Weight & Metabolic

Elecoglipron: what the VISTA and SOLSTICE phase 2 trials actually showed, and when phase 3 reports

By the US Health Digest editorial team · Published September 2, 2026 · Every claim linked to its primary source

The 11.8% weight-loss figure attached to AstraZeneca's elecoglipron (AZD5004, licensed from Eccogene) is a week-36 secondary endpoint from a 310-person phase 2 trial, VISTA, measured under the efficacy estimand — not a phase 3 primary result. VISTA's co-primary endpoint was week 26, where the 75 mg arm averaged −10.5% versus −0.6% on placebo. Elecoglipron is not FDA-approved, appears in no DailyMed label, and cannot be prescribed. Its first phase 3 obesity trial, EMBOLD, has a primary completion date of July 31, 2028.

This page reports what trial documents, regulators' databases and company releases say. It is not advice to start, stop or change any medicine.

What elecoglipron is, and how AstraZeneca got it

Elecoglipron is an oral, once-daily, small-molecule GLP-1 receptor agonist. It carries three names in the literature — the international non-proprietary name elecoglipron, AstraZeneca's development code AZD5004, and the originator code ECC5004 — and all three refer to the same compound. Unlike semaglutide, it is not a peptide, which is the same structural distinction that separates Foundayo (orforglipron) from the Wegovy pill.

AstraZeneca did not discover it. On November 9, 2023 the company licensed ECC5004 from Shanghai-based Eccogene for an upfront payment of $185m, with up to a further $1.825bn in clinical, regulatory and commercial milestones plus tiered royalties. AstraZeneca took exclusive rights everywhere except China, where Eccogene retains co-development and co-commercialisation rights; Eccogene's own announcement states the same terms. The molecule was then in a US phase 1 trial.

VISTA: the obesity trial, and exactly where 11.8% comes from

VISTA (NCT06579092) was a double-blind, randomised, placebo-controlled phase 2 dose-ranging study in 310 adults with obesity (BMI ≥30 kg/m²) or overweight (BMI ≥27 kg/m²) plus at least one weight-related condition, and without type 2 diabetes. Participants were recruited in Australia, Canada, Germany, Japan, Taiwan, the UK and the USA and randomised between October 8, 2024 and February 18, 2025. Mean age was 48·4 years, 73% were female, mean bodyweight was 106·9 kg and mean BMI 38·2 kg/m². Five elecoglipron regimens were tested: fixed 5 mg and 15 mg, plus escalations to 50 mg (every four weeks), and to 75 mg by either weekly or every-two-week steps.

The two numbers that matter sit at two different timepoints. The trial's dual primary endpoints were both measured at week 26: percent change in bodyweight, and the proportion reaching at least 5% weight loss. Total treatment duration was 36 weeks, and the week-36 weight result is registered on ClinicalTrials.gov as a secondary outcome. AstraZeneca's June 8, 2026 release reports both, and states plainly: "Weight loss in participants receiving elecoglipron did not plateau, reaching 11.8% at 36 weeks (75mg) versus 0.3% with placebo." The 11.8% is the week-36 secondary figure. The published paper's headline number is the week-26 co-primary: between −2·6% (5 mg) and −10·5% (75 mg with weekly titration) compared with −0·6% with placebo, with 40·4–88·8% of elecoglipron participants reaching at least 5% loss against 15·6% on placebo.

VISTA weight results, from AstraZeneca's press release of June 8, 2026, which labels this table "VISTA Efficacy Estimand Results at Weeks 26 and 36". Week 26 is the co-primary endpoint; week 36 is a registered secondary outcome. Retrieved September 2, 2026.
DoseMean bodyweight change, week 26 (co-primary)Mean bodyweight change, week 36 (secondary)≥10% loss at week 36≥15% loss at week 36
Elecoglipron 5 mg−2.6%−2.7%8.8%4.9%
Elecoglipron 15 mg−5.6%−6.5%35.1%14.0%
Elecoglipron 50 mg (4-weekly escalation)−8.1%−9.2%52.0%14.1%
Elecoglipron 75 mg (weekly escalation)−10.5%−11.8%61.9%39.3%
Elecoglipron 75 mg (2-weekly escalation)−10.0%−11.1%62.5%39.8%
Placebo−0.6%−0.3%4.5%2.7%

The estimand, and why 11.8% cannot be set beside 13.6%

Every weight number above comes from one analysis, and AstraZeneca names it: "The trial used the efficacy estimand to measure the full weight-loss potential of elecoglipron in patients who followed treatment as directed, providing the biological data needed to confirm optimal dosing for Phase III trials." An efficacy estimand asks what the drug does when it is taken. It is a legitimate question, and a systematically friendlier one than the analyses that anchor the two approved pills, which count everyone randomised regardless of whether they stopped.

