ORIGINAL CONTRIBUTION

Effects of Initiating Carvedilol in Patients

With Severe Chronic Heart Failure
Results From the COPERNICUS Study
Henry Krum, MB, BS, PhD Context ␤-Blockers remain underused despite their established utility for improving
Ellen B. Roecker, PhD outcome in heart failure. Concerns that initiation of treatment produces few imme-
diate benefits and may have important risks may be deterring widespread use.
Paul Mohacsi, MD
Objective To evaluate the early effects of the ␤-blocker carvedilol in patients with
Jean L. Rouleau, MD severe heart failure.
Michal Tendera, MD Design, Setting, and Patients Randomized, double-blind, placebo-controlled trial
Andrew J. S. Coats, MD conducted from October 28, 1997, to March 20, 2000, at 334 hospital centers in 21
countries among 2289 patients with symptoms of heart failure at rest or with minimal
Hugo A. Katus, MD
exertion who were clinically euvolemic and had a left ventricular ejection fraction of
Michael B. Fowler, MD less than 25%.
Milton Packer, MD Intervention Patients were randomly assigned to receive carvedilol, with start dos-
age of at 3.125 mg twice daily with uptitration to a target dosage of 25 mg twice daily
for the Carvedilol Prospective
(n=1156), or placebo (n=1133), in addition to their usual medications for heart failure.
Randomized Cumulative Survival
(COPERNICUS) Study Group Main Outcome Measures Death, hospitalization, or permanent withdrawal from
study drug, as well as adverse events during the first 8 weeks of treatment.

B
ETA-BLOCKERS PROLONG LIFE
Results The carvedilol group experienced no increase in cardiovascular risk but instead
and reduce the risk of disease had fewer patients who died (19 vs 25; hazard ratio [HR], 0.75; 95% confidence interval
progression in patients with [CI], 0.41-1.35); who died or were hospitalized (134 vs 153; HR, 0.85; 95% CI, 0.67-
chronic heart failure,1-6 but they 1.07); or who died, were hospitalized, or were permanently withdrawn from treatment
remain underutilized in clinical prac- (162 vs 188; HR, 0.83; 95% CI, 0.68-1.03). These effects were similar in direction and
tice despite their established benefits.7 magnitude to those observed during the entire study, and were apparent particularly in
In part, this underutilization is related the 624 patients with recent or recurrent decompensation or a very depressed left ven-
to physician concerns that initiation of tricular ejection fraction. Differences in favor of carvedilol became apparent as early as
14 to 21 days following initiation of treatment. Worsening heart failure was the only se-
treatment with a ␤-blocker is difficult and rious adverse event with a frequency greater than 2% and was reported with similar fre-
requires special expertise.8 Patients who quency in the placebo and carvedilol groups (6.4% vs 5.1%).
start taking a ␤-blocker may experi-
Conclusions These data suggest that, in clinically euvolemic patients, the relation
ence decreases in blood pressure as well
of benefit to risk during initiation of treatment with carvedilol is similar to that seen
as retention of sodium,9,10 which can during long-term therapy with the drug. Our findings should provide the reassurance
cause symptomatic hypotension and/or needed to encourage the high levels of use that are warranted by the results of long-
worsening heart failure during the first term clinical trials.
4 to 8 weeks of therapy.9-14 Further- JAMA. 2003;289:712-718 www.jama.com
more, many physicians have assumed
that the beneficial effects of ␤-blockers Author Affiliations: Monash University, Mel- Physicians and Surgeons, Columbia University, New
are delayed9,15 so that a favorable effect bourne, Victoria, Australia (Dr Krum); University of York, NY (Dr Packer).
Wisconsin, Madison (Dr Roecker); University Hospi- The COPERNICUS Study Investigators and Coordi-
of treatment on symptoms, hospitaliza- tal, Bern, Switzerland (Dr Mohacsi); Silesian School nators are listed at the end of the article.
tions, or death may not become appar- of Medicine, Katowice, Poland (Dr Tendera); Uni- Corresponding Author and Reprints: Henry Krum,
versity Health Network and Mt Sinai Hospital, MD, National Health and Medical Research Council
ent for many months.2-4,16 Concerns that Toronto, Ontario, Canada (Dr Rouleau); Royal Center of Clinical Research Excellence in Therapeu-
Brompton Hospital, London, England (Dr Coats); tics, Dept of Medicine and Epidemiology and Preven-
Universitaets Klinikum Luebeck, Luebeck, Germany tive Medicine, Monash University, Alfred Hospital, Mel-
See also pp 730 and 754. (Dr Katus); Stanford University Medical Center, bourne, Victoria, 3181 Australia, (e-mail: henry.krum
Stanford, Calif (Dr Fowler); and the College of @med.monash.edu.au).

712 JAMA, February 12, 2003—Vol 289, No. 6 (Reprinted) ©2003 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

initiation of ␤-blocker therapy carries mize the degree of volume retention)

