Relaxin for Treatment of Acute Heart Failure: Making the Case for Treating Targeted Patient Profiles
- First Online:
- Cite this article as:
- Hernandez-Montfort, J.A., Arora, S. & Slawsky, M.T. Curr Heart Fail Rep (2013) 10: 198. doi:10.1007/s11897-013-0148-6
Patients presenting with acute heart failure (AHF) represent a heterogeneous population with respect to demographics, clinical profiles, and precipitating factors. Despite this, most clinical trials have treated the study population as a homogeneous group in an attempt to achieve adequate statistical power for endpoint analysis. This approach has proven to be of little value in the development of new agents for treatment of AHF. By contrast, the phase III clinical trial of relaxin focused on a subset of AHF patients who were normotensive or hypertensive and who had moderate renal impairment. The study patients, who were primarily from Eastern Europe, represented a population that would be expected to have less genetic variability than the study populations in larger multinational AHF trials. A focused study design targeting specific patient profiles should be considered for future clinical AHF trials that investigate new therapies or compare the effectiveness of existing therapies.
KeywordsRelaxinAcute heart failureClinical trialsVasodilatorBlood pressure
Acute Heart Failure: Scope of the Problem and Current Management
Acute heart failure (AHF) is a major global health challenge with a high prevalence, significant morbidity and mortality, and a high cost of care [1–3]. In patients older than 65 years, AHF is the most common cause of hospitalization and represents a pivotal event in the patient’s life expectancy, with an estimated 1-year mortality rate reaching 40 % [4, 5]. The presence of end-organ damage and persistent congestion during the first days after admission for AHF is also associated with a poor prognosis [3, 6–8].
Administration of diuretics has been the mainstay of acute (and chronic) heart failure (HF) management. The use of diuretics together with salt and fluid restriction is based on the therapeutic concept of lost homeostasis via fluid and sodium retention . However, there is scant evidence to support the benefit of diuretics beyond symptomatic relief . Studies incorporating the use of vasodilators (ASCEND) or inotropic agents with vasodilatory properties (OPTIME-CHF, REVIVE-II) have also failed to show additional benefit in relief of dyspnea or in morbidity/mortality endpoints [11–13]. The failure of previous clinical AHF trials to demonstrate the superiority of a study drug has usually been attributed to the drug’s ineffectiveness or flaws in study design; however, the possibility that drug efficacy could be diluted by population heterogeneity has not been addressed . The large AHF registries clearly demonstrated that the AHF population is heterogeneous [1, 15]. This has led to awareness of distinct clinical AHF profiles that may benefit from the use of targeted therapies [16, 17••].
Relaxin for Treatment of AHF
It has been suggested that the clinical profile of vascular stiffness in AHF patients with HF and preserved ejection fraction (HFpEF) and/or hypertension would be better suited to vasodilator, rather than diuretic therapy [18•]. The early use of vasodilator therapy has been promoted for treatment of AHF in normotensive or hypertensive patients, but not in patients with “low blood pressure.” According to the 2012 ESC guidelines, the recommended systolic blood pressure (SBP) cutoff for AHF therapy with intravenous nitroglycerin is <110 mmHg . Clinical trials examining the efficacy of vasodilator therapy with relaxin in AHF have targeted this patient profile.
Relaxin, a 6KDa peptide hormone, with structural similarities to insulin, is produced by the corpus luteum . The increase in relaxin concentration that occurs during pregnancy results in systemic and renal vasodilation . The underlying mechanism involves relaxin binding to a G-protein coupled receptor pathway located on vascular endothelial cells. This results in activation and upregulation of the endothelin B receptor, leading to increased production of the vasodilator, nitric oxide [22•].
In a small, dose-ranging hemodynamic study of an 8-h infusion of relaxin in stable HF patients, the most pronounced effects on right atrial pressure, pulmonary capillary wedge pressure, mean pulmonary artery pressure, and systemic vascular resistance were at lower doses of 10–100 mcg/kg/day. Patients in the lower relaxin dose group tended to have a higher blood pressure at baseline (mean SBP 129 mmHg, as compared with 109 mmHg in the higher dose group) and a more marked blood pressure decrease in response to drug (−5 to −13 mmHg vs. +2 to −3 mmHg) .
