|Home | About | Journals | Submit | Contact Us | Français|
To investigate whether the beneficial and harmful effects of platelet glycoprotein IIb/IIIa receptor blockers in non‐ST elevation acute coronary syndromes (NSTE‐ACS) depend on age.
A meta‐analysis of six trials of platelet glycoprotein IIb/IIIa receptor blockers in patients with NSTE‐ACS (PRISM, PRISM‐PLUS, PARAGON‐A, PURSUIT, PARAGON‐B, GUSTO IV‐ACS; n=31402) was performed. We applied multivariable logistic regression analyses to evaluate the drug effects on death or non‐fatal myocardial infarction at 30 days, and on major bleeding, by age subgroups (<60, 60–69, 70–79, 80 years). We quantified the reduction of death or myocardial infarction as the number needed to treat (NNT), and the increase of major bleeding as the number needed to harm (NNH).
Subgroups had 11155 (35%), 9727 (31%), 8468 (27%) and 2049 (7%) patients, respectively. The relative benefit of platelet glycoprotein IIb/IIIa receptor blockers did not differ significantly (p=0.5) between age subgroups (OR (95% CI) for death or myocardial infarction: 0.86 (0.74 to 0.99), 0.90 (0.80 to 1.02), 0.97 (0.86 to 1.10), 0.90 (0.73 to 1.16); overall 0.91 (0.86 to 0.99). ORs for major bleeding were 1.9 (1.3 to 2.8), 1.9 (1.4 to 2.7), 1.6 (1.2 to 2.1) and 2.5 (1.5–4.1). Overall NNT was 105, and overall NNH was 90. The oldest patients had larger absolute increases in major bleeding, but also had the largest absolute reductions of death or myocardial infarction. Patients 80 years had half of the NNT and a third of the NNH of patients <60 years.
In patients with NSTE‐ACS, the relative reduction of death or non‐fatal myocardial infarction with platelet glycoprotein IIb/IIIa receptor blockers was independent of patient age. Larger absolute outcome reductions were seen in older patients, but with a higher risk of major bleeding. Close monitoring of these patients is warranted.
Platelet glycoprotein IIb/IIIa receptor blockers decrease the risk of death or non‐fatal myocardial infarction at 30 days in patients with non‐ST elevation acute coronary syndromes (NSTE‐ACS) who are not routinely scheduled for early revascularisation.1,2,3,4 Age is an important risk factor for these patients, and if the relative benefits of effective interventions are the same across age groups, doctors should treat older patients even more aggressively than younger patients, as the absolute benefit may be larger.5 However, in clinical practice, utilisation of platelet glycoprotein IIb/IIIa receptor blockers is lower among older patients.6
Older patients may be undertreated because of several reasons; they may be under‐represented or excluded from randomised clinical trials, or clinicians may believe that benefits in younger patients may not generalise to the older patients or they may be worried about harmful effects in older patients.5 Researchers have variously reported that the benefit of platelet glycoprotein IIb/IIIa receptor blockers is greater in younger patients,7 similar in older and younger patients,8 or greater in older patients given their higher baseline risk.5,9 However, across it is difficult to determine how the efficacy of platelet glycoprotein IIb/IIIa receptor blockers varies across age subgroups because most trials are not large enough to provide a reliable answer. Individual acute coronary syndrome (ACS) trials have been inconclusive or even conflicting regarding the presence or absence of relative differences in drug effects across ages.10,11,12,13,14,15 Usually, the patient population was split only into two age groups (eg, <65 years, 65 years),11,13,14,15 and different primary end points were considered. An evaluation of the drug effects across age groups in a meta‐analysis using individual data can better define its relative and absolute efficacies in older versus younger patients.
One more issue is relevant in the interpretation of the effects of platelet glycoprotein IIb/IIIa receptor blockers by age groups. The incorporation of harmful major bleeding rates in the evaluation of effects should be considered to further understand the net drug effectiveness across age strata.5,9,16
We investigated whether the relative effects of platelet glycoprotein IIb/IIIa receptor blockers were consistent across age subgroups in patients with NSTE‐ACS. Further, we evaluated whether the absolute benefits and harms differed across age subgroups.