Figure 1. Patient Flow During Initiation and
important risks and few immediate ben- and treated with an angiotensin- Up-titration of Study Drug
efits has contributed to the underuti- converting enzyme inhibitor or an
lization of these drugs in the manage- angiotensin II receptor antagonist (un- 3106 Screened
ment of heart failure.7 less these were not tolerated). Treat-
Most of the information we have about ment with digitalis, spironolactone, va- 2289 Randomized
the responses to ␤-blocker therapy have sodilators, and amiodarone were
been derived from uncontrolled stud- allowed, but not required. 1133 Assigned to 1156 Assigned to
Receive Placebo Receive Carvedilol
ies,9,13,15 and thus, it has been difficult to Patients were excluded if they had a
determine if the effects reported were re- reversible or correctable cause of heart 59 Withdew from 51 Withdew from
lated to treatment or to underlying dis- failure; had severe primary pulmonary, Study Drug Study Drug
ease. To date, only 2 controlled trials renal, or hepatic disease; had a contra- 0 Lost to follow-up
of Vital Status
0 Lost to follow-up
of Vital Status
(with metoprolol and bucindolol) have indication to ␤-blocker therapy; or had 25 Died 19 Died
described in detail the clinical events an acute illness that required continued
occurring following the initiation of hospitalization. In addition, patients 1133 Included in 1156 Included in
Analysis Analysis
therapy.16-18 In these studies, initiation of were not allowed to have had within
␤-adrenergic blockade appeared to be the past 2 months cardiac surgery or an-
well-tolerated in patients with mild heart gioplasty, a myocardial or cerebral is-
failure but was associated with an early chemic event, or sustained or hemody- tients were asked about the occurrence
increase in risk of worsening heart fail- namically destabilizing ventricular of any clinical event or adverse effect, vi-
ure and drug withdrawal in patients with tachyarrhythmia. Patients also were ex- tal signs and body weight were mea-
severe heart failure. cluded if they had received an ␣-blocker, sured, the dose of the study drug was re-
We describe the initiation of treat- calcium channel blocker, or class I an- corded, and patients were administered
ment with the ␣, ␤-adrenergic blocker tiarrhythmic drug within 4 weeks; a the next level of the study drug if they
carvedilol in the Carvedilol Prospec- ␤-blocker within 2 months; or an intra- were tolerating the drug at a dosage less
tive Randomized Cumulative Survival venous positive inotropic agent or vaso- than 25 mg twice daily and had not re-
(COPERNICUS) study.19 The primary dilator within 4 days of screening. Other ceived a higher dose. The intent of the
objective of the COPERNICUS trial was exclusion criteria included systolic blood up-titration phase was to identify the
to evaluate the long-term effects of carve- pressure less than 85 mm Hg, heart rate highest dose of carvedilol that each pa-
dilol on the survival of patients with se- less than 68/min, serum creatinine level tient could tolerate, and patients were
vere heart failure. In the overall study greater than 2.8 mg/dL (213.5 µmol/L), considered to have completed the up-
(mean follow-up, 10.4 months), carve- or a serum potassium level less than 3.5 titration phase when they were able to
dilol reduced the risk of death by mEq/L or greater than 5.2 mEq/L. tolerate this dose for 2 weeks. Thus, the
35% compared with placebo. The duration of the up-titration period was
COPERNICUS study provides an ideal Study Design expected to be 8 weeks, although the ra-
setting in which to evaluate the early Details of the study design have been pidity of up-titration could be slowed if
benefits and risks of treatment, since this previously published.19 In this double- deemed appropriate.
trial focused on patients with severe blind trial, eligible patients were ran- If warranted by clinical circum-
heart failure who might be expected to domly assigned to receive either carve- stances, the dose of carvedilol or pla-
have the greatest difficulty starting treat- dilol or placebo (in a 1:1 ratio provided cebo could be reduced or temporarily
ment with a ␤-blocker.8,9,13 as capsules identical in size and shape), discontinued, the doses of all concomi-
in addition to their usual medications for tant drugs could be adjusted, and the
METHODS heart failure. Study medication was la- investigator could implement any new
Study Participants beled with sequential randomization treatments, except for open-label treat-
COPERNICUS study patients were en- numbers linked up to a block random- ment with a ␤-blocker. Following
rolled from 334 hospital centers and 21 ization scheme; at the randomization completion of the up-titration phase,
countries between October 28, 1997, visit, each patient was assigned the low- patients entered the maintenance phase
and March 20, 2000. Patients were eli- est number available at each site. The and continued in the double-blind
gible if they had dyspnea or fatigue at starting dosage was 3.125 mg of carve- therapy until the entire trial ended. The
rest or on minimal exertion for at least dilol or placebo twice daily, which if tol- COPERNICUS trial was stopped on
2 months and a left ventricular ejec- erated then was increased to 6.25 mg March 20, 2000, when the finding of a
tion fraction less than 25% due to is- twice daily after 2 weeks, to 12.5 mg marked beneficial effect of carvedilol on
chemic or nonischemic cardiomyop- twice daily after 4 weeks, and finally to survival led to a recommendation by the
athy. All patients were treated with a a target dosage of 25 mg twice daily or trial’s data and safety monitoring board
diuretic (which was adjusted to mini- placebo after 6 weeks. At each visit, pa- for early termination.19
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, February 12, 2003—Vol 289, No. 6 713

