative ECG tracings were available; (iii) there was no electrocardiographic or clinical evidence of AF or other atrial tachycardias in
all follow-up ECG’s, Holter exams, and patient interviews and (iv) all
demographic and clinical data were available.
Of the entire maze-operated cohort, 158 (67%) patients had no
evidence of AF or other atrial tachyarrhythmias throughout the follow-up
period, while 33% developed atrial arrhythmias or AF during followup. Eight patients (all foreign countries residents) were lost to follow-up,
leaving 150 patients to be included in the current analysis. Antiarrhythmic
agents (exclusive beta blockers) were discontinued at the 3-month postoperative clinic visit and not renewed in all patients who maintained sinus
rhythm. Of the 78 patients who developed atrial arrhythmia at follow-up,
28 had normal sinus rhythm at the baseline 3-month visit; of these
28 patients, 20 developed atrial arrhythmia in the 6-month post-surgical
period, and the remaining 8 developed it during the 9-month postsurgical period. The median follow-up time for the entire study population was 18 months (range: 7 – 39 months).
The study was approved by the Institutional Review Board.
Electrocardiogram analysis and
echocardiographic methods
All 150 patients that comprised the study cohort were at AF rhythm prior
to surgery. At least three ECG tracings were performed during the
follow-up period for the entire cohort. All available tracings were evaluated for the current analysis. Standard 12-lead ECG tracings were distinctly evaluated with the aid of an electronic magnifier by two
independent readers (J.B. and A.G.). In case of disagreement between
the readers, adjudication by a third reader was performed. The tracings
were analysed to determine the cardiac rhythm and the following parameters: PR interval, P-wave duration, axis, amplitude, and morphology
in all leads. Following the P-wave amplitude analysis in all leads, a subdivision into three anatomical regions was performed: the septal-anterior
region (consisting of leadsV1 – 4), inferior region (leads II, III, aVF), and
lateral region (leads I, aVL, V5 – 6). In each of these regions, the average
P-wave amplitude was calculated and used for the analysis. In the case
that the P-wave morphology consisted of a negative deflection only,
this deflection was used for the calculation of the average value, and in
the case of a biphasic deflection, the absolute sum of the deflections
were added and used for the analysis.4,9 P-wave morphology was
divided into four patterns: positive (+), negative (2), positive followed
by a negative deflection (+, 2), and a negative followed by a positive deflection (2, +). The investigators were not aware of the echocardiographic LAMC profile at the time of the readings. Echocardiography
was performed using commercially available systems (Philips IE 33);
VIVID 7 (General Electric); VIVID I (General Electric); or Sonos 7500
(Hewlett Packard) and revised by an experienced echocardiographist
(R.K.). Measurements and evaluation of the LA size and function were
made according to the American Society of Echocardiography recommendations for chamber quantification10 as previously reported.3 The
absence of A wave on the tracings was considered as the absence of
mechanical atrial contraction. All patients were in sinus rhythm during
the echocardiographic examinations. Left atrial volume (LAV) was measured using the Simpson’s method from the four chamber apical view,
580
J. Buber et al.
corrected for the body surface area (LAV index or LAVI), and divided into
quartiles at the post-surgical measurements—≤21 mL/m2, 21.1 ≤
LAVI ≤ 27, 27.1 ≤ LAV ≤ 32.9, and ≥33 mL/m2.
Left atrial mechanical contraction assessment
by P-wave characteristics
The electrocardiographic parameters that were examined were evaluated for their relationships with the LAMC profile on echocardiography.
In a secondary analysis, P-wave parameters were also evaluated for relation with LAVI and with the clinical outcome of the occurrence of ischaemic stroke suspected to be of thrombo-embolic origin at any time during
follow-up.
The P-wave parameters were consistent in all tracings obtained, beginning from the tracing obtained 3 months after the surgery (time 0), except
for the amplitude in the three anatomical area. In the case of variability in
the P-wave amplitudes, the tracing performed at the closest calendar
proximity to the final echocardiographic examination was used for interpretation. Echocardiographic parameters used for analyses were
obtained at the baseline time 0 studies and each study thereafter. The
presence or absence of LAMC was consistent in all readings. P-wave amplitude in the three anatomical areas was dichotomized at the upper and at
the lower quartile, and compared with the three lower and upper quartiles, respectively, for our primary analysis.
Statistical analysis
The clinical and electrocardiographic P-wave characteristics of study subjects stratified by the presence or absence of LAMC, by upper quartile
LAVI, and by the occurrence of stroke were compared using x 2 tests
for categorical variables and t-tests or Mann– Whitney – Wilcoxon
tests for continuous variables.
