Albiglutide
Albiglutide (Tanzeum, Eperzan), once-weekly GLP-1 receptor agonist fused to human albumin
Written by Aaron CuhaReviewed Sep 2026
Also known as: Tanzeum, Eperzan, GSK716155
A once-weekly GLP-1 receptor agonist approved in 2014 and withdrawn from every market by 2018. In the 9,463 person Harmony Outcomes trial it reduced cardiovascular death, heart attack or stroke by 22 percent against placebo, and in a head to head trial it lowered HbA1c less than liraglutide. The drug is no longer available anywhere.
Overview
Albiglutide is a recombinant fusion protein: two tandem copies of human GLP-1 (7-36), each with glycine replacing alanine at position 8 to resist DPP-4, fused to human albumin. The albumin gives it a half-life of about 5 days and once-weekly dosing, and a molecular weight of 72,970 daltons, which makes it by far the largest molecule ever sold as a GLP-1 drug. GlaxoSmithKline developed it as GSK716155; the FDA approved it as Tanzeum on 15 April 2014 and the European Medicines Agency as Eperzan in March 2014, both for glycaemic control in adults with type 2 diabetes.
The defining trial came after the approvals. Harmony Outcomes (Lancet, 2018) randomised 9,463 adults aged 40 and over with type 2 diabetes and established cardiovascular disease at 610 sites in 28 countries to albiglutide 30 to 50 mg weekly or placebo on top of standard care. Over a median 1.6 years, the composite of cardiovascular death, myocardial infarction or stroke occurred in 338 of 4,731 (7 percent, 4.6 per 100 person-years) on albiglutide against 428 of 4,732 (9 percent, 5.9 per 100 person-years) on placebo: hazard ratio 0.78 (95 percent CI 0.68 to 0.90), P below 0.0001 for non-inferiority and P equals 0.0006 for superiority. Acute pancreatitis (10 against 7), pancreatic cancer (6 against 5) and medullary thyroid carcinoma (0 in each) did not differ. The trial was funded by GlaxoSmithKline. Before that, the phase 3 programme had included a randomised, open-label head to head against liraglutide (HARMONY 7, Lancet Diabetes and Endocrinology 2014): in 841 adults over 32 weeks, HbA1c fell 0.78 percentage points on albiglutide against 0.99 on liraglutide, a difference of 0.21 (95 percent CI 0.08 to 0.34) that did not meet the prespecified non-inferiority margin (P equals 0.0846). Albiglutide caused more injection site reactions (12.9 against 5.4 percent) and fewer gastrointestinal events (35.9 against 49.0 percent).
The drug was withdrawn anyway. The EMA's Eperzan page records that on 29 October 2018 the marketing authorisation was withdrawn at the request of the holder, GlaxoSmithKline Trading Services Limited, which had notified the European Commission of its decision to permanently discontinue the marketing of the product for commercial reasons. Drugs@FDA lists Tanzeum's marketing status as Discontinued. Two registry entries say the same in the sponsor's words: a planned paediatric phase 3 (NCT03015519) was withdrawn because albiglutide would have been withdrawn from the market before the study ended, and a small extension (NCT02750930) was terminated as a result of GSK business considerations and not due to quality, safety or efficacy concerns. No other reason appears in any primary source, and this profile states none.
Mechanism of action
Albiglutide activates the GLP-1 receptor: glucose-dependent insulin secretion, suppression of glucagon, slower gastric emptying and reduced appetite, the same mechanism as every drug in the class. Two features set it apart. The glycine-for-alanine substitution at position 8 of each GLP-1 copy blocks cleavage by DPP-4, and the fusion to human albumin, produced in Saccharomyces cerevisiae, keeps the molecule in circulation for days. Its size may also limit penetration into the brain relative to smaller lipidated analogues, which is one hypothesis offered for its comparatively modest weight loss, but that is inference rather than demonstrated in people. In the head to head trial, the practical consequences of the albumin design were visible: fewer gastrointestinal adverse events than liraglutide (35.9 against 49.0 percent) and more injection site reactions (12.9 against 5.4 percent), alongside a smaller HbA1c reduction. The FDA label carries the class boxed warning about thyroid C-cell tumours seen in rodents with other GLP-1 agonists, noting that carcinogenicity of albiglutide itself could not be assessed in rodents.
Human evidence
Among the largest human records of any compound on this site: a 9,463 person cardiovascular outcome trial, a full phase 3 glycaemic programme including a randomised head to head against liraglutide in 841 people, an 814 person insulin-replacement trial, and four years of marketed use. All trials funded by GlaxoSmithKline.
