- Email: [email protected]
Rate Control vs Rhythm Control in Patients With Nonvalvular Persistent Atrial Fibrillation
Rate Control vs Rhythm Control in Patients With Nonvalvular Persistent Atrial Fibrillation* The Results of the Polish How to Treat Chronic Atrial Fibrillation (HOT CAFE) Study Grzegorz Opolski, MD, PhD; Adam Torbicki, MD, PhD; Dariusz A. Kosior, MD, PhD; Marcin Szulc, MD, PhD; Beata Woz˙akowska-Kapłon, MD, PhD; Piotr Kołodziej, MD, PhD; and Piotr Achremczyk, MD, PhD; for the Investigators of the Polish HOT CAFE Trial
Study objectives: The relative risks and benefits of strategies of rate control vs rhythm control in patients with atrial fibrillation (AF) remain to be fully explored. Design: The How to Treat Chronic Atrial Fibrillation (HOT CAFE) Polish trial was designed to evaluate in a randomized, multicenter, and prospective manner the feasibility and long-term outcomes of rate control vs rhythm control strategies in patients with persistent AF. Patients: Our study population comprised 205 patients (134 men and 71 women; mean [ⴞ SD] age, 60.8 ⴞ 11.2 years) with a mean AF duration of 273.7 ⴞ 112.4 days. The mean observation period was 1.7 ⴞ 0.4 years. One hundred one patients were randomly assigned to the rate control group and received rate-slowing therapy guided by repeated 24-h Holter monitoring. Direct current cardioversion and atrioventricular junctional ablation with pacemaker placement were alternative nonpharmacologic strategies for patients with tachycardia that was resistant to medical therapy. One hundred four patients were randomized to sinus rhythm restoration and maintenance using serial cardioversion supported by a predefined stepwise antiarrhythmic drug regimen (ie, disopyramide, propafenone, sotalol, and amiodarone). In both groups, thromboembolic prophylaxis followed current guidelines. Measurements and results: At the end of follow-up, 63.5% of patients in the rhythm control arm remained in sinus rhythm. No significant differences in the composite end point (ie, all-cause mortality, number of thromboembolic events, or major bleeding) were found between the rate control group and the rhythm control group (odds ratio, 1.98; 95% confidence interval, 0.28 to 22.3; p > 0.71). The incidence of hospital admissions was much lower in the rate control arm (12% vs 74%, respectively; p < 0.001). New York Heart Association functional class improved in both study groups, while mean exercise tolerance, as measured by the maximal treadmill workload, improved only in the rhythm control group (5.2 ⴞ 5.1 vs 7.6 ⴞ 3.3 metabolic equivalents, respectively; p < 0.001). The rhythm control strategy led to an increased mean left ventricular fractional shortening (29 ⴞ 7% vs 31 ⴞ 7%, respectively; p < 0.01). One episode of pulmonary embolism occurred in the rate control group despite oral anticoagulation therapy, while three patients in the rhythm control arm of the study experienced ischemic strokes (not significant). Conclusions: The Polish HOT CAFE study revealed no significant differences in major end points between the rate control group and the rhythm control group. (CHEST 2004; 126:476 – 486) Key words: antiarrhythmic therapy; atrial fibrillation; electrocardioversion rate control; rhythm control Abbreviations: AF ⫽ atrial fibrillation; AFFIRM ⫽ Atrial Fibrillation Follow-up Investigation of Rhythm Management; HOT CAFE ⫽ How to Treat Chronic Atrial Fibrillation; INR ⫽ international normalized ratio; NYHA ⫽ New York Heart Association; PIAF ⫽ Pharmacological Intervention in Atrial Fibrillation; RACE ⫽ Rate Control vs Electrical Cardioversion for Persistent Atrial Fibrillation
476
Clinical Investigations
goal in patients with atrial fibrilT helationtherapeutic (AF) can be either the restoring and maintaining of sinus rhythm (mainly with antiarrhythmic therapy) or allowing persistent AF to become permanent, while controlling the ventricular rate and preventing systemic thromboembolism.1,2 The debate over rate control vs rhythm control in the treatment of AF still is ongoing. It has been a standard practice to maintain sinus rhythm, primarily through the use of antiarrhythmic drugs and cardioversion, which relieves symptoms and may retard or prevent progression to permanent AF. Rhythm control, however, carries a risk of proarrhythmias and negative inotropic effects due to the use of antiarrhythmic agents.3–5 The prevention of AF recurrence with antiarrhythmic drug therapy, regardless of the agent, is achieved in only approximately 50% of patients at 6 months to 1 year.6,7 Moreover, optimal thromboembolic prevention in the rhythm control strategy has not yet been defined. While rate control is easier to achieve and leads toward improved hemodynamics, it allows AF to persist and may not eliminate AF symptoms (eg, palpitations, dyspnea, chest pain, fatigue, and dizziness).8 The How to Treat Chronic Atrial Fibrillation (HOT CAFE) study was designed to compare the strategy of sinus rhythm restoration and maintenance with that of ventricular rate control and chronic thromboembolic prophylaxis. We report the results of the initial phase of the study, which involved six clinical centers and a limited number of patients. This phase was implemented to confirm the study protocol and its methods before the planned main phase of the study, which was ultimately abandoned due to the results of other trials.9 –11
*From the Departments of Cardiology (Drs. Opolski and Kosior) and Internal Medicine and Hypertension (Dr. Szulc), Medical University of Warsaw, Warsaw, Poland; the Department of Chest Medicine (Dr. Torbicki), National Institute of Tuberculosis and Pulmonary Disease, Warsaw, Poland; the Department of Cardiology (Dr. Woz˙akowska-Kapłon), Municipal Hospital, Kielce, Poland; the Department of Cardiology (Dr. Kolodziej), Municipal Hospital, Siedlce, Poland; and the Department of Cardiology (Dr. Achremczyk), Municipal Hospital, Radom, Poland. The study results were presented at the 24th Congress of the European Society of Cardiology, Berlin, Germany, in October 2002. The study was supported by Polish Government research grant No. PO5B06012. Manuscript received August 8, 2003; revision accepted February 11, 2004. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: [email protected]). Correspondence to: Grzegorz Opolski, MD, PhD, Chair, Department of Cardiology, Medical University of Warsaw, Warsaw, Poland; e-mail: [email protected] www.chestjournal.org
Materials and Methods The HOT CAFE study12 was a prospective, randomized, open multicenter clinical trial that was designed to evaluate the effects of rhythm control vs rate control in patients with persistent AF. The study involved six cardiology centers in central Poland, five of which enrolled patients. The study was approved by the institutional review board or ethics committee at those centers, and consecutive, eligible patients provided written informed consent. Recruitment began in March 1997, randomization was concluded in December 2000, and follow-up was terminated in December 2002. Inclusion and Exclusion Criteria Eligible patients had to meet the following inclusion criteria: 50 to 75 years of age; and AF had to be known to be present for at least 7 days, but not for ⬎ 2 years. Only patients with a first clinically overt persistent episode of AF were enrolled. Persistent AF was defined as a non-self-terminating arrhythmia requiring electrical cardioversion to restore sinus rhythm. The following were exclusion criteria: documented inefficiency; intolerance to or contraindications for treatment with antiarrhythmic drugs; presence of arrhythmia associated with an acute reversible condition; thyroid dysfunction; pregnancy or lactation; history of myocardial infarction within 3 months preceding enrollment into the study; acute myocarditis; cardiac surgery during the previous 30 days; severe cardiac disability (ie, New York Heart Association [NYHA] functional class IV); severe systemic hypertension not responding to treatment (diastolic pressure, ⬎ 115 mm Hg); hypotension (systolic pressure, ⬍ 90 mm Hg); history of transient ischemic attack (with significantly marked vascular pathology requiring surgical intervention); history of hemorrhagic stroke; ischemic stroke during the 3 months preceding entrance into the trial; any mitral stenosis or other valvular disease suitable for surgical treatment; R-R intervals exceeding 3 s; ventricular response to AF of ⬍ 90 beats/min (unrelated to drugs used to reduce ventricular rate); bundle branch block or QT-segment prolongation (ie, corrected QT interval of ⬎ 480 ms or uncorrected QT interval of ⬎ 500 ms); alcoholism; contraindications to anticoagulation therapy; liver, kidney, or CNS damage; advanced chronic lung disease, malignancy; or any noncardiac illness associated with a life expectancy of ⬍ 1 year; or participation in another study. In addition, premenopausal women who were still capable of procreation and had not undergone tubal ligation or hysterectomy also were excluded from the study. Randomization, Therapy, and Follow-up Participating patients were randomly assigned to one of the two groups following a predefined sequential management (Fig 1). The randomization of eligible patients was performed by permuted block design with equal allocation and was stratified centrally by a steering center of the Chair and Department of Cardiology. The trial was conducted on an intention-to-treat basis. All patients were seen in the outpatient department at least once a month during the first 6 months of follow-up and every 3 months until the end of the study. All patients enrolled in the trial were potentially eligible for either rate control or rhythm control therapy at entrance into the study. Rate Control Group: The therapeutic strategy in the rate control group was rate control and chronic thromboembolic prophylaxis, with heart rate as the therapeutic target. Heart rate was considered to be adequately controlled when the ventricular response ranged between 70 and 90 beats/min during rest CHEST / 126 / 2 / AUGUST, 2004
477
Figure 1. The HOT CAFE trial design. SR ⫽ sinus rhythm; DC ⫽ direct current; * ⫽ disopyramide, propafenone, or sotalol.