Identification of electrocardiographic parameters
associated with lack of left atrial mechanical contraction
We included 12 potential ECG risk factors (presented in Table 2) for the
model that assessed the relation between ECG factors and lack of LAMC.
Thresholds for categorization of numerical variables were pre-specified
Table 1 Baseline characteristics of the study patients by LAMC
Variable
LAMC absent (n 5 47)
LAMC present (n 5 103)
P value
Age at surgery
Female
62 + 12
30 (64%)
63 + 11
41 (40%)
0.88
,0.01
Body mass index
27 + 5
28 + 5
0.67
Clinical variables
CHA2DS2-VASc score ≥2
...............................................................................................................................................................................
31 (66%)
57 (55%)
0.22
Hypertension
29 (61%)
65 (64%)
0.81
Diabetes
Congestive heart failure
12 (25%)
0
20 (19%)
5 (5%)
0.41
0.09
Peripheral arterial disease
9 (19%)
21 (20%)
0.83
2 (5%)
4 (4%)
0.91
0.17
Paroxysmal
9 (19%)
22 (21%)
Persistent
Permanent
16 (34%)
22 (47%)
32 (31%)
48 (47%)
Beta-blocker treatment
12 (25%)
30 (29%)
29 + 13
28 + 10
0.40
13 (27%)
19 (18%)
0.20
LA diameter after surgery (cm)
LA appendage removed
4.7 + 0.6
44 (93%)
4.5 + 0.7
99 (96%)
0.07
0.50
Left ventricular ejection fraction (%) after surgery
56 + 7
Prior stroke
Atrial fibrillation type
Echocardiographic parameters
Post-surgery LA volume index
Post-surgery LA volume index ≥35 mL/m2
Left ventricular ejection fraction ≤35% after surgery
Surgery-associated parameters
CABG during surgery, n
0.73
56 + 7
0.95
1 (2%)
3 (3%)
0.78
9 (19%)
21 (20%)
0.86
Mechanical prosthetic valve implantation
11 (23%)
30 (29%)
0.85
Bioprosthetic valve implantation
Maze-only, n (%)
27 (54%)
0
46 (44%)
6 (6%)
0.09
Left atrial maze only
37 (78%)
84 (81%)
0.68
Months from surgery to last echocardiogram
Total follow-up in months
12 (6–22)
17 (10– 37)
12 (5 –26)
19 (7 –39)
0.81
0.61
Values mean + SD, n (%), or median (IQR). Percents may not total 100%, because of rounding.
AF, atrial fibrillation; CABG, coronary artery bypass grafting; CHA2DS2-VASc, congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke or transient ischaemic
attack, vascular disease, age 65 –74 years, and sex category (female); LA, left atrial; LAA, left atrial appendage; LAMC, left atrial mechanical contraction; LAVI, left atrial volume index.
581
Morphological features of the P-waves at surface ECG as surrogate to mechanical function of LA
utilizing accepted ECG criteria. Univariate relationships between candidate covariates and lack of LAMC (as defined above) were assessed by
t-tests (x 2 for binary responses). The covariates with P values ,0.20
were further evaluated by carrying out a best-subset logistic regression
analysis, examining the models created from all possible combinations
of ECG predictor variables, and using a penalty of 3.84 on the likelihood
ratio x 2 value for any additional factor included (corresponds to a P value
of 5% for one df x 2 test).
Table 2 P-wave parameters at surface ECGs by LAMC
Variable
LAMC absent (n 5 47)
LAMC present (n 5 103)
P value
...............................................................................................................................................................................
