Serelaxin lowered organ-injury markers but not deaths in RELAX-AHF-2

What is serelaxin, and how was it supposed to help heart failure?

Serelaxin is recombinant human relaxin-2, the hormone that rises during pregnancy and drives maternal cardiovascular and renal adaptation through the RXFP1 receptor.

The RELAX-AHF-2 design paper describes the rationale as tissue-specific vasodilation plus end-organ protection. The idea is that a heart failure patient undergoing rapid decongestion needs both symptom relief and protection of the kidneys, liver, and heart from hypoperfusion injury during that process ^[6].

That is coherent, receptor-level pharmacology, not speculation. What was never settled is whether activating the pathway for 48 hours during a hospital stay changes whether a patient is alive six months later.

Does serelaxin reduce mortality in acute heart failure?

Serelaxin does not reduce mortality in acute heart failure: in the 6,545-patient RELAX-AHF-2 trial, 180-day cardiovascular death was 8.7% on serelaxin versus 8.9% on placebo ^[2].

RELAX-AHF-2 was explicitly built to confirm the RELAX-AHF mortality and worsening-heart-failure signals, per its own design paper ^[6]. It enrolled roughly six times as many patients as the original trial, with two co-primary endpoints: cardiovascular death at 180 days and worsening heart failure at day 5. Both were null. The cardiovascular death result had a hazard ratio of 0.98 (95% CI 0.83-1.15, p=0.77) ^[2].

A dedicated cause-of-death analysis went further. Unlike RELAX-AHF, the RELAX-AHF-2 analysis found no apparent effect of serelaxin on any category of death, cardiovascular or otherwise, among the 11.5% of patients who died by 180 days ^[4].

A meta-analysis pooling eight randomized trials and 8,477 patients reached the same conclusion across mortality timepoints, describing serelaxin as unrelated to mortality ^[1]. An age subgroup analysis of RELAX-AHF-2 found no meaningful shift in treatment effect across age bands from under 65 to over 85, so the null result was not an artifact of an unusually old or unusually sick trial population ^[3].

RELAX-AHF-2 is about as clean a confirmatory failure as clinical trial data gets: a large, well-powered, event-driven trial designed specifically to retest a secondary finding, and the finding did not survive the retest. Development of serelaxin for this indication stopped after RELAX-AHF-2 was null on both co-primary endpoints ^[2]⁠^[6].

Where did serelaxin's mortality-benefit claim come from?

The serelaxin mortality claim comes from RELAX-AHF, a 1,161-patient trial in which a 37% reduction in 180-day mortality was a secondary finding, not the primary endpoint ^[6].

RELAX-AHF met its primary endpoint, dyspnea improvement through day 5 ^[6]. Reported alongside that result were a 47% reduction in worsening heart failure and, most consequentially, a 37% reduction in both all-cause and cardiovascular mortality through day 180 ^[6].

That mortality number migrated into pipeline databases, cardiology review paragraphs, and eventually longevity-supplement marketing copy. It was never the trial's primary endpoint. For longevity claims in general, see the lifespan evidence behind popular longevity peptides.

Why did the RELAX-AHF mortality signal not hold up?

The RELAX-AHF mortality signal was uneven: a post-hoc analysis found a hazard ratio of 0.42 in patients with atrial fibrillation versus 0.90 in patients without it ^[7].

The interaction p-value was 0.0643, a pattern consistent with a subgroup carrying the overall effect rather than an effect distributed evenly across the trial population ^[7]. A secondary finding that concentrates in one subgroup within a modestly sized trial is exactly the kind of result statisticians expect to shrink, or vanish, when retested in a much larger, adequately powered sample. RELAX-AHF-2 was that retest.

For another case of earlier secondary-endpoint signals meeting a confirmatory trial, see davunetide's PSP trial.

Does serelaxin relieve shortness of breath in heart failure?

Serelaxin met its dyspnea endpoint in RELAX-AHF ^[6], but pooled data across the serelaxin program show no consistent dyspnea improvement ^[1].

RELAX-AHF-2 was powered around mortality and worsening heart failure, not dyspnea. The meta-analysis of eight randomized trials described serelaxin as without effect on dyspnea relief ^[1].

Did serelaxin protect the heart and kidneys in RELAX-AHF-2?

Serelaxin reduced heart and kidney injury markers in a 1,020-patient RELAX-AHF-2 biomarker substudy, but it still did not reduce 180-day mortality in those patients ^[9].

Compared with placebo, serelaxin reduced the percentage rise in troponin T through day 14, reduced rises in creatinine, cystatin C, and uric acid at day 2, and produced a larger early drop in NT-proBNP ^[9]. These are the biomarker changes the RXFP1 mechanism predicts: less cardiac and renal strain during aggressive decongestion.

Biomarker movement without a matching hard-outcome benefit is the most accurate read on serelaxin. The drug moved measurable organ-injury markers, but whatever it does was not enough, at this dose and in this population, to change survival. The pattern points to a real but clinically insufficient signal, not an inert pathway ^[9].

What did the RELAX-AHF-EU trial show?

RELAX-AHF-EU, an open-label trial, found a lower rate of worsening heart failure or death through day 5 with serelaxin, 5.0% versus 6.9% ^[8].

The hazard ratio was 0.71, with a number needed to treat of 53. The trial was stopped early once RELAX-AHF-2's neutral topline came in, and the effect did not hold for the day-14 composite ^[8].

Given the open-label design and early termination, RELAX-AHF-EU fits the biomarker substudy's picture of a modest, short-window physiological effect. It is not an independent confirmation of a mortality benefit. For another peptide judged by a cardiovascular outcomes trial, see liraglutide's cardiovascular outcomes trial.

What is still unknown about serelaxin?

Serelaxin's open questions concern dose, patient subgroups, and uses outside acute heart failure:

• Dose, infusion length and patient selection. Whether a different dose, a longer infusion, or a narrower population, such as the RELAX-AHF atrial fibrillation subgroup ^[7], would show a real mortality effect has not been tested. With development for this indication halted, it is unlikely to be.
• Comorbidity clusters. Exploratory work continues to mine the RELAX-AHF-2 dataset. A comorbidity-cluster analysis identified five distinct multimorbidity subgroups among the 6,545 patients, with treatment allocation appearing to modify outcomes differently across clusters. That is hypothesis-generating, not confirmatory, evidence ^[5].
• Relaxin beyond heart failure. None of the trials cited below tests relaxin biology outside acute heart failure, so they can neither support nor refute claims about anti-fibrotic effects in other organ systems or in healthy aging.