OUP user menu

Efficacy, safety and tolerability of β-adrenergic blockade with metoprolol CR/XL in elderly patients with heart failure

Prakash C Deedwania, Stephen Gottlieb, Jalal K Ghali, Finn Waagstein, John CM Wikstrand
DOI: http://dx.doi.org/10.1016/j.ehj.2004.05.022 1300-1309 First published online: 1 August 2004

Abstract

Aim To study the efficacy and tolerability of β-blockade in elderly patients with heart failure in the MERIT-HF study.

Methods and results Cox proportional hazards model was used to calculate hazard ratios (HR) with 95% confidence intervals (CI). Risk reduction was defined as (1-HR). In patients ⩾65 years total mortality was reduced by 37% (95% CI 17% to 52%; Math), sudden death by 43% (95% CI 17% to 61%; Math), and death from worsening heart failure by 61% (95% CI 32% to 77%; Math). Hospitalisations for worsening heart failure was reduced by 36% Math. Elderly patients with severe heart failure (NYHA class III/IV with ejection fraction Math0.25; Math, and patients above 75 years Math showed similar risk reductions. Metoprolol CR/XL was safe and well tolerated both during initiating therapy and during long-term follow-up.

Conclusions Metoprolol CR/XL was easily instituted, safe and well tolerated in elderly patients with systolic heart failure. The data suggest that these are the patients in whom treatment will have the greatest impact as shown by number of lives saved and number of hospitalisations avoided. The time has come to overcome the barriers that physicians perceive to β-blocker treatment, and to provide it to the large number of elderly patients with heart failure in need of this therapy.

  • Heart failure
  • β-Blockade
  • Prognosis
  • Tolerability
  • Elderly

Introduction

Heart failure is primarily a disorder of the elderly population, with 90% of all new cases occurring in patients above the age of 65 years.1,2 Its impact as a major and growing public health problem is evident by the estimated 1% prevalence in people older than 65 years, and 10% prevalence in patients above 75 years.1,2 Hospitalisation for heart failure is the most common cause in patients over 65 years of age, and the prediction is that the number of hospitalisations for heart failure will double in the next 10 years.2,3

Based on several large-scale clinical trials4–8 the current treatment guidelines for heart failure due to left ventricular (LV) systolic dysfunction recommend β-blockers as routine treatment, that is a treatment that should be used in all patients who tolerate this medicine.2,9 Despite these recommendations, and strong clinical evidence supporting the use of β-blockers in heart failure, these agents are underutilised, particularly in the elderly.10,11 This is probably attributable largely to concern about the safety and tolerability of β-blockers in the elderly, as well as the lack of published data regarding their beneficial effects on mortality and hospitalisations in the elderly.

The Metoprolol CR/XL Randomized Intervention Trial in Chronic Heart Failure (MERIT-HF) is the largest randomised clinical trial of β-blocker therapy in patients with heart failure due to LV systolic dysfunction.7,8 The MERIT-HF study had a pre-defined Data Analysis Plan, in which subgroup analyses were pre-specified, including an analysis of the elderly subgroup for safety reasons. Accordingly, and considering the large number of patients ⩾65 years randomised in MERIT-HF Math, we performed an analysis to study the efficacy and tolerability of the β-blocker metoprolol succinate controlled release/extended release (CR/XL) in this elderly population. We also analysed separately the effect of metoprolol CR/XL in elderly patients with severe heart failure as defined by NYHA functional class III/IV with ejection fraction (EF) Math0.25.5,12 Finally, some data are provided for the smaller group of patients aged 75 years or older.