PR interval .200 ms
26 (55)
76 (74)
0.02
P-wave duration .120 ms
25 (53)
75 (73)
0.02
P-wave duration in ms, median (interquartile range)
Abnormal P-wave axisa
112 (95– 132)
18 (38)
125 (105–140)
36 (35)
0.08
0.72
P-wave morphology at V1
Positive
37 (78)
26 (25)
,0.001
Negative
+, 2
1 (2)
9 (19)
3 (3)
74 (72)
0.78
,0.001
2, +
0
P-wave morphology at II
Positive
Negative
+, 2
2, +
28 (59)
0
0
26 (25)
0
N/A
,0.001
N/A
19 (40)
0
70 (68)
7 (7)
,0.001
0.07
Positive
Negative
27 (58)
0
24 (23)
5 (5)
,0.001
0.09
+, 2
18 (38)
68 (66)
,0.001
2 (4)
6 (6)
0.52
Positive
28 (59)
27 (26)
,0.001
Negative
+, 2
1 (2)
27 (38)
0 (0)
75 (73)
N/A
,0.001
1 (2)
1 (1)
P-wave morphology at III
2, +
P-wave morphology at aVF
2, +
P-wave morphology at aVL
Positive
Negative
+, 2
2, +
N/A
5 (10)
12 (12)
0.87
33 (70)
17 (16)
,0.001
0
9 (19)
0
74 (72)
N/A
,0.001
6 (13)
19 (40)
15 (15)
41 (40)
P-wave morphology at aVR
Positive
Negative
+, 2
2, +
P-wave amplitude septal-anterior leads b
0
0
0.86
1.0
N/A
22 (47)
46 (45)
0.89
Upper quartile: ≥0.25 mV
6 (12)
59 (57)
,0.001
Lower quartile: ≤0.05 mV
P-wave amplitude inferior leads b
23 (50)
8 (8)
,0.001
Upper quartile: ≥0.15 mV
3 (6)
5 (5)
0.90
22 (48)
15 (15)
0.007
Upper quartile: ≥0.10 mV
3 (6)
13 (12)
0.25
Lower quartile: ≤0.02 mV
33 (70)
46 (45)
0.005
Lower quartile: ≤0.03 mV
P-wave amplitude lateral leads b
Values are n (%). Percents may not total 100%, because of rounding.
a
Defined as ,08 –758.
b
Septal-antrior leads are V1 –4, inferior leads are II, III, aVF, lateral leads are V5, 6, I, aVL, amplitude measurement was obtained by adding the positive and the negative components in
leads with positive –negative P-wave morphology. (2, +), negative–positive morphology; (+, 2), positive –negative morphology.
582
Development of electrocardiogram score for lack of left
atrial mechanical contraction
After selection of binary covariates, each was assigned a numerical value
based on the relative value of its logistic regression coefficient in the
multivariate logistic regression model (specifically, the ECG factor with
the lowest logistic regression coefficient among the three factors in the
model was assigned a numerical value of 1, and the other two factors
were assigned numerical values based on the relative values of their logistic regression coefficients to that of the lowest value). An ECG score was
constructed in each patient by adding the assigned numerical values of the
factors identified in each patient, and the study population was categorized based on the distribution of the ECG score (assessed as an ordinal
measure) among patients.
Relation of electrocardiogram score to clinical outcomes
during follow-up
Kaplan – Meier estimates for stroke by ECG score counts were determined and statistically evaluated with the log-rank test. Cox proportional
hazards regression analyses were carried out in the total population for
the assessment of the risk of stroke by ECG score counts. The two additional covariates that were included in the model were the LA indexed
volume and the CHA2Ds2-VASc score. Each covariate was first assessed
separately using univariate analysis. All three had P values ≥0.10.
All P values were two-sided and a P value ,0.05 was considered
significant. Analyses were performed with the use of SAS software
(version 9.20).
J. Buber et al.
tracings was 0.02 mV at the tracings obtained 6 months after the procedure and 0.01 mV at the tracings obtained 9 months and 1 year after
the procedure.
An abnormal P-wave axis (defined as either leftward or rightward
of 08 –758) existed to similar degree between patients in both groups.
The median P-wave amplitude in the entire study cohort was largest
in the anterior leads [0.1 mV, interquartile range (IQR) 0.05–0.25],
followed by the inferior leads (median, 0.05 mV, IQR 0.03– 0.15),
and the lateral leads (median, 0.03 mV, IQR 0.02 –0.1).
P-wave parameters, left atrial activity
and left atrial volume
As shown in Table 2, significant differences existed between the
groups in terms of both amplitude and morphology of the P-waves
at the different pre-defined anatomical areas. Patients with absent
LAMC more frequently had a positive-only deflection of the
P-waves at V1 (78 vs. 25%, P , 0.001), as well as at leads II, II, and
aVF (P , 0.001 for all three leads). In addition, a negative
deflection-only morphology at aVL existed more frequently for this
group (70 vs. 16%, P , 0.001). Conversely, patients with preserved
LAMC more frequently had a negative–positive morphology at
aVL (72 vs. 19%, P , 0.001) and a positive –negative morphology at
V1 and II, III, aVF (P , 0.001 for all four leads, Table 2, Figure 1).
Results
Lower-quartile (i.e. ≤0.05 mV) average
Forty-seven patients (31%) had no evidence of LAMC at the echocardiographic assessment performed 3 months after the index surgery.
Follow-up ECGs obtained at 12 months and yearly thereafter
showed consistent findings. Similarly, consistent P-wave-related
ECG findings existed for the patients in both groups (LAMC
present/absent) in all the tracings that were obtained during followup, excluding the P-wave amplitudes as noted above.
