Preoperative Tissue Doppler Imaging-Derived Atrial Conduction Time Can Predict Postoperative Atrial Fibrillation in Patients Undergoing Aortic Valve Replacement for Aortic Valve Stenosis

Abstract

Background: Postoperative atrial fibrillation (POAF) is a common complication of cardiac surgery and may result in stroke or heart failure and poor prognosis. This study aimed to evaluate a novel index of total atrial conduction time derived from the P-wave onset (lead II) to the peak A’ wave on tissue Doppler imaging (PA-TDI duration). The PA-TDI duration was compared with previously reported predictors of POAF, and the optimal cutoff value of PA-DTI was calculated in patients undergoing aortic valve replacement (AVR) for AV stenosis (AS).

Methods and Results: We enrolled 63 patients undergoing isolated AVR. They underwent transthoracic echocardiography with TDI preoperatively and were monitored postoperatively with continuous electrocardiographic telemetry for 7 days. The hospital stay was significantly longer in the 41 patients with POAF than in the 22 without POAF (33.8±19.7 vs. 24.1±8.1 days, P=0.03). Multivariate analysis revealed that PA-TDI duration (odds ratio [OR], 1.07; 95% confidence interval [CI], 1.02–1.13; P=0.0072) and age (OR, 1.14; CI, 1.03–1.28; P=0.016) were significant independent predictors of POAF. Receiver-operating characteristic curve analysis showed the optimal cutoff values of PA-TDI duration and age were 147.3 ms and 74 years, respectively.

Conclusions: The PA-TDI duration was an independent predictor of POAF after AVR for AS. Patients with PA-TDI duration >147 ms should be considered high risk and treated appropriately to improve outcomes. (Circ J 2014; 78: 2173–2181)

Atrial fibrillation (AF) is a common complication after cardiac surgery, occurring in 10–65% of the patients.¹ Several studies have found adverse outcomes associated with postoperative AF (POAF), including increased risk of hospital stay, higher risk of stroke, permanent pacemaker implantation, and increased in-hospital and long-term mortality rates.² The incidence of POAF is higher with valvular surgery.³ Obviously, in aortic valve (AV) surgery, the incidence of AF is related to these postoperative complications.⁴ Moreover, Filardo et al reported that new-onset POAF was significantly associated with increased long-term risk of mortality in patients undergoing AV replacement (AVR) surgery.⁵ These results indicate that POAF should be prevented and controlled postoperatively.

In the past 2 decades, many clinical variables, including preoperative characteristics, comorbidities, intraoperative findings and postoperative complications and requirements, have been evaluated as predictors of POAF; however, most of these predictors were not consistent among studies and only advanced age was a consistent predictor.⁶ Recently, echocardiographic assessments have achieved much better prediction of POAF by measuring the left atrium-related values including the left atrial volume index (LAVI), LA strain and total atrial conduction time (TACT). Although there are several measurements of TACT, the time interval from the P-wave onset on the electrocardiogram (ECG) to the peak of the A’ lateral wave on TDI (PA-TDI duration) provides a reliable estimation of TACT, and the PA-TDI duration has been shown to be significantly related to new-onset AF.^7,8 Our previous study revealed that the PA-TDI duration was predictive of AF in patients undergoing off-pump coronary artery bypass grafting (CABG).⁹

In the present study, we aimed to evaluate the efficacy of PA-TDI duration for predicting POAF in patients undergoing AVR for AV stenosis (AS).

Methods

Subjects

We retrospectively enrolled 63 consecutive patients with severe isolated AS, who underwent AVR at Hiroshima University Hospital between February 2009 and January 2014. The exclusion criteria were: emergency surgical procedure; prior implantation of a permanent pacemaker or implantable cardioverter defibrillator; prior resynchronization therapy; use of class I or III antiarrhythmic agents. Based on these criteria, we excluded 36 patients who underwent AVR for severe AS with/without concomitant procedures including CABG, other valve procedures, Maze procedure, replacement of the ascending aorta or aortic root implantation.

The institutional review board approved the retrospective use of the patients’ data. All patients underwent 12-lead ECG, transthoracic echocardiography and coronary angiography before surgery. All patients were monitored with continuous ECG telemetry for 1 week or more following AV surgery. A patient was determined to have AF when any episode lasted >5 min. Cardiologists in the cardiovascular intensive care unit (ICU), who were not informed of the study, confirmed the diagnosis of AF and initiated prompt treatment using β-blockers, calcium antagonists, antiarrhythmic drugs and/or defibrillation. Patients were divided into 2 groups according to the presence of POAF. Patients’ characteristics, ECG and echocardiographic parameters, medications, and perioperative parameters were compared between groups.

Echocardiography

All patients underwent transthoracic echocardiography, including 2D, M-mode, pulsed wave, continuous wave, color flow, and tissue Doppler imaging (TDI), using an iE33 system (Philips Medical Systems, Best, The Netherlands) equipped with a 3.5-MHz transducer at a depth of 16 cm. All patients were imaged in the left lateral decubitus position. The 2D and color Doppler images were obtained in the parasternal short- and long-axis views, and the apical 2- and 4- chamber views according to the American Society of Echocardiography guidelines.¹⁰ The left ventricular diameter and wall thickness were measured on 2D echocardiography. The mitral valve inflow pattern (E wave, A wave, E wave deceleration time, E/A ratio) was measured on pulsed wave Doppler. The left ventricular ejection fraction (LVEF) was calculated from the apical 2- and 4-chamber views at end-diastole and end-systole using Simpson’s method.¹⁰ The LA volumes were measured during 2 phases of the cardiac cycle: LA maximum volume during the end-systolic phase (just before mitral valve opening) and LA minimum volume during the end-diastolic phase (just before mitral valve closure). The volumes were then calculated using the method of disks.¹⁰ The LAVI was calculated by dividing the maximum LA volume by the body surface area (calculated using body height and weight in the Dubois formula). LV mass was measured using the area-length formula¹⁰ and indexed to the body surface area. Mitral regurgitation (MR) severity was graded on a 0–4 scale.¹¹ TDI was performed with transducer frequencies of 3.5–4.0 MHz by adjusting the spectral pulsed Doppler signal filters to acquire the Nyquist limit of 15–20 cm/s and using the minimal optimal gain. Spectral pulsed Doppler was used to measure myocardial TDI velocities (early diastolic and late diastolic velocities) for the LV lateral wall from the apical 4-chamber view. The PA-TDI interval was defined as the time interval from the P-wave onset (lead II) to the peak of the A’ wave on the tissue Doppler tracing of the LA lateral wall.⁷ The PA-TDI interval was measured in 3 cardiac cycles and averaged. Clinical echocardiographic evaluation on transthoracic echocardiography was performed <1 month before surgery (Figure 1). Experienced echocardiographers (blinded to the clinical information) conducted all echocardiographic examinations and analyzed the results.

