← All trial deep-dives

AMPLITUDE-O trial deep-dive: efpeglenatide CV outcomes in T2D (Gerstein 2021 NEJM)

Last verified 2026-05-28 · Phase 3 · Primary completion December 2020; results published NEJM September 2021. Trial record subsequently marked Terminated on ClinicalTrials.gov after Sanofi discontinued the efpeglenatide development program; the drug was never submitted for FDA or EMA approval. · NCT03496298 ↗

By Eli Marsden · Founding Editor
Editorially reviewed (not clinically reviewed) · How we verify contentLast reviewed

AMPLITUDE-O (Effect of Efpeglenatide on Cardiovascular Outcomes) was a multicenter, double-blind, placebo-controlled, event-driven phase 3 trial sponsored by Sanofi and run at 344 sites across 28 countries. Investigators randomized 4,076 adults with type 2 diabetes and either a history of cardiovascular disease or current kidney disease (eGFR 25.0 to 59.9 mL/min/1.73 m²) plus at least one additional cardiovascular risk factor to weekly subcutaneous efpeglenatide 4 mg, efpeglenatide 6 mg, or volume-matched placebo in a 1:1:1 design, stratified by baseline SGLT2-inhibitor use. The primary endpoint was the first occurrence of a three-point major adverse cardiovascular event composite — nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular or undetermined causes. Median follow-up was 1.81 years, the shortest of any positive GLP-1 CVOT to date. Results published by Gerstein and colleagues in the New England Journal of Medicine on September 2, 2021 reported a hazard ratio of 0.73 for the primary endpoint — the first positive cardiovascular outcomes trial for an exendin-based, structurally non-human GLP-1 receptor agonist and the validation that the cardioprotective class effect extends beyond the human-GLP-1 backbone shared by liraglutide, dulaglutide, and semaglutide.

Enrollment
4,076
Duration
Median follow-up 1.81 years; trial was event-driven and stopped after the pre-specified MACE accrual was reached
Drug
Efpeglenatide 4 mg and 6 mg (weekly subcutaneous)
Population
Adults with type 2 diabetes and either a history of cardiovascular disease (myocardial infarction, ischemic stroke, unstable angina with electrocardiographic changes, myocardial ischemia on imaging or stress test, or coronary, carotid, or peripheral arterial revascularization) or current kidney disease defined as estimated glomerular filtration rate of 25.0 to 59.9 mL/min/1.73 m² plus at least one additional cardiovascular risk factor. Per the NEJM publication: mean age 64.5 years, 33% female, mean HbA1c 8.9%, mean BMI 32.7 kg/m², mean diabetes duration 15.4 years. 89.6% had established cardiovascular disease at randomization. Background therapy included metformin (74.4%), insulin (61.7%), SGLT2 inhibitors (15.2% — the highest baseline SGLT2 use of any GLP-1 CVOT at that point), and antihypertensives, statins, and antiplatelet agents per standard of care.

Primary endpoint

Three-point MACE: first occurrence of nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular or undetermined causes (time-to-first-event, pooled efpeglenatide 4 mg + 6 mg vs placebo)

Treatment arm

189 of 2,717 (7.0%); 3.9 events per 100 person-years

Comparator

125 of 1,359 (9.2%); 5.3 events per 100 person-years

Treatment difference: Hazard ratio 0.73 (95% CI 0.58 to 0.92); P<0.001 for noninferiority; P=0.007 for superiority

27% relative reduction in three-point MACE with pooled efpeglenatide vs placebo over a median 1.81 years — the shortest follow-up among positive GLP-1 CVOTs and the first positive cardiovascular outcomes result for an exendin-based (non-human GLP-1 backbone) receptor agonist. The pooled-arms design was pre-specified as the primary regulatory question; both noninferiority and superiority were met.

Secondary endpoints

EndpointTreatmentComparatorDifference
Three-point MACE — 6 mg dose vs placebo (pre-specified dose-comparison secondary, reported in detail by Gerstein 2023 Circulation)

35% relative reduction in MACE with the 6 mg dose alone — the larger of the two dose effects and the driver of the pooled-arm result.

