Author + information
- Received December 1, 2006
- Revision received December 7, 2007
- Accepted December 11, 2007
- Published online May 13, 2008.
- L. Michael Prisant, MD, FACC⁎,⁎ (, )
- Kevin L. Thomas, MD†,
- Eldrin F. Lewis, MD, MPH‡,
- Zhen Huang, MS†,
- Gary S. Francis, MD, FACC§,
- W. Douglas Weaver, MD, FACC∥,
- Marc A. Pfeffer, MD, PhD, FACC‡,
- John J.V. McMurray, MD, FACC¶,
- Robert M. Califf, MD, MACC† and
- Eric J. Velazquez, MD, FACC†
- ↵⁎Reprint requests and correspondence:
Dr. L. Michael Prisant, HB-2010, 1467 Harper Street, Medical College of Georgia, Augusta, Georgia 30912.
Objectives African Americans have a high incidence of heart failure (HF). Limited retrospective observational subgroup analyses of patients with left ventricular systolic dysfunction (LVSD) suggest marginal benefit of angiotensin-converting enzyme inhibitors in the prevention of HF hospitalizations or total mortality in African Americans.
Background Very few data exist concerning the effectiveness of angiotensin receptor blockers in this population.
Methods Baseline characteristics, treatments, and outcomes of patients from the U.S. (3,390 white and 340 African-American patients) in the VALIANT (VALsartan In Acute myocardial iNfarcTion) trial were compared. This trial included patients with an acute myocardial infarction (MI) after initial stabilization and documented LVSD and/or HF. Patients were randomly assigned to receive treatment with valsartan, captopril, or the combination; follow-up continued for up to 3 years (median 24.7 months).
Results African Americans had more coronary risk factors, more markers of poor outcome after MI, and were less likely to be revascularized when compared with white patients. After adjusting for treatment assignment, baseline characteristics, and post-infarction parameters, no difference was found in the 3-year rate of all-cause mortality, cardiovascular mortality, rehospitalization for HF, recurrent MI, or stroke between the 2 groups.
Conclusions African Americans sustaining an acute MI with LVSD and/or HF had similar clinical outcomes compared with white Americans. Valsartan, captopril, or the combination had comparable effects on cardiovascular morbidity and mortality in African Americans and white Americans.
Heart failure (HF) and/or left ventricular systolic dysfunction (LVSD) frequently complicate acute myocardial infarction (MI), and significantly worsen the prognosis of patients with coronary heart disease (CHD) (1). African Americans with CHD have the highest mortality rates and have shown slower declines in CHD mortality relative to white Americans (2). African Americans also have disproportionately higher rates of HF, and the diagnosis carries higher morbidity (3–5). Compared with white patients, overall mortality among African Americans with HF is higher. Because the incidence of MI in African Americans is thought to be lower than in white patients, a higher incidence of hypertension is often cited as the dominant cause of HF.
Several large randomized controlled trials have shown that angiotensin-converting enzyme (ACE) inhibitors reduce mortality and other adverse outcomes in the treatment of acute MI with LVSD (6–12). The efficacy of neurohormonal antagonists, specifically ACE inhibitors, in African-American patients has been questioned given several clinical trials showing a diminished effect on blood pressure and prevention of HF in African-American patients treated with ACE inhibitors compared with white Americans (13). Additionally, retrospective analyses have suggested that ACE inhibitors are less effective in African Americans with chronic HF based on higher hospitalization rates on treatment for African Americans compared with white Americans (4,14). Prior studies have not examined the effect of angiotensin receptor blockers (ARBs) or the combination of ACE inhibitors and ARBs in African-American HF patients relative to similarly treated white Americans. The VALIANT (VALsartan In Acute myocardial iNfarcTion trial) studied 14,703 patients and compared the effectiveness of an ACE inhibitor, ARB, and their combination among patients with MI complicated by LVSD, HF, or both (12). Acute MI was diagnosed by a typical clinical presentation and/or electrocardiographic changes and abnormal cardiac markers. For overall mortality and the rate of death from cardiovascular causes, reinfarction, or hospitalization for HF, valsartan was as effective as captopril. There was no added benefit for combining the drugs.