Baseline characteristics of the study patients stratified by the presence of LAMC are presented in Table 1. Median follow-up periods
were similar for the two groups. Left atrial maze was performed in
121 (81%) of the patients, whereas biatrial maze was performed
in the remaining patients. Six cases (4%) were standalone maze. Both
LAMC contractile profiles existed in a similar proportion among
patients who underwent the biatrial maze procedure. P-wave parameters stratified by LAMC are presented in Table 2. The P-waves
and the PR segments were more frequently prolonged (.120 and
.200 ms, respectively) in the preserved LAMC group. There was a
trend of P-wave duration shortening as time elapsed from the index
procedure that did not reach statistical significance between the
groups: median P-wave duration for the entire study population was
125 ms, and decreased to a median of 112 ms 6 months following
the procedure (data not shown). Seventy-seven patients (51% of the
entire study population) had a detectable variability in the amplitude
of the P-waves at repeated assessment of the 6-month post-operative
ECG tracings, 56 patients (36% of the entire study population) had a
detectable variability detected at the 9-month post-operative ECG
tracings, and 29 patients (19% of the entire study population) had a variability detected at the 12-month post-operative ECG tracings. Detectable variability was defined as a change of .0.01 mV in the P-wave
amplitude. The median change in the P-wave amplitude at follow-up
P-wave amplitude at the septal-anterior leads was more frequent
among patients with absent LAMC compared with the group of
patients with preserved LAMC (50 vs. 8%, P , 0.001), as well as
at the inferior (≤0.03 mV) and the lateral (≤0.02 mV) leads
(P ¼ 0.007 and P ¼ 0.005, respectively). In a multivariate Cox analysis
performed, three parameters were found to be independently associated with absent LAMC (Table 3): positive-only P-wave morphology
at V1 [odds ratio (OR): 2.4; 95% confidence interval (CI): 1.21–5.60;
P ¼ 0.03), negative-only P-wave morphology at aVL (OR: 7.14; 95%
CI: 2.50–18, P , 0.001), and average P-wave amplitude ≤0.05 mV
(lower quartile value) amplitude at the septal-anterior leads (OR:
7.50; 95% CI: 3.12–19; P , 0.001). The sensitivity, specificity, and predictive values for each of these three parameters were evaluated, as
well as the combination of all three (Table 4). A positive-only morphology at V1 showed the highest sensitivity and negative predictive value
(NPV, 78.7 and 88.5%, respectively), while average P-wave amplitude
≤0.05 mV at the sepatal-anterior leads showed the highest specificity
and positive predictive value (PPV, 94 and 74.2%, respectively). For the
combination of the three parameters, the sensitivity was 75%, specificity 81.5%, PPV 67%, and the NPV was 87.5%.
In the analysis performed to evaluate for association between the
various P-wave parameters and LAVI, no statistically significant associations were established (P . 0.1 for all parameters evaluated, data
not shown). We also evaluated for an association between the type of
maze procedure performed (i.e. biatrial vs. LA maze only) and of the
various P-wave parameters. As compared with LA maze only, biatrial
maze was associated with more frequent lower-quartile amplitude of
the P-waves at all anatomical regions (P ¼ 0.04 for the septal-anterior
region, P ¼ 0.02 for both inferior and lateral regions). In addition, a
significantly lower proportion of patients who underwent a biatrial
Morphological features of the P-waves at surface ECG as surrogate to mechanical function of LA
583
Figure 1 (A and B) ECG tracings from two of the study patients: a patient with absent LAMC (A) and a patient with intact LAMC (B). Patients with
absent LAMC had more frequently small P-wave amplitude at V1 – 4, a positive-only deflection at V1, and a negative-only deflection at aVL compared
with patients with intact LAMC.
maze had a P-wave duration .120 ms and a PR interval .200 ms
(P , 0.01 for both).
P-wave parameters and the occurrence
of ischaemic stroke
Fifteen patients (10% of the entire study population) experienced
ischaemic strokes during follow-up, while no haemorrhagic strokes
occurred. Ten patients without LAMC (21%) had strokes, compared
with five (5%) with preserved LAMC. As previously reported,3 absent
LAMC was associated with a nearly 5-fold (P ¼ 0.02) increase in
the risk for stroke for the entire study population and with a
4.1-fold (P ¼ 0.04) increase in risk for patients not on anticoagulation
treatment. In a secondary comparative analysis aimed to evaluate the
association between P-wave parameters and the occurrence of
thrombo-embolic stroke, a terminal positive deflection at aVL
existed less frequently in patients who had a stroke (33 vs. 62%,
P ¼ 0.03, data not shown), while a positive-only morphology at
lead V1 was more frequent in this group (47 vs. 22%, P ¼ 0.05). No
further parameters were found to be associated with an increased
risk of stroke (P . 0.1 for all amplitudes and morphologies, data
not shown), and no association between the type of maze procedure
and stroke was observed (P ¼ 0.77).