- Harmony Outcomes: 9,463 people, median 1.6 years. Cardiovascular death, heart attack or stroke in 7 percent against 9 percent; hazard ratio 0.78 (95 percent CI 0.68 to 0.90); superiority P equals 0.0006. No excess pancreatitis, pancreatic cancer or thyroid cancer detected within the follow-up.
- HARMONY 7 head to head: 841 people, 32 weeks. HbA1c fell 0.78 points on albiglutide against 0.99 on liraglutide; albiglutide failed the non-inferiority test. Fewer gastrointestinal events (35.9 against 49.0 percent), more injection site reactions (12.9 against 5.4 percent).
- Insulin replacement trial: 814 people; 54 percent on weekly albiglutide stopped all mealtime insulin with equivalent HbA1c, 4.4 kg less weight and less hypoglycaemia than continuing insulin, at the cost of doubled gastrointestinal adverse events.
- Post hoc age analysis of Harmony Outcomes: point estimates favoured albiglutide in every age band; the confidence interval crossed 1 for those 65 and over and 75 and over.
- Approved on the basis of the phase 3 glycaemic programme in 2014; the FDA label noted that HbA1c reductions were smaller than with some comparators and that the drug was not recommended first-line.
What this does not tell you: Median follow-up of 1.6 years is short for a cardiovascular outcome trial and says nothing about longer horizons. The cardiovascular result is against placebo; no outcome trial compared albiglutide with another GLP-1 agonist, so the head to head evidence is glycaemic only and it favoured liraglutide. The drug is unavailable everywhere, so none of this can be acted on by a patient. Everything was sponsor funded and the sponsor stopped selling it.
Reading the research record
Albiglutide is the counterexample this site needs against a comfortable assumption: that a drug which wins a large outcome trial on hard endpoints stays on the market. It reduced cardiovascular death, heart attack and stroke by 22 percent in 9,463 people, published in the Lancet in October 2018, and its European authorisation was withdrawn the same month, at the company's request, for what the EMA records as commercial reasons. Drugs@FDA lists it as discontinued. The only reasons on record are the sponsor's own phrases in the EMA notification and two registry entries: commercial reasons, business considerations, not quality, safety or efficacy. This profile does not go beyond them.
The evidence does, however, show what the drug was up against. In its own head to head trial it lowered HbA1c less than liraglutide and failed non-inferiority; its weight effect in the phase 3 programme was smaller than that of the lipidated analogues that followed; and by 2018 dulaglutide and semaglutide offered weekly dosing with larger glycaemic and weight effects. A drug can be genuinely protective and still be the weakest member of its class on the endpoints prescribers use to choose between members. Evidence and availability are different things, and albiglutide is the cleanest demonstration in the GLP-1 literature that they can come apart.
The evidence, charted
Fig. 1 · evidence composition
6of 8 citations (75%) are in people
Counted from the citation list on this page. The Human count is the same number shown in the badge at the top. Both understate any literature larger than the sources we cite.
Fig. 2 · evidence over time
Evidence spans 6 distinct years, 2014 to 2024, counted from the citation list on this page.
Fig. 3 · legal status at a glance
US
Approved
UK
Approved
AU
Approved
CA
Not approved
Approved in 3 of 4, prescription route in 0, not approved in 1. A jurisdiction's classification is a regulatory fact, not a verdict on the science; see Legal status below for the full text and any notes.
Key studies & citations
- Human2018
Albiglutide and cardiovascular outcomes in patients with type 2 diabetes and cardiovascular disease (Harmony Outcomes): a double-blind, randomised placebo-controlled trial
9,463 adults aged 40 and over with type 2 diabetes and cardiovascular disease, 610 sites, 28 countries, randomised 1:1 to albiglutide 30 to 50 mg weekly or placebo, median follow-up 1.6 years. Primary composite (cardiovascular death, myocardial infarction, stroke): 338 of 4,731 (7 percent) against 428 of 4,732 (9 percent), hazard ratio 0.78 (95 percent CI 0.68 to 0.90), superiority P equals 0.0006. Pancreatitis 10 against 7, pancreatic cancer 6 against 5, medullary thyroid carcinoma 0 against 0. Funded by GlaxoSmithKline.
Lancet - Human2014
Once-weekly albiglutide versus once-daily liraglutide in patients with type 2 diabetes inadequately controlled on oral drugs (HARMONY 7): a randomised, open-label, multicentre, non-inferiority phase 3 study
841 adults randomised to albiglutide 30 mg weekly (titrated to 50 mg) or liraglutide titrated to 1.8 mg daily, 32 weeks, open-label. HbA1c change minus 0.78 percentage points against minus 0.99; difference 0.21 (95 percent CI 0.08 to 0.34), non-inferiority not met at the 0.3 margin (P equals 0.0846). Injection site reactions 12.9 against 5.4 percent; gastrointestinal events 35.9 against 49.0 percent. Funded by GlaxoSmithKline. A head to head result in which the comparator won on the primary endpoint.