(12-lead resting ECG) and not exceeding 140 beats/min during moderate exercise. The circadian pattern and mean rate over a 24-h period were monitored by ECG Holter recordings performed at study entry, and were repeated 1 and 3 months after randomization, and every 6 months thereafter. Exercise testing was used to evaluate heart rate during submaximal exercise at the beginning of observation, at 2, 12, and 24 months after randomization, and at the end of the study. The pharmacologic therapies approved for use in this study arm included the following: -blockers; nondihydropyridine calcium blockers; digoxin; or a combination of these drugs. Cardioversion and atrioventricular junctional ablation with pacemaker placement were alternative nonpharmacologic strategies in patients with resistant tachycardia. Anticoagulation therapy was considered in all patients who were assigned to the rate control group. Following American College of Chest Physicians guidelines,13 oral anticoagulation therapy with a target international normalized ratio (INR) of 2.5 (range, 2.0 to 3.0) was recommended for patients at high risk of stroke. Systemic embolization risk factors included prior stroke or transient ischemic attack, left ventricular dysfunction, congestive heart failure, diabetes mellitus, history of hypertension, and coronary artery disease. Patients between 65 and 75 years of age who had no risk factors were offered therapy with 325 mg aspirin daily or, in patients with proven aspirin intolerance, ticlopidine, 500 mg per day. Patients ⬍ 55 years of age without clinical or echocardiographic evidence of cardiovascular disease, diabetes, or a history of stroke or systemic embolism could decline pharmacologic thromboembolic prophylaxis.13 Rhythm Control Group: Patients randomly assigned to the rhythm control group underwent electrical cardioversion without prior treatment with antiarrhythmic drugs. Cardioversion was 478
considered to be successful if sinus rhythm was maintained for 2 h after the procedure, at which point the first dose of an antiarrhythmic drug was administered. Following sinus rhythm restoration, patients received antiarrhythmic drug therapy. Due to potential proarrhythmic effects, the choice of drug was open, based on arrhythmia etiology, concomitant heart disease, and the patient’s age.14 As a first step the patients received propafenone (450 to 600 mg per day), disopyramide (300 to 600 mg per day), or sotalol (160 to 320 mg per day). Patients with ischemic heart disease usually received sotalol or disopyramide. Class I drugs were avoided in patients with left ventricular dysfunction on echocardiographic examination or signs of congestive heart failure. Propafenone and sotalol were the drugs of first choice in patients with systemic hypertension. -blockers were allowed in addition to disopyramide or propafenone when considered clinically indicated. A recurrence of AF, which was defined as AF lasting at least 12 h, mandated a repeat cardioversion. Once the sinus rhythm was restored, another predefined antiarrhythmic drug was administered in the sequence of the above-mentioned first-line drug. When a recurrence occurred after 6 months of antiarrhythmic therapy, cardioversion was attempted. If successful, the same antiarrhythmic regimen was continued. In situations in which there was no response to such a pharmacologic treatment and AF recurred, an oral daily dose of amiodarone (600 mg) was administered over 3 weeks, followed by 400-mg doses until the loading dose of at least 12.0 to 16.0 g was reached prior to another cardioversion. Where successful, the restored sinus rhythm was maintained through daily maintenance doses of 100 to 200 mg amiodarone. However, in patients in whom there was no response to the initial direct current cardioversion, they received amiodarone, as Clinical Investigations
above, with cardioversion attempted again. Patients in whom sinus rhythm was restored continued their amiodarone treatment in daily maintenance doses. The amiodarone therapy started at the hospital was continued on an outpatient basis with subjects being assessed at least weekly. In patients in whom there was no response to the first and second cardioversions, with the second cardioversion preceded by amiodarone loading, or there was a recurrence of AF without an identifiable cause while receiving amiodarone, the arrhythmia was considered permanent, and rate control became the goal of therapy. Only electrocardiographically confirmed episodes lasting ⬎ 12 h despite additional antiarrhythmic drug therapy were considered to be clinically significant. Patients experiencing such episodes received anticoagulation prophylaxis, with ventricular rhythm control as additional therapy. Patients who required pacing for rate support and/or to maintain normal sinus rhythm remained in the study. Current guidelines15 were followed when pacing was used. In the rhythm control arm of the study, thromboembolic prophylaxis (with acenocoumarol) titrated to achieve an INR of 2.0 to 3.0 was started 4 weeks before cardioversion.13 Therapy was continued in all patients until normal sinus rhythm had been maintained for at least 4 weeks. Decisions regarding prolonged or chronic anticoagulation or antiplatelet therapy were left to the discretion of the local clinical teams but had to be reported.13 Assessment and Follow-up A physician and a nurse coordinator assessed patients initially and then at least monthly for 6 months, and every 3 months thereafter. Twelve-lead ECGs were obtained at each visit. Chest radiograph examinations were obtained at the first follow-up appointment and were repeated annually thereafter. The level of serum thyroid-stimulating hormone was measured every 6 months in patients receiving amiodarone. All study patients had their exercise tolerance estimated by a stress test that was performed at study entry, 2 months after randomization, and then annually. In addition, atrial and ventricular morphology and function were echocardiographically assessed during the followup period. The primary end point was a composite of death from any cause, thromboembolic complications (especially disabling ischemic stroke), and intracranial or other major hemorrhage. Disabling stroke was defined as an acute neurologic deficit for ⬎ 24 h that affected the ability to perform activities of daily life or resulted in death. Cerebrovascular events required diagnosis by CT scan and neurology consultation. Major bleeding was defined as a drop in hemoglobin levels of ⬎ 2 g/L, requiring a blood transfusion, needing surgical intervention, or resulting in death. All events occurring between randomization and the end of the study were recorded. All deaths were considered to be due to cardiovascular causes unless an unequivocal noncardiac cause could be identified. Secondary end points included rate control, sinus rhythm maintenance, discontinuation of therapy (especially if due to a potential proarrhythmic effect), hemorrhage, hospitalization, new or worsening congestive heart failure, or changes in exercise tolerance.
measurements were averaged for more than five cardiac cycles recorded during relaxed expiration. Echocardiography was performed at study entry, 2 months after randomization, and at the end of the study. In patients assigned to rhythm control, echocardiographic examination was performed prior to and 30 days after successful cardioversion, and at the end of the observation period. Exercise Testing On entry into the study, all patients underwent a treadmill exercise stress test, which was performed according to the symptom-related modified Bruce I protocol. This was repeated at 2, 12, and 24 months after randomization, and at the end of the study. In addition, in the rhythm control arm, this testing was performed prior to and 30 days after successful cardioversion. Exercise was continued until the patient was exhausted or developed progressive angina, hypotension, or ventricular arrhythmia. Age-predicted maximal heart rate, shifts in ST segments, the emergence of ventricular ectopic beats, or the development of atrial arrhythmia did not constitute end points of the treadmill test. Heart rate during AF was determined by measuring the average heart rate for a period of 60 s. Maximal workload and exercise duration also were considered. An exercise stress test was repeated as needed in the rate control group to achieve optimal heart rate control. 24-h ECG Holter Monitoring The circadian pattern and the mean 24-h heart rate were monitored with an ECG Holter recording performed at the study entry, at 1 and 3 months after randomization, and every 6 months thereafter. In addition, 24-h ECG monitoring was performed in patients assigned to the rhythm control strategy 7 and 30 days after successful cardioversion or after every antiarrhythmic drug change. Statistical Analysis Summary data are expressed as the mean ⫾ SD or as the numbers and percentages of patients. Analyses were performed using the intention-to-treat principle. Continuous variables were tested by analysis of variance for between-group comparisons (with the baseline value as the covariable). Dependent on analysis of variance results, the Student t test (in modification for paired measurements if necessary) was used to compare mean parameter values in each group on one visit or in one group on each visit. The Bonferroni correction was used to avoid multipletesting problems. The change in time of NYHA functional class was tested using the Wilcoxon test, and between-group comparisons were tested using the Kruskal-Wallis test on each visit. All pair comparisons, if needed, were made by the Dunn test. The cumulative risk of AF recurrence was estimated by the KaplanMeier product-limit method. Data were censored if the patient died, reached the end of the follow-up period, or was lost to follow-up without the occurrence of a primary end point. The difference between treatment groups was assessed with the log-rank test with a p value of ⬍ 0.05 considered to be statistically significant.
Echocardiography M-mode echocardiographic measurements including left atrial diameter and left ventricular diameter, were made at enddiastole in the parasternal long-axis view using standard techniques.16 Left ventricular fractional shortening was determined on two-dimensional imaging. The left atrial long axis was measured at end-diastole from the apical four-chamber view. All www.chestjournal.org
Results Baseline Characteristics Of 738 screened patients with AF from March 1997 to December 2000, 205 patients (134 men and CHEST / 126 / 2 / AUGUST, 2004
479
71 women) met the inclusion criteria and were enrolled into the study. Valvular heart disease (21%), thyrotoxicosis (13%), and advanced stages of concomitant diseases influencing the natural history of AF (11%) were the main exclusion criteria limiting enrollment. One hundred one consecutive patients were randomly assigned to the heart rate control group. The restoration and maintenance of sinus rhythm for a period of at least 12 months was attempted in 104 patients. Patients were followed up for a mean of 1.7 ⫾ 0.4 years (maximum follow-up, 2.5 years). There were no significant differences in the duration of follow-up between the two study groups. The mean age was 60.8 ⫾ 11.2 years. The mean arrhythmia duration prior to enrollment was 273.7 ⫾ 112.4 days. The characteristics of the study patients were typical of a population of subjects affected by persistent AF. There were no significant