Based on the results of the multivariate model performed for the
association between the P-wave parameters and LAMC profile, we
developed a scoring system to predict freedom from stroke. In accordance with the OR values established at the multivariate model,
each parameter was attributed the following scoring points:
average P-wave amplitude ≤0.05 mV at the septal-antrior leads
and a negative-only deflection at aVL: 2 points each; positive-only
deflection at V1: 1 point. As shown in Figure 2, the event-free survival
curve for stroke was 100% for patients with 0 risk points, 70% for
patients with 1–3 risk points, and 30% for patients with 4–5 risk
points (P log-rank ,0.001). Multivariate Cox proportional hazards
regression analysis showed that after adjusting to the LAVI and
CHA2DS2-VASc score, the presence of 4–5 risk points on the
584
J. Buber et al.
ECG was independently associated with a 2.3 increase in the risk of
stroke (95% CI, 1.01–5.7, P ¼ 0.05, Table 5).
Finally, in a subanalysis aimed to evaluate the association between
P-wave parameters and the occurrence of ischaemic stroke in the
group with absent LAMC only, no statistically significant associations
were established, possibly because this study was inadequately
powered to detect such association due to the low number of events.
Discussion
Main findings
In this single-centre study, we aimed to assess the various P-wave
changes on the surface ECG of patients who underwent the modified
Cox-maze III procedure at surgery and to relate them to the contractile
profile of the LA, to its size, and to the occurrence of ischaemic stroke
at follow-up. Our main findings are that: (i) compared with patients
with preserved LAMC, absent LAMC profile was characterized by
three distinctive P-wave morphological features on the surface ECG:
average P-wave amplitude ≤0.05 mV at the septal-anterior leads,
positive-only morphology at V1, and a negative-only amplitude (as
opposed to a negative–positive morphology) at aVL; (ii) in a risk
score that was developed based on these three parameters, patients
with a high number of risk points had only 30% freedom of stroke at
follow-up as compared with a 70% freedom of stroke among patients
with intermediate number of risk points and 100% freedom of stroke
among patients with no risk points; and (iii) there was no statistically
Table 3 Multivariate analysis: P-wave parameters on
the surface ECG and absent LAMCa
P-wave covariate
Odds
ratio
95% CI
Positive-only at V1
2.40
1.21– 5.60
Negative-only at aVL
Lower-quartile amplitude
septal-anterior leads
7.14
7.50
2.50– 18
3.12– 19
P value
................................................................................
0.03
,0.001
,0.001
a
Further adjusted for the P-wave duration .120 ms, performance of a biatrial maze
operation, positive-only morphology at II, at III, at aVF, lower-quartile amplitude at
the inferior leads and at the lateral leads.
CI, confidence interval.
significant relationship between the LA size (expressed as LAVI) and
the P-wave parameters on the ECG.
Activation patterns following the
Cox-maze procedure,
electrocardiographic findings and left
atrial mechanical function
The morphology of the P-wave on the ECG is the end-result
of a complex interplay between electrophysiological and
anatomical –geometrical features of the atria through which the
sinus beat travels. These considerations may become even more
complex following the Cox-maze procedure, which obscures the
electro-anatomical relationship in the atria due to the multiple lines
of ablation.