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

Statistical Analysis present report focused on the effects of were assessed in a very high risk sub-
A major clinical event was defined as treatment during initiation and up- group consisting of patients with re-
death, hospitalization, or permanent titration, the period of principal inter- cent or recurrent cardiac decompensa-
withdrawal of the study medication est was 8 weeks, which corresponded tion or very depressed cardiac function.
for any reason. Cumulative incidence to the expected duration of the up- These high-risk patients were charac-
curves for the occurrence of these events titration period. These specific out- terized by 1 or more of the following:
were constructed by the Kaplan-Meier comes and the 8-week period also were the presence of pulmonary rales, asci-
method, using a time-to-first-event ap- the focus of an earlier analysis of the up- tes, or edema at randomization; 3 or
proach.20 Cox proportional hazards re- titration period in the Metoprolol CR/XL more hospitalizations for heart failure
gression models were used to estimate Randomised Intervention Trial in Con- within the last year; hospitalization at
hazard ratios (HRs) and 95% confi- gestive Heart Failure Study Group.16 the time of screening or randomiza-
dence intervals (CIs).21 The analyses of Because earlier studies had raised tion; need for intravenous positive ino-
major outcome variables included all concerns that patients at highest risk tropic agent or vasodilator drug within
randomly assigned patients according to might respond poorly to ␤-adrenergic 14 days before randomization; or left
the intention-to-treat principle. Since the blockade,14,22 the effects of carvedilol ventricular ejection fraction of 15% or
less.19 The baseline variables that de-
fined this high-risk group were iden-
Figure 2. Effect of Carvedilol on Risk of Major Clinical Events in All Randomizly Assigned
Patients and in Patients at Highest Risk During First 8 Weeks and During the Entire Trial tified a priori without knowledge of
their influence on the treatment effect.
All Randomized Patients First 8 Weeks Entire Trial The safety of carvedilol was assessed
Placebo Carvedilol by changes in vital signs (summarized as
No./ Total Kaplan-Meier No./ Total Kaplan-Meier Hazard Favors Favors a mean [SE] change from baseline) and
Event Rate, % Event Rate, % Ratio Carvedilol Placebo by reports of adverse events with onset
(95% CI)
All-Cause Mortality within 8 weeks of randomization. All re-
First 8 Weeks 25/1133 2.3 19/1156 1.7 0.75 (0.41-1.35) ports of adverse events were included
Entire Trial 0.65 (0.52-0.81) whether or not they were deemed by the
Death or Hospitalization for Any Reason investigator to be related to treatment.
First 8 Weeks 153/1133 14.4 134/1156 12.3 0.85 (0.67-1.07)
Entire Trial 0.76 (0.67-0.87)
Adverse events with a frequency of at
Death, Hospitalization, or Permanent Study Drug
least 2% among all randomly assigned
Withdrawal for Any Reason patients in either treatment group, and
First 8 Weeks 188/1133 17.5 162/1156 14.8 0.83 (0.68-1.03) differences between treatment groups of
Entire Trial 0.76 (0.67-0.86)
at least 2% in the frequencies of the event
.05 0.1 1.0 2.0
were considered clinically significant. An
Hazard Ratio adverse event was defined in the study
protocol as serious if it was fatal or life-
High-Risk Patients threatening, required or prolonged hos-
Placebo Carvedilol pitalization, or resulted in persistent or
No./ Total Kaplan-Meier No./ Total Kaplan-Meier Hazard Favors Favors significant disability or incapacity. Sta-
Event Rate, % Event Rate, % Ratio Carvedilol Placebo
(95% CI)
tistical analyses were performed using
All-Cause Mortality SAS (versions 6.12 and 8.0; SAS Insti-
First 8 Weeks 15/316 5.0 3/308 1.0 0.20 (0.06-0.70) tute, Cary, NC).
Entire Trial 0.61 (0.41-0.89)
Death or Hospitalization for Any Reason RESULTS
First 8 Weeks 63/316 20.9 44/308 15.0 0.71 (0.48-1.04)
Entire Trial 0.71 (0.56-0.89)
Of the 2289 patients who were enrolled
Death, Hospitalization, or Permanent Study Drug
into the trial, 1133 were randomly as-
Withdrawal for Any Reason signed to the placebo group and 1156 to
First 8 Weeks 76/316 25.0 51/308 17.3 0.67 (0.47-0.96) the carvedilol group (FIGURE 1). Of these,
Entire Trial 0.68 (0.54-0.84)
624 (27.3%) patients fulfilled the crite-
.05 0.1 1.0 2.0 ria for recent or recurrent cardiac de-
Hazard Ratio compensation or very depressed car-
diac function, of whom 316 were
CI indicates confidence interval. Patients at highest risk are patients with recent or recurrent cardiac decom-
pensation or very depressed cardiac function and characterized by 1 or more of the following: the presence of
randomly assigned to placebo and 308
pulmonary rales, ascites, or edema at randomization; greater than 3 hospitalizations for heart failure within to carvedilol. As reported previously,18
the last year; hospitalization at the time of screening or randomization; need for intravenous positive inotropic the 2 treatment groups were similar with
agent or vasodilator drug within 14 days of randomization; or left ventricular ejection fraction less than 15%.
respect to all baseline characteristics. En-
714 JAMA, February 12, 2003—Vol 289, No. 6 (Reprinted) ©2003 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

rolled patients had a mean age of 63.3 Kaplan-Meier curves suggest that Changes in Vital Signs
years, a median left ventricular ejection the differences between the carvedilol During the First 8 Weeks
fraction of 20%; 79.7% were male and and placebo groups begin to appear as There were small changes in mean (SE)
67.2% had ischemic heart disease. early as 14 to 21 days following initia- systolic blood pressure (placebo group,
The majority of patients were suc- tion of treatment for both all-cause –2.0 [0.5] mm Hg and carvedilol group,
cessfully titrated to the target doses of mortality and for the combined end –3.6 [0.5] mm Hg) and in diastolic blood
the study medication specified for each point of death, hospitalization, or pressure (placebo group, –1.8 [0.3]
visit. At 2 weeks, 97.2% of placebo pa- withdrawal, in the analysis of all ran- mm Hg and carvedilol group, –2.7 [0.3]
tients and 97.1% of carvedilol patients domly assigned patients and in the mm Hg) at the end of 8 weeks. At 8
were receiving at least 3.125 mg twice analysis of patients at highest risk weeks, heart rate slowed progressively
daily. At 4 weeks, 87.6% of placebo pa- (FIGURE 3). in the carvedilol group as the dose of the
tients and 84.0% of carvedilol patients
were receiving at least 6.25 mg twice
daily. At 6 weeks, 79.1% of placebo pa- Figure 3. Kaplan-Meier Analysis for All Randomly Assigned Patients and for Patients at
tients and 71.7% of carvedilol patients Highest Risk During the First 8 Weeks Following Initiation of Therapy
were receiving at least 12.5 mg twice All Randomized Patients
daily. At 8 weeks, 70.9% of placebo pa- All-Cause Mortality Death, Hospitalization,
tients and 58.6% of carvedilol patients or Study Drug Withdrawal
were receiving 25 mg twice daily. The 3 20

