- Email: [email protected]
Analysis of risk factors for development of atrial fibrillation early after cardiac valvular surgery
Analysis of Risk Factors for Development of Atrial Fibrillation Early After Cardiac Valvular Surgery Craig R. Asher, MD, Dave P. Miller, MS, Richard A. Grimm, Delos M. Cosgrove III, MD, and Mina K. Chung, MD
DO,
Atrial fibrillation (AF) commonly develops after cardiac valvular surgery. The objective of this study was to identify risk factors for postoperative AF following valvular surgery. A cohort of 915 consecutive adult patients undergoing isolated valvular surgery with preoperative sinus rhythm was analyzed. Univariate and independent multivariate risk factors for postoperative AF were determined. A second cohort of 305 patients with the same inclusion criteria was used to validate the multivariate predictors. Patients studied had a mean age of 56.1 6 14.7 years, 57.9% were men, 79.6% had a normal left ventricular ejection fraction, and their mean left atrial size was 46.2 6 9.3 mm. The incidence of
postoperative AF was 36.7%. Independent predictors of postoperative AF included: advanced age (odds ratio [OR] 1.506 per decade, 95% confidence interval, [CI] 1.35 to 1.68, p 5 0.0001); mitral stenosis (OR 2.066, CI 1.21 to 3.52, p 5 0.0077); left atrial enlargement (OR 1.468, CI 1.07 to 2.01, p 5 0.0165); use of systemic hypothermia (OR 0.572, CI 0.422 to 0.776, p 5 0.0003); and a history of cardiac surgery (OR 0.676, CI 0.465 to 0.981, p 5 0.0393). Among these variables, advanced age, mitral stenosis, and left atrial enlargement were confirmed as independent risk factors in the validation cohort. Q1998 by Excerpta Medica, Inc. (Am J Cardiol 1998;82:892– 895)
trial fibrillation (AF), the most common complication after cardiac surgery, occurs with an inciA dence after valvular surgery that exceeds that after
Data collection: In general, patients were monitored for arrhythmias continuously for $3 days. Electrocardiograms were performed routinely in all patients on the day of surgery, and the first, second, and fifth postoperative days. Postoperative AF was defined when any episode of this arrhythmia was recorded on an electrocardiogram or telemetry strip. The variables considered for review included .40 preoperative and intraoperative factors, including demographic, clinical, electrocardiographic and echocardiographic variables, as well as surgical and myocardial protection methods. Surgical and myocardial protection variables included the surgical approach (medial vs lateral thoracotomy), type of surgery (valve location and procedure), use of systemic hypothermia, type and method of cardioplegia delivery, type of cannulation, use of circulatory arrest, total and maximum aortic occlusion, and cardiopulmonary bypass times. Statistical analysis: Statistical analysis was performed using the SAS program (Cary, North Carolina). Continuous variables are reported as mean 6 SD and compared using 2-tailed Student’s t tests. Categorical variables are reported as percentages or frequencies and were analyzed using chi-square tests. Univariate predictors of postoperative AF were determined. Clinically significant univariate factors (p ,0.1) and all factors with p ,0.05 were entered into a multivariate model in a stepwise fashion to identify factors independently associated with the development of postoperative AF. Factors were considered significant in the multivariate model at p #0.05. To validate the independent predictors, a second cohort of patients was selected. The CVIR database identified 305 patients between December 1995 and October 1996 who met the same inclusion criteria as the initial study population.
coronary artery bypass surgery. In comparison to the 17% to 33% incidence of postoperative AF following coronary bypass surgery,1–7 AF develops after cardiac valvular surgery in 38% to 64% of patients.2,8 The objective of this study was to identify factors that predispose patients undergoing valvular surgery to the development of postoperative AF, so that high-risk patients can be targeted for aggressive prophylactic trials and low-risk patients spared from unnecessary treatment.
METHODS A retrospective analysis of patients undergoing cardiac valvular operations performed at the Cleveland Clinic Foundation between January 1993 and October 1996 was performed. Data were obtained from the Cardiovascular Information Registry (CVIR), a regularly updated, quality-controlled clinical database. Patient population: The study population consisted of patients with the following inclusion criteria: (1) age $18 years; (2) sinus rhythm on preoperative electrocardiogram; and (3) isolated cardiac valvular surgery without concomitant coronary bypass grafting or other surgical procedures. Most operations were performed by 6 cardiac surgeons. The baseline characteristics of the population studied are shown in Table I. From the Departments of Cardiology, Biostatistics and Epidemiology, and Thoracic and Cardiovascular Surgery,The Cleveland Clinic Foundation, Cleveland, Ohio. Manuscript received February 13, 1998; revised manuscript received and accepted May 14, 1998. Address for reprints: Mina K. Chung, MD, Department of Cardiology, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk F-15, Cleveland, Ohio 44195.
892
©1998 by Excerpta Medica, Inc. All rights reserved.
