The Pediatric Randomized Carvedilol Trial in Children with Heart Failure

Linda Brookes, MSc


July 20, 2006

Editorial Collaboration

Medscape &

Presenter: Robert E Shaddy, MD (University of Utah, Salt Lake City)

Carvedilol, the nonselective beta1,2- and alpha1-adrenergic receptor blocker, is approved in the United States, Europe, and other parts of the world for treatment of chronic heart failure (HF) in adults, but its use in children to date has been based on the findings of adult clinical trial studies and only anecdotal and uncontrolled data in younger patients. The challenges that have hindered HF trials in children are the smaller numbers of patients, the lack of endpoints for pediatric HF studies, the widespread reluctance of investigators, parents, and institutional review boards (IRBs) to participate in placebo-controlled trials, and the unknown history of HF in children enrolled in clinical trials.

However, a number of reports documenting the use of carvedilol in children with HF have reported improvement in functional status, and its efficacy in the pediatric population has been regarded as having been confirmed.[1] It was therefore surprising to many pediatric cardiologists and other specialists to learn that the results of the first randomized controlled clinical trial of any chronic HF therapy in children, carried out with carvedilol, has failed to show any treatment benefit in patients up to 17 years of age.[2] The trial was carried out in children with symptomatic, systemic ventricular systolic dysfunction, using a composite measure of clinical outcomes after 8 months of treatment. The failure to show a treatment response in the trial was mainly due to a substantially larger placebo improvement response than expected.

Study Design

The study was a prospective, randomized, double-blind, placebo-controlled, parallel-group, multicenter trial in which patients were randomized 1:1:1 to twice-daily placebo, low-dose carvedilol (target dose 0.2 mg/kg/dose or 12.5 mg/dose for subjects weighing > 62.5 kg), or high-dose carvedilol (target dose 0.4 mg/kg/dose or 25 mg/dose for subjects > 62.5 kg).[3]


Male and female children from birth through age 17 years were eligible for the trial. Other inclusion criteria were:

  • Chronic symptomatic HF resulting from systemic ventricular systolic dysfunction due to dilated cardiomyopathy or congenital heart disease and receiving standard HF therapy such as diuretic, digoxin, or angiotensin-converting enzyme (ACE) inhibitor (all patients had to be on ACE inhibitors unless contraindicated or intolerant);

  • Diagnosis of New York Heart Association (NYHA) class II-IV HF (older children) or Ross class II-IV HF (infants and younger children) for at least 1 month;

  • Estimated ejection fraction (EF) < 40% in patients with systemic left ventricular (LV) dysfunction, or qualitative evidence of a dilated ventricle with moderate systemic systolic dysfunction in patients with right ventricular or single ventricular physiology (non-left ventricle; NLV).

A total of 161 children were enrolled at 26 US centers over a period of 4 years: 48% were male, and median age was 3 years (range, 3 months to 17 years, with 45% aged < 2 years). Systemic ventricular disease morphology was LV in 74% and NLV in 26%. HF class was NYHA class II in 71% and class III in 27%. Median EF was 26%, and median plasma brain natriuretic peptide (BNP) was 111 pg/mL.

Forty-four patients (80%) in the placebo group and 82 (77%) in the combined carvedilol group completed the 8-month study. Sixteen percent of each group withdrew due to adverse events, of whom 5 (9%) of the patients on placebo and 13 (12%) of those on carvedilol underwent heart transplantation.

Primary Endpoint

The primary objective of this study is to evaluate the efficacy of carvedilol administered twice daily for 8 months in terms of its effect compared with placebo on a composite measure of clinical outcomes in children with symptomatic systemic systolic dysfunction and HF. Because no primary outcome measure has been validated in children with HF, the investigators selected a composite that included measurement of functional assessment, patient/parent global assessment, and major clinical events as the primary endpoint of the study. This composite measure of clinical chronic HF outcomes after 8 months was designated "worsened," "improved," or "unchanged," based on the following definitions:

  • Death;

  • Hospitalization for at least 24 hours for worsening HF requiring intravenous medication;

  • Withdrawal due to worsening HF, treatment failure, or lack or insufficiency of therapeutic response;

  • Discontinuation due to withdrawal of consent or other administrative reason with worsening HF at the time of withdrawal;

  • Worsening in NYHA or Ross HF class and/or moderate to marked worsening in the global assessment score

  • No worsening (as defined above), with improved NYHA class or Ross class of HF or moderate to marked improvement in the global assessment score.

  • Patient neither improved nor worsened.

