Liraglutide heart trial

Overview

Marso SP et al.. N Engl J Med 2016 · PMID 27295427 · DOI 10.1056/NEJMoa1603827

Fewer heart deaths and deaths overall; pancreatitis not increased, but more gallstone disease and a small, non-significant excess of pancreatic cancer.

Does the study support the specific claim? No · Methodological quality: high

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Marso SP et al.

Journal

N Engl J Med

Year

2016

Identifier

PMID 27295427 DOI 10.1056/NEJMoa1603827

Verification

verified – PubMed efetch 2026-10-25 (this agent)

Retracted

no

Corrections

none

Extent checked

full text

Conflicts of interest and funding

Funded by Novo Nordisk (manufacturer) and NIH grants. Steering committee of 11 academics and 4 sponsor employees; data analysed by Statogen Consulting and the sponsor; sponsor-funded medical writer; four authors are Novo Nordisk employees/shareholders with liraglutide-related patents; most academic authors report industry fees.

2. Research question

CV outcomes of liraglutide

3. Study design and population

Study design

Randomised controlled trial – Randomised, double-blind, placebo-controlled multicentre cardiovascular outcome trial (LEADER, NCT01179048) with 2-week placebo run-in; non-inferiority then superiority hierarchical testing

Population

Type 2 diabetes, HbA1c >=7.0%, age >=50 with established CVD/CKD stage >=3/heart failure NYHA II-III, or age >=60 with >=1 CV risk factor; 81.3% had established CVD and/or CKD; mean diabetes duration 12.8 years; mean HbA1c 8.7%

Country

410 sites in 32 countries

Period

Randomisation September 2010 to April 2012; closeout August 2014 to December 2015; median follow-up 3.8 years (median exposure 3.5 years)

Sample size

9,340 randomised (4,668 liraglutide, 4,672 placebo)

Inclusion and exclusion criteria

Excluded: type 1 diabetes; use of GLP-1 RA, DPP-4 inhibitor, pramlintide or rapid-acting insulin; personal/family history of MEN2 or medullary thyroid cancer; acute coronary or cerebrovascular event within 14 days.

4. Methods and endpoints

Exposure / intervention

Liraglutide 1.8 mg (or maximum tolerated dose) subcutaneously once daily plus standard care (median dose 1.78 mg)

Comparator

Placebo on top of standard care

Endpoints

Primary: first CV death, nonfatal (incl. silent) MI or nonfatal stroke. Exploratory (no multiplicity adjustment): expanded CV composite, all-cause death, renal/retinal microvascular composite, neoplasms, pancreatitis (all adjudicated blinded).

Statistical methods

Cox model, ITT; non-inferiority margin HR 1.30 then superiority; sensitivity analyses incl. per-protocol; no multiplicity adjustment for exploratory outcomes; adverse events compared with chi-square tests.

5. Actual results

Results

Primary 608/4668 (13.0%) vs 694/4672 (14.9%), HR 0.87 (95% CI 0.78-0.97); CV death 4.7% vs 6.0%, HR 0.78 (0.66-0.93); all-cause death 8.2% vs 9.6%, HR 0.85 (0.74-0.97); microvascular composite HR 0.84 (0.73-0.97), nephropathy HR 0.78 (0.67-0.92); retinopathy HR 1.15 (0.87-1.52). Safety: acute pancreatitis 18 (0.4%) vs 23 (0.5%); chronic pancreatitis 0 vs 2; pancreatic carcinoma 13 (0.3%) vs 5 (0.1%), p=0.06; acute gallstone disease 145 (3.1%) vs 90 (1.9%), p<0.001; acute cholecystitis 0.8% vs 0.4%; any malignant neoplasm 6.3% vs 6.0%; medullary thyroid carcinoma 0 vs 1; severe hypoglycaemia 2.4% vs 3.3% (rate ratio 0.69, 0.51-0.93); discontinuation for AE 9.5% vs 7.3% (mainly nausea, vomiting, diarrhoea); heart rate +3.0 bpm; weight -2.3 kg; HbA1c -0.40 points at 36 months.

