Refractory Celiac Disease, Type I: Which Drug Buys the Remission, and What Holds It
A single patient, a year into a gluten-free diet that should have worked and hasn’t. The disagreement isn’t just which drug induces remission faster — it’s whether either plan on the table actually accounts for what happens after remission is reached.
Colleen B., a 58-year-old church choir director, has led the same forty-voice ensemble through two Christmas Eve services and a diocesan competition since her celiac diagnosis fourteen months ago, and has followed her dietitian's gluten-free plan with the same discipline she brings to rehearsal attendance — label-checking, a dedicated kitchen shelf, no cross-contamination shortcuts she’s aware of. None of it has resolved her symptoms. She still runs three to four loose stools most days, has lost eleven pounds she didn’t have to spare, and a repeat duodenal biopsy at month twelve, done specifically because a strict diet should have produced histologic improvement by then and hadn’t, still showed Marsh IIIb villous atrophy. Flow cytometry on that same biopsy found a normal intraepithelial lymphocyte population — no aberrant clonal expansion — confirming refractory celiac disease Type I rather than the more ominous Type II.
That Type I confirmation is the fact her whole care team is reading in slightly different directions. It rules out the lymphoma-adjacent surveillance intensity Type II would require, but it doesn’t settle how urgently to treat, and her nutritional trend has kept moving in the wrong direction while everyone confirmed the diagnosis: albumin down from 3.6 to 3.1 g/dL over the past two months, hemoglobin drifting to 10.8. Open-capsule budesonide has real, disease-specific evidence in exactly this population — Brar and colleagues reported both symptomatic and histologic response to open-capsule budesonide in refractory celiac disease specifically, not a finding borrowed from IBD or extrapolated from systemic steroid data — with its high first-pass hepatic metabolism limiting the systemic exposure a course of prednisone would carry. Azathioprine works on a genuinely different timeline, typically weeks longer to reach effect, but is also the drug her team would reach for to hold any remission budesonide achieves, since induction-dose budesonide isn’t meant to run indefinitely. The actual question in front of the group isn’t only which drug gets her into remission fastest — it's whether starting only one of them, with no stated plan for the other, quietly commits her to relitigating this exact decision again once the first drug does its job.
GI clinic, after the repeat biopsy
I want to start open-capsule budesonide. This is the one drug in front of us with trial evidence generated specifically in refractory Type I disease, not borrowed from somewhere else, and its first-pass hepatic metabolism means she isn’t taking on the systemic steroid burden a prednisone course would carry. There’s no reason to reach for the drug with the slower onset and heavier immunosuppressive profile first, when the better-tolerated option has real disease-specific data behind it.
I agree with the evidence on budesonide in general, but I’m less comfortable treating this as a straightforward sequential trial given where her numbers actually are. Her albumin dropped half a point in two months and her hemoglobin is trending down on top of that — that’s not a flat refractory picture, that’s active deterioration happening while we’ve been confirming a diagnosis she already had. If budesonide doesn’t work and we only start azathioprine after that trial fails, weeks from now, I think we’re gambling with a nutritional trajectory that has real momentum in the wrong direction.
I’m not disputing budesonide’s evidence base — I’m disputing whether "start with the gentler drug and escalate if it fails" is actually the conservative choice here, or whether it just looks conservative because it avoids committing to immunosuppression today.
I want to name something neither position has actually addressed: what happens after remission, however we get there. Induction-dose budesonide is not meant to run indefinitely — if it works, someone has to have already decided what holds the gain. If the plan is "budesonide now, azathioprine later, once we see how she responds," that’s a real and defensible plan, but it needs to be said out loud and planned for now, with a specific taper-and-transition timeline, not treated as a decision we’ll make again once we get there.
Given her nutritional trajectory, I think the honest answer synthesizes both of you rather than choosing between you: start budesonide now for the faster, better-tolerated induction attempt, but start it with an explicit four-to-six-week reassessment built in from day one — not an open-ended trial — and have azathioprine already queued as the maintenance plan the moment budesonide gets her into remission, not as a fallback we only reach for if it fails.
Agreed: budesonide 9mg/day starts today, with a hard four-to-six-week reassessment of symptoms, albumin, and hemoglobin already scheduled — not an open-ended trial — and azathioprine named in the chart today as the maintenance plan to start once remission is reached.
Not agreed: how much weight her current nutritional trajectory should carry if the four-week mark shows only partial improvement. The hematologist wants a lower bar for switching to azathioprine early, given how fast her albumin has already moved; the gastroenterologist wants the full window honored before concluding budesonide has failed.