Volenrelaxin in chronic kidney disease: a Phase 2 withdrawn on the heart failure readout

What was tried

Eli Lilly and Company registered NCT06598631 in September 2024 as a Phase 2, randomised, double-blind, placebo-controlled study of volenrelaxin in adults with chronic kidney disease. Three subcutaneous dose arms were planned against placebo. The primary endpoint was percent change from baseline in urine albumin-to-creatinine ratio at week 12, with change in estimated glomerular filtration rate at week 12 as a secondary endpoint. Eligibility required a KDIGO diagnosis by eGFR and albuminuria category plus 90 days on a maximally tolerated ACE inhibitor or angiotensin receptor blocker.

No participant was randomised. The registry records actual enrollment of 0 and the status WITHDRAWN, with the posted reason: "Study was terminated due to a lack of foreseeable clinical benefit in the proposed chronic kidney disease population, following the termination of a related heart failure study that demonstrated no benefit in an overlapping patient group." The record was last updated on 8 September 2026.

Volenrelaxin, also LY3540378 and CHEMBL6068449 at ChEMBL maximum phase 2, is a long-acting form of human relaxin-2 acting at RXFP1, the relaxin family peptide receptor 1, encoded by ENSG00000171509.

The biological hypothesis

Relaxin-2 is the pregnancy hormone that reorganises maternal haemodynamics. Danielson and colleagues showed in 2001 that relaxin is required for the renal vasodilation of pregnancy in conscious rats, the origin of the idea that RXFP1 agonism could raise renal plasma flow and glomerular filtration on demand (doi:10.1172/JCI11975). A second strand treats relaxin as an antifibrotic acting through RXFP1 to suppress TGF-beta driven collagen deposition, running from early work on relaxin and renal fibrosis (doi:10.1046/j.1523-1755.2001.0590031184.x) through reviews of kidney disease progression (doi:10.1097/MNH.0b013e32831b7096) and antifibrotic action (doi:10.1111/bph.13529).

The case for a CKD trial was that RXFP1 agonism should raise eGFR, lower albuminuria, and slow fibrosis over longer exposure. The trial was built to read the first two.

Genetic anchoring was thin. The Open Targets overall association score for RXFP1 and chronic kidney disease is 0.0033, carried entirely by literature evidence with no genetic datatype contributing. The strongest RXFP1 associations there are cardiovascular disorder at 0.372 and heart failure at 0.370.

What actually happened

The adjacent study that decided this one was NCT05592275, a Phase 2 randomised double-blind trial of volenrelaxin at 25 mg, 50 mg and 100 mg weekly against placebo in worsening heart failure with preserved ejection fraction. It enrolled 332 participants between February 2023 and January 2025 and was stopped early after an interim review. Results are posted and published (Borlaug BA et al., Nature Medicine 2025).

On the primary endpoint of left atrial reservoir strain at week 26, the 25 mg arm beat placebo by 3.853 percentage strain points (95% CI 1.14 to 6.57, p equal to 0.0058), 50 mg gave 1.286 (95% CI -1.33 to 3.90, p equal to 0.3321) and 100 mg gave 0.881 (95% CI -1.83 to 3.59, p equal to 0.5211). There is no dose ordering in that sequence. Pooled across doses, volenrelaxin increased NT-proBNP by 24.5 percent (95% CI 2.0 to 51.8) and had no significant effect on eGFR, at 2.2 mL/min/1.73 m2 (95% CI -1.8 to 6.3). The authors report a non-significant increase in heart failure hospitalisation, hazard ratio 2.64 (95% CI 0.93 to 7.56, p equal to 0.070), a signal for more cardiovascular and renal serious adverse events, odds ratio 2.52 (95% CI 0.95 to 6.68, p equal to 0.056), and a conclusion of worsening congestion.

Failure mechanism, best guess

This is translational mismatch, not a portfolio squeeze. The kidney hypothesis depended on RXFP1 agonism producing a measurable renal haemodynamic effect in chronic disease. In a 332-patient trial with weekly dosing across a fourfold dose range, the eGFR confidence interval spanned zero. That is the most direct falsification available, because eGFR was also the withdrawn trial's own secondary endpoint.

Two features deepen the problem. Absence of dose ordering leaves the one nominally positive comparison, at the lowest dose, with no pharmacological gradient behind it. RXFP1 signalling is subject to receptor desensitisation and bell-shaped concentration response behaviour, so continuous long-acting agonism can underperform pulsatile exposure. The NT-proBNP rise points the same way, because genuine unloading should move that biomarker down.