The goals of this analysis were to assess the effect of captopril, valsartan, or captopril plus valsartan on multiple outcomes as a function of race. The specific outcomes of interest include: all-cause mortality, cardiovascular mortality, HF hospitalizations, recurrent MI, stroke, adverse events, and medication discontinuation.
The methods for this trial have been previously described, and the main results were reported (12,15). Briefly, this was a randomized, double-blind trial conducted in 24 countries. The entire trial included 14,703 patients. The primary outcome was death from any cause. Valsartan was compared with captopril and valsartan plus captopril for the primary end point of overall mortality and the secondary cardiovascular end points. If valsartan was not superior or inferior to captopril, a noninferiority analysis (using both intention-to-treat and per-protocol analyses) was planned to test the null hypothesis that valsartan and captopril were comparable for the primary end point. There was no difference in total mortality or the combined cardiovascular end point for valsartan compared with captopril or valsartan plus captopril versus captopril. Thus, valsartan was as effective as captopril and combination therapy offered no added benefits, but did increase the rate of adverse events.
This subset analysis includes 3,730 patients (3,390 white and 340 African-American patients) enrolled in the U.S. Men and women ≥18 years of age who sustained an acute MI complicated by clinical or radiographic signs of HF and/or evidence of LVSD by echocardiography, contrast angiography, or radionuclide ventriculography were enrolled. Exclusion criteria included: systolic blood pressure <100 mm Hg, serum creatinine >2.5 mg/dl, previous intolerance or contraindication to ACE inhibitors or ARBs, significant valvular heart disease, or a concomitant disease severely limiting life expectancy. All patients were required to provide written informed consent. The VALIANT study used 4 race/ethnicity categorizations that were obtained at enrollment: white/Caucasian, black/African American, Asian, and other/unknown. Race was self-reported, and only patients identified as African American or white were included in this analysis.
Consenting eligible patients were randomly assigned 1:1:1 to treatment with valsartan, captopril, or a combination of captopril plus valsartan. The starting doses were valsartan 20 mg, captopril 6.25 mg, or valsartan 20 mg plus captopril 6.25 mg once daily. The dose could be titrated up to valsartan 160 mg twice daily, captopril 50 mg 3 times daily, or valsartan 80 mg twice daily plus captopril 50 mg 3 times daily. The doses used in each group could be titrated according to the patient's clinical status. A clinical end point committee blinded to treatment assignment adjudicated primary end points. A single independent data and safety monitoring board and the institutional review board or ethics committee at each participating site approved the protocol.
All statistical analyses were performed at the Duke Clinical Research Institute. African-American patients were compared with white American patients on 3-year clinical outcomes of all-cause mortality, cardiovascular mortality, hospitalization for HF, recurrent MI, and stroke. The null hypothesis was that there is no difference between African-American and white patients on a given 3-year clinical outcome.
Comparison of baseline characteristics and adverse events between African Americans and white Americans were carried out using a Pearson chi-square or Cochran-Mantel-Haenszel chi-square test, and a Wilcoxon rank sum test for categorical and continuous data, respectively. Clinical outcomes at 3 years were analyzed using Cox regression models stratified by race. Cumulative event rates through 3 years were estimated for both African Americans and white Americans, after adjustment for baseline characteristics. Stepwise selection was used to choose baseline characteristics and create a parsimonious model for each clinical outcome. A p value of 0.10 was required for a variable to enter and stay in the final model. Longitudinal analysis was carried out to examine the effects of covariates to variables of interest over time. Trajectories were plotted for visual examination. We checked the assumption of normality as well as the possibility of including additional polynomial terms in the model. A final linear mixed model was used to analyze these repeated measures over time.
The list of control variables for each clinical outcome is reported in the Online Appendix. Standard errors of adjusted event rates were estimated by bootstrapping from the original data with 500 resamples of the entire cases of data. Differential treatment effects on African Americans versus white Americans were tested by interactions between race and treatments adjusting for baseline characteristics.
Continuous variables were summarized as medians with 25th and 75th percentiles, and categorical variables as frequencies and percentages. For a single comparison, a p value of 0.05 was considered statistically significant evidence to reject a null hypothesis. For multiple comparisons, the significance level was adjusted by the Bonferroni method.
The SAS software package (version 8, SAS Institute Inc., Cary, North Carolina) was used for data analysis; SAS Jackboot macro (SAS Institute, Inc.) was used for bootstrapping standard errors.
