Has the colchicine / GLP-1 symptom overlap ever produced a documented late-recognised toxicity? A case-report and FAERS search
Not medical advice. This is a literature and adverse-event-database search for the founder's own reading and for sharpening a question list for his prescriber. Nothing here is a dose, a timing, or a change recommendation. Tags used below: [STUDY] = published paper; [FAERS] = my own counts from the public openFDA adverse-event API, which are unvalidated spontaneous reports; [REGULATOR] = a medicines-agency publication; [REASONING] = my inference, the weakest tier.
The question
"Is there any published case report or pharmacovigilance signal of colchicine toxicity being recognised late in GLP-1 or GLP-1/GIP receptor agonist users because the gastrointestinal warning symptoms were attributed to the weight-loss drug?"
This is the first open follow-up of [[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]], which found that the tirzepatide/colchicine risk is not kinetic but a possible signal-masking overlap, and left "has it produced real harm" unsearched. Context: the founder is on tirzepatide and allopurinol; colchicine has still never been confirmed as prescribed to him, so this remains a what-if for a future flare dose or bridge.
What we already know (from the vault)
- The overlap is label-documented on both sides. Colchicine's gastrointestinal effects appear within 24 hours in up to 20% of patients and "can herald the onset of more significant toxicity"; tirzepatide's most common adverse reactions are the same symptoms, with severe gastrointestinal reactions in 3.1% at 15 mg versus 1% on placebo. [[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]]
- The stronger chain in that brief was renal, not gastric. GLP-1-driven vomiting or diarrhoea leads to volume depletion, then to a renal-function change, and renal function governs exposure to both allopurinol and colchicine. Same brief; it runs through the founder's confirmed dehydration trigger in [[gout-flare-log]].
- Colchicine is a recommendation in the protocol, not a prescription. [[2026-06-24-gout-management-protocol]] pairs urate-lowering initiation with low-dose colchicine or an anti-inflammatory; [[gout-flare-log]] shows the 2026-07-06 visit produced prednisone plus allopurinol, and no colchicine dispensing has been recorded since.
- No study of tirzepatide with any gout drug exists. [[2026-07-20-tirzepatide-allopurinol-gout-maintenance-monitoring]] and the September brief both established this; the present search extends the negative to the safety-report literature.
- Dose data is still missing. [[zepbound-log]] has no entry past dose 1, so the current tirzepatide dose (and therefore where he sits on the gastrointestinal-reaction curve) is not in the vault.
What the web says
- [STUDY] No published case report of this scenario was found. A PubMed title/abstract search for colchicine with any of semaglutide, tirzepatide, liraglutide, dulaglutide, exenatide or "GLP-1" returned 28 records. I read all 28 titles: every one is a cardiovascular co-therapy review, guideline, trial-era commentary or animal study (the two drugs are both used for residual cardiovascular risk), plus one unrelated 2008 retina paper. None is a toxicity case. A Europe PMC full-text search for colchicine toxicity/poisoning/intoxication/overdose terms together with any GLP-1 drug name, restricted to case reports, returned one hit (a 2011 letter titled "A subtle case of iron-deficiency anemia", PMID 20532731, not retrieved, title unrelated) and zero hits when the toxicity term was required in the title.
- [STUDY] Late or missed recognition of colchicine toxicity is a real, documented phenomenon - just never with a GLP-1 drug as the decoy. A 2023 BMJ Case Reports paper describes colchicine toxicity "mimicking septic shock" after clarithromycin co-prescription and concludes that "failure to recognise drug toxicity can result in poor outcomes" (PMID 37813551). A 2002 Pediatrics report is titled "Late diagnosis of severe colchicine intoxication" (PMID 11986465). The documented decoys are sepsis, gastroenteritis and unknown ingestion, not another drug's side-effect profile.
- [FAERS] The co-reported population exists and is growing, so the absence is not for lack of exposure. openFDA (data through 2026-10-28, 14.7 million reports): 26,665 reports list colchicine; 444,751 list at least one of the five GLP-1-class drugs queried; 584 list both (semaglutide 191, liraglutide 116, dulaglutide 113, tirzepatide 103, exenatide 94; some overlap). Co-reports by year: 22 in 2022, 55 in 2023, 75 in 2024, 106 in 2025, 106 so far in 2026. Roughly 543 are distinct on a crude age/sex/reaction-set dedupe.
