Larazotide eased celiac symptoms but missed its permeability endpoint

What is larazotide acetate?

Larazotide acetate (AT-1001, also listed as INN-202 and INN-1100) is a synthetic eight-amino-acid peptide taken orally and designed to act locally in the small intestine rather than enter systemic circulation.

Its proposed mechanism is regulation of intestinal tight junctions, the protein complexes that control how permeable the gut lining is to molecules passing between, rather than through, epithelial cells.

The rationale traces to zonulin, a signaling protein hypothesized to loosen tight junctions in response to gluten exposure in people with celiac disease. Larazotide is described as a zonulin antagonist: block zonulin's signal, keep the junctions closed, and limit the downstream immune cascade that gluten triggers in celiac patients.

Is the zonulin mechanism behind larazotide settled?

No. The commercial ELISA assay historically used to measure zonulin levels has been challenged for cross-reacting with unrelated proteins, including by researchers tied to the original zonulin research.

That challenge does not make the tight-junction mechanism false. It means the biomarker used to support the underlying biology is contested, which is a different evidence position from the confident "blocks zonulin" language most coverage uses. Contested assays undermine whole literatures elsewhere too, as irisin's ELISA problem shows.

Does larazotide reduce celiac symptoms in trials?

Yes, on symptom scales, across four RCTs in diagnosed celiac disease. The trials ran in sequence, and the symptom finding held while the permeability finding did not:

• Proof of concept (n=21). A single 12 mg oral dose blunted the intestinal permeability rise normally seen after gluten challenge and reduced GI symptoms significantly (P = 0.018) ^[1]. This trial is the one most often cited as if it were the whole story. It is a strong signal in a very small sample, and it set expectations the larger trials did not fully meet.
• Phase 2a dose-ranging (n=86). Lower doses appeared to soften gluten-induced symptom worsening on a validated symptom scale (GSRS), while the trial's primary permeability outcome missed ^[2].
• Gluten-challenge trial (n=184). At 1 mg three times daily, larazotide significantly reduced GI symptoms (GSRS, P = 0.002) and attenuated the rise in anti-transglutaminase IgA, an immune marker, during a six-week gluten challenge ^[3].
• Largest completed trial (n=342). The 0.5 mg dose met its primary endpoint, a validated celiac-specific symptom scale (CeD-GSRS), with a 26% reduction in symptomatic days and a 31% increase in improved-symptom days (ANCOVA P = 0.022) ^[4].

The largest trial produced an oddity worth sitting with: higher doses of 1 mg and 2 mg showed no benefit ^[4]. That inverse dose-response suggests a narrow therapeutic window rather than "more peptide, more effect."

A pooled meta-analysis of 626 patients across four trials found statistically significant symptom improvement on GSRS and CD-GSRS scales in the gluten-challenged subgroup, including a reduction in gluten-related diarrhea ^[5].

Did larazotide ever reduce intestinal permeability?

Not as a trial endpoint. The LAMA-ratio permeability endpoint failed in every larger trial that set it ^[2]⁠^[3]⁠^[5], while the n=21 proof-of-concept study did blunt the permeability rise after gluten challenge ^[1].

In the Phase 2a dose-ranging trial, the primary endpoint, the lactulose-to-mannitol urinary ratio (LAMA ratio), showed no significant difference from placebo ^[2]. The larger gluten-challenge trial's permeability endpoint again did not reach significance ^[3]. Across the pooled dataset of four trials, larazotide did not significantly improve the intestinal permeability primary endpoint ^[5].

The pattern is the first and largest crack in the "it fixes leaky gut" framing. The permeability biomarker did not move in the larger trials, even where symptoms did.

What happened to larazotide's Phase 3 trial?

The confirmatory Phase 3 trial, CD303 (n=307), was terminated in 2022 before efficacy was established. That trial would have moved larazotide toward regulatory approval.

Most existing coverage omits or buries the termination, and it changes the shape of the whole evidence trail. The story is not small promising trials confirmed at scale. It is small promising trials, a permeability endpoint that never succeeded across four attempts, and a large confirmatory trial that did not reach a clean readout.

Evidence weakening as rigor and sample size increase is the opposite of what a maturing therapeutic candidate usually looks like, and it is the core reason larazotide rates Grade C, moderate, rather than higher.

Who were larazotide's trials actually run in?

Every larazotide trial enrolled people with a celiac diagnosis, mostly already on a gluten-free diet, dosed around episodic gluten exposure, either an experimental challenge or accidental cross-contamination.

No trial enrolled people without celiac disease. None tested daily long-term use for general gut wellness. None measured outcomes outside the GI tract: no skin, joint, mood, or cognitive endpoints appear anywhere in the larazotide trial record.

The strongest, most consistent finding across four RCTs is narrow: larazotide, taken orally around a defined gluten exposure, reduces gastrointestinal symptoms in diagnosed celiac patients already on a gluten-free diet. That is a real, replicated, imperfect finding on a symptom scale. It is not a validated claim that larazotide seals the gut, blocks zonulin in a mechanistically confirmed way, or does anything measurable for permeability ^[1]⁠^[2]⁠^[3]⁠^[4]⁠^[5].

Moving from "reduced GI symptoms after gluten exposure in diagnosed celiac patients" to "fixes leaky gut" changes the population, the exposure pattern, and the outcomes measured. The absent evidence for teduglutide and leaky gut covers the same leap for a different peptide.

Can larazotide affect skin, joints, or brain fog?

No. Larazotide is not meaningfully entering circulation, so it has no proposed pathway to influence skin barrier function, joint inflammation, or brain fog, however often those symptoms are colloquially blamed on leaky gut.

The gut-restricted action cuts both ways. Low systemic absorption limits opportunities for systemic toxicity, which is why larazotide's trial safety profile looks clean. The same restriction is a hard ceiling on plausibility for any effect outside the intestine.

Is injected larazotide safe?

Injected larazotide has no safety data. Larazotide was designed and tested exclusively for local, oral, gut-restricted action, and injection is a different exposure route.

In trials, oral larazotide was well tolerated, with adverse events similar to placebo, in diagnosed celiac patients. Why peptides can't be swallowed explains why oral activity is the exception that makes larazotide's design unusual.

What is still unknown about larazotide?

The target, the biomarker, and every population outside diagnosed celiac disease remain open questions:

• The target. Whether intestinal permeability is even the right target is unresolved, because no trial has demonstrated the LAMA-ratio effect the rationale predicts.
• The biomarker. Whether the zonulin biology behind the "antagonist" framing measures what researchers think it measures is actively contested.
• Other populations. No clinical data exist, positive or negative, in non-celiac populations, in people outside a gluten-challenge context, or in anyone using larazotide outside the oral route studied.