International Journal of Cardiology 161 (2012) 160–165

Contents lists available at SciVerse ScienceDirect

International Journal of Cardiology

j ourna l homepage: www.e lsev ie r .com/ locate / i j ca rd

Sinus rhythm versus atrial fibrillation in elderly patients with chronic heartfailure — Insight from the Cardiac Insufficiency Bisoprolol Study in Elderly

Ivan Stankovic a,⁎, Aleksandar N. Neskovic a, Biljana Putnikovic a, Svetlana Apostolovic b, Mitja Lainscak c,d,Frank Edelmann e, Wolfram Doehner f, Goetz Gelbrich g,h, Simone Inkrot i, Thomas Rau j,Christoph Herrmann-Lingen k, Stefan D. Anker d,l, Hans-Dirk Düngen i

a Department of Cardiology, Clinical Hospital Center Zemun, Faculty of Medicine, University of Belgrade, Belgrade, Serbiab University Clinical Center Nis, Department of Cardiology, Nis, Serbiac University Clinic of Respiratory and Allergic Diseases, Department of Cardiology, Golnik, Sloveniad Charité-Universitätsmedizin, Campus Virchow-Klinkum, Department of Cardiology, Applied Cachexia Research, Berlin, Germanye Göttingen University Medical Center, Department of Cardiology and Pneumology, Göttingen, Germanyf Charité-Universitätsmedizin, Campus Virchow-Klinkum, Center for Stroke Research, Berlin, Germanyg University of Leipzig, Clinical Trial Center, Leipzig, Germanyh Institute of Clinical Epidemiology and Biometry, Julius Maximilian University of Würzburg, Würzburg, Germanyi Charité-Universitätsmedizin, Campus Virchow-Klinikum, Department of Internal Medicine — Cardiology, Berlin, Germanyj University Medical Center Hamburg Eppendorf, Institute for Experimental and Clinical Pharmacology, Hamburg, Germanyk Göttingen University Medical Center, Department of Psychosomatic Medicine and Psychotherapy, Göttingen, Germanyl Centre for Clinical and Basic Research, IRCCS San Raffaele, Rome, Italy

Abbreviations: AF, atrial fibrillation; bpm, beats pInsufficiency Bisoprolol Study in Elderly; HF, heart failventricular ejection fraction; NYHA, New York Heart ASR, sinus rhythm.⁎ Corresponding author at: Clinical Hospital Center Zem

Vukova 9, 11070 Belgrade, Serbia. Tel.: +381 62 403016;E-mail address: future.ivan@gmail.com (I. Stankovic

0167-5273/$ – see front matter © 2012 Elsevier Irelanddoi:10.1016/j.ijcard.2012.06.004

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 12 January 2012Received in revised form 30 May 2012Accepted 7 June 2012Available online 21 June 2012

Keywords:Atrial fibrillationBeta-blockersChronic heart failure

Background: It has been suggested that patients with chronic HF and atrial fibrillation (AF) may respond dif-ferently to beta-blockers than those in sinus rhythm (SR).Methods: In this predefined analysis of the CIBIS-ELD trial, a total of 876 chronic HF patients (164 patients withAF) were randomized to bisoprolol or carvedilol. During the 12-week-treatment phase, beta-blockers were dou-bled fortnightly up to the target dose or maximally tolerated dose, which was maintained for 4 weeks.Results: Patients with AF had lower left ventricular ejection fraction (LVEF), exercise capacity, self-rated health,quality of life (QoL) scores for both SF36 physical and psychosocial component, and higher NYHA class thanthose in SR. Beta-blocker titration was associated with clinical improvement in both AF and SR patients:LVEF, 6-minute walk distance, physical and psychosocial components of QoL scores, self-rated health and

NYHA class (pb0.05, for all). The extent of improvement did not differ between patients with AF and in SRand did not differ between bisoprolol and carvedilol. Heart rate (HR) at baseline was higher in the AF group,and remained higher until the end of the trial. Patients with higher baseline HR had larger reductions in HR,regardless of rhythm. AF patients more frequently reached target beta-blocker dose compared to those in SR(pb0.005).Conclusions: Elderly patients with chronic HF and AF derive comparable clinical benefits from beta-blocker titrationas those in SR. Patients with AF tolerate higher beta-blocker doses than those in SR, which appears to be related tohigher baseline HR.

© 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Atrial fibrillation (AF) is the most common rhythm disorder in pa-tients with chronic heart failure (HF), affecting 15–30% of cases, and its

er minute; CIBIS-ELD, Cardiacure; HR, heart rate; LVEF, leftssociation; QoL, quality of life;

un, Department of Cardiology,fax: +381 11 3168878.).