The presence of a P-wave terminal deflection on the surface ECG’s
inferior leads indicates the presence of LA electrical activation,
whether it is a positive terminal deflection as observed during
normal sinus rhythm, or a negative deflection, as may be observed
in the case of reverse conduction (i.e. in the caudo-cranial
direction) due to blocked conduction at the level of the Bachman
bundle.4,6,11 – 13 In our study cohort, a negative terminal deflection at
the inferior leads occurred more frequently in patients with an preserved LAMC, whereas patients with absent LAMC more frequently
lacked any terminal deflection at these leads, probably due to complete
absence or a very delayed electrical conduction to the LA, leading to
its lack of mechanical activation. Similarly, we assume that the
absence of a positive deflection at lead aVL represents absent
LAMC: a negative-positive morphology of the P-wave at aVL was
reported to occur in patients after atrial compartment operation
who had a delayed, yet preserved activation of the LA, while a
negative-only P-wave morphology occurred more in patients with
absent electrical conduction to the LA.4 The likely explanation for
the positive-only deflection of the P-waves at V1 and absent LAMC
is conduction in the posterior–anterior direction, possibly via posterior interatrial connections (rather than via Bachmann’s bundle). In their
study of P-wave morphology in patients with sinus rhythm before
radiofrequency ablation of AF, Holmqvist et al. analysed unfiltered
signal-averaged P-waves in LA activation mapping to determine
orthogonal P-wave morphology. The authors reported a clear
association between biphasic P-waves in the sagittal plane and anterior–posterior LA activation vector.14 It can thus be assumed that
isolation of the LA will result in disappearance of the terminal negative
phase of the P-waves. Similar observations were made by other
Table 4 Sensitivity, specificity, and predictive values of the independent P-wave parameters associated with absent LAMC
Parameter
Sensitivity (%)
Specificity (%)
Positive predictive
value (%)
Negative predictive
value (%)
...............................................................................................................................................................................
Positive-only at V1
Amplitude ≥0.05 mV at septal-anterior leadsa
a
78.7
50
74.7
92
58.7
74.2
88.5
80.5
Negative-only at aVL
70
84.5
67
86
All three parameters
74
81.5
67
87.5
Amplitude measurement was obtained by adding the positive and the negative components in leads with positive –negative P-wave morphology.
Morphological features of the P-waves at surface ECG as surrogate to mechanical function of LA
585
In a study that described changes in P-wave morphology and dispersion following thoracoscopic surgery for AF, Nassif et al.18
described an initial increase in the duration, the area, and the dispersion of the P-waves after the procedure, followed by a gradual
decrease to values below the baseline measurements. This pattern
was at least partially attributed to reverse remodelling after restoration of sinus rhythm. The findings of our study come in some agreement with this observation, as the P-wave amplitude in our patient
cohort also continued to decrease over time, yet there was no
similar change in the duration of the P-waves. P-wave dispersion
was not a part of our analysis.
P-wave morphology and the risk
of ischaemic stroke
Figure 2 Kaplan– Meier estimates of the probability of survival
free of stroke by the number of risk points in all study patients. ‘Baseline echo’ refers to the echocardiographic study performed 3
months after index surgery. See text for score calculation.
Table 5 Multivariate analysis: the effect LA size, patients
risk profile, and high-risk ECG features on the risk of
stroke in study populationa
Study population
Hazard
ratio
95% CI
P value
LAVI . 33 mL/m2
3.08
1.05–9.04
0.03
High-risk ECG featuresb
2.35
1.01–5.7
0.05
................................................................................
Entire study population (n ¼ 150)
a
Further adjusted for CHA2DS2-VASc score .2.
Defined as 4–5 risk points attributed by the ECG scoring system. See text for
details.
LAVI, left atrial volume index; CI, confidence interval.
b
investigators who studied LA activation patterns using various
methods, from surface ECGs to high-density mapping of LA endocardial activation.15 – 17
The association between very low P-wave amplitude and either
the mechanical or electrical profile of the LA has not been described
before, and hence is a novel finding. Very small P-waves may represent inadequate electrical conduction within the atria, leading to its
mechanical impaired or complete inactivity. We do not assume,
however, that an association exists between P-wave amplitude and
the actual LA size. In our study, as well as in previous studies,4,5,7
no such association has been established. The suggested explanation
is that the LA ablation lines made during the operation lead to the
low-amplitude morphology of the P-wave, thus dissociating its
actual anatomical properties (i.e. its size) from the electrical ones.
In accordance with previous studies, our results indicate that the
intra-atrial conduction becomes prolonged following the maze
procedure, leading to prolongation of the P-wave without actual
prolongation of the PR segment, since the area of the atrioventricular
node is not involved in this procedure.4,5,7
The rate of thrombo-embolic strokes was relatively high in this study,
and the possible explanations were previously discussed.3 One
theory is that the high stroke rates can be at least partially attributed
to Coumadin discontinuation 3 months following the procedure, especially for patients with a mechanically inactive LA. In their longterm follow-up of a large cohort of patients who underwent the
maze procedure, Weimar et al. 19 report that Coumadin treatment
was continued if patients had atrial tachyarrhythmias or an echocardiogram that showed atrial stasis or thrombus. As discussed previously by our group,3 such practice may indeed prove to be of high
benefit in terms of protection from thrombo-embolic strokes.
Upon utilization of the risk score for the prediction of stroke proposed in this analysis, patients with a high number of risk points (i.e.