differences in baseline characteristics between the rate control group and the rhythm control group. A comparison of the clinical and echocardiographic characteristics of study patients on enrollment is shown in Table 1. Primary End Point None of the patients was lost to follow-up. Allcause mortality, and the number of thromboembolic and major bleeding complications, the primary composite end point of the study, were not significantly different between the rate control group and the rhythm control group (odds ratio, 1.98; 95% confidence interval, 0.28 to 22.3; p ⬎ 0.71). One patient (1.0%) assigned to the rate control group died due to leukemia diagnosed 6 months after enrollment. Three patients (2.9%) died in the rhythm control
Table 1—Baseline Characteristics of the Study Groups* Parameter Age, yr Gender Female Male AF duration 7 d–1 mo 1 mo–2 yr Mean AF duration, d AF etiology Ischemic heart disease Myocardial infarction CABG Hypertension Valvular heart disease Lone AF Diabetes Congestive heart failure, NYHA scale I II III Echocardiographic parameters LAsax, mm LAlax, mm LVEDD, mm FS, % Pharmacologic therapy, % -adrenolitics Verapamil/diltiazem Digoxin -adrenolitics ⫹ digoxin ACE-I Amlodipine Nitrates Diuretics
Rate Control (n ⫽ 101)
Rhythm Control (n ⫽ 104)
Significance
61.4 ⫾ 17.6
60.4 ⫾ 7.9
NS
38 (37.6) 63 (62.4)
33 (31.7) 71 (68.3)
NS NS
17 (16.8) 84 (83.2) 243.2 ⫾ 137.3
16 (15.4) 88 (84.6) 220.4 ⫾ 148.6
NS NS NS
38 (37.6) 7 (6.9) 0 (0.0) 60 (59.4) 15 (14.8) 25 (24.8) 18 (17.8)
52 (50.0) 7 (6.7) 1 (1.0) 72 (69.2) 16 (15.4) 18 (17.3) 15 (14.4)
NS NS NS NS NS NS NS
48 (47.5) 48 (47.5) 5 (5.0)
30 (28.8) 59 (56.7) 15 (14.4)
NS NS NS
48.0 ⫾ 4.7 64.0 ⫾ 6.8 50.8 ⫾ 5.9 32.8 ⫾ 6.6
47.4 ⫾ 5.3 62.5 ⫾ 7.9 52.2 ⫾ 6.8 29.9 ⫾ 6.9
NS NS NS NS
49.5 7.9 3.0 39.6 71.2 12.1 9.1 16.1
52.1 7.7 6.5 31.7 69.2 14.4 8.7 14.4
NS NS NS NS NS NS NS NS
*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. NS ⫽ not significant; CABG ⫽ coronary artery bypass grafting; ACE ⫽ angiotensin-converting enzyme; LAsax ⫽ left atrial anteroposterior axis; LAlax ⫽ left atrial longitudinal axis; LVEDD ⫽ left ventricular end diastolic diameter; FS ⫽ left ventricular fractional shortening. 480
Clinical Investigations
group, two (1.9%) from strokes. One female patient (1.0%) from the rhythm control arm died in a motor vehicle accident in which she was a passenger, in the 10th month of observation. Three patients (2.9%) assigned to the rhythm control group developed massive cerebral strokes. Two of the strokes (1.9%), both resulting in death, were observed on the third day following successful cardioversion despite appropriate anticoagulant therapy (INR, 2.8 ⫾ 0.7) during a sufficient period of time (35.3 ⫾ 11.2 days) prior to the procedure. In both cases, a cardiogenic etiology was supported by the exclusion of all other possible causes. The third stroke (0.9%), which was not disabling, affected a patient with AF recurrence during amiodarone therapy. The patient was receiving long-term aspirin therapy. Pulmonary embolism occurred despite oral anticoagulation in one patient (1.0%) in the rate control group. One patient from the rhythm control arm withdrew his consent after 6 months of observation. Secondary End Points Rate Control and Maintenance of Sinus Rhythm: The treatment goal in the rate control group was an optimal heart rate, as defined by 24-h Holter monitoring. Calcium antagonists (7.9%), -blockers (49.5%), -blockers plus digoxin (39.6%), or digoxin alone (3.0%) were used. Two patients (2.0%) in this study arm developed symptomatic, sustained tachycardia despite receiving pharmacologic therapy, and ultimately underwent atrioventricular node ablation with pacemaker implantation. One patient (1.0%) in the rate control arm required a pacemaker due to symptomatic bradycardia and pauses of ⬎ 3.5 s. Planned electrical cardioversion was performed in all 104 patients assigned to receive rhythm control. Following the first electrical cardioversion, sinus rhythm was restored in 56 patients (53.8%). Fortyeight patients failed to convert and received loading doses of amiodarone prior to the next cardioversion attempt. Sinus rhythm recovery occurred in 10 patients (20.8%) during amiodarone loading. The second electrical cardioversion, preceded by amiodarone loading, restored sinus rhythm in 24 of 38 patients (63.2%). The first and second cardioversions, the latter preceded by amiodarone loading, enabled sinus rhythm restoration in 90 patients (86.5%). Sixty-six patients (63.5%) in the rhythm control arm were in sinus rhythm at the end of the study. In 27 patients, sinus rhythm was successfully maintained with the first antiarrhythmic compound following the initial effective cardioversion. Sequential electrical cardioversion and antiarrhythmic treatment resulted in sinus rhythm maintenance in an www.chestjournal.org
additional 15 patients who previously had been subjected to a successful initial procedure. Amiodarone administered in loading doses prior to the second cardioversion and maintenance therapy following the successful procedure prevented AF recurrence in the remaining 24 subjects whose initial cardioversion had failed. Altogether, 37 patients (56.0%) received amiodarone as AF prophylaxis following sinus rhythm restoration. Two patients (1.9%) in the rhythm control arm required pacemaker placement due to bradycardia. In both patients, antiarrhythmic therapy was continued. There was a continuous decline in heart rate during the follow-up period in patients who were assigned to receive rhythm control. The mean heart rate measured 30 days after electrical cardioversion was significantly lower than that at randomization (86.0 ⫾ 11.6 vs 77.8 ⫾ 12.6 beats/min, respectively; p ⬍ 0.001). There were no other significant differences in measured parameters during further observation. However, the mean heart rate during subsequent visits was significantly lower than that at entry in the sinus rhythm group. The mean heart rate in 24-h Holter monitoring was significantly lower during rhythm control (79.1 ⫾ 8.6 beats/min) than during rate control (85.8 ⫾ 7.5 beats/min; p ⬍ 0.003). There were no significant changes in mean heart rate in the rate control arm of the study during the follow-up. Rate control therapy maintained the ventricular response rate within satisfactory ranges during the follow-up (at randomization 86.2 ⫾ 7.8 beats/ min; at the end of follow-up, 83.1 ⫾ 4.3 beats/min; p ⬎ 0.093). Discontinuation of Therapy: Two of 25 patients who had received sotalol in the sinus rhythm group required drug withdrawal due to abdominal pain, diarrhea, or eczema, all of which disappeared on drug termination. One patient experienced torsade de pointes during a routine 24-h Holter monitoring session, and another patient had QT prolongation (ie, QTc ⬎ 0.6). Two of the 38 patients treated with propafenone experienced drug intolerance (ie, headaches, nausea, and vomiting). One patient presented with complex ventricular arrhythmias during routine Holter monitoring after 3 weeks of propafenone therapy. Another patient treated with propafenone presented with atrial flutter requiring electrical cardioversion. One of 10 patients receiving disopyramide demonstrated dyspeptic symptoms, and another patient who experienced a dry mouth and urination difficulties refused to change drugs. Potential proarrhythmic effects were noted in 4 of 83 patients (4.8%) receiving drugs other than amiodarone. CHEST / 126 / 2 / AUGUST, 2004
481
Clinical and biochemical symptoms of hyperthyroidism and AF recurrence were observed in 1 of 59 study patients receiving long-term amiodarone therapy. One patient had decreased thyroid-stimulating hormone levels without clinical symptoms. Bleeding Complications: Oral acenocoumarol was used as thromboembolic prophylaxis in 74.3% patients in the rate control arm. Aspirin and ticlopidine were used in 19.8% and 1.0% of patients, respectively. At total of 4.9% of study patients received no anticoagulant agents because of a perceived low risk of systemic embolization.13 All patients assigned to sinus rhythm restoration received anticoagulation therapy (mean INR, 2.43 ⫾ 0.72) at least 3 weeks before and 4 weeks after undergoing cardioversion. In 15.6% of patients in whom cardioversion was successful, oral anticoagulants were prescribed for ⬎ 4 weeks after the procedure. In 62.2% of study subjects, thromboembolic prophylaxis was continued with aspirin, and one patient (0.9%) received ticlopidine. Nineteen patients (21.1%) who maintained sinus rhythm for ⬎ 1 month were prescribed no additional antithrombotic agents.13 Thirteen patients, all of whom were receiving acenocoumarol at the time, had minor bleeding complications outside the CNS (rate control group, 5 patients [5.0%]; rhythm control group, 8 patients [7.7%]). Hospitalization: Five patients (4.8%; 0.05 hospitalizations per person) who were assigned to ventricular rate control required hospitalization (mean length of hospital stay, 12.3 ⫾ 4.2 days). There were 211 hospitalizations in the group in which sinus rhythm restoration was attempted (mean length of hospital stay, 2.5 ⫾ 1.1 days). Excluding 104 hospitalizations required by protocol for cardioversion, there were 1.03 hospitalizations per person in the sinus rhythm group. The most frequent cause of hospital admission was recardioversion (94 cases; 87.9%). Thirteen patients (12.1%) required inpatient care due to treatment side effects and/or new or worsening concomitant disease. Clinical outcomes of the follow-up period in both study groups are presented in Table 2. Congestive Heart Failure: Both therapeutic strategies prevented the development and/or progression of heart failure (NYHA functional class). In the rate control group, heart failure regression was observed during the first 2 months of observation (p ⬍ 0.02), but no further improvement in NYHA functional class occurred during follow-up. Almost 2 years of observation revealed statistically significant heart failure regression in the rate control arm of the study 482
(p ⬍ 0.05). In the rhythm control group, a similar improvement of hemodynamic status (NYHA functional class) was observed during follow-up (p ⬍ 0.001) [Fig 2]. No difference in NYHA functional class between patients initially randomized to the two strategies was found at the end of the follow-up period. Exercise Tolerance: The rate control strategy did not result in a statistically significant difference in mean maximal workload during the treadmill test (5.3 ⫾ 1.1 vs 4.8 ⫾ 2.5 metabolic equivalents, respectively; p ⬎ 0.42). Conversely, in the rhythm control arm, exercise tolerance improvement and a significant increase in maximal workload or were observed (from 5.2 ⫾ 5.1 to 7.6 ⫾ 3.3 metabolic equivalents, respectively; p ⬍ 0.001). Both treatment strategies led to significantly increased exercise duration, but much more so in the rhythm control group (rate control group: 118.0 ⫾ 87.5 vs 157.8 ⫾ 126.2 s; p ⬍ 0.03; rhythm control group: 125.3 ⫾ 115.7 vs 294.7 ⫾ 216.7 s; p ⬍ 0.001). At the end of follow-up, both maximal workload and exercise duration were higher in the rhythm control arm (p ⬍ 0.001 and p ⬍ 0.001, respectively). Echocardiographic Parameters: In the rate control group, both right atrial enlargement (22.1 ⫾ 4.1 vs 23.3 ⫾ 3.8 cm2, respectively; p ⬍ 0.001) and left atrial enlargement (25.7 ⫾ 5.0 vs 26.4 ⫾ 5.0 cm2, respectively; p ⬍ 0.05) were observed during the follow-up period. Our observation demonstrated a significant decrease in right atrial size (21.8 ⫾ 4.1 vs 20.3 ⫾ 3.4 cm2, respectively; p ⬍ 0.01) and left atrial size (26.2 ⫾ 4.7 vs 25.4 ⫾ 5.0 cm2, respectively; p ⬍ 0.02) in the rhythm control arm. A significant increase in left ventricular fractional shortening (29.9 ⫾ 6.9% vs 34.5 ⫾ 8.9%, respectively; p ⬍ 0.001) was found in patients assigned to the sinus rhythm restoration group, while the rate control strategy failed to significantly affect left ventricular fractional shortening (32.8 ⫾ 6.6 vs 35.6 ⫾ 7.4%, respectively).
