Late-Breaking Trials from the European Society of Cardiology Congress in Vienna, Austria

The 3 CHARM studies (Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity) and their combined results were presented at the ESC meeting and published in Lancet in September.1-4 Since angiotensin II type 1 receptor blockers (ARB) have pharmacologic effects beyond that of angiotensin converting enzyme inhibitors (ACEI), there is the potential to further improve outcomes in patients with heart failure beyond that observed with ACEI, beta blockers, and spironolactone. Also, ARBs may be the ideal agent for those intolerant to ACEI. Thus, the CHARM investigators compared candesartan administration to placebo in 3 groups of patients with symptomatic heart failure (NYHA class II-IV): those with left ventricular ejection fraction (LVEF) > 40% (CHARM-preserved); < 40% and on ACEI (CHARM-added); and < 40% but ACEI-intolerant (CHARM-alternative). Overall, 7601 patients were randomized to candesartan titrated to a target dose of 32 mg/d or placebo and followed for at least 2 years (mean, 38 months). The primary end point for all 3 trials was cardiovascular death or hospital admission for heart failure management. Patients were excluded for abnormal renal function, elevated serum potassium, and symptomatic hypotension, as well as other standard exclusions. Total mortality and other secondary end points were also assessed.

Overall, the candesartan-treated patients showed an absolute reduction in total mortality of 1.6%, which was statistically significant after adjustment for covariates (hazard ratio, .90, 95% CI .82-.94; P = .032). The combined end point of cardiovascular death and hospitalization for heart failure was highly significantly reduced by candesartan (unadjusted HR .84, CI l77-.91; P < .0001). There was also a reduction in the number of patients developing diabetes (6% vs 7.4%; P = .02). Subgroup analysis showed similar benefits regardless of age, sex, NYHA class, and other drug therapy. Adverse effects surveillance showed more renal insufficiency and hyperkalemia in the candesartan group: Creatinine doubled in 6% vs 4%; P =.002; and potassium > 6 mmol/L 2% vs 1%; P = .017.

Candesartan also lowered blood pressure an average of 5/3 mm Hg (P < .001). In CHARM-added, the primary end point was also reduced by candesartan (HR .85, CI .75-.96; P = .01) as it was in CHARM-alternative (HR .77, CI .67-.89; P = .0004). However, in CHARM-preserved it was not. Only heart failure admissions were reduced in CHARM-preserved (230 vs 279; P = .017).

The investigators concluded that candesartan is generally well tolerated and reduces cardiovascular deaths and hospitalizations for heart failure in patients with symptomatic heart failure and LVEF < 40% regardless of other therapy including ACEI and beta blockers. It has a moderate effect on preventing hospital admissions in those with heart failure and LVEF > 40%.

Comment by Michael H. Crawford, MD

This is the largest heart failure trial undertaken to date, and it used the novel approach of prospectively combining 3 complementary trials to determine the effect on total mortality. It is also another example of the danger in relying on subgroup analyses in other trials to make recommendations. In Val-Heft patients on ACEI, beta blockers and valsartan had increased mortality, suggesting that poly-pharmacy in heart failure therapy may have its limits. CHARM used the same entry criteria as Val-Heft and prospectively tested this hypothesis in a larger group of patients only to find that there was no difference in the effect of candesartan in those on ACEI and beta blockers. In addition, this is the first trial to show a mortality benefit with an ARB. This was not the case for studies involving losartan or valsartan, although these trials were smaller.

This raises the issue of whether all ARBs are the same or whether some are superior. Finally, this is the first trial to show that ARBs have any benefit in patients with heart failure and LVEF > 40%. The reason for this benefit is not clear from the study since no measures of diastolic function were performed.

Candesartan therapy in this setting did result in adverse effects: 2.4% developed serum potassiums > 6 mmol/L, and 6.5% doubled their creatinine levels. Interestingly, 39 of the patients randomized to candesartan had a history of angioedema on ACEI, yet only 3 experienced mild angioedema on candesartan. Thus, angioedema on ACEI does not appear to be a contraindication to candesartan.

Other issues with this study include the potential beneficial effects of blood pressure lowering, especially in the LVEF > 40% group. Further lowering blood pressure could also be of value in the low EF groups. On the other hand, trying to get a heart failure patient on ACEI, beta blockers, spironolactone, and ARB at doses used in trials without excessive blood pressure lowering will be a challenge. Of interest, in CHARM-added, 100% were on ACEI, but only 55% were on beta blockers and 17% were on spironolactone. Also, the combination of 3 drugs that block the rennin-angiotensin system if applied in an unselected population could result in even more dangerous hyperkalemia than was observed in this study. Finally, the addition of candesartan to patients with heart failure and low LVEF on treatment with other agents will result in 1 death prevented per 63 cases treated, which is within the range we usually consider cost effective. It appears that heart failure treatment is getting more difficult and complicated with each new trial.

Dr. Crawford, Professor of Medicine, Associate Chief of Cardiology for Clinical Programs, University of California, San Francisco, is Editor of Clinical Cardiology Alert.


1. Pfeffer MA, et al. Lancet. 2003;362:759-766.

2. McMurray JJV, et al. Lancet. 2003;362:767-771.

3. Granger CB, et al. Lancet. 2003;362:772-776.

4. Yusuf S, et al. Lancet. 2003;362:777-781.