On the basis of these definitions, there was no statistically significant difference (P = .74) between the placebo group and the combined carvedilol group in the percentage of patients who improved, remained unchanged, or worsened during the study. The 55.6% rate of improvement seen in the placebo group was similar to that of the combined carvedilol group (56.3%) (Table 1).

Table 1. Analysis of HF Composite Outcome
Placebo Carvedilol
Improved (%) 55.6 56.3
Unchanged (%) 14.8 19.4
Worsened (%) 29.6 24.3
HF = heart failure

Since at the time of randomization patients were stratified into study treatment arms by ventricular morphology, a prespecified secondary analysis was performed of the interaction between treatment and ventricular morphology (116 LV patients vs 41 NLV patients). This analysis showed a significant interaction between treatment and LV status (P = .02), with 51% of the LV patients on placebo improved compared with 64% of those on carvedilol (OR 0.28, 95% CI 0.07-1.12), whereas 66% of the NLV placebo patients improved compared with only 35% of those on carvedilol (OR 1.71, 95% CI 0.78-3.76). However, the study was not powered to detect differences within the strata.

Secondary Endpoints

The secondary objectives of the study were to determine the effect of carvedilol on individual components of the composite of clinical outcomes (hospitalizations for worsening HF, all-cause mortality and cardiovascular hospitalizations, HF symptoms, and patient and physician global assessment); to determine the effect of carvedilol on echocardiographic indices of ventricular function and remodeling; to characterize the pharmacokinetics of carvedilol in pediatric patients with HF; and to characterize the effects of carvedilol on neurohormonal systems.

No significant differences were seen between the carvedilol and placebo groups in terms of mortality or hospitalizations since the study was not powered for these endpoints, although the hazard ratios for each endpoint favored carvedilol (Table 2).

Table 2. Mortality and Hospitalizations
All-cause mortality 0.68 .53
Cardiovascular mortality 0.59 .59
All-cause mortality or HF hospitalization 0.75 .76
All-cause mortality or cardiovascular hospitalization 0.80 .80
HF = heart failure; HR = hazard ratio

An increase in EF was seen in each group (8.0 EF units with placebo, 10.1 EF units with low-dose carvedilol, and 12.9 EF units with high-dose carvedilol). Within each group, these differences were statistically different compared with screening (P < .001); however, there was no statistical difference between the 3 groups (P = .09) (Table 3).

Table 3. Ejection Fraction
Placebo Carvedilol
Low-dose High-dose
Screening (%) 25.4 27.4 26.8
Endpoint (%) 33.3* 37.6* 39.7*
*P < .001 vs screening.

Mean plasma BNP levels were similar in the 3 groups, but only the placebo group showed a statistically significant decrease between screening and follow-up.


At least 1 adverse event was reported in 98% of patients on placebo and in 88% of those on carvedilol. The most common events were upper respiratory tract infection, vomiting, and cough. The most common cardiovascular event was worsening HF, which occurred in 22% of those on placebo and 12% of those on carvedilol. Rates of study withdrawal due to adverse events were similar in the placebo and carvedilol groups (13%). HF worsening was the most common event leading to withdrawal (13% of those on placebo and 11% of those on carvedilol withdrew from the study due to HF worsening).


Commenting on the unexpected results, particularly the high spontaneous improvement rate (55.6%) in the placebo group, Dr. Shaddy suggested that a much larger study will be required to detect a difference between the drug and placebo groups. This high level of response to placebo, as well as the finding of the interaction between systemic ventricular anatomy and the primary outcome, are important considerations for the design of future clinical trials, he stressed. Other reasons why this trial may not have shown any treatment benefit for carvedilol, he suggested, might be the high incidence of NYHA class II HF patients enrolled (71%). He also noted the slow rate of enrollment -- 161 subjects over 4 years.

  1. Buck ML. Use of carvedilol in children with cardiac failure. Pediatr Pharm. 2005;11(2). 2005 Children's Medical Center, University of Virginia. Available online at:

  2. Shaddy R, Boucek M, Hsu D, et al; Pediatric Carvedilol Study Group. Multicenter, randomized, placebo-controlled, double-blind trial of carvedilol in children with heart failure. Program and abstracts of the American College of Cardiology 55th Annual Scientific Session; March 11-14, 2006; Atlanta, Georgia. Smaller Trial Late-Breaking Clinical Trials II. Abstract 418-6.

  3. Shaddy RE, Curtin EL, Sower B, et al; The Pediatric Randomized Carvedilol Trial in Children with Heart Failure Study. The pediatric randomized carvedilol trial in children with chronic heart failure: Rationale and design. Am Heart J. 2002;144:383-389. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.