Review of the CIBIS-II Trial

Bisoprolol in systolic heart failure

The Lancet Volume 353, Issue 9146 p9-13 January 02, 1999

Background: Accumulating data at the time suggested functional benefits of antagonism of beta-adrenoreceptors in patients with heart failure. Multiple specific beta-blockers were being tested in trials. The CIBIS 1 trial found a trend towards 20% lower mortality in the bisoprolol (a highly cardio-selective beta-blocker) group and 30% fewer admissions to hospital for worsening heart failure. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II) trial was designed to test this evidence further.

Patients Eligible patients had New York Heart Association Class III-IV symptoms with LVEF ≤ 35% and were stable on diuretics and ACE-inhibitors. Exclusion criteria included recent MI or coronary intervention, AV block or resting heart rate less 60 bpm and systolic BP < 100 mmHg. Patients already on beta-blockers or with planned therapy with beta-blockers were also not enrolled.

Baseline Characteristics The mean age of patients was 61 years, 81% male, and 83% Class III. The mean LVEF was 28%. About half the patients had ischemic heart disease, 12% primary dilated cardiomyopathy and nearly 40% had a mixture of valvular heart disease, hypertensive heart disease or unproven ischemic disease.

The mean SBP on enrollment was 130 mmHg and resting HR was 80 bpm. The mean duration of heart failure before enrollment was 3.5 years. About 20% had AF at baseline. Nearly all patients were on ACE-I and half were on digoxin.

Trial Procedures There was no run-in period. CIBIS II was double blinded. Slightly more than 2,600 patients were randomized 1:1 to bisoprolol or placebo in 274 hospitals across 18 countries.

Patients in the bisoprolol group were started at 1.25 mg daily and titrated up weekly to as high as 10 mg daily. The goal was to attempt the highest tolerated dose. Patients were seen every 3 months.

Endpoints The primary endpoint was all-cause mortality. Secondary endpoints included all-cause hospital admissions, cardiovascular mortality, combined CV death and CV hospital admissions, and premature treatment withdrawals.

The authors estimated a 11.2% mortality in the placebo group and powered the trial to find a 25% reduction in death in the bisoprolol arm over 2 years.

Results The trial was sopped early (mean follow-up 1.3 years) after the planned second interim analysis for benefit. The primary outcome of all-cause death occurred in 11.8% in the bisoprolol group vs 17.3% in the placebo arm (HR 0.66 (95% CI 0.54-0.81, p < 0.0001)).

Bisoprolol reduced sudden death (3.6% vs 6.3%), all-cause hospitalization (33% vs 39%), CV death (9% vs 12%). Permanent treatment withdrawal occurred in 15% of both arms.

The subgroup analysis showed no substantial treatment heterogeneity. The most common dose was 10 mg daily reached in 43% of patients.

Conclusion The 34% reduction in death was clinically meaningful and statistically robust. Our confidence in such a large effect size stems from a) previous data on beta-blockers, which found similar effects, b) the 42% reduction in sudden death in the bisoprolol arm and c) the large reductions in all-cause hospitalization. In addition, the trial conduct appeared strong with almost no lost-to-follow up. The lack of run-in period strengthens the external validity of CIBIS II.

The same caveats seen in the US carvedilol trial also apply to CIBIS II, namely that patients were ambulatory, outpatients, mostly with Class III symptoms. Patients enrolled in the trial had a mean SBP of 130 mmHg and a resting heart rate of 80. Nearly all patients were tolerating ACE-I and half were taking digoxin. In addition, patients were started on low-dose and gradually titrated higher. The majority of patients were on higher than 5 mg daily.

The authors warned against applying these results to non-ambulatory patients with Class IV symptoms, especially if there was recent instability.