- [FAERS] Colchicine is almost never the accused drug in those reports. Colchicine is coded as suspect or interacting in 27 of 584 (4.6%); a GLP-1 drug is suspect in 169. In the rest both are bystanders in reports about something else.
- [FAERS] There is no disproportionality signal for toxicity; the crude number points the other way. Among colchicine reports, I compared those with a GLP-1 drug co-listed against those without. Toxicity-type coding (toxicity, overdose, drug level increased, drug interaction): 28 of 584 (4.8%) versus 2,746 of 26,081 (10.5%), reporting odds ratio (ROR) 0.43, 95% CI 0.29-0.63. Marrow/muscle end-organ terms (pancytopenia, rhabdomyolysis, myopathy, neuromyopathy, agranulocytosis, multi-organ dysfunction, marrow failure): 6 (1.0%) versus 1,667 (6.4%), ROR 0.15 (0.07-0.34). Fatal outcome: 5.8% versus 12.5%. Nausea/vomiting/diarrhoea: 15.8% versus 14.6%, ROR 1.09 (0.87-1.37), no difference. Acute kidney injury/renal failure/dehydration: 12.3% versus 9.7%, ROR 1.30 (1.01-1.67), a marginal elevation.
- [FAERS] Reading the toxicity-coded co-reports one by one finds other culprits. I pulled all 584 and read every report carrying a toxicity-type term (31 on a slightly wider term list). An 11-report US cluster of acute kidney injury plus "drug interaction" is one case duplicated across manufacturers, attributed to nifedipine with nirmatrelvir/ritonavir. A 6-report Belgian cluster is one polypharmacy patient with spironolactone, metformin and others as suspects. Three Spanish pancytopenia reports are one patient on azathioprine, rituximab and allopurinol. Others implicate lenalidomide, methotrexate with pegloticase, tobramycin, and a morphine overdose. In none of these is colchicine the suspect.
- [FAERS] The closest thing to the scenario is one French case, and it shows recognition rather than masking. Two 2026 reports (IDs 26756444 and 26839973, same patient, physician-reported via the French agency network): a 43-year-old man on semaglutide 0.25 mg weekly who started colchicine 1 mg daily for pericarditis and metformin on the same day, developed diarrhoea, coded "potentiating drug interaction" with colchicine, metformin and semaglutide all marked as interacting; colchicine was withdrawn and he recovered. No toxicity term, not hospitalised. One older report (2011, France) has colchicine as a suspect with liraglutide as a bystander in a case of diarrhoea, colitis, rhabdomyolysis, lactic acidosis and renal failure, alongside atorvastatin, metformin, mycophenolate and ramipril as co-suspects. FAERS carries no narrative text, so timing of recognition cannot be read from either.
- [STUDY] [REGULATOR] The published signal work and regulator guidance on colchicine toxicity do not mention the class. The main FAERS signal-detection study of colchicine interactions (Gomez-Lumbreras et al., Ann Pharmacother 2023) found 37 ROR signals, all among CYP3A4/P-glycoprotein inhibitors, strongest for atazanavir with rhabdomyolysis/myopathy (ROR 35.4); GLP-1 drugs were outside its design. New Zealand's Medsafe published "Factors contributing to colchicine toxicity" in March 2025 after two toxicity case reports: the factors named are renal and hepatic impairment, older age and low weight, interacting inhibitors, marrow-suppressing co-medication and patient misunderstanding. No GLP-1 drug and no symptom-misattribution mechanism appears. Its patient advice is to "stop colchicine immediately and seek medical advice if they experience gastrointestinal side effects." My one web query for UK, EU, Australian and French agency communications on this pair surfaced none.
Convergences and contradictions
- Convergence: the vault's September inference and every source searched agree that the overlap is real on paper and undocumented as harm. Three independent lenses (case-report literature, spontaneous reports, regulator guidance) return the same negative.
