Pii: s0140-6736(98)11181-9

Articles
The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II):a randomised trial IntroductionExperimental and clinical trials have shown beneficial Background In patients with heart failure, β-blockade has effects with β-blockade in heart failure.1–3 There is improved morbidity and left-ventricular function, but the reluctance to use β-blockade therapy, however, and impact on survival is uncertain. We investigated the unequivocal evidence of benefit from randomised placebo efficacy of bisoprolol, a β selective adrenoceptor blocker controlled trials is needed to convince the medical in decreasing all-cause mortality in chronic heart failure.
community of its safety and efficacy.
Clinical trials in heart failure have tested compounds Methods In a multicentre double-blind randomised placebo- with different pharmacological profiles.2,3 Meta-analyses of controlled trial in Europe, we enrolled 2647 symptomatic placebo-controlled trials of β-blockers have suggested an patients in New York Heart Association class III or IV, with overall effect on mortality of 32%.4–6 The Cardiac left-ventricular ejection fraction of 35% or less receiving Insufficiency Bisoprolol Study (CIBIS) studied bisoprolol, standard therapy with diuretics and inhibitors of a highly selective antagonist of β adrenoceptors, which are angiotensin-converting enzyme. We randomly assigned found mainly in the heart and especially in ventricular patients bisoprolol 1·25 mg (n=1327) or placebo (n=1320) tissue.7 That trial showed a non-significant trend towards daily, the drug being progressively increased to a maximum 20% lower mortality in the bisoprolol group and 30% of 10 mg per day. Patients were followed up for a mean of fewer admissions to hospital for worsening heart failure.8 1·3 years. Analysis was by intention to treat.
We designed the CIBIS-II trial to test this evidence Findings CIBIS-II was stopped early, after the second further, based on the CIBIS trial results.
interim analysis, because bisoprolol showed a significant mortality benefit. All-cause mortality was significantly lower with bisoprolol than on placebo (156 [11·8%] vs 228 The study design and protocol of CIBIS has been published.9We did a double-blind placebo-controlled randomised trial, [17·3%] deaths with a hazard ratio of 0·66 (95% CI 0·54–0·81, p<0·0001). There were significantly fewersudden deaths among patients on bisoprolol than in those on placebo (48 [3·6%] vs 83 [6·3%] deaths), with a hazard Eligible patients were ambulatory, aged 18–80 years, and had a ratio of 0·56 (0·39–0·80, p=0·0011). Treatment effects left-ventricular ejection fraction, measured within 6 weeks of were independent of the severity or cause of heart failure.
randomisation, of 35% or less. Symptoms had to includedyspnoea on exertion, orthopnoea, or paroxysmal nocturnal Interpretation β-blocker therapy had benefits for survival in dyspnoea, with or without oedema, and fatigue, corresponding stable heart-failure patients. Results should not, however, to class III or IV of the New York Heart Association (NYHA).
be extrapolated to patients with severe class IV symptoms We recruited patients from 274 hospitals in 18 countries in and recent instability because safety and efficacy has not Patients had to have a diagnosis of chronic heart failure, made at least 3 months previously, with clinical stability during the preceding 6 weeks for heart failure or 3 months for acute myocardial infarction or unstable angina. Cardiovascular therapyhad to have been unchanged in the 2 weeks beforerandomisation. Treatment had to include a diuretic and anangiotensin-converting-enzyme (ACE) inhibitor, although weallowed other vasodilators if patients were intolerant of ACEinhibitors; the use of digoxin was optional. We measured left-ventricular ejection fraction by echocardiography with theTeicholz formula for M-mode recordings, or the modifiedSimpson’s rule for measurements of end-diastolic and end-systolic volume on apical two-dimensional views. If adequateviews could not be obtained by echocardiography, we usedcontrast or radionuclide ventriculography.