Ltd. All rights reserved.

proportion increases with age and New York Heart Association (NYHA)functional class [1–3]. Despite the well established beneficial role ofbeta-blockers in the treatment of both AF and chronic HF, it has beensuggested that patientswith chronic HF and coexisting AFmay not equal-ly benefit from beta-blocker therapy as those in sinus rhythm (SR) [4].

While limited data from previous studies suggested comparableimprovement in left ventricular (LV) systolic function with beta-blocker treatment in patients with coexisting chronic HF and AF,exercise capacity and symptoms did not improve to the same extentas in those in SR [5,6].

Further, it is unknown whether different pharmacological charac-teristics of beta-blockers may influence their clinical effects in

Table 1Baseline characteristics.

Atrial fibrillation(n=164)

Sinus rhythm(n=712)

p Value

Woman, no. (%) 39 (23.8) 290 (40.7) b0.001Age, yrs 72.1±5.3 72.4±5.6 0.126Heart rate, bpm 86±17 71±12 b0.001Systolic blood pressure, mm Hg 133±18 138±22 0.003Diastolic blood pressure, mm Hg 82±12 80±12 0.036LVEF, % 37±11 43±14 b0.001Preserved LVEF (>45%), no. (%) 25 (15.2) 225 (31.6) b0.001

NYHA class, no. (%) 0.016I 3 (1.8) 31 (4.4)II 96 (58.5) 479 (67.3)III 64 (39.0) 194 (27.2)IV 1 (0.6) 8 (1.1)

6MWD, m 292±106 329±110 b0.001NTproBNP, median (IQR), pg/ml 1540 (924–2747) 503 (207–1433) b0.001Hemoglobin, mean (SD), g/dl 13.6±1.6 13.9±1.7 0.299a

FEV1, ml 2218±661 2186±663 0.651

Co-morbidities, no. (%)Hypertension 128 (78) 596 (84) 0.087Diabetes mellitus 40 (24) 183 (26) 0.766Hyperlipidemia 83 (56) 465 (65) b0.001Renal dysfunction [eGFRb60 ml/min/1.73 m2]

58 (35) 270 (38) 0.592

Cardiovascular medication, no. (%)Beta-blocker (pretreatment) 97 (59) 430 (60) 0.791ACE inhibitor and/or ARB 136 (83) 605 (85) 0.549Aldosterone receptor antagonist 73 (45) 202 (28) b0.001Diuretic 138 (84) 511 (72) 0.001Cardiac glycoside 72 (44) 57 (8) b0.001Antiarrhythmic 28 (17) 67 (9) 0.008Statin 43 (26) 299 (42) b0.001Antiplatelet 86 (52) 495 (70) b0.001Anticoagulant 107 (65) 112 (16) b0.001

6MWD = 6-minute walk distance. ACE = angiotensin converting enzyme. ARB =angiotensin receptor blocker. eGFR = estimated glomerular filtration rate. FEV1 =forced expiratory volume in first second. LVEF = left ventricular ejection fraction.NTproBNP = N-terminal pro b-type natriuretic peptide.

a Adjusted for sex.

161I. Stankovic et al. / International Journal of Cardiology 161 (2012) 160–165

patients with chronic HF and AF. Previous studies suggested thatcarvedilol's nonselectivity and additional (vasodilatatory) propertiesmay be important in patients with chronic HF [3,7]. On the otherhand, it has been shown that beneficial effects of beta-blockers are re-lated to the magnitude of heart rate reduction [8], which favors theselective beta1-adrenoceptor-blocker bisoprolol [9].

Therefore, we prospectively analyzed data from The Cardiac Insuffi-ciency Bisoprolol Study in Elderly (CIBIS-ELD) [9] to compare theresponse to beta-blocker titration in elderly chronic HF patients with AFversus those in SR. This study is the first head-to-head comparison of car-vedilol and bisoprolol in patients with chronic HF and coexisting AF.

2. Methods

The CIBIS-ELD trial was an investigator-initiated, multicenter, double-blind trial,where 876 patients ≥65 years with systolic or diastolic chronic HF (NYHA class II/III)were randomized to receive bisoprolol or carvedilol. Eligible patients had to be beta-blocker naïve or on ≤25% of the recommended target dose. Study design and mainresults have been published previously [9,10]. Briefly, the beta-blocker dose was dou-bled fortnightly to reach the target dose (10 mg bisoprolol once daily or 25 mg carve-dilol twice daily within 6 weeks) or maximally tolerated dose. Patients weighing>85 kg were scheduled to reach the target dose of carvedilol of 50 mg twice dailywithin 8 weeks. Investigators were free to delay titration or reduce the dose if clinicallyindicated. The titration phase was followed by a maintenance period lasting 4 weeksand the final visit was at 10 weeks (12 weeks for patients >85 kg).

The primary endpoint of tolerability of CIBIS-ELD was defined as reaching the tar-get dose through the process of fortnightly doubling with no more than one delayedincrease and with the target dose maintained for at least 10 days.

