Adult Autoimmune Enteropathy: Choosing Among Options With No Real Trial Behind Any of Them
A single patient, forty pounds lighter than he was six months ago, with a diagnosis so rare that no drug proposed for him tonight has ever been tested against another in a real trial. The disagreement is about which thin evidence to trust first.
Owen F., a 45-year-old man, has lived alone since his divorce three years ago, close enough to his two teenage sons that he still coaches their travel soccer team on weekends — or did, until the past six months made standing on a sideline for two hours increasingly hard. He teaches high school shop class on weekdays, a job that kept him on his feet most of the day until his stamina began giving out there too. What started as intermittent loose stools became eight to ten watery bowel movements daily, and forty pounds of weight loss that has left him too weak to stand at a workbench for a full class period. Celiac serology and genetic testing were both negative, colonoscopy and upper endoscopy showed diffuse villous blunting without the pattern celiac disease produces, and a panel sent for anti-enterocyte and anti-goblet cell antibodies came back positive — findings that, together with the exclusion of celiac disease, infection, and lymphoma, support a diagnosis of adult-onset autoimmune enteropathy, a condition overwhelmingly described in children and only rarely reported starting in adulthood.
That rarity is exactly what makes tonight's decision genuinely difficult rather than routine. Adult AIE has no randomized trial evidence behind any treatment — the entire published literature is case reports and small retrospective series, most drawn from pediatric IPEX-related disease, with the adult, non-syndromic experience resting largely on two single-center Mayo Clinic series: Akram and colleagues, who described fifteen adults and wrote the diagnostic criteria still in use, and Sharma and colleagues, who later compared adult AIE against refractory celiac disease. Corticosteroids appear most often as the reported first agent across those series, which is a real if modest form of evidence in a disease this rare, but response rates and durability vary widely case to case, and Owen's albumin, already at 2.1 g/dL with a hemoglobin of 8.9, reflects a nutritional trajectory that has real momentum before any drug has even been tried. Tacrolimus and vedolizumab both have documented case-level responses in the same literature, drawn from comparably small numbers of reported patients, meaning the choice in front of the team is less about which option has better evidence and more about which reasoning — fastest, best-suited to his current severity, or best-matched to the disease's actual mechanism — the group trusts most when the evidence itself cannot arbitrate.
GI/nutrition joint clinic, new AIE diagnosis
I want to start systemic corticosteroids. This is the agent that appears most often as the reported first step in Akram's series and the adult case literature that followed it, and in a disease this rare, where no comparative trial exists or ever will, I think following the most commonly reported approach is the most defensible default we have.
I don’t disagree that steroids have the most case-report precedent, but I’m looking at his albumin and hemoglobin and I don’t think a sequential steroid-first trial is free of real cost. If steroids don’t work, declaring that failure and escalating to tacrolimus takes weeks he may not have much room to spare on a trajectory already this steep. I’d start tacrolimus now, with a shorter concurrent steroid taper, rather than waiting to see steroids fail first.
I’m not arguing tacrolimus has stronger evidence than steroids — it doesn’t, they’re comparably thin — I’m arguing that when the evidence can’t break the tie, his own deterioration curve should.
I want to name what I think both of you are actually doing: reaching for the more familiar tool in a situation where familiarity is close to the only thing distinguishing the options. I’d put vedolizumab on the table seriously. AIE’s pathology is a T-cell attack concentrated specifically in the gut mucosa — a gut-selective agent is mechanistically well matched to that in a way neither steroids nor tacrolimus’s broader systemic immunosuppression specifically is, and it carries real case-level response reports of its own.
I don’t think mechanistic fit should be mistaken for proof — it isn’t, and I want to be honest that this is a genuinely close call across three options with comparably thin direct evidence. But given his severity, I’d favor starting tacrolimus now, as the hematologist proposed, specifically because it acts faster than vedolizumab typically does, while keeping vedolizumab explicitly on the table as the next step if tacrolimus doesn’t hold him, rather than a full corticosteroid course tried and failed first.
Agreed: tacrolimus starts today with a short concurrent corticosteroid taper, aggressive nutritional support continued in parallel, and vedolizumab named explicitly in the chart as the next step if tacrolimus does not achieve remission within a defined window.
Not agreed: the gastroenterologist remains genuinely unpersuaded that departing from the most commonly reported first-line approach (corticosteroids alone) is justified here, given how comparably thin all three options’ evidence actually is — a position the group noted rather than resolved, given the shared acknowledgment that no trial exists to settle it either way.