Figure 1.

Measurement of P-wave onset (lead II) to the peak A’ wave on tissue Doppler imaging (PA-TDI). A sample volume is placed on the lateral wall of the left atrium just above the mitral valve and the PA-TDI duration is the time from the P-wave onset to the peak of the A’ wave in the velocity graph. (A) Representative patient without postoperative atrial fibrillation shows a short PA-TDI duration of 130 ms. (B) Representative patient with postoperative atrial fibrillation shows a longer PA-TDI duration of 186 ms.

Surgical Technique and Postoperative Management

All preoperative cardiac medications, including β-blockers, calcium-channel antagonists, angiotensin-converting enzyme inhibitors and angiotensin-receptor blockers (except for nonsteroidal antiinflammatory drugs), were continued until the day before surgery. The patients underwent AVR using standard surgical techniques. General anesthesia was induced and maintained by intravenous infusion of fentanyl and propofol. Muscle relaxation was achieved by pancronium. Through a median sternotomy, cardiopulmonary bypass was established by ascending aortic and bicaval right atrial cannulae. Cardiac arrest was obtained by antegrade blood cardioplegia followed by intermittent retrograde cardioplegia. Through a transverse aortotomy, the diseased aortic cusps with calcification were resected using forceps and the Cavitron Ultrasonic Surgical Aspirator (SonoSurg, Olympus, Tokyo, Japan). In this series, bioprosthetic, mechanical and autologous pericardial valves were used in 48, 4 and 11 cases, respectively. All the patients were observed in the ICU and recovered from anesthesia approximately 2 h after the operation or on the following morning. Pulmonary capillary wedge pressure (PCWP) and central venous pressure (CVP) were measured with a pulmonary artery catheter every morning. An anesthesiologist blinded to the echocardiographic data acquired the pressure measurements. Each parameter was measured at the end of expiration and averaged over 6 respiratory cycles. Fluid-filled transducers were balanced before the study with the zero level at the mid-axillary line. The total volumes administrated (volume-in) and eliminated via all routes (volume-out) were recorded, and the net fluid balance (NFB) was calculated by subtracting the total volume-out from the volume-in during the operation, on the day after surgery (postoperative day (POD) 1) and from POD2 to POD5. The patients were transferred to a general ward on POD2. Anticoagulation with heparin was started on POD1, followed by warfarin therapy within 4 days after the operation. The prothrombin time (international normalized ratio) was controlled between 2.0 and 2.5.

Statistical Analysis

Data are presented as the mean±standard deviation. Continuous variables were analyzed using Student’s t-test. Categorical variables were analyzed using a Chi-square test or Fisher’s exact test. Univariate and multivariate logistic regression analyses were used to identify the predictors of POAF, and the odds ratio (OR) and 95% confidence interval (CI) were calculated. Receiver-operating-characteristic (ROC) curves were calculated to determine optimal cutoff values for predicting POAF. All statistical analyses were performed using SPSS version 22.0 (SPSS, Chicago, IL, USA). All P values <0.05 were considered significant.

Results

Preoperative Patient Characteristics

The overall incidence of POAF was 65% (41 of 63 patients). In 28 of the patients with POAF (68%), onset occurred on POD2 or 3, and the incidence gradually decreased by the POD7. In 1 patient (2%), AF persisted at discharge. All the patients were divided into 2 groups: with and without POAF. The preoperative characteristics of patients are shown in Table 1. Patients with POAF were significantly older than those without POAF (78.5±6.1 vs. 71.6±10.3 years, P=0.0014). The other preoperative characteristics were similar between the 2 groups, including a history of prior paroxysmal AF.

Table 1. Preoperative Characteristics of the Groups of Patients Undergoing Aortic Valve Replacement for Aortic Valve Stenosis With POAF or Sinus Rhythm (No POAF)