0002-9149/98/$19.00 PII S0002-9149(98)00498-6
TABLE I Baseline Characteristics of Study Population (n 5 915) Age (yrs) Men Unstable preoperative status Prior cardiac surgery Preoperative history Systemic hypertension Diabetes mellitus Rheumatic heart disease Congestive heart failure Left ventricular ejection fraction Normal Mild/mild-moderate/moderate Moderate–severe/severe Left atrial size (mm) Total aortic occlusion time (mins) Total cardiopulmonary bypass time (min)
TABLE II Incidence of Postoperative Atrial Fibrillation (AF) Following Cardiac Surgery
56.1 6 14.7 530 (57.9%) 23 (2.5%) 205 (22.4%) 359 196 128 299
(42.0%) (23.7%) (14.0%) (32.7%)
657 (79.6%) 135 (16.4%) 33 (4.0%) 46.2 6 9.3 65.7 6 30.2 91.6 6 50.7
RESULTS During the study period between January 1993 and November 1995, 3,163 patients underwent valvular surgery. Excluded from the study were 1,457 patients who had concomitant coronary revascularization or other procedures. Of the 1,706 patients in which isolated valvular surgery was performed, 29 patients were excluded due to age ,18 years, 760 patients due to preoperative electrocardiograms showing a rhythm other than sinus rhythm, and 2 patients due to a missing outcome variable. Sinus rhythm was present preoperatively in 915 patients, representing 54% of the population undergoing isolated valvular surgery. Incidence of postoperative atrial fibrillation: The incidence of postoperative AF by valve type and procedure is summarized in Table II. The combined incidence of AF for all patients following isolated valvular operations was 36.7%. For comparison, a cohort of 4,314 patients that underwent isolated coronary revascularization during the same time period as the study population of valvular surgery patients, was selected from the CVIR database. The incidence of postoperative AF for all patients undergoing isolated valvular operations in our study population exceeded the 27.5% incidence for the cohort of coronary revascularization patients (p 5 0.001). This occurred despite the younger age of patients undergoing isolated valve surgery (56 and 62 years of age for the isolated valve and isolated coronary bypass patients, respectively; p ,0.001). Comparisons between the incidences of valvular surgeries were limited by the small number of these operations performed. Univariate analysis: Univariate predictors of postoperative AF after isolated valvular surgery were identified. Variables with a p value ,0.10 are listed in Table III. The incidence of postoperative AF in unstable patients was significantly lower than in stable patients (13.0% vs 37.4%, p 5 0.02). However, the low incidence of postoperative AF among unstable patients may be attributable to the 45% perioperative mortality occurring in this population. Significant valvular or structural factors included mitral stenosis and left atrial enlargement. The only significant intraoperative factor was the use of systemic hypothermia as
Type of Operation AV repair AV replacement MV repair MV replacement MV repair/AV replacement AV/MV replacement AV/PV replacement* MV replacement/TV repair Other repair or replacement† Isolated valve Isolated CABG‡ CABG/valve§
n
%AF
49 333 296 80 40 26 37 18 36 915 4,314 569
24.5 36.9 38.2 35.0 52.5 50.0 18.9 27.8 39.0 36.7#\ 27.5# 41.1\
*Aortic and pulmonic valve replacement 5 Ross procedure. † All other combinations of repair and replacement involving 2 or 3 valves including tricuspid (TV), pulmonic (PV), aortic (AV), and mitral (MV) valves. ‡ CABG 5 isolated coronary bypass grafting performed during the same study period as the isolated valve population. § CABG/valve 5 combined coronary bypass and valve surgery performed during the same study period as the isolated valve population. # p 5 0.001, isolated valve surgery versus CABG. \ p 5 0.09, isolated valve surgery versus CABG/valve.
part of the myocardial protection delivery method, which was associated with a lower incidence of postoperative AF. Multivariate analysis: Independent risk factors associated with postoperative AF following isolated valvular surgeries were identified (Table IV). These 5 independent variables were further assessed individually. There was an increasing incidence of postoperative AF with advancing age. The incidence of AF was 14.0% in patients ,35 years of age, but over 3-fold higher at 50.5% in patients .75 years of age. The presence of left atrial enlargement as determined by echocardiographic or surgical observation was an independent risk factor for postoperative AF. The severity of left atrial enlargement was correlated with increased risk of postoperative AF. Transthoracic or transesophageal echocardiograms within 6 months of surgery were available for 73% of patients. The incidence of AF in patients with normal left atrial dimensions (#40 mm), was significantly less than for those with abnormal dimensions (.40 mm; 24.8% vs 37.4%, p ,0.005). Among patients with severe left atrial enlargement (dimensions .60 mm ), the incidence of postoperative AF increased to 44.4%. Mitral stenosis was a significant risk factor for postoperative AF independent of left atrial enlargement. Of the 71 patients with echocardiographic mitral stenosis who underwent mitral valve surgery, 36 (50.7%) developed postoperative AF compared with 35.6% of patients with all other valvular disorders (p ,0.01). Reoperation and the use of systemic hypothermia were found to be associated with a reduced incidence of postoperative AF. At the time of a first operative procedure, 61.1% of patients were in sinus rhythm. However, sinus rhythm was present in 43.5% at the
VALVULAR HEART DISEASE/POSTOPERATIVE AF AND VALVULAR SURGERY
893
TABLE III Univariate Predictors of Postoperative Atrial Fibrillation (AF) After Valvular Surgery
DISCUSSION