Methods

Study design

The MERIT-HF was a prospective, placebo-controlled, double-blind clinical trial that randomised a total of 3991 patients. None was lost to follow-up. The study design and main results have been published earlier.7,8 The study had a planned mean follow-up time of 2.4 years, but was stopped early after recommendation from the Safety Committee. The mean follow-up time was one year. The present subgroup analysis focuses on patients ⩾65 years of age at randomisation Math. Patients enrolled in MERIT-HF were 40–80 years of age, with EF ⩽0.40 and in NYHA class II-IV heart failure for at least 3 months before enrollment (start of placebo run-in), with a heart rate at or above 68 beats/min (bpm) at enrollment, and receiving optimum standard therapy of diuretics and an ACE-inhibitor. If an ACE-inhibitor was not tolerated, other vasodilators, preferably angiotensin II receptor blockers were used. Digitalis could also be prescribed.

The recommended starting dose of metoprolol CR/XL/matching placebo was 25 mg once daily in NYHA class II patients and 12.5 mg in patients in NYHA class III and IV. It was recommended that the dose be doubled after each 2-week period until the target dose of 200 mg of metoprolol CR/XL once daily or matching placebo, or the highest tolerated dose, was reached.

The two primary outcome variables for the study were total mortality, and the combined endpoint of all-cause mortality or all-cause hospitalisation (time to first event). The pre-defined secondary endpoints were all-cause mortality or hospitalisation due to worsening heart failure (time to first event), cardiac death or non-fatal myocardial infarction, total number of hospitalisations due to cardiovascular causes; and to worsening heart failure, and withdrawal of study drug for any cause, and for worsening heart failure.

Statistical methods

Student's t-test for continuous variables and Fishers Exact Test for categorical variables were used when analysing differences in baseline variables between the two age subgroups (⩾65 years vs. Math65 years). The Cox proportional hazards model was used to calculate hazard ratios (HR), for convenience expressed as relative risks, and 95% confidence intervals. When comparing absolute risk in the two placebo age groups (⩾65 years vs. Math65 years) adjustments for the following baseline variables were made: sex, EF, NYHA class, ischaemic aetiology, history of myocardial infarction, hypertension and diabetes mellitus, systolic blood pressure, heart rate, and smoking status. Absolute risk has been expressed as number of events per patient year of follow-up. When calculating patient years of follow-up for the different endpoints patients were censored at the time of the first event. Risk reduction was defined as (1-HR). P-values over 0.20 have been given as Math.

Results

Of the 1982 elderly patients, 992 were randomly assigned to receive placebo and 990 to receive metoprolol CR/XL; of the 2009 patients <65 years of age, 1009 were assigned to receive placebo and 1000 patients to receive metoprolol CR/XL (Table 1). The two randomisation groups were well balanced in both age groups. Compared with patients below 65 years of age (mean age 55.8 years; Math), patients ⩾65 years (mean age 71.8 years, Math) had a higher prevalence of ischaemic aetiology Math, were more often in NYHA class III and IV Math, had slightly higher systolic blood pressure Math, lower diastolic blood pressure and heart rate Math, lower BMI Math, higher serum creatinine Math, and more patients were in atrial fibrillation Math. The elderly group also included a slightly higher proportion of women Math. Similar differences as described above were observed between the two age groups with severe heart failure (defined as NYHA class III/IV and EF Math0.25). For further characteristics in this subgroup, see below and Ref. 12.

