Eloralintide vs Cagrilintide: Selective and Dual Amylin Agonism Compared

Amylin is the hormone your beta cells release alongside insulin at every meal to signal that the meal is over. Both of these compounds are engineered long-acting versions of it, and both are dosed once weekly by subcutaneous injection. That is where the similarity ends.

Cagrilintide is a Novo Nordisk molecule and a dual amylin and calcitonin receptor agonist. It is furthest along in development as the amylin half of CagriSema, where it is paired with semaglutide.

Eloralintide (LY3841136) is Eli Lilly's answer, and it is selective. In the discovery pharmacology it preferentially activated the human AMY1 receptor by 12-fold over the plain calcitonin receptor and 11-fold over AMY3R (Briere et al., 2025).

The selectivity is the entire thesis. Amylin receptors are not distinct gene products: they are the calcitonin receptor complexed with RAMP1, RAMP2 or RAMP3 (Christopoulos et al., 1999). A compound that also activates the bare calcitonin receptor is doing something extra, and the prevailing explanation for cagrilintide's heavier nausea burden is that extra calcitonin activity. Whether removing it produces a better drug in humans is what the current trials are meant to answer.

Neither is approved anywhere. Cagrilintide is in Phase 3 as part of CagriSema. Eloralintide has completed Phase 2 monotherapy and is in Phase 2 combination trials.

Receptor profile. Cagrilintide activates both amylin and calcitonin receptors. Eloralintide is selective for the amylin receptors, with measured 12-fold and 11-fold preferences. In rats, eloralintide produced significantly less conditioned taste avoidance than cagrilintide, which is the animal proxy for nausea (Briere et al., 2025).

Development strategy. Cagrilintide has always been developed as a partner for semaglutide. Eloralintide has a standalone Phase 2 monotherapy result and is separately being trialled with tirzepatide. That difference in strategy is why direct efficacy comparison is so awkward: one has been tested mostly in combination, the other mostly alone.

Published monotherapy efficacy. In REDEFINE 1, cagrilintide alone produced 11.8 percent weight loss at 68 weeks (Garvey et al., 2025). Eloralintide alone produced 9 to 20 percent across dose arms at 48 weeks, reaching 20 percent at 9mg (Billings et al., 2025). These are different trials with different populations and durations, so the comparison is directional at best, but eloralintide's top arms are clearly in a higher band than cagrilintide monotherapy.

Tolerability. This is where the selectivity claim gets complicated. Eloralintide's 6mg fixed-dose arm produced 64 percent nausea against 14 percent on placebo. Its escalating 3 to 9mg arm produced 25 percent. So the advantage is real but conditional on titration rather than intrinsic to selective agonism.

Body composition. Neither has a peer-reviewed human body composition result, and this is the single most important thing to know about the pair. Eloralintide has no such data at all. Cagrilintide has one dataset, a DXA substudy of 252 REDEFINE 1 participants presented at ObesityWeek 2025 and not yet published in a journal, and it does not support the muscle-sparing story: cagrilintide alone split its weight loss 62.9 percent fat and 37.1 percent lean, against 69.7 and 30.3 for semaglutide alone. The amylin arm lost proportionally more lean tissue than the GLP-1 arm.

That fits the broader pattern. A network meta-analysis of 262 trials covering 99,791 participants found the agent producing the most fat loss also produced the most lean loss (tirzepatide, 25.7 percent fat and 8.3 percent lean), which is what you would expect if lean loss follows the size of the deficit rather than the receptor engaged. The widely repeated claim that amylin analogs spare lean mass rests on rat studies.

Neither is legitimately available, so this is a question about which programme to follow rather than which to buy.

Cagrilintide is the one with a Phase 3 readout coming, as part of CagriSema, and it is the closest to something a doctor could eventually prescribe. It also has the only human body composition data in the amylin class, and that data is not flattering.

Eloralintide has the better monotherapy numbers and the better tolerability story, with the caveat that the tolerability advantage depends on slow escalation and evaporates at a fixed 6mg dose.

For anyone tracking this class for a future cut, the more useful question is not which amylin analog wins but whether the class adds anything over a GLP-1 you can actually obtain. REDEFINE 1 suggests the amylin component added roughly 4 to 7 percentage points on top of a maximal semaglutide dose, depending on which estimand you read. That is a real increment and a long way from transformative.