Irbesartan in Patients with Heart Failure and Preserved Ejection Fraction

Patients

We enrolled patients from centers in 25 countries. All patients were at least 60 years of age and had heart failure symptoms and a left ventricular ejection fraction of at least 45%.^11,12 In addition, we required patients to have been hospitalized for heart failure during the previous 6 months and have current New York Heart Association (NYHA) class II, III, or IV symptoms with corroborative evidence; if they had not been hospitalized, they were required to have ongoing class III or IV symptoms with corroborative evidence. Such evidence could include findings of pulmonary congestion on radiography, left ventricular hypertrophy or left atrial enlargement on echocardiography, or left ventricular hypertrophy or left bundle-branch block on electrocardiography. Treatment with an angiotensin-converting–enzyme (ACE) inhibitor was permitted only when such therapy was considered essential for an indication other than uncomplicated hypertension.

Exclusion criteria included previous intolerance to an angiotensin-receptor blocker; an alternative probable cause of the patient's symptoms (e.g., significant pulmonary disease); any previous left ventricular ejection fraction below 40%; a history of acute coronary syndrome, coronary revascularization, or stroke within the previous 3 months; substantial valvular abnormalities; hypertrophic or restrictive cardiomyopathy; pericardial disease; cor pulmonale or other cause of isolated right heart failure; a systolic blood pressure of less than 100 mm Hg or more than 160 mm Hg or a diastolic blood pressure of more than 95 mm Hg despite antihypertensive therapy; other systemic disease limiting life expectancy to less than 3 years; substantial laboratory abnormalities (such as a hemoglobin level of less than 11 g per deciliter, a creatinine level of more than 2.5 mg per deciliter [221 μmol per liter], or liver-function abnormalities); or characteristics that might interfere with compliance with the study protocol.

Study Procedures

The trial was approved by the ethics committee at each participating center; all patients provided written informed consent. Eligible patients were treated with single-blind placebo for 1 to 2 weeks before randomization; those who successfully completed this run-in phase and whose condition remained clinically stable were randomly assigned in a 1:1 ratio to receive irbesartan or matching placebo. The randomization schedule was implemented with the use of an interactive voice-response system. The randomization block size was two and was stratified according to site. Patients were also stratified according to their use of an ACE inhibitor at randomization. Therefore, for each site, separate blocks of two were designated for patients who were taking an ACE inhibitor and for those who were not taking an ACE inhibitor. Randomization of patients who were taking an ACE inhibitor at baseline was capped at 33% at each site.

Patients were started on 75 mg of irbesartan or placebo once daily. The dose was doubled to 150 mg after 1 to 2 weeks and was doubled again to 300 mg after an additional 1 to 2 weeks, according to a forced-titration protocol as tolerated. In addition to the titration visits, patients were seen 8 weeks, 14 weeks, and 6 months after randomization and every 4 months thereafter. The score on the Minnesota Living with Heart Failure scale¹³ and the plasma level of N-terminal pro-B–type natriuretic peptide (NT-proBNP) were recorded at randomization, at 6 and 14 months, and at the final study visit. Serum creatinine and potassium were measured before randomization and at weeks 2 and 8, at month 6, and annually thereafter and, along with NT-proBNP, were analyzed in a central laboratory (Esoterix Belgium).

The executive committee designed and oversaw the trial in collaboration with representatives of the study sponsors (Bristol-Myers Squibb and Sanofi-Aventis), with assistance from an international steering committee. The sponsors or a contract research organization collected the trial data, which were then analyzed at the Statistical Data Analysis Center at the University of Wisconsin, Madison, independently of the sponsors and according to a predefined statistical analysis plan. All investigators and committee members who were involved in the conduct of the study (except for members of the data and safety monitoring board) were unaware of study-group assignments. The manuscript was prepared and submitted for publication by members of the executive committee, who had unrestricted access to the study data and who vouch for the accuracy and completeness of the reported analyses.

Study Outcomes and Definitions

The primary outcome, which was analyzed as the time from randomization to the first event, was a composite of death from any cause or hospitalization for a protocol-specified cardiovascular cause. Reasons for such hospitalizations included worsening heart failure, myocardial infarction, stroke, unstable angina, ventricular or atrial dysrhythmia, or myocardial infarction or stroke that occurred during any hospitalization. The secondary outcomes were the components of the primary outcome (death from any cause and hospitalization for cardiovascular causes), a composite heart failure outcome (death due to worsening heart failure or sudden death or hospitalization due to worsening heart failure), a change in the total score on the Minnesota Living with Heart Failure scale at 6 months, a change in the plasma level of NT-proBNP at 6 months, a composite vascular-event outcome (death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke), and death from cardiovascular causes. Deaths and hospitalizations were adjudicated by members of an independent end-point committee who were unaware of study-group assignments and used prespecified criteria.

