Your Family Physician

Changes in natriuretic-peptide levels over the first five days of hospitalization for acute decompensated heart failure (ADHF) significantly and independently predict both short- and intermediate-term survival, probably better than mere baseline biomarker levels, conclude researchers, on the basis of their post hoc analysis of a randomized trial [1].

The group reviewed in-hospital changes in levels of brain-type natriuretic peptide (BNP) in the Survival of Patients With Acute Heart Failure in Need of Intravenous Inotropic Support (SURVIVE) trial and compared them with all-cause mortality at one and six months in the study's patients as a whole, rather than by treatment group. SURVIVE was a randomized comparison of levosimendan (Simdax, Orion) vs dobutamine in ADHF patients in need of inotropic support [2].

The authors, led by Prof Alain Cohen-Solal (Hôpital Lariboisière, Université Denis Diderot, Paris, France), found significant independent associations between mortality and change in BNP from baseline to days 1, 3, and 5. A >30% BNP decline or an absolute drop of at least 800 pg/mL by day 5 provided the best prediction, according to their report in the June 23, 2009 issue of Journal of the American College of Cardiology.

"Such early and reliable prediction information offers the advantage of achieving early optimization of treatment," according to the group. "Consistent with our current findings in patients hospitalized with ADHF, we hypothesize that survival can be improved when the in-hospital treatment algorithm is aimed at decreasing BNP" by at least 30% or 800 pg/mL over five days, they write, cautioning that the strategy remains to be tested in prospective trials.

BNP: Biomarker, But Also Mortality Surrogate?

An accompanying editorial accepts that in-hospital BNP reductions seemed to predict one- and six-month mortality in the analysis but is skeptical of the group's resulting hypothesis that therapy aimed at BNP-lowering in the first days of ADHF hospitalization might improve later survival [3].

"Although the authors clearly state that this requires confirmatory testing, such a hypothesis has important implications for clinical practice and clinical trials in acute heart-failure syndromes because the number of patients hospitalized with heart failure continues to grow," according to Drs Mihai Gheorghiade and Peter S Pang (Northwestern University Feinberg School of Medicine, Chicago, IL).

Most patients with ADHF "appear to improve" during hospitalization with available treatments, yet they still have a high risk of rehospitalization, and their 60-day mortality "can be as high as 30% to 40%," they write. "Attempts to improve clinical outcomes with novel therapies have largely failed in terms of efficacy and/or safety." A surrogate marker for outcomes would in fact be welcome, they say, considering the difficulty and cost of conducting large mortality trials in ADHF.

But neither SURVIVE nor the levosimendan studies before it, write Gheorghiade and Pang, support the idea that "reductions in BNP are a reliable surrogate for postdischarge mortality as a result of early therapeutic optimization." A lot of data, in fact, suggest that "decreases in BNP may be associated with no benefit or even worse outcomes."

Commenting to heartwire , Cohen-Solal called the editorial "unfair" for its focus on his group's "hypothesis" that BNP reductions might lead to improved outcomes, while their analysis and actual conclusions were limited to the predictive value of BNP changes.

SURVIVE had randomized 1327 ADHF patients to receive infusions of either levosimendan, a calcium-sensitizing agent with both inotropic and vasodilating properties, or the familiar inotrope dobutamine. Entry to the study required an LVEF <30%>heartwire , the two groups fared the same with respect to the primary end point of all-cause mortality at 180 days; the rates were 26% and 28%, respectively.

Levosimendan is marketed in Europe and Latin America, but plans to gain approval in the US were abandoned following completion of SURVIVE [4].

In the current analysis, patients with >30% drop in BNP levels over five days had a 67% reduction in adjusted risk of death from any cause over the next month and a 47% decrease over the next six months. The risk reductions were independent of randomization group as well as such baseline features as BNP level, systolic and diastolic blood pressures, use of ACE inhibitors and beta blockers, and creatinine.

Rates and Hazard Ratios (Responders vs Nonresponders) for All-Cause Mortality at 31 and 180 Days

End point	Responders (%)	Nonresponders (%)	HR (95% CI)	p
31-d mortality	4.6	13.1	0.33 (0.21–0.52)	0.001
180-d mortality	17.9	30.3	0.53 (0.41–0.68)	0.001

*Responders defined as those with >30% BNP decrease from baseline at day 5 (of the trial's 1038 patients with available day-5 BNP levels).