84 of 1,358 (6.2%) on efpeglenatide 6 mg125 of 1,359 (9.2%) on placeboHazard ratio 0.65 (95% CI 0.50 to 0.86); P=0.0027
Three-point MACE — 4 mg dose vs placebo (pre-specified dose-comparison secondary)

Directional 18% reduction at the 4 mg dose that did not reach statistical significance on its own. Combined with the 6 mg result, the trial reported a clear dose-response across primary and all secondary outcomes (P for trend ≤0.018 across endpoints).

105 of 1,359 (7.7%) on efpeglenatide 4 mg125 of 1,359 (9.2%) on placeboHazard ratio 0.82 (95% CI 0.63 to 1.06); P=0.14
Composite renal outcome: sustained new macroalbuminuria, ≥40% decline in eGFR, or renal failure (pre-specified secondary, pooled efpeglenatide vs placebo)

32% relative reduction in the kidney composite — one of the largest renal-protection signals reported in a primary GLP-1 CVOT publication, foreshadowing the dedicated semaglutide kidney trial FLOW that reported in 2024.

353 of 2,717 (13.0%)250 of 1,359 (18.4%)Hazard ratio 0.68 (95% CI 0.57 to 0.79); P<0.001
Composite renal outcome — 6 mg dose vs placebo (dose-comparison secondary)

37% relative reduction in the kidney composite at the higher dose. The dose-response on renal endpoints was steeper than on MACE.

Efpeglenatide 6 mg armPlacebo armHazard ratio 0.63 (95% CI not reported in abstract); P<0.0001
Composite renal outcome — 4 mg dose vs placebo (dose-comparison secondary)

27% relative reduction at the lower dose — formally significant on the kidney composite even though the same dose was not formally significant on MACE alone.

Efpeglenatide 4 mg armPlacebo armHazard ratio 0.73 (95% CI not reported in abstract); P=0.0009
Expanded cardiovascular composite: MACE, coronary revascularization, or hospitalization for unstable angina — 6 mg dose vs placebo

27% relative reduction in the broader CV composite at the high dose. The 4 mg dose showed a directional 15% reduction (HR 0.85, P=0.17).

Efpeglenatide 6 mg armPlacebo armHazard ratio 0.73; P=0.011
MACE or any-cause death — 6 mg dose vs placebo

33% relative reduction at the high dose. The 4 mg dose showed a directional 19% reduction (HR 0.81, P=0.08).

Efpeglenatide 6 mg armPlacebo armHazard ratio 0.67; P=0.0021
Kidney function outcome: sustained ≥40% eGFR decline, renal failure, or death — 6 mg dose vs placebo

39% relative reduction in the harder kidney-function-plus-mortality composite at the high dose. The 4 mg dose was neutral on this endpoint (HR 0.97, P=0.83) — the only secondary outcome where the lower dose showed no signal.

Efpeglenatide 6 mg armPlacebo armHazard ratio 0.61; P=0.0072
Composite of MACE, any death, heart-failure hospitalization, or kidney function outcome — 6 mg dose vs placebo

37% relative reduction in the broadest pre-specified composite at the high dose. The 4 mg dose showed a directional 19% reduction that fell just outside formal significance (HR 0.81, P=0.067).

Efpeglenatide 6 mg armPlacebo armHazard ratio 0.63; P=0.0002
Dose-response trend across all primary and secondary outcomes

Statistically significant graded dose-response — the AMPLITUDE-O dose-exploration analysis (Gerstein 2023, Circulation, PMID 36802715) is the cleanest evidence in the GLP-1 class that titrating to higher doses may amplify both cardiovascular and renal benefit, a hypothesis that informed later high-dose strategies for semaglutide and tirzepatide.

Placebo < 4 mg < 6 mg ordering across endpointsReference (placebo)P for trend ≤0.018 across all primary and secondary cardiovascular and kidney outcomes

Adverse events

EventTreatment rateComparator rate
Diarrhea, constipation, nausea, vomiting, or bloating (category-level GI events)

The primary NEJM publication reports these gastrointestinal adverse events as the dominant tolerability signal, qualitatively consistent with the rest of the GLP-1 class. Numerical AE breakdowns by arm are in the NEJM full text and supplement (not freely accessible) — for verbatim AE percentages, see Gerstein et al., NEJM 2021;385:896-907 directly.