Lancet Diabetes and Endocrinology - Human2020
Impact of a Weekly Glucagon-Like Peptide 1 Receptor Agonist, Albiglutide, on Glycemic Control and on Reducing Prandial Insulin Use in Type 2 Diabetes Inadequately Controlled on Multiple Insulin Therapy: A Randomized Trial
814 people on basal plus prandial insulin randomised to albiglutide plus glargine with lispro withdrawn (402) or continued lispro plus glargine (412), 26 weeks. HbA1c 6.7 against 6.6 percent, non-inferior; 54 percent on albiglutide stopped all prandial insulin; weight minus 2.0 against plus 2.4 kg; documented hypoglycaemia 57.2 against 75.0 percent; gastrointestinal adverse events 26 against 13 percent. Published after the drug was withdrawn. Sponsor funded.
Diabetes Care - Human2024
Effect of albiglutide on cardiovascular outcomes in older adults: A post hoc analysis of a randomized controlled trial
Post hoc analysis of Harmony Outcomes by age: hazard ratio 0.66 (95 percent CI 0.53 to 0.82) under 65 and 0.86 (0.71 to 1.04) at 65 and over (interaction P equals 0.07); 0.78 under 75 and 0.70 (0.48 to 1.01) at 75 and over. Age as a continuous variable did not modify the effect. Post hoc, so hypothesis-generating.
Diabetes, Obesity and Metabolism - Human2019
Effect of Albiglutide, When Added to Standard Blood Glucose Lowering Therapies, on Major Cardiovascular Events in Subjects With Type 2 Diabetes Mellitus (Harmony Outcomes)
9,463 participants, completed February 2018, results posted 6 March 2019. Sponsor GlaxoSmithKline.
ClinicalTrials.gov - Review2018
Eperzan (albiglutide): European Public Assessment Report, authorisation withdrawn
Marketing authorisation issued 20 March 2014 and withdrawn 29 October 2018 at the request of the holder, GlaxoSmithKline Trading Services Limited, which notified the European Commission of its decision to permanently discontinue the marketing of the product for commercial reasons. The primary regulatory record of the withdrawal.
European Medicines Agency - Review2017
TANZEUM (albiglutide) for injection, prescribing information (2017 revision)
Indicated as an adjunct to diet and exercise for glycaemic control in adults with type 2 diabetes; not recommended as first-line therapy. Recombinant fusion of two DPP-4 resistant GLP-1 copies to human albumin, molecular weight 72,970 daltons, half-life 5 days. Boxed warning for thyroid C-cell tumours by class. Drugs@FDA now lists the product as Discontinued.
U.S. Food and Drug Administration, Drugs@FDA label, BLA 125431 - Human2018
A Study to Evaluate Pharmacokinetics, Safety and Efficacy of Albiglutide in Pediatric Subjects With Type 2 Diabetes (withdrawn)
Phase 3, WITHDRAWN before enrolment. Sponsor's reason: the study did not start recruiting as albiglutide would have been withdrawn from the market prior to study end. A separate extension (NCT02750930) was terminated as a result of GSK business considerations and not due to quality, safety or efficacy concerns.
ClinicalTrials.gov
Frequently asked questions
Can I still get albiglutide?
No. GlaxoSmithKline withdrew it from every market. The European authorisation was withdrawn on 29 October 2018 at the company's request and the FDA lists Tanzeum as discontinued. It has not been available since 2018.
Why was it withdrawn?
The only reasons on the record are the company's own: the EMA notes that GlaxoSmithKline chose to permanently discontinue marketing the product for commercial reasons, and two trial registry entries cite business considerations and not quality, safety or efficacy concerns. No safety problem was identified, and its outcome trial was positive.
Did it reduce heart attacks and strokes?
Yes, in one trial. Harmony Outcomes randomised 9,463 people with type 2 diabetes and cardiovascular disease to albiglutide or placebo for a median 1.6 years. Cardiovascular death, heart attack or stroke occurred in 7 percent on albiglutide against 9 percent on placebo, a hazard ratio of 0.78. The trial was funded by the manufacturer.
How did it compare with liraglutide?
In HARMONY 7, 841 people were randomised to weekly albiglutide or daily liraglutide for 32 weeks. HbA1c fell 0.78 percentage points on albiglutide against 0.99 on liraglutide, and albiglutide did not meet the non-inferiority margin. It caused fewer gastrointestinal side effects but more injection site reactions.