The main exclusion criteria were uncontrolled hypertension, myocardial infarction or unstable angina pectoris in the previous3 months, percutaneous transluminal coronary angioplasty or *Investigators and committee members listed at end of paper coronary-artery bypass graft in the previous 6 months, previous Correspondence to: Philippe Lechat, Pharmacology Department, or scheduled heart transplant, atrioventricular block greater than AP-HP, Pitié Salpêtrière Hospital, 75013 Paris, France first degree without a chronically implanted pacemaker, resting (e-mail: [email protected]) heart rate of less than 60 beats per min, systolic blood pressure THE LANCET • Vol 353 • January 2, 1999 Mean (SD) systolic blood pressure (mm Hg) Mean (SD) diastolic blood pressure (mm Hg) Mean (SD) left-ventricular ejection fraction (%) diameter (cm)Mean (SD) left-ventricular end-systolic Mean (SD) left-ventricular fractional shortening All important clinical circumstances were analysed by the members of the critical events committee, who were masked to treatment status. They classified all endpoints according to strict definitions of critical events, only the most important of whichare defined here.
*Coronary angiography unavailable or no history of myocardial infarction.
Sudden death was defined as death occurring within 1 h Table 1: Baseline characteristics of patients without previous worsening of symptoms of heart failure. We at rest of less than 100 mm Hg, renal failure (serum creatinine also took unexpected deaths occurring during sleep to be sudden у300 µmol/L), reversible obstructive lung disease, or pre- when patients were found dead by family members sharing the existing or planned therapy with β-adrenoreceptor blockers.
same room in the morning. We generally classified other We did not allow treatment with β-blockers (including eye unwitnessed deaths as unknown. We classified pump failure drops), calcium antagonists, inotropic agents except digitalis, when death occurred as a consequence of progressive and antiarrhythmic drugs other than amiodarone during the trial.
deterioration of heart failure, acute pulmonary oedema, or We identified three mutually exclusive aetiological groups: cardiogenic shock. We recorded non-cardiovascular death if patients with definite ischaemic heart disease with at least one cardiovascular events were excluded as cause of death. Death important coronary arterial stenosis of 70% or more of luminal was classified as being due to unknown cause when there was diameter seen on angiography, or a confirmed myocardial insufficient evidence to confirm cardiovascular or non- infarction; those with idiopathic dilated cardiomyopathy if normal coronary arteries were seen on angiography; and patients We recorded permanent treatment withdrawal when a medical with valvular heart disease or hypertension, together with those need for a β-blocker arose, when intolerance to study medication with suspected but unproved ischaemic heart disease or occurred despite increases in baseline therapy, if study-drug dose cardiomyopathy. We used a non-parallel structure to describe was decreased or temporarily withdrawn, if patients experienced intolerance to first dose, and for all other circumstances in which Randomisation was done by random numbers generated oncomputer at the independent statistical centre, sent to study centres by telefax. The code was kept at the statistical centre andwas not broken until the trial was stopped.
The study treatments were identical in appearance. Patients were started on bisoprolol 1·25 mg (n=1327) or placebo (n=1320) daily, the drug being increased successively to 2·50 mg, 3·75 mg, 5·00 mg, 7·50 mg, and 10·00 mg, according to tolerance. Patients received the first three concentrations of each dose for 1 week, and the higher concentrations for 4 weeks.
Investigators were asked to ensure that the highest tolerated dose was reached and maintained, if possible, for the duration of the trial. In patients with worsening heart failure, we recommended that the baseline heart-failure treatments were increased before the study drug was decreased. We followed up until the end of the study all patients in whom study medication was withdrawn, with follow-up visits every 3 months. There was no run-in period.
The primary endpoint was all-cause mortality. Secondary Numbers refer to patients who presented at least once with given event. For hospital endpoints were all-cause hospital admissions, cardiovascular admissions, numbers refer to patients admitted at least once with any cause.
