Is the evidence base for post-myocardial infarction beta-blockers outdated?

Perspectives
    ()

    sporsmal_grey_rgb
    Article

    Post-myocardial infarction beta-blocker therapy is based on studies from the early 1980s. Since then, the diagnostics for myocardial infarction have become more sensitive, revascularisation has improved survival and there are now improved secondary prophylaxis drugs. We therefore believe it is time for new research, and that the indication for routine beta-blocker therapy following myocardial infarction should be reviewed.

    It was pointed out as early as in 2007 that (1) post-myocardial infarction beta-blocker therapy is based on three studies from 1981 – 1982 (2) – (4) (Table 1). The clinical manifestations of myocardial infarction are the same now as in the early 1980s, but the diagnostics, treatment and patient population have all changed. The introduction of high-sensitivity troponin assays has changed the diagnostics. The pathogenesis and therapeutic effect may both be different now from in the past, because the relative distribution of various causal factors has changed. A patient with acute myocardial infarction in the 1970s was often a smoker, slender and with high total cholesterol. In 2015, the patient is an overweight, pre-diabetic non-smoker with low HDL cholesterol and high triglyceride levels. In the past, left-ventricular remodelling after severe myocardial infarction was a major problem and involved loss of contractile tissue followed by dilation of the left ventricle, a reduced ejection fraction and heart failure development. Today such infarctions are less common, as most patients are revascularised early and survive the infarction with a well preserved heart function. Are we actually looking at the same disease?

    Table 1

    Long-term studies of post-infarction beta-blocker therapy. GMT = Gothenburg Metoprolol Trial; NTS = Norwegian Timolol Study; BHAT = The β-blocker Heart Attack Trial; LVEF = Left ventricular ejection fraction ; n.a. = not available; mo. = months CAPRICORN = Carvedilol Post-infarct Survival Control in LV dysfunction study; RR = Relative risk

    GMT, n = 1 395 from 1981 (3)

    NTS, n = 1 884 from 1981 (2)

    BHAT, n = 3 838 from 1982 (4)

    CAPRICORN, n = 1 959 from 2001

    Patients with LVEF < 40 % (5)

    Therapy/follow-up

    Drug

    Metoprolol

    Timolol

    Propranolol

    Carvedilol

    Time from infarction to inclusion (average)

    < 48 hours (11 ± 3)

    6 – 27 days (14)

    5 – 12 days (14)

    3 – 21 days (n.a.)

    Initially intravenously

    15 mg on admission

    no

    no

    no

    Post-infarction target dose

    100 mg × 2

    10 mg × 2

    60 – 80 mg × 3¹

    25 mg × 2

    Time to target dose

    2 days

    2 days

    2 days

    4 – 6 weeks

    Average follow-up (min–max)

    3 mo. (n.a.)

    17 mo. (12 – 33)

    25 mo. (n.a.)

    15 mo. (n.a.)

    Concomitant therapy

     

    Thrombolysis/primary PCI

    no

    no

    no

    45 %

    Acetyl salicylic acid

    n.a.

    n.a.

    22 %

    86 %

    Anticoagulation

    n.a.

    n.a.

    9 %

    n.a.

    ACE inhibitor

    no

    no

    no

    98 %

    Statin

    no

    no

    no

    n.a.

    Mortality reduction

    RR 36 % (p < 0.03)

    RR 45 % (p < 0.001)

    RR 27 % (p < 0.005)

    HR 0.77 (0.60 – 0.98) (p = 0.03)

    [i]

    [i] ¹   Patients with serum propranolol ≥ 20 mg/l at 80 mg × 3 had the dose reduced to 60 mg × 3 at 4-week visit (82 %)

    The efficacy of any treatment depends on the point of departure in relation to the maximum that can be achieved, and the lower limit for a clinically relevant result. Reduced efficacy may be due to the treatment only working when there is a certain relative distribution of causal factors, or to the factors that are affected being already almost maximally reduced.

    New prevalence data show that a large number of coronary patients remain on beta-blocker therapy for several years after an infarction. This applies to both genders of all ages, with and without heart failure.

    In the NOR-COR Study for Coronary Prevention, the use of beta-blockers after myocardial infarction, percutaneous coronary intervention (PCI) or aortocoronary bypass surgery was registered on discharge and after a follow-up period of 1.7 years (5) (Table 2). The figures for prescription of beta-blockers were 85 % on discharge and 72 % after 1.7 years, which correspond to figures from the big European registry survey EUROASPIRE IV (6). Is the evidence base for routine beta-blocker treatment following myocardial infarction good enough?

