Compounds with a graded row

Ordered by rung, strongest first. The rung rates the study design; the chip beside it says what the study measured. Both are needed.

1. Liraglutide

   human observational Direct outcome Not approved for this condition

   Who was studied

   adults with ulcerative colitis who started liraglutide or semaglutide for a metabolic indication and stayed on it at least 12 weeks, forming part of a 150-patient treated arm matched 1:1 to 150 untreated controls, single US academic health system, 2022 to 2024; the liraglutide subgroup size and its sex breakdown are not reported in the abstract

   In the same retrospective matched cohort, the liraglutide subgroup reached 60% symptomatic remission at 12 weeks (partial Mayo 2 or less with rectal bleeding subscore zero) against 72.5% for semaglutide, while the whole GLP-1 receptor agonist arm reached 66.7% against 25.3% in matched controls. The comparison between the two drugs was not randomised and no head-to-head trial exists. Treatment was started for metabolic indications, not for colitis, the design cannot exclude confounding by indication, and no randomised trial of liraglutide in ulcerative colitis is on file.

   Source PubMed 1 primary source read for this row

2. Semaglutide

   human observational Direct outcome Not approved for this condition

   Who was studied

   adults with ulcerative colitis who started semaglutide or liraglutide for a metabolic indication and stayed on it at least 12 weeks, n=150, matched 1:1 to 150 untreated controls at one US academic health system, 2022 to 2024; sex breakdown not reported in the abstract

   A retrospective matched electronic-health-record cohort measured symptomatic remission at 12 weeks, defined as a partial Mayo score of 2 or less with a rectal bleeding subscore of zero: 66.7% of GLP-1 receptor agonist users against 25.3% of controls, adjusted odds ratio 5.90 (95% CI 3.52 to 9.86), with weight loss not associated with remission. Within the treated arm the semaglutide subgroup reached 72.5% against 60% for liraglutide, but the two were not randomised against each other and subgroup sizes are not given in the abstract. Everyone treated started the drug for a metabolic indication rather than for colitis, so the arms differ by indication as well as by drug, and no randomised trial of semaglutide in ulcerative colitis is on file.

   Source PubMed 1 primary source read for this row

3. Guanylin

   human observational Mechanistic only Not approved for this condition

   Who was studied

   adults with ulcerative colitis, n=60, graded 1 to 3 on a modified Mayo disease activity index, against 20 normal controls; colonic mucosal expression measured by quantitative RT-PCR and Western blot, nothing administered

   Recorded, and not evidence for this condition

   Guanylin messenger RNA and protein in colonic mucosa were lower in ulcerative colitis than in controls and fell further as the disease activity index rose, alongside guanylate cyclase-C and uroguanylin. Nothing was given to anyone: no study has administered guanylin to a person with ulcerative colitis. A separate experiment in Balb/c mice delivered a guanylin overexpression vector once daily for a week and reported lower intestinal permeability and histopathology scores - that is gene delivery in a mouse, not the peptide given as a treatment, and it does not measure the condition in a human.

   Source PubMed 2 primary sources read for this row

4. Omentin

   human observational Mechanistic only Not approved for this condition

   Who was studied

   consecutive adults with ulcerative colitis, n=126, stratified by Montreal extent into distal E1/E2 (n=84) and extensive E3 (n=42); a 36-patient subgroup assessed at week 14 of infliximab induction; age and sex not reported in the abstract

   Recorded, and not evidence for this condition

   Serum omentin-1 was quantified by ELISA and was lower in extensive than in distal colitis, median 48.60 against 62.70 ng/mL (p<0.01), and higher at baseline in the 23 of 36 patients who responded to infliximab. Omentin-1 was only measured; no study has administered omentin to anyone with ulcerative colitis, so nothing here tests whether giving it changes disease activity. This is a biomarker association read off blood samples, and the step from it to an effect is an argument rather than a result.

   Source PubMed Central 1 primary source read for this row

5. Uroguanylin

   human observational Mechanistic only Not approved for this condition

   Who was studied

   adults with ulcerative colitis, n=60, graded 1 to 3 on a modified Mayo disease activity index, against 20 normal controls; colonic mucosal expression measured by quantitative RT-PCR and Western blot, nothing administered

   Recorded, and not evidence for this condition

   Uroguanylin messenger RNA and protein in colonic mucosa were lower in ulcerative colitis than in controls and fell in proportion to the disease activity index, alongside guanylate cyclase-C and guanylin. This is an expression measurement in removed tissue, not an intervention: no study has administered uroguanylin to a person with ulcerative colitis, and the step from a downregulated endogenous ligand to a treatment effect is an argument, not a result.

