Long-term impact of secondary preventive treatments in patients with stable angina.

We assessed the independent effects of beta blockers, calcium antagonists, lipid-lowering drugs, angiotensin converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), anti-platelet drugs, vitamin K antagonists, percutaneous coronary intervention (PCI) and coronary artery by-pass grafting (CABG) on mortality and on the composite endpoint of death, myocardial infarction, stroke or heart failure in patients with stable angina pectoris. We estimated the effects of the interventions used at baseline by multivariate Cox regression and during follow-up by G-estimation in 7,665 patients followed for a mean of 5 years in the ACTION trial. Adjusted hazard ratios (95% confidence intervals) comparing all cause mortality among users during follow-up to non-users were 1.01 (0.91, 1.09) for beta blockade, 0.82 (0.75, 0.89) for ACEIs or ARBs, 0.93 (0.87, 0.98) for calcium antagonists, 0.54 (0.49, 0.62) for lipid-lowering drugs, 0.49 (0.42, 0.53) for anti-platelet drugs, 0.74 (0.69, 0.78) for PCI, and 0.91 (0.82, 0.98) for CABG. Effects on the composite endpoint were less marked. This observational study confirms that ACEIs or ARBs, lipid-lowering and anti-platelet drugs as used in the everyday management of stable angina have independent secondary preventive effects. Calcium antagonists, PCI and CABG also appear to improve outcome.

Subgroup effects despite homogeneous heterogeneity test results.

BACKGROUND: Statistical tests of heterogeneity are very popular in meta-analyses, as heterogeneity might indicate subgroup effects. Lack of demonstrable statistical heterogeneity, however, might obscure clinical heterogeneity, meaning clinically relevant subgroup effects. METHODS: A qualitative, visual method to explore the potential for subgroup effects was provided by a modification of the forest plot, i.e., adding a vertical axis indicating the proportion of a subgroup variable in the individual trials. Such a plot was used to assess the potential for clinically relevant subgroup effects and was illustrated by a clinical example on the effects of antibiotics in children with acute otitis media. RESULTS: Statistical tests did not indicate heterogeneity in the meta-analysis on the effects of amoxicillin on acute otitis media (Q = 3.29, p = 0.51; I2 = 0%; T2 = 0). Nevertheless, in a modified forest plot, in which the individual trials were ordered by the proportion of children with bilateral otitis, a clear relation between bilaterality and treatment effects was observed (which was also found in an individual patient data meta-analysis of the included trials: p-value for interaction 0.021). CONCLUSIONS: A modification of the forest plot, by including an additional (vertical) axis indicating the proportion of a certain subgroup variable, is a qualitative, visual, and easy-to-interpret method to explore potential subgroup effects in studies included in meta-analyses.

Gender differences in cardiac remodeling and clinical outcomes in chronic stable angina pectoris (from the ACTION Trial).

Our objective was to determine the gender differences in the relation between the echocardiographic parameters of cardiac remodeling and clinical outcomes in patients with chronic stable angina. The baseline ejection fraction (EF), end-diastolic volume, and end-systolic volume were assessed in 7,016 patients in the study "A Coronary disease Trial Investigating Outcomes with Nifedipine gastrointestinal therapeutic system" (ACTION). All-cause and cardiac mortality and incident heart failure were determined after a median of 5.0 years. Cox proportional hazard models were fit to determine the effect of gender on the relation between the echocardiographic parameters and clinical outcomes (interaction p <0.10). The association between the EF and mortality differed significantly between men and women, with women demonstrating a marked increase in risk as the EF decreased, compared to men (interaction p = 0.03, adjusted p = 0.07). Also, a significant interaction by gender was seen for the association between the end-systolic volume and the risk of heart failure (interaction p = 0.01, adjusted p = 0.05). In conclusion, the relation between EF and mortality differed according to gender in patients with chronic coronary disease, with women having a greater risk of adverse outcomes as the EF decreased. Similar findings were observed with the end-systolic and end-diastolic volumes and the risk of heart failure. These findings may reflect inherent gender-based differences in ischemic heart disease and cardiac remodeling and might help to identify women at high risk.