	No POAF (n=22)	POAF (n=41)	t-test/chi-square	Univariate logistic regression analysis
			P value	OR	95% CI	P value
Preoperative patient characteristics
Age	71.6±10.3	78.5±6.1	0.0014	1.12	1.03–1.22	0.0057
Male sex	9 (41)	14 (34)	0.60	0.75	0.26–2.318	0.60
BSA (m2)	1.49±0.12	1.49±0.20	1.00	0.99	0.049–20.1	1.00
BMI (kg/m2)	22.4±2.4	22.7±3.6	0.73	1.03	0.88–1.21	0.72
History of PAF	1 (5)	3 (7)	0.67	1.66	0.16–17.0	0.67
Hypertension	16 (73)	25 (61)	0.35	0.59	0.19–1.81	0.35
Hyperlipidemia	16 (73)	22 (54)	0.14	0.43	0.14–1.33	0.14
Diabetes	4 (18)	8 (20)	0.90	1.09	0.29–4.13	0.90
COPD	2 (9)	2 (5)	0.51	0.51	0.067–3.92	0.52
Hemodialysis	2 (9)	4 (10)	0.93	1.08	0.18–6.43	0.93
Smoking	3 (14)	1 (2)	0.082	0.15	0.015–1.63	0.12
Medications
ACEI/ARB	7 (32)	16 (39)	0.57	1.37	0.46–4.10	0.57
CCB	6 (27)	14 (34)	0.58	1.38	0.44–4.32	0.58
β-blocker	2 (9)	8 (20)	0.28	2.42	0.47–12.6	0.29
Statins	5 (23)	11 (27)	0.72	1.25	0.37–4.19	0.72
Electrocardiography
P-wave duration (ms)	108.7±11.8	119.6±13.8	0.0027	1.08	1.02–1.14	0.0061
Echocardiography
Aorta diameter (mm)	28.4±5.0	29.1±3.7	0.54	1.04	0.92–1.19	0.53
LA dimension (mm)	37.5±7.8	40.3±5.5	0.11	1.08	0.98–1.20	0.11
LVDd (mm)	46.7±4.8	46.4±5.8	0.81	0.99	0.90–1.09	0.81
LVDs (mm)	30.3±5.0	31.0±6.7	0.67	1.02	0.93–1.12	0.66
IVS thickness (mm)	10.9±1.9	11.7±2.1	0.15	1.22	0.93–1.60	0.15
PW thickness (mm)	11.1±1.9	11.1±1.8	0.90	1.02	0.76–1.36	0.89
E wave (m/s)	82.0±23.3	78.1±26.8	0.57	0.99	0.97–1.01	0.56
A wave (m/s)	100.9±26.3	96.2±26.6	0.50	0.99	0.97–1.01	0.49
E/A ratio	0.86±0.31	0.90±0.52	0.80	1.17	0.35–3.92	0.79
Deceleration time (ms)	295.6±89.6	276.7±91.5	0.44	1.00	0.99–1.00	0.44
LA volume (ml)	63.6±19.5	74.9±21.8	0.047	1.03	1.00–1.07	0.057
LAVI (ml/m2)	42.6±12.3	51.0±16.8	0.043	1.05	1.00–1.10	0.053
LVEF (%)	63.5±7.8	62.4±10.4	0.67	0.99	0.93–1.05	0.67
LVEDV	90.6±29.2	81.5±23.7	0.19	0.99	0.97–1.01	0.19
LVESV	34.7±18.5	32.3±18.8	0.63	0.99	0.97–1.02	0.63
LV mass index (g/m2)	148.8±50.6	156.4±49.7	0.57	1.00	0.99–1.01	0.56
AS
AVA (2D)	0.81±0.22	0.88±0.21	0.26	6.27	0.25–154.8	0.26
Peak velocity (m/s)	5.2±1.2	5.0±1.0	0.69	0.90	0.55–1.48	0.69
PG (max)	111.6±59.5	105.5±43.5	0.64	1.00	0.99–1.01	0.64
PG (mean)	67.4±32.9	58.4±24.6	0.24	0.99	0.97–1.01	0.24
MR
0	8 (36)	16 (39)
+1	11 (50)	15 (37)
+2	3 (14)	10 (24)	0.48
Tissue Doppler imaging
e’ wave peak (cm/s)	5.6±1.6	5.2±2.0	0.39	0.88	0.66–1.18	0.39
A’ wave peak (cm/s)	10.5±2.5	8.7±2.3	0.0098	0.72	0.56–0.94	0.015
E/e’ ratio (septal)	17.3±5.6	21.3±12.7	0.18	1.05	0.98–1.12	0.19
E/e’ ratio (lateral)	15.3±6.4	18.1±11.8	0.33	1.03	0.97–1.10	0.33
PA-TDI duration (ms)	136.7±12.9	154.6±18.7	0.0005	1.07	1.02–1.12	0.0026

ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-II receptor blocker; AS, aortic stenosis; AVA, aortic valve area; BMI, body mass index; BSA, body surface area; CCB, calcium-channel blocker; CI, confidence interval; COPD, chronic obstructive pulmonary disease; EF, ejection fraction; IVS, interventricular septum; LA, left atrial; LAVI, LA volume index; LV, left ventricular; LVDd, LV end-diastolic dimension; LVDs, LV end-systolic dimension; MR, mitral regurgitation; OR, odds ratio; PA-TDI, total atrial conduction time derived from P-wave onset (lead II) to the peak A’-wave on tissue Doppler imaging; PAF, paroxysmal atrial fibrillation; PG, pressure gradient; POAF, postoperative atrial fibrillation; PW, posterior wall.

Preoperative Electrocardiographic and Echocardiographic Variables

In the electrocardiographic assessment, patients with POAF had a significantly longer P-wave duration than those without POAF (119.6±13.8 ms vs. 108.7±11.8 ms, P=0.0027). In the echocardiographic assessment, LA volume, LAVI, PA-TDI and A’ peak velocity were significantly different between the 2 groups (Table 1). The degree of MR was grade 2 in 13 patients and in them, the mitral apparatus was normal or had slightly thickened cusps with or without calcification. Postoperative echocardiography showed that grade 2 mitral valve regurgitation improved to grade 1 in 8, improved to grade 0 in 1 and did not change in 4 patients. The degree of MR was not associated with the incidence of POAF.

Perioperative and Postoperative Factors

The type of prosthesis, operation time, cardiopulmonary bypass time, aortic cross-clamp time and blood loss were not significantly different between the 2 groups (Table 2). Intraoperative and postoperative medications, including the use of catecholamines, β-blockers and carperitide, were not related to the incidence of POAF. Postoperative PCWP was higher in patients with than without POAF (13.6±3.4 vs. 11.5±2.5 mmHg, P=0.016). NFB was greater on POD1 and POD2 in patients with than without POAF (POD1, 441.6±769.3 vs. −39.7±996.9, P=0.037; POD2, −693±818 vs. −1,313±1,067, P=0.013)

Table 2. Intraoperative and Postoperative Characteristics of the Groups of Patients Undergoing Aortic Valve Replacement for Aortic Valve Stenosis With or Without POAF