A meta‐analysis of individual patient data was performed, including trials reported since 1990 with the following characteristics: randomisation of patients with NSTE‐ACS, comparison of a platelet glycoprotein IIb/IIIa receptor blocker with placebo or control therapy, no recommendation for early (<48 h) coronary revascularisation during study–drug infusion, and enrolment of at least 1000 patients. Six trials met the inclusion criteria (PRISM, PRISM‐PLUS, PARAGON‐A, PURSUIT, PARAGON‐B and GUSTO IV‐ACS)10,11,12,13,14,15 with a total of 31402 patients. Details of the trial designs are available elsewhere.3
An electronic database consisting of data from individual patients in all eligible trials was available.3 These data were checked for completeness, for internal consistency of patients' records, and for consistency with the published reports. For this analysis, baseline characteristics regarded as important predictors of the outcome for which information was almost complete (ie, <1% missing) were age, sex, diabetes, smoking, previous myocardial infarction, previous heart failure, previous coronary artery bypass surgery (CABG), previous percutaneous coronary intervention (PCI) and ST segment depression. Other important predictors had >20% of missing data: blood pressure and heart rate were not recorded in the GUSTO IV‐ACS trial (n=7800, 25%), and baseline creatine kinase myocardial band (CK‐MB) was missing in 7469 patients (24%) across different trials. Blood pressure, heart rate and CK‐MB were used in addition to the other predictors in secondary analyses that yielded largely similar results.
For this analysis, the primary efficacy end point was defined a priori as the composite of death from any cause or non‐fatal myocardial infarction at 30 days. Myocardial infarction was a part of the composite outcome of all trials. The myocardial infarction definitions had subtle differences across trials regarding the CK‐MB threshold3 (table 11).). However, all trials had prespecified definitions of myocardial infarction.17,18 Secondary end points were: death, non‐fatal myocardial infarction, CABG, PCI, and CABG or PCI. The primary harm end point was major bleeding within 30 days. Individual trial definitions of major bleeding also had subtle differences at most, and trial‐specific definitions were retained.3 We should acknowledge that death or non‐fatal myocardial infarction and major bleeding do not have the same utility, and therefore are not comparable events. A few patients with major bleeding died or had a myocardial infarction within 30 days, and not all of the remaining patients had long‐term negative outcomes. Determining the relative weights of these events is largely subjective. A recent review identified that the weight of a major bleeding related to a drug in the context of an ACS was 0.87, compared with the weight of death, which was equal to zero.19
We divided the patient data into four subgroups according to age: <60, 60–69, 70–79 and 80 years. The decision to group patients in these intervals was made a priori, and was taken on the basis of decade intervals of common clinical use. The choice of other cut‐off points (eg, quartiles) yielded similar results (not shown). Relative differences between platelet glycoprotein IIb/IIIa receptor blockers and placebo/control on the primary end point by age subgroups were assessed, within each trial and across all trials. Logistic regression models were used, and ORs and corresponding 95% CIs were calculated. To evaluate platelet glycoprotein IIb/IIIa receptor blocker effect modification by age in each individual trial and in all trials, interaction tests were used.20 These tests also evaluated heterogeneity of effects across trials. The effects of platelet glycoprotein IIb/IIIa receptor blockers and the interactions were adjusted for the previously described predictors, for trial, and for potential differences in age‐related trends between trials. These effects were combined using random effects calculations.21 Heterogeneity of interactions across trials was evaluated with the random effects inverse variance model (with trial being the random effect).22
We performed analyses that incorporated the relationship between the baseline risk (eBR, proportion of patients in the placebo/control group with the primary efficacy end point), the efficacy odds ratio (eOR) and the respective number needed to treat (NNT). The calculation of NNT was performed using eBR and eOR, with the formula:22
NNT is the number of patients who need to be treated in order to prevent one additional death or non‐fatal myocardial infarction, and is the inverse of the absolute risk reduction. Furthermore, we looked at the relationship among the baseline proportion of the primary harm end point in the placebo/control group (hBR), the harm odds ratio (hOR) and the respective number needed to harm (NNH). The NNH was calculated using hBR and hOR, with the formula:23
The NNH is the number of patients who need to be treated in order to cause one major bleeding, and is the inverse of the absolute risk increase. The NNT and NNH calculations were performed overall and by age subgroups.
