An oral relaxin agonist moves the haemodynamics, not yet the outcomes

Design

international, multicentre, double-blind, placebo-controlled, dose-ranging phase 2 trial (LUMINARA)

Population

375 adults with chronic heart failure across 57 sites in 10 countries: 235 with ejection fraction 35% or below, 140 with 41% to 55%; mean age 65 and 70 years

Primary outcome

change in end-systolic volume index at 24 weeks (reduced ejection fraction) and in systemic vascular resistance index (mildly reduced or preserved)

Effect

end-systolic volume index -5.4 mL/m2 at 20 mg (95% CI -10.9 to 0.1, P=0.054); significant reductions in systemic vascular resistance index at all doses (all P<0.05)

AZD5462 is an oral once-daily agonist at the relaxin family peptide receptor 1, a target that experimentally lowers systemic vascular resistance and afterload and encourages reverse remodelling. LUMINARA tested it in a double-blind, placebo-controlled, dose-ranging trial across 57 sites in 10 countries, in two cohorts: 235 patients with ejection fraction 35% or below, and 140 with 41% to 55%. Background guideline-directed therapy was described as excellent.

The primary endpoints were mechanistic, not clinical: change in end-systolic volume index at 24 weeks in the reduced ejection fraction cohort, and change in systemic vascular resistance index in the mildly reduced or preserved cohort. In the first, the 20 mg dose produced a placebo-adjusted change of -5.4 mL/m2 (95% CI -10.9 to 0.1, P=0.054) - a result that did not reach significance and did not show a dose gradient. In the second, all doses significantly reduced systemic vascular resistance index. Tolerability was comparable to placebo.

This is phase 2 and should be read as such. The vascular resistance finding is consistent with the drug doing what the receptor predicts; the ventricular volume finding is a near-miss on a surrogate at one of three doses, which is the kind of result that has often not replicated. Nothing here changes prescribing, and the authors say plainly that larger, longer trials with clinical outcomes are needed. What makes it worth a cardiologist's attention is that it is an oral agent in a class that has otherwise required infusion, in a population where the pipeline for mildly reduced and preserved ejection fraction remains thin.

• Treat this as early-phase evidence - both primary endpoints are surrogates, and one was not met.
• The reduced ejection fraction result, -5.4 mL/m2 with P=0.054 at one dose, showed no dose gradient and should not be quoted as positive.
• The systemic vascular resistance reduction across all doses is the more consistent signal.
• Tolerability was comparable to placebo on top of good background therapy, which is the practical prerequisite for a phase 3.
• Nothing here alters guideline-directed medical therapy, which is what these patients were already receiving.

The statistics, in plain English

A P value of 0.054 with a confidence interval running from -10.9 to 0.1 means the data are compatible with anything from a substantial reduction in end-systolic volume index to none at all. In a dose-ranging trial, the fact that the effect appeared at 20 mg and not at 80 or 360 mg matters: a real drug effect usually shows a gradient, and its absence raises the possibility that this is the play of chance across three comparisons. The vascular resistance result is stronger because it appeared across every dose group. Both endpoints are surrogates, and in heart failure the graveyard of drugs that improved haemodynamics without improving outcomes is well populated.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.