	No POAF (n=22)	POAF (n=41)	t-test/chi-square	Univariate logistic regression analysis
			P value	OR	95% CI	P value
Type of prosthesis
Bioprosthesis	15 (68)	33 (80)
Mechanical	1 (5)	3 (7)
Auto-pericardium	6 (27)	5 (12)	0.31
Operation time (min)	294.5±88.9	295.7±77.1	0.95	1.00	0.99–1.01	0.95
Cardiopulmonary bypass time (min)	157.8±45.6	144.0±40.7	0.22	0.99	0.98–1.01	0.23
Aortic clamp time (min)	96.9±30.9	91.9±34.8	0.57	1.00	0.98–1.01	0.57
Blood loss (ml)	1,331±1,293	1,388±1,687	0.89	1.00	1.00–1.00	0.89
Fluid balance (ml)
Intraoperative	1,140±1,179	1,100±1,397	0.91	1.000	1.000–1.000	0.90
DOS	162±1,212	267±1,078	0.72	1.000	1.000–1.001	0.72
POD1	–39.7±996.9	441.6±769.3	0.037	1.001	1.000–1.001	0.045
POD2	–1,313±1,067	–693±818	0.013	1.001	1.000–1.001	0.019
POD3	–728±958	–709±1,116	0.95	1.000	1.000–1.001	0.94
POD4	–198±834	–160±839	0.87	1.000	0.999–1.001	0.87
POD5	–342±1,298	–122±809	0.44	1.000	1.000–1.001	0.44
Intraoperative medications
Dopamine	22 (100)	38 (93)	0.19			0.98
Dobutamine	3 (14)	2 (5)	0.22	0.33	0.050–2.11	0.24
Noradrenaline	2 (9)	7 (17)	0.39	2.06	0.39–10.9	0.40
Mirislol	7 (32)	16 (39)	0.57	1.37	0.46–4.10	0.57
Landiolol	2 (9)	1 (2)	0.24	0.25	0.021–2.93	0.27
Nicardipine	14 (64)	34 (83)	0.087	2.78	0.84–9.13	0.093
Carperitide	14 (64)	25 (61)	0.84	0.89	0.31–2.61	0.84
Postoperative medications
Dopamine	20 (91)	34 (24)	0.39	0.49	0.092–2.57	0.40
Dobutamine	7 (32)	7 (17)	0.18	0.44	0.13–1.48	0.19
Mirislol	8 (36)	12 (29)	0.56	0.72	0.24–2.17	0.56
Nicardipine	5 (23)	10 (24)	0.88	1.10	0.32–3.74	0.88
Carperitide	10 (45)	23 (56)	0.42	1.53	0.54–4.35	0.42
Landiolol	1 (9)	4 (10)	0.47	2.27	0.24–21.7	0.48
Oral β-blocker	11 (50)	22 (54)	0.78	1.16	0.41–3.27	0.78
Right atrial pacing	10 (45)	20 (49)	0.80	1.14	0.40–3.23	0.80
Postoperative PCWP (mmHg)	11.5±2.5	13.6±3.4	0.016	1.26	1.04–1.53	0.021
Postoperative CVP (mmHg)	6.5±2.6	7.7±3.0	0.11	1.17	0.96–1.42	0.12

CVP, central venous pressure; DOS, day of surgery; PCWP, pulmonary capillary wedge pressure; POD, postoperative day. Other abbreviations as in Table 1.

Patients with POAF had a significantly longer duration of hospital stay than those without POAF (33.8±19.7 vs. 24.1±8.1 days, P=0.03), because 6 patients in the POAF group had heart failure and required prolonged intravenous catecholamine infusion. POAF was controlled in these patients by antiarrhythmic drugs. In addition, permanent pacemaker implantation for complete atrioventricular (AV) block in 1 patient and Mobitz type 2 AV block in 1 patient in the POAF group also prolonged the hospital stay. Antiarrhythmic drugs, including classes Ia and Ic and β-blockers, were used to control POAF in these patients. No patients were treated with amiodarone.

The hospital mortality rate was 2.4% (1/41) in patients with POAF and 0% (0/22) in those without POAF (P=0.46). An 85-year-old woman developed persistent POAF immediately after the operation with poor control of INR (1.2–1.8) by warfarin. She subsequently developed LA thrombus arising from the left upper pulmonary vein and congestive heart failure. After that, strict warfarin control with heparinization was started, which resulted in unintended subcutaneous bleeding in the abdominal wall. No antiplatelet drug was administered. Despite aggressive treatment, she had a superior mesenteric artery thrombosis and multiple organ failure and died on POD107. One patient with POAF had perioperative stroke and required intensive rehabilitation. No patients in the study suffered perioperative myocardial infarction.

Predictors of POAF

Results of univariate analysis are shown in Table 1. Univariate analysis revealed that age, P-wave duration, PA-TDI duration, A’ peak velocity, postoperative PCWP and fluid balance on POD1 and POD2 were predictive of POAF (age: OR, 1.12; CI, 1.03–1.22; P=0.0057; P-wave duration: OR, 1.08; CI, 1.02–1.14; P=0.0061; PA-TDI duration: OR, 1.07; CI, 1.02–1.12; P=0.0026; A’ peak velocity: OR, 0.72; CI, 0.56–0.94; P=0.015; postoperative PCWP: OR, 1.26; CI, 1.04–1.53; P=0.021; POD1: OR, 1.001; CI, 1.000–1.001; P=0.045; POD2: OR, 1.001; CI, 1.000–1.001; P=0.019).

A positive correlation was seen between PA-TDI duration and P-wave duration (r=0.42; P=0.0018) and between PA-TDI duration and age (r=0.28; P=0.043); however, no correlation was observed between PA-TDI duration and A’ wave velocity (r=−0.26; P=0.056). There were no correlations among P-wave duration, A’ wave velocity and age.

Stepwise multivariate analysis showed that PA-TDI duration (OR, 1.07; CI, 1.02–1.13; P=0.0072) and age (OR, 1.14; CI, 1.03–1.28; P=0.0164) were significant independent predictors of POAF (Table 3).

Table 3. Multivariate Logistic Regression Analysis

	Predictors
	PA-TDI	Age
OR	1.07	1.15
95% CI	1.02–1.13	1.03–1.28
P value	0.0072	0.016
HL test		P=0.62

HL, Hosmer and Lemeshow. Other abbreviations as in Table 1.

