First-Line Eosinophilic Esophagitis: A Guideline Written Before One of Its Options Existed
A single patient, newly diagnosed after a food-impaction visit. The disagreement isn't about which therapy works — it's about how much weight to put on a guideline that was finished two years before one of the three options it should be weighing was ever approved.
Priya S., a 26-year-old woman finishing a PhD in molecular biology, spent three years quietly restructuring her diet around a problem she never named: she cut bread into small pieces, avoided chicken breast unless it was drowned in sauce, and always kept a glass of water at hand during meals. None of it seemed strange to her until a chicken sandwich lodged in her esophagus at a lab retreat and required endoscopic disimpaction that evening. The scope that freed the sandwich also found linear furrows, mucosal rings, and whitish exudates running most of the esophagus's length, and biopsies came back at 42 eosinophils per high-power field — more than double the 15/hpf threshold that separates a normal esophagus from one worth calling eosinophilic esophagitis, and consistent with the years of slow, careful eating she'd never mentioned to anyone as a symptom.
The actual disagreement in her case isn't about whether treatment is warranted — it clearly is — but about which of three real options to start with, and how much a specific document should be allowed to settle that question. The 2020 AGA/Joint Task Force guideline treats dietary elimination, topical corticosteroids, and proton pump inhibitors as roughly interchangeable first-line choices, each carrying a conditional recommendation rather than a ranked preference. That guideline was finalized before dupilumab's 2022 approval for EoE, so a literal reading of it doesn't rank dupilumab against the other three — it simply doesn't contain the option. Priya's team has to decide how much weight a guideline still deserves on the exact question it was never actually asked.
There's also a second, more mundane fact shaping the decision that has nothing to do with evidence at all: her insurance is a high-deductible plan tied to her graduate stipend, and an injectable biologic carries a cost none of the other two options come close to before any prior-authorization process even starts. That isn't a reason to rule dupilumab out on principle, but it does mean a treatment-naive patient's actual first move is shaped as much by what she can afford to try while waiting on an appeal as by what the literature says should come first — a constraint the guideline, written for a hypothetical patient with hypothetical coverage, has nothing to say about either.
At diagnosis, choosing where to start
Start with the six-food elimination diet. Lucendo's group put empiric SFED's histologic remission rate at roughly 70% in adults, the strongest single number belonging to any of the three options in front of us, and it's the only one that asks nothing pharmacologic of her at all.
If she’d already failed a diet trial elsewhere, I’d be having a different conversation — this is specifically about what a treatment-naive patient should try first.I don't dispute Lucendo's number, but that 70% assumes she completes the protocol, and completing it means an endoscopy after every food goes back in — up to six separate procedures to find her actual triggers. She's finishing a dissertation on someone else's grant timeline.
The efficacy argument is real, but it’s measuring people who stayed on the diet long enough to reintroduce foods in sequence. That’s not a guarantee for every patient who starts it, and I’d rather start her on something her actual life can sustain.
There's a third framing neither of you has used yet: dupilumab isn't absent from the 2020 guideline because it was tested and found wanting — it's absent because it didn't exist when the guideline was written. Reading that silence as a demotion is a mistake most people make with any two-year-old guideline in a field moving this fast.
That said, I'm not arguing she should start there. She's treatment-naive, and starting with the newest, most expensive, injectable option before trying either of two established oral or dietary approaches isn't proportionate to what we actually know about her yet — it's proportionate to what dupilumab's own trial covered, which was patients who'd already been through less-invasive options first.
Agreed: start budesonide oral suspension, with a repeat endoscopy and biopsy at 12 weeks to confirm histologic response before deciding on maintenance.
Not agreed, and worth stating plainly rather than smoothing over:
The gastroenterologist would move to dupilumab next, given her impaction history and the argument that a second treatment-naive attempt at diet after one drug failure adds delay without changing the underlying adherence risk.
The allergist would still want the six-food elimination diet tried once, arguing that a single steroid failure doesn't tell you whether a drug-free approach would have worked, and that her circumstances — dissertation deadline aside — may look different in three months.
Nobody treated this as resolved by assigning a tiebreaker. The plan for now is budesonide; the plan for a non-response is explicitly unsettled, carried forward to the 12-week visit rather than decided today on a hypothetical.