Original research•

Colchicine in acute myocardial infarction: cardiovascular events at 1-year follow up

Abstract

Objective In the COVERT-MI randomised placebo-controlled trial, oral administration of high-dose colchicine at the time of reperfusion and for 5 days in acute ST-elevated myocardial infarction did not reduce infarct size but was associated with a significant increase in left ventricular thrombus (LVT) in comparison to placebo. We aimed to assess the 1-year clinical outcomes of the study population.

Methods This study is a follow-up analysis of the COVERT-MI study on prespecified secondary clinical endpoints at 1 year. The primary endpoint of this study was a composite of major adverse cardiovascular events (MACEs), including all-cause death, acute coronary syndromes, heart failure events, ischaemic strokes, sustained ventricular arrhythmias and acute kidney injury at 1-year follow-up. The quality of life (QOL) and the drug therapy prescription were also assessed.

Results At 1 year, 192 patients (101 patients in the colchicine group, 91 in the placebo group) were followed up. Seventy-six (39.6%) MACEs were reported in the study population. There was no significant difference regarding the number of MACEs between groups: 36 (35.6%) in the colchicine group and 40 (44.1%) in the placebo group (p=0.3). There were no differences in the occurrence of ischaemic strokes between the colchicine group and the control group (3 (3%) vs 2 (2.2%), respectively, p=0.99). There was a trend towards fewer heart failure events in the colchicine group compared with the placebo group (12 (11.9%) vs 18 (19.8%), p=0.20). There was no significant difference in QOL scores at 1 year (75.8±15.7 vs 72.7±16.2 respectively, p=0.18).

Conclusions There was no significant difference between the colchicine and placebo groups at 1 year regarding MACEs, especially concerning deaths or ischaemic strokes. No excess of ischaemic adverse events was observed despite the initial increase in LVT in the colchicine group.

Trial registration number NCT0315681.

What is already known on this topic

In the randomised multicentre double-blinded COVERT-MI trial, a 5-day oral colchicine therapy compared with placebo did not show any difference in infarct size between study groups at 5 days and 3 months as assessed by cardiac magnetic resonance. However, there was an unexpected threefold increase in left ventricle thrombi incidence (LVT) in the colchicine group compared with placebo.

What this study adds

There was no significant difference between the colchicine and placebo groups at 1 year regarding major adverse cardiovascular events, especially concerning deaths or ischaemic strokes. No excess of ischaemic adverse events was observed despite the initial increase in LVT in the colchicine group.

How this study might affect research, practice or policy

This study does not support the use of a short-term colchicine treatment of 1 mg/day at the acute phase of myocardial infarction to reduce infarct size and improve outcomes. Further studies are needed to explore the potential benefit of anti-inflammatory treatment strategies to reduce ischemia-reperfusion myocardial damage.

Introduction

Myocardial infarction generates an inflammatory response,1 identified as a critical mediator of reperfusion injury in ST-elevation myocardial infarction (STEMI).2 3

Colchicine, a well-known anti-inflammatory drug,4 5 was tested in the randomised multicentre double-blinded COVERT-MI trial6 7 (Colchicine for Left Ventricular Remodeling Treatment in Acute Myocardial Infarction) with a short-term oral colchicine therapy compared with placebo. There was no difference in infarct size between study groups at 5 days and 3 months as assessed by cardiac magnetic resonance (CMR). However, there was an unexpected threefold increase in left ventricle thrombi incidence (LVT) in the colchicine group compared with placebo.

It is unclear whether the significant increase in LVT is associated with major adverse cardiovascular events (MACEs) and quality of life (QOL) in the longer term and if the absence of effect on infarct size translates into the absence of effect on any hard clinical endpoint.

Therefore, this research aimed to assess MACEs and QOL using the EQ-5D score at 1-year follow-up and compare them between the colchicine and placebo groups.

Methods

The study protocol was reported previously and compliant with the Consolidated Standards of Reporting Trials research reporting guidelines (online supplemental file 1).6

In summary, we included 192 adults with a first STEMI referred for primary percutaneous coronary intervention, from 20 July 2018, to 28 July 2020 in a multicentre, double-blinded placebo-controlled academic funded, prospective study. An institutional review committee approved the study (CPP: 17/084), and all patients gave written informed consent (ClinicalTrials.gov, NCT03156816).

The patients in the colchicine group received a 2 mg loading dose followed by 0.5 mg two times a day for 5 days. Colchicine treatment was stopped after the fifth day. The patients in the placebo group received a 2 mg loading dose of placebo followed by 0.5 mg two times a day for 5 days. Colchicine or placebo treatment was stopped after the fifth day.