Mean baseline total creatine phosphokinase and troponin I were not significantly different between African-American and white patients. African Americans were younger, more often female, and also more likely than white Americans to have a history of diabetes mellitus, chronic HF, and hypertension (Table 1). Moreover, African Americans had a lower estimated glomerular filtration rate (eGFR), higher systolic and diastolic blood pressures, faster heart rate, and higher Killip class. Conversely, white patients were more likely to have been previously hospitalized, to have undergone prior percutaneous coronary intervention (PCI), and to have a history of unstable angina. During the index hospitalization, HF and renal insufficiency were observed in more African Americans than white Americans.
African-American patients (84.1%) were as likely as white patients (88.6%) to undergo coronary angiography after acute MI. Among African Americans, the percentage of the number of diseased coronary vessels was 5.9% (none), 30.4% (single), 24.5% (double), and 39.2% (triple). The percentage among white Americans was 1.7% (none), 30.8% (single), 27.2% (double), and 40.3% (triple). However, after adjusting for the presence of single-vessel, double-vessel, and triple-vessel disease, African Americans were less likely than white Americans to receive PCI (p = 0.003) or coronary artery bypass grafting (CABG) (p = 0.021).
African Americans were more likely to be treated with an ACE inhibitor, long-acting nitrates, insulin, and non–potassium-sparing diuretics (Table 2). White Americans were more likely to be treated with antiplatelet therapies (other than aspirin) and oral anticoagulants. At 1 year of follow-up, comparisons by race of concomitant pharmacotherapies showed significant differences in several evidence-based medications. African Americans were more often treated with beta-blockers, calcium-channel blockers, non–potassium-sparing diuretics, long-acting nitrates, and other vasodilators. White Americans were more likely to be treated with aspirin, statins, or other lipid modulating therapies (Table 2).
There were no significant differences in the adjusted 3-year rates of all-cause mortality (22.1% for African Americans vs. 20.3% for white Americans; p = 0.45) or cardiovascular mortality (18.4% for African Americans vs. 16.7% for white Americans; p = 0.31) (Fig. 1). The recurrent MI rate at 3 years was similar between African Americans and white Americans. There was a trend toward higher rates of HF hospitalizations for African-American patients compared with white Americans at 3 years (p = 0.09). The stroke rate for African Americans was approximately twice as high as for white Americans (8.0% vs. 3.7%; p = 0.06). The outcomes by race for all-cause mortality, cardiovascular mortality, hospitalizations for HF, recurrent MI, and stroke were independent of treatment assignment (p > 0.101 for tests of interactions between race and treatment assignment across different outcomes).
Comparative treatment effect in African Americans
African Americans treated with valsartan compared with African Americans treated with captopril alone or captopril plus valsartan had similar cumulative all-cause mortality rates (Fig. 2). Similarly, for the outcomes of cardiovascular mortality, hospitalization for HF, recurrent MI, and stroke, there were no differences among African-American patients treated with any of the 3 medical regimens.
Adverse events and medication discontinuation
White patients were more likely than African Americans to experience hypotension (32.8% vs. 20.9%; p < 0.0001). There were no racial differences in the incidence of dry cough, which occurred in approximately 19% of all patients (20.2% for white patients vs. 18.8% for African Americans; p = 0.6), most of which was treated with captopril. Angioedema was rare; however, among patients treated with captopril, African Americans were almost 2 times more likely than white patients to develop angioedema, a result that was not statistically significant (1.2% for white patients vs. 2.1% for African Americans; p = 0.2). Adverse events led to discontinuation of treatment in <11% of the population treated. Angioedema was the most common reason for discontinuing captopril among African-American patients. Dry cough was the most prevalent adverse event that led to medication discontinuation in white Americans.
There was a nonstatistically significant trend for the development of renal dysfunction in African-American patients for all treatments compared with white patients (12.4% for white patients vs. 15.9% for African Americans; p = 0.07). A longitudinal analysis showed that diuretic use over time and baseline renal insufficiency were strongly associated with renal dysfunction during the trial (p < 0.0001 for both factors). In addition to having more renal insufficiency at baseline, African Americans had a significantly higher number of individuals who developed renal insufficiency from the post-qualifying MI to randomization compared with white patients; the absolute difference was 5.58%. Furthermore, there were a significantly greater number of African Americans taking diuretics overall as well (p < 0.001). African Americans have significantly (p < 0.0001) higher rates of diuretic use before discontinuation of study drug (63.3%) compared with white patients (47.2%).