Discussion The results of the HOT CAFE study have provided data that are relevant to the management of patients with AF. Our study demonstrated that the rate control of AF is as acceptable as sinus rhythm restoration and maintenance. Despite the relatively small number of enrolled patients and the short observation period, our study suggests that, at least with currently available pharmacologic therapy, there is no distinct advantage associated with the rhythm control strategy. Moreover, the control of Clinical Investigations
Table 2—Side Effects* Side Effect Mean follow-up, yr Deaths From cardiovascular causes Total No. Total bleeding complications Major bleeding Minor bleeding Nasal cavity bleeding Urinary tract bleeding Oral cavity bleeding Epitaxis Ecchymosis Intracranial bleeding Thromboembolic complications Ischemic stroke Pulmonary embolism Antiarrhythmic therapy Proarrhythmia Ventricular tachycardia Bigeminy/trigeminy QTc prolongation Atrial flutter Somatic symptoms Headache Abdominal pain/dyspeptic symptoms Oral cavity dryness Urination difficulties Hyperthyroidism Others Impotence (during atenolol therapy) Invasive therapy Implantation of pacemaker AV nodal ablation Hospitalizations Total
Rate Control (n ⫽ 101)
Rhythm Control (n ⫽ 104)
Significance
1.6 ⫾ 0.9
1.8 ⫾ 0.3
NS
1 (1.0) 5
2 (2.0) 3 (2.9) 8
NS NS NS NS
1 (1.0) 3 (2.9) 1 (1.0) 1 (1.0) 2 (1.9)
NS NS NS NS NS
3 (2.9)
NS NS
4 (3.8) 1 (1.0) 1 (1.0) 1 (1.0) 1 (1.0) 8 (7.7) 1 (1.0) 3 (2.9) 1 (1.0) 1 (1.0) 2 (1.9)
NS NS NS NS NS NS NS NS NS NS NS
3 (3.0) 2 (2.0)
1 (1.0)
1 (1.0) 3 (3.0) 2 (2.0) 0.05/person 8
NS 2 (1.9) 1.03/person 32
NS NS 0.001 0.05
*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. AV ⫽ atrioventricular. See Table 1 for other abbreviations not used in the text.
ventricular response was easily achieved, and thus allowed for the good management of symptoms and exercise tolerance. One noted advantage of rate control is the obviation of antiarrhythmic drug therapy and its potential side effects. Furthermore, whereas antiarrhythmic drug prophylaxis must often be initiated at a hospital, this is not mandatory when the aim is rate control. Successful arrhythmia suppression leads to the significant improvement of hemodynamic parameters, exercise tolerance, and congestive heart failure relief. In the HOT CAFE study, patients in the rhythm control group had better rate control and a significant decrease in mean heart rate during follow-up. Ventricular response to the pharmacologic therapy protocol in the rate control group remained within satisfactory ranges during follow-up, although there were no significant differences in ventricular rate www.chestjournal.org
frequency between the following visits. Our results support those of the Pharmacological Intervention in Atrial Fibrillation (PIAF) study.10 In our Polish study, both the rhythm control and rate control strategies, which were monitored by exercise testing and Holter recordings, resulted in symptomatic improvement. Exercise tolerance improvement and congestive heart failure relief in both arms of the study reemphasizes the point that the careful control of ventricular response in patients with AF can result in a substantial benefit for the patient. However, while similar at baseline, patients in the rhythm control arm had a higher maximal workload and could exercise longer than patients in the rate control arm at the end of follow-up. The outcomes of our trial revealed that the rhythm control strategy resulted in an increase in left ventricular fractional shortening, which has been found CHEST / 126 / 2 / AUGUST, 2004
483
Figure 2. Congestive heart failure intensity estimated by NYHA functional class in both groups during the follow-up period.
in other studies.17 The right and left atria increased their dimensions during follow-up in the rate control group, while the opposite was found in the rhythm control group, suggesting a partial reversal of atrial remodeling and, possibly, a better chance to maintain sinus rhythm. On the other hand, the strategy of rate control was associated with significantly fewer hospital admissions. Cardioversion, AF recurrence, and the side effects arising from therapeutic strategy in our study group have significant economic implications. This is particularly relevant given the number of people with AF. Besides the benefits of sinus rhythm restoration (eg, the elimination of AF symptoms and improved 484
hemodynamics), the risk of thromboembolism occurrence also was expected to be reduced. Interestingly, so far no studies have shown that the restoration of normal sinus rhythm eliminates the risk of thromboembolic complications.9,18 In our study, strokes occurred in the sinus rhythm restoration group, but thromboembolic prophylaxis seemed to effectively protect rate control group patients. Longterm anticoagulation therapy, both in AF patients and in all patients with successfully restored sinus rhythm, as was proposed in the PIAF study,10 is not an attractive alternative. Oral anticoagulant therapy decreases the appeal of a rhythm control strategy, and not only in high-risk patients. Other factors, including more hospitalizations and more frequent Clinical Investigations
new arrhythmias in the rhythm control group, also indicated that rate control is at least as useful as sinus rhythm restoration in the management of patients with AF. Our results are similar to those of the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial9 and the Rate Control vs Electrical Cardioversion for Persistent Atrial Fibrillation (RACE) trial,11 however, the patient populations included in those studies differed slightly from that in ours. Our study population comprised younger patients, with a lower risk of systemic embolization, and mostly with the first episode of AF. The AFFIRM trial,9 the largest study to evaluate rate control vs rhythm control for AF found no differences between the two strategies. All-cause mortality, the primary end point of the study, did not differ significantly between the two groups (p ⫽ 0.056), although the survival curves separated at around 1.5 to 2 years in favor of the rate control strategy. The AFFIRM trial revealed no differences in several of the secondary end points, including functional status, quality of life, or major bleeding. The rate of ischemic stroke over the average 3.5-year follow-up period showed a trend favoring the rate control group. Also, hospitalization and new arrhythmia rates were higher in the rhythm control group. The RACE trial11 was the second study to address the issue of a rate control vs a rhythm control strategy in recurrent persistent AF patients. The RACE protocol was based on the hypothesis that rate control with medical therapy would prove inferior to rhythm control using cardioversion in tandem with therapy using antiarrhythmic drugs. The primary composite end point of the study was a combination of cardiovascular death, hospitalization for heart failure, thromboembolic complications, severe bleeding, pacemaker implantation, and severe drug side effects. Over an average of 3 years of follow-up, Van Gelder and colleagues11 found no differences in mortality rate between the study groups, although the incidence of the composite primary end point was higher in the rhythm control group. The difference also lay in some nonfatal end points (eg, more thromboembolic complications, more frequent heart failure, and more adverse drug effects in the rhythm control group). The results of these studies have major implications for the way that persistent AF is managed. Above all, the decision to restore and maintain normal sinus rhythm should be driven by patient symptoms. Some patients remain symptomatic even after the ventricular response was well-controlled. For example, there is a small subgroup of patients with reduced left ventricular compliance in whom atrial systole is important for ventricular filling. In www.chestjournal.org
these patients, the restoration and maintenance of normal sinus rhythm may be an important strategy for symptom control. In any case, when we elect to use a rate control and anticoagulation strategy, it is important to document (with ambulatory ECG recordings) that rate control remains satisfactory during everyday activities. Considering the results of these studies,9 –11,19 we consider rate control in AF to be an acceptable primary therapy. The two strategies seem equal in terms of mortality rate as well as thromboembolic and hemorrhagic risk. A rhythm control strategy without long-term anticoagulation therapy might result in a higher stroke rate, which implies that anticoagulation therapy should be extended after cardioversion, even in patients with a stable sinus rhythm. Our study has several limitations, one of which is the relatively small number of patients. This was due to the restrictive selection criteria that we applied in order to achieve a homogenous group of patients. This could have affected our results and prevented the detection of more differences between the two treatment strategies. Also, the number of our patients receiving anticoagulation therapy after sinus rhythm restoration and maintenance was relatively low. Yet, only a single ischemic stroke, unrelated to cardioversion, occurred in this subgroup of patients. Thus, rhythm control may still be a valid option for patients with persistent AF and relative contraindications to long-term anticoagulation therapy. Despite the limited number of patients, our findings are similar to those reported in other trials.9 –11,19 The HOT CAFE trial, being a randomized trial with well-defined end points and an almost 2-year follow-up period, may be a useful element in a future meta-analysis. We hope that our results will contribute to the discussion on rate control vs rhythm control to the benefit of patients with AF.
Conclusions The HOT CAFE study suggests that the rate control of persistent AF is equivalent to rhythm control in terms of mortality rate, stroke, and thromboembolic and severe bleeding complications. Therefore, our trial supports the results of other studies, although our population was younger and possibly had AF relatively longer than other populations (for up to 2 years). The HOT CAFE trial, like the PIAF, AFFIRM, RACE, and other trials, showed no significant differences in the two proposed treatment strategies for AF, so we still have to choose the best therapeutic approach based on an individual patient’s needs.9 –11,19 A meta-analysis of all randomCHEST / 126 / 2 / AUGUST, 2004
485
ized trials might offer the possibility of subgroup analysis and might help to optimize treatment procedures in everyday clinical practice. References 1 Pritchet ELC. Management of atrial fibrillation. N Engl J Med 1992; 326:1264 –1271 2 Stafford RS, Robson DC, Misra B, et al. Rate control and sinus rhythm maintenance in atrial fibrillation. Arch Intern Med 1998; 158:2144 –2148 3 Stanton MS, Prystowsky EN, Fineberg NS, et al. Arrhythmogenic drugs effects: a study of 506 patients treated for ventricular tachycardia or fibrillation. J Am Coll Cardiol 1989; 14:200 –215 4 Flaker GC, Blackshear JL, McBride R, et al. Antiarrhythmic drug therapy and cardiac mortality in atrial fibrillation. J Am Coll Cardiol 1992; 20:527–532 5 Friedman LP, Stevenson WG. Proarrhythmia. Am J Cardiol 1998; 82:50N–58N 6 Coplen SE, Antman EM, Berlin JA, et al. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion: a meta-analysis of randomized control trials. Circulation 1990; 82:1106 –1116 7 Crijns HJGM, Van Gelder IC, Van Glist WH, et al. Serial antiarrhythmic drug treatment to maintain sinus rhythm after electrical cardioversion for persistent atrial fibrillation or flutter. Am J Cardiol 1991; 68:335–341 8 Grogan M, Smith HC, Gersh BJ, et al. Left ventricular dysfunction due to atrial fibrillation in patients initially believed to have idiopathic dilated cardiomyopathy. Eur Heart J 1992; 13:1290 –1295 9 Wyse DG, Waldo AL, DiMarco JP, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002; 347:1825–1833 10 Hohnloser SH, Kuck KH, Lilienthal J, et al. Rhythm or rate control in atrial fibrillation: Pharmacological Intervention In Atrial Fibrillation (PIAF); a randomized trial. Lancet 2000; 356:1789 –1794
