Angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) have shown to decrease mortality and cardiovascular morbidity especially in high-risk patients after acute myocardial infarction (AMI). Aim of this study was to assess the association between ACEI or ARB treatment (ACEI/ARB) at hospital discharge and long-term survival after AMI in real-life patient care. From a German population–based AMI registry, 3,544 patients (75.4% men), aged 28 to 74 years, hospitalized with an incident AMI between 2000 and 2008, surviving at least 24 hours and treated with β blockers and antiplatelet agents at discharge were included in this study. All data were collected by standardized interviews and chart review. End point of this study was all-cause mortality at 3 follow-up periods: 1, 3, and 5 years after AMI. Mortality was assessed for all registered patients in 2010. Survival analyses and multivariable Cox regression analyses were conducted. Of the 3,544 patients, 83.7% received ACEI/ARB and 90.1% were treated with statins at hospital discharge. During a median follow-up period of 5.0 years (interquartile range 1.0 years), 9.3% patients died. In the multivariable Cox models adjusting for a number of covariates, use of ACEI/ARB showed a significantly inverse relation with 1-, 3-, and 5-year mortality (e.g., 5-year mortality: hazard ratio 0.74, 95% confidence interval 0.59 to 0.94, p = 0.015), and the hazard ratios for mortality did not differ significantly between the 3 examined follow-up periods. In conclusion, use of ACEI/ARB at hospital discharge is independently associated with long-term survival benefit in patients with incident AMI already treated with other guideline-recommended cardiovascular drugs.
Recent guidelines recommend long-term treatment with angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) in patients with ACEI intolerance, in all acute coronary syndrome (ACS) patients regardless of medical history and individual cardiovascular risk factors, but further uncertainty exists around the long-term benefits of these agents in low-risk acute myocardial infarction (AMI) patients without hypertension, diabetes mellitus, chronic kidney disease, heart failure (HF), or severe left ventricular (LV) systolic dysfunction. A recent Australian longitudinal population-based study over 11 years found a similar long-term mortality risk for patients with an incident AMI treated with “β blockers and statins” compared with those on “β blockers, statins and ACEI, or ARB.” Therefore, further studies are necessary to assess the long-term impact of ACEI or ARB in the modern percutaneous coronary intervention era in addition to other guideline-recommended cardiovascular drugs. This study assesses the association between ACEI or ARB prescribed at hospital discharge and long-term survival of 1, 3, and 5 years in patients hospitalized with an incident AMI between 2000 and 2008, who were treated with β blockers and antiplatelet agents, using data of a population-based myocardial infarction (MI) registry including all consecutive AMI cases and taking into account a number of relevant comorbidities and other confounders.
Methods
The population-based MI registry in Augsburg, Germany, was established in 1984 as part of the World Health Organization MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Project. After the termination of MONICA in 1995, the MI registry became part of the framework of KORA (Cooperative Health Research in the Augsburg Region). Since 1984, all cases of coronary deaths and nonfatal (at least 24 hours surviving) AMIs of the 25- to 74-year old study population in the city of Augsburg and the 2 adjacent counties (about 600,000 inhabitants mainly of European origin) have been continuously registered. About 80% of all AMI cases of the study region are treated in the region’s major hospital, Klinikum Augsburg, a tertiary care center offering interventional cardiovascular procedures and heart surgery facilities. The methods of case identification, diagnostic classification of events, and data quality control have been described in detail elsewhere.
Patients were interviewed during hospital stay by trained nurses using a standardized questionnaire to collect sociodemographic characteristics, cardiovascular risk factors, comorbidities, history of medical and drug treatment, and information on the acute event. Further information on laboratory data, type of AMI, procedures and complications during hospital stay, vital signs, medical history, and medication use were collected by review of medical chart and discharge report. Data collection of the MONICA/KORA MI registry has been approved by the Ethics Committee of the Bavarian Chamber of Physicians, and the patients gave written informed consent.