mortality, cardiovascular mortality and cardiovascular hospital Table 2: Primary and secondary endpoints and exploratory admissions, a composite endpoint, and permanent premature THE LANCET • Vol 353 • January 2, 1999 Based on the CIBIS survival curves, we estimated the annual mortality rate to be about 11·2% in the placebo group. Toobtain a minimum of 25% lower mortality in the bisoprolol group in a 1-year recruitment period and 2-year follow-up, wecalculated that for an ␣ risk of 5% and a power of 95%, we We planned two interim analyses at 2500 patient-years and 5000 patient-years. The study could be stopped according to Peto’s rule10 if a significant difference in all-cause mortalitywas seen between the two groups at p<0·001 (two-tailed log- We did analyses by intention to treat. We calculated Kaplan- Figure 3: Relative risk of treatment effect on mortality by Meier survival curves on total mortality, and assessed differences aetiology and functional class at baseline between the treatment groups with the log-rank test (time to event). Hazard ratios and 95% CIs were calculated with Cox’sproportional hazards regression model. We used the Breslow- groups for any subgroup of aetiology of heart failure or Day test to calculate homogeneity of odds ratios between treatment groups, according to NYHA class and cause of Circumstances and causes of deaths are shown in heart failure. We compared baseline variables between the two table 2. There were 48 sudden deaths in the bisoprolol groups with Student’s t or Wilcoxon’s rank-sum tests for groups compared with 83 in the placebo group, a continuous variables, and Fisher’s exact or ␹2 tests for categorical The difference in admissions to hospital for worsening heart failure between the two groups was 32% (p<0·0001, table 2). There were, however, more admissions to 2647 patients were enrolled into the study and followed hospital for stroke in the bisoprolol group than in the up for a mean of 1·3 years. Baseline characteristics were placebo group (31 vs 16, p=0·04). Hospital admissions similar in the two groups (table 1).
were significantly fewer in the bisoprolol group than in The trial was stopped early because all-cause mortality the placebo group for ventricular tachycardia and was significantly less in the bisoprolol group than in the ventricular fibrillation (six vs 20, p=0·006) and for placebo group (figure 1). In the bisoprolol group, 156 hypotension (three vs 11, p=0·03), but were more (11·8%) patients died, compared with 228 (17·3%) in the common for bradycardia (14 vs two, p<0·004). The rate placebo group (p<0·0001). The estimated annual of heart transplantation was low and similar in the two mortality rate was 8·8% in the bisoprolol group and groups. The number of hospital admissions did not differ 13·2% in the placebo group (hazard ratio 0·66 [95% CI significantly for angina, myocardial infarction, supraventricular arrhythmias, cardiogenic shock, or There were significantly fewer cardiovascular deaths among patients on bisoprolol than among those on The most common dose of bisoprolol during the placebo (p=0·0049). Significantly fewer patients on maintenance phase was 10·0 mg, which was reached in bisoprolol were admitted to hospital for all causes than 564 patients; 152 reached 7·5 mg and 176 reached patients on placebo (p=0·0006) as well as for the combined endpoint of cardiovascular death and admssion Treatment effect did not differ between the to hospital for cardiovascular events (p=0·0004). The number of permanent treatment withdrawals was similarin the two groups (table 2).
We did subgroup analyses by cause of heart failure and β-blockade had benefits for all-cause mortality in patients severity of disease at baseline (figure 3). Mortality and with chronic heart failure. Benefits were also seen for admissions to hospital did not differ significantly between morbidity, assessed by admissions to hospital for allcauses, especially for worsening heart failure.
The magnitude of the treatment effect (a 32% lower risk of mortality and admission to hospital for heartfailure) is in accordance with findings from meta-analyses of previous randomised placebo-controlled trials.4 Ourresults were obtained in patients already taking diuretics and ACE inhibitors and not patients selected for tolerance of bisoprolol, since we had no run-in period.
Benefit occurred irrespective of the cause of heart failureor NYHA class of severity. The greatest effect was,however, seen in patients with ischaemic heart disease With the inclusion of our results, the cumulative experience with β-blockade therapy in chronic heart failure (more than 6000 patients in randomised trials) approaches that of ACE inhibitors in heart-failurepatients with symptoms.11 Benefit from the addition of THE LANCET • Vol 353 • January 2, 1999 blockade to ACE inhibitors after acute myocardial with our trial impossible. In theory, blockade of β - infarction has also been suggested in post-hoc analyses,12 adrenoceptors and β adrenoceptors should provide more but this strategy has not been supported by randomised complete protection against the harmful effects of controlled trials with sufficient power to address outcome catecholamines, but our results show that selective in patients with left-ventricular dysfunction, with or inhibition of β receptors is sufficient to lower the rate of sudden death presumed to be associated with arrythmia.
