Atrialfibrilation is a serious condition affecingmiionseofpeope worldwide. In fact, aeial fibrillation is the most common chronic tachyarrhythmi accountingfor 10% ofpatients who are admitted with circulatory probklms. Furthermore, tbis arrkytmia is a very 'exensive propossition' both in trms -of patients with recurrence of perstent increased frequency and duration of atrial fibriUlation following eectical hospitalsation as-well as in the peonal cardiovermsms. cost to patient in reduced quality of left. W. van Gist and CA. Visr, DrsHagens and Van Gelder tackkd Directors ofICIN. a clinicaly important qestion, wer repeated ekctrial cardioversien, in connt om junction witb antiarrtfihmic drugs, to Rt or maintain sinus rkythm is mandat. In Atrial fibrilltion (AF) ai&cts 0.5 to 1.0% other words, is rate control not infteior to of the geral population. The prevarhthm control inpatients with pertnt lcnce increaseswith age, reacing neadry i.e. non.selflimitngatrialfibrillation. Tbeir study, comprising thirty-one 10% of indiniduals ovcr the age of 80 years Despite this eno centres in the Nethrlands demonstrated population that rate control is-not in or to rym with AF the optima teatmet strttegy contrlfor the prvention of death and remains uncertain. The first- choice of morbidity from cardiovascular causes. therapy is the rhythm-control strategy Hence, rate control is appropriate in with restoration of sinus rhythm.2 A
severe drawback to this approach is the success rate for maintenance of sinus rhythm. Outcome will improve with the use of antiarrhythmic drugs
low
aftcr electrical cardioversion, but this unfortunately exposes the patient to the rsks of lif-threateng proarrhythmia. The second alternative, a rate-control strategy, is easy to achieve but it is not known whether this treatment strategy results in higher morbidity and mortaliy rates. Recently, scvcral randomised trials were published in which the issue of rate or rhythm control for atrial fibrillation was studied: the Dutch RACE study (RAte Control versus Electrical cardioversion for persistent atrial fibrillation), the North American AFFIRM study (Atrial Fibrillation Follow-up Investigation of Rhythm Management) and the smaller German PIAF (Pharmacological Intervention in Arial Filion) and STAF (Strategies of Treatment of Atrial Fibrillation) studiesA34
TIs 1. Randomised studies of rate- and rhythmcontrol straegies in atrial fibrillation. pm ehioht ho Rut Vs
(Yerw)
P R
522
2.3
10% vs 39%
Composite endpoinrt (cardiovascular death, heart failure, thromboembolism, bleeding, pacemaker implantation, severe advrse effects of drugs) 17.2% versus 22.6%
3.5
35% vs 63%
Alkause mortality 25.9% versus 26.7% (p=0.08)
nt
RACE
hS
(n)
Study
AFFIRM 4060
Fotwu
s
y
*
PIAF
252
1
10% vs 56%
Improvement of AF-related symptoms 61% versus 55% (p=0.317)
STAF
200
2
11% vs 26%
Composite endpoint (all-cause mortality, cerebrovascular events, thromboembolism, cardiopulmonary resuscitation) 5% versus 4.5% (p=0.99)
* Sinus
rhthm at the end of foUlow-up
146
146-NsddHm
JouulVclm
11,
Number 3, Mard
2003
The designs of these studies were essentially the same (table 1) and have been described elsewhere.t Rate control was performed with digoxin, verapamil or diltiazem and a n-blocker, alone or in combination. The target heart rate varied per study or was not specified. In the rhythm-control arms patients received prophylactic antiarrhythmic drugs and electrical cardioversion if necessary. Antithrombotic treatrnent consisted of oral anticoagulation or aspirin, depending on the patient's risk factors for stroke. Patients were allowed to stop anticoagulation when chronic sinus rhythm was obtained. Endpoints varied between studies (table 1). The main goal of each strategy is to reduce the risks assodated with atrial fibrillation, i.e. stroke, heart failure and syncope; at the same tine side effects of the intervention should be avoided as much as possible. Important side effects indude intracranial bleeding and drug proarrhythmia. Considering the above, it is dear that a morbidity and morality endpoint was chosen rather than an arrhythmia endpoint.