All medications were recorded and classified according to the international Anatomical Therapeutic Chemical Classification (ATC) System of the WHO Collaborating Center for Drug Statistics Methodology ( http://www.whocc.no/atc_ddd_index/ ). The hospital discharge medication considered for this analysis were antiplatelet agents (ATC code: B01AC), β blockers (ATC code: C07), statins (ATC code: C10AA, C10BA), ACEIs (ATC code: C09A, C09B), ARBs (ATC code: C09C, C09D), diuretics (ATC code: C03, C07B, C07C, C08G, C09BA, C09DA), anticoagulants (ATC code: B01AA, vitamin K antagonists; B01AB, heparin group), antiarrhythmics (ATC code: C01B), and nitrates (ATC code: C01DA). As the use of ARB is recommended in patients who are ACEI intolerant, the number of patients treated with an ARB is expected to be low. Therefore, we studied both drug classes acting on the renin-angiotensin system together as 1 group (ACEI/ARB).
The end point of this study was all-cause mortality 1, 3, and 5 years after the AMI. Mortality was assessed by checking the vital status of all registered patients of the KORA MI registry through the population registries inside and outside the study area in 2010; this procedure guaranteed that the vital status of cohort members who had moved out of the study area could also be assessed. Death certificates were obtained from local health departments.
In the present study, all patients registered with an incident AMI from January 1, 2000 to December 31, 2008 who survived at least 24 hours (n = 4,708) were included. Patients who died within hospitalization (n = 285) and patients with no data on discharge medication (n = 16) were excluded. To reduce residual confounding, we restricted the study population to patients taking both β blockers and antiplatelet agents to create a more homogeneous population. From the total of 4,407 patients with an incident AMI discharged alive, we excluded patients without treatment with β blockers and/or antiplatelet agents (n = 370; 8.4%) and patients whose data on any of the relevant covariables included in the final regression model were incomplete (n = 493), leaving 3,544 patients for analysis. Excluded patients with missing covariable information were older (median age 64 vs 60 years, p = 0.003), were less likely to receive any reperfusion therapy (74.0% vs 86.9%, p <0.0001) and treatment with ACEI/ARB (78.3% vs 83.7%, p = 0.003) and statins (78.7% vs 90.1%, p <0.0001), and showed a higher rate of in-hospital complications (12.8% vs 8.4%, p = 0.001), history of diabetes mellitus (38.3% vs 28.4%, p <0.0001), and asthma/chronic obstructive pulmonary disease (7.9% vs 5.0%, p = 0.02) compared with patients included in the study population. On the contrary, excluded patients had less frequently a left ventricular ejection fraction (LVEF) <30% (6.1% vs 7.4%, p = 0.01), but use of anticoagulants (17.7% vs 8.4%, p <0.0001) and 1-year mortality was higher (6.7% vs 2.9%, p <0.0001).
Categorical variables were expressed as percentages and continuous data as median values with interquartile range. The 2 patient groups (with or without ACEI/ARB treatment at hospital discharge) were compared using chi-square test or Fisher’s exact test for categorical variables and the t test for continuous variables. Survival times for the 3 follow-up periods (1, 3, and 5 years after the AMI) were calculated according to the method of Kaplan-Meier; statistical significance was determined by log-rank chi-square test.
