Barrett's Esophagus: Adding Aspirin to PPI for Chemoprevention
A patient already on PPI therapy for confirmed non-dysplastic Barrett's esophagus asks a genuinely reasonable question after reading about it himself: does adding aspirin actually lower his risk of progression, or is the real evidence behind that idea weaker than it's often made to sound.
Walter B., 67, is three years into retirement from a career running a small hardware store, and now spends most mornings at the same diner counter reading the paper cover to cover — which is where, six weeks ago, he read a health column mentioning aspirin and Barrett's esophagus and brought the clipping to his appointment folded into quarters. He was diagnosed with non-dysplastic Barrett's esophagus four years ago on surveillance endoscopy after a decade of reflux symptoms, has been on omeprazole 40mg daily since, and his most recent surveillance biopsy eighteen months ago showed no dysplasia. He has never taken daily aspirin and has no personal history of cardiovascular disease, peptic ulcer disease, or GI bleeding.
The trial his clipping was almost certainly drawing on, AspECT, randomized Barrett's patients to combinations of high- or low-dose PPI with or without aspirin and found that high-dose PPI plus aspirin reduced a composite endpoint of death, esophageal adenocarcinoma, or high-grade dysplasia compared to low-dose PPI alone over roughly nine years of follow-up — a real, positive finding, but one built around high-dose PPI as part of the tested combination, in a trial not specifically designed to isolate aspirin's own individual contribution to that reduction. Walter is already on standard-dose PPI, not the high-dose arm the trial's clearest signal came from, and he has no separate cardiovascular indication of his own that would make aspirin's bleeding risk more clearly worth accepting on its own terms.
His clipping didn't mention AspECT's dosing arms. The trial randomized across four combinations — high- or low-dose PPI, with or without aspirin — and the clearest signal came specifically from the high-dose-PPI-plus-aspirin arm compared against low-dose PPI alone, not from aspirin added to a standard dose the way a casual reading of the headline result might suggest. That distinction is the entire reason this conversation is happening at all rather than a simple yes-or-no answer to what he read.
Reading the trial he brought in himself
Walter's clipping is almost certainly pointing at AspECT, and it's a real, positive trial — high-dose PPI plus aspirin reduced a composite endpoint of death, esophageal adenocarcinoma, or high-grade dysplasia compared to low-dose PPI alone, over about nine years of follow-up in Barrett's patients. If we're going to act on that evidence, I'd want to match the arm that actually showed the benefit: step his PPI up to the high dose and add low-dose aspirin, rather than add aspirin on its own to what he's already taking.
I want to be precise about what that trial can and can't tell us. It tested a combination — high-dose PPI plus aspirin against low-dose PPI alone — and wasn't designed to isolate how much of the benefit came from the PPI dose increase versus the aspirin itself. Walter has no cardiovascular indication of his own for aspirin. Adding a bleeding-risk drug on the strength of an effect that might be mostly the PPI dose is a real gap between what the trial showed and what we'd be doing.
I don't think that makes the trial worthless, but it does mean 'add aspirin because of AspECT' overstates what AspECT actually isolated.
I'd lean toward the PPI dose increase without aspirin, at least for now. It captures the part of the trial's finding we can attribute with more confidence, and it avoids adding a new bleeding-risk medication to a 67-year-old with no independent indication for it. If Walter, having heard the actual trial design rather than the headline, still wants aspirin added, that's a reasonable informed choice — but I don't think we should present it as a settled recommendation.
Agreed: PPI dose increased to omeprazole 40mg twice daily, the part of AspECT's finding the team could attribute with the most confidence. Aspirin was not recommended as a standing default, but was explained in full — including the trial's combination design and its real limits — and Walter chose to start it himself, informed rather than persuaded.
Not agreed: whether AspECT's finding should be presented to future patients as "add aspirin for Barrett's" or with the fuller caveat the pharmacologist insisted on. That disagreement wasn't settled at this visit and stays a live one for how the group counsels the next patient who brings in the same clipping.