Clinical Trials of Relaxin in Acute Heart Failure
Phase II Study: Pre-RELAX-AHF
Pre-RELAX-AHF was a double blind, multicenter, placebo-controlled randomized dose-ranging study of relaxin in 234 AHF patients. Patients were randomized within 16 h of presentation with AHF, defined by the presence of dyspnea at rest or on minimal exertion, pulmonary congestion on a chest radiograph, and increased natriuretic peptide levels (BNP ≥350 pg/ml or NT-proBNP ≥1,400 pg/ml), after receiving at least 40 mg of intravenous furosemide or the equivalent. Inclusion criteria included systolic blood pressure ≥125 mmHg and impaired renal function (estimated GFR of 30–75 mL/min/1.73 m2). The study population was elderly (mean age 70 years) and hypertensive, with a mean SBP of 147 mmHg.
Enrolled patients received an infusion of relaxin (10, 30, 100, or 250 mcg/kg/day) or placebo over 48 h. Clinical endpoints included improvement in dyspnea as assessed by a Likert scale or a visual analogue scale (VAS), worsening of HF or renal function from baseline to day 5, persistence of worsened renal function between days 5 and 14, length of hospital stay, days alive and out of hospital to day 60, death due to cardiovascular causes or readmission for HF or renal failure to day 60, and cardiovascular mortality to day 180.
This study was not powered to assess for differences in endpoints; however, decongestion, as assessed by improvement in Likert dyspnea scores through 24 h, appeared greatest in patients treated with relaxin 30 mcg/kg/day (40 % with moderate to marked improvement in dyspnea, as compared with 23 % in the placebo group). More patients in the relaxin 30 mcg/kg/day group had resolution of systemic and venous congestion on examination, as compared with placebo. Given these findings, a relaxin infusion dose of 30 mcg/kg/day was chosen for the phase III clinical trial, RELAX-AHF.
With regard to other clinical endpoints, no significant improvement in renal function was observed in relaxin-treated groups, as compared with placebo. Length of stay tended to be shorter in patients treated with relaxin. The combined incidence of death from cardiovascular causes or readmissions at 60 days appeared to be lower in patients treated with relaxin (6.1 % vs. 17.2 % in the placebo group; p = .13). Similarly, there was an apparent reduction in 180-day cardiovascular mortality in the relaxin group (3 % vs. 14.3 % in the placebo group; p = .14) .
Phase III Study: RELAX-AHF
RELAX-AHF was a prospective, randomized, double blind, placebo controlled, parallel group trial in AHF patients that compared a 48-h intravenous infusion of relaxin (30 mcg/kg/day) with placebo, in addition to standard care. The inclusion criteria and definition of AHF were the same as in Pre-RELAX-AHF. Enrolled subjects (n = 1,161) had a mean age of 72 years, mean systolic blood pressure of 142 mmHg, and, on average, at least one hospitalization for HF in the year prior to enrollment.
Primary endpoints included change in dyspnea from baseline to day 5 using a VAS and moderate or marked improvement in dyspnea at 6, 12, and 24 h after drug administration as measured by a 7-point Likert scale. Secondary clinical endpoints were days alive and out of hospital up to day 60 and cardiovascular death or readmission for heart or renal failure before day 60.
There was significant improvement in dyspnea at day 5, as compared with baseline, for patients treated with relaxin, as defined by the area under the curve of the VAS. No significant difference in moderate or marked dyspnea relief was observed on evaluation with the Likert scale at 6, 12, and 24 h. Symptoms and clinical signs of congestion did, however, improve on day 2 after scheduled completion of study drug infusion.
There were no significant differences between the relaxin and placebo groups with regard to secondary endpoints of days alive and out of the hospital to day 60 or cardiovascular death or readmission to hospital for HF or renal failure before day 60. A reduction in cardiovascular death and all-cause mortality at 180 days was observed in the relaxin group, as compared with placebo (6.1 % vs. 9.6 %, p = .028, and 7.3 % vs. 11.3 %, p = .02, respectively).
Relaxin resulted in a greater reduction of SBP both during and for 24 h after infusion. More patients on relaxin, as compared with placebo (29 % vs. 18 %, p < .001), had to undergo protocol-defined blood-pressure-related study drug dose adjustment, resulting in 50 % dose reduction or discontinuation of the study drug. Most episodes resolved without requiring treatment with intravenous fluids. There was no rebound hypertension following discontinuation of relaxin. More patients in the placebo group (9 % vs. 6 % in the relaxin group, p = .03) had adverse events related to renal impairment .