African Americans were more likely than white patients to develop renal dysfunction and hyperkalemia requiring valsartan discontinuation (Table 3). After adjusting for chronic renal insufficiency (eGFR <60 ml/min/1.73 m2) at baseline, the difference between African Americans and white American patients on developing renal dysfunction during the study was not significant (p = 0.1346). However, the difference between African Americans and white Americans on discontinuing valsartan treatment for renal causes (Table 3) persisted (p < 0.0001) after adjusting for baseline renal insufficiency.
In a high-risk population with HF or LVSD after an acute MI, African-American patients treated with captopril, valsartan, or both had similar clinical outcomes to white American patients, after adjusting for differences in comorbidities. Despite having significantly more coronary risk factors and more markers of poor outcome post-infarction, African Americans were less likely to be revascularized percutaneously or surgically after MI. This disparity remained after adjusting for the extent of coronary artery disease. The differential use of revascularization procedures according to race has been extensively reported and is associated with worse outcomes in several studies (16–19). Because there was a trend toward higher rates of HF hospitalizations for African Americans compared with white Americans at 3 years, we can only speculate as to whether this is attributable to treatment with calcium-channel blockers, more coronary risk factors and poor post-infarction outcome markers, or fewer revascularization procedures.
Conflicting results from several clinical trials involving pharmacological blockade of the renin-angiotensin-aldosterone system have led to questions about the efficacy of these agents in African Americans. In the AASK (African American Study of Kidney Disease and Hypertension), patients treated with the ACE inhibitor ramipril as the first drug of combination therapy, as compared with a beta-blocker or calcium antagonist as initial therapy, had a decrease in the progression of their renal disease (20,21). The ALLHAT (Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial) reported African Americans had more strokes, more HF, and less blood pressure reduction on an ACE inhibitor compared with a diuretic, but not a higher total mortality (13). In the LIFE (Losartan Intervention For End Point Reduction in Hypertension) study, losartan was less effective than atenolol as a component of antihypertensive therapy among African Americans in reducing fatal and nonfatal strokes (22). Retrospective analyses of 2 heart failure trials concluded that ACE inhibitors were less effective in African Americans (14,23). In the V-HeFT (Vasodilator-Heart Failure Trial) II study, enalapril decreased mortality in white patients, but not in African-American patients (14). In the SOLVD (Studies Of Left Ventricular Dysfunction) study, African-American patients had a poorer prognosis and did not have a reduction in total mortality or hospitalization for HF with enalapril compared with placebo (23). Subsequent reanalysis of the SOLVD dataset that adjusted for ventricular function was able to show only a weak nonsignificant trend toward excess mortality, and a persistent trend in increased hospitalizations for HF among African-American patients (4). There have been a number of trials with ACE inhibitors conducted in the post-infarction setting (6–10,24–26). However, only 1 trial, the SAVE (Survival And Ventricular Enlargement) study, included at least 125 African-American patients (27). Unfortunately, the number of African-American participants did not permit an analysis of risk reduction for total mortality or cardiovascular morbidity and mortality (28).
African Americans compose approximately 10% of the U.S. VALIANT study population, which is consistent with or larger than most clinical trials examining treatment outcomes by race. Excluding the AASK study, the A-HeFT (African American Heart Failure) trial (21,29), (both of which enrolled only African Americans), the ALLHAT study (13), and the BEST (Beta-Blocker Evaluation of Survival Trial) (30), very few large clinical trials have prospectively included enough African-American patients to conclude meaningful observations. The landmark HOPE (Heart Outcomes Prevention Evaluation) trial (31) included so few African-American patients that no definitive recommendations could be made regarding the efficacy of ramipril in African-American patients at risk for cardiovascular events (32). A retrospective analysis of the LIFE trial that included only 6% African Americans reported a lack of benefit in this subgroup for losartan compared with atenolol in the primary composite end point (cardiovascular death, stroke, and MI) (22). This observation ultimately led to a product label stating the benefits of losartan may not apply to African Americans (32). The data on the efficacy of ACE inhibitors in African-American patients with HF are largely derived from retrospective analyses of the SOLVD study, in which 12% of the population was African American (4,5,23).