The precedent was on the table. Serelaxin, recombinant human relaxin-2 given as a 48-hour infusion, failed its confirmatory acute heart failure trial (doi:10.1056/NEJMoa1801291). The reading available in 2022 was that serelaxin failed on exposure and setting rather than on target. The volenrelaxin result removes that defence for the sustained-exposure version of the idea.

How to prevent this next time

Endpoint-level data exist here, so the levers are quantitative.

Gate the indication on the adjacent trial's renal secondary endpoint before finalising the protocol. That gate cost nothing and was available. An eGFR interval of -1.8 to 6.3 mL/min/1.73 m2 excludes the effect size a CKD programme needs.

Pre-register dose ordering as a pass or fail criterion. A monotonicity requirement across the three doses would have marked the heart failure primary endpoint uninterpretable on the day it read out.

Pre-specify biomarker direction. A relaxin programme claiming haemodynamic unloading should require NT-proBNP to fall. It rose.

Weight the base rate. Open Targets gives RXFP1 and chronic kidney disease 0.0033 with no genetic support, and the prior entrant at this receptor had already failed a confirmatory trial. A literature-only target with that record should carry a higher bar for a new indication, not a lower one on the argument that the indication differs.

The single highest leverage change would have been treating the eGFR secondary endpoint of the heart failure trial as the pre-registered go or no-go gate for the kidney programme, because it measured the exact pharmacology the CKD hypothesis required and it read out flat.

What this means for similar programs

Tectonic Therapeutics is developing TX45, a relaxin fusion protein. Leerink Partners analysts told BioSpace that the Lilly discontinuation caused significant concern about its likelihood of success before reiterating confidence on differences in approach and indication, and Tectonic stock fell as much as 40 percent.

Programmes at RXFP1 now inherit three data points: a failed acute heart failure confirmatory trial with short infusion, a terminated chronic heart failure trial with weekly dosing and no dose ordering, and a kidney trial cancelled on read-across before dosing. A new entrant should state which of the three it expects to differ from, in terms of receptor occupancy over time rather than indication label.

Read-across used this way goes under-recorded, because a withdrawn study leaves no results. Withdrawing before enrolment spared a Phase 2 cohort an exposure that showed congestion signals in a related population.

Open questions

Is this a receptor failure or an exposure failure? Continuous agonism may desensitise RXFP1 in a way pulsatile dosing would not, and no clinical dataset separates the two.

Would albuminuria have moved without an eGFR effect? Antifibrotic and haemodynamic mechanisms can dissociate, and the withdrawn trial's primary endpoint was UACR.

Is there a genetically anchored subgroup? Open Targets records no genetic datatype linking RXFP1 to chronic kidney disease, so there is nothing to enrich on.

Does the congestion signal reflect fluid handling or vascular permeability? The published report does not resolve it, and the answer sets how much safety concern transfers outside heart failure.

Sources

1. • ClinicalTrials.gov, NCT06598631, volenrelaxin in chronic kidney disease, WITHDRAWN. https://clinicaltrials.gov/study/NCT06598631 - ClinicalTrials.gov, NCT05592275, LY3540378 in worsening chronic HFpEF, TERMINATED, posted results. https://clinicaltrials.gov/study/NCT05592275 - Borlaug BA, Testani JM, Petrie MC, et al. Effects of volenrelaxin in worsening heart failure with preserved ejection fraction: a phase 2 randomized trial. Nature Medicine 2025;31:3853-3861.- Teerlink JR, Voors AA, Ponikowski P, et al. Effects of Serelaxin in Patients with Acute Heart Failure. New England Journal of Medicine 2019.- Danielson LA, Conrad KP. Relaxin is essential for renal vasodilation during pregnancy in conscious rats. Journal of Clinical Investigation 2001.- Samuel CS, Royce SG, Hewitson TD, et al. Anti-fibrotic actions of relaxin. British Journal of Pharmacology 2017.- Garber SL, Mirochnik Y, Brecklin CS, et al. Relaxin and renal fibrosis. Kidney International 2001.- Yoshida T, Kumagai H, Suzuki A, et al. Relaxin and the progression of kidney disease. Current Opinion in Nephrology and Hypertension 2009.- Open Targets Platform, RXFP1 (ENSG00000171509) and chronic kidney disease (MONDO_0005300), overall association score 0.0033. https://platform.opentargets.org/target/ENSG00000171509 - ChEMBL, VOLENRELAXIN, CHEMBL6068449, maximum phase 2. https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL6068449/ - BioSpace, Lilly Cuts Mid-stage Relaxin Study in Chronic Kidney Disease for Lack of Foreseeable Clinical Benefit. https://www.biospace.com/drug-development/lilly-cuts-mid-stage-relaxin-study-in-chronic-kidney-disease-for-lack-of-foreseeable-clinical-benefit.