    Table 2

    Percentage of patients in the NOR-COR study who were using beta-blockers on discharge and at the follow-up visit. NSTEMI = non-ST-segment elevation myocardial infarction, STEMI = ST-segment elevation myocardial infarction

    Discharge, %¹

    Follow-up, %¹

    All, N = 1 127

    85

    72

    Gender

    Men, n = 890

    85

    72

    Women, n = 237

    84

    73

    Age, years

    < 60, n = 364

    85

    71

    60 – 75, n = 625

    85

    73

    > 75, n = 138

    86

    74

    Time since index hospitalisation

    8 weeks – 6 mo., n = 233

    72

    6 – 12 mo., n = 242

    74

    12 – 24 mo., n = 347

    74

    24 – 36 mo., n = 305

    71

    Diagnosis on discharge

    NSTEMI, n = 561

    83

    70

    STEMI, n = 335

    89²

    74

    Stable coronary disease, n = 231

    82

    75

    Angiography during index hospitalisation

    Open vessels/wall changes, n = 69

    74²

    61²

    Single-vessel disease, n = 622

    84

    72

    Multiple-vessel disease, n = 435

    88

    75

    Revascularisation during index hospitalisation

    Percutaneous coronary intervention, n = 748

    84

    72

    Aortocoronary bypass, n = 147

    91²

    81²

    Patients without heart failure

    All, N = 979

    83

    70

    Drammen Hospital, n = 496

    86²

    77³

    Vestfold Hospital, n = 483

    81

    63

    Percutaneous coronary intervention, n = 816

    83

    69

    Aortocoronary bypass, n = 126

    90²

    79²

    [i]

    [i] ¹   All p values for differences at time of discharge versus at time of inclusion were significant (< 0.001)

    ²   Differences in sub-groups at time of discharge and time of inclusion: p < 0.05

    ³   Differences in sub-groups at time of discharge and time of inclusion: p < 0.01

    What do new studies indicate?

    What do new studies indicate?

    A single literature search revealed no major studies with follow-up of ≥ three months of patients with a normal left ventricular ejection fraction.

    The Clopidogrel and Metoprolol in Myocardial Infarction Trial (COMMIT) (n = 45 852 patients, metoprolol versus placebo within 24 hours, four-week follow-up) revealed no effect on total mortality, and a reduction of reinfarction and ventricular fibrillation was counterbalanced by increased risk of cardiogenic shock (7). Nor have beta-blockers been found to influence cardiovascular events in patients with stable coronary disease (8).

    Two new registry studies have brought the problem to the fore:

    In the biggest study, based on the Reduction of Atherothrombosis for Continued Health (REACH) Registry, 44 708 patients from general practice were included in the period 2003 – 2004 (9) (Table 3). The patients were from 44 countries worldwide, and 31 % had suffered myocardial infarction. After a follow-up period of 3.6 years, there was no difference in cardiovascular mortality between patients with and without beta-blockers when propensity score matching was used. In this analysis, attempts are made retrospectively to eliminate bias factors in the two groups (Table 4).

    Table 3

    Beta-blocker after myocardial infarction. Two registry studies. n.a. = not available; PCI = percutaneous coronary intervention

    REACH: infarction 100 % (9)

    CORONOR: infarction 62 % (10)

    β-blocker +

    β-blocker –

    p

    β-blocker +

    β-blocker –

    p

    n ( %)

     9 451 (67)

     4 592 (33)

    < 0.001

    3 320 (79)

    864 (21)

    < 0.001

    Age, average (SD)

    66 (10)

    69 (10)

    < 0.001

    66 (12)

    69 (11)

    < 0.001

    Men, %

    76

    78

    n.s.

    78

    78

    n.s.

    Concomitant therapy, %

    Acetyl salicylic acid

    81

    73

    < 0.001

    n.a.

    n.a.

    Other platelet inhibitors

    27

    25

    0.02

    n.a.

    n.a.

    Platelet inhibitors combined

    n.a.

    n.a.

    97

    93

    < 0.001

    Angiotensin-converting-enzyme inhibitors

    57

    46

    < 0.001

    62

    49

    < 0.001

    Angiotensin-2 blockers

    17

    21

    < 0.001

    n.a.

    n.a.

    Statins

    82

    69

    < 0.001

    93

    89

    < 0.001

    Previous PCI, %

    n.a.

    n.a.

    79

    74

    n.s.

    Previous aortocoronary bypass

    n.a.

    n.a.

     

     

    21

    21

    n.s.

    [i]

    [i] ¹   only specified p value for previous revascularisation: < 0.001

    Table 4

    Propensity score analysis to correct for confounding with and without beta-blockers

    REACH (9)

    CORONOR (10)

    β-blocker +

    β-blocker ÷

    β-blocker +

    β-blocker ÷

    n (%)

    3 379

    3 379

    839

    839

    Age, average (SD), years

    69 (10)

    69 (10)

    69 (11)

    69 (11)

    Men, %

    75

    75

    79

    78

    Suffered infarction, %

    100

    100

    49

    50

    Concomitant therapy, %

    Acetyl salicylic acid

    76

    76

    n.a.

    n.a.

    Another platelet inhibitor

    23

    24

    n.a.

    n.a.

    Platelet inhibitors combined

    n.a.

    n.a.