4 –5) were found to be at an increased risk for stroke during followup. The presence of 4–5 risk points remained independently associated with this risk in a multivariable model that included the LA
size and the CHA2DS2-VASc score. Kohsaka et al.20 described in a
retrospective evaluation of 146 patients, who did not undergo
cardiac surgery, that negative terminal deflection of the P-wave at
lead V1 is associated with an increased risk of ischaemic stroke, regardless of the actual size of the LA. The results of our study indicate
an association between negative terminal deflection pattern at V1
and preserved LAMC, whereas a positive-only deflection was more
common among patients with absent LAMC, probably due to complete or very delayed electrical inactivation of the LA. It is likely
that our study was inadequately powered to detect statistically significant relationships between a single P-wave morphological feature
and the occurrence of stroke, given the low event number.
Limitations
This study was conducted at a single centre and the data retrospectively analysed. P-waves following the Cox-maze procedure may
appear very small on the surface ECG, and sometimes even the
usage of a magnifier may not suffice to determine the exact morphology. In our attempts to overcome this obstacle, we used the electronic magnifier. Right precordial leads (V3R–V6R) were not
obtained for the study patients, and P-wave may appear larger and
clearer in these leads. In the case of biphasic P-wave morphology
on the surface ECG, P-waves amplitude was measured and reported
by adding the positive and the negative components. Even though this
method is of anatomical sense, it is novel and has not been officially
utilized in prior similar studies. Thus, our results regarding P-wave
586
amplitudes should be interpreted in the light of using this technique
only. Electrophysiological studies were not performed in the study
patients to establish the electrical conduction within the RA and
the LA. The data derived from such studies could potentially have
provided us with the exact electronic conduction pattern within
the atria. Right atrial maze alone was not performed in any of the
study patients, and such procedure might cause P-wave morphologies different from the ones described in this study. Asymptomatic
AF could have occurred since AF recognition was based on symptoms and periodic Holter examinations. Recurrent AF events may
by themselves alter the morphology of the P-waves during sinus
rhythm. The association between a decreased, rather than absent,
LA contractile profile and the P-wave morphology was not assessed
in the current study. The overall number of ischaemic strokes was
small, and so this study was possibly inadequately powered to identify
interactions between P-wave covariates and the occurrence of
stroke.
The follow-up protocol for the occurrence of atrial arrhythmias was
developed prior to the publication of the HRS/EHRA/ECAS consensus
document on catheter and surgical ablation of AF. However, the
follow-up performed for this study’s cohort is in good agreement
with the follow-up protocol recommended in the consensus document, which suggests a follow-up ECG every 6 months in the first
2 years unless symptoms suggestive of AF necessitate more frequent
assessment.21 Follow-up duration varied in the study population,
ranging from 7 to 39 months, and later atrial arrhythmias could hypothetically occur for some of the shorter follow-up patients. However,
we showed that the correlation between P-wave morphology and
LAMC profile (the main aim of this study) was independent of the
follow-up duration.
Conclusions
In this study, we identified three independent parameters of the
P-waves on the ECG that accompanied the absence of the LAMC
on the echocardiographic examination: a positive-only deflection at
V1, a negative-only deflection (as opposed to a negative–positive
morphology) at aVL, and low average amplitude (≤0.05 mV) of the
P-waves at the septal-anterior leads. We assume that these morphological features are associated with either lack of or very delayed electrical conduction to the LA, and hence its mechanical standstill.
Our study adds to the understanding of the complex relations
between the mechanical and the electrical properties of the LA in
the presence of scar lesions performed at the maze procedure. Although we aimed to establish an association between the electrocardiographic parameters and clinical events, our report constitutes one
of very few such attempts. Thus, more validated parameters for the
evaluation of the LA performance and prediction of stroke should
be utilized, at least until additional investigations in this relatively unexplored field will be undertaken.
Conflict of interest: none declared.
J. Buber et al.
References
1. Yashima N, Nasu M, Kawazoe K, Hiramori K. Serial evaluation of atrial function by
Doppler echocardiography after the maze procedure for chronic atrial fibrillation.
Eur Heart J 1997;18:496 –502.
2. Feinberg MS, Waggoner AD, Kater KM, Cox JL, Lindsay BD, Pérez JE. Restoration of
atrial function after the maze procedure for patients with atrial fibrillation: assessment by Doppler echocardiography. Circulation 1994;90:II285 –92.
3. Buber J, Luria D, Sternik L, Raanani E, Feinberg MS, Goldenberg I et al. Left atrial contractile function following a successful modified Maze procedure at surgery and the
risk for subsequent thromboembolic stroke. J Am Coll Cardiol 2011;58:1614 –21.