Secondary endpoints included adverse events and clinical parameters of patientstatus which were measured at baseline, prior to dose titration at each visit, and atfollow-up.

Patients with AF were analyzed as a predefined subgroup to determine potentialdifferences in reaching primary endpoint and to compare the effects of beta-blockertitration on exercise capacity (6-minute walk distance — 6MWD), symptoms (NYHAfunctional class), left ventricular ejection fraction (LVEF), parameters of diastolic func-tion and quality of life (the short-form quality of life health survey) to those in SR.Heart rate and blood pressure changes as well as relationship between heart ratechanges and achieved beta-blocker dose in both rhythm groups were also studied.

The effects of beta-blocker titration on additional echocardiographic parameters(LV end-diastolic diameter, LV mass index, left atrial volume index), N-terminal prob-type natriuretic peptide (NT-proBNP) plasma levels and self-rated health scorewere also compared between patients with AF versus those in SR.

2.1. Electrocardiographic recordings

As per CIBIS-ELD study protocol, an electrocardiogram (ECG) was recorded at base-line, fortnightly at each titration visit, as well as at follow up. The standard 12-lead ECGwas recorded after 2 to 3 min of rest in the supine position, at the paper speed of25 mm/s and standardization of 10 mm/1 mV. Baseline rhythm was used to classifypatients as belonging to the AF or the SR group. Fifty-six patients with pacemakers (8 inAF and 48 in SR group) were excluded from the analysis.

2.2. Measurement of left ventricular ejection fraction

Left ventricular ejection fraction was calculated using biplane Simpson's rule, asthe mean value of three consecutive cardiac cycles.

2.3. Statistical analysis

Differences in baseline characteristics between patients with AF and SR were ana-lyzed using Student t-test for continuous variables and Fisher's exact test for categori-cal variables. The difference of rates in reaching the maximum recommended beta-blocker dose between AF and SR group was tested by two-sided Fisher's exact test.

Logistic regressionwas used to examine the increase of the odds for reaching the tar-get dose associated with AF, baseline heart rate (odds ratio per additional beat/min) andboth AF and baseline heart rate combined.

The significance of changes in clinical endpoints was assessed within each groupby paired t-test. Comparison across groups was carried out by the analysis of covari-ance (ANCOVA) with the follow-up measurement as dependent variable, the heartrhythm (AF or SR) at baseline ECG as factor, and the baseline measurement as covariate(or as categorical co-factor in case of NYHA class). When significant differences be-tween groups were observed, additional comparison (bisoprolol vs. carvedilol) wascarried out by 2-way ANCOVA, with the type of beta-blocker as cofactor. A separateanalysis was done to test the interaction between observed differences and baselineLV systolic function (preserved vs. reduced LVEF). Analyses were performed usingSPSS Version 15 (SPSS Inc., Chicago, IL, USA).

The authors of this manuscript have certified that they comply with the Principlesof Ethical Publishing in the International Journal of Cardiology [11].

3. Results

Of 876 patients randomized in CIBIS-ELD, 164 (18.6%) had AF atbaseline. Demographic and clinical patients' characteristics are sum-marized in Table 1. Patients with AF were more frequently male,higher resting HR, lower systolic and higher diastolic blood pressurethan those in SR.

PatientswithAFhad lower LVEF, presentedwith higherNYHAclasses,higher NTproBNP levels and lower exercise capacity, as indicated by6MWD, than those in SR. Left ventricular end-diastolic dimension, LVmass index and left atrial volume index were all significantly larger inthe AF group. Patients with AF also had lower quality of life scores(both physical and psychosocial components), higher depression sumscores (Patient Health Questionnaire — PHQ) and more frequently ratedtheir health as poor. Ischemic cardiomyopathy was the most commonetiology of chronic HF in both groups (44% in AF and 54% in SR group).

More patients with AF were receiving diuretics, aldosterone recep-tor antagonists, anticoagulants, cardiac glycosides and anti-arrhythmicdrugs at baseline, while there was no difference in angiotensin conver-ting enzyme inhibitors/angiotensin receptor blockers and beta-blockerpre-treatment.

3.1. Heart rate and blood pressure changes

Patients with higher baseline heart rates had larger reductions inheart rate, regardless of rhythm (Fig. 1, negative slope of the solidlines). As baseline heart rates were higher in the AF group, the

Fig. 1. After adjusting changes in heart rate for baseline values, patients with AF hadfewer heart rate reduction, compared to those in SR with the same baseline heartrate (the regression line for AF group lying above the line for SR group). Patientswho had higher baseline heart rate had more reduction, illustrated by the negativeslope of the regression lines; patients with AF are overrepresented in the high baselineheart rate domain (blue–gray points ratio increases from the left to the right). AF =atrial fibrillation. ECG = electrocardiogram. SR = sinus rhythm.