mean dosages of placebo at 2, 4, 6 and

8 weeks were 3.5, 6.7, 12.5, and 19.8 15
% of Patients With Event

Placebo Placebo
mg twice daily, respectively; the mean 2
dosages of carvedilol at 2, 4, 6, and 8
10
weeks were 3.5, 6.5, 11.6, and 17.8 mg
twice daily, respectively. 1
Carvedilol Carvedilol
5
Death, Hospitalization,
or Permanent Withdrawal
0 Hazard Ratio (95% CI), 0.75 (0.41-1.35) 0 Hazard Ratio (95% CI), 0.83 (0.68-1.03)
During First 8 Weeks
During the first 8 weeks, the carvedilol 0 2 4 6 8 0 2 4 6 8
Weeks After Randomization Weeks After Randomization
group, compared with the placebo group, No. at Risk
Placebo 1133 1100 1054 1023 986 1133 1071 970 896 837
had fewer patients with a major clinical Carvedilol 1156 1119 1079 1048 1009 1156 1087 1005 944 876
event and had fewer patients who died,
who died or were hospitalized, or who High-Risk Patients
died, were hospitalized, or were perma- All-Cause Mortality Death, Hospitalization,
nently withdrawn from double-blind or Study Drug Withdrawal
treatment (FIGURE 2). The direction and 6 30

magnitude of these effects during the first

8 weeks were similar to those observed
% of Patients With Event

Placebo
during the entire study. 4 Placebo 20
Similar effects were observed for all 3
end points when the analyses were con-
fined to patients at highest risk, that is, 2 10
those patients with recent or recurrent Carvedilol
Carvedilol
decompensation or a very depressed left
ventricular ejection fraction (Figure 2). 0 0
Hazard Ratio (95% CI), 0.20 (0.06-0.70) Hazard Ratio (95% CI), 0.67 (0.47-0.96)
The carvedilol group , when compared
with the placebo group, had a lower risk 0 2 4 6 8 0 2 4 6 8
Weeks After Randomization Weeks After Randomization
of death, of death or hospitalization, and No. at Risk
Placebo 316 309 295 281 270 316 294 257 227 214
of death, hospitalization, or with- Carvedilol 308 299 290 282 268 308 286 258 241 221
drawal of double-blind treatment. Again,
Patients at highest risk are patients with recent or recurrent cardiac decompensation or very depressed cardiac
the direction and magnitude of these ef- function and characterized by 1 or more of the following: the presence of pulmonary rales, ascites, or edema
fects seen in this cohort during the first at randomization; greater than 3 hospitalizations for heart failure within the last year; hospitalization at the
8 weeks were similar to those observed time of screening or randomization; need for intravenous positive inotropic agent or vasodilator drug within
14 days of randomization; or left ventricular ejection fraction less than 15%.
during the entire study.
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, February 12, 2003—Vol 289, No. 6 715

