Barrett’s esophagus is a metaplastic change of the esophageal epithelium which increases the risk of esophageal adenocarcinoma by 10 to 30 fold. It is thought to progress via the development of dysplasia to esophageal adenocarcinoma (EAC). EAC is a lethal malignancy with poor outcomes when diagnosed after the onset of symptoms despite recommendations for screening and surveillance to detect the presence of Barrett’s esophagus and dysplasia. Theoretically, screening followed by surveillance should provide the opportunity to both diagnose EAC early (and treat it endoscopically with excellent outcomes) and/or prevent EAC by the endoscopic treatment of dysplasia. Unfortunately, most EAC continues to be diagnosed outside of screening and surveillance programs. The uptake of Barrett’s esophagus screening remains low and surveillance is beset with several challenges.1
Chemoprevention of EAC in those with Barrett’s esophagus appears attractive due to several reasons. Given the presumed theoretical stepwise progression of Barrett’s esophagus to EAC, there is likely a time window in which those with Barrett’s esophagus can be identified and placed on a chemopreventive agent to reduce EAC incidence. Several agents have been investigated in this regard with the most common drug being studied being aspirin. From a pathophysiologic basis, aspirin makes excellent sense given its ability to inhibit the COX 1 and 2 dependent proinflammatory and neoplastic pathways that are known to mediate carcinogenesis in Barrett’s esophagus.
However, there are several weaknesses in this theoretical argument. First, the rate of progression in those with non-dysplastic Barrett’s esophagus (80% or more in population based studies) is extremely low (0.3% per year).2 Hence in addition to only a small proportion of patients who could benefit from this approach, this makes the development and testing of any chemopreventive agent challenging. Second, there is evidence to suggest that most (80%) cases of EAC or high grade dysplasia (HGD) in Barrett’s esophagus are diagnosed within the first two years of diagnosis with only the remaining 20% being diagnosed over the next several years during surveillance (truly incident).3 This would further weaken the case for chemoprevention which is based on preventing incident HGD or EAC. An alternative approach would be to identify those at higher risk of progression and focus chemoprevention on this subset. However the identification of those who may progress remains a work in progress with clinical and biomarker variables being identified and validated but not yet ready for clinical application.
The lack of conclusive evidence on aspirin being a chemopreventive agent in those with Barrett’s esophagus to reduce the incidence of EAC is perhaps the biggest obstacle to recommending it in everyone with Barrett’s esophagus. The strongest evidence on the ability of COX inhibition to reduce EAC comes only from animal studies. Human retrospective studies have shown a reduction in the incidence of EAC in those consuming aspirin. In a pooled analysis of samples from the BEACON consortium, aspirin and NSAID use was associated with a 32% lower risk of EAC compared to controls. A meta-analysis of nine observational studies also reported a 36% reduced risk of EAC in those with BE.4
The lack of conclusive evidence on aspirin being a chemopreventive agent in those with Barrett’s esophagus to reduce the incidence of EAC is perhaps the biggest obstacle to recommending it in everyone with Barrett’s esophagus.
However, the evidence of chemoprevention in Barrett’s esophagus in prospective trials is much weaker. A prior Phase IIb randomized clinical trial (RCT) of a COX2 inhibitor (Celecoxib 200 mg/day) did not show a significant effect on the primary endpoint (change in the proportion of biopsy samples with dysplasia or carcinoma) or secondary endpoints (tissue prostaglandin levels) after a treatment duration of 48 weeks.5
Most importantly, the results of the recent AspECT RCT were also not supportive of the use of aspirin in the chemoprevention of EAC in Barrett’s esophagus.6 In this trial 2,557 Barrett’s esophagus patients were randomized in a two-by-two factorial design to high or low dose proton pump inhibitor (PPI) with or without aspirin. Patients were followed for a mean of 8.9 years. The primary outcome was a composite outcome of all-cause mortality, HGD or EAC. High dose PPI was superior to low dose PPI for preventing the composite primary outcome. Most of the survival advantage was however driven by all-cause mortality. High dose PPI was superior to low dose PPI in delaying the onset of EAC or HGD. Unfortunately, aspirin use was not associated with a decreased risk of EAC or HGD (combined outcome) and neither was high or low dose PPI, in a per protocol analysis. It was only in a secondary analysis, when patients who had used NSAIDs were censored at the time of their first use, the use of aspirin showed benefit for the primary outcome. Hence the evidence base for the use of aspirin in the chemoprevention of EAC in Barrett’s esophagus is weak at best.