162 I. Stankovic et al. / International Journal of Cardiology 161 (2012) 160–165

mean reduction of heart rate (not adjusted for baseline) was larger inthe AF group (−9.2 vs. −6.2 bpm, pb0.001; dashed lines in Fig. 1).

When comparing patients with the same baseline heart rates, pa-tients with AF had lesser mean heart rate reduction than those in SR(ANCOVA difference in Table 2, solid lines in Fig. 1). Accordingly,heart rate remained higher in patients with AF compared to thosein SR until the end of the trial (76 vs. 64 bpm; pb0.001).

Systolic blood pressure (SBP) remained lower while diastolicblood pressure (DBP) remained higher in patients with AF until theend of the study. The reductions of both SBP and DBP from baselineto follow-up were comparable in both groups and there was no sig-nificant correlation between blood pressure and heart rate changes(p=0.244).

3.2. Heart rhythm, heart rate and achieved beta-blocker dose

The patients with AF were more likely to reach maximum beta-blocker dose compared to those in SR (odds ratio (OR) 1.76, 95% con-fidence interval (CI): 1.23–2.53; p=0.002) (Fig. 2). Maximumguideline-recommended beta-blocker dose was reached by 42% ofpatients with AF vs. 29% of patients in SR (p=0.003).

Higher baseline heart rate was associated with greater chancesto reach beta-blocker target doses (OR 1.033 per 1 bpm increase,95% CI: 1.022–1.044).

When both AF and heart rate were adjusted for each other, base-line heart rate remained significant at an odds ratio of 1.031 pereach beat/min for reaching the beta-blocker target dose (pb0.001)independently of AF, while AF was associated with an odds ratio of1.11 for achieving the target dose (p=0.60) when controlling forheart rate. Therefore, it appears that reaching the beta-blocker targetdose in patients with AF more frequently than in patients in SR is me-diated predominantly by higher baseline heart rate.

There was no significant relationship between digoxin treatment(AF group 44%, SR group 8%) and beta-blocker dose achieved.

3.3. Changes in exercise capacity and echocardiographic parameters

After 12 weeks of beta-blocker titration, LVEF and exercise capac-ity (as indicated by 6MWD) significantly improved in both groups to

a similar degree (Table 2). There was no significant change from base-line to follow-up in parameters reflecting LV diastolic function(E wave peak velocity, E wave deceleration time, and E/Ea ratio) ineither group.

A significant reduction of LVmass index was observed in patients inSR, but not in those with AF. On the other hand, a trend towards LAvolume index reduction was observed in patients with AF (p=0.059),but not in those in SR.

3.4. Changes in symptoms

A highly significant and comparable reduction of symptoms wasobserved from baseline to follow-up in both rhythm groups. Impor-tantly, the significant improvement was observed both by physicians(as NYHA class assessment) and patients (as measured by quality oflife and self-rated health scores). However, when changes in self-rated health were adjusted for baseline scores, patients with AF im-proved less than those in SR with the same baseline self-rated healthscores (p=0.038, Table 2). In addition, the extent of the improve-ment in quality of life was equal in AF and SR groups.

3.5. The role of preserved vs. reduced left ventricular ejection fraction

We tested all the differences in clinical and echocardiographic pa-rameters for interaction with baseline LV ejection fraction and we didnot find any. Moreover, adjustment of the analyses for LV function didnot substantially alter any of the results. Nevertheless, when the anal-ysis was restricted to the subgroup with reduced LVEF, the signifi-cance between AF and SR with respect to change in LVEF wasreached (Table 2). In contrast, there were almost no changes inLVEF in the patients with preserved LVEF, and no difference betweenAF and SR in this subgroup (pb0.05).

3.6. Relationship between observed effects and type of beta-blocker

All variables for which significant changes from baseline were ob-served (heart rate, systolic and diastolic blood pressure, LVEF, NYHAclass, 6MWD, self-rated health, QoL SF-36 psychosocial and physicalcomponent), were additionally analyzed to determine whether thechanges can be attributable predominantly to one of the testedbeta-blockers. Table 3 shows changes of clinical parameters with re-gard to substance interaction. No interaction was found.

4. Discussion

Our data revealed that elderly chronic HF patients with AF derivedsimilar clinical benefits from beta-blocker therapy as patients in SR.After 12 weeks of beta-blocker titration, patients in both rhythmgroups demonstrated comparable improvement in LV systolic func-tion, exercise capacity, NYHA class and quality of life. Effect size wassimilar in both groups, although patients with AF were sicker at base-line and remained at poorer functional level at the end of follow-up.Importantly, no interaction effect related to the type of beta-blocker(bisoprolol or carvedilol) was found.