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

study medication and duration of treat- any reason other than death. There was quency greater than 2%, namely wors-
ment increased (placebo group, − 2.2 no difference between placebo and ening heart failure, and it was reported
[0.4] bpm and carvedilol group, −12.5 carvedilol in the number of patients with a similar frequency with placebo
[0.4] bpm), but body weight did not withdrawn for worsening heart failure and carvedilol (6.4% vs 5.1%, respec-
change in either group (placebo group, (0.7% vs 0.6%, respectively, for all pa- tively, among all randomly assigned pa-
0 [0.07] kg and carvedilol group, 0.1 tients and 1.9% vs 1.6%, respectively, tients; and 11.4% and 8.8%, respec-
[0.07] kg). for highest risk patients). In addition, tively, in patients at highest risk).
fewer patients in the carvedilol group Patients in the carvedilol group were
Adverse Events During than in the placebo group experi- more likely than in the placebo group
the First 8 Weeks enced a serious adverse event (13.8% to report dizziness, hypotension, edema,
Overall, 59 patients (5.2%) in the pla- vs 15.0% among all randomly as- and bradycardia (TABLE). In general,
cebo group and 51 patients (4.4%) in signed patients and 16.2% vs 22.2% these reactions were not considered se-
the carvedilol group permanently with- among high-risk patients). Only 1 se- rious, but in a small number of cases
drew from double-blind medication for rious adverse event occurred with a fre- required withdrawal of double-blind
medication. Clinically significant dif-
ferences (ie, greater than 2% differ-
Table. Adverse Events During the First 8 Weeks* ence) between the treatment groups
No. (%) were not observed during the up-
Placebo Carvedilol titration period for any other adverse
All Randomized Patients event.
(n = 1133) (n = 1156)
Bradycardia COMMENT
Serious adverse event 1 (0.1) 10 (0.9)
Many physicians believe that for pa-
Trial drug decreased due to adverse event 3 (0.3) 27 (2.3)
tients with severe chronic heart fail-
Withdrawn due to adverse event 0 4 (0.3)
Dizziness
ure to experience the long-term ben-
Serious adverse event 4 (0.4) 5 (0.4) efits of ␤-adrenergic blockade, they
Trial drug decreased due to adverse event 22 (1.9) 67 (5.8) must undergo a period of initiation and
Withdrawn due to adverse event 3 (0.3) 10 (0.9) up-titration that may be trouble-
Edema some.8 There is concern that the with-
Serious adverse event 5 (0.4) 3 (0.3) drawal of sympathetically mediated ino-
Trial drug decreased due to adverse event 3 (0.3) 11 (1.0) tropic support following the start of
Withdrawn due to adverse event 2 (0.2) 1 (0.1) treatment with a ␤-blocker carries a
Hypotension high risk of worsening heart failure, pul-
Serious adverse event 2 (0.2) 6 (0.5)
Trial drug decreased due to adverse event 11 (1.0) 38 (3.3)
monary edema, or cardiogenic shock.
Withdrawn due to adverse event 0 (0.0) 5 (0.4)
Patients with severe heart failure are
considered most likely to experience
Patients at Highest Risk
early worsening and delayed benefit of
(n = 316) (n = 308)
Bradycardia treatment,8,9,13,14 since such individu-
Serious adverse event 0 3 (1.0) als show the most marked activation of
Trial drug decreased due to adverse event 2 (0.6) 8 (2.6) the sympathetic nervous system23 and
Withdrawn due to adverse event 0 2 (0.6) are assumed to be the most dependent
Dizziness on adrenergically mediated circula-
Serious adverse event 1 (0.3) 0
tory support.24
Trial drug decreased due to adverse event 7 (2.2) 17 (5.5)
The findings of the COPERNICUS
Withdrawn due to adverse event 2 (0.6) 2 (0.6)
Edema
study with carvedilol challenge beliefs
Serious adverse event 1 (0.3) 2 (0.6) about the efficacy and safety of ␤-block-
Trial drug decreased due to adverse event 0 3 (1.0) ade during the first several weeks of treat-
Withdrawn due to adverse event 0 1 (0.3) ment. During both initiation and up-
Hypotension titration, patients treated with carvedilol
Serious adverse event 1 (0.3) 1 (0.3) had no increase in the risk of worsen-
Trial drug decreased due to adverse event 3 (0.9) 14 (4.5) ing heart failure, pulmonary edema, car-
Withdrawn due to adverse event 0 2 (0.6) diogenic shock, or other serious ad-
*Includes adverse events with a frequency of at least 2% in all randomly assigned patients in either treatment group
and differences between treatment groups in the frequency of the event of at least 2%. An adverse event was de- verse cardiovascular events, including
fined in the study protocol as serious if it was fatal or life-threatening; required or prolonged hospitalization or re- death. The principal adverse events at-
sulted in persistent or significant disability or incapacity.
tributable to carvedilol during the first
716 JAMA, February 12, 2003—Vol 289, No. 6 (Reprinted) ©2003 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