View this table:
Table 1

Baseline characteristics in patients in the two age groups

Characteristics<65 years Formula⩾65 years Formula<65 vs ⩾ 65
Placebo FormulaMetoprolol CR/XL FormulaPlacebo FormulaMetoprolol CR/XL Formulap-value
Mean age (yr)56 (6.5)56 (6.5)72 (4.1)72 (4.0)Not done
Caucasian (%)92929796<0.0001
Female sex (%)19212624<0.001
Ischemic etiology (%)56557575<0.0001
NYHA class (%)<0.0001
II46453736
III51525960
IV3.33.04.33.9
Ejection fraction0.28 (0.07)0.28 (0.07)0.28 (0.08)0.28 (0.07)>0.2
Systolic blood pressure (mmHg)128 (16)129 (16)132 (18)132 (18)<0.0001
Diastolic blood pressure (mmHg)80 (9.2)80 (9.1)77 (8.9)77 (9.1)<0.0001
Heart rate (bpm)84 (10)84 (10)81 (10)81 (10)<0.0001
BMI (kg/m2)28.0 (4.9)28.3 (5.0)26.5 (4.1)26.2 (4.1)<0.0001
Serum creatinine (Formulag/L)100 (28)100 (27)113 (37)115 (37)<0.0001
Medical History
Previous myocardial infarction (%)42405556<0.0001
Hypertension (%)42444644>0.2
Diabetes mellitus (%)23252625>0.2
Atrial fibrillation (%)13122019<0.0001
Medications
Diuretics (%)89899193>0.2
ACE inhibitors (%)91918787<0.0001
ACE inhibitor or AII blocker (%)97979594<0.001
Digitalis (%)666462630.08
ASA (%)43434948<0.01
Statin (%)25242321<0.05
Symptoms
Peripheral edema (%)14161515>0.2
Jugular venous distention (%)14141313>0.2
Pulmonary rales (%)101012120.10
Third heart sound (%)22242423>0.2
  • For continuous variables standard deviations have been given within brackets.

Fig. 1 illustrates with Kaplan–Meyer estimates survival curves for total and cause specific mortality in the two age groups, and Fig. 2 point estimates for relative risk and 95% CI for total mortality, cause specific mortality, and combined endpoints for the different subgroups analysed. Further details on these endpoints are given in Tables 4 and 5.

Fig. 1

Cumulative percentages of all-cause mortality (top), sudden death (middle), and death from worsening heart failure (bottom) in the two age groups.

Fig. 2

Point estimates for relative risks and 95% of confidence intervals for cause-specific mortality (left-hand panel), and combined endpoints (right-hand panel) by age group for all patients randomised, and for those with severe heart failure (NYHA III/IV and ejection fraction Embedded Image0.25).

View this table:
Table 4

Number of clinical events for cause-specific mortality, and combined endpoints in the two age groupsa

EndpointFormula65 yearsFormula65 years
Placebo Formula (%)Metoprolol CR/XL Formula (%)Risk reduction (95% CI) %p-valuePlacebo FormulaFormula (%)Metoprolol CR/XL Formula (%)Risk reduction (95% CI) %Formula-value
Cause-specific mortality
Total83 (8.1)58 (5.6)30 (3–50)0.034134 (14.1)87 (8.9)37 (17–52)0.0008
Cardiovascular (CV)80 (7.8)54 (5.2)33 (5–52)0.024123 (12.9)74 (7.6)41 (22–56)0.0002
Sudden death60 (5.8)37 (3.6)38 (7–59)0.01972 (7.6)42 (4.3)43 (17–61)0.0032
Worsening heart failure13 (1.3)12 (1.2)8 (–)>0.245 (4.7)18 (1.8)61 (32–77)0.0005
Combined endpointsa
All-cause mortality/all-cause hospitalisation342 (40.1)265 (80.2)24 (11–35)0.0007425 (57.0)376 (48.6)14 (1–25)0.03
All-cause mortality/hosp. due to worsening heart failure193 (20.1)134 (13.8)31 (14–45)0.0008246 (28.0)177 (19.5)30 (16–43)0.0002
Cardiac death/non-fatal acute MI88 (8.6)57 (5.5)35 (10–54)0.0094137 (14.6)82 (8.5)42 (23–56)0.0001
  • CI, confidence interval; hosp., hospitalisation; MI, myocardial infarction.

  • a Only the first endpoint that occurred in each patient counted.