ROC analysis showed that the area under the curve for PA-TDI duration was 0.804 (95% CI, 0.69–0.92; P<0.0001) and for age it was 0.701 (95% CI, 0.572–0.810; P=0.0064) (Figure 2). The respective optimal cutoff values of PA-TDI duration and age were >147.3 ms (sensitivity 77.1%; specificity, 79.0%; positive predictive value, 79.0%) and >74 years (sensitivity, 70.7%; specificity, 63.6%; positive predictive value, 78.4%).

Figure 2.

(A) Area under the receiver-operating-characteristic (ROC) curve for PA-TDI duration is 0.804 (95% CI, 0.69–0.92; P<0.0001), that for P-wave duration measured on the 12-lead surface ECG is 0.756 (95% CI, 0.625–0.888; P=0.0001) and that for A’ wave velocity is 0.697 (95% CI, 0.547–0.846; P=0.0098). The results indicate a similar predictive value for the 3 parameters. (B) Area under the ROC for age is 0.701 (95% CI, 0.572–0.810; P=0.0064). CI, confidence interval; PA-TDI, P-wave onset (lead II) to the peak A’ wave on tissue Doppler imaging.

Discussion

This study demonstrated that the preoperative PA-DTI duration predicted postoperative AF in patients who underwent solitary AVR for AS.

POAF is one of the most common complications after cardiac surgery and contributes to worsening of both the postoperative status and prognosis, and increases hospital costs.¹² Banach et al reported an association between POAF and increased risk of stroke and hospital mortality in patients undergoing isolated AVR.¹³ Filardo et al reported that POAF was significantly associated with increased long-term risk of mortality, which was 48% higher in patients with than in those without POAF.⁵ In the present study, of the 64 patients who underwent AVR, 42 (65.6%) developed POAF. Furthermore, in the POAF group, the length of hospital stay was prolonged and 1 patient suffered a superior mesenteric artery thrombosis that resulted in multiple organ failure. Preoperative assessment and preventive management of POAF would be beneficial for improving prognosis.

Many risk factors for the development of POAF have been identified. The preoperative risk factors include older age, male sex, previous history of AF, LV failure, LA enlargement, chronic obstructive pulmonary disease, diabetes mellitus, obesity and reoperation. The perioperative and postoperative risk factors include intraoperative and postoperative catecholamine use, respiratory failure and postoperative LV diastolic dysfunction.^13,14 That suggests POAF is multifactorial, and that all these predictors should be considered preoperatively.

Several recent studies showed that POAF could be predicted by LA size, LA volume, LA area, LAEF, LA strain and TACT. LA is a known risk factor of POAF, as reported in several previous studies. Osranek et al found that LAVI >32 ml/m² was associated with a nearly 5-fold increase in the risk of POAF in cardiac and ascending thoracic surgery.¹⁵ Naito et al reported that LAVI ≥52 ml/m² was a predictor of POAF in patients with AS without coronary artery disease.¹⁶ In that study, LAVI was larger in patients with than in those without POAF (P=0.047); however, logistic regression analysis did not show that LAVI was a significant predictor (P=0.053), which may have been related to the small sample. Leung et al reported that LA area and LAEF calculated from the LA area were predictive of POAF, and that the postoperative atrial filling fraction, assessed by mitral inflow on Doppler echocardiography, independently increased the risk of POAF.¹⁷

Speckle-tracking echocardiography (STE) has been used to evaluate myocardial function and recently was used to measure LA strain and strain rate to assess LA function.¹⁸ Cameli et al reported that in patients undergoing AVR for AS, LAVI and the peak LA longitudinal strain measured by STE were predictors of POAF in their univariate analysis.¹⁹ Furthermore, in their multivariate regression analysis, age and peak LA longitudinal strain (but not LAVI) were independent predictors of POAF. Moreover, ROC curve analysis showed that the AUC of peak LA longitudinal strain (0.89) showed higher diagnostic accuracy than the AUC of LAVI (0.78). Thus, peak LA longitudinal strain is a promising tool; however, it must be measured off-line by an experienced echocardiographer using high-quality images.

TACT is defined as the time between the initial and last portions of the atrial myocardium depolarizing, which usually reflects the interval from the sinus node to LA wall activation.⁷ The simplest measurement of TACT is the P-wave duration on the 12-lead surface ECG. Buxton and Josephson reported that a prolonged P-wave duration predicted POAF; however, the longest P-wave duration in the standard leads (invariably lead II) was not predictive, but the total P-wave duration, which was measured by superimposing P-waves from the 3 standard limb leads, was predictive of AF.²⁰ The present study demonstrated that P-wave duration was a univariate predictor of AF, but in some patients the end of P-wave was obscure and could vary among observers. In addition to measuring P-wave duration on the standard surface leads, Guidera and Steinberg reported that the P-wave duration measured from the signal-averaged ECG (SA-ECG) was predictive of AF.²¹ The SA-ECG is measured using 3 unique orthogonal ECG leads and the P-wave is averaged over a large number of cardiac cycles (range 84–842 beats) to reduce noise and amplify the portions of the P-wave that may not be visible on the 12-lead ECG.

Recently, alternative approaches to determine TACT by echocardiography have been reported. Fuenmayor et al measured the time from the P-wave onset (lead II) to the A’-wave onset determined by TDI with the transducer over the mitral valve (PA-mv), and demonstrated that this value significantly correlated with TACT determined by right atrial and coronary sinus catheterization.²² As a more precise evaluation of LA wall activation, Merckx et al measured PA-DTI duration and demonstrated a significantly better correlation with SA-ECG than with PA-mv or P-wave duration on the surface ECG.⁷

The PA-TDI duration reflects both electrical and structural remodeling in the LA.²³ Electrical remodeling is known to appear at an earlier phase than structural remodeling.^24,25 de Vos et al showed that a prolonged PA-TDI duration was the most important predictor of new-onset AF.⁸ Another report suggested that a prolonged PA-TDI duration was associated with the recurrence of AF after catheter ablation.²³ Bertini et al found that the PA-TDI duration may be useful for stratifying the risk of AF occurrence in heart failure patients with and without a history of AF.²⁶ Recently, Özlü et al showed that the LA maximum volume and PA-TDI duration were independent predictors of POAF after conventional CABG.²⁷ The new information from our previous study was that the PA-TDI duration was also an independent predictor of POAF in patients undergoing off-pump CABG.⁹

