Long-term colchicine for the prevention of vascular recurrent events in non-cardioembolic stroke (CONVINCE): a randomised controlled trial
Peter Kelly,Robin Lemmens,Christian Weimar,Cathal Walsh,Francisco Purroy,Mark Barber,Ronan Collins,Simon Cronin,Anna Czlonkowska,Philippe Desfontaines,Adinda De Pauw,Nicholas Richard Evans,Urs Fischer,Catarina Fonseca,John Forbes,Michael D Hill,Dalius Jatuzis,Janika Kõrv,Peter Kraft,Christina Kruuse,Catherine Lynch,Dominick McCabe,Robert Mikulik,Sean Murphy,Paul Nederkoorn,Martin O'Donnell,Peter Sandercock,Bernadette Schroeder,Gek Shim,Katrina Tobin,David J Williams,Christopher Price
DOI: https://doi.org/10.1016/S0140-6736(24)00968-1
2024-07-13
Abstract:Background: Anti-inflammatory therapy with long-term colchicine prevented vascular recurrence in coronary disease. Unlike coronary disease, which is typically caused by atherosclerosis, ischaemic stroke is caused by diverse mechanisms including atherosclerosis and small vessel disease or is frequently due to an unknown cause. We aimed to investigate the hypothesis that long-term colchicine would reduce recurrent events after ischaemic stroke. Methods: We did a randomised, parallel-group, open-label, blinded endpoint assessed trial comparing long-term colchicine (0·5 mg orally per day) plus guideline-based usual care with usual care only. Hospital-based patients with non-severe, non-cardioembolic ischaemic stroke or high-risk transient ischaemic attack were eligible. The primary endpoint was a composite of first fatal or non-fatal recurrent ischaemic stroke, myocardial infarction, cardiac arrest, or hospitalisation (defined as an admission to an inpatient unit or a visit to an emergency department that resulted in at least a 24 h stay [or a change in calendar date if the hospital admission or discharge times were not available]) for unstable angina. The p value for significance was 0·048 to adjust for two prespecified interim analyses conducted by the data monitoring committee, for which the steering committee and trial investigators remained blinded. The trial was registered at ClinicalTrials.gov (NCT02898610) and is completed. Findings: 3154 patients were randomly assigned between Dec 19, 2016, and Nov 21, 2022, with the last follow-up on Jan 31, 2024. The trial finished before the anticipated number of outcomes was accrued (367 outcomes planned) due to budget constraints attributable to the COVID-19 pandemic. Ten patients withdrew consent for analysis of their data, leaving 3144 patients in the intention-to-treat analysis: 1569 (colchicine and usual care) and 1575 (usual care alone). A primary endpoint occurred in 338 patients, 153 (9·8%) of 1569 patients allocated to colchicine and usual care and 185 (11·7%) of 1575 patients allocated to usual care alone (incidence rates 3·32 vs 3·92 per 100 person-years, hazard ratio 0·84; 95% CI 0·68-1·05, p=0·12). Although no between-group difference in C-reactive protein (CRP) was observed at baseline, patients treated with colchicine had lower CRP at 28 days and at 1, 2, and 3 years (p<0·05 for all timepoints). The rates of serious adverse events were similar in both groups. Interpretation: Although no statistically significant benefit was observed on the primary intention-to-treat analysis, the findings provide new evidence supporting the rationale for anti-inflammatory therapy in further randomised trials. Funding: Health Research Board Ireland, Deutsche Forschungsgemeinschaft (German Research Foundation), and Fonds Wetenschappelijk Onderzoek Vlaanderen (Research Foundation Flanders), Belgium.