   Source PubMed 1 primary source read for this row

6. BPC-157

   animal in vivo Direct outcome Not approved for this condition

   Who was studied

   rats given 400 mg/kg cysteamine intrarectally, and a separate colon-colon anastomosis model, sacrificed at days 3, 5, 7 and 14; strain, sex and n per group not reported in the abstract obtained

   BPC 157 at 10 mcg/kg or 10 ng/kg intraperitoneally, or 0.16 mcg/mL in drinking water, was reported to heal cysteamine colitis and colon-colon anastomosis in rats where saline controls did not. Cysteamine colitis is a chemical injury model, not a model of ulcerative colitis, and the same paper's other half is a cuprizone brain-injury model, so the colitis arm is one part of a broad claim. The BPC-157 colitis literature is large but almost entirely review articles from the one Zagreb group; the phase II inflammatory bowel disease programme they cite (PL 14736, Pliva) has no published result, and ClinicalTrials.gov lists only three registered BPC-157 studies - healthy volunteers, hamstring strain and a peptide gummy - none with a gastrointestinal endpoint. There is no human colitis evidence of any rung.

   Source PubMed 2 primary sources read for this row

7. KPV

   animal in vivo Direct outcome Not approved for this condition

   Who was studied

   female C57BL/6 mice, 8 weeks old, 18-22 g, 10 per group; DSS colitis (3% w/v in drinking water, assessed at 8 days) and TNBS colitis (150 mg/kg in 50% ethanol intrarectally, assessed at 48 hours)

   Oral KPV at 100 micromolar in the drinking water reduced body weight loss, colonic myeloperoxidase activity, histological inflammation and pro-inflammatory cytokine mRNA in both chemical colitis models. The rest of the paper is in vitro, in Caco2-BBE and HT29-Cl.19A epithelial lines and Jurkat T cells, and shows the effect runs through the PepT1 di/tripeptide transporter. DSS and TNBS colitis are chemical injury models and not ulcerative colitis, the tripeptide was delivered in water rather than as a dosed medicine, and no human trial of KPV in ulcerative colitis or in any inflammatory bowel disease is on file.

   Source PubMed Central 1 primary source read for this row

8. PEA

   animal in vivo Direct outcome Not approved for this condition

   Who was studied

   mice with DSS-induced colitis, strain, sex and n per group not reported in the abstract obtained; the human material was colonic biopsies from ulcerative colitis patients and primary human enteric glial cell cultures, not treated patients

   Palmitoylethanolamide improved the macroscopic signs of DSS colitis in mice - disease activity index, colon length, spleen weight, macrophage and neutrophil infiltration - and lowered the pro-inflammatory markers tested, with the effect abolished by a PPAR-alpha antagonist but not a PPAR-gamma one. The human half of the study exposed ex vivo ulcerative colitis biopsies and enteric glial cultures to the compound in a dish. No patient was given palmitoylethanolamide, no clinical endpoint was measured in a person, and no trial of it in ulcerative colitis is on file.

   Source PubMed 1 primary source read for this row

Recorded absences

1 compound is listed here without a rung. Each row exists to mark a distinction a reader would otherwise draw wrongly, and each says what is missing rather than leaving the compound off the page.

1. Carnosine

   No rung — nothing on file to grade Not approved for this condition

   Recorded as a named absence. The one randomised, placebo-controlled, investigator-blinded ulcerative colitis trial in this area tested polaprezinc, a zinc-L-carnosine chelate, and not carnosine: 28 patients, 18 given a 150 mg polaprezinc enema and 10 not, with modified Matts' endoscopic and Mayo scores read at one week on top of usual induction therapy. The zinc is half the molecule that was tested and that design cannot separate it from the carnosine, so the trial grades as evidence for polaprezinc. No trial of carnosine itself in ulcerative colitis is on file.

Checked, and nothing recorded

17 compounds were checked against ulcerative colitis and produced no gradeable row.

• Cholecystokinin (CCK) — checked, no row on file
• Dulaglutide — checked, no row on file
• Exendin-4 — checked, no row on file
• Galanin — checked, no row on file
• Gastrin — checked, no row on file
• Ghrelin — checked, no row on file
• GIP (Gastric Inhibitory Peptide) — checked, no row on file
• GLP-1 (7-36) — checked, no row on file
• GLP-2 (Teduglutide) — checked, no row on file
• Ipamorelin — checked, no row on file
• Motilin — checked, no row on file
• Neurotensin — checked, no row on file
• PYY3-36 — checked, no row on file
• Secretin — checked, no row on file
• Thymosin alpha-1 — checked, no row on file
• VIP — checked, no row on file
• Wheat Peptides — checked, no row on file

A compound is here because someone looked and the sources did not support a row — not because nobody looked. Under the rubric, uncertainty resolves to no row, and a row resting on a vendor page or a clinic blog is worse than no row at all.