8 weeks of therapy were those expected treatment of heart failure, and they se- Tendera, Coats, Katus, Fowler, and Packer have served
as consultants for Roche Pharmaceuticals and/or Glaxo-
as a result of the inhibitory effects of the lected patients carefully and followed SmithKline Ltd. Dr Roecker has received salary sup-
drug on ␣-receptors (dizziness and hy- them closely during the course of the port from a research contract with GlaxoSmithKline
potension) and ␤-receptors (bradycar- trial. Furthermore, the protocol speci- Ltd. Dr Fowler has received honararia from Glaxo-
SmithKline Ltd, Roche Pharmaceuticals, Astra Zen-
dia and peripheral edema).10,24 How- fied that patients were to be clinically eca and has served as a consultant to Bristol-Meyers
ever, these adverse reactions were mild euvolemic before they were randomly Squibb, GlaxoSmithKline Ltd, and Scios Inc. Dr Packer
has served as a consultant to GlaxoSmithKline Ltd and
and infrequent, occurring in 3 to 7 more assigned, and every effort was made to Roche Pharmeceuticals.
patients in the carvedilol group per 100 maintain euvolemia during initiation Funding/Support: This study was supported by grants
from Roche Pharmaceuticals and Glaxo SmithKline Ltd.
patients treated. Importantly, because and up-titration of the study medica- Acknowledgment: We thank Christoph Staiger, MD,
these adverse reactions were self- tion. Patients were encouraged to re- Ildiko Amann-Zalan, MD, and Diethelm Messinger,
MS, of Roche Pharmaceuticals; Ellen L. Curtin, MD,
limited and not considered serious, they port any adverse effects or weight gain, Terry L. Holcslaw, PhD, and Neil Shusterman, MD, of
rarely led to the discontinuation of ef- and the dose of other medications could GlaxoSmithKline, Ltd; and Melissa K. Schultz, MS, and
fective treatment. be modified or the rapidity of upward Barbara Kowalcyk, MS, of the University of Wiscon-
sin for their invaluable contributions to the study.
The results of the COPERNICUS titration of the dose of the study drug COPERNICUS study group coordinators: Steering
study also challenge the belief that the could be decreased, if such adjust- Committee: M. Packer (Chair), A. Castaigne, A. Coats,
M. Fowler, H. Katus, H. Krum, P. Mohacsi, J-L. Rou-
benefits of ␤-adrenergic blockade in pa- ments were clinically warranted. These leau, M. Tendera. Data and Safety Monitoring Board:
tients with heart failure are delayed. Dur- approaches will need to be followed in K. Swedberg (Chair), E. Angermann, R. Campbell (de-
ing the first 8 weeks of treatment, fewer clinical practice; similar precautions ceased), J. Cohn, A. Maseri, S. Pocock. Biostatistics
Center: D. DeMets, E. Roecker, M. Schultz. End-
patients died or were hospitalized in the have been recommended for general use point Committee: P. Carson (Chair), V. Bernstein, C.
carvedilol group, and the magnitude of in recent guidelines.28 O’Connor, M. Haass, V. Mareev, A. Miller, S. Per-
rone, B. Rauch, G. Sutton. Roche/GlaxoSmithKline Op-
risk reduction by carvedilol during this In conclusion, the relation of ben- erating Committee: I. Amann-Zalan, E. Curtin, M.
early phase of therapy was similar to that efit to risk during initiation of treat- Harsch, T. Holcslaw, E. Kroener-Bentel, D. Mess-
inger, C. Staiger.
seen during the entire study. The abil- ment with carvedilol is similar to that List of investigators: Argentina: F. Diez, E. Kuschnir,
ity of carvedilol to produce beneficial ef- seen during long-term treatment with S. Perrone. Australia: P. Garrahy, J. Horowitz, I. Jef-
fects early during the course of treat- the drug. In clinically euvolemic fery, J. Karrasch, H. Krum, P. McDonald, J. Waites.
Austria: B. Eber, F. Schmalzl, J. Slany, R. Spinka,
ment was particularly striking in the patients with advanced heart failure, W. Weihs. Canada: P. Alain, M. Arnold, R. Baigrie,
patients at highest risk, that is, those with initiation of treatment with carvedilol M. Bentley-Taylor, J. Bonet, J. Champagne, P. Costi,
T. Cuddy, D. Dion, D. Fell, D. Gossard, M. Gupta,
recent or recurrent decompensation or was well-tolerated and was associated W. Hui, J. Howlett, D. Humen, J. Hynd, T. Kashour,
a very depressed left ventricular ejec- with fewer major adverse events than M. Khouri, P. Klinke, S. Kouz, M. Langlais, M. H. Le-
tion fraction. These observations indi- initiation of treatment with placebo. If blanc, S. Lepage, B. Lubelsky, D. Manyari, M. Matangi,
G. Moe, A. Morris, J. Nasmith, M. Palaic, P. Pflug-
cate that the clinical benefits of sympa- concerns about efficacy and safety felder, D. C. Phaneuf, A. Rajakumar, T. Rebane, J. Ricci,
thetic antagonism are not necessarily during the initiation of ␤-blocker J. Rouleau, F. Sestier, S. Smith, J. Stone, P. Talbot, M.
White. Czech Republic: P. Bocek, I. Gajdosová, J. Gre-
delayed and suggest that the mecha- therapy have caused physicians to gor, P. Gregor, I. Kotik, A. Linhart, J. Lukl, P. Petr,
nisms by which such benefits are me- deny or delay the use of these drugs, J. Popelova, B. Semrad, V. Stanek, R. Stipal. France:
A. Gabriel, J. L. Guermonprez, G. Mougeot, J. Puel,
diated are not of necessity dependent on our findings should provide the reas- R. Roudaut. Germany: T. Beyer, A. Costard-Jäckle, W.
changes in left ventricular function or surance needed to encourage the high Döring, F. Freytag, H. Katus, H. Koch, F. Menzel,
geometry, which are known to require levels of use that are warranted by the S. Peters, U. Sechtem, W. Sehnert, H. F. Vöhringer, E.
Wunderlich, H. Zebe, R. Zotz. Great Britain: R. Bain,
months to become apparent.15,25 It also results of clinical trials. P. Bennett, A. Coats, D. Davies, S. Gibbs, T. Green-
is noteworthy that the Kaplan-Meier wood, M. Heber, A. Lahiri, R. Mattu, J. McComb, I.
Author Contributions: Dr Roecker has had full ac- McLay, D. Nichols, R. Northcote, B. Silke, S. Ste-
curves for the placebo and carvedilol cess to the data for the COPERNICUS study and takes phens, J. Swan, C. Weston. Hungary: M. Csanády,
groups (Figure 3) began to separate af- responsibility for the accuracy of the data analysis pre- L. Cserhalmi, I. Édes, T. Gesztesi, E. Kaló, A. Katona,
sented in this article. A. Jánosy, F. Poór, M. Rusznák, K. Simon, F. Szabóki,
ter about 21 days of treatment; this was J. Tarján, J. Tenczer, S. Timár, P. Vályi, K. Zámoly. Israel:
Study Concept and Design:Krum, Roecker, Mohacsi,
at a time when patients were generally Rouleau, Tendera, Coats, Katus, Fowler, Packer. G. Avinader, A. Caspi, A. Darausha, D. David, Y. Kishon,
receiving a dosage of only 6.25 mg of Acquisition of Data: Krum, Mohacsi, Rouleau, Tendera, E. Klainman, B. Lewis, A. Marmor, M. Mitelman, M.
Coats, Katus, Fowler, Packer. Omary, L. Reisin, T. Rosenfeld, S. Shasha,
carvedilol twice daily. This finding in Analysis and Interpretation Data: Krum, Roecker, Z. Vered, R. Zimlichman. Italy: E. Arosio, A. Branzi,
patients with severe heart failure is Mohacsi, Rouleau, Tendera, Coats, Katus, Fowler, C. Campana, M. Casaccia, L. Dei Cas, A. Di Lenarda,
Packer. P. Fioretti, M. Frigerio, A. L’Abbate, M. Modena.
consistent with the results of an earlier Drafting of Manuscript: Krum, Roecker, Packer. Lithuania: A. Kibarskis, P. Serpytis, D. Vasiliauskas,
study, which showed that even Critical Revision of Manuscript for Important Intel- P. Zabiela. Mexico: N. Garcia-Hernández. The Neth-
6.25 mg of carvedilol twice daily was lectual Content: Krum, Roeker, Mohacsi, Rouleau, erlands: R. Breedveld, J. Cornel, M. Daniels, P. Dun-
Tendera, Coats, Katus, Fowler, Packer. selman, B. Hamer, L. van Kempen, G. Linssen, A. Maas,
effective in patients with mild-to- Statistical Expertise: Roecker. P. de Milliano, S. Twisk, A. Willems. Poland: L. Cer-
moderate symptoms.26,27 Obtained Funding: Packer. emuzynski, A. Cieslinski, M. Dalkowski, J. Dubiel, B.
Administrative, Technical or Material Support: Krum, Filipek, H. Halaczkiewicz, M. Janion, K. Kawecka-
The findings of the present study Roeker, Mohacsi, Rouleau, Tendera, Coats, Katus, Jaszcz, M. Krzeminska-Pakula, B. Kusnierz, K. Loboz-
should be interpreted in light of the fact Fowler, Packer. Grudzien, A. Malinski, T. Mandecki, W. Musial, W.
Study Supervision: Krum, Mohacsi, Rouleau, Tendera, Piotrowski, W. Pluta, W. Prastowski, W. Ruminski, A.
that the investigators and coordina- Coats, Katus, Fowler, Packer. Rynkiewicz, W. Smielak-Korombel, M. Tendera, R. Tro-
tors were highly experienced in the Financial Disclosures: Drs Krum, Mohacsi, Rouleau, jnar, M. Ujda, J. Wodniecki, K. Wrabec, M. Zalewski.

