EloraTZP vs Retatrutide: What the Three Datasets Show
A Lilly slide shows 29% weight loss in 31 weeks with eloralintide plus tirzepatide. Only 3 of 12 had data, and it can't be compared with retatrutide's 28.3%.
A Lilly EASD slide showing 29.0% weight loss with eloralintide plus tirzepatide at 32 weeks has been compared with retatrutide's 28.3%, but only 3 of the 12 participants in that arm had data at week 32, and the retatrutide figure is an 80-week efficacy-estimand result (25.0% on the primary analysis). The larger phase 2b trial reported 23.3% at 48 weeks in type 2 diabetes versus 14.8% for tirzepatide alone, with up to 27.0% discontinuing for adverse events. Both are investigational and unapproved.
A slide from Lilly’s EASD presentation has been circulating with a striking number: eloralintide plus tirzepatide, 29.0% average weight loss in 32 weeks. Social posts set it against retatrutide’s 28.3% in TRIUMPH-1 and asked whether Lilly had just outrun its own triple agonist. The number is real. The comparison is not, and a larger trial published the day before this article adds the context the slide leaves out. Everything below is investigational and unapproved, reported as scientific context only.
What does the slide actually show?
Eloralintide (LY3841136) is a once-weekly, selective amylin receptor agonist from Eli Lilly; in a 12-week phase 1 study it produced mean weight reductions of 2.6% to 11.3% across dose groups,4 and as a single agent roughly 9% to 20% at 48 weeks in a 263-person phase 2 trial against 0.4% on placebo; phase 3 trials are under way.31 The slide comes from a small phase 1b study that added it to tirzepatide, a GIP and GLP-1 receptor agonist. It summarises two phase 1b multiple-ascending-dose studies in adults without diabetes (Study A, 96 participants, mean BMI 32.8; Study B, 64 participants, 75% men, mean BMI 30.5), registered as NCT06345066 and NCT06297616. The figures are modelled least-squares means, not raw averages of everyone enrolled, and the abstract calls the headline numbers “maximum mean” reductions.9
| Study | Weeks | Arm | Randomised (N) | With valid data (n) | Weight change | Difference vs comparator (90% CI) |
|---|---|---|---|---|---|---|
| A | 16 | PBO + tirzepatide 5 | 12 | 10 | −10.0 | — |
| A | 16 | Eloralintide 3 + tirzepatide 5 | 12 | 10 | −17.0 | −7.0 (−8.9 to −5.0) |
| A | 16 | Eloralintide 6 + tirzepatide 5 | 12 | 8 | −18.2 | −8.2 (−10.2 to −6.2) |
| A | 16 | Eloralintide 9 + tirzepatide 5 | 12 | 6 | −20.5 | −10.5 (−12.6 to −8.4) |
| A | 32 | PBO + tirzepatide 15 | 8 | 5 | −17.8 | — |
| A | 32 | Eloralintide 9 (start 1.5) + tirzepatide 15 | 12 | 7 | −23.6 | −5.8 (−8.9 to −2.7) |
| A | 32 | Eloralintide 9 (start 3) + tirzepatide 15 | 12 | 3 | −29.0 | −11.2 (−14.7 to −7.7) |
| B | 24 | PBO + PBO | 16 | 16 | −0.5 | — |
| B | 24 | PBO + tirzepatide 15 | 16 | 14 | −16.4 | −16.0 (−18.3 to −13.7) |
| B | 24 | Eloralintide 9 (start 3) + PBO | 16 | 16 | −14.2 | −13.7 (−16.0 to −11.5) |
| B | 24 | Eloralintide 9 (start 3) + tirzepatide 15 | 16 | 12 | −25.5 | −25.0 (−27.3 to −22.7) |
Mean percent change in body weight from baseline to the end of each treatment period (least-squares means). Study A comparators are PBO + tirzepatide; Study B comparators are double placebo. Source: EASD 2026 abstract 76, Table.9 Doses are trial-arm labels, reported as trial facts and not as guidance of any kind.
Participants with valid data at week 32 in the arm that reached 29.0% (the slide plots to week 31). Neither the slide nor the abstract says why the other nine did not.9
The headline arm is also the arm with the fewest people left. In the 16-week cohort the count with data fell from 10 of 12 to 6 of 12 as the eloralintide dose rose. The slide does not give reasons, and it would be wrong to assume them, but a mean built on three people is a signal, not an estimate. The comparison that matters inside Study A is against tirzepatide alone: 29.0% versus 17.8%, an 11.2-point gap, in arms of 3 and 5 evaluable participants. Study B gives a steadier read: at 24 weeks the same combination reached 25.5% against 16.4% for tirzepatide alone, with 12 of 16 and 14 of 16 participants evaluable, a gap of 9.1 points in a cohort that was three-quarters male.