Correlates of coronary angiography in patients with stable angina and geographical differences in its utilisation: the ACTION experience.

BACKGROUND: Utilisation of coronary angiography (CAG) varies between different countries. For patients with stable angina, the present study aimed to assess whether such differences could be explained by differences in patient characteristics, and whether these differences were related to outcome. METHODS: Using data from the ACTION trial, which compared long-acting nifedipine GITS with placebo in 7665 patients with stable angina from 19 countries, we determined by country the ratio of the observed (O) and the expected (E, based on multivariate models) number of patients who had a history of CAG before entry, or underwent CAG during a mean follow-up of 5 years. Similarly, we determined corresponding O/E ratios for the combined occurrence of any death, myocardial infarction (MI) or debilitating stroke (DS) during follow-up. RESULTS: O/E ratios for a history of CAG before entry ranged from 0.68 [95% confidence interval (CI) 0.60-0.77) for Sweden to 1.43 (95%CI 1.36-1.44) for Belgium, and were significantly correlated (p=0.04) to the corresponding O/E ratios for CAG during follow-up. The combined O/E ratio for CAG either before entry or during follow-up was not correlated (p=0.7) to the O/E for death, MI or DS, which ranged from 0.38 (95%CI undetermined) for Austria to 1.34 (95%CI 0.80-1.89) for France. CONCLUSIONS: The degree to which CAG is utilised in patients with stable angina varies between countries but is not related to the occurrence of death, MI or stroke. This supports the notion that percutaneous coronary intervention does not reduce the risk of events.

Unconventional end points in cardiovascular clinical trials: should we be moving away from morbidity and mortality?

BACKGROUND: Mortality and irreversible or major morbid events are the end points conventionally chosen for cardiovascular clinical trials because they are considered to reflect the effects of intervention on the natural history of disease. Other end points are now being considered and implemented because of the recognized limitations associated with using mortality and morbidity as the sole measures of therapeutic efficacy. METHODS AND RESULTS: This article reflects the discussion and recommendations regarding nontraditional end points for cardiovascular trials generated from a meeting of clinical trial experts convened to discuss this issue. Less common end points that have been used in cardiovascular clinical trials include composite clinical scores integrating measures of quality of life with mortality and morbidity or using the function of vital organs as end points. Appropriate measurement and applications of such end points is controversial. CONCLUSIONS: More experience is needed in applying and analyzing results with these nontraditional end points to enable their optimal use in clinical trials in cardiology, but such approaches have the potential to redress many of the conceptual and actual deficiencies inherent in conventional measures of outcome.

A propensity-matched study of hypertension and increased stroke-related hospitalization in chronic heart failure.

Hypertension is a risk factor for heart failure and stroke. However, the effect of hypertension on stroke in patients with heart failure has not been well studied. In the Digitalis Investigation Group trial, 3,674 (47%) of the 7,788 patients had a history of hypertension. Probability or propensity scores for a history of hypertension were calculated for each patient through use of a multivariable logistic regression model and were then used to match 2,386 pairs of patients with and without a history of hypertension. Kaplan-Meier and matched Cox regression analyses were used to estimate associations of a history of hypertension hospitalization for stroke during 37 months of median follow-up. After matching, patients without and with a history of hypertension had a mean systolic blood pressure of 127 mm Hg. Hospitalization for stroke occurred in 90 patients (rate, 129/10,000 person-years of follow-up) without a history of hypertension and 121 patients (rate, 178/10,000 person-years of follow-up) with a history of hypertension (hazard ratio when hypertension was compared with no hypertension=1.52; 95% confidence interval=1.11 to 2.08; p=0.010). This association was also observed among patients with baseline systolic blood pressure <140 mm Hg (hazard ratio=1.35; 95% confidence interval=1.01 to 1.81; p=0.044). In conclusion, a history of hypertension was associated with increased risk of hospitalization for stroke among patients with heart failure who were balanced in all measured baseline covariates, including blood pressure.

Meta-analyses of chronic disease trials with competing causes of death may yield biased odds ratios.