drugs and amiodarone was 39, 27 and 31, respectively. The other six patients withdrew before the end of the study while they were in sinus rhythm. Ten percent (n= 26) of the patients in the rate-control group were in sinus rhythm at the end of the follow-up. Results of the R4CE study Spontaneous conversion to sinus rhythm The study design and rcsults have been occurred in 13 patients, and 13 patients publishedprevious.37 The 522 patients were in sinus rhythm after electrical included in RACE study represented a cardioversion which was indicated for typical population with persistent AF.8 AF-related symptoms. Mean age was 69 years, and most patients The primary endpoint occurred in 44 had an underlying diseas ofwhich hyper- ofthe 256 rate-control patients (17.2%) tension (49%) and coronary artery disease and in 60 of the 266 rhythm-control (27%) were most common. At baseline, patients (22.6%) (table 2), which inthere was a slight over-representation of dicated that rate control is not inferior hypertension in the rhythm-control to rhythm control. The components of group: 55% against 43% of the patients the primary endpoint were well balanced (p-0.007). Only 21% ofthe patients had between the two groups, except for AF without underlying heart disea. adverse effects of antiarrhythmic drugs After a mean follow-up of 2.3 years and pacemaker implantation, which were sinus rhythmwaspresentin 39% (n=103) more frequently observed under rhythm of the patients in the rhythm-control control. Post-hoc analysis revealed that group after a median of two electrical hypertension and female sex was associcardioversions (figure 1). The number of ated with more (nonfatal) endpoints in patients onsotalol, cass IC antarrhythmnic the rhythm-control group.
Rate control
100%
100% Sinus rhythm
Sinus rhythm
Atrial fibrillation
Atrial fibrillation
0%tJ
0% t 0.0
RhY conr
0.5
1.0 1.5 2.0 2.5 3.C Tim since madomleatlon (yean 1)
1.0 1.5 2.0 2.5 3.0 Tim lIwo raidemleatlon (years)
Figure 1. Heart rfrtm in fellow-up.
Nedtherlds Heat Journal, Volume 11, Numbers3, Mardi 2003
147
The AFFiRM study in short Iike RACE, AFFIRM was desgned to evaluate rate and rhythm control in patent with atial fibditon.4,9 Patient characteristics were comparable with those ofthe RACE study. After a mean follow-up of3.5 years, the primary endpoint, which was overall mortity, oc- AF and these risk factors should be cuffed in 310 of the 2027 gned to adeoyaelyatcauatedirsecativ of The out ofthe rate control. Even more endpoints were the heart's obsrved in the rhythm-control stegy: RACE, AFFIRM, PIAF and STAF are 356 of 2033 patients. Mortaity at five shown in tabk 1. yea amounted to 21.3% versus 23.8%, t ofAF which resulted in a trend towards sPer- Anticouleton in the t iority of the rate-contol strategy More thn halfofthe componentofhe (pu0.08). Thromboembolic complica- prisnazyendpointsin RACE were relted tions occurred frequently, predominant- to trmboembolic complication and ly after cessation oforal anticoagulion bkedings (table 2). Most of these ocn o d ratio cuncd atan INRoutuide the t c orwitheite (INK) at subtherapeutic evels. All range. Th numbrOfpatent receiving AFFIRM- patents had one or more risk oral anticogulat ion in RACE ranged factors for strokce and therefore the under rhytIn control from 228 (86%) authors also stat hat all patients with to 263 (99%) against 246 (96%) to 254