That distinction is not academic here, because a quarter of VISTA's participants did not finish the drug: 288 participants (93%) completed the study and 231 (75%) completed the assigned treatment. An analysis that sets aside what happened to the other 25% after they stopped will report a larger number than one that does not.

One point of honesty about sourcing. Neither Lancet full text is reachable from this system: thelancet.com returns 403 to our requests, neither paper is deposited in PubMed Central, and Europe PMC lists no open full text for either. Everything we report from the two papers is the structured abstract carried in the PubMed record, and nothing else. We searched the complete text of both PubMed records for the string "estimand" and found zero occurrences — the abstract says only "estimated mean change from baseline". The word "efficacy estimand" comes from AstraZeneca's own release, not from the paper's abstract, and we have not read the paper's statistical analysis section. We flag that rather than imply otherwise. The same search found zero occurrences of "11.8" and of "discontinu" in the VISTA abstract; those figures are the company's.

The four oral GLP-1s, side by side

This is the comparison a reader actually needs, and the estimand column is the reason it is usually done badly. Two of these four are approved and on pharmacy shelves; two are investigational, and both investigational rows rest on phase 2 data analysed under a friendlier estimand than the approved rows.

The four oral GLP-1 receptor agonists in or past phase 3, as of September 2, 2026. Each row is a different trial with its own population, duration, sample size and analysis method. These are not head-to-head comparisons and the results cannot be ranked against one another. Sources linked in each cell; approval status verified against DailyMed and openFDA on September 2, 2026.
DrugSponsorMolecule typeApproval status todayBest (or latest) weight result — dose, duration, comparator, estimandTrial, PMID / NCT
Orforglipron (Foundayo) Eli Lilly Small molecule, non-peptide FDA-approved April 1, 2026 for chronic weight management; not approved for type 2 diabetes −11.2% at the top dose vs −2.1% on placebo at 72 weeks, in 3,127 adults with obesity and without diabetes; treatment-regimen estimand (counts everyone randomised); phase 3 ATTAIN-1, PMID 40960239
Oral semaglutide 25 mg (Wegovy tablets) Novo Nordisk Peptide, formulated as a tablet FDA-approved; in US pharmacies since early January 2026 −13.6% vs −2.2% on placebo at 64 weeks, in 307 adults with overweight or obesity and without diabetes; treatment-policy estimand (counts everyone randomised); phase 3 OASIS 4, PMID 40934115
Aleniglipron (GSBR-1290) Structure Therapeutics Small molecule, non-peptide Investigational. Not approved. Phase 3 ACCOMPLISH-1 began July 13, 2026 −11.3% placebo-adjusted LS mean (95% CI −13.9 to −8.6) at 120 mg at 36 weeks, in 230 adults; efficacy (hypothetical) estimand; phase 2b ACCESS, PMID 42249138; NCT06693843. Phase 3 ACCOMPLISH-1 is NCT07654361
Elecoglipron (AZD5004 / ECC5004) AstraZeneca, in-licensed from Eccogene Small molecule, non-peptide Investigational. Not approved anywhere, not available, cannot be prescribed. Phase 3 began June 29, 2026 −10.5% at 75 mg vs −0.6% on placebo at 26 weeks (co-primary); −11.8% vs −0.3% at 36 weeks (secondary), in 310 adults; efficacy estimand — AstraZeneca's description of the analysis, a word that does not appear in the paper's abstract; phase 2 VISTA, PMID 42259337; NCT06579092
Four bars, four different questions Mean body-weight loss from baseline (%). Each pair is one randomised comparison. Pairs are from different trials, phases, durations and estimands and are not comparable to each other. VISTA — phase 2, obesity/overweight, no diabetes, week 26 CO-PRIMARY, efficacy estimand 10.5% Elecoglipron 75 mg (weekly esc.) 0.6% Placebo VISTA — same trial, week 36 SECONDARY, efficacy estimand — this is the widely quoted 11.8% 11.8% Elecoglipron 75 mg (weekly esc.) 0.3% Placebo OASIS 4 — phase 3, obesity/overweight, no diabetes, week 64 primary, treatment-policy estimand 13.6% Oral semaglutide 25 mg 2.2% Placebo ATTAIN-1 — phase 3, obesity, no diabetes, week 72 primary, treatment-regimen estimand 11.2% Orforglipron 36 mg 2.1% Placebo 0% 5% 10% 15% Mean body-weight loss from baseline, % — every bar drawn at exactly 34 px per percentage point from a shared zero at x=230 Sources: VISTA, Lancet 2026 (PMID 42259337) and AstraZeneca release, 8 June 2026; OASIS 4, NEJM 2025 (PMID 40934115); ATTAIN-1, NEJM 2025 (PMID 40960239). Retrieved September 2, 2026.
Bars are drawn at exactly 34 px per percentage point from a shared zero at x=230 — so 10.5% is 357.0 px, 11.8% is 401.2 px, 13.6% is 462.4 px and 11.2% is 380.8 px. These are cross-trial comparisons, not head-to-head results. Only the two bars inside each labelled pair were randomised against one another; the four pairs come from different trials, phases, populations, durations and estimands, and the ordering between pairs is not a measurement of which drug is stronger. Aleniglipron is deliberately absent: the ACCESS paper reports a placebo-adjusted least-squares mean difference (−11.3%) rather than a mean change from baseline in the drug arm, so it is a different quantity and drawing it on this axis would be wrong. It appears in the table above instead.