In the present study, 18.6% of patients were in AF. This is consistentwith other heart failure studies, where the prevalence of AF rangedfrom 15 to 30% [4–6]. In contrast to our sample, where AF was associ-ated with significantly lower LVEF (37±11% in AF vs. 43±14 in SRgroup), the level of systolic dysfunction in similar studies was compa-rable between SR and AF groups, with LVEF varying between 26 and28%.

4.1. Efficacy of beta-blockers in patients with CHF and AF

Several large scale clinical trials have proven that beta-blockertreatment in patients with chronic HF and SR reduces mortality and

Table 2Clinical endpoints.

Atrial fibrillation Sinus rhythm Difference AF-SR from ANCOVA

Heart rate, bpm −9.2 −6.2 +6.2Mean change (95%CI) (−12.0 to −6.4) (−7.2 to −5.3) (+4.1 to +8.3)p-Value b0.001 b0.001 b0.001

Systolic BP, mm Hg −7.9 −9.7 −1.3Mean change (95%CI) (−11.1 to −4.6) (−11.4 to −8.1) (−4.2 to +1.7)p-Value b0.001 b0.001 0.400

Diastolic BP, mm Hg −4.4 −4.4 +1.6Mean change (95%CI) (−6.3 to −2.4) (−5.4 to −3.5) (−0.1 to +3.3)p-Value b0.001 b0.001 0.062

NYHA class −0.28 −0.27 +0.06Mean change (95%CI) (−0.38 to −0.19) (−0.31 to −0.22) (−0.02 to +0.15)p-Value b0.001 b0.001 0.158

6MWD, m +14.0 +16.0 −7.0Mean change (95%CI) (+4 to +25) (+11 to +21) (−19 to +4)p-Value 0.007 b0.001 0.216

NTproBNP, pg/ml ×0.99 ×1.02 ×1.19Mean change (95%CI) (×0.85 to ×1.15) (×0.94 to ×1.1) (×1.01 to ×1.41)p-Value 0.863 0.681 0.042

Self rated health (1=excellent; 5=poor) −0.32 −0.33 +0.15Mean change (95%CI) (−0.46 to −0.18) (−0.39 to −0.26) (+0.01 to+0.28)p-Value b0.001 b0.001 0.038

SF-36 Physical component +2.3 +2.2 −0.8Mean change (95%CI) (+0.9 to +3.7) (+1.5 to +2.9) (−2.1 to +0.6)p-Value 0.002 b0.001 0.246

SF-36 Psychosocial component +4.2 +2.7 0Mean change (95%CI) (+2.3 to +6.1) (+1.9 to +3.6) (−1.7 to +1.7)p-Value b0.001 b0.001 0.990

Echocardiographic parametersLVEF, % +4.6 +2.4 +1.1

Mean change (95%CI) (+3.4 to +5.9) (+1.9 to +3.4) (−0.1 to +2.3)p-Value b0.001 b0.001 0.076

LVEF, % (patients with reduced LVEF only) +5.3 +3.4 +1.6Mean change (95%CI) (+4.0 to +6.6) (+2.8 to +4.0) (+0.3 to +2.9)p-Value b0.001 b0.001 0.018

LV EDD, mm −0.23 −0.31 +0.697Mean change (95%CI) (−0.92 to +0.46) (−0.69 to +0.76) (−0.001 to +1.49)p-Value 0.510 0.116 0.086

LA volume index, ml/m2 −1.9 +0.61 +0.55Mean change (95%CI) (−3.9 to 0.07) (−0.21 to +1.42) (−1.57 to +2.66)p-Value 0.059 0.146 0.611

LV mass index −1.9 −2.44 +2.48Mean change (95%CI) (−6.2 to +2.4) (−4.51 to −0.36) (−1.75 to +6.71)p-Value 0.390 0.021 0.251

E wave peak velocity, cm/s −0.62 +2.8 +0.92Mean change (95%CI) (−5.1 to +3.8) (−1.6 to +7.2) (−9.75 to +11.59)p-Value 0.783 0.216 0.866

E wave deceleration time, ms +6.5 +2.8 −16.9Mean change (95%CI) (−4.2 to +17.1) (−3.8 to +9.4) (−31.9 to −2.1)p-Value 0.231 0.402 0.026

E/Ea +0.57 −0.57 +1.24Mean change (95%CI) (−1.5 to +2.6) (−1.0 to +0.28) (−0.20 to +2.67)p-Value 0.063 0.586 0.092

6MWD = 6-minute walk distance. BP = blood pressure. LA = left atrium. LVEDD = left ventricular end-diastolic diameter. LVEF = left ventricular ejection fraction. NTproBNP =N-terminal pro b-type natriuretic peptide. NYHA = New York Heart Association.

163I. Stankovic et al. / International Journal of Cardiology 161 (2012) 160–165

symptom burden [12–15]. However, only few trials have dealt withthe efficacy of beta-blockers in chronic HF with coexisting AF.