View this table:
Table 5

Number of clinical events for cause-specific mortality, and combined endpoints in the two age groups with severe heart failure (NYHA III/IV and ejection fraction <0.25)

Endpoint<65 years>65 years
Placebo Formula (%)Metoprolol CR/XL Formula (%)Risk reduction (95% CI) %Formula-valuePlacebo Formula (%)Metoprolol CR/XL FormulaRisk reduction (95% CI) %Formula-value
Cause-specific mortality
Total24 (13.2)13 (7.0)47 (−5 to 73)0.06548 (24.7)32 (16.0)36 (0–59)0.051
Cardiovascular (CV)23 (12.6)13 (7.0)44 (−10 to 72)0.08847 (24.2)27 (13.5)45 (11–65)0.013
Sudden death14 (7.7)6 (3.2)58 (−10 to 84)0.06925 (12.9)16 (8.0)38 (−16 to 73)0.13
Worsening heart failure8 (4.4)6 (3.2)26 (−)>0.220 (10.3)7 (3.5)66 (20–86)0.0092
Combined endpointsa
All-cause mortality/all-cause hospitalisation81 (59.6)59 (39.1)35 (7–52)0.018122 (86.5)96 (64.0)26 (3–43)0.029
All-cause mortality/hosp. due to worsening heart failure61 (40.1)35 (20.5)49 (23–67)0.001183 (48.5)53 (29.0)40 (16–58)0.0031
Cardiac death/non-fatal acute MI25 (13.8)13 (7.0)49 (0–74)0.04749 (25.5)28 (14.1)45 (12–65)0.011
  • CI, confidence interval; EF, ejection fraction; hosp., hospitalisation; MI, myocardial infarction.

  • a Only the first endpoint that occurred in each patient was counted.

Primary outcome clinical events

Total mortality

The mortality in the placebo group was higher in those ⩾65 years of age compared to those <65 years after adjustment for differences in baseline characteristics (excluding age): for total mortality the relative risk was 1.82 (95% CI 1.37 to 2.42; Math); and for mortality from heart failure 4.18 (CI 2.21 to 7.92; Math).

In patients ⩾65 years there were 134 deaths (14.1% per patient year of follow-up) on placebo vs. 87 deaths (8.9%) on metoprolol CR/XL, a 37% risk reduction in total mortality Math. Corresponding figures in those below 65 years were 83 deaths (8.1% per patient year of follow-up) vs. 58 deaths (5.6%) corresponding to a 30% risk reduction in total mortality Math.

All-cause mortality or all-cause hospitalisation (time to first event)

The combined endpoint of all-cause mortality or all-cause hospitalisation (time to first event) was reduced by 14% in patients ⩾65 years Math, and by 24% in patients below 65 years of age Math.

Secondary outcome clinical events

Cause-specific mortality

Sudden death was reduced by 43% in those ⩾65 years Math, and by 38% in patients below 65 years Math. In patients ⩾65 years there were 45 deaths (4.7%) due to worsening heart failure in the placebo group, and 18 deaths (1.8%) in the metoprolol CR/XL group (risk reduction 61%; Math). In patients below 65 years the risk for death from worsening heart failure was low in both randomisation groups, 13 deaths (1.3%) occurred in the placebo group and 12 deaths (1.2%) in the metoprolol CR/XL group.

Cause-specific combined endpoints

The combined endpoint of all-cause mortality or hospitalisation due to worsening heart failure (time to first event) was reduced by 30% in patients ⩾65 years Math, and by 31% in patients below 65 years of age Math. In addition, there was also a reduction in the combined endpoint of cardiac death or non-fatal acute myocardial infarction in both age groups (Fig. 2).

Hospitalisations

Table 2 gives cause-specific data for number of patients hospitalised at least once and total number of hospitalisations in the two age groups. In patients ⩾65 years there were 252 hospitalisations for worsening heart failure in the placebo group vs. 161 hospitalisations in the metoprolol CR/XL group Math; in patients Math65 years the corresponding figures were 199 and 156 Math.