The aim of the present study was to determine the optimal cutoff value of PA-DTI duration to predict POAF. de Vos et al showed that in 249 patients with cardiac disease, the PA-TDI duration ranged from 103 ms to 223 ms, and the 2-year incidence of AF was 33% in patients with a PA-TDI duration >190 ms vs. 0% in patients with a PA-TDI duration <130 ms.⁸ Furthermore, patients with a PA-TDI duration >165 ms had a reasonable chance of developing AF.⁸ In our previous study in patients undergoing off-pump CABG, a PA-TDI duration >141 ms was found to be a predictor of POAF.⁹ In the present study, we found that the occurrence of POAF in patients with a PA-TDI duration <130 ms (6/22 cases (27%)) and the optimal cutoff value of the PA-TDI duration was >147 ms. The difference in the optimal PA-TDI cutoff values in our 2 studies may have be related to the extent of invasive surgical manipulation and the underlying disease (coronary artery disease or AS). The optimal cutoff value of PA-TDI duration to predict POAF should be evaluated separately for each situation and disease.

Postoperative characteristics were examined in this study, and NFB on POD1 and POD2, and postoperative PCWP were predictors of POAF in the univariate analysis. These findings are consistent with the results of previous studies.^28,29 Kalus et al reported that NFB on the POD2 was an independent predictor of POAF.²⁸ Other reports also showed that early postoperative NFB was a predictor of POAF.^29,30 Although the timing of fluid overload was different than in our study, the results indicate that excessive fluid administration after surgery increased the incidence of POAF.^29,30 Wu et al reported that higher PCWP on POD1 was an independent predictor of POAF (11.4±2.3 vs. 10.7±2.7 mmHg, P=0.047).³¹ Frost et al noted that CVP on admission to the ICU in the POAF group was 2 mmHg higher than that in the group without POAF.³² These differences in PCWP and CVP were statistically significant, but in the clinical setting they should be carefully monitored because the actual differences were small.

Prevention of POAF has been achieved by several therapeutic modalities, including β-blockers, sotalol, amiodarone, angiotensin-II receptor antagonists and temporary pacing.^33–36 Crystal et al performed a meta-analysis of 27 randomized controlled trials that included 3,840 patients and found a reduced incidence of POAF in patients undergoing CABG who were treated postoperatively with β-blockers (OR, 0.39; 95% CI, 0.28–0.52).³⁶ One of the current strategies (described later) to prevent POAF in cardiac surgery is the prophylactic use of β-blockers. However, in the present study, oral β-blockers were used in 33 patients (53%) and the use of β-blockers was not associated with the incidence of POAF. In some cases, we did not start β-blocker therapy immediately after the operation, for several reasons, including AV block and heart failure requiring catecholamine infusion, which should be considered a discrete problem after AVR surgery for AS. Prophylactic administration of a β-blocker is already recommended.³⁷ Sezai et al reported that the perioperative infusion of landiolol before weaning from cardiopulmonary bypass and the initiation of oral bisoprolol on POD1 significantly lowered the incidence of POAF.³⁸ Most of the information about the effect of β-blockers on POAF has been obtained from patients undergoing CABG.^33,34,36 Large-scale randomized studies of patients with AV disease are needed to confirm the best way to prevent POAF after AVR.

Study Limitations

First, we examined a small number of patients in a single institution and the study design was retrospective without randomization. The present findings should be validated in a larger prospective study. Second, we did not perform an electrophysiology study to examine parameters related to atrial electrical remodeling or record the SA-ECG from the body surface. The PA-TDI duration might have overestimated TACT. Third, this study included patients who underwent isolated AVR; however, some patients had mitral valvular disease as well, which may have increased the risk of POAF. However, MR is often observed in cases of AS because the sustained pressure overload in AS produces concentric LV hypertrophy and an increased transmitral pressure gradient, which worsens existing structural MR and produces functional MR. A meta-analysis of 3,053 patients undergoing AVR for AS concluded that the severity of MR following AVR improved in 55.5% and remained unchanged in 37.7%.³⁹ Improved long-term survival was seen at 3, 5 and 10 years in patients with nil-to-mild MR compared with those who had moderate-to-severe MR.³⁹ In the present study, 62% (39/63 cases) of patients had trivial or mild MR that was not surgically treated and it reduced or ceased after the operation in most cases. Patients with obvious structural MR were not included in this study, because in such cases MV repair was performed together with AVR. However, it was not possible to completely exclude patients with any MR from this study.

In this study, the incidence of POAF was higher than in previous studies. One reason may be that the definition of AF was shorter (5 min) and differed from that used in other studies. Patients were monitored by continuous ECG telemetry for 1 week or more, and POAF occurred from 8 to 11 days after the operation in 4 cases. Another reason may be that we included patients with paroxysmal AF (4 cases) and those with MR, although neither of these variables was significantly associated with the incidence of POAF. However, patients with grade 2 MR tended to have a higher incidence of POAF than those with grade 0 or 1 (77% (10/13) vs. 62% (31/50), P=0.31). POAF has been reported to occur in older patients,^14,40 and the age of patients in this study was higher than that in previous studies.^5,19,41 These differences may account for the higher incidence of POAF in the present study.

Our current strategy to prevent POAF is to treat patients conventionally if the PA-TDI duration is <147.3 ms and/or age is <74 years, because we consider these patients at low risk for POAF. However, if the PA-TDI duration is >147.3 ms and/or age is >74 years, patients are considered at high risk for POAF and treated with β-blocker and/or amiodarone. Intravenous β-blockade with randiolol (5 μg·kg^–1 ·min^–1) is started during or after the operation and followed by oral bisoprolol or carvedilol, starting at a low dose that is increased up to the maximal dose. Intravenous amiodarone (200 mg/kg) is started after the operation followed by oral administration of 100 mg/day. The choice of drugs is influenced by comorbidities (bronchial asthma, interstitial pneumonia, etc) and electrocardiographic findings (prolonged AV interval, prolonged QT interval).

