eloralintide

lilly's selective amylin agonist. 20% weight loss at 48 weeks in one 263-person phase 2, and 64.3% nausea in the fixed 6 mg arm. five phase 3 trials are enrolling and none reads out before 2028.

tier B · weight loss · 20% Billings 2025 phase 2

verdict

an amylin receptor agonist from eli lilly, code LY3841136. one published 48-week phase 2 (NCT06230523, n=263) produced up to 20% mean weight loss without an incretin anywhere in the molecule. every trial on the record is lilly-run, the phase 3 programme has not read out, and the tolerability picture is messier than the selectivity pitch suggests.

on whether eloralintide works — one trial says yes and says it loudly. the 48-week phase 2 (Billings et al., Lancet 2025, NCT06230523, 263 adults across 46 US centres) reported mean bodyweight change of -9% at 1 mg, -12% at 3 mg, -18% at 6 mg and -20% at 9 mg, against -0.4% on placebo, on the prespecified efficacy estimand. that is incretin-scale weight loss from a mechanism with no incretin in it. it is also one trial, funded and authored by the sponsor, with a 10-week follow-up and an interim database lock.

on whether selectivity means no nausea — the phase 1b supported that idea and the phase 2 complicated it. across 12 weeks with no escalation, nausea in the multiple-ascending-dose study topped out at 13.0%. across 48 weeks in phase 2, nausea hit 64.3% in the 6 mg arm and 54.2% in the 6/9 mg arm, with 25.0% vomiting at 6 mg. the rate is not monotonic in dose (33.3% at 9 mg, below the 6 mg arm), so the cleanest reading is that a slow four-week escalation moderates the gastrointestinal load, and the receptor selectivity by itself does not.

on where the material in circulation comes from — the molecule is investigational and holds no authorization at FDA or EMA. vendor listings do exist, generally quoting CAS 2883634-40-8 and a purity figure by HPLC. that CAS number, the molecular formula and the mass all trace to supplier catalogs. the ChEMBL entry for eloralintide (CHEMBL6068462) carries no deposited structure at all, no SMILES and no InChIKey, so there is no registry-grade reference to check a vendor claim against. no independent lot analysis for this compound was found anywhere in the literature.

based on published evidence and registered trial records. not medical advice.

why B-tier

B-tier because the evidence is drug-grade and the record is one trial deep. what earns the B is real: a 48-week randomised, double-blind, placebo-controlled phase 2 in 263 adults across 46 centres, a hard primary endpoint, a clean dose-response, published in the Lancet, sitting on top of a published phase 1 and a published phase 1b that agree with it. what holds it at B is equally real: no completed phase 3, no result posted from any of the four completed phase 1 studies, no dataset that Eli Lilly did not sponsor and did not author, and a 48-week adverse-event table that undercuts the selectivity argument the whole design rests on. when ENLIGHTEN-1 or ENLIGHTEN-2 reads out, and if the 20% holds at phase 3 scale with a tolerable escalation schedule, the tier moves. that cannot happen before February 2028.

the core tension

twenty percent mean weight loss at 48 weeks from a single-mechanism amylin agonist is a real result on a real trial. the story attached to it is that AMY1 selectivity buys that weight loss without the nausea that sank the amylin class, and the 48-week adverse-event table does not carry that story: 64.3% nausea at 6 mg, 25.0% vomiting at 6 mg, and a rate that falls rather than rises at 9 mg. what actually moved tolerability was a four-week escalation schedule, the same lever every incretin drug already uses. underneath both readings sits the same structural fact: one published efficacy trial, one sponsor, four completed phase 1 studies with nothing posted, and a phase 3 programme that cannot report before 2028.

what it is

A 37-amino-acid analog of human amylin, engineered by Eli Lilly and carried in the literature as LY3841136 (registry synonyms AMY-1176 and AMY1176). Amylin is the pancreatic hormone co-secreted with insulin that signals meal-related satiety through the area postrema in the brainstem. Native amylin lasts minutes and aggregates readily; eloralintide replaces the native disulfide with a methylene thioacetal bridge for chemical stability and carries a C20 fatty diacid on Lys26, which binds albumin and stretches the terminal half-life to roughly two weeks. Three non-coded residues sit at positions 11, 15 and 22. The design goal was potency at the AMY1 receptor relative to the calcitonin receptor. In Lilly's own translational work the human AMY1R EC50 was 23.9 pM against 291.0 pM at the human calcitonin receptor, the roughly 12-fold figure Lilly quotes on its slides. It holds no authorization at FDA or EMA.