An important point that cannot be ignored is the potential toxicity of aspirin. Despite the relatively low adverse event rates in the AspECT trial, low (81 mg/day) and regular dose (300-325 mg/day), aspirin therapy is associated with not negligible risks of bleeding (GI and non-GI) as well as small bowel ulcerations and strictures. In three recent randomized trials of low dose aspirin versus placebo (ASPREE, ASCEND and ARRIVE), the risk of major bleeding (GI and cerebrovascular) doubled without substantial effects on cardiovascular events (0.97% in the aspirin group and 0.46% in the placebo group in the ARRIVE trial).7
Hence taken together, aspirin cannot be recommended for chemoprevention in all Barrett’s esophagus patients at this time. Its use in those who otherwise meet criteria for secondary prevention of cardiovascular disease and are at a higher risk of neoplastic progression (such as older age men, ex- or current smokers and those with longer Barrett’s esophagus segments) may be considered on an individualized basis.
- Patients with Barrett’s esophagus without dysplasia have low risk of progression and COX inhibition with aspirin to prevent progression is theoretically intriguing.
- Retrospective studies seem to support the potential of EAC chemoprevention with aspirin use.
- However, evidence of the efficacy of COX inhibition in prospective studies is weak. The large AspECT trial did not show evidence of reduction in the incidence of EAC or HGD with the use of 300 mg/day over a mean follow up of almost nine years.
- The risk of serious bleeding (gastrointestinal and cerebrovascular) is roughly doubled in patients on low dose aspiring in large community based primary prevention RCTs.
Dr. Iyer has received research funding from Exact Sciences, C2 Therapeutics (Pentax Medical), Medtronic and Nine Point Medical. He as consulted for Medtronic, C2 Therapeutics (Pentax Medical) and CSA Medical.
1. Shaheen NJ, Falk GW, Iyer PG, et al. ACG Clinical Guideline: Diagnosis and Management of Barrett’s Esophagus. Am J Gastroenterol. 2016;111:30-50; quiz 51.
2. Desai TK, Krishnan K, Samala N, et al. The incidence of oesophageal adenocarcinoma in non-dysplastic Barrett’s oesophagus: a meta-analysis. Gut. 2011.
3. Visrodia K, Singh S, Krishnamoorthi R, et al. Systematic review with meta-analysis: prevalent vs. incident oesophageal adenocarcinoma and high-grade dysplasia in Barrett’s oesophagus. Aliment Pharmacol Ther. 2016;44:775-84.
4. Zhang S, Zhang XQ, Ding XW, et al. Cyclooxygenase inhibitors use is associated with reduced risk of esophageal adenocarcinoma in patients with Barrett’s esophagus: a meta-analysis. Br J Cancer. 2014;110:2378-88.
5. Heath EI, Canto MI, Piantadosi S, et al. Secondary chemoprevention of Barrett’s esophagus with celecoxib: results of a randomized trial. Journal of the National Cancer Institute. 2007;99:545-57.
6. Jankowski JAZ, de Caestecker J, Love SB, et al. Esomeprazole and aspirin in Barrett’s oesophagus (AspECT): a randomised factorial trial. Lancet. 2018;392:400-408.
7. Patrono C, Baigent C. Role of aspirin in primary prevention of cardiovascular disease. Nat Rev Cardiol. 2019;16:675-686.