It would be simplistic to ascribe the clinical benefit observed with relaxin in RELAX-AHF solely to its effect on lowering blood pressure. A recent study with the intravenous vasodilator nitroprusside suggested that lowering blood pressure might have a detrimental hemodynamic effect in hypertensive patients with HFpEF. In this study, mean SBP was higher in the HFpEF group (166 mmHg) than in the comparison HF with reduced ejection fraction (HFrEF) group (113 mmHg); however, mean pulmonary artery pressure and pulmonary capillary wedge pressure were not different at baseline. With a similar decrease in filling pressures, there was a greater drop in SBP and less increase in stroke volume and cardiac output in the HFpEF group. The HFpEF patients were 4 times more likely to experience a drop in stroke volume with nitroprusside. The difference in response to lowering filling pressures in HFpEF patients appeared to be due to decreased ventricular-arterial compliance in this group [26•]. The results of this study emphasized important differences in HF phenotypes that will require clarification when assessing response to vasodilator therapy.
RELAX-AHF: Biomarkers as Predictors of Drug Efficacy and Clinical Outcomes
In RELAX-AHF, plasma levels of specific biomarkers of organ injury were obtained at baseline, 24 and 48 h, days 3 and 4, if in hospital, and days 5, 14, and 60 after admission. Myocardial injury was defined by high-sensitivity troponin T (hs-cTnT) level ≥20 % from baseline; renal damage was defined as increases in cystatin-C ≥0.3 mg/dL and creatinine ≥0.3 mg/l from baseline. Hepatic damage was assessed as an increase in aspartate transaminase and alanine transaminase ≥20 % from baseline to day 2. Decongestion was defined as a decrease of NT-proBNP ≥30 % relative to baseline at day 2.
The use of relaxin in AHF patients in the RELAX-AHF trial was not associated with improvement in 60-day cardiovascular mortality and readmission rates, as compared with placebo. Consistent with previous findings from PRE-RELAX AHF, there was a significant reduction in cardiovascular death and all-cause mortality at 180 days in the relaxin-treated group that persisted even after adjustment for baseline characteristics. The study authors noted that the mortality curves separated during the initial hospitalization (after 5 days following study entry), suggesting an effect from early protective intervention with relaxin. In support of this contention, biomarker analysis demonstrated a significant decrease in levels of prespecified biomarkers of organ injury (heart, liver, and kidney) in the relaxin-treated group, as compared with control. There was also a significant decrease in proBNP, a marker of cardiac wall stress. Within 2 days of randomization, patients treated with relaxin had significant reductions of hs-cTnT, as compared with placebo (16.5 % vs. 27.3 %, p ≤ .0001). Patients receiving relaxin also had significantly lower NT-proBNP levels at day 2, when compared with placebo. However, there was no significant difference in hs-cTnT or NT-proBNP between the relaxin-treated group and placebo at day 5. By contrast, markers of renal injury and renal dysfunction—cystatin-C, BUN, and creatinine—were significantly greater from day 2 through day 5 in patients on placebo. Cystatin-C levels remained significantly elevated at day 14 in the placebo versus relaxin group. In addition, serum levels of uric acid, a marker of inflammation and oxidative stress, were significantly increased in the placebo group from day 1 through 5 . The relationship of increased uric acid levels to renal dysfunction in this group is uncertain and presents an interesting avenue for future study.
Both Pre-RELAX-AHF and RELAX-AHF reported an association between relaxin treatment during the index AHF admission and reduced cardiovascular and all-cause mortality at 180 days. While this is an interesting finding, causal effect has been questioned, given the lack of benefit in 60-day morbidity and mortality endpoints. Similarly, one cannot infer that the improvement in markers of myocardial damage and renal dysfunction observed with relaxin during the index AHF admission resulted in mortality benefit at 180 days without knowledge of intervening events. Further analysis of the relationship between biomarker elevation, renal dysfunction, readmission at 60 days, and mortality at 180 days is warranted.
In summary, the RELAX-AHF study represents a novel approach to evaluation and management of AHF by introducing therapy targeting a specific patient profile and incorporating biomarker data as an adjunct to assessment of clinical efficacy and as a possible predictor of outcomes. The importance of clinical decongestion and improvement in biomarkers of myocardial injury and renal dysfunction on clinical outcomes remains uncertain. Important unanswered questions remain regarding the efficacy of relaxin, as compared with other intravenous vasodilators (nitroglycerin, nitroprusside, or nesiritide), and the use of relaxin as an adjunct or alternative to diuretics. Further studies addressing these questions will help better define a clinical role for relaxin.