In the only prospective, double-blind, placebo controlled study [6],the combination of carvedilol and digoxin was shown to improveventricular function and reduce symptoms, while it had no impacton exercise capacity. Of note, the study was significantly limited bya relatively small sample (47 patients) and was specifically designedto test the superiority of the combination of carvedilol and digoxinto either compound alone.

Furthermore, retrospective studies [5,7,16,17] have confirmed thebeta-blocker efficacy in the improvement of LV systolic function inboth rhythm groups, whereas results on exercise capacity, symptoms,quality of life and mortality varied.

Fung et al. [5] reported the efficacy of carvedilol and metoprolol inimproving exercise capacity and quality of life only in the SR group. Ofnote, this observation could be the result of chance only, as study waslimited by the small sample size, especially in the AF group (12patients). In another retrospective analysis of two prospective CHFtrials, carvedilol and metoprolol were shown to increase LVEF inpatients with AF and SR, while data on quality of life and exercisecapacity were not reported [16].

In the post hoc analysis of the US Carvedilol study, an improve-ment in systolic function and physician global assessment alongwith a trend toward a reduction in the combined endpoint of deathor HF hospitalization in patients with AF receiving carvedilol was ob-served [3]. This opposes the CIBIS II post hoc report where the benefit

Fig. 2. Patients with AF more often reached maximum beta-blocker dose compared tothose in SR. * for achieving the European Society of Cardiology target dose.

164 I. Stankovic et al. / International Journal of Cardiology 161 (2012) 160–165

of bisoprolol in reduction of all-cause mortality, cardiovascular mor-tality and hospitalization was observed only in patients with SR [4].Excessive systolic blood pressure decrease was linked with highermortality in AF group, whereas the effects of bisoprolol on LV function,exercise capacity and symptoms in patients with AF were not reported.As suggested by CIBIS II investigators, the observation could be due tochance [4], and it was not confirmed in other large-scale trials such asMERIT [13] and COPERNICUS [14]. These studies found similar benefitof metoprolol and carvedilol among patients with SR and AF [17].

Moreover, it is uncertain whether the conflicting results from USCarvedilol [3] and CIBIS II [4] substudies are due to distinct pharma-cological properties of carvedilol and bisoprolol. Both carvedilol, anonselective beta and alpha1-adrenergic receptor antagonist andbisoprolol, selective beta1-adrenergic receptor antagonist, demon-strated efficacy in patients with chronic HF and SR, but it is unknownwhether either of them could perform better in HF patients with AF.

Table 3Clinical endpoints and relationship with the type of beta-blocker.

Overall biso–carve change D

Heart rate, bpm −2.1 −Mean change (95%CI) (−3.6 to −0.5) (p-Value (biso vs carve) 0.008

Systolic BP, mm Hg +0.6 −Mean change (95%CI) (−1.7 to +2.9) (p-Value (biso vs carve) 0.597

Diastolic BP, mm Hg −0.3 +Mean change (95%CI) (−1.6 to +1.1) (p-Value (biso vs carve) 0.686

LVEF, % +0.4 −Mean change (95%CI) (−0.5 to +1.4) (p-Value (biso vs carve) 0.363

NYHA class −0.01 +Mean change (95%CI) (−0.08 to +0.06) (p-Value (biso vs carve) 0.712

Self rated health (1=excellent; 5=poor) +0.06 +Mean change (95%CI) (−0.05 to +0.17) (p-Value (biso vs carve) 0.280

6MWD, m +5.3 +Mean change (95%CI) (−3.9 to +14.4) (p-Value (biso vs carve) 0.253

SF-36 Physical component +0.4 −Mean change (95%CI) (−0.7 to +1.5) (p-Value (biso vs carve) 0.485

SF-36 Psychosocial component +0.4 −Mean change (95%CI) (−1.0 to +1.7) (p-Value (biso vs carve) 0.610

6MWD= 6-minute walk distance. BP = blood pressure. LVEF = left ventricular ejection fracAssociation.

Our study was the first to compare carvedilol and bisoprolol in pa-tients with chronic HF complicated by AF, and it did not demonstratedrug-related differences in achieving beneficial clinical effects of thebeta-blocker titration.

4.2. Heart rate response to beta-blocker treatment

Apart from exercise capacity, functional status and survival, thereis inconsistency in reported heart rate response to beta-blocker treat-ment in patients with chronic HF and AF. In previous studies, heartrate reduction with beta-blockers was found to be lower [4,16] orsimilar [5] in patients with AF compared with those in SR.

Retrospective analyses of CIBIS II showed that both baseline heartrate and its change over time are prognosticmarkers, with lower base-line heart rate and greater reductions being better [4]. However, it wasalso shown that beta-blocker treatment adds a mortality benefit inde-pendent of baseline heart rate and of its reduction with time [4].