Neuroendocrine activation may underlie this Differences in effects according to the pharmacological therapeutic benefit by inhibition of the potential harmful profiles of β−blockers is, however, important and effects of compensatory mechanisms, the renin- continuing trials of drugs such as bucindolol,20 carvedilol, angiotensin-aldosterone system for ACE inhibitors, and and metoprolol with carvedilol will provide essential sympathetic activity with β-blockers. Cardiac work and energy consumption are decreased by unloading with The lack of difference in treatment effects on mortality ACE inhibitors, slowing of heart rate with β-blockers, and and secondary endpoints by cause or severity of disease lowering of blood pressure with both. The contrasts with the findings of CIBIS, in which bisoprolol neuroendocrine hypothesis in heart failure includes the had greatest benefits in patients with non-ischaemic possibility that ACE inhibitors prevent direct toxic heart failure. Given the consistent and striking benefit of cardiac effects of angiotensin II and aldosterone, and β-blockers in secondary prevention after myocardial β-blockers prevent the toxic effects of catecholamines.
infarction,7,21 there is no plausible scientific explanation The greatest effect on mortality was a 42% lower rate for this apparent anomaly. This observation did, however, of sudden deaths among patients on bisoprolol, as well as result from a post-hoc analysis and highlights the non-significantly fewer deaths related to pump failure.
This finding suggests that bisoprolol was acting We saw benefits of bisoprolol for patients in NYHA principally as an antiarrhythmic agent rather than having class IV; however, we included only stable patients and a specific effect on myocardial function. Given what is the use of β-blocker treatment in non-ambulatory known already about the positive effects of β-blockade on patients with class IV symptoms, especially those with cardiac structure and function, the way in which we recent instability, needs to be defined.
classified the cause of death should be taken into account.
The addition of a β-blocker to standard therapy with a Sudden deaths or those associated with pump failure had diuretic and an ACE inhibitor can be recommended in strict definitions. Many deaths had, therefore, to appropriate, stable, ambulatory patients who have heart be classified for the purposes of the trial as being failure caused by impaired left-ventricular systolic unknown cause. Unwitnessed or insufficiently function. The limited use of β-blocker therapy after documented deaths classified as unknown were probably myocardial infarction, despite the cumulative evidence of sudden and some deaths associated with pump failure, double-blind, randomised, controlled trials, suggests that which can also seem to be sudden. Moreover, bisoprolol anxiety about safety or lack of clarity about the target led to a significantly lower death rate in the unknown population are common. The continued accumulation of group, which strongly implies that most of the deaths in information about β-blockers in heart failure is, therefore, this category were cardiac. These difficulties of important, since the population of patients with heart classification reinforce the value of all-cause mortality as failure is much less well-defined than that for patients the major mortality endpoint in similar trials.
with myocardial infarction. Without further information The strikingly lower frequency of sudden deaths among from large randomised controlled trials, the uptake of patients on bisoprolol in our trial suggests an important β-blockade in clinical practice outside specialist antiarrhythmic effect. Although a similar difference was not seen in CIBIS, a significant trend was found in the For all heart-failure patients, administration of β-blocker USA carvedilol studies.13 In CIBIS, the rate of ventricular therapy should be gradual and progressive, starting with tachycardia episodes was lower in a substudy, in which low doses. The optimum rate of dose increase and the vagally dependent variability in heart rate was higher in maximum dose need to be more accurately defined. Use the bisoprolol group than in the placebo group,14,15 an of the maximum tolerated dose seems acceptable; at effect that has been linked to improved long-term present, recommendations on rates of dose increase can prognosis after myocardial infarction and in heart be based only on those adopted in clinical trials.
failure.16,17 This finding and the significantly lower rate of Patients with severe class IV heart failure, those with admission to hospital for ventricular tachycardia or heart failure after acute myocardial infarction, and those fibrillation in the CIBIS-II bisoprolol group supports the with symptomless left-ventricular dysfunction are being drug’s potential antiarrhythmic effect.
studied in the continuing clinical trials COPERNICUS, Improvement or preservation of left-ventricular CAPRICORN, and CARMEN with carvedilol.
function could also improve long-term prognosis.