(99%) in the rate-control group. In the rhythm-control group it was allowed to stop oral anticogulaion fsinus rhythm was present for longer than onc month caud after cardiNovsion. This may have excess strke in the rhythm-control ann in RACE since thrombsrisk probably persists dcspite sinus rhythm.'0 Thoracic aortic atherosclerosis is also a wellrecognised risk factor for stroke in these paient which may havc contributed." In additon, asymptomatic episodes of AF may add to the continued stroke risk.'2 Also the AFFIRM investigators reported more thromboembolic complications under rhythm control.4 A meta-analysis of the AFFIRM, RACE, PIAF and STAF furither substantiated these observations: the incidence of an ischaemic strok was sigficany higher under rhythm control than under rate -control: 6.3% versus 4.7% (ps0.04).'3 In this respect, it is important to note that
Tab 2. Incidence of the prmary endpoint and its componentts*
N*nbe(%)
Pdmvy .n*holnts In the RACE
Total cardiovascular mortality - sudden death/nonsudden cardiovascular Heart failure
fatal/nonfatal Thromboembolic complications
-
- fatal/nonfatal
Bleeding - fatal/nonfatal
Sevre adverse effects of antiarrhythmic drugs fatal/nonfatal Implantation of a pacemaker - fat/nonfatal
-
* Some
Rsb coibd
(-=2n)
(n=266)
4 (17.2) 18 (7.0) 8/10 9 (3.5) 4/5 14 (5.5) 0/14 12 (4.7) 6/6 2 (0.8) 0/2 3 (1.2) 0/3
60 (22.6)
18 (6.8) 8/10 12 (4.5) i1/l 21 (7.9) 6/15 9 (3.4) 3/6 12 (4.5) 0/12 8 (3.0) 0/8
petet lid more thn one endpoint.
148
148NalandHa rt Jounal, Veunm 11, Number 3, March 2003
iWsh2S.R''
O N,
all of the 35 patients with a thromboembolic complication in RACE had one or more risk factors for stroke. Thcse risk factors were age above 65 years, hypertension, diabtes, atrial enlargement, left entricular dysfunction and a previous thromboembolic event. In AFFIRM, patient were induded only if they had one or more risk factors for stroke. It must be noted that among the wellknown risk factors for stroke the rhythm is not included.10 Therefore, the main lesson leamed from the randomised studies is that antcoaglaton must be riskctors are prent continued if even ifpatients mat snus rhythm.
"
*E---' 1 1ri- i,
Gelder IC van, Hagens VE, Bosker HA, Kingma JH, Kap 0, Kingma T, et al A Comparison ofRate Control and Rhythm Control in Padients with Rccurrent Persitent Atrial Fibrillation. NEvngJMed2002;347:1834-40. 4 Wyse DG, Waldo AL, DiMarco JP, Domanlki MJ, Rosnberg Y, Schron EB, et al. A compadsonofrate control and rhythnicontrol inpadents with attial fibrillation. N Enji J Med 2002;347:1825-33. 5 Hohnloser SH, Kuck KR, Lilienthal J. Rhythm or rate contrd in atial fibrillation - Pharmacological Intrvention inAiri Fibriaion (PIAF): a randomised trial. Lancet2000;356:1789-94. 6 Carlson J, Tebbe U. Rhythm control versus rate contrl im atial fibrijlation: results from the 3
safer and more effective methods of mantaining sius rhythm are needed to reduce morbidity related to palpitations and AF-induced heart failure. U
The RACE tWiswas supported by grants from the Interuniversity Cardiology Institute, the Ccntre for Health Care In ce and 3M Pharma, the Nether-
lands. Clinicl implications First of all, as mentioned above, the VE. Hagens. randomised studies show that in the L C. van Gelder. presence of stroke risk factors, antiUniversity Hosital Groningen. coagulaon cannot be stopped even if chronic sinus rhythm can be mainid. J.G.P. Tijssen. Therefore the bottom line here is that Academic Medical Centre Amstedam. HA BoeAr. cardiologs can no longer sell the cadioversion to their patients using the Rijnstate Hospital A rnem. J.H. Kiugma. argument that anticoagulation can be stopped after a successl shock. SecondSt. Antonius Hosita Nieuwegein. 