The trials included in this analysis were sponsored by several pharmaceutical companies, which are mentioned in the main trial reports,10,11,12,13,14,15 and in the acknowledgements. This study was designed, conducted and interpreted independently of the sponsors. They had the right to review the manuscript, but not to censor the findings. No separate industrial grant was obtained for this investigation.
Overall, 11155 (35%) patients were <60 years old, 9727 (31%) were aged between 60–69 years, 8468 (27%) were aged between 70–79 years, and 2049 (7%) were 80 years old. Table 22 shows baseline characteristics for all age subgroups.
The proportion of women and of all patients with a history of diabetes, myocardial infarction or heart failure, and ST depression increased with age. Further, patients 80 years had lower proportions of previous revascularisation procedures than younger patients. The proportion of patients >70 years old ranged between 30% in the PURSUIT and PRISM trials and 40% in the GUSTO IV‐ACS trial.
The overall adjusted relative reduction in the odds of death or myocardial infarction at 30 days was 9% (OR 0.91; 95% CI 0.85 to 0.99). There was no difference in the relative benefit of platelet glycoprotein IIb/IIIa receptor blockers across age subgroups (p for interaction=0.5), and this was also true for secondary efficacy end points (table 33).). Interestingly, the ratio of non‐fatal myocardial infarction over death decreased with increasing age. The overall adjusted relative increase in the odds of major bleeding was 83% (OR 1.83 (1.5 to 2.2)). This was especially high for patients aged 80 years (OR 2.5 (1.5 to 4.1)), but there were no significant differences between ages (p for interaction=0.3); (table 33).
With regard to the incidence of death or non‐fatal myocardial infarction, two trials showed significantly different relative effects across age subgroups, but in opposite directions (table 44).). The PRISM trial patients had a clear gradient of platelet glycoprotein IIb/IIIa receptor blocker effect relative to age: older patients had larger odds reductions than younger patients (p for interaction=0.01). Conversely, younger PURSUIT trial patients had larger odds reductions than the older patients (p for interaction=0.03). The interactions between platelet glycoprotein IIb/IIIa receptor blockers and age subgroup were heterogeneous across trials (p=0.002).
The absolute risk of death or myocardial infarction at 30 days correlated with age, varying from 8% in the youngest (<60 years) to 21% in the oldest group (80 years). Major bleeding at 30 days also correlated with age, from 0.8% in the youngest to 2.3% in the oldest. For the overall relative reduction in the odds of death or myocardial infarction of 9%, the NNT was 105. For the overall relative increase in the odds of major bleeding of 83%, the NNH was 90.
The oldest patients had the largest absolute reductions of death or myocardial infarction, but also had larger absolute increases in major bleeding. Patients <70 years had higher NNTs and NNHs (149 and 163 for those <60 years, and 105 and 110 for those between 60 and 69 years) than those >70 years (87 and 55 for those between 70 and 79 years, and 67 and 56 for those >80 years). Figure 11 shows the absolute event rate difference between platelet glycoprotein IIb/IIIa receptor blocker and placebo/control arms for all age subgroups. We noted somewhat greater harm in patients 70 years and a somewhat variable benefit across all age subgroups.
In patients with ACS without ST elevation, the relative reduction in the odds of death or myocardial infarction at 30 days with platelet glycoprotein IIb/IIIa receptor blockers was largely independent of age. The oldest patients had about threefold greater baseline risk than the youngest patients, not only for death or myocardial infarction but also for major bleeding. In the oldest patients, the use of platelet glycoprotein IIb/IIIa receptor blockers yielded larger absolute reductions in death/myocardial infarction, but also larger absolute increases in major bleeding rates compared with the youngest patients.