In our trial the mean age of patients was 61 years, at Increased left-ventricular ejection fraction has been seen least a decade younger than that of patients seen in with other β-blockers,4 which may be dose-dependent.18 clinical practice. In most clinical trials in heart failure, We did not measure left-ventricular function sequentially, there is, therefore, inadequate information about the but in CIBIS prognostic improvement was significantly effects of treatment in older patients and more data in the linked to increased left-ventricular ejection fraction.19 A meta-analysis of randomised trials showed a trend towards better survival with non-selective compounds.4 Austria—Klein W Brunhuber, R Hofmann, P Kühn, H-J Nesser, J Slany, This finding was mainly related to the strikingly lower W Weihs, C Wiedermann, H Wimmer. Belgium—W van Mieghem, mortality rate seen in the US trials of carvedilol;8 J Boland, J M Chaudron, L Jordaens, J P Melchior. Czech Republic—M Aschermann, J Bruthansl, M Hradec, F Kölbel, B Semrád. Denmark— differences in design, especially the presence of a run-in T Haghfelt, J Fischer-Hansen, C O Goetzsche, P Hildebrandt, E Kassis, phase in the carvedilol trials, makes comparison of results V Rasmussen, J Rokkedal, A Thomassen. Finland—K Groundstroem, THE LANCET • Vol 353 • January 2, 1999 P Uusimaa. France—J Y Le Heuzey, M C Aumont, J F Aupetit, N Baille, Waagstein F, Hjalmarson A, Varnauskas E, Wallentin I. Effect of P Baudouy, A Belin, A Bonneau, G Bonneric, J P Bousser, B Citron, chronic beta-adrenergic receptor blockade in congestive P Dary, E Decoulx, P De Groote, T Denolle, F Dievart, P Duriez, cardiomyopathy. Br Heart J 1975; 37: 1022–36.
J C Eicher, G Enjuto, M Ferrière, E Fournier, M Garandeau, J Gauthier, The Metoprolol in Dilated Cardiomyopathy (MDC) Trial Study M Genest, A Gerbe, J P Godenir, B Guillot, J P Guillot, P Guillot, Group. 3-year follow-up of patients randomised in the Metoprolol in P Heno, C d’Ivernois, M Jean, S Kacet, R Kalle, M Komajda, A Lacroix, Dilated Cardiomyopathy Trial. Lancet 1998; 351: 1180–81.
R Lallemand, H Lardoux, M Marquet, M Martin, O Martin, D Méry, Australia and New Zealand Heart Failure Research Collaborative R Mossaz, P Mothes, T Olive , M Ostorero, F Paganelli, E Page, Group. Randomised, placebo-controlled trial of carvedilol in patients C Pauly-Laubry, J Puel, J F Rousseau, J J Roux, A Schenowitz, with congestive heart failure due to ischaemic heart disease. Lancet K Sourdais, F Tremel, A Verdun, S Witchiz, J E Wolf. Germany— 1997; 349: 375–80.
V Hombach, I Assmann, T Beyer, K O Bischoff, H Darius, G Ertl., Doughty RN, Rodgers A, Sharpe N, MacMahon S. Effects of E Fleck, K Förster, F Freytag, U Gleichmann, R Haasis, R Henßge, beta-blocker therapy on mortality in patients with heart failure: a D Hey, P Hesse, T Höfs, M Keck, H Klein, E T Kromer, J Krüls-Münch, systematic overview of randomised controlled trials. Eur Heart J 1997; B Lüderitz, B Maisch, V Mitrovic, S Neubauer, K J Osterziel, H Simon, 8: 560–65.
