|
|
||||||||
J Thorac Cardiovasc Surg 2007;134:1569-1576
© 2007 The American Association for Thoracic Surgery
Surgery for Acquired Cardiovascular Disease |
a Department of Cardiology, Medical University of Lodz, Poland
c Second Department of Family Medicine, Medical University of Lodz, Poland
d Department of Cardiac Surgery, Medical University of Lodz, Poland
b Chair of Statistics of University of Lodz, Poland.
Received for publication June 2, 2007; revisions received August 8, 2007; accepted for publication August 17, 2007. * Address for reprints: Maciej Banach, MD, PhD, Department of Cardiology, 1st Chair of Cardiology and Cardiac Surgery, Medical University of Lodz, Sterlinga 1/3; 91-425 Lodz, Poland. (Email: m.banach{at}termedia.pl).
| Abstract |
|---|
|
|
|---|
Methods: The study comprised 300 patients with aortic valve defects of either aortic stenosis (n = 150) or regurgitation (n = 150) who underwent aortic valve replacement. For each patient, 2-mode and Doppler echocardiographic examinations were performed in the preoperative period, early postoperative period, and long-term observation, and selected hemodynamic parameters were analyzed.
Results: Factors significantly associated with atrial fibrillation in patients with aortic stenosis were heart failure (odds ratio = 5.5), age 70 years or more (4.5), low (3.9) and high body mass index (1.7), maximal transvalvular gradient (3.7), low left ventricular ejection fraction (5.1), end-systolic (2.9) and end-diastolic intraventricular septum thickness (1.5), and insignificant mitral regurgitation (1.9) in the preoperative period; and left ventricular ejection fraction (4.4) and end-systolic intraventricular septum thickness (1.8) in the early postoperative period. In the aortic regurgitation group, factors significantly associated with atrial fibrillation were age (1.8), left ventricular ejection fraction (3.7), left ventricular end-systolic diameter (1.7), end-diastolic intraventricular septum thickness (1.7), left atrium dimension (4.1) and insignificant mitral regurgitation (2.5) in the postoperative period; essential arterial hypertension (3.3), diabetes mellitus (2.6), and heart failure in the history (4.5) in the preoperative period; and left ventricular ejection fraction (1.9) and left atrium dimension (2.9) in the early postoperative period.
Conclusion: On the basis of the separated risk factors, all patients should be preoperatively classified to applicable groups of risk of postoperative atrial fibrillation appearance, and the prophylactic treatment should be administered in the group of patients with the highest risk. It may essentially decrease the rate of complications and deaths, and, consequently, the costs of postoperative medical care.
| Introduction |
|---|
|
|
|---|
Many authors have tried to evaluate the predictors of postoperative AF. On the basis of available studies, they singled out inter alia the following postoperative risk factors: left ventricle failure in the preoperative period, age more than 75 years, reoperation, intraoperative and postoperative catecholamines application, and respiratory failure.2,7-10
The aim of the study was to evaluate the risks factors of paroxysmal AF in patients who underwent aortic valve replacement (AVR).
| Materials and Methods |
|---|
|
|
|---|
The mean age of the patients was 61.5 ± 5.4 years; there were 143 men (47.67%) and 157 women (52.33%). The mean body surface area was 1.87 ± 0.21, body mass index (BMI) was 26.75 ± 3.42, and mean preoperative ejection fraction was 49.13% ± 7.8%. The following concomitant diseases were observed in this group of patients: arterial hypertension (n = 195; 64.7%), diabetes mellitus (n = 60; 20%), renal failure (n = 11; 3.67%), and severe heart failure with a left ventricular ejection fraction (LVEF) of 35% or less (n = 51; 17%).
All patients with the following conditions were excluded from the study: aortic valve defect in the course of infective valve defect, aortic valve defect as a result of myocardial infarction complication, history of myocardial infarction or cerebrovascular event, history of cardiac surgery, 1-step surgery of other valve replacement/plastics and/or surgical revascularization, surgery in emergency/urgent mode, history of preoperative arrhythmias (eg, AF/flutter, ventricular arrhythmias, receiving antiarrhythmic drugs, including beta-blockers), or other significant coexistent conditions (eg, severe renal, pulmonary diseases, and neoplasms).