View this table:
Table 2

Cause-specific data for number of patients hospitalised at least once, and total number of hospitalisations in the two age groups

Hospitalizations<65 years⩾ 65 years
Placebo Formula (rate)*Metoprolol CR/XL Formula (rate)p-valuePlacebo Formula (rate)Metoprolol CR/XL Formula (rate)p-value
All-cause
Patients with any hospitalisation300 (0.352)236 (0.269)0.0019368 (0.493)345 (0.446)>0.2
Total No. of hospitalisations5094730.00726405480.14
Cardiovascular cause
Patients with any hospitalisation225 (0.248)169 (0.182)0.0023269 (0.330)225 (0.267)0.024
Total No. of hospitalisations3483160.00544253330.015
Worsening heart failure
Patients with any hospitalisation134 (0.140)91 (0.094)0.0030160 (0.182)109 (0.120)0.0009
Total No. of hospitalisations1991560.00352521610.0006
  • No. = number.

  • * rate = per patient year of follow-up; not given for Total No. of hospitalisations since more than one hospitalisation can occur in a patient.

Patients with severe heart failure (NYHA Class III/IV and EF Math0.25)

There were 425 patients ⩾65 years with severe heart failure in total (NYHA Class III/IV and EF Math0.25; mean age 72 (4.1) years; mean ejection fraction 0.18 (0.04)) of whom 210 were randomised to placebo and 215 to metoprolol CR/XL. The risk for all-cause mortality or hospitalisation due to worsening heart failure in the placebo groups increased from 20.1% in all patients below 65 years of age to 48.5% in patients with severe heart failure ⩾65 years at randomisation, see Fig. 3 .

Fig. 3

Bar charts illustrating yearly risk for the combined endpoint of all-cause mortality or hospitalisations for worsening heart failure by age group in all patients randomised (upper panel) and in those with severe heart failure (NYHA III/IV and ejection fraction Embedded Image0.25; lower panel).

In those ⩾65 years 48 deaths totally occurred on placebo and 32 deaths on metoprolol CR/XL (36% reduction; Math, Fig. 3); corresponding figures for those Math65 years were 24 deaths vs. 13 deaths (47% reduction; Math). The combined endpoint of all-cause mortality or hospitalisation due to worsening heart failure was reduced by 40% in those ⩾65 years Math, and by 49% in patients below 65 years of age Math.

Patients aged 75 years or older

There were 490 patients ⩾75 years of age in total (mean age 77 years (1.5); mean ejection fraction 0.27 (0.07)) of whom 247 were randomised to placebo and 243 to metoprolol CR/XL. Of these, 34 patients died in the placebo group and 24 in the metoprolol CR/XL group (relative risk 0.71; 95% CI 0.42–1.19); corresponding data for sudden death was 17 vs. 8 deaths (0.47; 0.20–1.10), and for death from heart failure 12 vs. 9 deaths (0.75; 0.32–1.77); for the combined endpoint of all-cause mortality or hospitalisation for worsening heart failure 67 vs. 53 patients (0.79; 95% CI 0.55–1.14). There were a total of 188 hospitalisations (all-cause) in the placebo group and 149 in the β-blocker group; corresponding figures for hospitalisation for worsening heart failure were 70 vs. 55 hospitalisations.

Dose of study medicine, heart rate and blood pressure

The mean daily dose of metoprolol CR/XL in patients ⩾65 years was 146 mg (81% ⩾100 mg, 54% on 200 mg) compared to 168 mg (90% ⩾100 mg, 71% on 200 mg) in those below 65 years (Math; corresponding placebo 172 vs. 178 mg). The corresponding doses of metoprolol CR/XL in those with severe heart failure (NYHA III/IV and EF Math0.25) were 140 mg (78% ⩾100 mg, 50% on 200 mg) and 157 mg (84% ⩾100 mg, 63% on 200 mg) in the two age groups, respectively (Math; corresponding placebo 155 vs. 167 mg). The mean metoprolol CR/XL dose in patients aged ⩾75 years was 140 mg (76% ⩾100 mg, 50% on 200 mg) once daily (164 mg on placebo).