S G Spitzer, R Stöhring, G Taubert, W Teichmann, K Theisen, Heidenrich PA, Lee TT, Massie BM. Effect of beta-blockade on W Wende, H Wieser, R Zotz. Hungary—I Preda, M Csanády, mortality in patients with heart failure: a meta-analysis of randomized L Cserhalmi, I Edes,T Gesztesi, P Kárpáti, K Simon, J Tarján. clinical trials. J Am Coll Cardiol 1997; 30: 27–34.
Italy—R Fogari, R Tramarin, N Galie, P Giani, U Milanese, S Scalvini, Lechat P, Packer M, Chalon S, Cucherat M, Arab T, Boissel JP. D Scrutinio, L A Sechi, F Tettamanti, F De Vito. Ireland—P Crean, Beta-blockers in heart failure: meta-analysis of randomized trials.
H McCann, D Mulcahy, D Sugrue. Netherlands— Circulation 1998, 98: 1184–91.
D C A van Hoogenhuyze, P H van der Burgh, R Ciampricotti,J M van Dantzig, F R DenHartog, J A Henneman, H A M van Kesteren, Haeusler G, Schliep HJ, Schelling P, et al. High β1 selectivity and J A Kragten, K L Liem, A Limburg, M R van der Linde, G C M Linssen, favourable pharmacokinetics as the outstanding properties of H Pasteuning, H J A M Penn, P Van Rossum, H J Schaafsma, bisoprolol. J Cardiovasc Pharmacol 1986; 8 (suppl II): S2–15.
A Schelling, R Sloos, J C L Wesdorp. Poland—J Korewicki, CIBIS investigators and committees. A randomized trial of β-blockade P Achremczyk, E Czestockowska, M Dowgird, A Dyduszynski, J Górski, in heart failure: the Cardiac Insufficiency Bisoprolol Study (CIBIS).
K Ilmurzynska, K Janicki, Z Kornacewicz-Jach, T Kraska, Circulation 1994; 90: 1765–73.
M Krzeminska-Pakula, J Kuch, E Nartowicz, T Petelenz, W Piwowarska, The CIBIS-II Scientific Committee. Design of the Cardiac I Rawczynska-Englert, W Ruzyllo, G Swiatecka, M Tendera, Insufficiency Bisoprolol Study II (CIBIS II). Fund Clin Pharmacol M Wierzchowiecki, J Wodniecki, D Wojciechowoski, K Wrabec, 1997; 11: 38–42.
H Wysocki. Portugal—R Seabra Gomes, M Fátima Ceia, N Lousada, 10 Peto R, Pike MC, Armitage P, et al. Design and analysis of J M Martins Campos, L A Providência, A L Zamith Cerveira de Moura.
randomised clinical trials requiring prolonged observation of Russia—V J Marejev, D M Aronov, G P Arutjunov, B J Bart, each patient, I: introduction and design. Br J Cancer 1976; 34:
S S Basechikin, J N Belenkov, J B Beloussov, O A Bokeria, R A Charchogljan, V Doschytsin, T A Fedorova, M G Glezer, 11 Garg R, Yusuf S, for the Collaborative Group on ACE Inhibitor A Gorbachenkov, VA Gorshkov, A L Gospodarenko, V T Ivashkin, Trials. Overview of randomized trials of angiotensin-converting A J Ivleva, A A Kyrichenko, A A Lavrov, L B Lazebnik, A Marynov, enzyme inhibitors on mortality and morbidity in patients with heart V P Mazaev, N R Polejev, A Shpektor, B A Sidorenko, K E Sobolev, failure. JAMA 1995; 273: 1450–56.
A K Starodoubtsev, G I Storozhakhov, A L Syrkin, V S Zodionchenko, 12 Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on T V Zvereva. Slovakia—J Murin, G Kaliská, R Rybar. Spain—V Valle, mortality among high risk and low risk patients after myocardial M Artaza, P Conthe, J M Cruz, M Garcia-Moll, J L Lopez-Sendon, infarction. N Engl J Med 1998; 339: 489–97.