SOLSTICE: the type 2 diabetes trial, and its open-label semaglutide arm

VISTA and SOLSTICE are different trials in different populations, and conflating them is the most common error in the coverage. SOLSTICE (NCT06579105) was a phase 2b trial in 406 randomised adults with type 2 diabetes across nine countries, managed on diet and exercise alone or on metformin or an SGLT2 inhibitor monotherapy, with entry HbA1c of 7·0–10·5% (6·5–10·5% in the USA). Its primary endpoint was change in HbA1c at 26 weeks, not weight. The paper reports HbA1c falling between −0·91% (95% CI −1·25 to −0·58; 5 mg) and −1·88% (−2·23 to −1·53; 75 mg with two-weekly escalation) against −0·15% (−0·42 to 0·12) with placebo. AstraZeneca reports the same top-dose result rounded to 1.9% against 0.2%, and adds a weight figure of −7.7% versus −1.7% on placebo at 26 weeks.

SOLSTICE also contained an oral semaglutide 14 mg arm, and this is where care is needed. That arm was open-label — participants and investigators knew what it was — while the elecoglipron and placebo arms were masked, and AstraZeneca describes it as included "for exploratory comparison". It is not a head-to-head efficacy result, it is not the 25 mg weight-management dose, and it should not be read as evidence that one drug beats the other. For what an actual randomised comparison of two GLP-1 pills looks like, see our account of ACHIEVE-3, orforglipron versus oral semaglutide.

Tolerability: where this class differentiates, and what has not been published

Gastrointestinal side effects are the practical limit on oral GLP-1 dosing, and VISTA's numbers are not small. In the 75 mg arm compared with placebo, AstraZeneca reports nausea in 55% versus 20%, constipation 41% versus 6%, diarrhoea 35% versus 25% and vomiting 29% versus 5%. Overall adverse events were reported by 84% (27 of 32) to 98% (48 of 49) of participants across elecoglipron doses compared with 84% (68 of 81) in the placebo group. In SOLSTICE the 75 mg-versus-placebo figures were nausea 37% versus 3%, constipation 29% versus 4%, diarrhoea 21% versus 15% and vomiting 18% versus 1%.

What is not published is the number readers most want. AstraZeneca's release says only that "Adverse events leading to discontinuation were infrequent in both trials and no liver safety signals were observed" — a characterisation with no percentage attached. We searched the complete text of both PubMed records for the string "discontinu" and found no occurrences in either abstract. The nearest available proxy in VISTA is that 231 of 310 participants (75%) completed the assigned treatment, which counts every reason for stopping, not only adverse events. Neither trial has posted results on ClinicalTrials.gov as of September 2, 2026. By comparison, the published discontinuation figures in this class are specific: treatment-related discontinuations of 10.4% across aleniglipron arms in ACCESS, and 5.3% to 10.3% on orforglipron in ATTAIN-1 versus 2.7% on placebo. Until elecoglipron's equivalent is published, the honest answer is that it is unknown to the public.