In the present study, there was no significant difference in heart ratereduction between the rhythm groups, but patients with AF weremorelikely to reach recommended beta-blocker doses. On the other hand, ithas been shown that the magnitude of heart rate reduction was associ-ated with the survival benefit of beta-blockers in HF, whereas the doseof beta-blocker was not [8]. In our study, patients with AF had higherbaseline heart rates, and despite more frequently reaching higherbeta-blocker dose levels, they also had higher HR at the end of thestudy. Hence, it is uncertain whether the overall benefits associatedwith beta-blocker titration observed in this study could be translatedinto improved survival in patients with chronic HF and AF.

Furthermore, this indirectly suggests that patients with AF mayhave a higher level of sympathetic tone which would require higherdoses to achieve a similar level of beta-blockade, which was also hy-pothesized by CIBIS II investigators [4]. Indeed, the development of AFhas been demonstrated to cause an increase in sympathetic nervousactivity, and this phenomenon is not only due to rapid heart rates,but partially to the irregular ventricular response itself [18].

ifference biso–carve in AF Difference biso–carve in SR Interaction

2.7 −1.8−6.2 to +0.7) (−3.4 to −0.1) p=0.614

0.8 +0.9−6.1 to +4.6) (−1.6 to +3.5) p=0.574

0.5 −0.4−2.6 to +3.5) (−1.9 to +1.0) p=0.611

0.05 +0.55−2.2 to +2.1) (−0.5 to +1.6) p=0.627

0.002 −0.02−0.154 to +0.157) (−0.09 to +0.06) p=0.846

0.03 +0.07−0.21 to +0.27) (−0.06 to +0.19) p=0.784

2.0 +6.0−19.1 to +23.1) (−4.1 to +16.2) p=0.736

0.8 +0.7 p=0.304−3.2 to +1.7) (−0.5 to +1.9)

0.02 +0.45 p=0.788−3.1 to +3.1) (−1.1 to +2.0)

tion. NTproBNP = N-terminal pro b-type natriuretic peptide. NYHA= New York Heart

165I. Stankovic et al. / International Journal of Cardiology 161 (2012) 160–165

4.3. Limitations

Our study has some important limitations that should be takeninto account when interpreting the results. First, patients with pre-served LVEF (36% in SR and 15% in AF group), were also included inthe present study. Not surprisingly, our data showed that no im-provement in LVEF can be expected with beta blockers in patientswith preserved LVEF, regardless of cardiac rhythm. Although convinc-ing data from prospective randomized trials are lacking, our findingsare concordant with results from recent studies [19–21] whichsuggested that the effects of beta-blockade (except for LVEF) mightbe similar in patients with preserved and impaired LVEF.

Second, as in real-world clinical practice and in some previousclinical trials (CIBIS II and US Carvedilol Heart Failure Trials Program),the resting HR was measured in both groups by means of 12-leadECG. A trial evaluating HR changes in patients with AF would ideallyuse a 24-hour ambulatory ECG monitoring. Finally, we cannot ruleout the possibility that beneficial effects of beta-blockade mighthave been even larger if the follow-up period had lasted longer thanstipulated in the CIBIS-ELD protocol.

4.4. Clinical implications

We showed that improvement of clinical parameters, exercisecapacity and quality of life after beta-blocker titration is similar inpatientswith chronicHF and coexistingAF compared to patients in SR. Al-though we cannot conclude on mortality benefits, beta-blocker prescrip-tion should not be avoided in patientswith chronic HF complicated by AF.Large-scale prospective, placebo-controlled trials are warranted to con-firmmortality benefits in patientswith chronic HF andAF. Chronic HF pa-tients with AF are more likely to reach guideline-recommended targetdoses. Higher doses of beta-blocker may be needed to achieve similarheart rate reduction when compared to patients in SR.

Disclosures

This work was supported by the Competence Network Heart Fail-ure, funded by the Federal Ministry of Education and Research(BMBF), FKZ 01GI0205. H.-D.D. reported receiving research grantsupport and travel support from Merck KGaA, and equipment provi-sion support from Merck KgaA, Roche, and Biosite. A.N.N. and B.P.are supported in part by the unrestricted research grant No 175099,from the Ministry of Education and Science of the Republic of Serbia.S.I. reported receiving travel support from Merck KGaA. ML receivedHeart Failure Association Research Fellowship. For all other authors,there is nothing to declare.

References

[1] Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. Atrialfibrillation is associated with an increase risk for mortality and heart failureprogression in patients with asymptomatic and symptomatic left ventricularsystolic dysfunction: a retrospective analysis of the SOLVD trials. J Am Coll Cardiol1998;32:695–703.