Higher with efpeglenatide than placebo (category-level finding reported in the NEJM abstract; arm-specific percentages reported in the full text are behind the NEJM paywall and not reproduced here)Lower than the efpeglenatide arms
Acute pancreatitis (adjudicated)

Consistent with the broader GLP-1 class CVOT pattern; no specific pancreatitis signal flagged in the primary publication's safety summary.

No imbalance versus placebo reported in the NEJM publicationReference
Pancreatic cancer (adjudicated)

No pancreatic-cancer safety signal reported over the 1.8-year median follow-up. Short observation window relative to LEADER (3.8 years) limits inference.

No imbalance versus placebo reported in the NEJM publicationReference
Diabetic retinopathy complications

Contrasts with the well-known retinopathy signal in SUSTAIN-6 for semaglutide. Short follow-up and a different molecular backbone may both contribute.

No retinopathy safety signal reported in the NEJM publicationReference
Discontinuation due to adverse events

Pattern of discontinuation driven primarily by gastrointestinal intolerance, consistent with the class. For arm-specific numbers, see NEJM 2021;385:896-907 directly.

Higher with efpeglenatide than placebo per the NEJM publication; specific percentages in the full text and supplementLower than the efpeglenatide arms

Subgroup analyses

  • Established cardiovascular disease at baseline (89.6% of cohort): Treatment effect on primary MACE consistent with the overall result; the trial enrolled predominantly secondary-prevention participants by design

    AMPLITUDE-O had the highest proportion of established CVD at entry among the major GLP-1 CVOTs, which contributed to the high event rate per person-year and allowed the trial to reach its event target in 1.8 years.

  • Baseline SGLT2-inhibitor use (15.2% — randomization was stratified by this variable): Cardiovascular benefit of efpeglenatide consistent in participants on and off baseline SGLT2 inhibitors (no significant interaction in the primary publication)

    First positive GLP-1 CVOT with prospectively stratified SGLT2 background; supported the now-standard combined GLP-1 plus SGLT2 cardio-renal strategy for high-risk T2D.

  • Baseline kidney disease (eGFR 25.0 to 59.9 mL/min/1.73 m²): Cardiovascular and kidney benefits observed across the reduced-eGFR subgroup; arm-specific subgroup effects detailed in the Gerstein 2024 Diabetes Obes Metab analysis (PMID 38116691)

    AMPLITUDE-O is one of the few GLP-1 CVOTs to enroll participants with eGFR as low as 25 mL/min/1.73 m², extending the GLP-1 evidence base into stage 3b-4 chronic kidney disease.

  • Dose comparison (4 mg vs 6 mg): Graded dose-response: 6 mg HR 0.65 for MACE, 4 mg HR 0.82 for MACE; P for trend ≤0.018 across primary and secondary outcomes

    The cleanest GLP-1 CVOT evidence that within-class titration to higher exposure amplifies cardiovascular and renal benefit. Detailed in Gerstein 2023 Circulation (PMID 36802715).

  • Sex (33% female): Treatment effect on primary MACE consistent across women and men in the primary publication

    Underrepresentation of women is typical of CVOTs in established cardiovascular disease populations.

Clinical significance

AMPLITUDE-O sits in an unusual position in the GLP-1 evidence base: a fully positive cardiovascular outcomes trial for a drug that was never marketed. The trial answered three questions the rest of the class had left open. First, it showed that an exendin-based GLP-1 receptor agonist — structurally derived from a lizard peptide rather than the human GLP-1 backbone of liraglutide, dulaglutide, and semaglutide — produces a MACE reduction of the same magnitude as the human-backbone agents, removing the molecular-backbone variable as an explanation for the class effect. Second, the trial reported the most explicit dose-response analysis in any GLP-1 CVOT, with the 6 mg arm driving a 35% MACE reduction while the 4 mg arm landed at 18% directional, and a 32% pooled-arm reduction in the kidney composite — evidence that informed the field's later move toward higher-dose semaglutide and tirzepatide. Third, AMPLITUDE-O reached its event target in 1.81 years of median follow-up, the shortest of any positive GLP-1 CVOT, because the cohort was enriched for established cardiovascular disease (89.6%) and concurrent kidney disease — a design choice that shortened the time-to-evidence but also narrowed the generalizability to lower-risk patients. Sanofi nonetheless discontinued efpeglenatide before any regulatory submission, and the drug has not been resurrected by its originator Hanmi Pharmaceutical or any subsequent licensee, leaving AMPLITUDE-O as a load-bearing class-effect paper that is cited far more often than the drug it tested is dispensed.

