Meta-Analysis of Large-Scale Randomized Trials to Determine the Effectiveness of Inhibition of the Renin-Angiotensin Aldosterone System in Heart Failure




Renin-angiotensin-aldosterone system (RAAS) inhibition is 1 of the most effective strategies for the management of heart failure with reduced systolic function. However, trials that included patients with preserved systolic function have not shown a clear beneficial effect. Pooling evidence from several heart failure trials provides the opportunity to better assess the differential effects of RAAS inhibition across the continuum of systolic function. The authors searched MEDLINE for large-scale trials published from 1966 to March 2014 that compared RAAS inhibitors against placebos. Studies were eligible for inclusion if they were conducted in heart failure populations with either clinical signs of heart failure or reduced ejection fractions. Inverse variance–weighted fixed-effects meta-analysis was used to pool outcomes of interest, with metaregression used to test for trends. In 16 trials with 54,621 randomized heart failure participants, RAAS inhibition reduced the risks for hospitalization for heart failure by 20% (relative risk [RR] 0.80, 95% confidence interval [CI] 0.77 to 0.83), cardiovascular mortality by 14% (RR 0.86, 95% CI 0.83 to 0.90), and all-cause mortality by 11% (RR 0.89, 95% CI 0.85 to 0.92). However, proportional effects decreased with increasing mean left ventricular ejection fraction (LVEF) for all outcomes (p for trend <0.01). Although there was no significant proportional effect on cardiovascular and all-cause mortality in trials with a mean LVEF >50%, RAAS inhibition was still found to decrease the risk for heart failure hospitalization in patients with preserved LVEFs (RR 0.88, 95% CI 0.80 to 0.97). In conclusion, the relative beneficial effects of RAAS inhibition in heart failure decreases with increasing left ventricular systolic function. Nonetheless, RAAS inhibition significantly reduces the risks for all-cause mortality and cardiovascular mortality in patients with moderately reduced LVEFs and the incidence of hospitalization in patients with preserved left ventricular function.


Despite recent advances in prevention and management, heart failure (HF) still causes significant mortality and morbidity worldwide. Renin-angiotensin-aldosterone system (RAAS) inhibitors, including angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), and aldosterone receptor antagonists (ARAs), reduce the risk for mortality and hospitalization in patients with existing HF. However, the benefit of these therapies has been limited largely to patients with HF with reduced ejection fraction, as no single trial has demonstrated a therapy to be effective in populations with HF with preserved ejection fraction (HFpEF). This lack of an observed effect in HFpEF populations may be due to the relatively small number of events in individual trials conducted in HFpEF populations. Alternatively, RAAS inhibition may benefit only certain subgroups of patients with HFpEF, an effect that is diluted in trials conducted in a broad range of patients with HFpEF. Although guidelines have only considered individual trials in isolation, a pooled analysis of the totality of evidence offers an ideal approach to investigate the overall benefits of RAAS inhibition across the spectrum of HF populations.


Methods


This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A search strategy, previously designed to identify blood pressure–lowering trials, was used to identify all trials that tested antihypertensive agents (regardless of whether they were used for blood pressure lowering) in any patient population. In brief, search terms were “anti-hypertensive agents,” “hypertension,” “diuretics, thiazide,” “angiotensin-converting enzyme,” “receptors, angiotensin/antagonists & inhibitors,” “tetrazoles,” “calcium channel blockers,” “vasodilator agents,” the names of all blood pressure–lowering drugs listed in the British National Formulary as keywords or text words, and the Medical Subject Headings term blood “pressure/drug effects.” Using this strategy, MEDLINE was searched by an experienced research librarian in May 2014 for randomized trials published from 1966 to March 2014. Additionally, bibliographies of identified trials, reviews, and meta-analyses were searched by hand to further identify trials. After identification of all large trials testing antihypertensive agents, trials were examined as to whether they (1) were conducted in HF populations and (2) compared RAAS inhibitors against placebos. We defined RAAS inhibition as treatment with ACE inhibitors, ARBs, or ARAs. Trials that compared different doses of RAAS inhibitors (which did not also include placebo arms) were excluded.


Although ACE inhibitors, ARBs, and ARAs act on different receptors, we sought to combine all 3 agents to comprehensively examine the effect of RAAS inhibition. ACE inhibitors and ARBs are often considered to be and are analysed as of the same class. However, we chose not to distinguish, a priori, among ACE inhibitors, ARBs, and ARAs, as their differential effects in HF populations are uncertain. Instead, we chose to pool all 3 classes of medications, to examine the totality of the evidence, and test for heterogeneity among classes of medications as appropriate.