Not approved, not available — and the phase 3 calendar

We checked this three ways on September 2, 2026. Elecoglipron returns no matches in the openFDA drug label API, none in openFDA's Drugs@FDA database, and no records in the DailyMed drug-name service (database published September 1, 2026). AstraZeneca describes it as investigational. There is no US prescribing information, no approved dose, no pharmacy supply and no legitimate route to obtain it outside a clinical trial.

The phase 3 programme is large, and the dates are public. AstraZeneca's release names the EMBOLD trials in obesity and the ELUMINATE trials in type 2 diabetes, plus cardiovascular and kidney outcome trials.

The elecoglipron phase 3 programme as registered on ClinicalTrials.gov, retrieved September 2, 2026. "Primary completion" is the date the sponsor expects final data collection for the primary endpoint — publication and any regulatory decision come later.
TrialPopulationEnrolment (planned)Primary endpointStatus and startEstimated primary completion
EMBOLD master protocol (NCT07667803)Obesity or overweight, with or without type 2 diabetes4,500Percent body weight loss at 72 weeksRecruiting; began June 29, 2026July 31, 2028 (study completion July 30, 2029)
EMBOLD-Asia (NCT07775404)Asian participants with obesity or overweight, with or without type 2 diabetes351Percent weight change and ≥5% weight loss at week 52Recruiting; began August 21, 2026June 4, 2027
ELUMINATE-2 (NCT07662213)Type 2 diabetes; open-label, versus oral semaglutide1,200Change in HbA1c at week 52Recruiting; began July 6, 2026May 23, 2028
ELUMINATE-5 (NCT07662109)Type 2 diabetes; elecoglipron plus dapagliflozin2,000Change in HbA1c at week 40Recruiting; began July 6, 2026July 5, 2028
ELUMINATE-1 (NCT07662044)Type 2 diabetes; alone or with dapagliflozin, versus placebo800Change in HbA1c at week 40Recruiting; began July 6, 2026July 14, 2028
ELUMINATE-4 (NCT07662135)Type 2 diabetes with impaired renal function, on background dapagliflozin900Change in HbA1c at week 40Recruiting; listed start July 6, 2026July 13, 2028
ELUMINATE-3 (NCT07664553)Type 2 diabetes on background insulin600Change in HbA1c at week 40Recruiting; began July 8, 2026May 23, 2028
ELEVATE-HF (NCT07761117)Heart failure with preserved or mildly reduced ejection fraction6,950CV death or hospitalisation/urgent visit for heart failureNot yet recruiting; listed start August 28, 2026September 3, 2029
ELEVATE-CKD (NCT07753993)Chronic kidney disease7,000≥50% sustained eGFR decline, end-stage kidney disease, or all-cause mortalityNot yet recruiting; listed start August 28, 2026October 4, 2030

Read the first row again. The obesity trial that would produce a number comparable to ATTAIN-1's −11.2% at 72 weeks is not scheduled to finish collecting its primary data until July 31, 2028. On any normal timetable, analysis, publication and a regulatory review follow after that.

Why phase 2 is not phase 3, with a precedent from this exact class

The most useful cautionary example is orforglipron itself, and the parallel is close enough to be uncomfortable. Its phase 2 trial enrolled 272 adults with obesity or overweight, ran 36 weeks, assessed bodyweight at week 26 as the primary endpoint and week 36 as a secondary endpoint — the same shape as VISTA. It reported mean weight change of −8.6% to −12.6% at week 26 and −9.4% to −14.7% at week 36. The 14.7% was the headline everywhere. The phase 3 trial that actually supported approval, ATTAIN-1, enrolled 3,127 people, ran 72 weeks and reported −11.2% under the treatment-regimen estimand.

Different durations, populations and analysis methods make that an illustration rather than a measurement. But the direction of travel from a phase 2 headline to a phase 3 primary result is routinely downward, and the same caution applies to aleniglipron. Broader context on the class is in our GLP-1 statistics page.

