Using a retrospective observational study design, we assessed whether all angiotensin-converting enzyme (ACE) inhibitors had similar mortality rates after acute myocardial infarction in patients aged 65 years or older. Our results showed that mortality rates for enalapril, fosinopril, captopril, and quinapril were higher than that for ramipril. This suggests that there is no class effect for ACE inhibitors. Further research is needed to confirm the results of our study.
Readers wishing to see the tables for this paper should consult the printed version.Several studies examining the effect of angiotensin-converting enzyme (ACE) inhibitors among elderly patients in the period following myocardial infarction (MI) have shown improved survival with these medications.1-7 No previous study, however, has examined the long-term mortality rate in association with the individual ACE inhibitors chosen after the acute event. The objective of our retrospective cohort study was to evaluate whether all ACE inhibitors are associated with a similar 1-year mortality rate in patients aged 65 years or older who have had an acute MI. To show whether it is appropriate to consider a class effect in this group of drugs, we compared the effects of enalapril (Vasotec), fosinopril (Monopril), captopril (Capoten), quinapril (Accupril), lisinopril (Prinivil, Zestril), perindopril (Aceon), and ramipril (Altace) on mortality rates.
Patients and methods
We identified 18,453 patients ≥65 years of age who had an acute MI within 109 hospitals in Quebec, Canada, between April 1, 1996, and March 31, 2000. The data were collected from administrative databases storing information on hospital discharge summaries and from prescription claims databases, and were merged using encrypted Medicare numbers.
Patients who filled at least 1 prescription for an ACE inhibitor within 30 days after hospital discharge and who continued to receive the same drug for at least 1 year after the acute infarction were included in the study. Among the first 18453 patients identified, 7512 met the inclusion criteria. Patients were subsequently divided into groups according to the ACE inhibitor prescribed at discharge.
Several confounders and variables were examined. For each patient, clinical, demographic, hospital, and physician characteristics were identified and compared among the different groups of ACE inhibitors. Included among the specific characteristics were 2 hospital characteristics (MI volume and availability of cardiac catheterization), physician specialty, the year of acute MI (to account for temporal trends), discharge medications (nitrates, β blockers, aspirin, lipid-lowering agents, warfarin, calcium channel blockers, and diuretics), procedures performed in the hospital (coronary artery bypass graft surgery, percutaneous coronary intervention [PCI], and cardiac catheterization), shock, malignant cancers, acute renal failure, cerebrovascular disease, cardiac dysrhythmia, diabetes, congestive heart failure (CHF), sex, and age. We also evaluated 2 time-dependent variables: (a) dosage category (no drug, below target, or above target), and (b) exposure and nonexposure for each patient.
Adherence to treatment was evaluated based on the proportion of patients filling at least 1 prescription during the period of 9 to 12 months after discharge, while taking into account the time for which a patient was covered by a specific drug. Overall, patients were followed for an average of 2.3 years, but results were truncated at 1 year because exposure information was not available beyond the first year.
We used Kaplan-Meier curves to measure unadjusted mortality throughout 1 year of follow-up for patients taking each ACE inhibitor and compared them using the log-rank test. We used a multivariable Cox proportional hazards model8 to account for the difference in follow-up period and to control for the differences among patient characteristics. The correlation between death rates in the first year and particular ACE inhibitors was adjusted for the previously mentioned fixed variables in the Cox models. In addition, because results from the Heart Outcomes Prevention and Evaluation (HOPE) study showed that ramipril is the drug being used most frequently in the immediate period following acute MI, this drug was used as an a priori reference category in our study, and the adjusted hazard ratios for the other ACE inhibitors were compared with ramipril.
Among the 7512 patients included in the study, 34% received enalapril, which was the drug prescribed most often, 29% received lisinopril, 12% received ramipril, 6% received captopril, 4% received quinapril, and 3% received perindopril. The demographic, clinical, physician, and hospital data for the patients are shown in Table 1, based on the drug that was prescribed at discharge. Ramipril was started at a later time than the other drugs; therefore, the follow-up period was shorter. As a result, patients taking ramipril underwent PCI and cardiac catheterization more often than those receiving the other drugs. They also received prescriptions for other cardiac drugs more often as a consequence of trends in the use of these procedures.
Table 2 shows a summary of the fixed and time-dependent multivariable Cox models for 1-year mortality, categorized by hospital, with adjusted hazard ratios and 95% confidence intervals (CIs) for 1-year mortality among patients taking the different drugs. Patients in the enalapril, fosinopril, captopril, quinapril, and lisinopril groups had markedly higher 1-year mortality rates compared with those in the ramipril reference group, after adjusting for all the previously mentioned confounders and variables. This table also shows results of the model with 2 time-dependent variables added: current exposure time and dosage category. Including these covariates in the model did not alter the results. Achieving the target dosage or being above the target dosage did not predict mortality (hazard ratio = 0.94 [95% CI, 0.80-1.10]).
Analyses to test for any effects based on the presence of CHF, physician specialty, and hospital volume were performed, but no significant correlations were shown ( > .12 for all interactions).
The results of our study indicate that all ACE inhibitors do not result in the same reductions in 1-year mortality following an acute MI. Compared with ramipril, enalapril, captopril, fosinopril, quinapril, and lisinopril had higher mortality rates using currently prescribed dosages. The comparison for lisinopril and ramipril, as well as for perindopril and ramipril, was not statistically significant. Despite the fact that the basic composition of all ACE inhibitors is the same, some drugs in this class have pharmacologic and structural differences that may affect bioavailability and potency, possibly causing the differences in effectiveness shown in our study.
One of the limitations of our study was that it was observational. The collection of data was from a database that provided limited clinical information on each patient. Possible comorbid secondary diagnoses were not always clearly identified. We also restricted subjects in the study to patients who did not change ACE inhibitors within the first year after discharge to avoid possible biases. We found no difference among individual drugs after adjusting for several confounders, including CHF. The possibility that certain ACE inhibitors were prescribed to sicker patients could not be eliminated.
Our data did not provide information about the medication patients received in the hospital, and we know that ACE inhibitors show beneficial effects if given within 36 hours of acute MI.9,10 We can assume, however, that in most cases the drugs were initiated in the hospital. We also took into account the exposure time to a drug by measuring when the prescription was filled without knowing the actual pill intake.11 Prescriptions were given for a period of 30 days for most patients. If prescriptions were interrupted, the time was brief, with a median of 3 days. Those adherence problems, however, would have been the same the ACE inhibitors. A final limitation was that the follow-up period for ramipril was shorter than it was for the other drugs because it was the last one added to the drug class recommended for patients after acute MI.
The current study shows that all ACE inhibitors should not be regarded as having the same benefits. Patients ≥ 65 years of age taking ramipril at the currently used dosage were shown to have a lower 1-year mortality rate after acute MI, compared with other ACE inhibitors. Although the reasons for the differences among drugs are not known, they are most likely because of pharmacologic and structural differences among the specific drugs.12-15 Physicians providing care to patients in the post-acute MI period should not presume a class effect among all ACE inhibitors. Prospective clinical studies or randomized clinical trials are needed to corroborate the results of our study.
The authors acknowledge the editorial assistance of Melissa Rousseau.