This study was prospectively designed to evaluate the effects of a long-acting form of human relaxin on cardiac function, circulatory congestion and kidney function in patients with recent worsening HFpEF, a group at very high risk for recurrent events. The trial was terminated before planned enrollment was completed, limiting power and, therefore, certainty about the conclusions, but, as compared to placebo, low-dose volenrelaxin improved the primary endpoint of LA reservoir strain, with no significant effect at higher dosages. Kidney function was improved with volenrelaxin with shorter duration of treatment, with no significant effect observed at 26 weeks. Despite these favorable signals, there was evidence of worsening congestion across multiple independent assessments, including higher NT-proBNP levels, estimated LA pressures (E/e′ ratio), cardiac output and estimated plasma volume and lower hematocrit, hemoglobin and serum sodium. This phase 2 trial was not powered to detect an effect of volenrelaxin on HF hospitalization or cardiovascular death, but there were signals for greater risk with volenrelaxin, particularly for HF hospitalization, which did not reach statistical significance. A large and rapid improvement in health status and congestion was observed in the placebo group, as expected in recently hospitalized patients who have been treated and are receiving frequent follow-up care, but this effect was muted in the volenrelaxin group, evidenced by lower likelihood of NYHA class improvement at 26 weeks, trends for a greater need for loop diuretic intensification and suggestive evidence of worsening symptom severity based upon changes in KCCQ scores. The evidence for worsening rather than improved congestion with volenrelaxin does not support further evaluation of this molecule as a treatment for patients with worsening HFpEF.
Relaxin is a peptide hormone synthesized by the placenta during pregnancy that has been shown in preclinical and clinical studies to have vasodilatory, antifibrotic and antiinflammatory effects, along with increases in renal plasma flow and potential improvements in eGFR, changes that help support the heightened circulatory and metabolic demands of pregnancy6,7,8. These effects could also be advantageous in patients with HF, forming the basis for clinical trials of short-term relaxin treatment in the setting of acute HF. In the RELAX-AHF trial, 48-hour treatment with intravenous relaxin reduced LA pressure and dyspnea in patients with acute HF but had no effect on hospital readmission10,12. In the larger RELAX-AHF-2 trial, the same 48-hour treatment with intravenous relaxin did not reduce the incidence of cardiovascular death at 180 days or worsening HF at 5 days compared to placebo13but, in a large meta-analysis of all six phase 3 trials (n = 11,359), 48-hour infusion with relaxin was shown to reduce the risk of worsening HF to day 5 while also reducing serum creatinine, NT-proBNP and troponin levels14.
These favorable signals with short-term administration suggest that longer-duration treatment with relaxin could produce salutary effects in patients with HF, but previously tested recombinant human relaxin formulations had very short half-life, requiring continuous infusion, making such trials impractical14. Volenrelaxin is a long-acting human relaxin analogue and potent agonist of the RXFP1 receptor that consists of human relaxin fused to the serum albumin-binding VHH domain, allowing for longer half-life and once-weekly dosing without substantial modification of the parent relaxin molecule15. Patients with HFpEF, particularly HFpEF with recent worsening congestion, display severe LA myopathy, manifested by atrial fibrillation, LA dilatation and LA dysfunction, along with greater congestion and kidney dysfunction3,4,5. This led us to hypothesize that volenrelaxin would improve LA function and, given the salutary effects with short-term relaxin9,10,11,12,13,14also reduce congestion and improve kidney function in patients with worsening HFpEF.
Contrary to this hypothesis, we found that congestion worsened with 26 weeks of volenrelaxin treatment. This was evidenced by increases in NT-proBNP, E/e′ ratio and estimated plasma volume (Fig. 2). Hematocrit and hemoglobin decreased, likely reflecting hemodilution from volume expansion. Red blood cell mass was not directly measured to verify this mechanism, but the observed reduction in serum sodium also supports expansion of plasma water. Cardiac output increased, likely related, at least in part, to the Frank–Starling mechanism accompanying volume expansion that accompanied vasodilation, and LA volumes also tended to increase, in keeping with greater congestion. GGT levels are known to increase in the setting of HF decompensation due to liver congestion16and these were also increased in the volenrelaxin group as compared to placebo. Although isolated changes in one or two of these markers of congestion could be explained by the play of chance, the totality and consistency of evidence across multiple independent measures strongly supports the interpretation of an increase in congestion with volenrelaxin.
Biochemical and echocardiographic evidence of volume expansion with volenrelaxin was corroborated by suggestions of worsening clinical status, including lower likelihood of NYHA class improvement at 26 weeks with volenrelaxin (28% versus 49%), even as loop diuretic intensification tended to be greater compared to placebo, and signals of worsening KCCQ scores. No statistically significant differences were observed in the risk for HF hospitalization, but the directionality of relationships favored placebo (HR = 2.64, 95% CI: 0.93–7.56, P = 0.070). The numbers of cardiac and renal serious adverse events and treatment-emergent adverse events were also greater in the volenrelaxin group.