AHF Vasodilator Trials: Different Populations and Different Endpoints
Key inclusion criteria
-AHF within the previous 16 h
-AHF within the previous 24 h
-AHF within the previous 24 h
-Dyspnea at rest or with minimal exertion
-Dyspnea at rest or with minimal activity
-Persistence of dyspnea at rest. (RR of at least 24/min)
-SBP > 125 mm Hg
One or more of the following signs:
At least 2 out of 4 criteria:
-Mild to moderate renal dysfunction , GFR 30–75 mL/min per 1.73 m2
-RR ≥20/min, or
-Elevated BNP or pro-BNP
-Pulmonary congestion on chest X-ray
-Pulmonary congestion/edema with rales one third of the way or more up the lung fields
-Pulmonary edema on physical exam
-BNP ≥350 pg/ml or NTproBNP ≥1,400 pg/mL
-Radiological pulmonary congestion or edema
One or more of the following objective measures of HF:
-LVEF <40 % or wall motion index ≤1.2. If patient had a PA catheter, CI = 2.5 L/min/m2 and PCWP ≥20 mm Hg
-Evidence of congestion/edema on CXR, or
-BNP ≥400 pg/mL or pro-BNP ≥1,000 pg/mL or
-PCWP >20 mm Hg, or
-LVEF <40 % in the previous 12 months
-IV diuretic use (at least 40 mg furosemide or equivalent)
-IV diuretic use (at least one dose) within 24 h
-IV diuretic use (at least one dose) within 24 h
Key exclusion criteria
-SBP <100 mm Hg or 100 mm Hg with the use of IV nitroglycerin
-SBP <100 mm Hg or <120 mm Hg with use of vasodilators
-Creatinine >2.5 mg/dL
Mean systolic blood pressure (mm Hg)
Median SBP: 123–124
Ejection fraction (%)
<40 % in 80 % of patients
Mean serum creatinine (mg/dL)
Mean eGFR 53–54 mL/min/ 1.73 m2
European: 66 %
European: 21 %
Multinational including North American (percentage not reported)
North American: 10 %
North American: 45 %
Sixty-six percent of the RELAX-AHF patients were from Europe, as compared with only 21 % in ASCEND-HF. Conversely, 45 % of the ASCEND-HF study population was from North America, as compared with 10 % in RELAX-HF. Such a geographical disparity could have an impact on trial results, as demonstrated in a meta-analysis of beta-blocker trials in chronic HFrEF. This study reported on significant geographic variation in response to beta-blocker therapy, with reduced survival benefit in the United States population, as compared with other regions. The authors suggested that this resulted from genetic variation in response to pharmacologic therapy .
Primary endpoints of the vasodilator trials included improvement in dyspnea/global symptoms after drug administration, utilizing VASs or Likert scales. Morbidity and mortality endpoints varied among the studies with regard to the chosen time points and reasons for hospital readmission. ASCEND-HF and VERITAS evaluated composite outcomes of death and readmission for HF at 30 days, which were 10.1 % and 33.2 %, respectively, in the placebo groups. There was a 60-day clinical composite outcome of cardiovascular death or readmission to hospital for HF or renal failure in RELAX-AHF. RELAX-AHF and VERITAS also reported on cardiovascular mortality at 180 days.
Conclusion: Disease Profiling and Clinical Trials in AHF
Our current understanding of AHF has evolved substantially. The premise that congestion from salt and water accumulation is the primary precipitant of AHF has been replaced by an understanding that complex processes occurring at the cellular and organ level contribute to the distinct clinical profiles observed in hospitalized patients with HF [30, 31].
The lack of success of clinical trials utilizing novel therapeutic strategies for AHF is generally attributed to factors such as the properties of the agents under investigation, study design, and trial execution. Most AHF trials to date have not considered the heterogeneity and etiology of distinct clinical profiles or associated comorbidities in trial design [17••, 32•].
One suggested approach to clinical trials in AHF has been the use of an “adaptive Bayesian design” for large studies involving multiple treatment approaches in a heterogeneous patient population. This approach provides flexibility for changing randomization schemes during the trial as data accumulate on patient subgroup responses to therapies [33•].
Another approach, utilized in RELAX-AHF, involves smaller trials targeting specific AHF patient profiles for evaluating new therapies and comparing the effectiveness of existing ones. Trial design should include specific geographic, demographic, clinical, and biomarker characteristics to define the study group and should incorporate the use of standardized clinical endpoints.
Compliance with Ethics Guidelines
Conflict of Interest
Jaime A. Hernandez-Montfort declares that he has no conflict of interest.
Sonali Arora declares that she has no conflict of interest.
Mara T. Slawsky has received compensation from St. Jude Medical for service as a consultant.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.