Eosinophilic Gastroenteritis Without a Trial to Guide Any of It
A single patient with a rare, under-studied diagnosis and a family history that makes her want to avoid the one option with the deepest evidence base. The disagreement is how to weigh a well-worn off-label drug against a barely-tested newer one when neither has ever been through a real trial for this disease.
Alicia F. has spent eight months explaining away a steadily worsening list of symptoms — bloating she blamed on stress, abdominal pain she attributed to a bad diet, nine pounds she assumed she'd simply been too busy to notice losing — until an endoscopy ordered to investigate the weight loss found marked eosinophilic infiltration of her stomach and duodenum, well outside the esophagus this time, and a low albumin confirming she's been losing protein through an inflamed gut rather than just eating less of it. Eosinophilic gastroenteritis, unlike its esophageal cousin, has no FDA-approved treatment of any kind and a genuinely thin evidence base overall — off-label systemic corticosteroids remain the most established option by a wide margin, while Arakawa and colleagues' systematic review of biologics targeting the same type 2 pathway found only case-report and small case-series support across five different agents, not a single controlled trial for any of them.
Alicia has a specific reason to want to avoid the best-established option. Her mother developed avascular necrosis of the hip after a prolonged course of steroids for an unrelated condition, and watching that recovery has left Alicia asking, before any prescription is written, whether there's a version of this treatment plan that doesn't involve steroids at all. The honest answer her team has to give her is that the evidentiary gap between corticosteroids and a biologic like dupilumab here isn't the gap between good evidence and no evidence — it's the gap between decades of off-label case-series experience and a handful of individually reported cases, a difference of degree rather than kind in a disease where nothing has actually been through a controlled trial.
Her mother's actual outcome adds real weight to what could otherwise read as an abstract worry: the avascular necrosis wasn't a minor complication, it required a hip replacement in her mid-fifties and left her with a lasting limp Alicia grew up watching. That's not the same as saying Alicia herself would develop the same complication — dose, duration, and individual susceptibility all vary — but it explains why she's asking the question before a single tablet is prescribed, rather than after a course has already started, and why the team is treating her preference as a serious input rather than an anxiety to be talked down from.
No trial for anything, and an active protein loss to manage today
I'd start prednisone today. Her albumin is 2.6 and she's lost nine pounds she didn't mean to lose — this is an active process, and steroids are the option with the deepest track record in this disease by a wide margin, off-label status aside.
I want to take her family history seriously, not as a soft preference but as a real, witnessed harm she's asking us to weigh. The honest evidence picture here is that dupilumab's case-report support and off-label steroid use are both built on the same kind of evidence — uncontrolled case series — just different denominator sizes.
I hear the urgency in her labs. But "more established" isn’t the same claim as "controlled-trial-proven" here, and I don’t think that distinction should disappear just because steroids have been used off-label longer.
I don't think this has to be one drug making both decisions. Her active protein loss is a today-problem, and steroids are the fastest, most established way to address it — that part of the gastroenterologist's argument is right regardless of the longer-term drug question.
But that doesn't mean steroids become her open-ended plan. I'd commit now, at this visit, to a defined dupilumab trial once she's stabilized — taking her preference seriously as the maintenance answer, without asking a single drug to solve both the urgent problem and the long-term one at once.
Agreed: start a tapering prednisone course now to address her active protein loss, with a planned transition to dupilumab as maintenance therapy once she's stabilized and the steroid taper is underway.
Not agreed, and left explicit:
The gastroenterologist would return to a longer, lower-dose steroid maintenance course, arguing the established option shouldn't be permanently abandoned on the strength of case-report-level biologic evidence alone.
The allergist would try a different biologic target before returning to steroids, arguing that one agent's failure in a disease this poorly characterized doesn't indict the whole class.
For now, both the urgent problem and her stated long-term preference have a plan. Which plan wins if the transition doesn't go smoothly is acknowledged as unsettled, not resolved by today's sequencing.