0. Kamp. ly, the RACE, AFFIRM, PIAF and Free Univerity Medical Centre STAF studies demonstrate that a rateHosita Amstrdam. control strategy is an acceptable alternative to rhythm control i paients with HJ.JG.M. Cnins,for the RACE recurrent AF. investiaos. However, the results ofthese studies University Hosital Mastritc do not make rhythm-control therapy reundant. Patients first presenting with AF should still get a chance to maintain I FeinbergWM, BlackshcarJL, LaupacisA, Kronmal R, Hart KG. Prevalence, age distribution, sinus rhythm in the long term. In a and gender of patients with atrial fibrilation. significnt number, sinus rhythm may Analysi and implications. Arch Intern Med appr asible and beneficial interms of 1995;155:469-73. reducing palpitations or dyspnoea. In 2 GelderICvan, Crins J, Tieleman KG, Brugemann J, KIamPJ de, Golink AT, et al. Chrnic additon, in patients who are severely atria fibriltion. Success ofserial cardioversion symptomatic ith AF continued rhythm therapy and safetyof oral anicoaguation. Arcb Intern Med 1996;156:2585-92. controlis unavoidable. For these patent
Ngshuhl_w Heart ournai Voume 11, Number 3, Mar 2003
STAFlotstudy(strategiesofteatmentofatial 7
8
9 10
11
12
fibrillaion).Pacis Clin Ekcnwp4rol2001;24 (part II):561. Gelder IC van, Hagens VE, Kingmna JH, Bosker HA, Kamp 0, Veeger NJGM, et al. Rate control versus electrical cardioversion fir atrial fibrillation. A rdomised compason of two ning morbidity, mortreatmentstrategies taiity, qualityoflife and cost-benefit-The RACE study deign. NethHartJ2002;10:118-24. LevyS, MarekM, Coumel P, Guize L, Leldeffe J, Medvcdowsky JL, et al. Characteration ofdifmntsbsbt ofaial fibriiladon in general practice in France: the iALFA study. The College of French Cardiologt. Circ tion 1999; 99:3028-35. Basline charcterscs of patients with atrial fibrllation: the AFFIRM Study. Am HeartJ 2002;143:991-1001. Hart KG, Pearce LA, Rothbart RM, McAnulty JH,AuingerRW, Halpein JL. Strokewithintermitnt ar fibrition: ncidence and predictrs during wirin therapy. Stroke Prevention in A"il Fibillai Invest orJAm Coil CardiuL2OW; 35:183-7. T ecpocaidiogphic correlates of thrmboembolmn in high-ri patients with nonvaivular atial fibriliadon. Ihe Stroke PreventioninAta ibiltion Investigators Committee on Echocardiogrhy. Ann Intern Med 1998;128:639-47. Falk RH. Maagement of atrial fibrillation radial refonn or modet modification? NEngI
JMed2002;347:1883-4-
13 VerhcugtFW, GelderICvan,Wyse DG, Hohnloser SH, Crijns HJ. Stroke prevention by rhydun versus rate control in atial fibrillation: inght from the randomized studies. Circula-
tion 2002;106(Suppl II):546.
149
severe drawback to this approach is the success rate for maintenance of sinus rhythm. Outcome will improve with the use of antiarrhythmic drugs
low
aftcr electrical cardioversion, but this unfortunately exposes the patient to the rsks of lif-threateng proarrhythmia. The second alternative, a rate-control strategy, is easy to achieve but it is not known whether this treatment strategy results in higher morbidity and mortaliy rates. Recently, scvcral randomised trials were published in which the issue of rate or rhythm control for atrial fibrillation was studied: the Dutch RACE study (RAte Control versus Electrical cardioversion for persistent atrial fibrillation), the North American AFFIRM study (Atrial Fibrillation Follow-up Investigation of Rhythm Management) and the smaller German PIAF (Pharmacological Intervention in Arial Filion) and STAF (Strategies of Treatment of Atrial Fibrillation) studiesA34
TIs 1. Randomised studies of rate- and rhythmcontrol straegies in atrial fibrillation. pm ehioht ho Rut Vs
(Yerw)
P R
522
2.3
10% vs 39%