To investigate the association of ACEI/ARB treatment (yes/no) and long-term mortality after 1, 3, and 5 years after AMI, relative risks were assessed using Cox proportional hazard regression model. The crude associations were sequentially adjusted for all variables that were significantly associated with either the outcome or the ACEI/ARB treatment. The stepwise adjustment was conducted using different models for each follow-up period. The associations of potential covariables with 1-, 3-, and 5-year mortality were assessed in univariate Cox proportional hazard models. To test for crude association with the primary independent variable, “ACEI/ARB treatment,” all potential confounding factors were cross-tabulated and chi-square tested. Only variables that were statistically significant at the 0.05 level with either the outcome or the primary independent variable were included in the multivariable regression analysis. Variables analyzed as potential confounding factors were gender (men/women), age (continuous), smoking (at time of the acute event) (yes/no), employed (yes/no), married (yes/no), body mass index >30 kg/m 2 (yes/no), medical history of stroke, diabetes mellitus, hyperlipidemia, hypertension, angina pectoris and asthma/chronic obstructive pulmonary disease (yes/no), LVEF (≥30%/<30%/missing), type of AMI (ST-segment elevation myocardial infarction [STEMI], non–ST-segment elevation myocardial infarction [NSTEMI], bundle branch block, or nonclassifiable/missing), in-hospital treatment (e.g., percutaneous coronary intervention, coronary artery bypass grafting, or thrombolysis) (yes/no), any in-hospital complication (cardiac arrest or cardiogenic shock or ventricular fibrillation or recurrent infarction or pulmonary edema) (yes/no), other medication at discharge (statins, diuretics, antiarrhythmics, nitrates, and anticoagulants), and the year of AMI to account for temporal trends in treatment strategies.
We considered a sparing model with only significant variables kept in the final model except age and gender (forced-in variables). The proportional hazard assumption of each predictor variable was tested graphically. The assumption was satisfied for all variables used in the Cox model except the variables such as smoking, history of hypertension and angina, body mass index >30 kg/m 2 , and gender, shown by parallel lines of log (−log(survival)) versus log of survival times. The 5 nonproportional variables were tested by generating time-dependent covariates. None of the time-dependent covariates were significant; therefore, further stratification was not necessary. Interaction effects of age, gender, diabetes, diuretics, year, and type of AMI and LVEF with the ACEI/ARB treatment were tested but failed to reach statistical significance. However, a stratified analysis by type of AMI (STEMI vs NSTEMI) was performed. In all analyses, a significance level of 5% was applied. All analyses were performed using SAS version 9.2 (SAS Institute Inc., Cary, NC).
Results
Overall, 3,544 patients (75.4% men) hospitalized with an incident AMI between 2000 and 2008 and treated with β blockers and antiplatelet agents at discharge were included in this study. There were 2,021 (57.0%) patients older than 65 years, 1,912 (54.0%) who presented with an NSTEMI, 3,081 (86.9%) who received any reperfusion therapy, and 3,194 (90.1%) who were treated with statin. Table 1 displays patient characteristics according to the use of ACEI/ARB. The 2,967 (83.7%) patients treated with ACEI/ARB were older, more often obese, suffered more often from hypertension and diabetes mellitus, received more often diuretics and statins, and showed differences in the type of AMI and in-hospital treatment compared with nonusers. On the contrary, more nonusers were employed and received nitrates. However, no significant difference was seen in terms of reduced LVEF. Over the study period, the ACEI/ARB prescribing rate was 83.7% and only 5% out of this represents ARB use. Prescribing rate of ACEI/ARB increased from 57.6% in 2000 to 88.0% in 2008.
