Relaxin receptor agonist did not meet primary endpoints in phase 3 study in acute heart failure
Paris, France | RELAX-AHF-2: Serelaxin in acute heart failure
Presented at ESC Heart Failure 2017 by: John R TEERLINK (San Francisco, United States of America) and Marco METRA (Brescia, Italy)
Relaxin is a naturally-occurring insulin-like peptide, found in both men and women. During pregnancy, it is found at elevated levels, as it contributes to the maternal haemodynamic adaptations to pregnancy. Serelaxin, a relaxin receptor agonist, is a recombinant form of human relaxin-2 hormone. Serelaxin causes vasodilation, thereby reducing myocardial overload and improving renal function. It also improves cell preservation by reducing inflammation and oxidative stress. In addition, it affects increases remodelling of the extracellular matrix through reducing fibrosis and collagen synthesis and enhancing collagen breakdown. This remodelling diminishes vessel stiffness.
In the RELAX acute heart failure (AHF) study, serelaxin has been shown to provide dyspnoea relief as compared to placebo (19% increase VAS AUC composite with serelaxin from baseline through day 5, mean difference: 448 mm-hr), in patients with dyspnoea and congestion on CXR, with elevated BNP/NT-proBNP, SBP >125 mmHg, eGFR 30-75 mL/min 1.73m2 and ≥40mg i.v. furosemide. Moreover, a significant reduction in the risk of in-hospital worsening HF (WHF) was seen by 47% through day 5 (HR: 0.53, 95%CI: 0.36-0.79, P=0.0016). Serelaxin was associated with significant improvements in the percentage of patients that showed increased hs-cTnT, cystatin-C or a decrease in NT-proBNP. Furthermore, a significant reduction in CV death (HR: 0.64, P=0.028) and all-cause mortality (HR: 0.83, P=0.02) through day 180 was seen.
Based on these results, the hypothesis was formulated that serelaxin may reduce end organ damage and improve symptoms in AHF resulting in decreased WHF and CV death. This was tested in the multicentre, randomised, double-blind, placebo-controlled phase 3 RELAX-AHF-2 study, conducted in over 500 sites in 34 countries, enrolling 6600 patients. Primary endpoints were the superiority of a 48-infusion of serelaxin 30µg/kg/day in reducing time to confirmed CV death in patients with AHF during a follow-up of 180 days, and time to first occurrence of WHF through day 5. Secondary endpoints included time to all-cause death, length of hospital stay during the index hospitalisation and time to composite endpoint of CV death or rehospitalisation due to HF/renal failure. Standard HF therapy continued throughout the trial. The protocol and inclusion criteria were largely similar to those of RELAX-AHF.
- CV mortality through day 180 was similar for patients receiving serelaxin (285 in 3274, 8.7%) or placebo (290 in 3271, 8.9%, HR: 0.98, 95%CI: 0.83-1.15, P-value: 0.3857).
- WHF through day 5 was nominally lower with serelaxin but this did not reach significance (serelaxin: 6.9%, placebo: 7.7%, HR: 0.89, 95%CI: 0.75-1.07, P=0.0968).
- All-cause mortality through day 180 did not differ significantly between treatment groups (11.2% with serelaxin vs. 11.9% with placebo, HR: 0.94, 95%CI: 0.81-1.08, P=0.3890).
- The Kaplan-Meier curves of CV death or rehospitalisation due to heart/renal failure through day 180 for both treatment largely overlapped (24.3% vs. 24.9%, HR: 0.97, 95%CI: 0.88-1.07, P=0.2744).
- Length of hospital stay did not significantly differ between treatments (median duration on serelaxin: 6.824 days, vs placebo: 6.857, mean difference: -0.183, 95%CI: -0.645 to 0.280).
- Safety analyses confirmed the safe use of serelaxin.
All primary and key secondary efficacy results of the RELAX-AHF-2 study were neutral, and no new safety concerns were observed. These results do not confirm a benefit of serelaxin for the treatment of AHF patients.
Dr. Teerlink more specifically discussed the results. He showed that no pre-specified subgroups were identified that preferentially benefit from treatment with serelaxin, with regard to adjudicated CV death, nor to WHF through day 5.
In search of explanations for the differences seen in the results of RELAX-AHF-1 and -2, he compared the observed systolic blood pressure (SBP) during study drug infusion in both treatment groups. In both studies, the SBPs for both treatment arms overlapped well. Thus, it was concluded that the pharmacological effect was the same in both studies, in terms of the vasodilatory effect.
When comparing key selected variables, only NT-proBNP differed between the populations of both studies, which was a difference by design. Patients in RELAX-AHF-2 had higher values, and were sicker.
Comparing outcome rates in the placebo groups of both studies, revealed that fewer patients on placebo in RELAX-AHF-2 showed WHF through day 5 as compared to the earlier, smaller RELAX-AHF-1 study (7.7% vs. 12.0%). All-cause death was a little higher in RELAX-AHF-2 (11.9% vs. 11.3% in AHF-1), with CV death accounting for 9.0% and 9.6% respectively in study 2 and 1, and non-CV death for 3.0% and 1.7%. During the discussion, dr. Alexandre Mebazaa (Paris, France) noted that clearly the placebo group is doing better every year. This is, however, not seen in daily practice. It was suggested that studies should select patients in worse conditions, to approach daily clinic more, and to find more benefit of the drug, as biological and clinical data do suggest that it works.
A biomarker substudy in about 1000 patients indeed confirm effects of serelaxin on cardiac and renal biomarkers (lower hs-TnT, lower NT-proBNP and lower Cystatin C). A posthoc analysis in this substudy showed a nominal improvement in WHF through day 5 (4.7% vs. 7.7%, HR: 0.60, 95%CI: 0.36-0.98, P=0.0286).
Dr. Frank Ruschitzka (Zurich, Switzerland) concluded that the field has been misled by the analogy to ACS. He stated that we do not know the enemy, and do not have a treatment. Treatment with serelaxin may remain a good rationale, as there are ‘glimpses of hope in the clinical data’. But heart failure is not one disease, but rather hundreds. So he thinks that we need to go more selective and identify patients better who might benefit from a specific treatment. This view was also expressed by others in this discussion. Dr. John Cleland (London, Great Britain) and Milton Packer (Dallas, USA) agreed that novel studies should take a different approach
In addition, the option to treat earlier was considered, possibly by identifying patients while still at home, rather than only starting treatment at presentation. Decompensation often starts in the weeks, not hours, before admission. Repeated admission might be worth exploring as well.
Our coverage of ESC HF 2017 is based on the information provided during the congress.