Dosing, food and water: the one place elecoglipron may genuinely differ

Both papers describe elecoglipron as taken once daily "without food or fluid restriction" (VISTA) and "with no food or fluid restrictions" (SOLSTICE). If phase 3 confirms it, that matters, because the two approved pills are not alike on this point.

The Wegovy tablet is a peptide and its label is strict: "Take one WEGOVY tablet orally once daily on an empty stomach in the morning with water (up to 4 ounces)", followed by "After taking a WEGOVY tablet, wait at least 30 minutes before eating food, drinking beverages or taking other oral medications". Foundayo, a small molecule, has no such requirement — its label says "Take FOUNDAYO orally once daily, with or without food." So the food-and-water advantage elecoglipron would bring is the advantage orforglipron already has; it separates small molecules from peptides, not elecoglipron from its competitors. Our guide to moving from an injection to a pill covers what those dosing conditions mean day to day.

The manufacturing argument is AstraZeneca's, and we report it as a claim: the company states that elecoglipron "has potential to expand treatment options for many patients due to easier scalability compared to peptides and no food or fasting restrictions." Small molecules are made by chemical synthesis rather than the biological processes peptides require, and none of the three pills here needs the refrigerated handling the Wegovy injection label specifies at 2°C to 8°C. That is an argument about supply and cold chain, not about how much weight anyone loses.

Questions worth raising with a prescriber

Frequently asked questions

Can I get elecoglipron now?

No. Elecoglipron is investigational. As of September 2, 2026 it returns no matches in the openFDA drug label API, no matches in openFDA's Drugs@FDA database, and no records in DailyMed's drug-name service. There is no US prescribing information and no approved dose. The only lawful way to receive it is enrolment in one of AstraZeneca's clinical trials.

Where does the 11.8% weight-loss figure come from?

From VISTA, a phase 2 trial in 310 adults. It is the mean bodyweight change at week 36 in the 75 mg arm escalated weekly, versus 0.3% on placebo, under the efficacy estimand. Week 36 is a registered secondary outcome. The trial's dual primary endpoints were both at week 26, where the same arm averaged −10.5% versus −0.6% on placebo.

Is elecoglipron better than the Wegovy pill or Foundayo?

That question cannot be answered from the published evidence. No trial has compared elecoglipron with either approved pill in obesity. VISTA is phase 2, ran 36 weeks and was analysed under what AstraZeneca calls an efficacy estimand; OASIS 4 is phase 3, ran 64 weeks and used a treatment-policy estimand; ATTAIN-1 is phase 3, ran 72 weeks and used a treatment-regimen estimand. Comparing those numbers directly compares four different questions.

When will phase 3 results be available?

The EMBOLD master protocol in obesity (NCT07667803, 4,500 participants, percent body weight loss at 72 weeks) lists an estimated primary completion date of July 31, 2028 and study completion of July 30, 2029. The smaller EMBOLD-Asia study lists June 4, 2027. The cardiovascular and kidney outcome trials list September 3, 2029 and October 4, 2030.

Does elecoglipron have to be taken on an empty stomach?

The published trials describe it as once daily with no food or fluid restriction. That is a trial description of an unapproved drug, not label instructions, because there is no label. By contrast the Wegovy tablet label directs an empty stomach in the morning with up to 4 ounces of water and a 30-minute wait, while the Foundayo label allows dosing with or without food.

How common were side effects, and how many people quit?

In VISTA's 75 mg arm versus placebo, AstraZeneca reports nausea 55% versus 20%, constipation 41% versus 6%, diarrhoea 35% versus 25% and vomiting 29% versus 5%. The discontinuation rate specifically due to adverse events has not been published as a number: the company calls it infrequent, and the word does not appear in either PubMed abstract. What is published is that 75% of VISTA participants completed the assigned treatment.

Are VISTA and SOLSTICE the same population?

No. VISTA enrolled adults with obesity or overweight and at least one weight-related condition without type 2 diabetes, and its primary endpoint was weight. SOLSTICE enrolled adults with type 2 diabetes and its primary endpoint was HbA1c. The 11.8% weight figure belongs to VISTA only.

Sources

This article is for information only and is not medical advice. Prescription weight-loss medication requires evaluation by a licensed clinician. See our medical disclaimer.