OBJECTIVE: To study the odds ratio (OR) as measure of treatment effect in the context of mutually exclusive causes of death. STUDY DESIGN AND SETTING: As example we consider meta-analyses of randomized trials of implantable cardioverter defibrillator implantation (ICD). We compare the pooled OR to the pooled cause-specific hazard ratio (HR) for each of the mutually exclusive outcomes "sudden cardiac death" (SCD) and "death other than SCD." RESULTS: The pooled OR and cause-specific HR for the reduction of SCD are similar (0.43 and 0.44, respectively) for nine included trials. However, the OR erroneously presumes a potential trend toward an adverse effect of the ICD on "death other than arrhythmia" (OR 1.11 [0.84-1.45]), whereas such an effect is small with the cause-specific HR (HR 1.03 [0.79-1.32]). In general, it is shown that a spurious association of treatment with "other death" may be seen when a substantial number of death from the cause of interest is postponed. CONCLUSION: The OR should be used with caution to study effects of treatment on mutually exclusive causes of death. Practically this concern applies primarily to meta-analysis where the use of the cause-specific HR, whenever available, is recommended.

Quality management of a large randomized double-blind multi-centre trial: the ACTION experience.

The ACTION (A Coronary disease Trial Investigating Outcome with Nifedipine GITS) study was an independent, investigator-initiated, multi-national trial comparing nifedipine GITS to placebo in 7665 patients with stable angina pectoris. The trial was sponsored by the manufacturer of the medication concerned. 291 centers in 19 countries participated. Results have been published. We defined quality management (QM) as all activities directed at ensuring data integrity and consistency; and ensuring appropriate trial conduct, including pro-active prevention of deviations from protocol. We describe the QM framework that was adopted for the ACTION trial and the key tools that were used. In the protocol, particular attention was paid to explicit definition of tasks and responsibilities of all participants, and to unequivocal operational definitions of terms such as 'randomized', 'follow-up', etc. that could be applied by investigators, on-site monitors and during data processing at the coordinating centre. A comprehensive clinical trial and study management system based on simultaneous display of scanned documents and data base content had a central role. We describe in detail how compliance with good clinical practice was ensured, how the intention-to-treat principle was implemented, how compliance with study medication and completeness of follow-up was achieved, how double blinding was maintained throughout the study structure, and how patient safety was protected. The protocol ruled out participation in any other study at the same time by ACTION participants. Our experience showed that the reasons for this are not always understood by investigators. Unequivocal operational definitions of the procedural concepts that characterize randomized clinical trials should not only be the basis of QM, but also of reporting results.

Uric acid and other renal function parameters in patients with stable angina pectoris participating in the ACTION trial: impact of nifedipine GITS (gastro-intestinal therapeutic system) and relation to outcome.

BACKGROUND: Little data is available concerning the prognostic implications of renal function abnormalities, their evolution over time and the effects of nifedipine on such abnormalities in patients with stable angina pectoris. METHODS: The previously published ACTION trial compared long-acting nifedipine GITS 60 mg once daily to placebo among 7,665 patients. Standard laboratory tests including creatinine and uric acid were assessed at baseline, after 6 months, 2 and 4 years, and at the end of follow-up. We assessed the impact of nifedipine on markers of renal dysfunction and determined whether evidence of renal failure alters the impact of nifedipine on the clinical outcome of patients with stable angina. RESULTS: Uric acid was not while creatinine level and estimated creatinine clearance were potent conditionally independent predictors of total mortality and of cardiovascular clinical events. Relative to placebo, nifedipine reduced 6-month uric acid levels by 3% (P < 0.001) of the baseline value. This difference was maintained during long-term follow-up, was present both in normotensives and in hypertensives, and was not explained by differences in diuretic therapy or allopurinol use. Nifedipine had no effect on the occurrence of clinical renal failure. Relative to placebo, the effects of nifedipine on cardiovascular death or myocardial infarction [hazard ratio (HR) = 1.01, 95% confidence interval (CI) 0.88-1.17], any stroke or transient ischaemic attack (HR = 0.73, 95% CI 0.60-0.88), new overt heart failure (HR = 0.72, 95% CI 0.55-0.95), and the need for any coronary procedure (HR = 0.81, 95% CI 0.75-0.88) were consistent across strata of markers of renal dysfunction. CONCLUSIONS: We conclude that, in patients with stable angina, nifedipine reduces uric acid levels and does not affect other markers of renal dysfunction. Renal dysfunction does not alter the effects of nifedipine on clinical outcome.