    93

    93

    Angiotensin-converting-enzyme inhibitors

    49

    48

    48

    49

    Angiotensin-2 blockers

    21

    20

    n.a.

    n.a.

    Statins

    74

    75

    89

    89

    Previous percutaneous coronary intervention

    n.a.

    n.a.

    84

    85

    Previous aortocoronary bypass

    n.a.

    n.a.

    21

    21

    Follow-up, average, years

    3.6

    3.6

    2.0

    2.0

    Cardiovascular mortality, n (%)

    273 (8.0)

    291 (8.6)

    13 (1.5)²

    32 (3.8) 2

    n.a.

    Annual rate, %

    2.2

    2.4

    0.8

    1.9

    n.a.

    Hazard rate for cardiovascular mortality

    With β-blockers (95 % CI)

    0.91 (0.76 – 1.09)

     

     

    0.43 (0.22 – 0.82)

     

    0.011

    [i]

    [i] ¹   No values given when differences were not significant

    ²   Figures extrapolated from Kaplan-Meyer curves for cumulative cardiovascular mortality in the two groups

    CORonariens stables en region NORd-Pas-de-Calais (the CORONOR study) included 4 184 patients from northern France, 2 612 (62 %) of whom had suffered infarctions (10). The follow-up time was two years, and a statistically significant reduction in cardiovascular mortality was seen. There were few events, and type 2 errors therefore cannot be excluded. The same statistical method was used as for the study based on REACH.

    The CORONOR and REACH studies are accompanied by commentaries by, respectively, Floyd (11) and Danchin & Laurent (12), who examine some weaknesses of the studies. Non-beta-blocker users may have stopped because of side effects or because of contraindications that affected their prognosis. The type of beta-blocker and indication for treatment are not known, and no information is provided about left ventricular function. The conclusion is that more research is needed.

    A meta-analysis of randomised, controlled studies included 102 003 patients from 60 studies, 12 of which were published in the revascularisation period (1991 – 2013) (13). Beta-blockers had no effect on mortality, whereas the 30-day incidence of reinfarction and angina were reduced at the expense of increased incidence of heart failure and cardiogenic shock. The other 48 studies were published during the pre-revascularisation period (1966 – 1999), when beta-blockers reduced mortality, infarction and angina alike.

    Another, recent meta-analysis included 10 observational studies of treatment with beta-blockers after PCI-treated infarction (14). Of the patients in this analysis, 25 168 had been given beta-blockers and 15 705 had not. The primary endpoint was total mortality. In adjusted analyses, this was reduced with beta-blockers, but the effect was only present in study populations with a reduced ejection fraction and low consumption of other medicinal secondary prophylaxis. There was no effect on cardiac mortality, reinfarction or hospitalisations for heart failure. The authors conclude that there is now no evidence for routine beta-blocker therapy for sufferers of myocardial infarction, and that new studies are necessary for further clarification.

    International guidelines

    International guidelines

    The European guidelines of 2013 on treatment of stable coronary disease argue that the efficacy of beta-blockers is uncertain because the studies on post-infarction patients were conducted prior to the implementation of other secondary prophylactic therapy (15).

    That same year it was pointed out in similar guidelines for ST elevation infarction that long-term beta-blocker therapy is well established, but that the uncertainties referred to above were recognised (16).

    In the 2015 guidelines for non-ST-elevation infarction, early beta-blocker therapy is recommended for patients with persistent ischaemic symptoms, provided this is not contraindicated (17). Patients without severe heart failure can be offered long-term therapy.

    In the American guidelines of 2012, it is pointed out that beta-blockers are efficacious in patients who have suffered myocardial infarction in the last three years and/or have left ventricular dysfunction (18).

    Adverse effects

    Adverse effects

    Frequently reported adverse effects of beta-blockers are fatigue, impaired exercise capacity, sleeping problems, life-threatening bradyarrhythmia and hypotension (15). However, a recent critical review showed that only five of 33 described adverse effects (hyperglycaemia, diarrhoea, dizziness, intermittent claudication and bradycardia) were more prevalent in patients treated with beta-blockers than in others (19).

    Side effects are disagreeable nonetheless, and their nature may be such as to mask another disease, which may remain undiagnosed for a long time as a result.

    Should patients today have beta-blockers?

    Should patients today have beta-blockers?

    Diagnostic criteria, initial treatment and follow-up of myocardial infarction have changed drastically since the 1980s. In view of the large numbers who use beta-blockers, the problems associated with side effects and masking of other disease are highly relevant. Therapy must therefore be based on the results of prospective, randomised studies performed on a relevant patient population.

    No such documentation exists for today’s post-infarction patients. The time is ripe for performing a study on patients who have been optimally treated by present standards. A central exclusion criterion must be severe infarctions that predispose the patient to left ventricular remodelling, where both an angiotensin convertase inhibitor and beta-blocker are indicated regardless.

    PDF
    Print

    Recent Articles

    Made by Ramsalt Using Ramsalt Media