Frequently asked questions

Why was efpeglenatide shelved if AMPLITUDE-O was positive?

Efpeglenatide was originally developed by Hanmi Pharmaceutical in South Korea and licensed to Sanofi in 2015. By the time AMPLITUDE-O reported in September 2021, the GLP-1 landscape had changed dramatically: weekly semaglutide (Ozempic) had already launched in 2017 with its own positive CVOT, oral semaglutide (Rybelsus) had launched in 2019, and tirzepatide phase 3 readouts in obesity were imminent. Sanofi had also exited diabetes research as a strategic priority during the same period, divesting from Lantus and Toujeo's franchises and reorganizing around vaccines, immunology, and rare diseases. A weekly exendin-based GLP-1 with a positive but unremarkable cardiovascular signal in a crowded market did not survive that portfolio review. Sanofi returned development rights to Hanmi after the trial, and no subsequent licensee has filed for FDA or EMA approval. The ClinicalTrials.gov record was eventually marked Terminated even though the primary endpoint was met — a reflection of program discontinuation, not trial failure.

Does AMPLITUDE-O affect current GLP-1 prescribing?

Not directly, because efpeglenatide is not commercially available. The trial does affect how clinicians and guideline writers reason about the GLP-1 cardiovascular class effect. Before AMPLITUDE-O, every positive GLP-1 CVOT (LEADER for liraglutide, SUSTAIN-6 for semaglutide, REWIND for dulaglutide, eventually SELECT for semaglutide 2.4 mg in obesity) tested a human-GLP-1-backbone molecule. AMPLITUDE-O is the only positive CVOT for an exendin-based agent, which strengthens the inference that the cardioprotection is a property of GLP-1 receptor activation rather than the specific peptide scaffold. AMPLITUDE-O is also the cleanest GLP-1 dose-response analysis in the published literature (Gerstein 2023 Circulation, PMID 36802715), which has been cited in support of titrating semaglutide to the higher 2.0 mg weekly dose and using maintenance doses of tirzepatide at 10 mg or higher when tolerated.

What was special about the exendin-4 backbone?

Efpeglenatide is built on exendin-4, a 39-amino-acid peptide originally isolated from the saliva of the Gila monster (Heloderma suspectum). Exendin-4 binds the human GLP-1 receptor with high affinity but shares only about 53% sequence identity with human GLP-1, which gives it natural resistance to dipeptidyl peptidase-4 (DPP-4) degradation. Exenatide (Byetta, twice daily; Bydureon, weekly) was the first FDA-approved exendin-based GLP-1 agonist. Lixisenatide (Adlyxin, once daily) was the second. Both failed to show cardiovascular superiority in their CVOTs (EXSCEL for exenatide was neutral; ELIXA for lixisenatide was neutral). That history created a working hypothesis in the field that exendin-based agents might be cardiovascularly inert compared with human-backbone agents — a hypothesis AMPLITUDE-O refuted. The likely reconciliation is that EXSCEL and ELIXA were underpowered (and ELIXA enrolled an acute-coronary-syndrome population with very different baseline risk), not that the molecular backbone matters.

How does AMPLITUDE-O compare to LEADER, SUSTAIN-6, and REWIND?

All four are positive GLP-1 cardiovascular outcomes trials in type 2 diabetes, and the MACE hazard ratios are remarkably similar across them. LEADER (liraglutide, 9,340 participants, 3.8 years median follow-up, Marso 2016): MACE HR 0.87. SUSTAIN-6 (semaglutide subcutaneous, 3,297 participants, 2.1 years median follow-up, Marso 2016): MACE HR 0.74. REWIND (dulaglutide, 9,901 participants, 5.4 years median follow-up, Gerstein 2019): MACE HR 0.88. AMPLITUDE-O (efpeglenatide, 4,076 participants, 1.81 years median follow-up, Gerstein 2021): MACE HR 0.73. AMPLITUDE-O is the shortest of the four because the cohort was enriched for established cardiovascular and kidney disease (89.6% with prior CV events versus 32% in REWIND), which produced a high baseline event rate and a fast accrual to the event target. The narrow spread of hazard ratios — 0.73 to 0.88 across four different molecules — is one of the strongest cardiovascular class-effect signals in modern cardiometabolic medicine.