Composite endpoinrt (cardiovascular death, heart failure, thromboembolism, bleeding, pacemaker implantation, severe advrse effects of drugs) 17.2% versus 22.6%
3.5
35% vs 63%
Alkause mortality 25.9% versus 26.7% (p=0.08)
nt
RACE
hS
(n)
Study
AFFIRM 4060
Fotwu
s
y
*
PIAF
252
1
10% vs 56%
Improvement of AF-related symptoms 61% versus 55% (p=0.317)
STAF
200
2
11% vs 26%
Composite endpoint (all-cause mortality, cerebrovascular events, thromboembolism, cardiopulmonary resuscitation) 5% versus 4.5% (p=0.99)
* Sinus
rhthm at the end of foUlow-up
146
146-NsddHm
JouulVclm
11,
Number 3, Mard
2003
The designs of these studies were essentially the same (table 1) and have been described elsewhere.t Rate control was performed with digoxin, verapamil or diltiazem and a n-blocker, alone or in combination. The target heart rate varied per study or was not specified. In the rhythm-control arms patients received prophylactic antiarrhythmic drugs and electrical cardioversion if necessary. Antithrombotic treatrnent consisted of oral anticoagulation or aspirin, depending on the patient's risk factors for stroke. Patients were allowed to stop anticoagulation when chronic sinus rhythm was obtained. Endpoints varied between studies (table 1). The main goal of each strategy is to reduce the risks assodated with atrial fibrillation, i.e. stroke, heart failure and syncope; at the same tine side effects of the intervention should be avoided as much as possible. Important side effects indude intracranial bleeding and drug proarrhythmia. Considering the above, it is dear that a morbidity and morality endpoint was chosen rather than an arrhythmia endpoint.
drugs and amiodarone was 39, 27 and 31, respectively. The other six patients withdrew before the end of the study while they were in sinus rhythm. Ten percent (n= 26) of the patients in the rate-control group were in sinus rhythm at the end of the follow-up. Results of the R4CE study Spontaneous conversion to sinus rhythm The study design and rcsults have been occurred in 13 patients, and 13 patients publishedprevious.37 The 522 patients were in sinus rhythm after electrical included in RACE study represented a cardioversion which was indicated for typical population with persistent AF.8 AF-related symptoms. Mean age was 69 years, and most patients The primary endpoint occurred in 44 had an underlying diseas ofwhich hyper- ofthe 256 rate-control patients (17.2%) tension (49%) and coronary artery disease and in 60 of the 266 rhythm-control (27%) were most common. At baseline, patients (22.6%) (table 2), which inthere was a slight over-representation of dicated that rate control is not inferior hypertension in the rhythm-control to rhythm control. The components of group: 55% against 43% of the patients the primary endpoint were well balanced (p-0.007). Only 21% ofthe patients had between the two groups, except for AF without underlying heart disea. adverse effects of antiarrhythmic drugs After a mean follow-up of 2.3 years and pacemaker implantation, which were sinus rhythmwaspresentin 39% (n=103) more frequently observed under rhythm of the patients in the rhythm-control control. Post-hoc analysis revealed that group after a median of two electrical hypertension and female sex was associcardioversions (figure 1). The number of ated with more (nonfatal) endpoints in patients onsotalol, cass IC antarrhythmnic the rhythm-control group.
Rate control
100%
100% Sinus rhythm
Sinus rhythm
Atrial fibrillation
Atrial fibrillation
0%tJ
0% t 0.0
RhY conr
0.5
1.0 1.5 2.0 2.5 3.C Tim since madomleatlon (yean 1)
1.0 1.5 2.0 2.5 3.0 Tim lIwo raidemleatlon (years)
Figure 1. Heart rfrtm in fellow-up.