At day 5, patients got a CMR scan to evaluate the infarct size (primary outcome of COVERT-MI trial).7

All the patients were followed up by phone, and trained research coordinators reviewed medical records to collect the 1 year occurrence of MACEs, medical prescriptions (ie, antiplatelet drug, anticoagulation drug, antihypertensive drug, diuretics, statin oral antidiabetic drug and insulin therapy). The QOL was also assessed at 1 year using the EQ-5D score.8

If the patient did not take the call, the reason was specified (ie, withdrawal, lost to follow-up, death since last contact)

Patients or the public were not involved in our research’s design, conduct, reporting, or dissemination plans.

MACEs were defined as all-cause death, acute coronary syndromes, heart failure events (new hospitalisation for heart failure), ischaemic strokes, sustained ventricular arrhythmias and acute kidney injury following the international definitions for each type of adverse event.9 10 The unexpected increased intraventricular thrombus incidence in the colchicine group compared with the control group was notified by our authorities as a serious adverse finding and was also considered as a MACE for the statistical analyses.

The primary prespecified endpoint for this follow-up study was the rate of MACEs at 1 year of follow-up. Secondary endpoints were the rate of each individual component of the primary endpoint and also the quality-of-life scores at 1 year compared between each study group.

Statistical analysis

Categorical variables are expressed as numbers (percentage), and quantitative variables as the mean±SD. Differences between groups were evaluated by the Wilcoxon rank-sum test for continuous variables and the Fisher exact test for proportions.

Survival curves were constructed using the Kaplan-Meier method. Exploratory survival curves considering the probability of heart failure events were also constructed. The statistical significance between groups was evaluated using the Log-rank test.

Statistical analyses were performed with R statistical software V.4.0.2 (22 June 2020) with two-tailed α set to 0.05.

Results

One hundred ninety-two patients (101 patients in the colchicine group, 91 in the placebo group) were followed up at 1 year, with a mean follow-up of 345±96 days. Six (3.1%) patients were lost to follow-up (3 (3%) in the colchicine group, 3 (3.3%) in the placebo group). Baseline characteristics in both groups were similar, with 19.5% women and a mean age of 60±10 years, 87% were Killip class I at admission, 13% were diabetic. Baseline characteristics are reported in table 1.

Table 1
|
Baseline characteristics

There were 18 (22.2%) LVT in the colchicine group versus 6 (7.4%) in the placebo group (p=0.01), as reported in the initial publication.7 All patients with LVT diagnosed at 5 days had a prescription of oral anticoagulant for at least 3 months, whatever the study group. At 3 months, five patients had a persistent LVT on the follow-up CMR. There was no reported persistent LV thrombus as assessed by routine echocardiography at 12 months follow-up.

There was no difference between the colchicine group and control group in drug therapy prescription at 1 year: antiplatelet drug (83 (95.4%) vs 83 (100%), respectively), oral anticoagulant (11 (12.6%) vs 9 (10.8%)), beta-blockers (74 (85.1%) vs 66 (79.5%)), angiotensin-converting enzyme inhibitors (65 (74.7%) vs 60 (72.3%)) and mineralocorticoid receptor antagonists (6 (6.9%) vs 3 (3.6%)).

No statistical difference was observed concerning statin (82 (94.3%) vs 77 (92.8%)), diuretics (11 (12.6%) vs 11 (13.3%)), oral antidiabetic drug (14 (16.1%) vs 9 (10.8%)) and insulin therapy (4 (4.7%) vs 2 (2.4%)).

There was no significant difference between groups regarding the QOL as assessed by the EQ-5D score at 1 year (75.8±15.7 vs 72.7±16.2 respectively, p=0.18).

At 1 year, 76 (39.6%) MACEs were reported in the study population, and 7 (3.6%) patients died within 1 year. There was no significant difference regarding the number of patients with at least one event between groups: 36 (35.6%) events in the colchicine group and 40 (44.1%) events in the placebo group (p=0.3). All-cause death rates in the colchicine and control groups were not statistically different (4 (4%) vs 3 (3%) respectively, p=0.99). All MACEs are reported in table 2.

Table 2
|
Major adverse cardiovascular events at 1 year in each group

Despite the significantly higher LVT incidence in the colchicine group compared with the placebo group, there were no significant differences in the occurrence of ischaemic strokes (3 (3%) vs 2 (2.2%), respectively, p=0.99) or acute coronary syndromes (p=0.78).

All events presented are reported as absolute number(n) and frequency in percentage (%). Heart failure events were defined as new hospitalisations for heart failure. Acute kidney injury was defined according to international definitions.9 Based on national health authorities’ decision, the left ventricular thrombus were considered as serious adverse events and are counted in the total number of events.