This meta‐analysis had more statistical power than individual trials to explore how the platelet glycoprotein IIb/IIIa receptor blocker effects vary by age.7,8,9,24 Individual trials did not report these effects in detail for similar age subgroups,10,11,13,14,15 and they analysed different end points. Previous analyses of the age effects in single trials have yielded inconclusive results.25 Only the PURSUIT and GUSTO IV‐ACS reported the same primary end point as we used in this paper. In addition, these analyses did not adjust for important predictors of the primary end point. We found that the PRISM and the PURSUIT trials showed significant differential relative effects of platelet glycoprotein IIb/IIIa receptor blockers across ages, but differences were in the opposite direction. We do not fully understand this phenomenon. We speculate that it could be related to the doses used and to the duration of the study drug infusion. This might have resulted in different levels of platelet inhibition in the PRISM trial (where the dose was later shown to produce suboptimal platelet inhibition in young patients compared with the PURSUIT trial (where the dose was not adjusted for older age or modest renal impairment), which might have had different consequences in younger and older patients.
The effects of other antithrombotics have been evaluated in elderly patients with unstable angina or NSTE–ACS.9 The low molecular weight heparin enoxaparin, compared with unfractionated heparin, seemed to have greater relative and absolute benefit in patients aged 65 years, compared with younger patients. When comparing clopidogrel plus aspirin with aspirin alone, there was a consistent 20% relative reduction in cardiovascular death, myocardial infarction or stroke in both elderly and younger patients. For platelet glycoprotein IIb/IIIa receptor blockers, we found an equivalent relative benefit across age subgroups, which translated into a greater absolute benefit in older patients compared with younger patients.
To describe the relative gain in primary efficacy end points by age subgroups, we defined a ratio of reduction of non‐fatal myocardial infarctions with reduction of deaths. For instance, a ratio >1 shows a larger benefit in reduction of non‐fatal myocardial infarctions compared with reduction of deaths. Given that the ratio of non‐fatal myocardial infarction to death decreased with increasing age, the use of platelet glycoprotein IIb/IIIa receptor blockers in the oldest patients probably prevented more deaths than non‐fatal myocardial infarctions.
Most trials, meta‐analyses and systematic reviews have neglected the contribution of major bleeding rates in the evaluation of the net platelet glycoprotein IIb/IIIa receptor blocker effectiveness across age subgroups in patients with NSTE–ACS.1,2,3,4,10,11,12,13,14,15,26,27,28,29,30,31 Elderly patients have a higher absolute risk of major bleeding.6,32 Therefore, the interpretation of the overall platelet glycoprotein IIb/IIIa receptor blocker efficacy needs to incorporate this harm. Although there was a trend towards increasing bleeding risk with increasing age, this was nowhere close to being statistically significant, and it should be interpreted cautiously given the small number of patients in the highest age category.
An appropriate dosing of platelet glycoprotein IIb/IIIa receptor blockers is a requisite to obtain greater benefit and less harm in elderly patients with NSTE–ACS. The CRUSADE registry showed that platelet glycoprotein IIb/IIIa receptor blockers were underutilised and misdosed in elderly patients, who are at higher risk for adverse cardiac events.33 An essential factor that increases the risk of major bleeding in elderly patients is poor renal function, which is associated with higher serum levels of platelet glycoprotein IIb/IIIa receptor blockers. Doses used in early trials were more aggressive than currently recommended doses, which are adjusted for renal dysfunction. Thus, elderly patients with NSTE–ACS should receive adequate doses of platelet glycoprotein IIb/IIIa receptor blockers to obtain the expected clinical benefit, and these doses should be adjusted for their level of renal function to avoid major bleeding events.