©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, February 12, 2003—Vol 289, No. 6 717

Downloaded From: https://jamanetwork.com/ on 11/06/2023

 INITIATING CARVEDILOL IN PATIENTS WITH SEVERE CHF

Portugal: M. Carrageta, R. Seabra-Gomes. Russia: field, J. Caplan, P. Carson, E. Carter, L. Christie, D. R. Kohn, M. Koren, D. Korn, K. Labresh, G. Lamas, L.
G. Arutyunov, R. Charchoglian, A. Gruzdev, A. Ivleva, Chromsky, M. Cishek, V. Corrigan, M. Costanza, Lancaster, C. Lawless, T. LeJemtel, C. Liang, G. Lit-
Y. Karpov, V. Kostenko, V. Mareev, V. Moisejev, C. Curry, J. Davia, P. Deedwania, E. de Marchena, G. man, E. Loh, B. Lorell, G. Luckesen, E. MacInerney, B.
L. Oblinskaya, V. Orlov, N. Perepech, E. Shlyhatkho, Dennis, R. DiBianco, S. Dunlap, E. Eichhorn, U. Elkayam, Massie, M. Mathier, F. McGrew, M. McIvor, H.
B. Sidorenko, A. Smirnov, A. Starodubsev, G. Storazha- J. English, N. Erenrich, C. Fallick, R. Feldman, D. Ferry, Meilman, F. Messineo, S. Meymandi, A. Miller, P.
kov. South Africa: P. Jordaan, P. Manga, D. Naidoo, D. Fishbein, L. Ford, D. Forman, M. Fowler, J. Ghali, Mohanty, J. Morledge, J. Neutel, M. Nocero, A. Onwua-
I. Radevski, N. Ranjith. Switzerland: B. Caduff, E. Gilbert, R. Gillespie, M. Givertz, S. Goldman, D. Gold- nyi, S. Oparil, R. Oren, G. Pennock, A. Poppas, C. Por-
P. Mohacsi, C. Röthlisberger, F. Widmer. Ukraine: scher, S. Goldsmith, R. Gordon, A. Gradman, B. Green- ter, C. Ramanathan, H. K. Reddy, R. Reeves, S. Rob-
E. Amosova, G. Dzyak, G. Knyshov, V. Kovalenko, berg, G. Hamroff, H. Haught, P. Hauptman, C. Heesch, erts, S. Restaino, R. Schwartz, R. Schneider, A. Seals,
V. Netyazhenko, S. Pavlyk, N. Seredjuk, Y. Serenko, T. Heywood, M. Higginbotham, R. Hobbs, R. Siegel, A. Smith, E. Smith, R. Smith, W. Smith, T.
L. Voronkov, A. Zmuro. United States: K. Aaronson, J. Hosenpud, C. Hunter, M. James, M. Johnson, J. Kal- Spaedy, L. Stevenson, S. Stowers, S. Teague, G. Tim-
W. Abraham, J. Alexander, J. Allen, J. Anderson, J. Ber- man, R. Karlsberg, E. Kasper, D. Kereiakes, V. Kinhal, mis, M. Tischler, N. Vijay, J. Walker, M. Walsh, C.
gin, P. Berman, P. Binkley, N. Bittar, J. Bowers, L. Brook- R. Kipperman, J. Kirkpatrick, P. Kirlin, M. Klapholz, Weaver, D. Weisshaar, V. Wilson.