The current study reports the largest analysis to date of AF occurring Variable Present (%AF) Absent (%AF) p Value after valvular surgery. Postoperative AF occurred in 36.7% of patients, Reoperation 59/205 (28.8) 277/710 (39.0) ,0.01 Systemic hypothermia 141/452 (31.2) 195/463 (42.1) ,0.01 exceeding rates reported at the same History of thyroid disease 35/76 (46.1) 301/839 (35.9) 0.08 institution after coronary bypass Stable preoperative status 333/891 (37.4) 3/23 (13.0) 0.02 grafting. Based on these estimates, Mitral stenosis 36/71 (50.7) 298/838 (35.6) 0.01 approximately one-third of the Left atrial enlargement 193/468 (41.2) 143/447 (32.0) ,0.01 Age .65 yrs 146/301 (48.5) 190/614 (30.9) ,0.01 70,000 patients that undergo valvular surgery annually will develop new postoperative AF.9 Because the pathogenesis and risk factors for AF TABLE IV Multivariate Predictors of Postoperative Atrial after valvular heart surgery are not well understood, Fibrillation Following Valvular Surgery development of preventive therapies have been limVariable OR (95% CI) p Value ited. Mitral stenosis 2.066 (1.21–3.52) 0.0077 Incidence: Despite the generally younger age of Age (per decade) 1.506 (1.35–1.68) 0.0001 patients undergoing isolated valvular surgery comLeft atrial enlargement 1.468 (1.07–2.01) 0.0165 pared with coronary bypass grafting, the incidences of Systemic hypothermia 0.572 (0.422–0.776) 0.0003 postoperative AF are greater in valve patients. The Reoperation 0.676 (0.465–0.981) 0.0393 greater inherent susceptibility to postoperative AF after valvular heart surgery may result from more prevalent structural and hemodynamic abnormalities, such time of a second operation and only 34.7% of patients as left atrial enlargement and pathologic changes in at the time of $3 operations. Patients in sinus rhythm the atria from aging and rheumatic mitral stenosis. before a reoperation had a lower incidence of postopPredictors of postoperative atrial fibrillation: In our erative AF than patients undergoing a first operation, analysis of 915 patients undergoing isolated cardiac respectively (59 of 205, 28.8% vs 277 of 710, 39.0%, valvular surgery, 5 pre- and intraoperative factors were identified that predicted the occurrence of postp 5 0.007). The use of systemic hypothermia was associated operative AF. The review by Creswell et al,2 of 297 with a lower incidence of postoperative AF compared patients undergoing isolated valvular surgery, found with no hypothermia (42.1% vs 31.2%, respectively, p age to be the only independent predictor of postoper,0.01). The degree of systemic hypothermia tended ative AF. A smaller analysis8 of 135 patients underto correlate with the lowest achieved temperature, going aortic valve replacement with coronary bypass with temperatures of 31° to 35°, 25° to 30°, and ,25° grafting discovered several correlations of hemodyassociated with incidences of postoperative AF of namic and clinical variables with postoperative AF. AGE: A consistent finding among studies of AF 32.1%, 30.9%, and 27.9%, respectively. occurring following coronary artery surgery is an inValidation of independent variables: Three of the 5 creasing incidence with advancing age.1– 4,6,7,10 Explaindependent risk factors were confirmed to be signifnations for the impact of age on this arrhythmia are icant factors among the validation cohort of 305 paspeculative and include structural atrophic or ischemic tients. Although only 16 patients had mitral stenosis in effects within the atria and related electrophysiologic the validation cohort, the incidence of postoperative changes. These age-related changes in the atria likely AF was markedly higher in these patients compared occur in patients undergoing either valvular surgery or with patients with other valvular disorders (69% vs. coronary bypass grafting. 27%, p ,0.001). An enlarged left atrium was again In patients undergoing isolated valvular surgery, predictive of higher postoperative AF rates (36% vs this factor has additional significance. Thirty-four per19%, p 5 0.002). Advanced age continued to be cent of patients were #50 years old in our study and strongly associated with postoperative AF, with rates had a low postoperative AF incidence of 19.4%. These of 12%, 21%, 37%, 37%, 46%, and 67% for patients younger patients represent an easily identifiable low30 to 39, 40 to 49, 50 to 59, 60 to 69, 70 to 79, and risk group for postoperative AF that may not require $80 years old, respectively. prophylactic treatment. The 2 variables that were protective factors in the LEFT ATRIAL ENLARGEMENT: Two prior studies of initial group were not confirmed in the validation coronary bypass surgery patients have considered the cohort. Systemic hypothermia was not associated with effect of left atrial enlargement as a predisposing a lower incidence of postoperative AF in the valida- factor for supraventricular tachyarrhythmias.6,11 Only tion cohort (32% with hypothermia vs 29% without the study of Dixon et al11 found that supraventricular hypothermia, p 5 0.701). Although lower event rates tachyarrhythmias in the postoperative period occurred were again observed in patients undergoing reopera- more frequently for patients with left atrial enlargetion versus first operations (26% vs 31% respectively, ment (RR 5 1.74). p 5 0.378), this difference did not achieve statistical Although it is apparent that postoperative AF is significance. more common with any left atrial enlargement, the 894 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 82
OCTOBER 1, 1998