Each patient who qualified for the AVR underwent coronary angiography (on average 33 ± 7.7 hours before the surgery, range 19–63 hours), and no significant changes in coronary arteries were found. For each patient included in the study, 2-mode and Doppler echocardiographic examinations were performed in the preoperative period (up to 48 hours before the operation), in the early postoperative period (4–21 days after surgery; on average after 9 days), and in long-term observation as a follow-up examination (18–24 months after the surgery; on average after 21 months). The following echocardiographic parameters were evaluated: LVEF, left ventricular end-systolic and end-diastolic diameters, end-systolic intraventricular septum thickness (ESIVST), end-diastolic intraventricular septum thickness (EDIVST), left atrium dimension (LAd), and mean and maximal transvalvular gradients for patients with aortic stenosis. All investigations were performed on Philips Hewlett Packard Sonos 2000AQ (Philips/Hewlett Packard, Andover, Mass) and ACUSON Sequoia Echo C256 (Siemens, New York, NY) ultrasound systems. The detailed characteristics of the patients are shown in Table 1.
|
Surgical Intervention
Each patient included in the study underwent AVR with extracorporeal circulation. After removal of the native valve, all patients were implanted with mechanical valves from St Jude Medical (St Paul, Minn). Valve sizes ranged from 19 to 31 mm. All patients underwent AVR at normothermia with the use of antegrade cold crystalloid St Thomas Hospital cardioplegic solution No. 2 (4°C-6°C).
Atrial Fibrillation Evaluation
Postoperative paroxysmal AF was diagnosed on the basis of multiple electrocardiographic examinations and confirmed in the 24-hour Holter monitoring. We considered paroxysmal AF when the arrhythmia reverted spontaneously or after treatment to sinus rhythm within 1 week.
Statistical Analysis
Statistical analyses were performed with STATISTICA PL 7.0 (StatSoft, Cracow, Poland) and SPSS 12.0 Software (SPSS Inc, Chicago, Ill). Normality was tested using the Shapiro–Wilk test. The association between the potential risk factors and the mortality rate was first evaluated by univariate analysis. For categoric variables, the chi-square test was used. The diagnostic utility of continuous risk factors was estimated through the use of receiver operating characteristic curves. The results were expressed in terms of the area under the curve with a 95% confidence interval for this area. Factors significant to at least a P value less than .10 were then analyzed using multivariate logistic regression (odds ratio, ±95% confidence interval, P value), which was used to identify the independent clinical predictors of postoperative AF.
| Results |
|---|
|
|
|---|
|
|
70 years (P < .001), BMI both
30 kg/m2 and
21 kg/m2 (P < .05 and P < .001, respectively), maximal transvalvular gradient
85 mm Hg (P < .001), LVEF
50% (P < .001), ESIVST
1.8 cm (P < .005) and EDIVST
1.4 cm (P < .05), insignificant mitral regurgitation (P < .02) in the preoperative period, and LVEF
50% (P < .001) and ESIVST
1.8 cm (P < .02) in the early postoperative period. Postoperative AF, in comparison with the conditions of the remaining 88 patients without postoperative AF, was associated with the following: increase in the length of intensive care unit (ICU) and hospital stay (3.26 ± 1.54 days vs 2.96 ± 1.17 days, respectively, P < .01, and 13.79 ± 6.27 vs 11.92 ± 5.64 days, respectively, P < .001); greater incidence of stroke (5 [8.1%] and 4 [4.5%] events in AF (+) and AF (–), respectively, P < .01); low output syndrome (17 [27.4%] and 22 [25%] events in AF (+) and AF (–), respectively, P < .05); and early postoperative death (5 [8.1%] and 3 [3.4%] deaths in AF (+) and AF (–), respectively, P < .001) (Table 2). Multivariate logistic regression analysis identified 7 independent predictors of postoperative paroxysmal AF in patients with aortic stenosis who underwent AVR: advanced age, history of heart failure, BMI
21 kg/m2, maximal transvalvular gradient
85 mm Hg, LVEF
50% in the pre- and early postoperative periods, and ESIVST
1.8 cm before the surgery (Table 3).
Patients With Aortic Regurgitation
According to statistical analysis, factors significantly associated with AF in patients who underwent AVR for aortic regurgitation were as follows: age
70 years (P < .05), LVEF
50% (P < .005), left ventricular end-systolic diameter
4.15 cm (P < .02), EDIVST
1.35 cm (P < .05), LAd
4.25 cm (P < .005), insignificant mitral regurgitation (P < .01) in the preoperative period, essential arterial hypertension
160/100 mm Hg (second-degree hypertension according to the European Society of Hypertension/European Society of Cardiology 2007 guidelines) (P < .001), diabetes mellitus (P < .001), heart failure in the history (P < .001), and LVEF
50% (P < .05) and LAd
4.15 cm (P < .005) in the early postoperative period. Postoperative AF, in comparison with the conditions of the remaining 81 patients without postoperative AF, was associated with an increase in the length of ICU and hospital stay (4.28 ± 1.38 days vs 3.45 ± 1.33 days, respectively, P < .001, and 15.14 ± 2.21 days vs 12.93 ± 3.36 days, respectively, P < .001); a greater incidence of stroke (5 [7.2%] and 3 [3.7%] events in AF (+) and AF (–), respectively, P < .001); low output syndrome (24 [34.8%] and 23 [28.4%] events in AF (+) and AF (–), respectively, P < .02); and early postoperative death (7 [10.1%] and 4 [4.9%] deaths in AF (+) and AF (–), respectively, P < .001). By comparing the long-term mortality, we observed a persistent significant difference between groups (6 [9.7%] and 3 [3.9%] deaths in AF (+) and AF (–), respectively, P < .001) (Table 4). Multivariate logistic regression analysis identified 7 independent predictors of postoperative paroxysmal AF in patients with aortic regurgitation: history of heart failure, diabetes mellitus and essential arterial hypertension, LVEF
50%, insignificant mitral regurgitation in the preoperative period, and LAd
4.25 cm and
4.15 cm in the pre- and early postoperative periods, respectively (Table 5).
