Colchicine reduces cardiovascular events in patients with coronary artery disease
Reassessed
Staleness check plus retirement of the contested status. No new material evidence was found. Re-weighed under the new scale: pooled evidence still favors a modest reduction, so the status moves from inconclusive (migrated) to supported, with confidence 0.6 and credence 0.65. Both arguments evaluated. Importance set to 0.6 and contestation to 0.6. The claim has no dependents to notify.
Reassessed: Inconclusive → Supported
verdict confidence 0.85 → 0.60 · credence 0.65
The evidence favors a modest reduction in cardiovascular events with low-dose colchicine in coronary artery disease, though the size of the effect and the populations it applies to are uncertain. Two large placebo-controlled trials found benefit: COLCOT after recent myocardial infarction (about 23 percent relative reduction) and LoDoCo2 in chronic coronary disease (about 31 percent). These trials led to FDA approval of low-dose colchicine in 2023 and to a class IIa recommendation in the ESC 2024 chronic coronary syndromes guideline. The causal role of inflammation in atherosclerosis, supported independently by CANTOS, makes the effect biologically plausible. The main evidence against comes from CLEAR SYNERGY, the largest trial (7,062 post-MI patients). It found no reduction in the primary composite (about 9.1 versus 9.3 percent) even though hs-CRP fell. Proposed explanations include a low event rate, high discontinuation, and only partial suppression of inflammation, but these are post hoc. Meta-analyses that include CLEAR SYNERGY still find a pooled reduction, with estimates of roughly 12 to 25 percent depending on method, and the effect is attenuated in acute coronary syndromes. A Bayesian reanalysis argues the evidence is consistent with a much smaller effect. The case is strongest in chronic coronary disease, where LoDoCo2 has not been contradicted by a comparable trial. It is weakest immediately after myocardial infarction, where the two large trials conflict. The question would be resolved by an individual-patient meta-analysis that explains why the trials disagree, or by a further large trial in chronic coronary disease.
No action needed
Trigger was structure_and_assess (first pass), but the claim is already fully structured and assessed: four subclaims (COLCOT, LoDoCo2, CLEAR SYNERGY, inflammation-causality) grouped under two named arguments, both with written forms and current evaluations; three affirming instances; a contested verdict (confidence 0.85, credence 0.6) recorded 2026-08-07 with thorough reasoning. The trigger appears to have re-fired on an already-processed claim. Rather than redo the structure, I ran a light staleness check: two web searches for developments since the assessment date. Nothing material surfaced. The 2026 Laudani ACS-focused meta-analysis, the 2025 cumulative-dose meta-analysis, and the EKSTROM surrogate-endpoint trial all either predate the existing assessment or are already reflected in its reasoning (it explicitly cites the dueling post-CLEAR SYNERGY meta-analyses and the 2025/2026 ACS-focused analysis). The contested status remains the right reading: credible randomized evidence and syntheses on both sides, with unresolved heterogeneity between COLCOT/LoDoCo2 and CLEAR SYNERGY. Canonical form, importance (0.65), and argument evaluations all remain apt. No assessment update, no structural change, no notification warranted.
Structured and assessed first pass
First pass. The claim arrived with a sound decomposition (COLCOT and LoDoCo2 under a "for" argument, CLEAR SYNERGY under an "against" argument, plus an ungrouped mechanistic support on inflammation's causal role); I judged it adequate and added no subclaims. The dueling post-2024 meta-analyses and the guideline/regulatory positions (FDA LODOCO 2023, ESC 2024 class IIa) are syntheses and endorsements of the same trial evidence, not independent dependencies, so they live in the assessment prose per §6. Canonical form kept: eleven words, neutral, both sides would accept it. Web evidence (3 searches): CLEAR SYNERGY null primary result confirmed (ACC, TCTMD, Radcliffe); post-CLEAR meta-analyses diverge (12-25% pooled MACE reduction vs. Bayesian reanalysis arguing negligible effect); ESC 2024 CCS guideline maintains class IIa. Verdict CONTESTED (confidence 0.85, credence 0.60): credible randomized evidence on both sides with no accepted reconciliation of the heterogeneity. This aligns with the seed (0.55) but was reached independently from the trial record and current syntheses. Importance set to 0.65 (contestation 0.85): major prescribing question for a large secondary-prevention population, actively argued. Both named arguments evaluated holds_with_caveats. No instances recorded this pass: sources read either reported the debate without asserting the CAD-wide claim in their own voice, or asserted only narrower trial-level propositions (post-MI, ACS) that belong to the subclaims. Marginal yield 0.4: an Elicit-grade literature pass or individual-patient-data synthesis, when available, could sharpen the effect-size question; the contested status itself is unlikely to move soon.
Assessed Inconclusive
verdict confidence 0.85 · credence 0.60
Whether low-dose colchicine reduces cardiovascular events in coronary artery disease is a live dispute within cardiology, with credible randomized evidence pointing in both directions. Two large placebo-controlled trials found a benefit: COLCOT, in patients with recent myocardial infarction, and LoDoCo2, in chronic coronary disease, together spanning the acute and stable ends of the disease. On their strength the FDA approved low-dose colchicine for cardiovascular risk reduction in 2023, and the European Society of Cardiology's 2024 chronic coronary syndromes guideline recommends it (class IIa). The biological rationale rests on the well-supported view that inflammation plays a causal role in atherosclerotic cardiovascular disease. Against this, CLEAR SYNERGY, the largest colchicine outcomes trial, found no reduction in cardiovascular events after myocardial infarction (about 7,000 post-MI patients, median follow-up around three years), despite lowering C-reactive protein, and colchicine trials in stroke populations (CONVINCE, CHANCE-3) were also null. Updated meta-analyses that include CLEAR SYNERGY still report a pooled reduction in major adverse cardiovascular events, with estimates ranging from roughly a 12 to 25 percent relative reduction depending on trial selection and endpoints, but Bayesian reanalyses and several commentators read the totality as consistent with little or no effect, and the size of any benefit is plainly smaller than the individual positive trials suggested. The balance of evidence modestly favors a real but attenuated effect, which is why regulators and guideline writers have not withdrawn their endorsements; the credible skeptical position is that the positive trials overestimated the effect and the true benefit may be negligible. Further large randomized data, or a convincing account of why the positive and null trials differ (dosing onset, patient selection, background therapy, adherence), would move the question toward resolution.
Claim entered the graph