Statistical Analysis

We originally anticipated an annual event rate of 18% for the primary outcome in the placebo group. A sample size of 3600 patients was planned to provide 1440 primary events, yielding a statistical power of 90% to detect a 14.5% reduction in risk with irbesartan, corresponding to a reduction in hazard of 15.75%, with a two-sided alpha of 0.05, assuming a recruitment period of 2 years and a minimum follow-up period of 2 years. A blinded review of event rates in 2004 indicated that outcomes had accumulated at a slower-than-anticipated rate. Consequently, to achieve the target number of events for the same decrease in the hazard in a reasonable time period, the sample size was increased to 4100 patients.

Data from all patients who underwent randomization were analyzed according to the intention-to-treat principle. The analyses of the primary outcome and other composites of death or hospitalization were performed with the use of Kaplan–Meier estimates, with the log-rank test for the comparison of the study groups, and a supportive Cox proportional-hazards model to calculate hazard ratios and 95% confidence intervals. Consistency of effects was assessed for eight prespecified subgroups, according to age (<65, 65 to 75, and >75 years), sex, ejection fraction (≤59% or >59%), the use or nonuse of ACE inhibitors and beta-blockers, the presence or absence of diabetes, hospitalization for heart failure within the previous 6 months, and geographic region (Europe, North America, or all other countries). Interactions were evaluated by fitting an interaction term between treatment and each of the eight covariates and then assessing significance with the use of a Wald test. The score on the Minnesota Living with Heart Failure scale and the log-transformed plasma level of NT-proBNP were studied by analysis of covariance, with the baseline value as a covariate. All analyses included the use of ACE inhibitors as a term in the model. To control for the global type I error, the study outcomes were examined in a prespecified sequence as described previously. If at any step superiority was not demonstrated at the 0.05 level, no conclusion would be drawn for subsequent outcomes. All P values are two-sided and were not adjusted for multiple testing.

The protocol specified that the data and safety monitoring board should conduct a single interim efficacy analysis for mortality from any cause after 50% of the total expected deaths had occurred. For this analysis, the Pocock approach was applied for harm and the O'Brien–Fleming approach was applied for benefit.

Patients

From June 2002 through April 2005, a total of 4563 patients were formally screened and 4128 underwent randomization at 293 sites in 25 countries in Western Europe, Eastern Europe, North America, South America, South Africa, and Australia. Of those patients, 2067 were assigned to receive irbesartan and 2061 to receive placebo. The common study termination date was set for April 17, 2008, when it was estimated that at least 1440 events of the primary outcome would have occurred. The mean follow-up time was 49.5 months, and the trial included 16,798 patient-years of follow-up.

Table 1. Table 1. Baseline Characteristics of the Patients.

The study groups did not differ significantly in baseline characteristics (Table 1). The mean age was 72 years, and 60% of the patients were women. The primary cause of heart failure was hypertension in 64% of the patients and ischemic heart disease in 25%, and hypertension was present in 88% overall. Atrial fibrillation was present in 29% and diabetes mellitus in 27%. Forty-one percent of the patients were obese, which was defined as a body-mass index (the weight in kilograms divided by the square of the height in meters) of more than 30. At baseline, the median level of NT-ProBNP was 339 pg per milliliter (interquartile range, 133 to 964). Baseline medications included diuretics (83%, including 52% who were taking a loop diuretic), beta-blockers (59%), calcium-channel blockers (40%), spironolactone (15%), and ACE inhibitors (25%).

Study-Drug Administration and Follow-up

At the end of the titration phase, 84% of the patients in the irbesartan group and 88% of those in the placebo group had reached the 300-mg dose (mean doses, 275 mg and 284 mg, respectively). The proportion of patients reaching the target dose did not differ according to the use of an ACE inhibitor. During the study, the proportion of patients receiving an ACE inhibitor rose from 25% in the two groups at baseline to 39% in the irbesartan group and 40% in the placebo group, the use of spironolactone rose from 15% in the two groups at baseline to 28% in the irbesartan group and 29% in the placebo group, and the use of beta-blockers rose from 59% in the irbesartan group and 58% in the placebo group to 73% in the two groups.