what it does

Activates amylin receptors in the area postrema and reinforces meal-related satiety, a pathway separate from GLP-1 and GIP. In a phase 1 single-ascending-dose study (n=48 healthy adults) day-29 bodyweight fell 4.4% at 12 mg against a 0.6% gain on placebo. In a 12-week phase 1b multiple-ascending-dose study with no dose escalation (n=100 with overweight or obesity), placebo-adjusted weight reduction ran 2.6% at 1.2 mg, 8.9% at 3 mg, 8.5% at 6 mg and 11.3% at 12 mg. In the 48-week phase 2 (NCT06230523, n=263) mean bodyweight change reached -20% at 9 mg against -0.4% on placebo. A DXA substudy put 60% to 70% of the mass lost as fat, which the sponsor's own slide describes as on par with other obesity treatments.

origin

Built inside Eli Lilly's metabolic pipeline, the same house that produced tirzepatide and orforglipron. The discovery-to-phase-1 package was published in Molecular Metabolism in 2025 by Briere and colleagues, all Lilly employees and shareholders. First-in-human dosing began 30 March 2022 (NCT05295940). The 48-week phase 2 ran from 5 February 2024 to 14 August 2025 under Lilly's chronic-weight-management master protocol (NCT06143956), was presented at Obesity Week in Atlanta in November 2025, and published in the Lancet the following month. The phase 3 ENLIGHTEN programme opened with ENLIGHTEN-2 on 15 December 2025. A Lilly patent family covering this chemotype (US-2022288168-A1, granted as US-12551532-B2 in February 2026) names Briere among its inventors, though the patent text never names the compound.

why researchers are interested

It is the amylin analog with the largest published monotherapy number, and 20% at 48 weeks from a single non-incretin mechanism is the reason the class comparisons now start here rather than with cagrilintide. Lilly is also running eloralintide alongside tirzepatide (NCT06603571) and alongside its own macupatide (NCT07215559, NCT07589608), on the theory that a satiety mechanism which does not touch the incretin axis stacks with one that does. And the phase 3 programme spans obesity, type 2 diabetes, obstructive sleep apnea, knee osteoarthritis pain and persistent obesity on a weekly incretin.

does it work

One trial says yes, and it is a good trial. The phase 2 was randomised, double-blind, placebo-controlled, ran 48 weeks across 46 US centres, and hit a hard primary endpoint with a clean dose-response; the 9 mg confidence interval ran -22.7% to -17.5%. The supporting phase 1 and phase 1b work is published, dose-proportional and consistent. What the record does not yet contain is the part that decides a grade. There is no completed phase 3; the earliest primary completion in the ENLIGHTEN programme is February 2028. There is no independent dataset, because every trial on this molecule is sponsored by Eli Lilly and every primary paper is Lilly-authored. Four completed phase 1 studies (n=148, 128, 188 and 30) have posted no results at all. And the tolerability story that justifies the design is only partly supported: gastrointestinal events were low over 12 weeks without escalation, then reached 64.3% nausea in the 6 mg arm over 48 weeks. The weight-loss claim is well supported for what it is. The claim that selectivity solves amylin tolerability is not.

claims vs the data

• up to 20% mean weight loss at 48 weeks — supported — Billings et al., Lancet 2025 (NCT06230523, n=263, 46 US centres). Efficacy estimand: -20% at 9 mg (95% CI -22.7 to -17.5) and -20% at 6/9 mg, against -0.4% on placebo. One trial, sponsor-funded and sponsor-authored.
• works without touching the incretin pathway — supported — Eloralintide is an amylin receptor agonist with no GLP-1 or GIP activity. Human AMY1R EC50 23.9 pM against 291.0 pM at the calcitonin receptor and 253.8 pM at AMY3R (Briere et al. 2025). The weight loss in phase 2 was monotherapy.
• selectivity means the amylin nausea problem is solved — contradicted — Over 48 weeks nausea reached 64.3% in the fixed 6 mg arm and 54.2% at 6/9 mg, with 25.0% vomiting at 6 mg. Nausea was not monotonic in dose (33.3% at 9 mg). The four-week 3/6/9 mg escalation moderated the load to 25.0%, which is a titration effect rather than a receptor-selectivity effect.
• phase 3 data supports eloralintide — contradicted — Five phase 3 trials are registered and recruiting (NCT07321886, NCT07282600, NCT07369011, NCT07353931, NCT07392190). None has posted results and the earliest primary completion is 29 February 2028.
• the best-in-class amylin analog on the numbers — partially true — The 2026 network meta-analysis ranked high-dose eloralintide first on absolute weight loss (-19.44 kg), BMI and waist circumference, and second on percent weight change (-18.01%) behind high-dose subcutaneous amycretin (-23.95%). Eloralintide contributed exactly one trial to that analysis, so the ranking re-uses the same 263 people. The authors call the evidence sparse and low-certainty.
• well characterised chemistry — unverified — The CAS number, molecular formula and mass in circulation are supplier-catalog values. The ChEMBL record (CHEMBL6068462) deposits no SMILES and no InChIKey. The structural description that is well sourced is qualitative: 37 residues, three non-coded residues, a Cys2-Cys7 methylene thioacetal bridge, and a C20 diacid on Lys26.

