Five questions worth testing
This discussion brief connects published celiac findings to five unresolved questions that researchers can challenge. It is intended to help identify a useful next analysis or preclinical experiment—not to announce a discovery, prescribe treatment or recruit patients.
Public discussion brief · Version 0.1 · September 23, 2026 · AI-assisted synthesis; not independently peer reviewed.
Explore the supporting research. Open the Celiac evidence library for study summaries, negative findings, models, datasets and the limits behind these questions.
Our research objective is durable, gluten-specific immune control together with preserved or restored intestinal function. We do not treat symptom relief, an improved blood test, reduced exposure or protection while taking a drug as interchangeable with that objective. The five questions are a research agenda, not a proven five-part treatment or a claim that one combination will work.
For patients and families: this is not something to try at home. Do not change a prescribed gluten-free diet, deliberately test gluten exposure or use an experimental drug because of this brief. Current care should remain with the treating clinician and dietitian. NIDDK describes gluten avoidance and appropriate follow-up as core care. [1]
What changes susceptibility into intestinal injury?
Published evidence and limit. A 2026 study implicated the regulatory RNA lnc13 in restraining gluten-related inflammatory responses in HLA-DQ8 mice. This is a mechanistic lead, not proof that restoring lnc13 treats human celiac disease. [2]
Question for criticism. Can a reversible immune or epithelial state explain progression better than inherited susceptibility or antibodies alone?
Potential decisive comparison. A researcher could examine appropriately consented, longitudinally characterized human samples, identify candidate pre-progression differences, and test causation in a suitable laboratory model. Exposure and other clinical differences must be considered rather than attributed automatically to the candidate mechanism.
What would weaken the idea. A difference that disappears after accounting for those other factors, or a manipulation that only suppresses immunity broadly, would weaken this explanation.
Useful expertise. Cohort investigators, mucosal immunologists and a laboratory able to evaluate candidate regulatory mechanisms.
Which cells are the important gluten-presentation routes?
Published evidence and limit. Human intestinal M cells were shown to process and present gluten antigen in organoid–T-cell assays. Demonstrating this capability does not establish that these cells are the sole or indispensable route in an intact human intestine. [3]
Question for criticism. Would selectively interrupting that route materially reduce the harmful response, or would other antigen-presenting cells compensate?
Potential decisive comparison. A qualified team could compare appropriately controlled human laboratory systems that include alternative antigen-presenting populations, rather than testing an isolated cell type and assuming it represents the whole disease.
What would weaken the idea. Little added effect when other presentation routes are present would argue against treating this as a dominant, stand-alone intervention point.
Useful expertise. Human intestinal organoid and antigen-specific T-cell expertise, plus independent criticism of model relevance.
Does protection during treatment persist after treatment ends?
Published evidence and limit. Teva reported reduced gluten-induced intestinal injury in a Phase 2a TEV-408 study. The source used here is a sponsor topline announcement, not an independent reanalysis or a demonstration of lasting treatment-free tolerance. [4]
Question for criticism. Does blocking the injury-producing response reset gluten-specific immunity, or protect tissue only while the treatment remains active?
Potential decisive comparison. Separate antigen-specific immune activity, tissue injury and persistence after the intervention has been removed in an appropriate nonclinical system. Any later human study would require its own scientific, ethical and regulatory approvals.
What would weaken the idea. Return of injury after removal, with the upstream gluten response unchanged, would support temporary protection rather than a durable immune reset.
Useful expertise. Mucosal immunology, suitable models or lawfully accessible biospecimens, and a credible assessment of residual treatment effects.
What immune change would count as genuine tolerance?
Published evidence and limit. TAK-101 was studied using a gluten-specific immune-response endpoint. A different antigen-specific approach, Nexvax2, failed to reduce acute gluten-induced symptoms in the terminated RESET CeD study. These are different products and endpoints, not a head-to-head comparison. [5, 6]
Question for criticism. Which immune-state changes, if any, track lasting tissue protection rather than a transient laboratory signal?
Potential decisive comparison. Where lawful samples and outcome data exist, compare immune-state findings against independent tissue outcomes and duration. Compare like-for-like endpoints and account for differences in interventions, populations and study design.
What would weaken the idea. A promising immune marker that does not predict independent tissue protection or persists only during active treatment would not establish the desired durable effect.
Useful expertise. Antigen-specific immune profiling, relevant sample access, trial-methodology expertise and willingness to examine negative findings.
Is recovery a separate problem after immune injury is controlled?
Published evidence and limit. A 2026 study found altered small-intestinal microbial fiber metabolism in celiac disease and faster healing with a dietary intervention in a sensitized mouse model during a gluten-free diet. This is not proof of a microbiome cure or demonstrated treatment-free tolerance in people. [7]
Question for criticism. Does supporting tissue recovery add a reproducible benefit after the inflammatory cause is equally controlled?
Potential decisive comparison. Compare recovery in an appropriate laboratory system while keeping the level of upstream immune injury comparable. Assess restored tissue function, not only an isolated biomarker.
What would weaken the idea. No added recovery benefit, a purely transient effect, or a result explained entirely by less initial injury would narrow this to an adjunct rather than an independent cure mechanism.
Useful expertise. Intestinal regeneration and barrier-function expertise, with metabolic or microbiome support only when justified.
A specific invitation to scientific criticism
The immediate request to researchers is written criticism: Which question is already answered? Which interpretation is wrong? Which existing dataset or laboratory model could decisively resolve the remaining uncertainty? A useful next response would identify an available model or lawful dataset, the most informative comparison, and the conditions under which the hypothesis should be rejected. A willingness to review is not an agreement to run a study. No funding, institutional endorsement, participant access or collaboration is represented as secured.
Sources and limits
Published findings belong to the cited researchers. The connecting questions and proposed comparisons are AI-assisted synthesis for expert review: AI-GENERATED / UNTESTED / NOT NOVELTY-CLEARED. No project experiment or patient-level reanalysis is reported here. Sources were checked at the accessible abstract, article or sponsor-announcement level; this is not a systematic review, full raw-data audit or complete trial-pipeline assessment. It does not determine whether a later study supersedes each historical result. Source links are provided; no third-party figures, full papers, private correspondence, partner proposals or proprietary implementation are reproduced.
- NIDDK
Treatment for Celiac Disease - Yang-Fischer et al. · Nature Immunology · 2026
The long noncoding RNA lnc13 restrains inflammatory responses to maintain oral tolerance to gluten - Wang et al. · Nature · online 2025; issue 2026
Human gut M cells resemble dendritic cells and present gluten antigen - Teva · sponsor announcement · September 2, 2026
Phase 2a topline results for investigational anti-IL-15 antibody TEV-408 - Kelly et al. · Gastroenterology · 2021
TAK-101 Nanoparticles Induce Gluten-Specific Tolerance in Celiac Disease: A Randomized, Double-Blind, Placebo-Controlled Study - Tye-Din et al. · Lancet Gastroenterology & Hepatology · 2023
Nexvax2 RESET CeD: interim analysis of a terminated randomized phase 2 study - Wulczynski et al. · Nature Communications · 2026
Small intestinal microbial fiber metabolism dysfunction in celiac disease