[2] Wang TJ, Larson MG, Levy D, et al. Temporal relations of atrial fibrillation andcongestive heart failure and their joint influence on mortality. The FraminghamHeart Study. Circulation 2003;107:2920–5.

[3] Joglar JA, Acusta AP, Shusterman NH, et al. Effect of carvedilol on survival andhemodynamics in patients with atrial fibrillation and left ventricular dysfunction:retrospective analysis of the US Carvedilol Heart Failure Trials Program. Am HeartJ 2001;142:498–501.

[4] Lechat P, Hulot JS, Escolano S, et al. Heart rate and cardiac rhythm relationshipswith bisoprolol benefit in chronic heart failure in CIBIS II Trial. Circulation2001;103:1428–33.

[5] Fung JW, Chan SK, Yeung LY, Sanderson JE. Is beta-blockade useful in heart failurepatients with atrial fibrillation? An analysis of data from two previously completedprospective trials. Eur J Heart Fail 2002;4(4):489–94.

[6] Khand AU, Rankin AC, Martin W, Taylor J, Gemmell I, Cleland JG. Carvedilol aloneor in combination with digoxin for the management of atrial fibrillation inpatients with heart failure? J Am Coll Cardiol 2003;42(11):1944–51.

[7] Swedberg K, Olsson LG, Charlesworth A, et al. Prognostic relevance of atrial fibrillationin patients with chronic heart failure on long-term treatment with beta-blockers:results from COMET. Eur Heart J 2005;26(13):1303–8.

[8] McAlister FA, Wiebe N, Ezekowitz JA, Leung AA, Armstrong PW. Meta-analysis:beta-blocker dose, heart rate reduction, and death in patients with heart failure.Ann Intern Med 2009;150:784–94.

[9] Düngen HD, Apostolović S, Inkrot S, et al. Titration to target dose of bisoprololversus carvedilol in elderly patients with heart failure: the CIBIS-ELD trial. Eur JHeart Fail 2011;13(6):670–80.

[10] Düngen HD, Apostolović S, Inkrot S, et al. Subproject Multicenter Trials in theCompetence Network Heart Failure. Bisoprolol vs. carvedilol in elderly patientswith heart failure: rationale and design of the CIBIS-ELD trial. Clin Res Cardiol2008;97(9):578–86.

[11] Coats AJ, Shewan LG. Ethics in the authorship and publishing of scientific arti-cles. Int J Cardio 2011;153:239–40.

[12] CIBIS II Investigators and Committees. The Cardiac Insufficiency Bisoprolol StudyII (CIBIS-II): a randomised trial. Lancet 1999;353:9–13.

[13] MERIT HF Study Group. Effect of metoprolol CR/XL in chronic heart failure:Metoprolol CR/XL Randomized Intervention Trial in Congestive Heart Failure(MERIT-HF). Lancet 1999;353:2001–7.

[14] Packer M, Coats AJ, Fowler MB, et al. Effect of carvedilol on survival in severechronic heart failure. N Engl J Med 2001;344:1651–8.

[15] Flather MD, Shibata MC, Coats AJ, et al. Randomized trial to determine the effect ofnebivolol on mortality and cardiovascular hospital admission in elderly patientswith heart failure (SENIORS). Eur Heart J 2005;26(3):215–25.

[16] Meng F, Yoshikawa T, Baba A, et al. Beta-blockers are effective in congestive heartfailure patients with atrial fibrillation. J Card Fail 2003;9(5):398–403.

[17] Lechat PP. Beta-blocker efficacy according to heart rate and rhythm in patientswith heart failure. Commentary on the Cardiac Insufficiency Bisoprolol Study IIanalysis. Card Electrophysiol Rev 2003;7(3):233–5.

[18] Wasmund SL, LI JM, Page RL, et al. Effect of atrial fibrillation and an irregularventricular response on sympathetic nerve activity in human subjects. Circulation2003;107(15):2011–5.

[19] van Veldhuisen DJ, Cohen-Solal A, Böhm M, et al. Beta-blockade with nebivolol inelderly heart failure patients with impaired and preserved left ventricular ejec-tion fraction: data From SENIORS (Study of Effects of Nebivolol Intervention onOutcomes and Rehospitalization in Seniors with Heart Failure). J Am Coll Cardiol2009;53(23):2150–8.

[20] Dobre D, van Veldhuisen DJ, DeJongste MJ, et al. Prescription of beta-blockers inpatients with advanced heart failure and preserved left ventricular ejectionfraction. Clinical implications and survival. Eur J Heart Fail 2007;9(3):280–6.

[21] Bergström A, Andersson B, Edner M, Nylander E, Persson H, Dahlström U. Effect ofcarvedilol on diastolic function in patients with diastolic heart failure andpreserved systolic function. Results of the Swedish Doppler-echocardiographicstudy (SWEDIC). Eur J Heart Fail 2004;6(4):453–61.