Hope Rekindled for ACE Inhibitors, ARBs in Preserved-EF Heart Failure

May 24, 2012

May 24, 2012 (Belgrade, Serbia) — Renin-angiotensin-system (RAS)-blocking agents were seen to significantly improve survival in patients with heart failure with preserved ejection fraction (HF-PEF) in an observational study of patients representing most of Sweden [1].

The finding is at odds with at least three major randomized trials that had all but dashed previously high expectations for this use of the drugs. But the cohort study's investigators allege that there are important limitations to those trials, which tested ACE inhibitors and angiotensin receptor blockers (ARBs), and that they showed signals of possible benefit that ought to keep the RAS-inhibitor door open for HF-PEF.

The investigators aren't arguing that their study, which included a propensity-matched cohort analysis, carries more weight than controlled trials. Still, it suggests that "one should consider RAS antagonists for a likely but not proven benefit" in HF-PEF until the question is settled by an "adequately powered randomized controlled trial," said Dr Lars H Lund (Karolinska Institutet, Stockholm, Sweden) here at the Heart Failure Congress 2012 sessions of the European Society of Cardiology Heart Failure Association.

"Representative of Sweden for Sure"

His group's analysis of >3000 propensity-matched pairs of patients with HF-PEF in the Swedish Heart Failure Registry saw a 9% (p=0.008) decline in all-cause mortality associated with treatment with either ACE inhibitors or ARBs.

The matched pairs were taken from >16 000 patients with LVEF >40% who were part of an overall cohort of >64 000 heart-failure patients in the registry. They included 12 543 who were on RAS inhibitors and 3673 not taking them.

"I think our population is representative of clinical practice. It covers almost all hospitals in the country. It's representative of Sweden for sure," Lund told heartwire . The unadjusted HR for the effect of RAS inhibition on all-cause mortality at one year in the overall cohort was 0.48 (p<0.001.)

The 3329 matched pairs were based on a propensity-scoring system that included 42 "relevant covariates" that were not only clinical and hemodynamic in nature but also socioeconomic, Lund said in his presentation. Adjusted that way, the HR for the same end point remained significant at 0.91 (p=0.008).

That compares with adjusted HRs of:

  • 0.86 (p=0.051) for the effect of the ARB candesartan on cardiovascular death or HF hospitalization in the CHARM-Preserved trial.

  • 0.92 (p=0.545) for the ACE inhibitor perindopril's effect on death or HF hospitalization in PEP-CHF.

  • 0.95 (p=0.60) with another ARB, irbesartan, for death or CV hospitalization in I-PRESERVE.

As the formal discussant following Lund's presentation, I-PRESERVE co–principal investigator Dr Barry M Massie (University of California and San Francisco VA Medical Center) critiqued the Swedish cohort analysis and defended the three randomized trials.

"I think probably the fact that most concerns me about this study is the definition of HF-PEF," he said. "I think very few people would say that an ejection fraction of 40% is normal in a chronic heart-failure population. Generally the trials that have been conducted have had cut points of 45% or 50%, and those are what I personally consider relatively preserved."

"There Were Trends"

Interviewed, Lund acknowledged their analysis is limited in being solely observational but said there were aspects of the three trials that were worth further exploration. "There were trends," he said. "Call it signals, interactions, whatever, but heart-failure hospitalizations were reduced." Also, patients in those trials had N-terminal pro–brain natriuretic peptide [NT-proBNP] levels "in the 300 to 400 [ng/L] range," he said. "That's barely heart failure."

Whether it was "stated or unstated," according to Lund, the designers of the three trials "didn't want a lot of comorbidity, so they had younger patients with normal renal function and relatively good functional status." He added, "It's a classic example of where the real-life populations get excluded from randomized trials."

In the matched cohort, he emphasized, the mean age was 79, while in the trials it was in the upper 60s to lower 70s. Also, renal function was more compromised, and the mean NT-proBNP levels were 4577 ng/L and 5192 ng/L for the two groups respectively. "We believe our registry [represents] older patients, sicker patients and--we propose--is more real life."

Massie acknowledged that PEP-CHF was underpowered in part because of a low overall rate of primary events and a substantial number of patients dropping out of the trial. But he was adamant that wasn't the case with I-PRESERVE.

"I know [Lund] thought it was underpowered, but [I-PRESERVE had] 1500 events for the primary end point. I don't believe it really was an underpowered trial. I think what we saw is what we got, and I think it's probably the right answer. I would not want to conclude from I-PRESERVE that there is an effect of RAS blockers."

As an aside, Massie pointed out that in the DIG trial HF-PEF arm, in which the LVEF cutoff was 45%, digoxin was associated with an HR for cardiovascular mortality or HF hospitalization of 0.88 (p=0.209) for the entire follow-up period but 0.75 (p=0.044) at two years. "Digoxin thus far has proved to be perhaps the most effective intervention in heart failure with preserved ejection fraction."

As for the near future, he said, the long-awaited Treatment of Preserved-Cardiac-Function Heart Failure (TOPCAT) trial, "which is over but not yet fully analyzed," entered 3200 patients with heart failure, an LVEF >45%, and at least one of the following indicators of increased risk: HF hospitalization within the past 12 months, brain-type natriuretic peptide (BNP) >100 pg/mL, or an NT-pro-BNP >360 pg/mL.

TOPCAT, funded by the National Heart, Lung, and Blood Institute, randomized the patients to placebo or the aldosterone antagonist spironolactone and followed them for an average of 3.25 years for the primary end point of cardiovascular death, aborted cardiac arrest, or HF hospitalization.

"All patients have completed their follow-up, and we expect to get a result toward the end of this year," Massie said. "And hopefully we will get a treatment, and it may well be a RAS blocker that's effective."

Lund discloses consulting for and receiving research funding and speaker's fees from AstraZeneca.