Randomised controlled trialindependent cross-checkfull text checked★ key study I-PPI-09 in-depth review
Barrett's oesophagus: high-dose PPI improves outcomes
Esomeprazole and aspirin in Barrett's oesophagus (AspECT): a randomised factorial trial.
Overview
Jankowski JAZ et al.. Lancet 2018 · PMID 30057104 · DOI 10.1016/S0140-6736(18)31388-6
Over about 9 years, high-dose esomeprazole improved a composite of death, cancer and severe dysplasia.
Does the study support the specific claim? No · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Barrett's chemoprevention.
- How was the study done?
RCT.
- How many people, animals or samples were studied?
2,557.
- What was actually found?
Benefit for high dose.
- How large was the observed effect?
TR 1.27; NNT 34.
- What do the results mean in everyday life?
PPIs help people with Barrett's.
- What are the study's limitations?
No placebo; open label.
- What does it say about Elawady's specific claim?
Contradicts universal ban.
- Do the results apply to Germany and Europe?
Applicable.
Scientific in-depth review
1. Full reference
- Authors
- Jankowski JAZ et al.
- Journal
- Lancet
- Year
- 2018
- Identifier
- PMID 30057104 DOI 10.1016/S0140-6736(18)31388-6
- Verification
- verified – PubMed efetch 2026-10-25
- Retracted
- no
- Corrections
ErratumIn: Lancet. 2018 Dec 15;392(10164):2552. doi: 10.1016/S0140-6736(18)32970-2.
- Extent checked
- full text
- Conflicts of interest and funding
Cancer Research UK (main), with an educational grant and free esomeprazole from AstraZeneca (Investigator-Sponsored Study Program); Wellcome Trust and NIHR/HTA support; sponsored by University of Oxford; funders had no role in design, analysis or writing. Several authors declare industry relationships.
2. Research question
Do high-dose PPI and aspirin improve outcomes in Barrett's oesophagus?
3. Study design and population
- Study design
- Randomised controlled trial – Phase 3, randomised, open-label (masked pathologists) 2x2 factorial trial (AspECT; EudraCT 2004-003836-77); comparison of two PPI doses, no PPI-free arm
- Population
Adults with histologically proven Barrett's oesophagus >=1 cm without high-grade dysplasia or adenocarcinoma and not on NSAIDs; ~80% men (women capped at ~500); median age 58-59; predominantly white
- Country
United Kingdom (84 centres) and Canada (1 centre)
- Period
Recruitment 10 March 2005 - 1 March 2009; follow-up to 1 March 2017; median 8.9 years (IQR 8.2-9.8); 20,095 patient-years
- Sample size
2557 randomised (PPI comparison 2535 analysed: 1270 high-dose, 1265 low-dose; aspirin comparison 2280); 313 primary events; 99.9% data completeness
- Inclusion and exclusion criteria
Barrett's >=1 cm; excluded prior adenocarcinoma/HGD and baseline NSAID use; 255 aspirin-intolerant/already-on-aspirin patients entered PPI randomisation only
4. Methods and endpoints
- Exposure / intervention
Esomeprazole 40 mg twice daily (high dose), with or without aspirin 300 mg (UK)/325 mg (Canada) daily
- Comparator
Esomeprazole 20 mg once daily (low dose), with or without aspirin
- Endpoints
Composite: time to all-cause death, oesophageal adenocarcinoma or high-grade dysplasia; components; serious adverse events
- Statistical methods
ITT accelerated failure time models adjusted for minimisation factors (time ratio >1 = benefit); Cox models in appendix; no multiplicity adjustment; sample size reduced in 2008 from 5000 to 2224 with longer follow-up
5. Actual results
- Results
High vs low-dose PPI: 139/1270 vs 174/1265 events; TR 1.27 (95% CI 1.01-1.58), P=0.038; NNT 34 (18-333). Components (high vs low PPI): all-cause death 79 vs 105, TR 1.36 (1.01-1.82); adenocarcinoma 40 vs 41, TR 1.04 (0.67-1.61); high-grade dysplasia 44 vs 59, TR 1.36 (0.92-2.02). Aspirin vs none: TR 1.24 (0.98-1.57), P=0.068; 1.29 (1.01-1.66) when censoring NSAID users. High-dose PPI + aspirin vs low-dose PPI alone: TR 1.59 (1.14-2.23). Serious adverse events related to treatment grade 3-5 in 28 (1%) participants (13 attributed to PPI, 15 to aspirin).
- Absolute risks
Composite events 10.9% (high-dose) vs 13.8% (low-dose PPI) over ~9 years; NNT 34 for high- vs low-dose PPI
- Authors' conclusion
High-dose PPI and aspirin, especially combined, significantly and safely improved outcomes in Barrett's oesophagus; authors acknowledge wide CIs with lower limits near 1, suggesting results are not robust.
6. Strengths
- Large, long-term randomised trial with near-complete follow-up
- Hard endpoints (death, cancer) and masked, double-reported pathology
- Systematic endoscopic surveillance per protocol
7. Weaknesses
- Open-label; no placebo or no-PPI arm (compares doses only)
- Benefit driven by all-cause mortality and dysplasia; no effect on adenocarcinoma itself
- Borderline significance; authors note results may not be robust
- Mid-trial sample-size reduction informed by external data
- No multiplicity adjustment; ~80% men, mostly white
- Partial industry support (AstraZeneca)
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
- some concerns
- Rationale
Single-reviewer structured RoB 2 judgement from full text; not formal (needs two independent reviewers and protocol).
| Domain | Judgement | Rationale |
|---|---|---|
| D1 Randomisation process | low | Central computer-generated minimisation with random element; concealed allocation; balanced baseline. |
| D2 Deviations from intended interventions | some concerns | Open-label; possible differential co-interventions (e.g. OTC NSAIDs/aspirin); ITT analysis. |
| D3 Missing outcome data | low | 99.9% data return. |
| D4 Measurement of the outcome | low | Death objective; pathology masked and double reported with central audit. |
| D5 Selection of the reported result | some concerns | Registered with prespecified composite, but sample-size amendment mid-trial and borderline results; NSAID-censored aspirin analysis emphasised. |
Full-text review (25 October 2026): Added component results (death TR 1.36; adenocarcinoma TR 1.04 null; HGD 1.36 NS), absolute event proportions, safety details, sample-size amendment, dose-comparison design (no PPI-free arm), funding structure; RoB 2 domains (overall unchanged: some concerns).
9. Assessment of the statistical interpretation
Modest borderline benefit of high- vs low-dose PPI on a composite driven by mortality and dysplasia, with no difference in adenocarcinoma. Shows long-term high-dose PPI was well tolerated (1% treatment-related severe SAEs) over ~9 years, but cannot estimate PPI vs no PPI.
10. Relevance to the specific claim
11. Independent contrary findings
12. Transferability to Germany and Europe
UK data; high.
13. Overall methodological judgement
Supports continued long-term (even high-dose) PPI therapy in Barrett's oesophagus as safe and modestly beneficial versus low dose; it does not directly test stopping PPIs.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | moderate |
| Directness | direct |
| Reproducibility | not replicated |
| Consistency with the overall evidence | unknown |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
PPI vs no PPI in Barrett's; confirmation of the mortality benefit; generalisability to women and non-white populations.
15. Linked fact checks
- All acid-suppressing medicines (PPIs and H2 blockers) are permanently forbidden; heartburn should stop through the diet, not through drugs. – Studies contradicting or limiting the claim