degree of dilation also adds to this risk. However, even among the patients with severe left atrial enlargement, the incidence of postoperative AF was not prohibitively high (44%) to forego consideration of the benefits of prophylactic therapy to prevent this arrhythmia. MITRAL STENOSIS: The importance of mitral stenosis as a predictor of postoperative AF, independent of left atrial enlargement, further demonstrates the potential multifactorial etiology of this arrhythmia in patients with mitral stenosis. The predominant cause of mitral stenosis in adults is rheumatic heart disease.12 This high incidence of arrhythmias is believed to be due to both the hemodynamic effects of elevated left atrial pressure and the structural changes that include myofibril hypertrophy and fibrosis of the atrial wall.13 REOPERATIONS: The lower incidence of postoperative AF following reoperations likely represented a natural selection bias. Increasing age and the progression of hemodynamic and structural effects on the left atrium may enhance the propensity to develop chronic AF over time, and patients with chronic AF were excluded from this study. Therefore, the patients undergoing reoperation in sinus rhythm represent a group that has already shown some resistance to chronic AF. SYSTEMIC HYPOTHERMIA: The importance of intraoperative atrial temperatures and mechanical activity has been widely studied. Data from experimental and small clinical studies have demonstrated an effect of intraoperative atrial and ventricular temperatures on supraventricular conduction.14,15 Although conflicting data exist, an association between intraoperative atrial activity detected by atrial electrograms and postoperative atrial arrhythmias has also been suggested.16 –18 The use of systemic hypothermia during isolated valvular surgery was associated with a reduced incidence of postoperative AF in our initial study population. However, this relation between the use of systemic hypothermia and postoperative AF was not present in the validation cohort. Thus, similar to other studies, the benefit of augmented hypothermia remains unproven.19 Conclusion: Postoperative AF is common following all types of isolated valvular surgery. In our study cohort, 5 independent variables predicted patients risk for postoperative AF. Three of the 5 independent variables were validated in a separate cohort and included advanced age, mitral stenosis, and left atrial
enlargement. Although effective prophylactic methods have not been well studied in patients undergoing valvular surgery, patients with these clinical characteristics are among those that may likely benefit from prophylactic efforts. 1. Leitch JW, Thomson D, Baird DK, Harris PJ. The importance of age as a predictor of atrial fibrillation and flutter after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1990;100:338 –342. 2. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539 –549. 3. Crosby LH, Pifalo WB, Woll KR, Burkholder JA. Risk factors for atrial fibrillation after coronary artery bypass grafting. Am J Cardiol 1990;66:1520 – 1522. 4. Fuller JA, Adams GG, Buxton B. Atrial fibrillation after coronary artery bypass grafting: is it a disorder of the elderly. J Thorac Cardiovasc Surg 1989;97:821– 825. 5. Vecht RJ, Nicolaides EP, Ikweuke JK, Liassides C, Cleary J, Cooper WB. Incidence and prevention of supraventricular tachyarrhythmias after coronary bypass surgery. Int J Cardiol 1986;13:125–134. 6. Mathew JP, Parks R, Savino JS, Friedman AS, Koch C, Mangano DT, Browner WS, for the MultiCenter Study of Perioperative Ischemia Research Group. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA 1996;276:300 –306. 7. Aranki SF, Shaw DP, Adams DH, Rizzo RJ, Couper GS, VanderVliet M, Collins JJ, Cohn LH, Burstin HR. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996;94:390 –397. 8. Douglas P, Hirshfeld JW, Edmunds LH. Clinical correlates of postoperative atrial fibrillation (abstr.). Circulation 1984;70 (Suppl II):II-165. 9. American Heart Association, Heart and Stroke Facts, Dallas, TX, 1993. 10. Frost L, Molgaard H, Christiansen EH, Hjortholm K, Paulsen PK, Thomsen PB. Atrial fibrillation and flutter after coronary artery bypass surgery: epidemiology, risk factors and preventive trials. Int J Cardiol 1992;36:253–261. 11. Dixon FE, Genton E, Vacek JL, Moore CB, Landry J. Factors predisposing to supraventricular tachyarrhythmias after coronary bypass grafting. Am J Cardiol 1986;58:476 – 478. 12. Gaasch WH, O’Rourke RA, Cohn LH, Rackley CE. Mitral valve disease. In: JW Hurst, ed. The Heart. 8th Edition. New York: McGraw-Hill; 1994:chapter 80. 13. Bailey GW, Braniff BA., Hancock EW, Cohn KE. Relation of left atrial pathology to atrial fibrillation in mitral valvular disease. Ann Intern Med 1968; 69:13–20. 14. Sato S, Yamauchi S, Schuessler RB, Boineau JP, Matsunaga Y, Cox JL. The effect of augmented atrial hypothermia on atrial refractory period, conduction and atrial flutter/ fibrillation in the canine heart. J Thorac Cardiovasc Surg 1992;104: 297–306. 15. Smith PK, Buhrman WC, Levett JM, Ferguson TB, Holman WL, Cox JL. Supraventricular conduction abnormalities following cardiac operations. A complication of inadequate atrial preservation. J Thorac Cardiovasc Surg 1983;85: 105–115. 16. Tchervenkov CI, Wynands JE, Symes JF, Malcolm ID, Dobell AR, Morin JE. Persistent atrial activity during cardioplegic arrest: a possible factor in the etiology of postoperative supraventricular tachyarrhythmias. Ann Thorac Surg 1983;36:437– 443. 17. Mullen JC, Khan N, Weisel RD, Christakis GT, Teoh KH, Madonik MM, Mickle DA, Ivanov J. Atrial activity during cardioplegia and postoperative arrhythmias. J Thorac Cardiovasc Surg 1987;94:558 –565. 18. Dewar ML, Rosengarten MD, Blundell PE, Chiu RC. Perioperative holter monitoring and computer analysis of dysarrhythmias in cardiac surgery. Chest 1985;87:593–597. 19. Cheung EH, Arcidi JM, Jackson ER, Hatcher CR, Guyton RA. Intracavitary right heart cooling during coronary bypass surgery: a prospective randomized trial. Circulation 1988;78(suppl III):III-173–III-179.