- Contradiction with the worry, mildly: the one on-point case shows the opposite of masking. The French clinician saw diarrhoea in a patient on both drugs, blamed the combination, and stopped colchicine. Additive gastrointestinal symptoms may make clinicians quicker to pull colchicine, not slower. One case cannot carry that conclusion; it only shows the masking story is not the only way the overlap plays out.
- Caveat that limits everything above: a spontaneous-report database is structurally nearly blind to this exact failure. If colchicine toxicity is missed because symptoms were blamed on a GLP-1 drug, the report (if any) gets filed against the GLP-1 drug as nausea or diarrhoea, or never gets filed. The 0.43 ratio is also not evidence of protection: GLP-1 co-reports are mostly reports about the GLP-1 drug with colchicine as a bystander, while colchicine-only reports are enriched for overdoses and poisonings. "No signal found" here means "not visible," which is weaker than "not happening."
Synthesis for RDCO
The answer to the question as asked is no. There is no published case report, no disproportionality signal, and no regulator communication describing colchicine toxicity recognised late because gastrointestinal warning symptoms were attributed to a GLP-1 or GLP-1/GIP agonist. The hypothesis the September brief raised remains a label-to-label inference with zero documented cases behind it. That is a meaningful update from "unsearched" to "searched across three source types and empty," and it should lower the weight this concern carries in the founder's gout thinking.
How strong is the negative? Moderate, not conclusive. On the reassuring side: more than 500 distinct reports now co-list the two drug classes, the count is rising every year, colchicine is the accused drug in under 5% of them, and the serious end-organ outcomes that define real colchicine toxicity (marrow suppression, rhabdomyolysis, multi-organ failure) show up in about 1% of co-reports against 6% of colchicine reports generally. If masking were producing a visible stream of late-caught toxicity, those end-organ terms are where it would surface, because late recognition is exactly what converts a gastrointestinal warning into marrow and muscle injury. They are not elevated. On the cautionary side: the comparison is confounded, the database has no narratives, co-prescription at scale is only two to three years old, and the mechanism is one that hides itself from this kind of surveillance. [REASONING] My read is that this is a low-probability, plausible-in-principle risk that would mostly matter in the settings where colchicine toxicity already happens: reduced kidney function, a CYP3A4 or P-glycoprotein inhibitor on board, or dosing above the label. Outside those settings, low-dose colchicine rarely progresses past gastrointestinal upset regardless of what else the patient takes.
The more useful finding is where the data does lean. The one marginally elevated term group in co-reports is the renal one (acute kidney injury, renal failure, dehydration; ROR 1.30, lower bound 1.01). That is weak, confounded by the chronic kidney disease common in this population, and partly inflated by duplicate reports, so I would not call it a signal. But it is directionally the same chain [[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]] ranked first: gastrointestinal losses, volume depletion, kidney function, drug exposure. Two briefs in a row have now looked for an exotic tirzepatide-gout interaction and come back pointing at hydration and kidney function instead. For [[2026-06-24-gout-management-protocol]] that argues for leaving the masking concern as a one-line note and keeping the emphasis where it already is.
What the founder could ask, phrased as questions and only relevant if colchicine is ever prescribed: (1) "Since I already get stomach symptoms from tirzepatide at times, how should I tell an ordinary bad week from a reason to stop colchicine and call you?" The regulator advice found here is simply to stop and seek advice on any gastrointestinal symptoms, which a prescriber may or may not want applied literally to someone with a baseline of them. (2) "Has my kidney function been checked since I started tirzepatide, and would you want it rechecked before or during a colchicine course?" (3) "Is anything else I take on the colchicine interaction list?" That list, not the GLP-1 class, is where every documented toxicity signal sits. These are conversation prompts, not instructions; the prescriber's answer governs.
Why this is in the vault
It closes the first open follow-up of [[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]] with a documented negative, so the tirzepatide risk section of [[2026-06-24-gout-management-protocol]] can carry the signal-masking concern as "theoretical, searched 2026-11-11, no cases" instead of as an open unknown, and so the question does not get re-queued.
Open follow-ups
- Does VigiBase or EudraVigilance, which hold the European reports where colchicine 1 mg dosing is standard, show a different picture from FAERS for colchicine co-reported with GLP-1 drugs? The single on-point case here was French, and openFDA under-represents non-US reports.