The net decrease in heart rate with metoprolol CR/XL was comparable in the two age groups (10.7 bpm ⩾65 years vs. 11.1 bpm Math65 at last follow-up visit). For patients receiving metoprolol CR/XL a net increase in systolic blood pressure of 2.1 mmHg (compared to placebo; Math) was recorded in the group ⩾65 years of age, compared to +1.0 mmHg in those Math65 years Math.

Safety and tolerability

Mortality and hospitalisations during up-titration of Math-blockade

During the up-titration phase of the study until the 8 week visit 14 deaths and 112 hospitalisations occurred in the placebo group of patients ⩾65 years, and 16 deaths and 90 hospitalisations in the group receiving metoprolol CR/XL. Corresponding figures in the subgroup of patients ⩾65 years with advanced heart failure (NYHA III/IV and EF Math0.25) were 6 vs. 5 deaths and 32 vs. 27 hospitalisations; and in those ⩾75 years 3 vs. 4 deaths and 31 vs. 24 hospitalisations.

Discontinuation of study medicine

The yearly discontinuation rate of study medicine (all-cause) by age and severity of heart failure is presented in Fig. 4 . Overall, the proportion of patients who discontinued study medication was lower in the metoprolol CR/XL group compared to those on placebo regardless of patients' age and severity of heart failure. Interestingly, the yearly discontinuation rate in the placebo group was doubled in patients ⩾65 years of age with severe heart failure (NYHA III/IV and EF Math0.25) compared to all patients on placebo below 65 years (26.6% vs. 14.1%). In patients aged ⩾75 years there were 52 discontinuations (all-cause) in the placebo group and 49 in the β-blocker group, and 10 vs. 9 patients discontinued because of worsening heart failure.

Fig. 4

Bar chart illustrating yearly discontinuation rate of study medicine (all-cause) by age group in all patients randomised (upper panel), and in those with severe heart failure defined as NYHA III/IV and ejection fraction Embedded Image0.25 (lower panel).

Table 3 presents the most commonly reported adverse events leading to discontinuation of study medicine in patients ⩾65 years. More patients were withdrawn because of worsening heart failure, angina pectoris, and myocardial infarction in the placebo compared to the metoprolol CR/XL group. Adverse events like AV-block, depression, bronchospasm and aggravation of chronic obstructive pulmonary disease were remarkably similar on metoprolol CR/XL and placebo. Slightly more patients in the β-blocker group discontinued because of bradycardia, hypotension, dizziness, fatigue and dyspnea. However, compared to placebo the net difference in discontinuation for any of these reasons was less than one patient per 100 treated during one year.

View this table:
Table 3

Tabulation of cause-specific adverse events leading to withdrawal of study medicine in the two randomization groups > 65 years of age. Adverse events listed according to total number reported in the two randomization groups combined. Net difference (metoprolol CR/XL-placebo) refers to number (No.) of patients for 100 treated for the first year of treatment (% per first year)a

Adverse EventbPlaceboMetoprolol CR/XLNet Difference (Meto CR/XL-Placebo) % per first year
No.%No.%
Heart failure627.2414.7−2.5
Angina pectoris121.460.7−0.7
Myocardial infarction111.360.7−0.5
Dyspnoea60.7101.1−0.4
Bradycardia40.5101.1−0.6
Fatigue60.680.90.3
Atrial fibrillation101.220.2−0.1
Dizziness30.370.80.5
AV-block50.640.5−0.1
Hypotension0080.90.9
Depression30.310.1−0.2
Bronchospasm20.220.20
COPD20.200−0.2
All patients with any AEc13215.312113.6−1.8
  • a 860 vs 890 patient years of follow-up until withdrawal of study medicine due to any adverse event or death in the placebo and metoprolol CR/XL group, respectively.

