CHARM: Candesartan benefits in heart failure

August 31, 2003

Vienna, Austria - The Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM) trials have shown a statistically significant reduction in cardiovascular death/CHF hospitalizations with the angiotensin II blocker candesartan in patients with left ventricular dysfunction, as both alternative therapy in patients intolerant of ACE inhibitors and when added to ACE inhibitors. However, results in a third group—CHF patients with preserved left ventricular function—were not so convincing.

Dr Marc Pfeffer

Dr Salim Yusuf

Results were presented today at the European Society of Cardiology Congress 2003 (ESC) by five of the leading CHARM investigators Dr Marc Pfeffer (Brigham and Women's Hospital, Boston, MA), Dr Karl Swedberg (Sahlgrenska University Hospital, Gothenburg, Sweden), Dr John McMurray (Western Infirmary, Glasgow, Scotland), Dr Salim Yusuf (McMaster University, Hamilton, ON), and Dr Chris Granger (Duke University, Durham, NC). The CHARM results are also published as four separate papers in the September 6 issue of the Lancet (which are to be made available online on September 1).

The 7601-patient CHARM trial evaluated candesartan on top of optimal treatment for heart failure and was made up of three smaller studies in separate populations:

  • Patients with LV dysfunction intolerant to ACE inhibitors (the CHARM Alternative trial).

  • Patients with LV dysfunction already taking ACE inhibitors (the CHARM Added trial).

  • Patients with preserved LV function (the CHARM Preserved trial).

While the first two groups have been well studied already, there is very little information about the treatment of CHF patients with preserved LV function, and this is the first major trial to focus on this population.

In each of the three trials, patients were randomized to candesartan (started at 4 mg and titrated up to 32 mg once daily) or placebo. Median follow-up was 37.7 months. The primary end point of the overall CHARM program was all-cause mortality in all three trials combined, which was reduced with candesartan by 9% (borderline statistical significance).

The primary end point of each of the three subtrials was cardiovascular death or CHF hospitalization. This was significantly reduced with candesartan in both the Alternative and Added studies and showed a strong positive trend in the Preserved trial.

Results "part decisive, part helpful, and part tantalizing"

The designated discussant of the CHARM trial at the ESC hot line session, Prof Philip Poole-Wilson (Royal Brompton Hospital, London, UK), said the results were "in part decisive, in part helpful, and in part tantalizing." He summarized the results of the three subtrials in the following table.

Effect of candesartan on major end points in the three CHARM trials

End point

Alternative trial (n=2028)

Added trial (n=2548)

Preserved trial (n=3025)

All-cause mortality Trend to benefit Trend to benefit No effect
CV death/CHF hospitalization Significant benefit Significant benefit Trend to benefit
CV death Significant benefit Significant benefit No effect
CHF hospitalization Significant benefit Significant benefit Trend to benefit

Poole-Wilson said that, in his view, the Alternative trial was "decisive" and confirmed previous data suggesting that CHF patients intolerant to ACE inhibitors should take an angiotensin II blocker instead. He described the Added trial as "reasonably convincing." "Candesartan, when added to all other drugs for heart failure, is clearly beneficial, but we must bear in mind that this is the eighth drug that CHF patients must now take," he said.

He said the Preserved trial was "less than convincing." Noting that there was no effect on all-cause or cardiovascular mortality but perhaps some effect on hospitalization, Poole-Wilson said that this was the sort of result he was expecting in this arm, as patients with preserved LV function were so difficult to treat. "We can certainly try candesartan in this patient group, but the results are not that convincing."

He noted that CHARM was the third major trial of an angiotensin II blocker in heart failure, following ELITE-2 and Val-HeFT. ELITE-2 showed no difference in efficacy between the angiotensin II blocker losartan and the ACE inhibitor captopril, whereas Val-HeFT showed a significant reduction in CHF hospitalizations when the angiotensin II blocker valsartan was added to ACE inhibition, although there was no effect on mortality.