A recent decision analysis evaluated the efficacy of an unspecified potential drug on survival in patients with myocardial infarction and unstable angina,5 and included serious adverse events (fatal complications) as an element of the evaluation of benefit–risk balance by age‐related baseline risks. The authors used a registry database, and a definite primary end point (mortality at 1 year). The estimate of effectiveness was larger than in our randomised data (relative risk reduction 25%, absolute risk reduction 2%), and the registry population was more heterogeneous in risk (baseline risk of 2.3% in the youngest vs 27% in the oldest). The authors defined a threshold beyond which the treatment benefit would be outclassed by the treatment harm, and found that the fatal complication rate would have to be sevenfold greater in the oldest compared with the youngest age group to outweigh the survival benefits associated with treatment. These results need to be interpreted cautiously, given that most major events in these patients do not lead to death. Moreover, retrospective observational data may sometimes inflate estimates of treatment efficacy.34
Some limitations of this study should be acknowledged. Firstly, even with over 30000 randomised patients, subtle age interactions could have been missed, especially for rare events such as death. We did not see any age interactions for death based on the available data (not reported), and the clinical significance of subtle interactions is debatable. Secondly, the total number of patients in the 80 age subgroup (n=2049) was small, and <25% of each of the other three groups (n>8400). Thirdly, a substantial number of missing values for a few important predictors (blood pressure, heart rate, CK‐MB) limited some possibilities of adjusted analysis. However, the results with imputed data yielded similar conclusions (not shown). Fourthly, additional research into the appropriate weighting of events is needed to allow a more direct comparison between benefits and harms.
Some nuances should be considered in interpreting these results. The trials included broad populations of patients with ACS. Through analysis of subgroups, it seems evident that higher risk patients, such as those with positive troponins, diabetes, and perhaps ST segment depression, achieve the greatest benefit. Furthermore, it is likely that patients treated with the aggressive revascularisation strategy achieve more benefit than those treated with the conservative strategy. The trials themselves were heterogeneous, as the GUSTO IV‐ACS showed no benefit and perhaps a detrimental effect of abciximab, and PURSUIT used a very liberal definition of myocardial infarction that minimised the differences between eptifibatide and placebo. Finally, the category of major bleeding overestimates risk relative to the risk of blood transfusion, which is a more direct measure of risk and occurs less frequently. The EARLY ACS trial enrolls patients without age limit, tests whether the benefit of antithrombotic drugs is similar between elderly and young patients, and is also addressing each of the above issues.35 Allowing for these caveats, our analysis provides estimates for NNTs and NNHs by age subgroups that may be used in clinical decision making for the use of platelet glycoprotein IIb/IIIa receptor blockers in patients with NSTE‐ACS.
In conclusion, the relative risk reduction of death or myocardial infarction with platelet glycoprotein IIb/IIIa receptor blocker is independent of age in patients with non‐ST‐elevation ACS. Larger absolute reductions of death or myocardial infarction, were observed in the oldest compared with the youngest patients, as well as larger absolute increases in major bleeding rates. Attention should be given to optimising the benefit to elderly patients without increasing bleeding, by ensuring that doses adjusted for renal function are given, and elderly patients should be monitored more intensively.
The data included in this subgroup meta‐analysis were provided by Merck, White House Station, NJ, USA (sponsor of the PRISM and PRISM‐PLUS trials); F Hoffman‐La Roche, Basel, Switzerland (sponsor of PARAGON‐A and PARAGON‐B trials); COR Therapeutics, San Francisco, CA, USA, and Schering‐Plough Inc, Kenilworth, NJ, USA (sponsors of the PURSUIT trial) and Centocor Inc, Malvern, PA, USA (sponsor of the GUSTO IV‐ACS trial).
Funding: Dr. AVH received support from The Netherlands Organisation for Scientific Research (ZON/MW 908‐02‐117).
Competing interests: DJM is a consultant for Merck, Centocor, and Eli Lilly, and has received honoraria from the Merck, Centocor, as well as from Roche. HW is a consultant for and has received honoraria from Merck. PT was principal investigator and chairman of the Steering Committee for the PRISM–PLUS trial. PWA has received research grants and honoraria from Eli Lilly and Schering–Plough. RMC has worked with Centocor, Lilly, COR, Schering‐Plough, and Merck. MLS is a consultant for Merck, Centocor and Lilly, and has provided paid expert testimony to Schering–Plough.