Absolute risks

NNT over 3 years: 66 for the primary outcome and 98 for all-cause death (authors' estimate). All-cause death 8.2% vs 9.6% (-1.4 percentage points). Pancreatitis 1.5 episodes per 1000 patient-years in both arms combined.

Authors' conclusion

Lower rate of primary outcome.

6. Strengths

• Large double-blind RCT with run-in, stratified randomisation and matching placebo
• Blinded external adjudication of CV events, pancreatitis and neoplasms
• Vital status 99.7%; 96.8% completed final visit, died or had primary outcome
• Pre-specified hierarchical testing for primary outcome; sensitivity analyses consistent

7. Weaknesses

• Industry sponsored; sponsor co-analysed data and four authors are sponsor employees with patents
• High-risk population (HbA1c >=7%, mostly established CVD); limited applicability to lower-risk patients
• Exploratory outcomes, including all-cause death and safety outcomes, not adjusted for multiplicity
• Trial not powered for rare harms such as pancreatic cancer or pancreatitis
• Follow-up limited to 3.5-5 years

8. Risk of bias

Orienting tool

RoB 2

Type of assessment

structured assessment based on the full text by a single reviewer (no formal double review)

Full formal assessment

no

Overall risk of bias

low

Rationale

Unchanged (low) for the primary outcome; exploratory and safety outcomes carry multiplicity and low-power limitations. Strong sponsor involvement in analysis noted. Single-reviewer structured judgement; a formal RoB 2 needs two independent reviewers.

Full-text review (25 October 2026): Added eligibility, design details, full safety table (pancreatitis 18 vs 23, pancreatic carcinoma 13 vs 5 p=0.06, gallstones 3.1% vs 1.9%, GI discontinuation), NNTs, sponsor roles/patents. Bias unchanged (low). Relevance refined for C-ODD-04/C-ODD-05.

9. Assessment of the statistical interpretation

Modest relative (13%) and absolute (1.9 percentage points) reduction in MACE with adequate precision. Pancreatitis was not increased (18 vs 23), but numbers are small; the pancreatic cancer imbalance (13 vs 5, p=0.06) is imprecise and neither confirms nor excludes a risk, as the authors state. Gallstone disease was clearly increased.

10. Relevance to the specific claim

Contradicts C-ODD-06 (net harm). For C-ODD-04: pancreatitis not increased (18 vs 23) but pancreatic carcinoma numerically higher (13 vs 5, p=0.06); cysts not studied. Supports part of C-ODD-05 (documented GLP-1 RA harm signals: gallstone disease, GI discontinuation).

11. Independent contrary findings

Adverse effects listed in archive studies.

12. Transferability to Germany and Europe

Multinational trial including European sites or directly applicable to EU practice.

13. Overall methodological judgement

High-quality evidence that liraglutide lowers cardiovascular events and death in high-risk T2DM. It does not show increased pancreatitis, but it documents increased gallbladder disease and GI intolerance and a non-significant pancreatic cancer imbalance; it cannot assess pancreatic cysts.

14. Open questions

Pancreatic cancer and pancreatitis risk with longer exposure (requires pooled/pharmacovigilance data); retinopathy signal; applicability to lower-risk and non-diabetic (obesity) users; pancreatic cysts were not studied.

15. Linked fact checks

• Injectable diabetes and weight-loss drugs (GLP-1 receptor agonists) can trigger pancreatitis and cysts. – Studies with mixed or unclear results
• Every major class of glucose-lowering drug (metformin, sulfonylureas, thiazolidinediones, DPP-4 inhibitors, SGLT2 inhibitors, GLP-1 agonists) carries a documented signal of serious harm in the peer-reviewed literature. – Studies with mixed or unclear results
• Glucose-lowering drugs are net harmful and belong among the 'forbidden' items; dapagliflozin (Forxiga) is among the 'dirtiest' drugs and kills the brain; leg amputations happen only to people taking diabetes drugs and insulin. – Studies contradicting or limiting the claim