Blood pressure reduction in stable angina by nifedipine was related to stroke and heart failure reduction but not to coronary interventions.

BACKGROUND AND OBJECTIVE: Whether blood pressure (BP) reduction is a necessary prerequisite for cardiovascular risk reduction or an epiphenomenon has not been definitively established. We used an innovative analytic method to address this question. METHODS: For 7,287 participants in a stable angina trial comparing long-acting nifedipine to placebo, we estimated the BP response after 2 weeks of treatment corrected for regression-to-the mean, and then related the latter and assigned treatment to subsequent cardiovascular outcomes. RESULTS: Subsequent stroke and heart failure was strongly related to 2-week corrected systolic BP response, but coronary angiography and bypass surgery was not. Adjustment for the 2-week corrected systolic BP response changed nifedipine effect estimates (relative to placebo) for subsequent stroke from 28% (P=0.04) to 21% (P=0.13) risk reduction, and for heart failure from 30% (P=0.02) to 21% (P=0.11) risk reduction; but did not alter the effect estimates for coronary angiography (27% reduction, P<0.001), and coronary bypass surgery (22% reduction, P=0.002). CONCLUSION: The stroke and heart failure risk reduction by nifedipine GITS in patients with stable angina can be attributed partly to its BP lowering effect, whereas effects on coronary procedures are likely to be related almost entirely to its antianginal effects.

Resource utilization implications of treatment were able to be assessed from appropriately reported clinical trial data.

BACKGROUND AND OBJECTIVE: Published clinical trial data rarely allow assessment of the health care resource utilization implications of treatment. We give an example of how these can be assessed given appropriate tabulation of data. METHODS: Data from a trial comparing long-acting nifedipine gastrointestinal therapeutic system to placebo in 7,665 patients with stable angina pectoris was analyzed. RESULTS: Relative to placebo, nifedipine significantly increased mean cardiovascular (CV) event-free survival by 41 days but had no effect on mean survival. Per 100 years of follow-up, 78.1 patient-years of double-blind nifedipine administration reduced use of another calcium antagonist, an angiotensin converting enzyme inhibitor, an angiotensin receptor blocker, a diuretic and a cardiac glycoside by 1.54, 3.73, 2.63, 2.23, and 0.64 years, respectively, whereas 0.21 less hospitalization for overt heart failure, 0.47 less hospitalization for any stroke or transient ischemic attack, 0.8 less coronary angiogram, 0.38 less coronary bypass procedure, and 0.13 additional orthopedic procedure was required. Combining resource utilization with cost data for one particular hospital showed that one additional year of CV event-free survival costs an average additional euro 3,036 in the setting considered. CONCLUSION: Appropriately tabulated clinical trial data allows clinicians to judge the resource utilization implications and economic effect of treatment decisions.

Clinical course of isolated stable angina due to coronary heart disease.

AIMS: To describe the clinical course of patients with stable angina due to coronary heart disease without a history of cardiovascular (CV) events or revascularization (isolated angina). METHODS AND RESULTS: Of 7,665 patients in a trial comparing long-acting nifedipine with placebo, 2170 (28%) had isolated angina. During a mean follow-up of 4.9 years, 147 of these died (1.4/100 patient-years), while 761 (8.7/100 patient-years) either died, or had a cardiac event or procedure. The first event was death in 82, myocardial infarction or heart failure in 112, coronary revascularization in 171, and chest pain requiring hospitalization in 396. Six hundred and twelve patients (6.8/100 patient-years) underwent coronary angiography (CAG), followed by revascularization in 371. Sixty-eight of 262 deaths or major cardiac events were preceded by chest pain requiring hospitalization or revascularization. Event-rates after CAG were higher than before. The stroke rate was 0.7/100 patient-years (75 patients). CONCLUSION: Patients with stable isolated angina have low rates of death and major cardiac events, but relatively high rates of chest pain requiring hospitalization despite contemporary management. Since the majority of deaths and major CV clinical events are not preceded by clear warning symptoms, the main clinical implication is that measures to prevent such events must target all patients.