VALVULAR HEART DISEASE/POSTOPERATIVE AF AND VALVULAR SURGERY
895
DO,
Atrial fibrillation (AF) commonly develops after cardiac valvular surgery. The objective of this study was to identify risk factors for postoperative AF following valvular surgery. A cohort of 915 consecutive adult patients undergoing isolated valvular surgery with preoperative sinus rhythm was analyzed. Univariate and independent multivariate risk factors for postoperative AF were determined. A second cohort of 305 patients with the same inclusion criteria was used to validate the multivariate predictors. Patients studied had a mean age of 56.1 6 14.7 years, 57.9% were men, 79.6% had a normal left ventricular ejection fraction, and their mean left atrial size was 46.2 6 9.3 mm. The incidence of
postoperative AF was 36.7%. Independent predictors of postoperative AF included: advanced age (odds ratio [OR] 1.506 per decade, 95% confidence interval, [CI] 1.35 to 1.68, p 5 0.0001); mitral stenosis (OR 2.066, CI 1.21 to 3.52, p 5 0.0077); left atrial enlargement (OR 1.468, CI 1.07 to 2.01, p 5 0.0165); use of systemic hypothermia (OR 0.572, CI 0.422 to 0.776, p 5 0.0003); and a history of cardiac surgery (OR 0.676, CI 0.465 to 0.981, p 5 0.0393). Among these variables, advanced age, mitral stenosis, and left atrial enlargement were confirmed as independent risk factors in the validation cohort. Q1998 by Excerpta Medica, Inc. (Am J Cardiol 1998;82:892– 895)
trial fibrillation (AF), the most common complication after cardiac surgery, occurs with an inciA dence after valvular surgery that exceeds that after
Data collection: In general, patients were monitored for arrhythmias continuously for $3 days. Electrocardiograms were performed routinely in all patients on the day of surgery, and the first, second, and fifth postoperative days. Postoperative AF was defined when any episode of this arrhythmia was recorded on an electrocardiogram or telemetry strip. The variables considered for review included .40 preoperative and intraoperative factors, including demographic, clinical, electrocardiographic and echocardiographic variables, as well as surgical and myocardial protection methods. Surgical and myocardial protection variables included the surgical approach (medial vs lateral thoracotomy), type of surgery (valve location and procedure), use of systemic hypothermia, type and method of cardioplegia delivery, type of cannulation, use of circulatory arrest, total and maximum aortic occlusion, and cardiopulmonary bypass times. Statistical analysis: Statistical analysis was performed using the SAS program (Cary, North Carolina). Continuous variables are reported as mean 6 SD and compared using 2-tailed Student’s t tests. Categorical variables are reported as percentages or frequencies and were analyzed using chi-square tests. Univariate predictors of postoperative AF were determined. Clinically significant univariate factors (p ,0.1) and all factors with p ,0.05 were entered into a multivariate model in a stepwise fashion to identify factors independently associated with the development of postoperative AF. Factors were considered significant in the multivariate model at p #0.05. To validate the independent predictors, a second cohort of patients was selected. The CVIR database identified 305 patients between December 1995 and October 1996 who met the same inclusion criteria as the initial study population.
coronary artery bypass surgery. In comparison to the 17% to 33% incidence of postoperative AF following coronary bypass surgery,1–7 AF develops after cardiac valvular surgery in 38% to 64% of patients.2,8 The objective of this study was to identify factors that predispose patients undergoing valvular surgery to the development of postoperative AF, so that high-risk patients can be targeted for aggressive prophylactic trials and low-risk patients spared from unnecessary treatment.
METHODS A retrospective analysis of patients undergoing cardiac valvular operations performed at the Cleveland Clinic Foundation between January 1993 and October 1996 was performed. Data were obtained from the Cardiovascular Information Registry (CVIR), a regularly updated, quality-controlled clinical database. Patient population: The study population consisted of patients with the following inclusion criteria: (1) age $18 years; (2) sinus rhythm on preoperative electrocardiogram; and (3) isolated cardiac valvular surgery without concomitant coronary bypass grafting or other surgical procedures. Most operations were performed by 6 cardiac surgeons. The baseline characteristics of the population studied are shown in Table I. From the Departments of Cardiology, Biostatistics and Epidemiology, and Thoracic and Cardiovascular Surgery,The Cleveland Clinic Foundation, Cleveland, Ohio. Manuscript received February 13, 1998; revised manuscript received and accepted May 14, 1998. Address for reprints: Mina K. Chung, MD, Department of Cardiology, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk F-15, Cleveland, Ohio 44195.
892
©1998 by Excerpta Medica, Inc. All rights reserved.