key facts

• molecular formula: C₂₀₁H₃₁₉N₄₉O₆₅S₂ (supplier-listed, not peer-reviewed)
• molecular weight: ~4526 Da (supplier-listed; no structure deposited in ChEMBL)
• amino acids: 37
• half-life: 310 to 366 h (12.9 to 15.3 days) across 0.4 to 12 mg in phase 1
• type: selective long-acting amylin receptor (AMY1R) agonist
• CAS: 2883634-40-8 (supplier-listed)
• -20% 9mg arm at 48 weeks, phase 2 (n=263)
• 1 published efficacy trial, sponsor-run
• 5,615 phase 3 participants, zero read out
• 64.3% nausea in the fixed 6mg arm

frequently asked questions

What is eloralintide?

Eloralintide (Eli Lilly code LY3841136) is an investigational 37-amino-acid amylin receptor agonist for chronic weight management. It is engineered for once-weekly subcutaneous dosing, with a methylene thioacetal bridge in place of amylin's native disulfide and a C20 fatty diacid that binds albumin and stretches the half-life to roughly two weeks. It is in phase 3 trials as of July 2026 and is not approved anywhere.

What does eloralintide do?

It activates amylin receptors in the area postrema and reinforces meal-related satiety, a pathway separate from the GLP-1 and GIP drugs. In the published 48-week phase 2 (NCT06230523, n=263) mean bodyweight change reached -20% in the 9 mg arm against -0.4% on placebo. A DXA substudy in that trial attributed 60% to 70% of the mass lost to fat.

How is eloralintide dosed in trials?

Once-weekly subcutaneous dosing. The phase 2 tested 1, 3, 6 and 9 mg as fixed doses, plus two escalation schedules (6/9 mg and 3/6/9 mg). The four-week 3/6/9 mg escalation reached 16% mean weight loss with 25.0% nausea, against 64.3% nausea in the fixed 6 mg arm. Phase 3 dose selection has not been disclosed in the registry records.

What are the side effects of eloralintide?

In the 48-week phase 2, the most common events in the pooled eloralintide arms were nausea (32.7%), fatigue (26.9%), constipation (14.9%), diarrhoea (14.9%) and vomiting (8.2%), against placebo rates of 13.5%, 11.5%, 5.8%, 9.6% and 0%. Gastrointestinal events were mostly mild to moderate, with no severe nausea, diarrhoea or vomiting reported. Alopecia appeared in 6.7% of pooled eloralintide participants and in no placebo participant. Pulse rate and both blood pressures fell across dose arms. Adverse events led to discontinuation in 10.1% of pooled eloralintide participants against 7.7% on placebo. There were no deaths.

Is eloralintide FDA approved?

No. There is no FDA application number, no EMA marketing authorization, and no orphan, fast-track or breakthrough designation on record. The phase 3 ENLIGHTEN programme began enrolling in December 2025 with the earliest primary completion in February 2028, so no approval decision can precede that.

How is eloralintide different from cagrilintide?

Both are long-acting amylin analogs; cagrilintide is Novo Nordisk's. Eloralintide was designed for potency at AMY1 relative to the calcitonin receptor, roughly 12-fold in Lilly's cell assays. In Lilly's head-to-head rat work eloralintide spared significantly more lean mass (p=0.0101) and needed more drug to produce conditioned taste avoidance (ED50 8.9 against 4.2 nmol/kg), at slightly less total weight loss (11.1% against 13.3%). No head-to-head human trial of the two exists, so every cross-molecule comparison in humans is indirect.

related peptides

• cagrilintide — novo's long-acting amylin analog, and the comparator in lilly's own rat head-to-head
• cagrisema — novo's amylin plus GLP-1 combination, third on percent weight change behind eloralintide in the 2026 network meta-analysis
• tirzepatide — lilly's approved benchmark, and eloralintide's combination partner in phase 1 and 2

reptides grades the research record and cites the literature behind every call. research reference only; not medical advice.