| Variable | Patients Receiving ACEI/ARB at Hospital Discharge | p Value | |
|---|---|---|---|
| yes (n = 2967) | No (n = 577) | ||
| Age (years) ∗ | 62 (54, 68) | 60 (51, 68) | 0.001 |
| Age >65 years | 1720 (58.0%) | 301 (52.2%) | 0.010 |
| Men | 2229 (75.1%) | 442 (76.6%) | 0.451 |
| Smoker | 1086 (36.6%) | 227 (39.3%) | 0.213 |
| Married | 2198 (74.1%) | 431 (74.7%) | 0.532 |
| Employed | 1118 (37.7%) | 249 (43.2%) | 0.014 |
| Body mass index >30 kg/m² | 785 (26.5%) | 106 (18.4%) | <0.0001 |
| Hypertension † | 2360 (79.5%) | 341 (59.1%) | <0.0001 |
| Diabetes mellitus † | 891 (30.0%) | 116 (20.1%) | <0.0001 |
| Stroke † | 172 (5.8%) | 35 (6.1%) | 0.801 |
| Hyperlipidemia † | 2091 (70.5%) | 397 (68.8%) | 0.651 |
| Angina pectoris † | 429 (14.5%) | 73 (12.7%) | 0.255 |
| Asthma/chronic obstructive pulmonary disease † | 157 (5.3%) | 20 (3.5%) | 0.183 |
| Left ventricular ejection fraction | 0.087 | ||
| ≥30% | 1848 (62.3%) | 365 (63.3%) | |
| <30% | 231 (7.8%) | 30 (5.2%) | |
| Missing data | 888 (29.9%) | 182 (31.5%) | |
| Type of acute myocardial infarction | 0.044 | ||
| ST-segment elevation myocardial infarction | 1201 (40.5%) | 199 (34.5%) | |
| Non-ST-segment elevation myocardial infarction | 1578 (53.2%) | 334 (57.9%) | |
| Bundle branch block | 136 (4.6%) | 34 (5.9%) | |
| Non classifiable/missing data | 52 (1.7%) | 10 (1.7%) | |
| Any in-hospital complication ‡ | 245 (8.3%) | 51 (8.8%) | 0.389 |
| Treatment in hospital | <0.0001 | ||
| Percutaneous coronary intervention | 2143 (72.2%) | 301 (52.2%) | |
| Coronary artery bypass grafting | 383 (12.9%) | 154 (26.7%) | |
| Thrombolysis | 73 (2.5%) | 27 (4.7%) | |
| No reperfusion therapy | 368 (12.4%) | 95 (16.4%) | |
| Other medication at discharge | |||
| Statins | 2697 (90.9%) | 497 (86.1%) | 0.001 |
| Diuretics | 1631 (55.0%) | 129 (22.4%) | <0.0001 |
| Antiarrhythmics | 44 (1.5%) | 14 (2.4%) | 0.102 |
| Nitrates | 90 (3.0%) | 49 (8.5%) | <0.0001 |
| Anticoagulants | 244 (8.2%) | 52 (9.0%) | 0.531 |
∗ Median (1st quartile, 3rd quartile).
† Patient-reported information on comorbidities was collected with a standardized interview during hospital stay and further data were gathered in a concluding chart review. If the information on comorbidities from patient-report and medical chart differed, the chart information was used.
‡ Complication during hospital stay: cardiac arrest or cardiogenic shock or ventricular fibrillation or recurrent infarction or pulmonary edema.
During a median follow-up period of 5.0 years (interquartile range 1.0 years), 331 (9.3%) patients died, of these 104 (2.9%) deaths occurred within 1 year and 230 (6.5%) within 3 years after first AMI. In the unadjusted analysis, survival at the end of the 5-year follow-up period was 90.1% (95% confidence interval [CI] 89.0% to 91.2%) in patients receiving the ACEI/ARB treatment and 89.0% (95% CI 86.4% to 91.7%) in patients not receiving ACEI/ARB (log-rank test, p = 0.28). Figure 1 presents Kaplan-Meier plots for 1-, 3-, and 5-year follow-up period after first AMI.
In the total sample, ACEI/ARB treatment was significantly inversely related to 1-, 3-, and 5-year mortality ( Table 2 ). After adjustment for age and gender (model 1), the hazard ratio (HR) was 0.81 (95% CI 0.65 to 1.02) for 5-year mortality. Further stepwise adjustment for significant or forced-in confounding variables strengthened the associations between ACEI/ARB treatment and 1-, 3-, and 5-year mortality (e.g., 5-year mortality: HR 0.74, 95% CI 0.59 to 0.94 after multivariable adjustment; model 4). To assess whether ACEI/ARB treatment is associated with mortality in patients with STEMI and NSTEMI, respectively, we analyzed the association between ACEI/ARB treatment and 1-, 3-, and 5-year mortality separately for the 2 groups ( Table 2 ). In this stratified analysis, a significant association was found only for patients with NSTEMI. However, a formal test for the interaction between ACEI/ARB use and the type of AMI revealed no significant interaction.