We included all large-scale (>1,000 patient-years) trials using RAAS inhibitors. Because we were interested in the effect of RAAS inhibition in HF populations, not the comparative effectiveness of RAAS inhibitors, we restricted our analysis to placebo-controlled trials. No restrictions were put in place on the definition of HF. As a result, we included trials that randomized patients with clinical signs of HF without reduced ejection fraction as well as those who had reduced ejection fractions with or without clinical signs. Trials conducted after acute myocardial infarction were also eligible for inclusion if they meet the broad definition of HF.


Abstracts were screened independently by 2 researchers (CAE and JC), with potentially eligible studies screened in full text by 2 researchers (CAE and TC). Data extraction was also performed independently by 2 researchers (CAE and JC). The number of individuals in the treatment and control arms, the mean left ventricular ejection fraction (LVEF), baseline blood pressure (systolic and diastolic), doses of medications, titration, inclusion criteria, and numbers of all-cause mortality events, cardiovascular mortality events, and HF hospitalizations in both arms were recorded. If a hazard ratio was provided, we used that ratio, because hazard ratios avoid censoring associated with crude relative risks. However, a relative risk was derived for trials that did not report hazard ratios, using reported numbers of events and numbers of individuals randomized to each arm.


For 2 trials conducted after acute myocardial infarction, HF outcomes were defined as progression to severe HF, which included hospitalization for HF and open-label use of ACE inhibitors. These outcomes were treated as hospitalization for HF for meta-analysis. For 1 trial, the Valsartan in Acute Myocardial Infarction Trial (VALIANT), hospitalization for HF was not published; as a result, this outcome was acquired directly from a member of the VALIANT steering committee. Although VALIANT compared a dual-inhibition arm (ACE inhibitor and ARB) with 2 single-inhibition arms (ACE inhibitor and ARB), this is equivalent to comparing an ARB against a placebo on a background therapy of ACE inhibitor and an ACE inhibitor against a placebo on a background therapy of ARB. Therefore, we chose to include the trial and randomly selected a single treatment arm (the ACE inhibition arm) to be compared with the combination (ACE inhibitor and ARB) treated arm (resulting in the ACE inhibitor and ARB arm being compared with the ACE inhibitor, arm equivalent to an ARB compared against a placebo against background ACE inhibitor therapy).


Methodologic quality was assessed using Cochrane’s risk for bias tool. Each trial was evaluated on 5 dimensions: selection bias (randomization and allocation concealment), performance bias (blinding of participants and investigators), detection bias (blinding and reliability of evaluation), reporting bias (selective outcome reporting), and attrition bias (incomplete data). After judging each dimension as either low, unclear, or high risk for bias, the trial, as a whole, was judged to be of low, unclear, or high risk for bias depending on whether individual bias was likely to have affected reported outcomes.


For all analyses, overall effect estimates were calculated using inverse variance–weighted fixed-effects analysis, with 95% confidence intervals (CIs). Heterogeneity among subgroups was quantified using the I 2 statistic and tested using Cochran’s Q statistic. Trend across subgroups was tested using metaregression with the LVEF as the modifier of interest, treated as a continuous variable. Sensitivity analyses using random-effects meta-analysis was performed in the presence of statistically significant heterogeneity (p <0.05). An additional sensitivity analysis removing trials testing combined ACE inhibition and ARB (Candesartan in Heart Failure Reduction in Mortality–Added, VALIANT, and the Valsartan in Heart Failure Trial) against single therapy, was also performed. An analysis excluding VALIANT alone was performed, because VALIANT used a half dose of valsartan in the combination treatment arm. Finally, a sensitivity analysis excluding trials conducted immediately after myocardial infarction was conducted.


Analyses were performed to address: (1) whether RAAS inhibition across all HF patients reduces the risk for fatal outcomes and hospitalization for worsening HF and (2) whether benefits of RAAS inhibition differ by mean levels of the LVEF, for which trials were classified by the mean ejection fraction into quintiles as well as categories of 20% to 29%, 30% to 39%, 40% to 49%, and >50%. All analyses were conducted with R version 3.1.1.




Results


A total of 10,143 abstracts were identified and screened ( Supplementary Figure 1 ). One hundred thirty-one randomized trials were identified, of which 115 were excluded for not fulfilling inclusion criteria. Sixteen trials comparing RAAS inhibitors against placebos were therefore identified. Twelve trials had mean LVEFs of ≤40%, while 4 trials had mean LVEFs >40% ( Table 1 ). In total, 54,621 patients with 11,488 fatal all-cause events, 8,309 fatal cardiovascular events, and 9,190 HF hospitalization events were included in the analysis. Fifteen trials were judged to possess low risk for bias ( Supplementary Table 1 ). One trial was judged to be of unclear risk for bias. Dosages, titrations, and other characteristics of trials are provided ( Supplementary Table 2 ).