He suggested that one possible reason that the CHARM results were better than the previous two trials could have been the high dose of candesartan used. "In this trial, the average dose of candesartan was around 24 mg, which would give a very powerful inhibition of the angiotensin II receptor," he commented.

Other observations of interest that Poole Wilson pointed out from the CHARM trial included the following:

  • An increase in fatal cancer was seen with candesartan, but as there was no increase in nonfatal cancer this was "almost certainly due to chance."

  • The benefit of candesartan was seen in patients also taking both a beta blocker and an ACE inhibitor, which contradicts results from the Val-HeFT trial and provides reassurance on this issue.

  • The vast majority of patients were on recommended doses of ACE inhibitors, and so the results cannot be explained by the fact that the ACE inhibitor dose was not high enough.

  • In the 39 patients who were intolerant of ACE inhibitors because of angioedema, there was just one case of angioedema on candesartan, showing that although a history of this side effect on an ACE inhibitor should signal caution, it should not be a complete contraindication to an angiotensin II blocker.

CHARM Alternative

The CHARM Alternative trial, presented by Granger, involved 2028 patients who were intolerant of ACE inhibitors[1]. Reasons for their intolerance were mainly cough (70%), hypotension (14%), and kidney disturbances (13%). They were followed for a median of 34 months.

Results showed a significant reduction in the primary end point of cardiovascular death or heart failure hospitalization and in both the individual components.

Major results of the CHARM Alternative trial

End point

Candesartan (n=1013) (%)

Placebo (n=1015) (%)

Adjusted hazard ratio (95% CI)


CV death/CHF hospitalization 33.0 40.0 0.70 (0.60-0.81) <0.0001
CV death 21.6 24.8 0.80 (0.66-0.96) 0.02
CHF hospitalization 20.4 28.2 0.61 (0.51-0.73) <0.0001

Granger said that these patients tolerated candesartan "quite well," with no statistical difference in overall drug discontinuations (21% candesartan vs 19% placebo), but there were significant increases in discontinuations due to hypotension (3.7% vs 0.9%), increased creatinine levels (6.1% vs 2.7%), and increased potassium levels (1.9% vs 0.3%) in the candesartan group.

Granger concluded: "Despite intolerance to ACE inhibitors, patients can tolerate candesartan and show a reduction in morbidity/mortality. But we still need to monitor hypotension, creatinine, and potassium."


The CHARM Added trial, presented by McMurray, involved 2548 patients who were followed for a median of 41 months[2].

Like the Alternative trial, results showed a significant reduction in the primary end point of cardiovascular death or heart failure hospitalization and in both the individual components.

Major results of the CHARM Added trial

End point

Candesartan (n=1276) (%)

Placebo (n=1272) (%)

Adjusted hazard ratio (95% CI)


CV death/CHF hospitalization 37.9 42.3 0.85 (0.75-0.96) <0.01
CV death 23.7 27.3 0.83 (0.71-0.97) 0.021
CHF hospitalization 24.2 28.0 0.83 (0.71-0.97) 0.018

McMurray noted that candesartan showed similar benefits in patients taking and not taking beta blockers and in patients taking higher and lower doses of ACE inhibitors. There were few adverse reactions, but, like the other arms of the CHARM program, candesartan was associated with increases in hypotension, hyperkalemia, and renal dysfunction. McMurray concluded that: "Adding candesartan to an ACE inhibitor brings about further reductions in cardiovascular morbidity and mortality."

CHARM Preserved

The CHARM Preserved trial, presented by Yusuf, involved 3025 patients who were followed for a median of 36 months[3]. This arm included more women (40%) than the other two arms, and patients were also a little older than in the other two trials. The effects of candesartan were less marked in this group of patients, as shown in the table.