Conclusions

Increased PA-TDI duration was an independent predictor of POAF after AVR for AS. PA-TDI is easily measured, and patients with a PA-TDI >147 ms should be considered at high risk for POAF and treated to reduce the incidence of stroke and heart failure, shorten the hospital stay and improve the outcome.

References

• 1.    Maisel  WH,  Rawn  JD,  Stevenson  WG. Atrial fibrillation after cardiac surgery. Ann Intern Med 2001; 135: 1061–1073.
• 2.    Echahidi  N,  Pibarot  P,  O’Hara  G,  Mathieu  P. Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol 2008; 51: 793–801.
• 3.    Almassi  GH,  Schowalter  T,  Nicolosi  AC,  Aggarwal  A,  Moritz  TE,  Henderson  WG, et al. Atrial fibrillation after cardiac surgery: A major morbid event? Ann Surg 1997; 226: 501–511.
• 4.    Kuduvalli  M,  Grayson  AD,  Au  J,  Grotte  G,  Bridgewater  B,  Fabri  BM; North West Quality Improvement Programme in Cardiac Interventions. A multi-centre additive and logistic risk model for in-hospital mortality following aortic valve replacement. Eur J Cardiothorac Surg 2007; 31: 607–613.
• 5.    Filardo  G,  Hamilton  C,  Hamman  B,  Hebeler  RF Jr,  Adams  J,  Grayburn  P. New-onset postoperative atrial fibrillation and long-term survival after aortic valve replacement surgery. Ann Thorac Surg 2010; 90: 474–479.
• 6.    Kaireviciute  D,  Aidietis  A,  Lip  GY. Atrial fibrillation following cardiac surgery: Clinical features and preventative strategies. Eur Heart J 2009; 30: 410–425.
• 7.    Merckx  KL,  De Vos  CB,  Palmans  A,  Habets  J,  Cheriex  EC,  Crijns  HJ, et al. Atrial activation time determined by transthoracic Doppler tissue imaging can be used as an estimate of the total duration of atrial electrical activation. J Am Soc Echocardiogr 2005; 18: 940–944.
• 8.    de Vos  CB,  Weijs  G,  Crijns  HJGM,  Cheriex  EC,  Palmans  A,  Habets  J, et al. Atrial tissue Doppler imaging for prediction of new onset atrial fibrillation. Heart 2009; 95: 835–840.
• 9.    Fujiwara  M,  Nakano  Y,  Hidaka  T,  Oda  N,  Uchimura  Y,  Sairaku  A, et al. Prediction of atrial fibrillation after off-pump coronary artery bypass grafting using preoperative total atrial conduction time determined on tissue Doppler imaging. Circ J 2014; 78: 345–352.
• 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: 1440–1463.
• 11.    Tribouilloy  C,  Shen  WF,  Quéré  JP,  Rey  JL,  Choquet  D,  Dufossé  H, et al. Assessment of severity of mitral regurgitation by measuring regurgitant jet width at its origin with transesophageal Doppler color flow imaging. Circulation 1992; 85: 1248–1253.
• 12.    Mahoney  EM,  Thompson  TD,  Veledar  E,  Williams  J,  Weintraub  WS. Cost-effectiveness of targeting patients undergoing cardiac surgery for therapy with intravenous amiodarone to prevent atrial fibrillation. J Am Coll Cardiol 2002; 40: 737–745.
• 13.    Banach  M,  Goch  A,  Misztal  M,  Rysz  J,  Jaszewski  R,  Goch  JH. Predictors of paroxysmal atrial fibrillation in patients undergoing aortic valve replacement. J Thorac Cardiovasc Surg 2007; 134: 1569–1576.
• 14.    Handa  N,  Miyata  H,  Motomura  N,  Nishina  T,  Takamoto  S; Japan Adult Cardiovascular Database Organization. Procedure- and age-specific risk stratification of single aortic valve replacement in elderly patients based on Japan Adult Cardiovascular Surgery Database. Circ J 2012; 76: 356–364.
• 15.    Osranek  M,  Fatema  K,  Qaddoura  F,  Al-Saileek  A,  Barnes  ME,  Bailey  KR, et al. Left atrial volume predicts the risk of atrial fibrillation after cardiac surgery: A prospective study. J Am Coll Cardiol 2006; 48: 779–786.
• 16.    Naito  Y,  Yamazaki  K. Preoperative left atrial volume index predicts postoperative atrial fibrillation in patients with severe aortic valve stenosis. J Anesth 2013; 27: 699–704.
• 17.    Leung  JM,  Bellows  WH,  Schiller  NB. Impairment of left atrial function predicts post-operative atrial fibrillation after coronary artery bypass graft surgery. Eur Heart J 2004; 25: 1836–1844.
• 18.    Mondillo  S,  Galderisi  M,  Mele  D,  Cameli  M,  Lomoriello  VS,  Zacà  V, et al; Echocardiography Study Group Of The Italian Society Of Cardiology (Rome, Italy). Speckle-tracking echocardiography: A new technique for assessing myocardial function. J Ultrasound Med 2011; 30: 71–83.
• 19.    Cameli  M,  Lisi  M,  Reccia  R,  Bennati  E,  Malandrino  A,  Solari  M, et al. Pre-operative left atrial strain predicts post-operative atrial fibrillation in patients undergoing aortic valve replacement for aortic stenosis. Int J Cardiovasc Imaging 2014; 30: 279–286.
• 20.    Buxton  AE,  Josephson  ME. The role of P wave duration as a predictor of postoperative atrial arrhythmias. Chest 1981; 80: 68–73.
• 21.    Guidera  SA,  Steinberg  JS. The signal-averaged P wave duration: A rapid and noninvasive marker of risk of atrial fibrillation. J Am Coll Cardiol 1993; 21: 1645–1651.
• 22.    Fuenmayor  AJ,  Ramírez  L,  Fuenmayor  AM. Validation of inter-atrial conduction time measurement by means of echo-Doppler. Arch Cardiol Mex 2002; 72: 125–128.
