Clinical trials for celiac disease therapies have a problem: they may be designed in ways that work against themselves. New research suggests that the way trials screen participants could be one reason so many promising treatments fail to show results.
What This Means for Celiac Families
For families like mine, this matters because clinical trials are the only path to new therapies. Every delay, every failed trial, every study that can’t show clear results pushes that timeline further out. And according to this research, some of those failures may not be because the drugs don’t work—they may be because the trials are selecting the wrong participants.
The issue is surprisingly simple: many trials only accept people whose intestines have healed completely on a gluten-free diet. The logic seems sound—you need healthy tissue at the start to measure damage from gluten. But that screening requirement may be filtering out the very patients most likely to show measurable responses.
People whose intestines heal perfectly may be biologically different from those who don’t fully heal. They might have milder disease or immune systems that respond differently. When researchers give this pre-selected group gluten, they may see less damage than expected—making it impossible to tell if the drug being tested actually works.
For my son and other patients waiting for therapies, this technical detail has real consequences. A trial that fails due to poor design—not because the drug doesn’t work—sets the field back years.
Key Takeaways
- Many celiac trials only accept people whose intestines healed completely on a gluten-free diet
- This screening requirement may filter out patients most likely to show strong responses to gluten
- When trials can’t detect damage clearly, they can’t tell if treatments work
- Researchers propose accepting patients with a range of healing levels to better represent the celiac population
- Better trial design could mean fewer failed studies and faster progress toward therapies
The Science
Want to understand the technical details behind this research? We’ll walk you through the methodology and terminology below, defining every term. No medical degree required.
The Standard Trial Blueprint
Most celiac disease clinical trials involving gluten challenge follow a similar design: recruit people with celiac disease who have been on a gluten-free diet, perform baseline biopsies, feed them gluten, then measure what happens to their intestines. The logic seems sound. But the screening process matters enormously.
The issue centers on the villous height-to-crypt depth (Vh:Cd) ratio—a measurement pathologists use to assess intestinal damage. In celiac disease, villi (the finger-like projections that absorb nutrients) become flattened while crypts (the pockets at the base of villi) deepen. The ratio quantifies that damage. Many trials require participants to have normalized or near-normalized villi at baseline, reasoning that you need healthy starting tissue to measure damage from gluten challenge.
But what if that screening criterion is filtering out the very patients most likely to respond in measurable ways?
The Healed-Villi Paradox
The researchers—Benjamin Lebwohl, Daniel Leffler, and Joseph Murray, publishing in Clinical Gastroenterology and Hepatology—argue that requiring near-normal histology (tissue structure) at trial entry creates a paradox. People whose intestines heal completely on a gluten-free diet may be different biologically from those who don’t fully heal. They might have milder disease, better dietary adherence, or immune systems that respond differently. When you gluten-challenge this pre-selected group, you may see less dramatic or less consistent damage than you’d see in a broader celiac population.
This becomes a statistical power problem—the ability of a study to detect real effects. If the drug being tested works by preventing gluten-induced damage, but your study population doesn’t show robust damage to begin with, you can’t measure the drug’s protective effect. Trials need clear signal-to-noise ratios (the difference between the treatment effect and background variation). If baseline healing status varies widely, and gluten response varies accordingly, the intervention effect gets muddied.
The Burden of Gluten Challenge
Gluten challenge trials ask people with celiac disease to do something that goes against every instinct and medical recommendation: eat gluten deliberately. As a parent, I can’t imagine asking my son to participate in a study that would make him sick on purpose, even for research. The symptoms, the intestinal damage, the weeks or months of feeling terrible—it’s a high bar for participation.
Earlier this year we covered the DAISY study, which is actively recruiting participants for gluten challenge research. Studies like DAISY recognize the burden on participants and offer compensation and support. But the fundamental challenge remains: you’re asking people to harm themselves temporarily in hopes of advancing science.
Given that sacrifice, trials need to be designed to extract maximum value from participants’ willingness to endure gluten exposure. Screening criteria that inadvertently reduce measurable outcomes waste that sacrifice.
What Optimal Criteria Might Look Like
The researchers propose rethinking entry criteria to better match trial goals. Instead of requiring near-complete healing, trials might benefit from enrolling patients with a range of baseline histology—provided they meet diagnostic criteria for celiac disease and have been adherent to a gluten-free diet.
This approach recognizes biological variability (natural differences between individuals) as real and meaningful rather than treating it as noise to be controlled away. It also acknowledges that the patients most representative of the broader celiac population—those whose healing is partial or incomplete despite dietary adherence—may be better study participants than outliers with perfect mucosal recovery (intestinal healing).
Another consideration is timing. How long should participants be on a gluten-free diet before baseline biopsy? Too short, and villi haven’t had time to heal. Too long, and you may preferentially enroll people whose intestines heal exceptionally well. The optimal window may vary depending on what the trial is testing.
Why This Matters for Drug Development
This isn’t just methodological fine-tuning. Drug development for celiac disease has been notoriously difficult. Multiple promising therapies have failed in late-stage trials, often due to inability to demonstrate efficacy (effectiveness) convincingly. Some failures reflect genuine lack of drug effect. But others may reflect trial design issues that made real effects impossible to detect.
The celiac community has been waiting for adjunct therapies—medications that could reduce accidental gluten exposure damage or allow small amounts of gluten in social situations—for decades. Every failed trial represents millions of dollars, years of work, and patient hopes that don’t materialize. If even some of those failures trace back to fixable design issues like entry criteria, that’s information the field desperately needs.
For pharmaceutical companies and researchers, this analysis offers a roadmap for designing trials with better chances of success. For patients and families, it’s a reminder that progress depends not just on discovering new drugs but on building the infrastructure to test them properly.
What Families Should Know About Trial Participation
If you’re considering enrolling yourself or your child in a celiac disease clinical trial, these technical details matter less than the basics: understanding the risks, knowing what’s required, and weighing potential benefits against guaranteed burden. Gluten challenge trials are not casual commitments.
That said, participation in research is how the field moves forward. Trials like DAISY and others depend entirely on people willing to contribute their time, their bodies, and their discomfort to science. The least researchers can do is design those trials to maximize the value of that contribution.
This paper doesn’t announce a breakthrough therapy or a diagnostic advance. But it does something equally important: it identifies a weak point in the research pipeline and proposes fixes. In a field where progress has been frustratingly slow, optimizing trial design is itself a form of progress.
The Long Game
My son is ten now. By the time any therapy currently in development reaches the market, he’ll be in his teens or beyond. The gluten-free diet works—it keeps him healthy, and we’ve built routines that make it manageable. But it’s exhausting, limiting, and socially isolating in ways that accumulate over years.
The promise of adjunct therapies isn’t that they’ll replace the gluten-free diet. It’s that they might reduce the stakes of mistakes. That a birthday party or a school field trip or a restaurant meal might carry less risk. That the cognitive load of vigilance might ease slightly.
That promise depends on clinical trials that work. And trials that work depend on getting the details right—including the details that determine who gets enrolled in the first place.
Related Coverage
References
Lebwohl B, Leffler DA, Murray JA. Optimizing Histologic Entry Criteria for Clinical Trials in Celiac Disease Involving Gluten Challenge. Clinical Gastroenterology and Hepatology. 2026 May 15:S1542-3565(26)00357-5. doi: 10.1016/j.cgh.2026.04.031. Available at PubMed