- In GLP-1 or GLP-1/GIP agonist users, how often does acute kidney injury follow a gastrointestinal adverse reaction, and at what point in dose escalation? The label warns about it and co-reports lean that way, but an incidence figure from trial or cohort data is not in the vault.
- Is there any published evidence that baseline gastrointestinal symptoms from one drug delay recognition of a second drug's toxicity in general (for example metformin or GLP-1 symptoms masking digoxin or lithium toxicity)? A general answer would price this mechanism better than a colchicine-only search can.
Related
- [[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]] - the parent brief; this closes its first open follow-up
- [[2026-06-24-gout-management-protocol]] - the active protocol whose tirzepatide risk note this brief settles
- [[2026-07-20-tirzepatide-allopurinol-gout-maintenance-monitoring]] - the grandparent brief on maintenance monitoring
- [[gout-flare-log]] - flare history and the still-unconfirmed colchicine status
- [[zepbound-log]] - tirzepatide dose log, stale after dose 1
- [[2026-06-15-tirzepatide-uric-acid-gout-flare-window]] - early-window urate mechanism brief
- [[2026-10-12-glp1-discontinuation-cohort-past-18-months]] - adjacent GLP-1 real-world-data brief
Sources
- Vault:
~/rdco-vault/06-reference/research/2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption.md([[2026-09-27-tirzepatide-gastric-emptying-oral-gout-drug-absorption]]) - Vault:
~/rdco-vault/01-projects/longevity/gout-flare-log.md([[gout-flare-log]]) - Vault:
~/rdco-vault/01-projects/longevity/zepbound-log.md([[zepbound-log]]) - Vault:
~/rdco-vault/01-projects/longevity/2026-06-24-gout-management-protocol.md([[2026-06-24-gout-management-protocol]]) - cited via the parent brief, not re-read on this run - Web (primary data, my own queries): openFDA drug adverse event API, https://api.fda.gov/drug/event.json, dataset last updated 2026-10-28, queried 2026-11-11. Drug matching used the harmonised
patient.drug.openfda.generic_namefield for colchicine, tirzepatide, semaglutide, liraglutide, dulaglutide and exenatide. All 584 co-reports were downloaded and the toxicity-coded ones read individually. Report IDs cited: 26756444, 26839973, 7783176-8. openFDA's own disclaimer applies: results are unvalidated and are not for medical decisions. - Web (search): PubMed E-utilities, title/abstract and all-field searches for colchicine with GLP-1 drug names and toxicity, delayed-diagnosis and pharmacovigilance terms (28 + 1 + 3 + 1 + 14 records, titles read)
- Web (search): Europe PMC REST full-text search, colchicine toxicity terms with GLP-1 drug names (1 case-report hit, 0 title hits)
- Web (abstract): Gomez-Lumbreras A, et al. "Drugs That Interact With Colchicine Via Inhibition of Cytochrome P450 3A4 and P-Glycoprotein: A Signal Detection Analysis Using a Database of Spontaneously Reported Adverse Events (FAERS)." Ann Pharmacother 2023;57(10):1137-1146, PMID 36688283
- Web (abstract): Ngeyvijit J, et al. "CYP3A4/P-glycoprotein inhibitors related colchicine toxicity mimicking septic shock." BMJ Case Rep 2023;16(10):e257186, PMID 37813551
- Web (abstract): Guven AG, et al. "Late diagnosis of severe colchicine intoxication." Pediatrics 2002;109(5):971-3, PMID 11986465
- Web (regulator, fetched): Medsafe, "Factors contributing to colchicine toxicity," Prescriber Update, 6 March 2025 - https://medsafe.govt.nz/Profs/PUArticles/March2025/Factors-contributing-to-colchicine-toxicity.html
- Web (not retrieved, flagged): PMID 20532731, "A subtle case of iron-deficiency anemia," Intern Emerg Med 2011 - the lone full-text case-report hit; title unrelated, body not read, so its relevance is unverified
- Web (not searched, flagged): VigiBase, EudraVigilance, and the MHRA, EMA, TGA and ANSM sites directly; lixisenatide, albiglutide and compounded products were not in the FAERS query