The worsening congestion observed with volenrelaxin may appear surprising considering earlier short-term studies with relaxin showing vasodilation, reduction in LA pressure and decreases rather than increases in NT-proBNP9,10,11,12,13,14,17. However, the present results with longer-term RXFP1 activation complement and are compatible with the acute studies, suggesting that longer-duration treatment may be necessary for congestion to manifest. Indeed, we observed no evidence of effect on NT-proBNP at the 3-week assessment, but, with longer-duration administration, an increase in NT-proBNP became evident (Fig. 1). Similar trends were observed for E/e′. Relaxin is a known vasodilator, which was confirmed by the reductions in total peripheral resistance at 12 weeks and 26 weeks. Acute vasodilation can improve congestion through redistribution of blood volume, but, if this vasodilation triggers a compensatory reduction in sodium excretion, congestion will worsen18. This increase in sodium avidity may precipitate sodium retention and plasma volume expansion, as shown in older studies with direct vasodilators19,20. Indeed, excessive vasodilation is also known to be a key cause of high-output HF21. Such effects may be even greater with relaxin (versus other vasodilators) given the teleologic need for volume expansion in the gravid state, although the specific role of relaxin to promote volume expansion compared to other pregnancy-associated hormones remains unclear6,7,8.
LA reservoir strain was improved with low-dose volenrelaxin but not with higher dosages or pooling dosages. RXFP1 receptors are expressed in atrial tissue, where their activation enhances cAMP–PKA signaling pathways to improve contractile and potentially lusitropic performance6,7,8. This could explain the trend for improved LA function, although it is also possible that Frank–Starling effects from volume loading contributed, which would be consistent with the fact that NT-proBNP levels also increased to the greatest extent in patients randomized to the 25-mg dose of volenrelaxin. Regardless, the deleterious consequences of worsening congestion overrode any favorable effect on LA function that might have been possible.
We observed that kidney function was improved with shorter durations of treatment with volenrelaxin, an effect that was no longer significant by 26 weeks. This is consistent with preclinical studies showing increases in renal plasma flow and GFR6,7,8 as well as human clinical trial evidence with short-term relaxin treatment, where serum creatinine was reduced with 48-hour infusion14. We observed that the early favorable effect on eGFR was muted with longer duration of treatment, and there were numerically more renal adverse events. This is consistent with an earlier study of acute intravenous relaxin where potential HF benefits were not found to be related to improvements in renal function22. Kidney function is known to worsen with increasing congestion in HF23and it may be that progressive congestion with longer treatment duration might have offset ostensible direct renal vasodilatory effects by the conclusion of the trial. Of note, beneficial effects on eGFR were not coupled with favorable effects on glomerular function, as albuminuria tended to worsen with longer duration of volenrelaxin. This may be related to worsening congestion, as previous studies showed that acute stressors, such as acute myocardial infarction, HF decompensation and even exercise or febrile illness, can be powerful drivers of proteinuria24,25,26,27,28.
The major limitation relates to the fact that this trial was stopped early before the target enrollment was achieved, decreasing power. Multiple dosages of volenrelaxin were evaluated, introducing heterogeneity, although there were not clear dose–response relationships or significant treatment interaction by dosage. Plasma volume was estimated using a validated formula, rather than direct measurement, and circulatory congestion was not directly assessed using invasive hemodynamics, which would not be feasible in a trial of this size. However, multiple studies have validated the biomarkers (NT-proBNP) and echocardiographic measures (E/e′) employed here to estimate circulatory congestion29,30,31. This study was not powered to evaluate effects on clinical HF events, so we cannot reach any conclusions regarding potential effects on hard outcomes. We evaluated an advanced HFpEF cohort with severe congestion, cardiac dysfunction and sodium avidity, as we hypothesized that potentially favorable effects of volenrelaxin might be most apparent in this cohort. Ongoing trials testing other agents in the class will provide more insight into effects in patients with less advanced HF32. Despite these limitations, this trial has several notable strengths, including the novelty of the intervention, which has shown promise in shorter trials but never been evaluated for this sustained duration, and the robust, multimodal assessment of cardiorenal function and hemodynamics, using complementary blood biomarkers and echocardiography, with multiple serial repeated measurements to provide greater physiologic insight.
Despite some evidence for improvement in LA function at low dose, treatment with the long-acting human relaxin agonist volenrelaxin was associated with worsening congestion across multiple independent imaging, laboratory and clinical domains in patients with recently decompensated HFpEF. These data do not support volenrelaxin as a useful therapy in this setting.