The survival curves by the Kaplan-Meier method reporting the probability of MACEs are presented in figure 1. There was no difference in MACE incidence at 1 year between the colchicine and the placebo groups (p=0.99 by log rank test).

Figure 1
Request permissionrequest permission icon
Figure 1

Kaplan-Meier curves for the cumulative probability of major cardiovascular adverse events up to 12 months among patients assigned to the colchicine or the placebo group.

Finally, the survival curves by the Kaplan-Meier method reporting the probability of heart failure events at 1 year are reported in figure 2. There was a non-significant trend towards fewer heart failure events in the colchicine group compared with the placebo group (12 (11.9%) vs 18 (19,8), p=0.10 by log rank test) (figure 2).

Figure 2
Request permissionrequest permission icon
Figure 2

Kaplan-Meier curves for the cumulative probability of heart failure events up to 12 months among patients assigned to the colchicine or the placebo group.

Discussion

In our study, a 5-day oral treatment with colchicine to reduce infarct size was not associated with significant differences in MACEs compared with placebo at 1 year. Despite a significant increase in the incidence of left ventricular thrombus associated with colchicine, this did not result in a significant increase in ischaemic strokes or acute coronary events. There was, however, a non-significant reduction in the incidence of heart failure events at 1 year in the colchicine group compared with placebo, which might be just related to chance.

In the COVERT-MI trial, an unexpected warning was issued regarding a significantly higher proportion of LVT in the colchicine group compared with the placebo group.7 No data in the literature reports such an association or any prothrombotic effect with colchicine. At a 1-year prospective follow-up with almost no patients lost to follow-up, there was no significant difference between groups regarding MACEs, especially concerning the number of deaths, ischaemic strokes or acute coronary syndromes. In terms of safety, this is important data, although all patients with LVT discovered during the study had an active prescription of oral anticoagulants for 3 months at least, which could have prevented any embolic ischaemic event.

Conversely to the COLCOT trial,11 where a long-term low-dose (0.5 mg daily) colchicine led to a significantly lower risk of ischaemic cardiovascular events than placebo in the following month of an acute myocardial infarction,11 our study with a higher dose (1 mg daily) but for only 5 days was not associated with any benefit on MACEs at 1 year. This difference does not surprise us, considering the duration of treatment and the low likelihood of any long-term effect related to this short acute treatment. In the COLCOT trial, the benefit was carried mainly on ischaemic events, especially ischaemic strokes, but there was no report on heart failure events.

Clinical and experimental studies have suggested a positive effect of colchicine at the acute phase of myocardial infarction with prolonged treatment on reducing adverse remodelling and, therefore, potential heart failure events.5 12 This was related to an anti-inflammatory effect of colchicine on MMP-9, NOX2 and TGF-beta1.5 The principal hypothesis in the COVERT-MI trial was to target the important proinflammatory reaction following ischaemic damage to the myocardium, which peaks 48 hours after reperfusion.2 3 13 Our study failed to demonstrate any significant effect on infarct size or adverse remodelling, with a low proportion of patients in both groups showing any adverse remodelling.7 Still, we were interested in exploring the heart failure outcomes in our study population. Our results show no significant effect on heart failure outcomes, and the trend in lower incidence of heart failure in the colchicine group is probably only related to chance. Further clinical studies, with different anti-inflammatory dosages and treatment durations, should be performed to assess this unmet and complex target of postreperfusion inflammation.

Study limitations

Our study has some limitations.

First, the main study, COVERT-MI, failed to demonstrate a difference concerning the infarct size between the colchicine and placebo groups. The follow-up at 1 year was prespecified and performed prospectively in the original version of the protocol. It is unlikely that a colchicine treatment, given during only 5 days without effect on infarct size, could impact hard cardiovascular outcomes at 1 year. Considering the anti-inflammatory effect of colchicine given at the acute phase and covering the inflammatory peak period after ischaemic injury, it was coherent to assess the heart failure outcomes at 1 year. The trend in difference of HF events between the colchicine group and placebo could hint in this direction with a lack of statistical power, but we think it is more related to chance.

However, special attention was needed for this time because of the unexpected increase in LV thrombus incidence in the colchicine group compared with the placebo group. One-year follow-up was mandatory to make sure there was no excess in ischaemic embolic events.

This study is a follow-up analysis of prespecified secondary clinical endpoints in a limited sample size. The small sample size limits the statistical power of all the comparisons performed. All results presented in this manuscript should be considered as hypothesis generating.

Article metrics
Altmetricopen-url
Dimensionsopen-url