Why can’t 29.0% at 32 weeks be set against 28.3% at 80?
Five things differ at once. Duration: 32 weeks against 80, in a curve that had not plateaued in either. Analysis: 28.3% is retatrutide’s efficacy estimand, which models everyone staying on treatment; TRIUMPH-1’s primary analysis, counting everyone randomised, was 25.0% against 3.9% on placebo.6 Size: TRIUMPH-1 randomised 2,339 adults; this cohort had 12. Population: the phase 1b participants were lighter, with a mean BMI of 32.8 and weight of 94 kg in Study A (102.3 kg in the 29.0% arm) against 40.0 and 112.7 kg in TRIUMPH-1.10 Design: TRIUMPH-1 was a registrational phase 3 trial; this was a phase 1b safety and tolerability study. A fair reading is that the slide shows eloralintide adds weight loss on top of tirzepatide, which is what an amylin agonist is meant to do. It does not rank two drugs.
What did the larger phase 2b trial add?
On 30 September 2026 Lilly reported the phase 2b trial of the combination, which it calls EloraTZP: 367 adults with obesity or overweight and type 2 diabetes, 48 weeks, double-blind, placebo-controlled, randomised across ten arms.12 The dose-matched arm of the slide above, eloralintide 9 mg with tirzepatide 15 mg, reached 23.3% on the efficacy estimand, against 14.8% for tirzepatide 15 mg alone and 3.0% for placebo.
| Arm (48 weeks, efficacy estimand) | Weight change | A1C change |
|---|---|---|
| Eloralintide 3 + tirzepatide 5 | −13.2% | −2.2 pts |
| Eloralintide 6 + tirzepatide 5 | −19.4% | −2.7 pts |
| Eloralintide 6 + tirzepatide 10 | −19.9% | −2.6 pts |
| Eloralintide 9 + tirzepatide 15 | −23.3% | −2.9 pts |
| Tirzepatide 15 alone | −14.8% | −2.4 pts |
| Eloralintide alone (3, 6, 9) | −8.2%, −12.3%, −11.1% | −1.1 to −1.4 pts |
| Placebo | −3.0% | −0.3 pts |
Phase 2b, NCT06603571, mean baseline 105.4 kg and A1C 8.1%. Trial-arm labels, not guidance.1
That is an 8.5-point gain over tirzepatide alone, smaller than the 11.2 points on the slide and measured in roughly 37 participants per arm rather than 3. The tolerability line is the part the weight figure hides: discontinuation because of adverse events was 10.8% to 27.0% across the combination arms, against 2.9% for tirzepatide alone and, as reported, 16.7% for placebo.1 Lilly’s own plan for the phase 3 programme is a co-formulated product “using an optimized escalation schedule”, which suggests the escalation schedule used in this trial is something Lilly wants to improve.1
Three datasets, side by side
| Dataset | Population | Duration | Headline result | Size |
|---|---|---|---|---|
| Eloralintide + tirzepatide, phase 1b | Obesity or overweight | 32 weeks | −29.0% vs −17.8% tirzepatide alone (modelled; 3 of 12 with data) | 12 per arm |
| EloraTZP, phase 2b | Obesity or overweight with type 2 diabetes | 48 weeks | −23.3% vs −14.8% tirzepatide alone (efficacy estimand) | 367 across 10 arms |
| Retatrutide, TRIUMPH-2 (phase 3) | Obesity or overweight with type 2 diabetes | 80 weeks | −20.8% (efficacy); −18.8% vs −5.1% placebo (primary) | 1,152 |
| Retatrutide, TRIUMPH-1 (phase 3) | Obesity, no diabetes | 80 weeks | −28.3% (efficacy); −25.0% vs −3.9% placebo (primary) | 2,339 |
Sources: Lilly slide and releases, NEJM and The Lancet.1679 Different trials; not a head-to-head comparison.
So is eloralintide plus tirzepatide ahead of retatrutide?