  • b One patient may have more than one reason for withdrawal.

  • c Refers to adverse events leading to withdrawal of study medicine.

Discussion

The results of this analysis of MERIT-HF show that metoprolol CR/XL improved survival, reduced hospitalisations due to worsening heart failure, and was well tolerated and safe in elderly patients with chronic symptomatic systolic heart failure. Similar results on risk reductions and for tolerability and safety were observed for elderly patients with advanced heart failure, and for patients above the age of 75 years compared to all patients randomised.

Age-related changes in patients with heart failure

There is a series of age-related changes in the activity of the sympathetic nervous system, including an increase in plasma norepinephrine13 and in muscle sympathetic nerve activity,14 a decrease in chronotropic and inotropic responses to catecholamines,15 and a decrease in myocardial β-adrenergic receptor density.15,16 Furthermore, certain changes that occur with aging may alter the pharmacokinetics and pharmacodynamics of β-blockers in the elderly, with profound decreases in cardiac β-receptor responsiveness,16 decreased baro-receptor activity, and decreased hepatic and renal clearance.17 Thus the finding from this analysis of MERIT-HF that treatment with metoprolol CR/XL was safe, well tolerated and beneficial in elderly patients with heart failure due to LV systolic dysfunction is of paramount importance.

Complications of heart failure in the elderly and response to Math-blockade

Cause-specific mortality

The risk of dying from heart failure increases steeply with age,1,2 which is also clearly illustrated in the figures from MERIT-HF. The risk of dying from heart failure was nearly four times higher in patients over the age of 65 years as compared with those below the age of 65 (see Fig. 1, lower panel, placebo group). It is noteworthy that treatment with metoprolol CR/XL almost completely abolished this increased risk of dying from heart failure in the elderly patients.

Regardless of age however, sudden death is the most common mode of death, occurring nearly twice as often as death from progressive heart failure in patients above the age of 65 years. It is noteworthy that the protective effect of the β-blocker on sudden death extended to those older than 65 years, as evident by a 43% reduction in sudden death in patients older than 65 years receiving metoprolol CR/XL.

Hospitalisations

The higher burden of heart failure in patients older than 65 years compared to those younger than 65 is also manifested by a 30% higher rate of hospitalisations for worsening heart failure in the elderly (placebo group), with multiple hospitalisations for worsening heart failure occurring more often in patients over 65 years compared to patients below 65 years. Not only did patients older than 65 years experience a significant reduction in hospitalisation for worsening heart failure with β-blockade, but the magnitude of reduction was at least as good as that seen in those below 65 years (36% in those ⩾65 years vs. 22% in those Math65 years, respectively).

Elderly patients with advanced heart failure

Although compared to the entire cohort of elderly patients in MERIT-HF the elderly cohort with severe heart failure was smaller, the subgroup analysis of the elderly with severe heart failure showed risk reductions similar to those observed for all patients randomised.

Patients older than 75 years

MERIT-HF randomised close to 500 patients above the age of 75 years. The findings in this subgroup echoed the overall results in regards both to the good tolerability and estimates for reduction in mortality and hospitalisations.

Societal benefits of treating elderly patients with heart failure

Age and severity of heart failure are important for absolute mortality risk as illustrated by the observation that in patients ⩾65 years with advanced heart failure every second patient died or was hospitalised for worsening heart failure, during one year of follow-up, compared to only one out of five patients below 65 years. However, regardless of absolute risk, in MERIT-HF relative risk was reduced by a similar extent, i.e., by 40% in those ⩾65 years and 31% in those <65 years (see Figs. 1 and 2).

Considering the high event rate in elderly patients, and the marked reduction in the rate of hospitalisation for worsening heart failure observed in this study, β-blockade given to older patients should result in significant cost savings and reductions in the health care expenditures for the care of elderly patients with heart failure.