Between baseline and 6 months, blood pressure declined by a mean (±SD) of 3.8±18.0 mm Hg systolic and 2.1±10.5 mm Hg diastolic in the irbesartan group and by a mean of 0.2±17.6 mm Hg systolic and 0.2±10.4 mm Hg diastolic in the placebo group; the decreases in the two groups persisted for the duration of the trial. Among the surviving patients, the discontinuation rates in the irbesartan group and in the placebo group, respectively, were 13% and 12% at 1 year, 21% and 20% at 2 years, and 34% and 33% at the end of the trial.

At the end of the study, vital-status data were not available for 29 patients (1%) in the irbesartan group and 44 patients (2%) in the placebo group. If contact could not be made at end of study, data for these patients were censored from the analysis at the date they were last known to be alive.

Primary Outcome

Figure 1. Figure 1. Kaplan–Meier Curves for the Primary Outcome.

The primary outcome of death from any cause or hospitalization for prespecified cardiovascular causes (worsening heart failure, myocardial infarction, stroke, atrial or ventricular arrhythmia, and myocardial infarction or stroke occurring during hospitalization for any cause) is shown for patients receiving irbesartan and those receiving placebo. The Kaplan–Meier curves illustrate the time to the first event (hazard ratio in the irbesartan group, 0.95; 95% confidence interval [CI], 0.86 to 1.05; P=0.35 by the log-rank test). The vertical lines represent 95% confidence intervals.

Table 2. Table 2. Primary Outcome with Component Events.

Figure 2. Figure 2. Primary Outcome According to Prespecified Subgroups.

The plot shows hazard ratios and 95% confidence intervals for the primary outcome, with patients stratified according to eight subgroups prespecified in the statistical analysis plan. No heterogeneity was observed for these subgroups.

The primary composite outcome occurred in 742 patients (36%) in the irbesartan group and in 763 patients (37%) in the placebo group. There were 100.4 end-point events per 1000 patient-years in the irbesartan group and 105.4 events per 1000 patient-years in the placebo group. The hazard ratio for the primary outcome in the irbesartan group, as compared with the placebo group, was 0.95 (95% confidence interval [CI], 0.86 to 1.05; P=0.35) (Figure 1 and Table 2). The neutral effect of treatment was consistent across all prespecified subgroups (Figure 2).

Secondary Outcomes

Table 3. Table 3. Secondary Outcomes.

Rates of death from any cause were 52.6 and 52.3 per 1000 patient-years in the irbesartan group and the placebo group, respectively (hazard ratio, 1.00; 95% CI, 0.88 to 1.14; P=0.98) (Table 3). The rates for protocol-specified hospitalization were 70.6 and 74.3, respectively (hazard ratio, 0.95; 95% CI, 0.85 to 1.08; P=0.44). There were also no significant differences between the study groups for any of the other prespecified secondary outcomes or for hospital admissions for specific cardiovascular indications or for any cause (Table 3). After 6 months, scores on the Minnesota Living with Heart Failure scale improved in both groups, but the difference in the magnitude of change between the two groups was not significant. There was also no significant difference between the groups in the change in the level of NT-proBNP after 6 months.

Adverse Events

Table 4. Table 4. Drug Discontinuations and Adverse Events.

During the course of the study, 16% of patients in the irbesartan group and 14% of patients in the placebo group discontinued a study drug because of an adverse event (P=0.07) (Table 4). At baseline, the mean levels of serum creatinine were 1.00±0.34 mg per deciliter (88.4±30.1 μmol per liter) in the irbesartan group and 1.00±0.32 mg per deciliter (88.4±28.3 μmol per liter) in the placebo group. At the final visit, the mean levels of serum creatinine were 1.02±0.46 mg per deciliter (90.2±40.7 μmol per liter) in the irbesartan group and 0.98±0.34 mg per deciliter (86.6±30.1 μmol per liter) in the placebo group (P=0.11). During the course of the study, a doubling of the serum creatinine level occurred in at least one measurement in 6% of patients in the irbesartan group and in 4% of patients in the placebo group (P<0.001). A serum potassium level of more than 6.0 mmol per liter occurred at least once in 3% of patients in the irbesartan group and in 2% of patients in the placebo group (P=0.01). However, the differences in the rates of serious adverse events due to hypotension, renal dysfunction, and hyperkalemia between the two groups were not significant (Table 4).