REFERENCES
1. Packer M, Bristow MR, Cohn JN, et al. The effect of 10. Epstein SE, Braunwald E. The effect of beta- patients with advanced chronic heart failure. N Engl
carvedilol on morbidity and mortality in patients with adrenergic blockade on patterns of urinary sodium J Med. 2001;344:1659-1667.
chronic heart failure. N Engl J Med. 1996;334:1349- excretion: studies in normal subjects and in patients 19. Packer M, Coats AJ, Fowler MB, et al. Effect of
1355. with heart disease. Ann Intern Med. 1966;75:20- carvedilol on survival in severe chronic heart failure.
2. CIBIS II Investigators and Committees. The Car- 27. N Engl J Med. 2001;344:1651-1658.
diac Insufficiency Bisoprolol Study (CIBIS II): a ran- 11. Packer M, Cohn JN, eds. Consensus recommen- 20. Kaplan EL, Meier P. Nonparametric estimation from
domised trial. Lancet. 1999;353:9-13. dations for the treatment of chronic heart failure. Am incomplete observations. J Am Stat Assoc. 1958;53:
3. Metoprolol CR/XL Randomised Intervention Trial J Cardiol. 1999;83(suppl 2A):1A-38A. 457-481.
in Congestive Heart Failure (MERIT-HF) Study Group. 12. The RESOLVD Investigators. Effects of metopro- 21. Cox DR. Regression models and life-tables.
Effect of metoprolol CR/XL in chronic heart failure. lol CR in patients with ischemic and dilated cardio- J R Stat Soc [B]. 1972;34:187-202.
Lancet. 1999;353:2001-2007. myopathy. Circulation. 2000;101:378-384. 22. Lechat P, Bouzamondo A, Sanchez P, Hulot JS,
4. Hjalmarson A, Goldstein S, Fagerberg B, et al. Ef- 13. Sackner-Bernstein J, Krum H, Goldsmith RL, et al. Eichhorn EJ, Cucherat M. Relationships between base-
fects of controlled-release metoprolol on total mor- Should worsening heart failure early after initiation of line risk and treatment effect of beta-blockers in heart
tality, hospitalizations, and well-being in patients beta-blocker therapy for chronic heart failure pre- failure [abstract]. Eur Heart J. 2000;21:297A.
with heart failure: the Metoprolol CR/XL Random- clude long-term treatment? [abstract]. Circulation. 23. Cohn JN, Levine B, Olivari MT, et al. Plasma nor-
ized Intervention Trial in Congestive Heart Failure 1995;92(suppl I):I-395. epinephrine as a guide to prognosis in patients with
(MERIT-HF). JAMA. 2000;283:1295-1302. 14. Anderson JL, Krause-Steinrauf H, Goldman S, et chronic congestive heart failure. N Engl J Med. 1984;
5. Packer M, Colucci WS, Sackner-Bernstein JD, et al. al. Failure of benefit and early hazard of bucindolol in 311:819-823.
Double-blind, placebo-controlled study of the ef- class IV heart failure [abstract]. Circulation. 2000;102 24. Gaffney TE, Braunwald E. Importance of adren-
fects of carvedilol in patients with moderate to se- (suppl II):II779-II780. ergic nervous system in the support of circulatory func-
vere heart failure. Circulation. 1996;94:2793-2799. 15. Hall SA, Cigarroa CG, Marcoux L, Risser RC, Gray- tion in patients with congestive heart failure. Am J Med.
6. Colucci WS, Packer M, Bristow MR, et al. Carve- burn PA, Eichhorn EJ. Time course of improvement in 1963;34:320-324.
dilol inhibits clinical progression in patients with mild left ventricular function, mass and geometry in pa- 25. Australia-New Zealand Heart Failure Research Col-
symptoms of heart failure. Circulation. 1996;94:2800- tients with congestive heart failure treated with beta- laborative Group. Randomized, placebo-controlled trial
2806. adrenergic blockade. J Am Coll Cardiol. 1995;25: of carvedilol in patients with congestive heart failure due
7. Krum H, Tonkin AM, Currie R, Djundjek R, Johns- 1154-1161. to ischemic heart disease. Lancet. 1997;349:375-380.
ton CI. Chronic heart failure in Australian general prac- 16. Gottlieb SS, Fischer ML, Kjekshus J, et al. MERIT- 26. Bristow M, Gilbert EM, Abraham WT, et al. Carve-
tice: The Cardiac Awareness Survey and Evaluation HF: tolerability of ␤-blocker initiation NYHA II, III, and dilol produces dose-related improvements in left ven-
(CASE) study. Med J Aust. 2001;174:439-444. IV CHF [abstract]. Circulation. 2000;102(suppl II):II- tricular function and survival in subjects with chronic
8. Eichhorn EJ, Bristow MR. Practical guidelines for ini- 778. heart failure. Circulation. 1996;94:2807-2816.
tiation of beta-adrenergic blockade in patients with 17. Gottlieb SS, Fischer ML, Kjekshus J, et al. Toler- 27. Packer M. Current role of beta-adrenergic block-
chronic heart failure. Am J Cardiol. 1997;79:794-798. ability of beta-blocker initiation and titration in the ers in the management of chronic heart failure. Am
9. Waagstein F, Caidahl K, Wallentin I, Bergh C, Hjal- Metoprolol CR/XL Randomised Intervention Trial in J Med. 2001;110(suppl 7A):81S-94S.
marson A. Long term a-blockade in dilated cardiomy- Congestive Heart Failure (MERIT-HF). Circulation. 28. Hunt SA, Baker DW, Chin MH, et al. ACC/AHA
opathy: effects of short- and long-term metoprolol 2002;105:1182-1188. guidelines for the evaluation and management of
treatment followed by withdrawal and readministra- 18. The Beta-Blocker Evaluation of Survival Trial In- chronic heart failure in the adult. Circulation. 2001;
tion of metoprolol. Circulation. 1989;80:551-563. vestigators. A trial of the beta-blocker bucindolol in 104:2996-3007.

Every quotation contributes something to the stabil-

ity or enlargement of the language.
—Samuel Johnson (1709-1784)

718 JAMA, February 12, 2003—Vol 289, No. 6 (Reprinted) ©2003 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ on 11/06/2023