On the other hand, the group writes, mortality at 31 days was tripled in patients with increased BNP levels from baseline to day 5, compared with those with a >30% decrease. The latter group, compared with those without a >30% BNP decrease, had significantly less cardiogenic shock and ventricular tachycardia over 180 days and similar rates of ACS, hypertension, renal failure, and acute MI.

BNP-Guided Therapy in Context

"There have been a lot of trials in chronic heart failure looking at whether tailoring therapy to reduce BNP levels would improve outcomes, but they used proven therapies like ACE inhibitors and beta blockers," Dr Monica R Shah (National Heart, Lung, and Blood Institute, Bethesda, MD) told heartwire . Some of those trials suggested that BNP-guided therapy was an improvement over standard management and some, like the STARBRITE trial for which Shah was principal investigator, showed no difference.

In ADHF, "a 30% decrease in BNP may just mark patients who are going to do better," she said. It may be possible to improve outcomes by pushing BNP down that far over five days, "but it's also a possibility that patients more likely to get better are also going to have a 30% decrease in their BNP simply because their disease isn't as advanced. Patients who are sicker may not be able to lower their BNP by that much regardless of the therapy used."

And if treatment with a drug actually does seem to lower biomarker levels, it could also have detrimental effects. "A drug may reduce BNP levels but may also increase the risk for ventricular arrhythmias and sudden death. That was the case with inotropes," Shah said.

"We all know that BNP is a good prognostic marker for outcomes. But we really don't have a marker that totally captures all of the effects of any drug."

Gheorghiade and Pang point out that in another levosimendan trial called REVIVE, which had a placebo control but enrollment criteria similar to those of SURVIVE, patients who received the active drug showed short-term BNP reductions but also more ventricular tachycardia, atrial fibrillation, and hypotension and a trend of increased mortality in the first five days. "It should be pointed out that the analysis by Cohen-Solal et al excluded the 69 patients (5.2%) who died within the first five days," they write. And in neither trial did five-day reductions in BNP with levosimendan translate into improved outcomes.

"We're not saying that we should aim to decrease BNP whatever the method used," said Cohen-Solal, who freely acknowledged that levosimendan was not associated with improved survival compared with dobutamine in SURVIVE. "We are saying that, overall, an early decrease in BNP is a very, very important predictor of outcome," whereas BNP as a treatment target was just a proposal for future research. There is now justification, he said, for a prospective trial comparing standard management with one in which "you do everything to decrease BNP by more than 30% by day 5."

SURVIVE was funded by Abbott Laboratories and Orion. The report discloses that coauthors Drs Bidan Huang and Danlin Cai are Abbott employees. Cohen-Solal and coauthors Dr Markku S Nieminen (University Central Hospital, Helsinki, Finland) and Dr Alexandre Mebazaa (Hôpital Lariboisière, Université Denis Diderot) report receiving honoraria from Abbott; Nieminen reports being a consultant for Orion; and Dr Damien Logeart (Hôpital Lariboisière, Université Denis Diderot) reports receiving fees from Biosite and Roche Diagnostics. Gheorghiade "is or has been a consultant for and/or received honoraria" from Abbott, Astellas, Bayer, AstraZeneca, Corthera, DebioPharm, Erre-Kappa Terapeutici, EKR Therapeutics, GlaxoSmithKline, Johnson & Johnson, Medtronic, Merck, Nile Therapeutics, Novartis, Otsuka, PeriCor Therapeutics, PDL BioPharma, Scios, Solvay Pharmaceuticals, and SigmaTau. Pang "is or has been a consultant" to Astellas, Bayer, the Medicines Company, Otsuka, Nile Therapeutics, PDL BioPharma, Pericor Therapeutics, and Solvay Pharmaceuticals; has received honoraria from Biogen Idec, Corthera, EKR Therapeutics, and Palatin Technologies; and has received research support from Corthera, Merck, and PDL BioPharma. Shah said she has no conflicts.

Source : http://www.medscape.com/viewarticle/704586?sssdmh=dm1.487338&src=nldne