• 23.    Uijl  DW,  Gawrysiak  M,  Tops  LF,  Trines  SA,  Zeppenfeld  K,  Schalij  MJ, et al. Prognostic value of total atrial conduction time estimated with tissue Doppler imaging to predict the recurrence of atrial fibrillation after radiofrequency catheter ablation. Europace 2011; 13: 1533–1540.
• 24.    Ausma  J,  Velden  HM,  Lenders  MH,  Ankeren  EP,  Jongsma  HJ,  Ramaekers  FC, et al. Reverse structural and gap-junctional remodeling after prolonged atrial fibrillation in the goat. Circulation 2003; 107: 2051–2058.
• 25.    Teh  AW,  Kistler  PM,  Lee  G,  Medi  C,  Heck  PM,  Spence  SJ, et al. Electroanatomic remodeling of the left atrium in paroxysmal and persistent atrial fibrillation patients without structural heart disease. J Cardiovasc Electrophysiol 2012; 23: 232–238.
• 26.    Bertini  M,  Borleffs  CJ,  Delgado  V,  Ng  AC,  Piers  SR,  Shanks  M, et al. Prediction of atrial fibrillation in patients with an implantable cardioverter-defibrillator and heart failure. Eur J Heart Fail 2010; 12: 1101–1110.
• 27.    Özlü  MF,  Erdem  K,  Kırış  G,  Parlar  Aİ,  Demirhan  A,  Ayhan  SS, et al. Predictive value of total atrial conduction time measured with tissue Doppler imaging for postoperative atrial fibrillation after coronary artery bypass surgery. J Interv Card Electrophysiol 2013; 37: 27–33.
• 28.    Kalus  JS,  Caron  MF,  White  CM,  Mather  JF,  Gallagher  R,  Boden  WE, et al. Impact of fluid balance on incidence of atrial fibrillation after cardiothoracic surgery. Am J Cardiol 2004; 94: 1423–1425.
• 29.    Hravnak  M,  Hoffman  LA,  Saul  MI,  Zullo  TG,  Whitman  GR,  Griffith  BP. Predictors and impact of atrial fibrillation after isolated coronary artery bypass grafting. Crit Care Med 2002; 30: 330–337.
• 30.    Nally  BR,  Dunbar  SB,  Zellinger  M,  Davis  A. Supraventricular tachycardia after coronary artery bypass grafting surgery and fluid and electrolyte variables. Heart Lung 1996; 25: 31–36.
• 31.    Wu  ZK,  Iivainen  T,  Pehkonen  E,  Laurikka  J,  Zhang  S,  Tarkka  MR. Fibrillation in patients subjected to coronary artery bypass grafting. J Thorac Cardiovasc Surg 2003; 126: 1477–1482.
• 32.    Frost  L,  Jacobsen  CJ,  Christiansen  EH,  Mølgaard  H,  Pilegaard  H,  Hjortholm  K, et al. Hemodynamic predictors of atrial fibrillation or flutter after coronary artery bypass grafting. Acta Anaesthesiol Scand 1995; 39: 690–697.
• 33.    Arsenault  KA,  Yusuf  AM,  Crystal  E,  Healey  JS,  Morillo  CA,  Nair  GM, et al. Interventions for preventing post-operative atrial fibrillation in patients undergoing heart surgery. Cochrane Database Syst Rev 2013; 1: CD003611, doi:10.1002/14651858.CD003611.pub3.
• 34.    Auer  J,  Weber  T,  Berent  R,  Puschmann  R,  Hartl  P,  Ng  CK, et al; Study of Prevention of Postoperative Atrial Fibrillation. A comparison between oral antiarrhythmic drugs in the prevention of atrial fibrillation after cardiac surgery: The pilot study of prevention of postoperative atrial fibrillation (SPPAF), a randomized, placebo-controlled trial. Am Heart J 2004; 147: 636–643.
• 35.    Dahl  JS,  Videbæk  L,  Poulsen  MK,  Pellikka  PA,  Veien  K,  Andersen  LI, et al. Prevention of atrial fibrillation in patients with aortic valve stenosis with candesartan treatment after aortic valve replacement. Int J Cardiol 2013; 165: 242–246.
• 36.    Crystal  E,  Connolly  SJ,  Sleik  K,  Ginger  TJ,  Yusuf  S. Interventions on prevention of postoperative atrial fibrillation in patients undergoing heart surgery: A meta-analysis. Circulation 2002; 106: 75–80.
• 37.    Sakamoto  A,  Kitakaze  M,  Takamoto  S,  Namiki  A,  Kasanuki  H,  Hosoda  S; JL-KNIGHT study group. Landiolol, an ultra-short-acting β-blocker, more effectively terminates atrial fibrillation than diltiazem after open heart surgery: Prospective, multicenter, randomized, open-label study. Circ J 2012; 76: 1097–1101.
• 38.    Sezai  A,  Nakai  T,  Hata  M,  Yoshitake  I,  Shiono  M,  Kunimoto  S, et al. Feasibility of landiolol and bisoprolol for prevention of atrial fibrillation after coronary artery bypass grafting: A pilot study. J Thorac Cardiovasc Surg 2012; 144: 1241–1248.
• 39.    Harling  L,  Saso  S,  Jarral  OA,  Kourliouros  A,  Kidher  E,  Athanasiou  T. Aortic valve replacement for aortic stenosis in patients with concomitant mitral regurgitation: Should the mitral valve be dealt with? Eur J Cardiothorac Surg 2011; 40: 1087–1096.
• 40.    Shibayama  K,  Watanabe  H,  Tabata  M,  Sasaki  S,  Fukui  T,  Umemura  J, et al. Impact of ejection fraction on long-term outcome after elective aortic valve replacement in octogenarians with aortic stenosis. Circ J 2012; 76: 1761–1767.
• 41.    Atarashi  H,  Inoue  H,  Okumura  K,  Yamashita  T,  Kumagai  N,  Origasa  H; J-RHYTHM Registry Investigators. Present status of anticoagulation treatment in Japanese patients with atrial fibrillation: A report from the J-RHYTHM Registry. Circ J 2011; 75: 1328–1333.