4. Lo HM, Lin JL, Lin FY, Tseng YZ. Characteristic P wave morphology in patients undergoing the atrial compartment operation for chronic atrial fibrillation with mitral valve
disease. PACE 2003;26:1864 –72.
5. Park HU, Kim KH, Kim KB, Ahn H, Choi YS, Oh S. Characteristics of P wave in
patients with sinus rhythm after the maze operation. J Korean Med Sci 2010;25:
712 –5.
6. Bagliani G, Michelucci A, Angeli F, Meniconi L. Atrial activation analysis by surface P
wave and multipolar esophageal recording after cardioversion of persistent atrial fibrillation. PACE 2003;26:1178 –88.
7. Kamata J, Nakai K, Chiba N, Hosokawa S, Sato Y, Nasu M et al. Electrocardiographic
nature of restored sinus rhythm after Cox maze procedure in patients with chronic
atrial fibrillation who also had other cardiac surgery. Heart 1997;77:50 –5.
8. Sternik L, Ghosh P, Luria D, Glikson M, Shpigelshtein D, Malachy A et al. Mid-term
results of the ‘hybrid maze’: a combination of bipolar radiofrequency and cryoablation for surgical treatment of atrial fibrillation. J Heart Valve Dis 2006;15:664 –70.
9. Michelucci A, Bagliani G, Colella A, Pieragnoli P, Porciani MC, Gensini G et al. P wave
assessment: state of the art update. Card Electrophysiol Rev 2002;6:215 – 20.
10. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA et al.
Chamber Quantification Writing Group; American Society of Echocardiography’s
Guidelines and Standards Committee; European Association of Echocardiography.
Recommendations for chamber quantification: a report from the American Society
of Echocardiography’s Guidelines and Standards Committee and the Chamber
Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am
Soc Echocardiogr 2005;18:1440 –63.
11. Antz M, Otomo K, Arruda M, Scherlag BJ, Pitha J, Tondo C. Electrical conduction
between the right atrium and the left atrium via the musculature of the coronary
sinus. Circulation 1998;98:1790 –5.
12. Kasai A, Anselme F, Saoudi N. Myocardial connections between left atrial myocardium and coronary sinus musculature in man. J Cardiovasc Electrophysiol 2001;12:
981 –5.
13. Chauvin M, Shah DC, Haı̈ssaguerre M, Marcellin L, Brechenmacher C. The anatomic
basis of connections between the coronary sinus musculature and the left atrium in
humans. Circulation 2000;101:647 –52.
14. Holmqvist F, Husser D, Tapanainen JM, Carlson J, Jurkko R, Xia Y et al. Interatrial conduction can be accurately determined using standard 12-lead electrocardiography:
validation of P-wave morphology using electroanatomic mapping in man. Heart
Rhythm 2008;5:413 –8.
15. Platonov PG. Atrial conduction and atrial fibrillation: what can we learn from surface
ECG? Cardiol J 2008;15:402 –7.
16. Markides V, Schilling RJ, Ho SY, Chow AW, Davies DW, Peters NS. Characterization
of left atrial activation in the intact human heart. Circulation 2003;107:733 – 9.
17. Betts TR, Roberts PR, Morgan JM. High-density mapping of left atrial endocardial activation during sinus rhythm and coronary sinus pacing in patients with paroxysmal
atrial fibrillation. J Cardiovasc Electrophysiol 2004;15:1111 –7.
18. Nassif M, Krul SP, Driessen AH, Deneke T, Wilde AA, de Bakker JM et al. Electrocardiographic P wave changes after thoracoscopic pulmonary vein isolation for atrial fibrillation. J Interv Card Electrophysiol. [Epub ahead of print].
19. Weimar T, Schena S, Bailey MS, Maniar HS, Schuessler RB, Cox JL et al. The cox-maze
procedure for lone atrial fibrillation: a single-center experience over 2 decades. Circ
Arrhythm Electrophysiol 2012;5:8 –14.
20. Kohsaka S, Sciacca RR, Sugioka K, Sacco RL, Homma S, Di Tullio MR. Electrocardiographic left atrial abnormalities and risk of ischemic stroke. Stroke 2005;36:2481 –83.
21. Calkins H, Kuck KH, Cappato R, Brugada J, Camm AJ, Chen SA et al. 2012 HRS/
EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of
Atrial Fibrillation: recommendations for patient selection, procedural techniques,
patient management and follow-up, definitions, endpoints, and research trial
design. Europace 2012;14:528 –606.