Absolute risk is of importance when considering the numbers needed to treat to save one life or avoid one hospitalisation. Based on observations in MERIT-HF, the number needed to treat for one year to save one life was 19 patients for those ⩾65 years of age compared to 40 patients for the group Math65 years of age. For the subgroup with severe heart failure, even fewer patients needed to be treated for one year to save one life, in total 11 patients for those ⩾65 years of age vs. 16 patients for those Math65 years. It is however important to emphasise that patients in NYHA class II to III represent the greatest number of patients with heart failure. It is in this population where the addition of a β-blocker to existing therapy will exert its largest public health benefit.

Safety and tolerability

More patients over 65 years of age were withdrawn from β-blocker treatment compared to those below 65 years, however, a similar pattern was seen in the placebo group with more elderly patients withdrawn from treatment compared to those Math65 years. Thus discontinuation does not necessarily reflect clinically important adverse events caused by β-blockade, the drug may have been well-tolerated in many patients in whom it was nevertheless discontinued.18 However, β-adrenergic receptor blockade may cause a transient decrease in EF in some patients, which could also increase the symptomology of heart failure.19

Metoprolol CR/XL was also tolerated well as judged by fewer drug withdrawals in comparison with placebo regardless of age and severity of heart failure, and also as judged from number of deaths and hospitalisations occurring during the up-titration phase of the study. The good tolerability may be due to both the careful up-titration program used, and the relative stability of the patients enrolled in the study. Patients with decompensated heart failure (pulmonary oedema, hypo-perfusion or hypo-tension) were not randomised until stabilised. Adverse events like AV-block, depression, bronchospasm and aggravation of chronic obstructive pulmonary disease were remarkably similar on placebo and metoprolol CR/XL. Interestingly, we found that systolic blood pressure increased by a few mmHg on metoprolol CR/XL compared to placebo in the elderly group, which, one may speculate, can indicate improved LV function on β-blockade during long-term treatment.

Limitations

To learn more about the benefits of β-blockade, we have in this report from MERIT-HF presented a number of analyses from elderly patients with systolic heart failure including subgroups of those with advanced heart failure, and also a subgroup including those above the age of 75 years at randomisation. Multiple testing has been performed and results in subgroups should therefore focus on consistency and 95% confidence intervals, rather than on p-values. We should expect some variation of the treatment effect around the overall estimate as we examine a large number of subgroups, due to smaller sample size in subgroups and due to chance. Thus, the best estimate of the treatment effect on total mortality and other clinical events for any subgroup is the estimate of the hazard ratio for the overall trial.

Conclusion

In conclusion the results of this analysis of MERIT-HF demonstrated that β-blockade with metoprolol CR/XL was safe and well tolerated in elderly patients with chronic symptomatic systolic heart failure. While physicians tend to be very cautious in these patients, the data suggest that these are the patients in whom treatment will have the greatest impact as seen by number of lives saved and number of hospitalisations avoided. An Editorial that accompanied the original publication of the MERIT-HF data concluded “it would be difficult to make an argument that Math-blockers should be withheld from any patient meeting the general criteria for entry into clinical trials”.20 The results of the present analysis of the large group of elderly patients included in MERIT-HF underlines the importance of this statement. It is time to overcome the barriers that physicians perceive to β-blocker treatment, and to provide it to the large number of elderly patients with heart failure in need of this therapy, this should both improve their quality and length of life.20

Acknowledgments

Data Management and Biostatistics was performed by Georgina Bermann and Peter Johansson, AstraZeneca, Mölndal, and Hans Wedel, Nordic School of Public Health, Göteborg, Sweden. Members of the MERIT-HF Study Group are given in Ref. [7]. The MERIT-HF study was supported by grants from AstraZeneca.

Footnotes

  • 1 J.W. also senior medical advisor, Clinical Science, AstraZeneca, Mölndal, Sweden.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
View Abstract