Diagnostic criteria and adjudication process both determine published event-rates: the ACTION trial experience.

OBJECTIVE: Few trials report event-adjudication procedures in detail. Using data from the ACTION (A Coronary disease Trial Investigating Outcome with Nifedipine GITS) study, we compared the impact on event-rates of an adjudication strategy based on systematic screening of all reported serious adverse events (SAEs) with a strategy based on investigator diagnoses. The final diagnosis was always made by a critical events committee (CEC) using standard criteria. METHODS: ACTION randomized 7665 patients with stable angina to either nifedipine or placebo. Pre-specified events included acute or procedural myocardial infarction (MI), refractory angina, heart failure and debilitating stroke. Clinically related SAEs including in-hospital procedures were combined into episodes independent from the investigator diagnoses entered on SAE reports. All fatal episodes and those episodes suggestive of pre-specified events were adjudicated by the CEC. RESULTS: During follow-up, 17,081 episodes were reported in 5312 patients. The SAE descriptions ruled out the occurrence of a pre-specified event in 28%. The remaining 72% were adjudicated by the CEC and 616 cases of MI, 361 of refractory angina, 275 of heart failure and 190 of debilitating stroke were diagnosed (total=1442). Had adjudication by the CEC been limited to the 3924 episodes (2397 patients) that were fatal or for which the investigator had reported any of the diagnoses mentioned, 98 cases of MI, 35 of refractory angina, 81 of heart failure and 14 of debilitating stroke would have been missed (total=228). CONCLUSION: Both the diagnostic criteria used and the adjudication process determine event-rates and conclusions about treatment effects in clinical trials. Published trial reports should always state if event-adjudication was independent of the diagnoses of investigators, and if all events of interest were adjudicated or only the first one.

Predictive value of local and core laboratory echocardiographic assessment of cardiac function in patients with chronic stable angina: The ACTION study.

AIMS: To evaluate the relationship between echocardiographic cardiac function and outcome in patients with stable symptomatic angina. METHODS: Baseline echo left ventricular ejection fraction and volume data measured in a central laboratory was available for 7016 patients (92% of the total) participating in the ACTION trial (A Coronary disease Trial Investigating Outcome with Nifedipine GITS). Ejection fraction was also measured by investigators. Evaluation of the different echocardiographic variables was based on adjusted hazard ratios comparing the unfavourable limit of the 90% range of the variable concerned to the favourable limit. RESULTS: The centrally measured ejection fraction was the most powerful predictor of all-cause death (adjusted hazard ratio=2.5), myocardial infarction, any stroke or transient ischaemic attack and overt heart failure (adjusted hazard ratio=4.5). The addition of either end systolic volume or end diastolic volume to ejection fraction did not materially affect the power of prediction. Compared to the central ejection fraction measurement, the investigator-measured ejection fraction was a less powerful predictor for all outcomes considered. CONCLUSION: Routine echocardiography carefully analysed by standardised methods provides useful prognostic information in patients with stable angina, including for total mortality.

The development of heart failure in patients with stable angina pectoris.

BACKGROUND: To describe the clinical characteristics of patients with stable angina pectoris who develop heart failure and the events preceding its onset. METHODS AND RESULTS: Of 7665 patients with stable angina in the ACTION trial, which compared long-acting nifedipine to placebo, 207 (2.7%) developed heart failure (HF) during a mean follow-up of 4.9 years. Those who developed HF were significantly (P<0.05) older, more often had diabetes, had a more extensive history of cardiovascular disease, lower ejection fractions, a higher serum creatinine and glucose, a lower haemoglobin, and were more often on blood pressure lowering drugs. A cardiac event or an intervention (n=155), a significant non-cardiac infection (n=19) or poor control of hypertension (n=12) preceded the development of HF in 186/207 cases (90%). There was no obvious precipitating factor in the remaining 21 patients (10%). Myocardial infarction increased the risk of the development of new HF within one week more than 100-fold. Nifedipine reduced the incidence of HF by 29% (P=0.015). CONCLUSIONS: The development of heart failure is uncommon in patients with stable angina, and even less so in the absence of an obvious precipitating factor.