Eosinophilic Esophagitis: Dupilumab Before or After the PPI Trial
A newly diagnosed patient with dense eosinophilic esophagitis and a heavy atopic history raises a genuine sequencing question the 2025 ACG guideline doesn't fully close: does her atopic burden justify starting on dupilumab, or does the stepwise pathway still call for a documented PPI trial first.
Maya T., 27, works as a freelance graphic designer and got engaged this spring; most of the wedding planning has happened around a habit she never used to think about — cutting every bite small and drinking water between them so nothing catches. She has had seasonal allergic rhinitis since childhood and mild persistent asthma diagnosed at 19, both controlled on an intranasal steroid and a low-dose inhaled combination, neither ever requiring an emergency visit. Eight months ago a piece of steak lodged in her chest long enough that a friend drove her to an emergency department; endoscopy that night showed the trachealization and linear furrowing of eosinophilic esophagitis, and biopsies confirmed it — 48 eosinophils/hpf proximally, 52 distally, well past the 15-eos diagnostic threshold.
Her peripheral eosinophil count returned at 620 cells/µL and total IgE at 340 IU/mL, a genuinely atopic-heavy profile. The 2025 ACG clinical guideline (Dellon and colleagues) suggests dupilumab for patients twelve and older who are non-responsive to PPI therapy, and its single early-use consideration is written narrowly — for moderate-to-severe asthma or eczema. Her asthma is mild persistent and well-controlled on a low-dose combination inhaler, she has no eczema, and allergic rhinitis appears nowhere in that carve-out. Her atopic burden is real, but on the guideline's own terms it does not reach the threshold that would license starting a biologic ahead of a PPI trial she has never had.
LIBERTY EoE TREET, the phase 3 trial behind dupilumab's approval, enrolled only patients still showing at least 15 eosinophils/hpf after eight weeks of high-dose PPI — every participant was a PPI non-responder by entry criterion, carried a mean of five years of disease, and roughly 40% had already required dilation. Maya is eight months from diagnosis, has taken no acid suppression at all, and has never been dilated. The trial that proved the drug works has nothing to say about a patient who has not yet done the thing every one of its participants had already failed.
Deciding where to start
I'd start dupilumab now, not after eight more weeks of dysphagia and food avoidance. Her asthma, her rhinitis, and an IgE of 340 aren't incidental — that's a systemic Th2 phenotype, and the esophagus is one organ in it. LIBERTY EoE TREET showed real histologic and symptomatic response on weekly dupilumab across 24 and 52 weeks, and the guideline does allow considering it early when atopic disease is prominent; I'd rather not make her wait through a step least likely to work in exactly her phenotype.
I read the same guideline, and its early-use language is narrower than you're using it. Dellon's 2025 ACG panel suggests dupilumab for patients non-responsive to PPI; the one early-use consideration names moderate-to-severe asthma or eczema. Maya's asthma is mild persistent and controlled, she has no eczema, and rhinitis was never in that carve-out — she doesn't meet it.
A meaningful minority of EoE patients, including plenty with real atopic comorbidity, achieve histologic remission on PPI alone.
'Systemic Th2 phenotype' is a real biological claim, but it's not a validated predictor of PPI non-response. And every patient in the trial you're citing had already failed eight weeks of high-dose PPI before randomization — that trial cannot tell us what dupilumab does in someone who hasn't taken any.
Both of you are arguing about which is more likely to work, and neither of you can actually know that without treating her. Eight weeks of PPI therapy is inexpensive, well-tolerated, and answers the empirical question directly — if she doesn't respond, that failure itself satisfies the guideline's refractory-disease criterion and clears dupilumab's pathway with nothing lost but time she's already been managing symptomatically for months. I'd run the trial, but pre-authorize the biologic now so there's no second delay if it fails.
Agreed: an 8-week omeprazole trial starts today, with a scheduled repeat endoscopy and biopsies at the end of it rather than a symptom-only reassessment, since histologic response is what actually adjudicates this. Prior authorization for dupilumab is submitted in parallel so a non-response doesn't cost a second waiting period.
Not agreed: the allergist's read that her atopic burden alone should have skipped the PPI step entirely. That question doesn't resolve today — it resolves in eight weeks, against her own biopsy result, and everyone at the table said they'd accept that answer either way.