Nedtherlds Heat Journal, Volume 11, Numbers3, Mardi 2003
147
The AFFiRM study in short Iike RACE, AFFIRM was desgned to evaluate rate and rhythm control in patent with atial fibditon.4,9 Patient characteristics were comparable with those ofthe RACE study. After a mean follow-up of3.5 years, the primary endpoint, which was overall mortity, oc- AF and these risk factors should be cuffed in 310 of the 2027 gned to adeoyaelyatcauatedirsecativ of The out ofthe rate control. Even more endpoints were the heart's obsrved in the rhythm-control stegy: RACE, AFFIRM, PIAF and STAF are 356 of 2033 patients. Mortaity at five shown in tabk 1. yea amounted to 21.3% versus 23.8%, t ofAF which resulted in a trend towards sPer- Anticouleton in the t iority of the rate-contol strategy More thn halfofthe componentofhe (pu0.08). Thromboembolic complica- prisnazyendpointsin RACE were relted tions occurred frequently, predominant- to trmboembolic complication and ly after cessation oforal anticoagulion bkedings (table 2). Most of these ocn o d ratio cuncd atan INRoutuide the t c orwitheite (INK) at subtherapeutic evels. All range. Th numbrOfpatent receiving AFFIRM- patents had one or more risk oral anticogulat ion in RACE ranged factors for strokce and therefore the under rhytIn control from 228 (86%) authors also stat hat all patients with to 263 (99%) against 246 (96%) to 254
(99%) in the rate-control group. In the rhythm-control group it was allowed to stop oral anticogulaion fsinus rhythm was present for longer than onc month caud after cardiNovsion. This may have excess strke in the rhythm-control ann in RACE since thrombsrisk probably persists dcspite sinus rhythm.'0 Thoracic aortic atherosclerosis is also a wellrecognised risk factor for stroke in these paient which may havc contributed." In additon, asymptomatic episodes of AF may add to the continued stroke risk.'2 Also the AFFIRM investigators reported more thromboembolic complications under rhythm control.4 A meta-analysis of the AFFIRM, RACE, PIAF and STAF furither substantiated these observations: the incidence of an ischaemic strok was sigficany higher under rhythm control than under rate -control: 6.3% versus 4.7% (ps0.04).'3 In this respect, it is important to note that
Tab 2. Incidence of the prmary endpoint and its componentts*
N*nbe(%)
Pdmvy .n*holnts In the RACE
Total cardiovascular mortality - sudden death/nonsudden cardiovascular Heart failure
fatal/nonfatal Thromboembolic complications
-
- fatal/nonfatal
Bleeding - fatal/nonfatal
Sevre adverse effects of antiarrhythmic drugs fatal/nonfatal Implantation of a pacemaker - fat/nonfatal
-
* Some
Rsb coibd
(-=2n)
(n=266)
4 (17.2) 18 (7.0) 8/10 9 (3.5) 4/5 14 (5.5) 0/14 12 (4.7) 6/6 2 (0.8) 0/2 3 (1.2) 0/3
60 (22.6)
18 (6.8) 8/10 12 (4.5) i1/l 21 (7.9) 6/15 9 (3.4) 3/6 12 (4.5) 0/12 8 (3.0) 0/8
petet lid more thn one endpoint.
148
148NalandHa rt Jounal, Veunm 11, Number 3, March 2003
iWsh2S.R''
O N,
all of the 35 patients with a thromboembolic complication in RACE had one or more risk factors for stroke. Thcse risk factors were age above 65 years, hypertension, diabtes, atrial enlargement, left entricular dysfunction and a previous thromboembolic event. In AFFIRM, patient were induded only if they had one or more risk factors for stroke. It must be noted that among the wellknown risk factors for stroke the rhythm is not included.10 Therefore, the main lesson leamed from the randomised studies is that antcoaglaton must be riskctors are prent continued if even ifpatients mat snus rhythm.