Clinical studies indicate that the reduction in blood pressure for African-American patients compared with white patients is less with a beta-blocker, an ACE inhibitor, or an ARB when used as monotherapy (33–37). Consistent with those studies, African Americans had a diminished blood pressure response to ACE inhibitors and ARBs compared with white Americans, and consequently had a lower rate of clinically significant hypotension. Concordant with prior reports, the rate of angioedema was greater among African-American study participants compared with white American participants (13,38). In the ALLHAT study, angioedema occurred in 0.7% of African-American patients assigned to lisinopril versus 0.3% of non-African Americans. In our analysis of the U.S. portion of the VALIANT study, angioedema by race was significantly different for patients treated with captopril, but did not reach statistical significance by race across all 3 treatments. The VALIANT study excluded previous intolerance or contraindication to an ACE inhibitor, and the number of African-American patients was relatively small.
The VALIANT study was not specifically designed to explore outcomes according to race, and results are limited by the sample size. Although the number of African Americans participating is relatively small, this report seems to be the largest cohort for an individual HF post-infarction trial treating patients with either an ACE inhibitor or an ARB. The African-American and white American patients compared in this analysis differed at baseline in several demographic and clinical factors. Statistical modeling adjusted for many factors, yet there is the possibility of unaccounted factors that may affect the results reported in this study. The use of hydralazine and isosorbide dinitrate has been shown to decrease mortality in African-American patients with chronic moderate-to-severe HF. The use of these medications was low in this study (<4%); however, the VALIANT trial is a post-MI study and the benefits of hydralazine and isosorbide dinitrate may not be generalizable to this population.
In this analysis of U.S. participants in the VALIANT study, there were no racial differences among post-MI patients with HF and/or LVSD treated with captopril, valsartan, or the combination in the 3-year rates of all-cause mortality, cardiovascular mortality, and recurrent MI. Hospitalization rates for HF and stroke were higher for African-American patients compared with white patients, although the p values for comparison were nonsignificant. A larger sample size may have yielded significant results. White patients experienced more hypotension than African Americans, and African Americans experienced more angioneurotic edema than white patients treated with an ACE inhibitor. African-American patients with acute MI complicated by HF and/or LVSD are as likely to benefit from treatment with ACE inhibitors or ARBs as white patients. It remains important to prospectively include minority populations in clinical trials to ensure that the potential benefits and adverse events of medications, devices, and other interventions are generalizable to all patients.
For the statistical methods to estimate adjusted event rates and a supplementary table, please see the online version of this article.
All of the authors have received grant support or consulting fees from the sponsor of the VALIANT (VALsartin In Acute myocardial iNfarcTion) study, Novartis, as well as from multiple other manufacturers of cardiovascular drugs.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- angiotensin receptor blocker
- coronary artery bypass grafting
- coronary heart disease
- estimated glomerular filtration rate
- heart failure
- left ventricular systolic dysfunction
- myocardial infarction
- percutaneous coronary intervention
- Received December 1, 2006.
- Revision received December 7, 2007.
- Accepted December 11, 2007.
- American College of Cardiology Foundation
- Gheorghiade M.,
- Bonow R.O.
- Cooper R.,
- Cutler J.,
- Desvigne-Nickens P.,
- et al.
- Pfeffer M.A.,
- Braunwald E.,
- Moye L.A.,
- et al.,
- The SAVE Investigators
- Ambrosioni E.,
- Borghi C.,
- Magnani B.
- Køber L.,
- Torp-Pedersen C.,
- Carlsen J.E.,
- et al.,
- Trandolapril Cardiac Evaluation (TRACE) Study Group
- Swedberg K.,
- Held P.,
- Kjekshus J.,
- Rasmussen K.,
- Rydén L.,
- Wedel H.
- Dickstein K.,
- Kjekshus J.
- ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group
- Julius S.,
- Alderman M.H.,
- Beevers G.,
- et al.
- ISIS-4 (Fourth International Study of Infarct Survival) Collaborative Group
- Moye L.A.,
- Pfeffer M.A.,
- Wun C.C.,
- et al.
- Yancy C.W.,
- Benjamin E.J.,
- Fabunmi R.P.,
- Bonow R.O.