A Martinez, F Monzón, M Ribas, E Roig, I Roldan Sweden—C Höglund, 13 Packer M, Bristow MR, Cohn J, et al for the US Carvedilol Heart S Ekdahl, L Hjelmaeus, K Lindberg, P Löfdahl, G Ulvenstam, Failure Study Group. The effect of carvedilol on morbidity and L Warselius. Switzerland—F Follath, W Anghern, P Dubach, P Erne, mortality in patients with chronic heart failure. N Engl J Med 1996; A Gallino, T Moccetti. UK—A Bridges, J Adgey, G Ambepitiya, N Boon, 334: 1349–55.
R M Boyle, A J Cowley, T Cripps, M K Davies, F Dunn, J Findlay, 14 Pousset F, Copie X, Lechat P, et al. Effects of bisoprolol on heart rate P Forsey, T Fyfe, B Gould, T W Greenwood, P Hubner, S Khan, variability in heart failure. Am J Cardiol 1996; 77: 612–17.
P Lewis, A Mackay, M Maltz, J McArthur, A McLeod, D McLeod, 15 Copie X, Pousset F, Lechat P, Jaillon P, Guize L, Le Heuzey JY, and M Metcalfe, M Millar-Craig, P Mills, J K Nelson, D Nicholls, the Cardiac Insufficiency Bisoprolol Study Investigators. Effects of G D Oakley, D L H Patterson, J E F Pohl, S Ray, B Silke, P R Wilkinson.
beta-blockade with bisoprolol on heart rate variability in advanced heart failure: analysis of scatterplots of RR intervals at selected heart rates. Am Heart J 1996; 132: 369–75.
Writing Committee—H J Dargie, P Lechat 16 Leiger RE, Miller JP, Bigger JT, Moss AJ and the multicenter Scientific committee—H J Dargie (Chairman, UK), E Erdmann post-infarction research group. Decreased heart rate variability and (Germany), F Follath (Switzerland), C Höglund (Sweden), P Lechat its association with increased mortality after myocardial infarction.
(France), J L Lopez Sendon (Spain), V Mareyev (Russia), W J Remme Am J Cardiol 1987; 59: 256–62.
(Netherlands), Z Sadowski (Poland), R J Seabra-Gomes (Portugal), 17 Nolan J, Batin P, Andrews R, et al. Prospective study of heart rate F Zannad (France), M Wehrlen-Grandjean (France). variability and mortality in chronic heart failure:results of the United Critical event committee—C Funck-Brentano (Chairman, France), Kingdom Heart Failure Evaluation and Assessment Trial (UK-Heart).
S Hansen (Sweden), S Hohnloser (Germany), E Vanoli (Italy). Circulation 1998; 98: 1510–16.
Advisory and safety committee—P Jaillon (Chairman, France), 18 Bristow M, Gilbert EM, Abraham WT, et al for the MOCHA G De Baker (Belgium), U Dahlstrˆm (SWeden), C Hill (France). investigators. Carvedilol produces dose-related improvements in left Independent statistical centre—A Leizorovicz, F Bugnard, C Rolland ventricular function and survival in subjects with chronic heart failure.
Circulation 1996; 4: 2807–16.
Merck KGaA—H Wiemann, P Verkenne. 19 Lechat P, Escolano S, Golmard JL, et al on behalf of CIBIS Coordinating centre—T Arab, N Cussac, V Dussous, S Haise, investigators. Prognostic value of bisoprolol-induced hemodynamic effects in heart failure during the Cardiac Insufficiency Bisoprolol Study. Circulation 1997; 96: 2197–205.
This study was sponsored by E Merck, Darmstadt.
20 The BEST Steering Committee. Design of the Beta-blocker Evaluation Survival Trial. Am J Cardiol 1995;75: 1220–23.
21 Yusuf S, Peto R, Lewis J, Collins R, Sleight P. Beta-blockade during THE LANCET • Vol 353 • January 2, 1999

Source: http://user.meduni-graz.at/michael.schimek/Wiss_GL_2012_2013/CibisII.pdf