Table 1

Summary of characteristics of trials comparing antihypertensive therapy against placebo or against active control

























































































































































































































Name Heart Failure Class EF Participants Intervention Control Mean EF Duration
(Yrs)
Age
(Yrs)
Male Baseline BP
(Mm Hg)
Heart failure with reduced ejection fraction
CHARM Added II to IV ≤40% 2548 Candesartan Placebo 28% 3.4 64 79% 125/75
CHARM Alternative II to IV ≤40% 2028 Candesartan Placebo 30% 2.8 67 68% 130/77
EMPHASIS HF II ≤30% 2737 Eplerenone Placebo 26% 1.8 69 78% 124/75
RALES III or IV ≤35% 1663 Spironolactone Placebo 25% 2 65 73% 122/75
SOLVD-prevention I ≤35% 4228 Enalapril Placebo 28% 3.1 59 89% 125/78
SOLVD-treatment II to IV ≤35% 2569 Enalapril Placebo 25% 3.5 61 80% 125/77
Val-HeFT II to IV <40% 5010 Valsartan Placebo 27% 1.9 63 80% 124/76
Post myocardial infarction with reduced ejection fraction
AIRE AMI with clinical evidence of HF NA 1986 Ramipril Placebo NA 1.3 65 74% NA
EPHESUS AMI with signs of HF or diabetes ≤40% 6632 Eplerenone Placebo 33% 1.3 64 71% 119/72
SAVE AMI ≤40% 2231 Captopril Placebo 31% 3.5 59 83% 113/70
TRACE AMI ≤35% 1749 Trandolapril Placebo 30% 2.2 68 72% 121/76
VALIANT AMI with signs of HF ≤40% 9794 Valsartan Control 35% 2.1 65 69% 123/72
Heart failure with preserved ejection fraction
CHARM preserved II to IV >40% 3020 Candesartan Placebo 54% 3.1 67 60% 136/79
I-PRESERVE II to IV ≥45% 4128 Irbesartan Placebo 60% 4.1 72 40% 137/79
PEP-CHF I To IV ≥40% 850 Perindopril Placebo 65% 2.2 75 45% 139/80
TOPCAT II to IV ≥45% 3445 Spironolactone Placebo 56% 3.3 69 48% 130/80


Overall, RAAS inhibition reduced the risk for all-cause mortality by 11% (relative risk [RR] 0.89, 95% CI 0.85 to 0.92), the risk for cardiovascular mortality by 14% (RR 0.86, 95% CI 0.83 to 0.90), and the risk for hospitalization for HF by 20% (RR 0.80, 95% CI 0.77 to 0.83) ( Figure 1 ). Although significant heterogeneity was observed among trials, random-effects estimates for all-cause mortality, cardiovascular mortality, and HF hospitalization did not differ materially from the fixed-effects estimates ( Supplementary Figures 2–4 ).




Figure 1


Effect of RAAS inhibition on all-cause mortality, cardiovascular mortality, and hospitalization for HF.


When trials were classified by the mean ejection fraction into quintiles, proportional benefits of therapy decreased with increasing mean LVEF in trials. Significant trends were observed for all-cause mortality ( Figure 2 ; p for heterogeneity = 0.04, p for trend = 0.03), cardiovascular mortality (p for heterogeneity = 0.006, p for trend = 0.005), and HF hospitalization (p for heterogeneity <0.0001, p for trend = 0.005). Despite decreasing proportional benefits with an increasing mean ejection fraction, significant benefit of RAAS inhibition on hospitalization for HF continued to be observed in the fourth quintile (mean ejection fraction 43%), with therapy reducing the risk for hospitalization by 8% (RR 0.92, 95% CI 0.85 to 0.99). A trend toward benefit of therapy was also observed in the fourth quartile of trials for all-cause mortality and cardiovascular disease mortality (mean ejection fractions 43% and 41%, respectively), with nonsignificant reductions of 6% for all-cause mortality (RR 0.94, 95% CI 0.87 to 1.02) and 6% for cardiovascular disease mortality (RR 0.94, 95% CI 0.86 to 1.01). When VALIANT, which used a half-dose ARB in the treatment arm, was excluded, significant benefit of RAAS inhibition on all-cause mortality and cardiovascular disease mortality was observed in the fourth quartile of trials (RR 0.87, 95% CI 0.80 to 0.95, and RR 0.83, 95% CI 0.78 to 0.98, respectively; mean LVEFs 42% and 44%, respectively; Supplementary Figure 5 ). Results did not differ materially when random-effects meta-analysis were used ( Supplementary Figure 6 ), when trials testing dual ACE inhibition and ARB were excluded ( Supplementary Figure 7 ), or when trials conducted after myocardial infarction were excluded ( Supplementary Figure 8 ).


Nov 28, 2016 | Posted by in CARDIOLOGY | Comments Off on Meta-Analysis of Large-Scale Randomized Trials to Determine the Effectiveness of Inhibition of the Renin-Angiotensin Aldosterone System in Heart Failure

Full access? Get Clinical Tree

Get Clinical Tree app for offline access