0002-9149/98/$19.00 PII S0002-9149(98)00498-6
TABLE I Baseline Characteristics of Study Population (n 5 915) Age (yrs) Men Unstable preoperative status Prior cardiac surgery Preoperative history Systemic hypertension Diabetes mellitus Rheumatic heart disease Congestive heart failure Left ventricular ejection fraction Normal Mild/mild-moderate/moderate Moderate–severe/severe Left atrial size (mm) Total aortic occlusion time (mins) Total cardiopulmonary bypass time (min)
TABLE II Incidence of Postoperative Atrial Fibrillation (AF) Following Cardiac Surgery
56.1 6 14.7 530 (57.9%) 23 (2.5%) 205 (22.4%) 359 196 128 299
(42.0%) (23.7%) (14.0%) (32.7%)
657 (79.6%) 135 (16.4%) 33 (4.0%) 46.2 6 9.3 65.7 6 30.2 91.6 6 50.7
RESULTS During the study period between January 1993 and November 1995, 3,163 patients underwent valvular surgery. Excluded from the study were 1,457 patients who had concomitant coronary revascularization or other procedures. Of the 1,706 patients in which isolated valvular surgery was performed, 29 patients were excluded due to age ,18 years, 760 patients due to preoperative electrocardiograms showing a rhythm other than sinus rhythm, and 2 patients due to a missing outcome variable. Sinus rhythm was present preoperatively in 915 patients, representing 54% of the population undergoing isolated valvular surgery. Incidence of postoperative atrial fibrillation: The incidence of postoperative AF by valve type and procedure is summarized in Table II. The combined incidence of AF for all patients following isolated valvular operations was 36.7%. For comparison, a cohort of 4,314 patients that underwent isolated coronary revascularization during the same time period as the study population of valvular surgery patients, was selected from the CVIR database. The incidence of postoperative AF for all patients undergoing isolated valvular operations in our study population exceeded the 27.5% incidence for the cohort of coronary revascularization patients (p 5 0.001). This occurred despite the younger age of patients undergoing isolated valve surgery (56 and 62 years of age for the isolated valve and isolated coronary bypass patients, respectively; p ,0.001). Comparisons between the incidences of valvular surgeries were limited by the small number of these operations performed. Univariate analysis: Univariate predictors of postoperative AF after isolated valvular surgery were identified. Variables with a p value ,0.10 are listed in Table III. The incidence of postoperative AF in unstable patients was significantly lower than in stable patients (13.0% vs 37.4%, p 5 0.02). However, the low incidence of postoperative AF among unstable patients may be attributable to the 45% perioperative mortality occurring in this population. Significant valvular or structural factors included mitral stenosis and left atrial enlargement. The only significant intraoperative factor was the use of systemic hypothermia as
Type of Operation AV repair AV replacement MV repair MV replacement MV repair/AV replacement AV/MV replacement AV/PV replacement* MV replacement/TV repair Other repair or replacement† Isolated valve Isolated CABG‡ CABG/valve§
n
%AF
49 333 296 80 40 26 37 18 36 915 4,314 569
24.5 36.9 38.2 35.0 52.5 50.0 18.9 27.8 39.0 36.7#\ 27.5# 41.1\
*Aortic and pulmonic valve replacement 5 Ross procedure. † All other combinations of repair and replacement involving 2 or 3 valves including tricuspid (TV), pulmonic (PV), aortic (AV), and mitral (MV) valves. ‡ CABG 5 isolated coronary bypass grafting performed during the same study period as the isolated valve population. § CABG/valve 5 combined coronary bypass and valve surgery performed during the same study period as the isolated valve population. # p 5 0.001, isolated valve surgery versus CABG. \ p 5 0.09, isolated valve surgery versus CABG/valve.
part of the myocardial protection delivery method, which was associated with a lower incidence of postoperative AF. Multivariate analysis: Independent risk factors associated with postoperative AF following isolated valvular surgeries were identified (Table IV). These 5 independent variables were further assessed individually. There was an increasing incidence of postoperative AF with advancing age. The incidence of AF was 14.0% in patients ,35 years of age, but over 3-fold higher at 50.5% in patients .75 years of age. The presence of left atrial enlargement as determined by echocardiographic or surgical observation was an independent risk factor for postoperative AF. The severity of left atrial enlargement was correlated with increased risk of postoperative AF. Transthoracic or transesophageal echocardiograms within 6 months of surgery were available for 73% of patients. The incidence of AF in patients with normal left atrial dimensions (#40 mm), was significantly less than for those with abnormal dimensions (.40 mm; 24.8% vs 37.4%, p ,0.005). Among patients with severe left atrial enlargement (dimensions .60 mm ), the incidence of postoperative AF increased to 44.4%. Mitral stenosis was a significant risk factor for postoperative AF independent of left atrial enlargement. Of the 71 patients with echocardiographic mitral stenosis who underwent mitral valve surgery, 36 (50.7%) developed postoperative AF compared with 35.6% of patients with all other valvular disorders (p ,0.01). Reoperation and the use of systemic hypothermia were found to be associated with a reduced incidence of postoperative AF. At the time of a first operative procedure, 61.1% of patients were in sinus rhythm. However, sinus rhythm was present in 43.5% at the
VALVULAR HEART DISEASE/POSTOPERATIVE AF AND VALVULAR SURGERY
893
TABLE III Univariate Predictors of Postoperative Atrial Fibrillation (AF) After Valvular Surgery
DISCUSSION