|
|
70 years (P < .05), LVEF
50% (P < .001), left ventricular end-systolic diameter
3.85 cm (P < .003) and end-diastolic diameter
5.5 cm (P < .02), ESIVST
1.75 cm (P < .001) in the preoperative period; heart failure in the medical history (P < .02), LVEF
50% (P < .001), left ventricular end-systolic diameter
3.75 cm (P < .003), and left ventricular end-diastolic diameter
5.35 cm (P < .05) in the early postoperative period; and LVEF
50% (P < .001) and left ventricular end-systolic and end-diastolic diameters
3.6 cm (P < .003) and
5.1 cm (P < .04), respectively, in the long-term observations. On the basis of univariate statistical analysis, we showed a significant correlation between postoperative paroxysmal AF and the type of aortic valve defect (P < .03 for aortic regurgitation), postoperative low cardiac output syndrome (P < .001), and early postoperative death (P < .002). Multivariate logistic regression analysis showed that postoperative AF was an independent risk factor of early postoperative death (odds ratio = 3.9; 95% confidence interval 1.7–9.0; P < .002). Postoperative AF was associated with an increase in the length of ICU and hospital stay (3.85 ± 1.23 days vs 3.12 ± 1.93 days, respectively, P < .001, and 14.75 ± 2.81 days vs 11.97 ± 2.78 days, respectively, P < .001). Multivariate logistic regression analysis identified 6 independent predictors of postoperative paroxysmal AF in patients who underwent AVR for an aortic valve defect: low BMI
21 kg/m2, LVEF
50%, left ventricular end-systolic diameter
3.85 cm, ESIVST
1.75 cm in the preoperative period, and LVEF
50% in early postoperative period and the type of aortic valve defect (aortic regurgitation) (Table 6).
|
| Discussion |
|---|
|
|
|---|
Few studies present predictors of postoperative AF in patients who underwent AVR. In Orlowska-Baranowska and colleagues study,17
the authors identified age, New York Heart Association functional class, history of preoperative paroxysmal AF, left ventricular mass index > 300 supraventricular beats on 24-hour electrocardiography before surgery, presence of supraventricular tachycardia, supraventricular tachycardia of > 5 beats, or a rate > 120 beats/min as risk factors of postoperative paroxysmal AF. By using multivariate analysis, they identified the following 4 variables as independent predictors: age, history of paroxysmal AF, presence of > 300 supraventricular beats/24 hours, and presence of supraventricular tachycardia. In Ducceschi and colleagues study,18
the authors included 302 patients and divided them into 2 groups according to the absence or evidence of post-AVR AF. Post-AVR paroxysmal AF occurred in 19% of patients. They noticed that AF post-AVR was associated with advanced age, left atrial enlargement, preoperative episodes of paroxysmal AF, use of a warm blood cardioplegic solution and normothermia, administration of inotropic agents, prolonged assisted ventilation, electrolyte imbalance, and atrioventricular and intraventricular conduction disorders. By using multivariate logistic regression analysis, the authors identified age, left atrial enlargement, a history of paroxysmal AF, and postoperative electrolyte imbalance as independent correlates of AF, whereas the use of hypothermia seemed to be a protective factor. These results are partially consistent with the outcomes obtained in our study. Both pre- and early postoperative LVEF
50% and history of heart failure were independent predictors of postoperative AF. We also noticed that changes of some hemodynamic parameters, such as ESIVST and EDIVST for patients with aortic stenosis and ESIVST, left ventricular end-systolic diameter, and LAd for aortic regurgitation, significantly influenced the appearance of postoperative AF.