Major results of the CHARM Preserved trial

End point

Candesartan (n=1514) (%)

Placebo (n=1509) (%)

Adjusted hazard ratio (95% CI)


CV death/CHF hospitalization 22.0 24.3 0.86 (0.74-1.00) 0.051
CV death 11.2 11.3 0.95 (0.76-1.18) 0.635
CHF hospitalization 15.9 18.3 0.84 (0.70-1.00) 0.047

Yusuf admitted that the primary end point was not statistically significant but said that the data from this arm of the CHARM trial did not show heterogeneity to the other two arms.

He commented to heartwire : "There is such sparse information on this group of patients that it was difficult to know how to power this arm of the trial. The mortality rate was lower than expected in this group, and so we may have been a little underpowered to show a significant result. But we did show some benefits. No, it's not absolutely conclusive, but it is highly persuasive. We can now treat all CHF patients with an angiotensin II blocker, without having to measure LV function."

Pfeffer gave a similar view. "We have no treatment for patients with preserved LV function, and we did show benefits so they should probably now receive candesartan. But we need more data on this group. This should be the beginning of many studies in this population," he told heartwire .

Granger was a little more cautious about whether these patients should receive treatment. "The effect was certainly not as strong as in the other groups, so I would say it would be reasonable to treat these patients, but not compulsory," he said. Several other experts questioned by heartwire echoed this view.

Another study in CHF patients with preserved LV function (I-PRESERVE) is under way with the angiotensin II blocker irbesartan. Pfeffer commented to heartwire that the I-PRESERVE investigators should be bolstered by the CHARM study. "Because our data showed a borderline result in this group, their study now becomes even more important."

CHARM Overall

Pfeffer presented data on all the three CHARM trials combinedthe CHARM Overall programshowing significant reductions in both the combined CV death/CHF hospitalizations and both individual components[4].

Major results of the CHARM Overall trial

End point

Candesartan (n=3803) (%)

Placebo (n=3796) (%)

Adjusted hazard ratio (95% CI)


All-cause mortality 23.3 24.9 0.90 (0.82-0.99) 0.032
CV death/CHF hospitalization 30.2 34.5 0.82 (0.75-0.88) <0.0001
CV death 18.2 20.3 0.87 (0.78-0.96) 0.006
CHF hospitalization 19.9 24.2 0.77 (0.70-0.84) <0.0001

He noted that there was no effect of candesartan on MI, stroke, or revascularization procedures but a significant reduction in the development of new diabetes was seen with the angiotensin II blocker (7.6% placebo vs 6.0% candesartan, hazard ratio 0.78; p=0.02).

In terms of side effects in the Overall program, as in the individual trials, candesartan was associated with significant increases in hypotension (3.5% vs 1.7%), serum creatinine levels (6.2% vs 3.0%), and hyperkalemia (2.2% vs 0.6%).

The Overall trial suggests that 23 patients need to be treated with candesartan for three years to prevent one CV death or CHF hospitalization, Pfeffer noted. Granger added that in the CHARM Alternative trial the equivalent number was 14. Pfeffer commented to heartwire that a quality-of-life analysis is still under way, but he is "pretty confident that patients do feel better on candesartan."

Commentary by Harvey White

Sep 4, 2003 - In an editorial in the Lancet[5], Dr Harvey White (Green Lane Hospital, Auckland, New Zealand) calculates that over three years the treatment effect seen in CHARM equates to one death prevented per 63 patients treated, one first hospitalization with heart failure prevented per 23 patients treated, and one new case of diabetes prevented per 71 patients treated.

On the basis of these results, he recommends that angiotensin-receptor blockers should be prescribed in addition to ACE inhibitors, beta blockers, and/or spironolactone in patients with ejection fractions of 40% or less and as an alternative to ACE inhibitors in patients with ACE-inhibitor intolerance.

On the results of the Preserved trial, he comments: "Use of an angiotensin-receptor blocker could be considered in patients with ejection fractions of more than 40% to reduce the risk of hospitalization with heart failure. More trials are needed to further investigate the effects of angiotensin-receptor blockers on relevant clinical outcomes in this large and important group of patients."


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.