Was there a kidney signal?

Yes, and it was one of the largest renal signals in any GLP-1 CVOT primary publication. The pre-specified renal composite — sustained new macroalbuminuria, sustained ≥40% decline in eGFR, or renal failure — fell 32% with pooled efpeglenatide versus placebo (353 of 2,717 events on efpeglenatide vs 250 of 1,359 on placebo; HR 0.68, 95% CI 0.57 to 0.79, P<0.001). The dose-response analysis (Gerstein 2023, Circulation, PMID 36802715) showed the 6 mg arm reduced the composite by 37% (P<0.0001) and the 4 mg arm reduced it by 27% (P=0.0009), making the kidney composite the only AMPLITUDE-O endpoint where the 4 mg dose was independently significant. The harder kidney-function-plus-mortality composite (≥40% sustained eGFR decline, renal failure, or death) was reduced 39% at the 6 mg dose (HR 0.61, P=0.0072) but was neutral at the 4 mg dose. AMPLITUDE-O's renal signal foreshadowed the dedicated semaglutide kidney trial FLOW (PMID 38785218), which reported in 2024 and confirmed a GLP-1 kidney-protection effect on harder endpoints.

Is there any chance efpeglenatide gets revived?

There is no public Hanmi or Sanofi program advancing efpeglenatide for type 2 diabetes or cardiovascular indications as of mid-2026. Hanmi has pivoted its GLP-1 portfolio toward HM15211 (efinopegdutide, a GLP-1 / glucagon dual agonist that Merck licensed and is developing for metabolic dysfunction-associated steatohepatitis) and HM15275 (a triple GLP-1 / GIP / glucagon agonist in earlier-phase development). The commercial logic that led Sanofi to discontinue efpeglenatide — a crowded weekly GLP-1 market dominated by semaglutide and tirzepatide — has only intensified. Efpeglenatide could in principle be repurposed as a lower-cost generic-style GLP-1 if its composition-of-matter patents lapse and a biosimilar pathway opens, but no such program has been announced. For now, AMPLITUDE-O remains a one-trial drug: the best-studied GLP-1 receptor agonist that patients cannot fill at a pharmacy.

References

  1. 1.Gerstein HC, Sattar N, Rosenstock J, et al.; AMPLITUDE-O Trial Investigators. Cardiovascular and Renal Outcomes with Efpeglenatide in Type 2 Diabetes. N Engl J Med. 2021. PMID: 34215025.
  2. 2.Gerstein HC, Branch K, Heenan L, et al.; AMPLITUDE-O Trial Investigators. Exploring the Relationship Between Efpeglenatide Dose and Cardiovascular Outcomes in Type 2 Diabetes: Insights From the AMPLITUDE-O Trial. Circulation. 2023. PMID: 36802715.
  3. 3.Gerstein HC, Sattar N, Rosenstock J, et al. Design and baseline characteristics of the AMPLITUDE-O cardiovascular outcomes trial of efpeglenatide, a weekly glucagon-like peptide-1 receptor agonist. Diabetes Obes Metab. 2021. PMID: 33026143.
  4. 4.Gerstein HC, Colhoun HM, Dagenais GR, et al.; AMPLITUDE-O Trial Investigators. Cardiovascular and renal outcomes with varying degrees of kidney disease in high-risk people with type 2 diabetes: An epidemiological analysis of data from the AMPLITUDE-O trial. Diabetes Obes Metab. 2024. PMID: 38116691.
  5. 5.Sanofi. Effect of Efpeglenatide on Cardiovascular Outcomes (AMPLITUDE-O). ClinicalTrials.gov, NCT03496298. 2021. https://clinicaltrials.gov/study/NCT03496298

Related trial deep-dives