The current study reports the largest analysis to date of AF occurring Variable Present (%AF) Absent (%AF) p Value after valvular surgery. Postoperative AF occurred in 36.7% of patients, Reoperation 59/205 (28.8) 277/710 (39.0) ,0.01 Systemic hypothermia 141/452 (31.2) 195/463 (42.1) ,0.01 exceeding rates reported at the same History of thyroid disease 35/76 (46.1) 301/839 (35.9) 0.08 institution after coronary bypass Stable preoperative status 333/891 (37.4) 3/23 (13.0) 0.02 grafting. Based on these estimates, Mitral stenosis 36/71 (50.7) 298/838 (35.6) 0.01 approximately one-third of the Left atrial enlargement 193/468 (41.2) 143/447 (32.0) ,0.01 Age .65 yrs 146/301 (48.5) 190/614 (30.9) ,0.01 70,000 patients that undergo valvular surgery annually will develop new postoperative AF.9 Because the pathogenesis and risk factors for AF TABLE IV Multivariate Predictors of Postoperative Atrial after valvular heart surgery are not well understood, Fibrillation Following Valvular Surgery development of preventive therapies have been limVariable OR (95% CI) p Value ited. Mitral stenosis 2.066 (1.21–3.52) 0.0077 Incidence: Despite the generally younger age of Age (per decade) 1.506 (1.35–1.68) 0.0001 patients undergoing isolated valvular surgery comLeft atrial enlargement 1.468 (1.07–2.01) 0.0165 pared with coronary bypass grafting, the incidences of Systemic hypothermia 0.572 (0.422–0.776) 0.0003 postoperative AF are greater in valve patients. The Reoperation 0.676 (0.465–0.981) 0.0393 greater inherent susceptibility to postoperative AF after valvular heart surgery may result from more prevalent structural and hemodynamic abnormalities, such time of a second operation and only 34.7% of patients as left atrial enlargement and pathologic changes in at the time of $3 operations. Patients in sinus rhythm the atria from aging and rheumatic mitral stenosis. before a reoperation had a lower incidence of postopPredictors of postoperative atrial fibrillation: In our erative AF than patients undergoing a first operation, analysis of 915 patients undergoing isolated cardiac respectively (59 of 205, 28.8% vs 277 of 710, 39.0%, valvular surgery, 5 pre- and intraoperative factors were identified that predicted the occurrence of postp 5 0.007). The use of systemic hypothermia was associated operative AF. The review by Creswell et al,2 of 297 with a lower incidence of postoperative AF compared patients undergoing isolated valvular surgery, found with no hypothermia (42.1% vs 31.2%, respectively, p age to be the only independent predictor of postoper,0.01). The degree of systemic hypothermia tended ative AF. A smaller analysis8 of 135 patients underto correlate with the lowest achieved temperature, going aortic valve replacement with coronary bypass with temperatures of 31° to 35°, 25° to 30°, and ,25° grafting discovered several correlations of hemodyassociated with incidences of postoperative AF of namic and clinical variables with postoperative AF. AGE: A consistent finding among studies of AF 32.1%, 30.9%, and 27.9%, respectively. occurring following coronary artery surgery is an inValidation of independent variables: Three of the 5 creasing incidence with advancing age.1– 4,6,7,10 Explaindependent risk factors were confirmed to be signifnations for the impact of age on this arrhythmia are icant factors among the validation cohort of 305 paspeculative and include structural atrophic or ischemic tients. Although only 16 patients had mitral stenosis in effects within the atria and related electrophysiologic the validation cohort, the incidence of postoperative changes. These age-related changes in the atria likely AF was markedly higher in these patients compared occur in patients undergoing either valvular surgery or with patients with other valvular disorders (69% vs. coronary bypass grafting. 27%, p ,0.001). An enlarged left atrium was again In patients undergoing isolated valvular surgery, predictive of higher postoperative AF rates (36% vs this factor has additional significance. Thirty-four per19%, p 5 0.002). Advanced age continued to be cent of patients were #50 years old in our study and strongly associated with postoperative AF, with rates had a low postoperative AF incidence of 19.4%. These of 12%, 21%, 37%, 37%, 46%, and 67% for patients younger patients represent an easily identifiable low30 to 39, 40 to 49, 50 to 59, 60 to 69, 70 to 79, and risk group for postoperative AF that may not require $80 years old, respectively. prophylactic treatment. The 2 variables that were protective factors in the LEFT ATRIAL ENLARGEMENT: Two prior studies of initial group were not confirmed in the validation coronary bypass surgery patients have considered the cohort. Systemic hypothermia was not associated with effect of left atrial enlargement as a predisposing a lower incidence of postoperative AF in the valida- factor for supraventricular tachyarrhythmias.6,11 Only tion cohort (32% with hypothermia vs 29% without the study of Dixon et al11 found that supraventricular hypothermia, p 5 0.701). Although lower event rates tachyarrhythmias in the postoperative period occurred were again observed in patients undergoing reopera- more frequently for patients with left atrial enlargetion versus first operations (26% vs 31% respectively, ment (RR 5 1.74). p 5 0.378), this difference did not achieve statistical Although it is apparent that postoperative AF is significance. more common with any left atrial enlargement, the 894 THE AMERICAN JOURNAL OF CARDIOLOGYT
VOL. 82
OCTOBER 1, 1998