Treatment with irbesartan did not reduce the risk of death or hospitalization for cardiovascular causes among patients who had heart failure with a preserved left ventricular ejection fraction, nor did it improve any of the secondary clinical outcomes, including disease-specific quality of life. These findings are in contrast to the benefits seen with inhibitors of the renin–angiotensin–aldosterone system, including angiotensin-receptor blockers, in patients with heart failure with a low left ventricular ejection fraction.^14-17 However, these findings are concordant with the results of two other studies involving patients who had heart failure with a preserved left ventricular ejection fraction, the Candesartan in Heart Failure: Assessment of Reduction in Mortality and Morbidity (CHARM)–Preserved trial (ClinicalTrials.gov number, NCT00634712)^17,18 and the Perindopril in Elderly People with Chronic Heart Failure (PEP-CHF) trial.¹⁹

The reasons for the lack of benefit are uncertain, but several explanations warrant consideration. One possibility is that many of the patients may not have had heart failure, since the diagnosis of heart failure with a preserved left ventricular ejection fraction is often not straightforward and can be mimicked by other conditions.²⁰ We believe that this is unlikely, since 44% of the patients had been hospitalized for heart failure within 6 months, and the remainder had at least NYHA class III symptoms and evidence of structural heart disease. The baseline NT-proBNP values in our trial were generally consistent with values used to diagnose heart failure with a preserved left ventricular ejection fraction in the outpatient setting.^21-24 However, the most compelling evidence that we enrolled a population with heart failure is the high rate at which patients had subsequent hospitalizations for heart failure. These hospitalizations occurred at rates approximately 4 to 8 times the rates observed in trials involving patients with hypertension and diabetes and those at high vascular risk.^25-28

A second consideration is whether the 300-mg target dose of irbesartan, although the highest approved dose, was suboptimal for efficacy in this disease. However, this dose reduced systolic and diastolic blood pressure by a mean of 3.8/2.1 mm Hg. Furthermore, the same dose of irbesartan reduced the onset of heart failure by 28% in the Irbesartan in Diabetic Nephropathy Trial.²⁹ On the basis of our study, we cannot assess whether a higher dose would have been beneficial.

Several other factors may have adversely affected the power of the trial. One factor was the high rate of study-drug discontinuation, which reached 34% by the end of the study, a proportion that was similar to that of other heart-failure trials of this duration. A second factor was the high rate of concomitant use of ACE inhibitors, which were taken at some time during the trial by 39% of patients in the irbesartan group and 40% of those in the placebo group, and spironolactone, which was taken by 28% of patients in the irbesartan group and 29% of those in the placebo group. In addition, 73% of patients in the two groups received a beta-blocker during the study. The treatment of a large proportion of patients with multiple inhibitors of the renin–angiotensin–aldosterone system might have left little room for further benefit from the addition of an angiotensin-receptor blocker.

The patients in our study closely resembled those with this syndrome in the community, and as expected their characteristics differed from those of patients who had heart failure with a low left ventricular ejection fraction.¹² In patients with a preserved ejection fraction, the pathophysiologic substrate of a dilated remodeled heart and clinically manifested atherosclerotic disease are either less evident or absent. Furthermore, in spite of the preponderance of patients with a history of hypertension, only a minority of patients had electrocardiographic evidence of left ventricular hypertrophy. Thus, important targets for renin–angiotensin blockade may have been absent in this population. Two previous large trials, PEP-CHF and CHARM-Preserved,^18,19 have also evaluated inhibitors of the renin–angiotensin system in patients with heart failure and a preserved left ventricular ejection fraction (predominantly in patients with a left ventricular ejection fraction of 50% or more), and neither showed an overall beneficial effect of such drugs.

In these patients, heart failure may be related to a variety of factors, including impairment of left ventricular diastolic dysfunction from myocardial hypertrophy and fibrosis or altered myocyte calcium handling,^30-32 abnormal ventricular–vascular coupling related to decreased vascular compliance,³³ impaired renal handling of salt and fluid, and other as-yet-poorly-characterized abnormalities. Although data from experimental models suggest that inhibitors of the renin–angiotensin–aldosterone system may affect these abnormalities,³⁴ it is unclear which of these potential mechanisms are primarily responsible for the clinical syndrome of heart failure with a preserved left ventricular ejection fraction. Although ACE inhibitors and angiotensin-receptor blockers have not proved to be beneficial in heart failure with a preserved left ventricular ejection fraction, aldosterone antagonists may have more success, since aldosterone plays a major role in stimulating myocardial collagen formation and in inhibiting the turnover of extracellular matrix.^35,36 This possibility is being tested in the ongoing Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trial (NCT00094302) being conducted by the National Heart, Lung, and Blood Institute.

In conclusion, we evaluated the effect of irbesartan versus placebo in patients who had heart failure with a preserved ejection fraction. No significant benefit of irbesartan was shown for a variety of cardiovascular outcomes, including death from any cause and hospitalization for cardiovascular causes.