The honest answer is that nobody can say yet, and the most defensible comparison is the narrowest one. In people with type 2 diabetes on the efficacy estimand, EloraTZP reached 23.3% at 48 weeks and retatrutide 20.8% at 80 weeks.18 That favours the combination on speed, but it compares a phase 2b trial of roughly 37 people per arm with a phase 3 trial of about 290, delivered as two separate injections against one, and it sits beside a discontinuation rate for adverse events of up to 27.0% against 3.8% to 11.6% for retatrutide in TRIUMPH-2.18 Neither is better on the evidence available. They are different bets on how to reach a similar place.
Four things would settle it: the phase 3 results for the EloraTZP co-formulation, which Lilly plans to start by the end of 2026; completer and intention-to-treat figures for the phase 1b cohorts; the full phase 2b paper, which has not yet appeared in a journal; and, ideally, a trial that puts both in the same population. We found no head-to-head announced in the sources reviewed. All of these trials are sponsor-funded. Both molecules are Lilly’s, so the framing of one outrunning the other says more about the way figures travel on social media than about the company’s pipeline. For the wider amylin picture see our review of amylin analogues after GLP-1; for the retatrutide side see our analysis of TRIUMPH-2 and TRIUMPH-3 and the retatrutide primer.
None of these compounds is approved by the FDA, the EMA or any other regulator. Nothing here is medical advice, a use recommendation or a dosing instruction; the doses named describe what was administered in supervised trials. Research-use-only reference materials are laboratory chemicals, not medicines, and are not intended for human or veterinary use. Eloralintide is not listed in Condor Research’s catalogue.
Condor Research · Scientific desk
Atrio Sciences s.r.o., IČO 57 669 651, Nitra (SK) · info@condorresearch.com
- Lilly's phase 1b slide shows eloralintide plus tirzepatide at 29.0% weight loss at 32 weeks versus 17.8% for tirzepatide alone, but the headline arm had data from 3 of 12 participants (modelled means).
- The 28.3% retatrutide figure quoted in comparisons is TRIUMPH-1's efficacy estimand at 80 weeks; its primary analysis was 25.0% versus 3.9% on placebo.
- The EloraTZP phase 2b trial (367 adults with type 2 diabetes, 48 weeks) reported 23.3% for the top combination versus 14.8% for tirzepatide 15 mg alone, an 8.5-point gain.
- Discontinuation for adverse events was 10.8% to 27.0% in the combination arms versus 2.9% for tirzepatide alone; Lilly plans a co-formulated phase 3 product with an optimised escalation schedule.
- No head-to-head trial of the combination and retatrutide has been announced; neither is approved, and nothing here is medical advice.
Does eloralintide plus tirzepatide beat retatrutide?
That has not been tested. The 29.0% figure comes from a phase 1b cohort in which only 3 of 12 participants had data at week 32, while retatrutide's 28.3% is an 80-week efficacy-estimand result from a 2,339-person phase 3 trial (25.0% on the primary analysis). The closest like-for-like figure, in type 2 diabetes on the efficacy estimand, is 23.3% at 48 weeks for the combination against 20.8% at 80 weeks for retatrutide, from different trials of different sizes.
What is eloralintide?
Eloralintide (LY3841136) is an investigational once-weekly selective amylin receptor agonist developed by Eli Lilly. In a 263-person phase 2 trial published in The Lancet in 2025, mean weight reductions at 48 weeks ranged from roughly 9% to 20% on the efficacy estimand versus 0.4% on placebo. Phase 3 trials of eloralintide are ongoing. It is not approved anywhere.
Why can't the 29% in 31 weeks be compared with the 28.3% in 80 weeks?
The two figures differ in duration, in the statistical analysis, in population and in size. The 28.3% is retatrutide's efficacy estimand, while the primary analysis of TRIUMPH-1 was 25.0%. The 29.0% is a modelled mean from a 12-person phase 1b cohort in which 3 participants had data at the final timepoint.
What did the EloraTZP phase 2b trial find?
In 367 adults with obesity or overweight and type 2 diabetes, the highest combination arm (eloralintide 9 mg with tirzepatide 15 mg) lost 23.3% of body weight at 48 weeks on the efficacy estimand versus 14.8% for tirzepatide 15 mg alone. Discontinuation because of adverse events ranged from 10.8% to 27.0% in the combination arms versus 2.9% with tirzepatide alone. Lilly plans phase 3 trials of a co-formulated product by the end of 2026.
Is eloralintide, or the eloralintide and tirzepatide combination, approved?
No. Both remain investigational and unapproved by any regulator. Research-use-only reference materials are laboratory chemicals, not for human or veterinary use, and eloralintide is not listed in Condor Research's catalogue.