degree of dilation also adds to this risk. However, even among the patients with severe left atrial enlargement, the incidence of postoperative AF was not prohibitively high (44%) to forego consideration of the benefits of prophylactic therapy to prevent this arrhythmia. MITRAL STENOSIS: The importance of mitral stenosis as a predictor of postoperative AF, independent of left atrial enlargement, further demonstrates the potential multifactorial etiology of this arrhythmia in patients with mitral stenosis. The predominant cause of mitral stenosis in adults is rheumatic heart disease.12 This high incidence of arrhythmias is believed to be due to both the hemodynamic effects of elevated left atrial pressure and the structural changes that include myofibril hypertrophy and fibrosis of the atrial wall.13 REOPERATIONS: The lower incidence of postoperative AF following reoperations likely represented a natural selection bias. Increasing age and the progression of hemodynamic and structural effects on the left atrium may enhance the propensity to develop chronic AF over time, and patients with chronic AF were excluded from this study. Therefore, the patients undergoing reoperation in sinus rhythm represent a group that has already shown some resistance to chronic AF. SYSTEMIC HYPOTHERMIA: The importance of intraoperative atrial temperatures and mechanical activity has been widely studied. Data from experimental and small clinical studies have demonstrated an effect of intraoperative atrial and ventricular temperatures on supraventricular conduction.14,15 Although conflicting data exist, an association between intraoperative atrial activity detected by atrial electrograms and postoperative atrial arrhythmias has also been suggested.16 –18 The use of systemic hypothermia during isolated valvular surgery was associated with a reduced incidence of postoperative AF in our initial study population. However, this relation between the use of systemic hypothermia and postoperative AF was not present in the validation cohort. Thus, similar to other studies, the benefit of augmented hypothermia remains unproven.19 Conclusion: Postoperative AF is common following all types of isolated valvular surgery. In our study cohort, 5 independent variables predicted patients risk for postoperative AF. Three of the 5 independent variables were validated in a separate cohort and included advanced age, mitral stenosis, and left atrial
enlargement. Although effective prophylactic methods have not been well studied in patients undergoing valvular surgery, patients with these clinical characteristics are among those that may likely benefit from prophylactic efforts. 1. Leitch JW, Thomson D, Baird DK, Harris PJ. The importance of age as a predictor of atrial fibrillation and flutter after coronary artery bypass grafting. J Thorac Cardiovasc Surg 1990;100:338 –342. 2. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 1993;56:539 –549. 3. Crosby LH, Pifalo WB, Woll KR, Burkholder JA. Risk factors for atrial fibrillation after coronary artery bypass grafting. Am J Cardiol 1990;66:1520 – 1522. 4. Fuller JA, Adams GG, Buxton B. Atrial fibrillation after coronary artery bypass grafting: is it a disorder of the elderly. J Thorac Cardiovasc Surg 1989;97:821– 825. 5. Vecht RJ, Nicolaides EP, Ikweuke JK, Liassides C, Cleary J, Cooper WB. Incidence and prevention of supraventricular tachyarrhythmias after coronary bypass surgery. Int J Cardiol 1986;13:125–134. 6. Mathew JP, Parks R, Savino JS, Friedman AS, Koch C, Mangano DT, Browner WS, for the MultiCenter Study of Perioperative Ischemia Research Group. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA 1996;276:300 –306. 7. Aranki SF, Shaw DP, Adams DH, Rizzo RJ, Couper GS, VanderVliet M, Collins JJ, Cohn LH, Burstin HR. Predictors of atrial fibrillation after coronary artery surgery: current trends and impact on hospital resources. Circulation 1996;94:390 –397. 8. Douglas P, Hirshfeld JW, Edmunds LH. Clinical correlates of postoperative atrial fibrillation (abstr.). Circulation 1984;70 (Suppl II):II-165. 9. American Heart Association, Heart and Stroke Facts, Dallas, TX, 1993. 10. Frost L, Molgaard H, Christiansen EH, Hjortholm K, Paulsen PK, Thomsen PB. Atrial fibrillation and flutter after coronary artery bypass surgery: epidemiology, risk factors and preventive trials. Int J Cardiol 1992;36:253–261. 11. Dixon FE, Genton E, Vacek JL, Moore CB, Landry J. Factors predisposing to supraventricular tachyarrhythmias after coronary bypass grafting. Am J Cardiol 1986;58:476 – 478. 12. Gaasch WH, O’Rourke RA, Cohn LH, Rackley CE. Mitral valve disease. In: JW Hurst, ed. The Heart. 8th Edition. New York: McGraw-Hill; 1994:chapter 80. 13. Bailey GW, Braniff BA., Hancock EW, Cohn KE. Relation of left atrial pathology to atrial fibrillation in mitral valvular disease. Ann Intern Med 1968; 69:13–20. 14. Sato S, Yamauchi S, Schuessler RB, Boineau JP, Matsunaga Y, Cox JL. The effect of augmented atrial hypothermia on atrial refractory period, conduction and atrial flutter/ fibrillation in the canine heart. J Thorac Cardiovasc Surg 1992;104: 297–306. 15. Smith PK, Buhrman WC, Levett JM, Ferguson TB, Holman WL, Cox JL. Supraventricular conduction abnormalities following cardiac operations. A complication of inadequate atrial preservation. J Thorac Cardiovasc Surg 1983;85: 105–115. 16. Tchervenkov CI, Wynands JE, Symes JF, Malcolm ID, Dobell AR, Morin JE. Persistent atrial activity during cardioplegic arrest: a possible factor in the etiology of postoperative supraventricular tachyarrhythmias. Ann Thorac Surg 1983;36:437– 443. 17. Mullen JC, Khan N, Weisel RD, Christakis GT, Teoh KH, Madonik MM, Mickle DA, Ivanov J. Atrial activity during cardioplegia and postoperative arrhythmias. J Thorac Cardiovasc Surg 1987;94:558 –565. 18. Dewar ML, Rosengarten MD, Blundell PE, Chiu RC. Perioperative holter monitoring and computer analysis of dysarrhythmias in cardiac surgery. Chest 1985;87:593–597. 19. Cheung EH, Arcidi JM, Jackson ER, Hatcher CR, Guyton RA. Intracavitary right heart cooling during coronary bypass surgery: a prospective randomized trial. Circulation 1988;78(suppl III):III-173–III-179.
VALVULAR HEART DISEASE/POSTOPERATIVE AF AND VALVULAR SURGERY
895