"
*E---' 1 1ri- i,
Gelder IC van, Hagens VE, Bosker HA, Kingma JH, Kap 0, Kingma T, et al A Comparison ofRate Control and Rhythm Control in Padients with Rccurrent Persitent Atrial Fibrillation. NEvngJMed2002;347:1834-40. 4 Wyse DG, Waldo AL, DiMarco JP, Domanlki MJ, Rosnberg Y, Schron EB, et al. A compadsonofrate control and rhythnicontrol inpadents with attial fibrillation. N Enji J Med 2002;347:1825-33. 5 Hohnloser SH, Kuck KR, Lilienthal J. Rhythm or rate contrd in atial fibrillation - Pharmacological Intrvention inAiri Fibriaion (PIAF): a randomised trial. Lancet2000;356:1789-94. 6 Carlson J, Tebbe U. Rhythm control versus rate contrl im atial fibrijlation: results from the 3
safer and more effective methods of mantaining sius rhythm are needed to reduce morbidity related to palpitations and AF-induced heart failure. U
The RACE tWiswas supported by grants from the Interuniversity Cardiology Institute, the Ccntre for Health Care In ce and 3M Pharma, the Nether-
lands. Clinicl implications First of all, as mentioned above, the VE. Hagens. randomised studies show that in the L C. van Gelder. presence of stroke risk factors, antiUniversity Hosital Groningen. coagulaon cannot be stopped even if chronic sinus rhythm can be mainid. J.G.P. Tijssen. Therefore the bottom line here is that Academic Medical Centre Amstedam. HA BoeAr. cardiologs can no longer sell the cadioversion to their patients using the Rijnstate Hospital A rnem. J.H. Kiugma. argument that anticoagulation can be stopped after a successl shock. SecondSt. Antonius Hosita Nieuwegein. 0. Kamp. ly, the RACE, AFFIRM, PIAF and Free Univerity Medical Centre STAF studies demonstrate that a rateHosita Amstrdam. control strategy is an acceptable alternative to rhythm control i paients with HJ.JG.M. Cnins,for the RACE recurrent AF. investiaos. However, the results ofthese studies University Hosital Mastritc do not make rhythm-control therapy reundant. Patients first presenting with AF should still get a chance to maintain I FeinbergWM, BlackshcarJL, LaupacisA, Kronmal R, Hart KG. Prevalence, age distribution, sinus rhythm in the long term. In a and gender of patients with atrial fibrilation. significnt number, sinus rhythm may Analysi and implications. Arch Intern Med appr asible and beneficial interms of 1995;155:469-73. reducing palpitations or dyspnoea. In 2 GelderICvan, Crins J, Tieleman KG, Brugemann J, KIamPJ de, Golink AT, et al. Chrnic additon, in patients who are severely atria fibriltion. Success ofserial cardioversion symptomatic ith AF continued rhythm therapy and safetyof oral anicoaguation. Arcb Intern Med 1996;156:2585-92. controlis unavoidable. For these patent
Ngshuhl_w Heart ournai Voume 11, Number 3, Mar 2003
STAFlotstudy(strategiesofteatmentofatial 7
8
9 10
11
12
fibrillaion).Pacis Clin Ekcnwp4rol2001;24 (part II):561. Gelder IC van, Hagens VE, Kingmna JH, Bosker HA, Kamp 0, Veeger NJGM, et al. Rate control versus electrical cardioversion fir atrial fibrillation. A rdomised compason of two ning morbidity, mortreatmentstrategies taiity, qualityoflife and cost-benefit-The RACE study deign. NethHartJ2002;10:118-24. LevyS, MarekM, Coumel P, Guize L, Leldeffe J, Medvcdowsky JL, et al. Characteration ofdifmntsbsbt ofaial fibriiladon in general practice in France: the iALFA study. The College of French Cardiologt. Circ tion 1999; 99:3028-35. Basline charcterscs of patients with atrial fibrllation: the AFFIRM Study. Am HeartJ 2002;143:991-1001. Hart KG, Pearce LA, Rothbart RM, McAnulty JH,AuingerRW, Halpein JL. Strokewithintermitnt ar fibrition: ncidence and predictrs during wirin therapy. Stroke Prevention in A"il Fibillai Invest orJAm Coil CardiuL2OW; 35:183-7. T ecpocaidiogphic correlates of thrmboembolmn in high-ri patients with nonvaivular atial fibriliadon. Ihe Stroke PreventioninAta ibiltion Investigators Committee on Echocardiogrhy. Ann Intern Med 1998;128:639-47. Falk RH. Maagement of atrial fibrillation radial refonn or modet modification? NEngI
JMed2002;347:1883-4-
13 VerhcugtFW, GelderICvan,Wyse DG, Hohnloser SH, Crijns HJ. Stroke prevention by rhydun versus rate control in atial fibrillation: inght from the randomized studies. Circula-
tion 2002;106(Suppl II):546.
149