486
11 Van Gelder J, Hagens VE, Bosker HA, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002; 347:1834 – 1840 12 Opolski G, Torbicki A, Kosior D, et al. Should sinus rhythm be restored in patients with chronic atrial fibrillation? Preliminary results from the Polish “Hot Cafe” study. Pol Arch Med Wewn 1999; 5:413– 418 13 Laupacis A, Albers G, Dalen J, et al. Antithrombotic therapy in atrial fibrillation. Chest 1995; 108:352–359 14 Reimold SC, Chalmers TC, Berlin JA, et al. Assessment of the efficacy and safety of antiarrhythmic therapy for persistent atrial fibrillation: observation on the role of trial design and implications of drug-related mortality. Am Heart J 1992; 124:924 –927 15 Dreifus LS, Fisch C, Griffin JC, et al. Guidelines for implantation of cardiac pacemakers and antiarrhythmic devices: a report of the American College of Cardiology/American Heart Association task Force of Diagnostic and Therapeutic Cardiovascular Procedures (Committee on Pacemaker Implantation). Circulation 1991; 84:455– 467 16 Schiller NB, Shah PM, Crawford M, et al. Recommendations for quantification of left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantification of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989; 6:358 –367 17 Gosselink ATM, Crijns HJGM, van den Berg MP, et al. Functional capacity before and after cardioversion of atrial fibrillation: a controlled study. Br Heart J 1994; 72:616 – 621 18 Van Gelder IC, Crijns HJGM, Hillege HI, et al. Electrical cardioversion strategy does not reduce risk of thromboembolism in atrial fibrillation [abstract]. Eur Heart J 1995; 16(suppl):282 19 Carlsson J, Miketic S, Windeler J, et al. Randomized trial of rate-control versus rhythm control in persistent atrial fibrillation: The Strategies of Treatment of Atrial Fibrillation (STAF) study. J Am Coll Cardiol 2003; 41:1690 –1696
Clinical Investigations
Study objectives: The relative risks and benefits of strategies of rate control vs rhythm control in patients with atrial fibrillation (AF) remain to be fully explored. Design: The How to Treat Chronic Atrial Fibrillation (HOT CAFE) Polish trial was designed to evaluate in a randomized, multicenter, and prospective manner the feasibility and long-term outcomes of rate control vs rhythm control strategies in patients with persistent AF. Patients: Our study population comprised 205 patients (134 men and 71 women; mean [ⴞ SD] age, 60.8 ⴞ 11.2 years) with a mean AF duration of 273.7 ⴞ 112.4 days. The mean observation period was 1.7 ⴞ 0.4 years. One hundred one patients were randomly assigned to the rate control group and received rate-slowing therapy guided by repeated 24-h Holter monitoring. Direct current cardioversion and atrioventricular junctional ablation with pacemaker placement were alternative nonpharmacologic strategies for patients with tachycardia that was resistant to medical therapy. One hundred four patients were randomized to sinus rhythm restoration and maintenance using serial cardioversion supported by a predefined stepwise antiarrhythmic drug regimen (ie, disopyramide, propafenone, sotalol, and amiodarone). In both groups, thromboembolic prophylaxis followed current guidelines. Measurements and results: At the end of follow-up, 63.5% of patients in the rhythm control arm remained in sinus rhythm. No significant differences in the composite end point (ie, all-cause mortality, number of thromboembolic events, or major bleeding) were found between the rate control group and the rhythm control group (odds ratio, 1.98; 95% confidence interval, 0.28 to 22.3; p > 0.71). The incidence of hospital admissions was much lower in the rate control arm (12% vs 74%, respectively; p < 0.001). New York Heart Association functional class improved in both study groups, while mean exercise tolerance, as measured by the maximal treadmill workload, improved only in the rhythm control group (5.2 ⴞ 5.1 vs 7.6 ⴞ 3.3 metabolic equivalents, respectively; p < 0.001). The rhythm control strategy led to an increased mean left ventricular fractional shortening (29 ⴞ 7% vs 31 ⴞ 7%, respectively; p < 0.01). One episode of pulmonary embolism occurred in the rate control group despite oral anticoagulation therapy, while three patients in the rhythm control arm of the study experienced ischemic strokes (not significant). Conclusions: The Polish HOT CAFE study revealed no significant differences in major end points between the rate control group and the rhythm control group. (CHEST 2004; 126:476 – 486) Key words: antiarrhythmic therapy; atrial fibrillation; electrocardioversion rate control; rhythm control Abbreviations: AF ⫽ atrial fibrillation; AFFIRM ⫽ Atrial Fibrillation Follow-up Investigation of Rhythm Management; HOT CAFE ⫽ How to Treat Chronic Atrial Fibrillation; INR ⫽ international normalized ratio; NYHA ⫽ New York Heart Association; PIAF ⫽ Pharmacological Intervention in Atrial Fibrillation; RACE ⫽ Rate Control vs Electrical Cardioversion for Persistent Atrial Fibrillation
476
Clinical Investigations
goal in patients with atrial fibrilT helationtherapeutic (AF) can be either the restoring and maintaining of sinus rhythm (mainly with antiarrhythmic therapy) or allowing persistent AF to become permanent, while controlling the ventricular rate and preventing systemic thromboembolism.1,2 The debate over rate control vs rhythm control in the treatment of AF still is ongoing. It has been a standard practice to maintain sinus rhythm, primarily through the use of antiarrhythmic drugs and cardioversion, which relieves symptoms and may retard or prevent progression to permanent AF. Rhythm control, however, carries a risk of proarrhythmias and negative inotropic effects due to the use of antiarrhythmic agents.3–5 The prevention of AF recurrence with antiarrhythmic drug therapy, regardless of the agent, is achieved in only approximately 50% of patients at 6 months to 1 year.6,7 Moreover, optimal thromboembolic prevention in the rhythm control strategy has not yet been defined. While rate control is easier to achieve and leads toward improved hemodynamics, it allows AF to persist and may not eliminate AF symptoms (eg, palpitations, dyspnea, chest pain, fatigue, and dizziness).8 The How to Treat Chronic Atrial Fibrillation (HOT CAFE) study was designed to compare the strategy of sinus rhythm restoration and maintenance with that of ventricular rate control and chronic thromboembolic prophylaxis. We report the results of the initial phase of the study, which involved six clinical centers and a limited number of patients. This phase was implemented to confirm the study protocol and its methods before the planned main phase of the study, which was ultimately abandoned due to the results of other trials.9 –11
*From the Departments of Cardiology (Drs. Opolski and Kosior) and Internal Medicine and Hypertension (Dr. Szulc), Medical University of Warsaw, Warsaw, Poland; the Department of Chest Medicine (Dr. Torbicki), National Institute of Tuberculosis and Pulmonary Disease, Warsaw, Poland; the Department of Cardiology (Dr. Woz˙akowska-Kapłon), Municipal Hospital, Kielce, Poland; the Department of Cardiology (Dr. Kolodziej), Municipal Hospital, Siedlce, Poland; and the Department of Cardiology (Dr. Achremczyk), Municipal Hospital, Radom, Poland. The study results were presented at the 24th Congress of the European Society of Cardiology, Berlin, Germany, in October 2002. The study was supported by Polish Government research grant No. PO5B06012. Manuscript received August 8, 2003; revision accepted February 11, 2004. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: [email protected]). Correspondence to: Grzegorz Opolski, MD, PhD, Chair, Department of Cardiology, Medical University of Warsaw, Warsaw, Poland; e-mail: [email protected] www.chestjournal.org
Materials and Methods The HOT CAFE study12 was a prospective, randomized, open multicenter clinical trial that was designed to evaluate the effects of rhythm control vs rate control in patients with persistent AF. The study involved six cardiology centers in central Poland, five of which enrolled patients. The study was approved by the institutional review board or ethics committee at those centers, and consecutive, eligible patients provided written informed consent. Recruitment began in March 1997, randomization was concluded in December 2000, and follow-up was terminated in December 2002. Inclusion and Exclusion Criteria Eligible patients had to meet the following inclusion criteria: 50 to 75 years of age; and AF had to be known to be present for at least 7 days, but not for ⬎ 2 years. Only patients with a first clinically overt persistent episode of AF were enrolled. Persistent AF was defined as a non-self-terminating arrhythmia requiring electrical cardioversion to restore sinus rhythm. The following were exclusion criteria: documented inefficiency; intolerance to or contraindications for treatment with antiarrhythmic drugs; presence of arrhythmia associated with an acute reversible condition; thyroid dysfunction; pregnancy or lactation; history of myocardial infarction within 3 months preceding enrollment into the study; acute myocarditis; cardiac surgery during the previous 30 days; severe cardiac disability (ie, New York Heart Association [NYHA] functional class IV); severe systemic hypertension not responding to treatment (diastolic pressure, ⬎ 115 mm Hg); hypotension (systolic pressure, ⬍ 90 mm Hg); history of transient ischemic attack (with significantly marked vascular pathology requiring surgical intervention); history of hemorrhagic stroke; ischemic stroke during the 3 months preceding entrance into the trial; any mitral stenosis or other valvular disease suitable for surgical treatment; R-R intervals exceeding 3 s; ventricular response to AF of ⬍ 90 beats/min (unrelated to drugs used to reduce ventricular rate); bundle branch block or QT-segment prolongation (ie, corrected QT interval of ⬎ 480 ms or uncorrected QT interval of ⬎ 500 ms); alcoholism; contraindications to anticoagulation therapy; liver, kidney, or CNS damage; advanced chronic lung disease, malignancy; or any noncardiac illness associated with a life expectancy of ⬍ 1 year; or participation in another study. In addition, premenopausal women who were still capable of procreation and had not undergone tubal ligation or hysterectomy also were excluded from the study. Randomization, Therapy, and Follow-up Participating patients were randomly assigned to one of the two groups following a predefined sequential management (Fig 1). The randomization of eligible patients was performed by permuted block design with equal allocation and was stratified centrally by a steering center of the Chair and Department of Cardiology. The trial was conducted on an intention-to-treat basis. All patients were seen in the outpatient department at least once a month during the first 6 months of follow-up and every 3 months until the end of the study. All patients enrolled in the trial were potentially eligible for either rate control or rhythm control therapy at entrance into the study. Rate Control Group: The therapeutic strategy in the rate control group was rate control and chronic thromboembolic prophylaxis, with heart rate as the therapeutic target. Heart rate was considered to be adequately controlled when the ventricular response ranged between 70 and 90 beats/min during rest CHEST / 126 / 2 / AUGUST, 2004
477
Figure 1. The HOT CAFE trial design. SR ⫽ sinus rhythm; DC ⫽ direct current; * ⫽ disopyramide, propafenone, or sotalol.
(12-lead resting ECG) and not exceeding 140 beats/min during moderate exercise. The circadian pattern and mean rate over a 24-h period were monitored by ECG Holter recordings performed at study entry, and were repeated 1 and 3 months after randomization, and every 6 months thereafter. Exercise testing was used to evaluate heart rate during submaximal exercise at the beginning of observation, at 2, 12, and 24 months after randomization, and at the end of the study. The pharmacologic therapies approved for use in this study arm included the following: -blockers; nondihydropyridine calcium blockers; digoxin; or a combination of these drugs. Cardioversion and atrioventricular junctional ablation with pacemaker placement were alternative nonpharmacologic strategies in patients with resistant tachycardia. Anticoagulation therapy was considered in all patients who were assigned to the rate control group. Following American College of Chest Physicians guidelines,13 oral anticoagulation therapy with a target international normalized ratio (INR) of 2.5 (range, 2.0 to 3.0) was recommended for patients at high risk of stroke. Systemic embolization risk factors included prior stroke or transient ischemic attack, left ventricular dysfunction, congestive heart failure, diabetes mellitus, history of hypertension, and coronary artery disease. Patients between 65 and 75 years of age who had no risk factors were offered therapy with 325 mg aspirin daily or, in patients with proven aspirin intolerance, ticlopidine, 500 mg per day. Patients ⬍ 55 years of age without clinical or echocardiographic evidence of cardiovascular disease, diabetes, or a history of stroke or systemic embolism could decline pharmacologic thromboembolic prophylaxis.13 Rhythm Control Group: Patients randomly assigned to the rhythm control group underwent electrical cardioversion without prior treatment with antiarrhythmic drugs. Cardioversion was 478
considered to be successful if sinus rhythm was maintained for 2 h after the procedure, at which point the first dose of an antiarrhythmic drug was administered. Following sinus rhythm restoration, patients received antiarrhythmic drug therapy. Due to potential proarrhythmic effects, the choice of drug was open, based on arrhythmia etiology, concomitant heart disease, and the patient’s age.14 As a first step the patients received propafenone (450 to 600 mg per day), disopyramide (300 to 600 mg per day), or sotalol (160 to 320 mg per day). Patients with ischemic heart disease usually received sotalol or disopyramide. Class I drugs were avoided in patients with left ventricular dysfunction on echocardiographic examination or signs of congestive heart failure. Propafenone and sotalol were the drugs of first choice in patients with systemic hypertension. -blockers were allowed in addition to disopyramide or propafenone when considered clinically indicated. A recurrence of AF, which was defined as AF lasting at least 12 h, mandated a repeat cardioversion. Once the sinus rhythm was restored, another predefined antiarrhythmic drug was administered in the sequence of the above-mentioned first-line drug. When a recurrence occurred after 6 months of antiarrhythmic therapy, cardioversion was attempted. If successful, the same antiarrhythmic regimen was continued. In situations in which there was no response to such a pharmacologic treatment and AF recurred, an oral daily dose of amiodarone (600 mg) was administered over 3 weeks, followed by 400-mg doses until the loading dose of at least 12.0 to 16.0 g was reached prior to another cardioversion. Where successful, the restored sinus rhythm was maintained through daily maintenance doses of 100 to 200 mg amiodarone. However, in patients in whom there was no response to the initial direct current cardioversion, they received amiodarone, as Clinical Investigations
above, with cardioversion attempted again. Patients in whom sinus rhythm was restored continued their amiodarone treatment in daily maintenance doses. The amiodarone therapy started at the hospital was continued on an outpatient basis with subjects being assessed at least weekly. In patients in whom there was no response to the first and second cardioversions, with the second cardioversion preceded by amiodarone loading, or there was a recurrence of AF without an identifiable cause while receiving amiodarone, the arrhythmia was considered permanent, and rate control became the goal of therapy. Only electrocardiographically confirmed episodes lasting ⬎ 12 h despite additional antiarrhythmic drug therapy were considered to be clinically significant. Patients experiencing such episodes received anticoagulation prophylaxis, with ventricular rhythm control as additional therapy. Patients who required pacing for rate support and/or to maintain normal sinus rhythm remained in the study. Current guidelines15 were followed when pacing was used. In the rhythm control arm of the study, thromboembolic prophylaxis (with acenocoumarol) titrated to achieve an INR of 2.0 to 3.0 was started 4 weeks before cardioversion.13 Therapy was continued in all patients until normal sinus rhythm had been maintained for at least 4 weeks. Decisions regarding prolonged or chronic anticoagulation or antiplatelet therapy were left to the discretion of the local clinical teams but had to be reported.13 Assessment and Follow-up A physician and a nurse coordinator assessed patients initially and then at least monthly for 6 months, and every 3 months thereafter. Twelve-lead ECGs were obtained at each visit. Chest radiograph examinations were obtained at the first follow-up appointment and were repeated annually thereafter. The level of serum thyroid-stimulating hormone was measured every 6 months in patients receiving amiodarone. All study patients had their exercise tolerance estimated by a stress test that was performed at study entry, 2 months after randomization, and then annually. In addition, atrial and ventricular morphology and function were echocardiographically assessed during the followup period. The primary end point was a composite of death from any cause, thromboembolic complications (especially disabling ischemic stroke), and intracranial or other major hemorrhage. Disabling stroke was defined as an acute neurologic deficit for ⬎ 24 h that affected the ability to perform activities of daily life or resulted in death. Cerebrovascular events required diagnosis by CT scan and neurology consultation. Major bleeding was defined as a drop in hemoglobin levels of ⬎ 2 g/L, requiring a blood transfusion, needing surgical intervention, or resulting in death. All events occurring between randomization and the end of the study were recorded. All deaths were considered to be due to cardiovascular causes unless an unequivocal noncardiac cause could be identified. Secondary end points included rate control, sinus rhythm maintenance, discontinuation of therapy (especially if due to a potential proarrhythmic effect), hemorrhage, hospitalization, new or worsening congestive heart failure, or changes in exercise tolerance.
measurements were averaged for more than five cardiac cycles recorded during relaxed expiration. Echocardiography was performed at study entry, 2 months after randomization, and at the end of the study. In patients assigned to rhythm control, echocardiographic examination was performed prior to and 30 days after successful cardioversion, and at the end of the observation period. Exercise Testing On entry into the study, all patients underwent a treadmill exercise stress test, which was performed according to the symptom-related modified Bruce I protocol. This was repeated at 2, 12, and 24 months after randomization, and at the end of the study. In addition, in the rhythm control arm, this testing was performed prior to and 30 days after successful cardioversion. Exercise was continued until the patient was exhausted or developed progressive angina, hypotension, or ventricular arrhythmia. Age-predicted maximal heart rate, shifts in ST segments, the emergence of ventricular ectopic beats, or the development of atrial arrhythmia did not constitute end points of the treadmill test. Heart rate during AF was determined by measuring the average heart rate for a period of 60 s. Maximal workload and exercise duration also were considered. An exercise stress test was repeated as needed in the rate control group to achieve optimal heart rate control. 24-h ECG Holter Monitoring The circadian pattern and the mean 24-h heart rate were monitored with an ECG Holter recording performed at the study entry, at 1 and 3 months after randomization, and every 6 months thereafter. In addition, 24-h ECG monitoring was performed in patients assigned to the rhythm control strategy 7 and 30 days after successful cardioversion or after every antiarrhythmic drug change. Statistical Analysis Summary data are expressed as the mean ⫾ SD or as the numbers and percentages of patients. Analyses were performed using the intention-to-treat principle. Continuous variables were tested by analysis of variance for between-group comparisons (with the baseline value as the covariable). Dependent on analysis of variance results, the Student t test (in modification for paired measurements if necessary) was used to compare mean parameter values in each group on one visit or in one group on each visit. The Bonferroni correction was used to avoid multipletesting problems. The change in time of NYHA functional class was tested using the Wilcoxon test, and between-group comparisons were tested using the Kruskal-Wallis test on each visit. All pair comparisons, if needed, were made by the Dunn test. The cumulative risk of AF recurrence was estimated by the KaplanMeier product-limit method. Data were censored if the patient died, reached the end of the follow-up period, or was lost to follow-up without the occurrence of a primary end point. The difference between treatment groups was assessed with the log-rank test with a p value of ⬍ 0.05 considered to be statistically significant.
Echocardiography M-mode echocardiographic measurements including left atrial diameter and left ventricular diameter, were made at enddiastole in the parasternal long-axis view using standard techniques.16 Left ventricular fractional shortening was determined on two-dimensional imaging. The left atrial long axis was measured at end-diastole from the apical four-chamber view. All www.chestjournal.org
Results Baseline Characteristics Of 738 screened patients with AF from March 1997 to December 2000, 205 patients (134 men and CHEST / 126 / 2 / AUGUST, 2004
479
71 women) met the inclusion criteria and were enrolled into the study. Valvular heart disease (21%), thyrotoxicosis (13%), and advanced stages of concomitant diseases influencing the natural history of AF (11%) were the main exclusion criteria limiting enrollment. One hundred one consecutive patients were randomly assigned to the heart rate control group. The restoration and maintenance of sinus rhythm for a period of at least 12 months was attempted in 104 patients. Patients were followed up for a mean of 1.7 ⫾ 0.4 years (maximum follow-up, 2.5 years). There were no significant differences in the duration of follow-up between the two study groups. The mean age was 60.8 ⫾ 11.2 years. The mean arrhythmia duration prior to enrollment was 273.7 ⫾ 112.4 days. The characteristics of the study patients were typical of a population of subjects affected by persistent AF. There were no significant
differences in baseline characteristics between the rate control group and the rhythm control group. A comparison of the clinical and echocardiographic characteristics of study patients on enrollment is shown in Table 1. Primary End Point None of the patients was lost to follow-up. Allcause mortality, and the number of thromboembolic and major bleeding complications, the primary composite end point of the study, were not significantly different between the rate control group and the rhythm control group (odds ratio, 1.98; 95% confidence interval, 0.28 to 22.3; p ⬎ 0.71). One patient (1.0%) assigned to the rate control group died due to leukemia diagnosed 6 months after enrollment. Three patients (2.9%) died in the rhythm control
Table 1—Baseline Characteristics of the Study Groups* Parameter Age, yr Gender Female Male AF duration 7 d–1 mo 1 mo–2 yr Mean AF duration, d AF etiology Ischemic heart disease Myocardial infarction CABG Hypertension Valvular heart disease Lone AF Diabetes Congestive heart failure, NYHA scale I II III Echocardiographic parameters LAsax, mm LAlax, mm LVEDD, mm FS, % Pharmacologic therapy, % -adrenolitics Verapamil/diltiazem Digoxin -adrenolitics ⫹ digoxin ACE-I Amlodipine Nitrates Diuretics
Rate Control (n ⫽ 101)
Rhythm Control (n ⫽ 104)
Significance
61.4 ⫾ 17.6
60.4 ⫾ 7.9
NS
38 (37.6) 63 (62.4)
33 (31.7) 71 (68.3)
NS NS
17 (16.8) 84 (83.2) 243.2 ⫾ 137.3
16 (15.4) 88 (84.6) 220.4 ⫾ 148.6
NS NS NS
38 (37.6) 7 (6.9) 0 (0.0) 60 (59.4) 15 (14.8) 25 (24.8) 18 (17.8)
52 (50.0) 7 (6.7) 1 (1.0) 72 (69.2) 16 (15.4) 18 (17.3) 15 (14.4)
NS NS NS NS NS NS NS
48 (47.5) 48 (47.5) 5 (5.0)
30 (28.8) 59 (56.7) 15 (14.4)
NS NS NS
48.0 ⫾ 4.7 64.0 ⫾ 6.8 50.8 ⫾ 5.9 32.8 ⫾ 6.6
47.4 ⫾ 5.3 62.5 ⫾ 7.9 52.2 ⫾ 6.8 29.9 ⫾ 6.9
NS NS NS NS
49.5 7.9 3.0 39.6 71.2 12.1 9.1 16.1
52.1 7.7 6.5 31.7 69.2 14.4 8.7 14.4
NS NS NS NS NS NS NS NS
*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. NS ⫽ not significant; CABG ⫽ coronary artery bypass grafting; ACE ⫽ angiotensin-converting enzyme; LAsax ⫽ left atrial anteroposterior axis; LAlax ⫽ left atrial longitudinal axis; LVEDD ⫽ left ventricular end diastolic diameter; FS ⫽ left ventricular fractional shortening. 480
Clinical Investigations
group, two (1.9%) from strokes. One female patient (1.0%) from the rhythm control arm died in a motor vehicle accident in which she was a passenger, in the 10th month of observation. Three patients (2.9%) assigned to the rhythm control group developed massive cerebral strokes. Two of the strokes (1.9%), both resulting in death, were observed on the third day following successful cardioversion despite appropriate anticoagulant therapy (INR, 2.8 ⫾ 0.7) during a sufficient period of time (35.3 ⫾ 11.2 days) prior to the procedure. In both cases, a cardiogenic etiology was supported by the exclusion of all other possible causes. The third stroke (0.9%), which was not disabling, affected a patient with AF recurrence during amiodarone therapy. The patient was receiving long-term aspirin therapy. Pulmonary embolism occurred despite oral anticoagulation in one patient (1.0%) in the rate control group. One patient from the rhythm control arm withdrew his consent after 6 months of observation. Secondary End Points Rate Control and Maintenance of Sinus Rhythm: The treatment goal in the rate control group was an optimal heart rate, as defined by 24-h Holter monitoring. Calcium antagonists (7.9%), -blockers (49.5%), -blockers plus digoxin (39.6%), or digoxin alone (3.0%) were used. Two patients (2.0%) in this study arm developed symptomatic, sustained tachycardia despite receiving pharmacologic therapy, and ultimately underwent atrioventricular node ablation with pacemaker implantation. One patient (1.0%) in the rate control arm required a pacemaker due to symptomatic bradycardia and pauses of ⬎ 3.5 s. Planned electrical cardioversion was performed in all 104 patients assigned to receive rhythm control. Following the first electrical cardioversion, sinus rhythm was restored in 56 patients (53.8%). Fortyeight patients failed to convert and received loading doses of amiodarone prior to the next cardioversion attempt. Sinus rhythm recovery occurred in 10 patients (20.8%) during amiodarone loading. The second electrical cardioversion, preceded by amiodarone loading, restored sinus rhythm in 24 of 38 patients (63.2%). The first and second cardioversions, the latter preceded by amiodarone loading, enabled sinus rhythm restoration in 90 patients (86.5%). Sixty-six patients (63.5%) in the rhythm control arm were in sinus rhythm at the end of the study. In 27 patients, sinus rhythm was successfully maintained with the first antiarrhythmic compound following the initial effective cardioversion. Sequential electrical cardioversion and antiarrhythmic treatment resulted in sinus rhythm maintenance in an www.chestjournal.org
additional 15 patients who previously had been subjected to a successful initial procedure. Amiodarone administered in loading doses prior to the second cardioversion and maintenance therapy following the successful procedure prevented AF recurrence in the remaining 24 subjects whose initial cardioversion had failed. Altogether, 37 patients (56.0%) received amiodarone as AF prophylaxis following sinus rhythm restoration. Two patients (1.9%) in the rhythm control arm required pacemaker placement due to bradycardia. In both patients, antiarrhythmic therapy was continued. There was a continuous decline in heart rate during the follow-up period in patients who were assigned to receive rhythm control. The mean heart rate measured 30 days after electrical cardioversion was significantly lower than that at randomization (86.0 ⫾ 11.6 vs 77.8 ⫾ 12.6 beats/min, respectively; p ⬍ 0.001). There were no other significant differences in measured parameters during further observation. However, the mean heart rate during subsequent visits was significantly lower than that at entry in the sinus rhythm group. The mean heart rate in 24-h Holter monitoring was significantly lower during rhythm control (79.1 ⫾ 8.6 beats/min) than during rate control (85.8 ⫾ 7.5 beats/min; p ⬍ 0.003). There were no significant changes in mean heart rate in the rate control arm of the study during the follow-up. Rate control therapy maintained the ventricular response rate within satisfactory ranges during the follow-up (at randomization 86.2 ⫾ 7.8 beats/ min; at the end of follow-up, 83.1 ⫾ 4.3 beats/min; p ⬎ 0.093). Discontinuation of Therapy: Two of 25 patients who had received sotalol in the sinus rhythm group required drug withdrawal due to abdominal pain, diarrhea, or eczema, all of which disappeared on drug termination. One patient experienced torsade de pointes during a routine 24-h Holter monitoring session, and another patient had QT prolongation (ie, QTc ⬎ 0.6). Two of the 38 patients treated with propafenone experienced drug intolerance (ie, headaches, nausea, and vomiting). One patient presented with complex ventricular arrhythmias during routine Holter monitoring after 3 weeks of propafenone therapy. Another patient treated with propafenone presented with atrial flutter requiring electrical cardioversion. One of 10 patients receiving disopyramide demonstrated dyspeptic symptoms, and another patient who experienced a dry mouth and urination difficulties refused to change drugs. Potential proarrhythmic effects were noted in 4 of 83 patients (4.8%) receiving drugs other than amiodarone. CHEST / 126 / 2 / AUGUST, 2004
481
Clinical and biochemical symptoms of hyperthyroidism and AF recurrence were observed in 1 of 59 study patients receiving long-term amiodarone therapy. One patient had decreased thyroid-stimulating hormone levels without clinical symptoms. Bleeding Complications: Oral acenocoumarol was used as thromboembolic prophylaxis in 74.3% patients in the rate control arm. Aspirin and ticlopidine were used in 19.8% and 1.0% of patients, respectively. At total of 4.9% of study patients received no anticoagulant agents because of a perceived low risk of systemic embolization.13 All patients assigned to sinus rhythm restoration received anticoagulation therapy (mean INR, 2.43 ⫾ 0.72) at least 3 weeks before and 4 weeks after undergoing cardioversion. In 15.6% of patients in whom cardioversion was successful, oral anticoagulants were prescribed for ⬎ 4 weeks after the procedure. In 62.2% of study subjects, thromboembolic prophylaxis was continued with aspirin, and one patient (0.9%) received ticlopidine. Nineteen patients (21.1%) who maintained sinus rhythm for ⬎ 1 month were prescribed no additional antithrombotic agents.13 Thirteen patients, all of whom were receiving acenocoumarol at the time, had minor bleeding complications outside the CNS (rate control group, 5 patients [5.0%]; rhythm control group, 8 patients [7.7%]). Hospitalization: Five patients (4.8%; 0.05 hospitalizations per person) who were assigned to ventricular rate control required hospitalization (mean length of hospital stay, 12.3 ⫾ 4.2 days). There were 211 hospitalizations in the group in which sinus rhythm restoration was attempted (mean length of hospital stay, 2.5 ⫾ 1.1 days). Excluding 104 hospitalizations required by protocol for cardioversion, there were 1.03 hospitalizations per person in the sinus rhythm group. The most frequent cause of hospital admission was recardioversion (94 cases; 87.9%). Thirteen patients (12.1%) required inpatient care due to treatment side effects and/or new or worsening concomitant disease. Clinical outcomes of the follow-up period in both study groups are presented in Table 2. Congestive Heart Failure: Both therapeutic strategies prevented the development and/or progression of heart failure (NYHA functional class). In the rate control group, heart failure regression was observed during the first 2 months of observation (p ⬍ 0.02), but no further improvement in NYHA functional class occurred during follow-up. Almost 2 years of observation revealed statistically significant heart failure regression in the rate control arm of the study 482
(p ⬍ 0.05). In the rhythm control group, a similar improvement of hemodynamic status (NYHA functional class) was observed during follow-up (p ⬍ 0.001) [Fig 2]. No difference in NYHA functional class between patients initially randomized to the two strategies was found at the end of the follow-up period. Exercise Tolerance: The rate control strategy did not result in a statistically significant difference in mean maximal workload during the treadmill test (5.3 ⫾ 1.1 vs 4.8 ⫾ 2.5 metabolic equivalents, respectively; p ⬎ 0.42). Conversely, in the rhythm control arm, exercise tolerance improvement and a significant increase in maximal workload or were observed (from 5.2 ⫾ 5.1 to 7.6 ⫾ 3.3 metabolic equivalents, respectively; p ⬍ 0.001). Both treatment strategies led to significantly increased exercise duration, but much more so in the rhythm control group (rate control group: 118.0 ⫾ 87.5 vs 157.8 ⫾ 126.2 s; p ⬍ 0.03; rhythm control group: 125.3 ⫾ 115.7 vs 294.7 ⫾ 216.7 s; p ⬍ 0.001). At the end of follow-up, both maximal workload and exercise duration were higher in the rhythm control arm (p ⬍ 0.001 and p ⬍ 0.001, respectively). Echocardiographic Parameters: In the rate control group, both right atrial enlargement (22.1 ⫾ 4.1 vs 23.3 ⫾ 3.8 cm2, respectively; p ⬍ 0.001) and left atrial enlargement (25.7 ⫾ 5.0 vs 26.4 ⫾ 5.0 cm2, respectively; p ⬍ 0.05) were observed during the follow-up period. Our observation demonstrated a significant decrease in right atrial size (21.8 ⫾ 4.1 vs 20.3 ⫾ 3.4 cm2, respectively; p ⬍ 0.01) and left atrial size (26.2 ⫾ 4.7 vs 25.4 ⫾ 5.0 cm2, respectively; p ⬍ 0.02) in the rhythm control arm. A significant increase in left ventricular fractional shortening (29.9 ⫾ 6.9% vs 34.5 ⫾ 8.9%, respectively; p ⬍ 0.001) was found in patients assigned to the sinus rhythm restoration group, while the rate control strategy failed to significantly affect left ventricular fractional shortening (32.8 ⫾ 6.6 vs 35.6 ⫾ 7.4%, respectively).
Discussion The results of the HOT CAFE study have provided data that are relevant to the management of patients with AF. Our study demonstrated that the rate control of AF is as acceptable as sinus rhythm restoration and maintenance. Despite the relatively small number of enrolled patients and the short observation period, our study suggests that, at least with currently available pharmacologic therapy, there is no distinct advantage associated with the rhythm control strategy. Moreover, the control of Clinical Investigations
Table 2—Side Effects* Side Effect Mean follow-up, yr Deaths From cardiovascular causes Total No. Total bleeding complications Major bleeding Minor bleeding Nasal cavity bleeding Urinary tract bleeding Oral cavity bleeding Epitaxis Ecchymosis Intracranial bleeding Thromboembolic complications Ischemic stroke Pulmonary embolism Antiarrhythmic therapy Proarrhythmia Ventricular tachycardia Bigeminy/trigeminy QTc prolongation Atrial flutter Somatic symptoms Headache Abdominal pain/dyspeptic symptoms Oral cavity dryness Urination difficulties Hyperthyroidism Others Impotence (during atenolol therapy) Invasive therapy Implantation of pacemaker AV nodal ablation Hospitalizations Total
Rate Control (n ⫽ 101)
Rhythm Control (n ⫽ 104)
Significance
1.6 ⫾ 0.9
1.8 ⫾ 0.3
NS
1 (1.0) 5
2 (2.0) 3 (2.9) 8
NS NS NS NS
1 (1.0) 3 (2.9) 1 (1.0) 1 (1.0) 2 (1.9)
NS NS NS NS NS
3 (2.9)
NS NS
4 (3.8) 1 (1.0) 1 (1.0) 1 (1.0) 1 (1.0) 8 (7.7) 1 (1.0) 3 (2.9) 1 (1.0) 1 (1.0) 2 (1.9)
NS NS NS NS NS NS NS NS NS NS NS
3 (3.0) 2 (2.0)
1 (1.0)
1 (1.0) 3 (3.0) 2 (2.0) 0.05/person 8
NS 2 (1.9) 1.03/person 32
NS NS 0.001 0.05
*Values given as mean ⫾ SD or No. (%), unless otherwise indicated. AV ⫽ atrioventricular. See Table 1 for other abbreviations not used in the text.
ventricular response was easily achieved, and thus allowed for the good management of symptoms and exercise tolerance. One noted advantage of rate control is the obviation of antiarrhythmic drug therapy and its potential side effects. Furthermore, whereas antiarrhythmic drug prophylaxis must often be initiated at a hospital, this is not mandatory when the aim is rate control. Successful arrhythmia suppression leads to the significant improvement of hemodynamic parameters, exercise tolerance, and congestive heart failure relief. In the HOT CAFE study, patients in the rhythm control group had better rate control and a significant decrease in mean heart rate during follow-up. Ventricular response to the pharmacologic therapy protocol in the rate control group remained within satisfactory ranges during follow-up, although there were no significant differences in ventricular rate www.chestjournal.org
frequency between the following visits. Our results support those of the Pharmacological Intervention in Atrial Fibrillation (PIAF) study.10 In our Polish study, both the rhythm control and rate control strategies, which were monitored by exercise testing and Holter recordings, resulted in symptomatic improvement. Exercise tolerance improvement and congestive heart failure relief in both arms of the study reemphasizes the point that the careful control of ventricular response in patients with AF can result in a substantial benefit for the patient. However, while similar at baseline, patients in the rhythm control arm had a higher maximal workload and could exercise longer than patients in the rate control arm at the end of follow-up. The outcomes of our trial revealed that the rhythm control strategy resulted in an increase in left ventricular fractional shortening, which has been found CHEST / 126 / 2 / AUGUST, 2004
483
Figure 2. Congestive heart failure intensity estimated by NYHA functional class in both groups during the follow-up period.
in other studies.17 The right and left atria increased their dimensions during follow-up in the rate control group, while the opposite was found in the rhythm control group, suggesting a partial reversal of atrial remodeling and, possibly, a better chance to maintain sinus rhythm. On the other hand, the strategy of rate control was associated with significantly fewer hospital admissions. Cardioversion, AF recurrence, and the side effects arising from therapeutic strategy in our study group have significant economic implications. This is particularly relevant given the number of people with AF. Besides the benefits of sinus rhythm restoration (eg, the elimination of AF symptoms and improved 484
hemodynamics), the risk of thromboembolism occurrence also was expected to be reduced. Interestingly, so far no studies have shown that the restoration of normal sinus rhythm eliminates the risk of thromboembolic complications.9,18 In our study, strokes occurred in the sinus rhythm restoration group, but thromboembolic prophylaxis seemed to effectively protect rate control group patients. Longterm anticoagulation therapy, both in AF patients and in all patients with successfully restored sinus rhythm, as was proposed in the PIAF study,10 is not an attractive alternative. Oral anticoagulant therapy decreases the appeal of a rhythm control strategy, and not only in high-risk patients. Other factors, including more hospitalizations and more frequent Clinical Investigations
new arrhythmias in the rhythm control group, also indicated that rate control is at least as useful as sinus rhythm restoration in the management of patients with AF. Our results are similar to those of the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) trial9 and the Rate Control vs Electrical Cardioversion for Persistent Atrial Fibrillation (RACE) trial,11 however, the patient populations included in those studies differed slightly from that in ours. Our study population comprised younger patients, with a lower risk of systemic embolization, and mostly with the first episode of AF. The AFFIRM trial,9 the largest study to evaluate rate control vs rhythm control for AF found no differences between the two strategies. All-cause mortality, the primary end point of the study, did not differ significantly between the two groups (p ⫽ 0.056), although the survival curves separated at around 1.5 to 2 years in favor of the rate control strategy. The AFFIRM trial revealed no differences in several of the secondary end points, including functional status, quality of life, or major bleeding. The rate of ischemic stroke over the average 3.5-year follow-up period showed a trend favoring the rate control group. Also, hospitalization and new arrhythmia rates were higher in the rhythm control group. The RACE trial11 was the second study to address the issue of a rate control vs a rhythm control strategy in recurrent persistent AF patients. The RACE protocol was based on the hypothesis that rate control with medical therapy would prove inferior to rhythm control using cardioversion in tandem with therapy using antiarrhythmic drugs. The primary composite end point of the study was a combination of cardiovascular death, hospitalization for heart failure, thromboembolic complications, severe bleeding, pacemaker implantation, and severe drug side effects. Over an average of 3 years of follow-up, Van Gelder and colleagues11 found no differences in mortality rate between the study groups, although the incidence of the composite primary end point was higher in the rhythm control group. The difference also lay in some nonfatal end points (eg, more thromboembolic complications, more frequent heart failure, and more adverse drug effects in the rhythm control group). The results of these studies have major implications for the way that persistent AF is managed. Above all, the decision to restore and maintain normal sinus rhythm should be driven by patient symptoms. Some patients remain symptomatic even after the ventricular response was well-controlled. For example, there is a small subgroup of patients with reduced left ventricular compliance in whom atrial systole is important for ventricular filling. In www.chestjournal.org
these patients, the restoration and maintenance of normal sinus rhythm may be an important strategy for symptom control. In any case, when we elect to use a rate control and anticoagulation strategy, it is important to document (with ambulatory ECG recordings) that rate control remains satisfactory during everyday activities. Considering the results of these studies,9 –11,19 we consider rate control in AF to be an acceptable primary therapy. The two strategies seem equal in terms of mortality rate as well as thromboembolic and hemorrhagic risk. A rhythm control strategy without long-term anticoagulation therapy might result in a higher stroke rate, which implies that anticoagulation therapy should be extended after cardioversion, even in patients with a stable sinus rhythm. Our study has several limitations, one of which is the relatively small number of patients. This was due to the restrictive selection criteria that we applied in order to achieve a homogenous group of patients. This could have affected our results and prevented the detection of more differences between the two treatment strategies. Also, the number of our patients receiving anticoagulation therapy after sinus rhythm restoration and maintenance was relatively low. Yet, only a single ischemic stroke, unrelated to cardioversion, occurred in this subgroup of patients. Thus, rhythm control may still be a valid option for patients with persistent AF and relative contraindications to long-term anticoagulation therapy. Despite the limited number of patients, our findings are similar to those reported in other trials.9 –11,19 The HOT CAFE trial, being a randomized trial with well-defined end points and an almost 2-year follow-up period, may be a useful element in a future meta-analysis. We hope that our results will contribute to the discussion on rate control vs rhythm control to the benefit of patients with AF.
Conclusions The HOT CAFE study suggests that the rate control of persistent AF is equivalent to rhythm control in terms of mortality rate, stroke, and thromboembolic and severe bleeding complications. Therefore, our trial supports the results of other studies, although our population was younger and possibly had AF relatively longer than other populations (for up to 2 years). The HOT CAFE trial, like the PIAF, AFFIRM, RACE, and other trials, showed no significant differences in the two proposed treatment strategies for AF, so we still have to choose the best therapeutic approach based on an individual patient’s needs.9 –11,19 A meta-analysis of all randomCHEST / 126 / 2 / AUGUST, 2004
485
ized trials might offer the possibility of subgroup analysis and might help to optimize treatment procedures in everyday clinical practice. References 1 Pritchet ELC. Management of atrial fibrillation. N Engl J Med 1992; 326:1264 –1271 2 Stafford RS, Robson DC, Misra B, et al. Rate control and sinus rhythm maintenance in atrial fibrillation. Arch Intern Med 1998; 158:2144 –2148 3 Stanton MS, Prystowsky EN, Fineberg NS, et al. Arrhythmogenic drugs effects: a study of 506 patients treated for ventricular tachycardia or fibrillation. J Am Coll Cardiol 1989; 14:200 –215 4 Flaker GC, Blackshear JL, McBride R, et al. Antiarrhythmic drug therapy and cardiac mortality in atrial fibrillation. J Am Coll Cardiol 1992; 20:527–532 5 Friedman LP, Stevenson WG. Proarrhythmia. Am J Cardiol 1998; 82:50N–58N 6 Coplen SE, Antman EM, Berlin JA, et al. Efficacy and safety of quinidine therapy for maintenance of sinus rhythm after cardioversion: a meta-analysis of randomized control trials. Circulation 1990; 82:1106 –1116 7 Crijns HJGM, Van Gelder IC, Van Glist WH, et al. Serial antiarrhythmic drug treatment to maintain sinus rhythm after electrical cardioversion for persistent atrial fibrillation or flutter. Am J Cardiol 1991; 68:335–341 8 Grogan M, Smith HC, Gersh BJ, et al. Left ventricular dysfunction due to atrial fibrillation in patients initially believed to have idiopathic dilated cardiomyopathy. Eur Heart J 1992; 13:1290 –1295 9 Wyse DG, Waldo AL, DiMarco JP, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002; 347:1825–1833 10 Hohnloser SH, Kuck KH, Lilienthal J, et al. Rhythm or rate control in atrial fibrillation: Pharmacological Intervention In Atrial Fibrillation (PIAF); a randomized trial. Lancet 2000; 356:1789 –1794
486
11 Van Gelder J, Hagens VE, Bosker HA, et al. A comparison of rate control and rhythm control in patients with recurrent persistent atrial fibrillation. N Engl J Med 2002; 347:1834 – 1840 12 Opolski G, Torbicki A, Kosior D, et al. Should sinus rhythm be restored in patients with chronic atrial fibrillation? Preliminary results from the Polish “Hot Cafe” study. Pol Arch Med Wewn 1999; 5:413– 418 13 Laupacis A, Albers G, Dalen J, et al. Antithrombotic therapy in atrial fibrillation. Chest 1995; 108:352–359 14 Reimold SC, Chalmers TC, Berlin JA, et al. Assessment of the efficacy and safety of antiarrhythmic therapy for persistent atrial fibrillation: observation on the role of trial design and implications of drug-related mortality. Am Heart J 1992; 124:924 –927 15 Dreifus LS, Fisch C, Griffin JC, et al. Guidelines for implantation of cardiac pacemakers and antiarrhythmic devices: a report of the American College of Cardiology/American Heart Association task Force of Diagnostic and Therapeutic Cardiovascular Procedures (Committee on Pacemaker Implantation). Circulation 1991; 84:455– 467 16 Schiller NB, Shah PM, Crawford M, et al. Recommendations for quantification of left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantification of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989; 6:358 –367 17 Gosselink ATM, Crijns HJGM, van den Berg MP, et al. Functional capacity before and after cardioversion of atrial fibrillation: a controlled study. Br Heart J 1994; 72:616 – 621 18 Van Gelder IC, Crijns HJGM, Hillege HI, et al. Electrical cardioversion strategy does not reduce risk of thromboembolism in atrial fibrillation [abstract]. Eur Heart J 1995; 16(suppl):282 19 Carlsson J, Miketic S, Windeler J, et al. Randomized trial of rate-control versus rhythm control in persistent atrial fibrillation: The Strategies of Treatment of Atrial Fibrillation (STAF) study. J Am Coll Cardiol 2003; 41:1690 –1696
Clinical Investigations