Despite the risk factors presented, we also found some other predictors that have not been mentioned in the available studies, for example, low BMI and maximal transvalvular gradient for patients with aortic stenosis, and a history of insignificant mitral regurgitation, essential atrial hypertension, and diabetes mellitus for patients with aortic regurgitation. Low BMI as an independent predictor of postoperative AF in the patients with aortic stenosis and the combined group of patients who underwent AVR, irrespective of the type of valve defect, seems to be exceptionally interesting, because these results confirmed previous reports suggesting that low BMI may worsen the prognosis and increase the risk of cardiovascular events and mortality in patients with heart diseases.19-22
Furthermore, in the other study we observed that low BMI and the previously stated essential arterial hypertension were also independent predictors of postoperative in-hospital death in patients who underwent AVR.23-25
In the present study, we also analyzed predictors of postoperative paroxysmal AF in the combined group of 300 patients who underwent AVR. We showed that in addition to few other significant predictors, the type of aortic valve defect, aortic regurgitation, was itself an independent risk factor of postoperative AF. There are no available studies indicating that the type of valve defect, aortic regurgitation, might influence postoperative AF occurrence irrespective of other pre- and postoperative risk factors. It means that patients with aortic regurgitation qualified for AVR should be treated as patients with a higher risk of postoperative AF. Further studies are necessary to confirm these results.
The present study has some limitations, which are mainly connected with the number of patients included in the study and consequently the number of patients with postoperative AF. We also did not analyze many other potential AF predictors, especially from the intraoperative period. However, we were aware that it would have been difficult to examine all possible risk factors, so we focused specifically on the hemodynamic parameters and the impact of postoperative AF on them.
| Conclusions |
|---|
|
|
|---|
The study may have an important influence on the management of patients who undergo AVR. On the basis of the separated risk factors, all patients should be preoperatively classified to applicable groups of risk of postoperative AF appearance, and the prophylactic treatment (pre- and postoperative) should be administered in the highest risk group of patients. It may essentially decrease the rate of complications and deaths, and the costs of postoperative medical care.
| References |
|---|
|
|
|---|
ski P, Rysz J, Piechowiak M, Zas
onka J. Prevention and treatment of atrial fibrillation following cardiac surgery. Pol Merkur Lekarski 2005;19:794-799.[Medline]
onka J, Iwaszkiewicz A, Barylski M, Irzma
ski R, et al. Is there a relation between preoperative atrial fibrillation and increased mortality after cardiac surgery?. Folia Cardiol 2005;12(Suppl D):102-104.
ski P, Ko
ider A, Irzma
ski R, Rysz J, Olszewski R, et al. Aortic valve replacement in patients with heart failure. Pol Merkur Lekarski 2006;20:642-645.[Medline]This article has been cited by other articles:
![]() |
K. O'Connor, J. Magne, M. Rosca, L. A. Pierard, and P. Lancellotti Left atrial function and remodelling in aortic stenosis Eur Heart J Cardiovasc Imaging, April 1, 2011; 12(4): 299 - 305. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Leon, N. Piazza, E. Nikolsky, E. H. Blackstone, D. E. Cutlip, A. P. Kappetein, M. W. Krucoff, M. Mack, R. Mehran, C. Miller, et al. Standardized Endpoint Definitions for Transcatheter Aortic Valve Implantation Clinical Trials: A Consensus Report From the Valve Academic Research Consortium J. Am. Coll. Cardiol., January 18, 2011; 57(3): 253 - 269. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Leon, N. Piazza, E. Nikolsky, E. H. Blackstone, D. E. Cutlip, A. P. Kappetein, M. W. Krucoff, M. Mack, R. Mehran, C. Miller, et al. Standardized endpoint definitions for transcatheter aortic valve implantation clinical trials: a consensus report from the Valve Academic Research Consortium Eur. Heart J., January 6, 2011; (2011) ehq406v2. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. E. Vardas, E. P. Koutalas, and E. N. Simantirakis Chapter 54 Atrial fibrillation and supraventricular arrhythmias The ESC Textbook of Acute and Intensive Cardiac Care, December 1, 2010; 1(1): med-9780199584314-chapter - med-9780199584314-chapter. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Filardo, C. Hamilton, B. Hamman, R. F. Hebeler Jr, J. Adams, and P. Grayburn New-Onset Postoperative Atrial Fibrillation and Long-Term Survival After Aortic Valve Replacement Surgery Ann. Thorac. Surg., August 1, 2010; 90(2): 474 - 479. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Kozlowski, S. Budrejko, G. Y H Lip, J. Rysz, D. P Mikhailidis, G. Raczak, and M. Banach Lone atrial fibrillation: what do we know? Heart, April 1, 2010; 96(7): 498 - 503. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Dizon, K. Chen, M. Bacchetta, M. Argenziano, D. Mancini, A. Biviano, J. Sonett, and H. Garan A Comparison of Atrial Arrhythmias After Heart or Double-Lung Transplantation at a Single Center: Insights Into the Mechanism of Post-Operative Atrial Fibrillation J. Am. Coll. Cardiol., November 24, 2009; 54(22): 2043 - 2048. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |