Highlights
-
•
Conflicting randomized control trial data exists for the use of colchicine for secondary prevention following acute myocardial infarction (AMI).
-
•
In a cohort study of >1,200 patients, uptake of colchicine for secondary prevention following AMI is low, at only 7%.
-
•
Prescription of colchicine following AMI was not associated with reductions in recurrent cardiovascular events in this real-world cohort.
Randomized controlled trials investigating colchicine for secondary prevention of cardiovascular events following acute myocardial infarction (AMI) have yielded conflicting results, and the real-world use and effectiveness of colchicine in this context remains unknown. As such, we sought to evaluate the use of colchicine following AMI in clinical practice and the associated outcomes. We performed a retrospective analysis of patients diagnosed with AMI and longitudinally followed in a large academic health system between 2018 and 2024 to describe the clinical use of colchicine for secondary prevention following AMI, as well as patient-level demographic and clinical characteristics associated with colchicine use. Next, using both multivariable logistic regression models with and without propensity matching, we examined the association between colchicine prescription following AMI and composite cardiovascular outcomes (comprised of recurrent AMI, any revascularization, stroke, and death). Kaplan-Meier Event-free Survival Analysis and Cox Proportional Hazards Models were performed. Of 1,796, 126 (7.0%) were prescribed colchicine after AMI. There was no association between use of colchicine and the composite cardiovascular events in either standard multivariable adjusted (Odds Ratio 1.00, 95% CI 0.66–1.50, p = 0.99) or propensity matched models (0.98, 0.57–1.66, p = 0.93). There was no difference in event-free survival between patients who were prescribed colchicine and those who were not. In summary, we report the first real-world data on the use and effectiveness of colchicine for prevention of cardiovascular events after AMI. Colchicine was infrequently prescribed for this indication and was not associated with lower rates of subsequent cardiovascular events.
Graphical Abstract
Following an acute myocardial infarction (AMI), patients remain at elevated risk for adverse cardiovascular events. Despite the adoption of appropriate pharmacotherapy to address traditional risk factors, there is significant residual risk of events such as recurrent AMI, urgent revascularization, stroke, and death, and there is a need for novel therapies to improve patient outcomes. Vascular inflammation has been hypothesized to be an important driver of the observed residual risk of atherothrombotic events. Anti-inflammatory agents, such as colchicine, have therefore been proposed as potential adjunctive pharmacotherapies for prevention of adverse cardiovascular events following AMI.
There have been 2 large randomized controlled trials, COLCOT (2019), and LoDoCo2 (2020) that demonstrated significant reductions in composite cardiovascular outcomes with the use of colchicine in patients with AMI and chronic coronary artery disease, respectively. Despite this accumulating data and subsequent support from guidelines and regulatory bodies, the uptake of colchicine prescription following AMI remains unclear, and real-world data describing its use or effectiveness outside of clinical trials is lacking. Further, the more recent CLEAR Trial (2024), which similarly randomized patients to colchicine or placebo following AMI, failed to demonstrate any difference in cardiovascular outcomes between groups. Given the conflicting data from the available randomized controlled trials and unclear penetration into clinical practice, we sought to evaluate the real-world use of colchicine among patients following AMI, as well as the clinical outcomes among those who do versus do not receive colchicine.
Methods
Due to the sensitive nature of the data collected for this study, requests to access the dataset from qualified researchers trained in protocols on the protection of human subjects may be sent to Cedars-Sinai Medical Center at biodatacore@cshs.org.
In a retrospective cohort analysis, we identified all adults diagnosed with AMI between 1/1/2018 and 12/5/2024 in our large academic healthcare system based on ICD-10 coding ( Supplementary Table 1 ). The study period was selected to include 1 year prior to the publication of the first large randomized controlled trial of colchicine in coronary artery disease, and to include the time period during which subsequent trials and guideline documents were released. Patients were eligible for study inclusion if they were (1) admitted to one of our 2 hospitals with a primary discharge code for AMI and were seen at least once in an outpatient clinic within 1 year of index AMI, or (2) presented to clinic with a new ICD-10 code for AMI (to capture AMIs that occurred at other institutions) and were seen again in clinic at least once more during the study period. For identified patients, we extracted demographic and clinical characteristics from the time of the index AMI, as well as the date of any subsequent AMI, coronary revascularization procedure, stroke, and death following index AMI. These outcomes were identified through the use of ICD-10 codes, documentation of coronary revascularization procedures in appropriate registries (NCDR-CathPCI and STS), and documentation of death in the electronic medical record ( Supplementary Table 1 ). Of note, we attempted to collect information on inflammatory markers such as CRP, however, found a 79% missingness rate; as such, inflammatory markers were not included in analyses. Patients were excluded if they had contraindications to colchicine use (documented colchicine allergy, irritable bowel syndrome, inflammatory bowel syndrome, chronic diarrhea, or prescription of strong CYP3A4 medications; Supplementary Table 1 ). The medical record was then reviewed for use of colchicine after AMI. The primary composite outcome was myocardial infarction, any coronary revascularization procedure (percutaneous coronary intervention or coronary artery bypass), stroke, and death. The secondary outcome was recurrent myocardial infarction, stroke, and death (i.e., revascularization procedures excluded). As some patients were on colchicine both prior to and after AMI, we also performed sensitivity analyses only on patients started on colchicine after index AMI.
Statistics
Demographic and clinical characteristics were analyzed overall and stratified by AMI diagnosis type (hospitalization vs outpatient coding). To compare demographic and clinical characteristics between groups, we analyzed normally distributed continuous variables using mean values with standard deviations and non‐normally distributed continuous variables using median and interquartile range, with t tests and nonparametric Kruskal–Wallis tests used for group comparisons, respectively. We analyzed the frequency distribution of categorical variables with chi squared tests used for group comparisons. A 2‐sided p-value < 0.05 was considered statistically significant, without correction for multiple testing. We used multivariable logistic regression to evaluate the association between demographic and clinical characteristics with the prescription of colchicine. We subsequently used univariable and multivariable logistic regression models to assess for associations between patient characteristics, including the prescription of colchicine, and the primary and secondary outcomes. We additionally assessed for associations between clinical characteristics, including the use of colchicine, and the primary composite outcome following propensity matching using 1-to-one nearest neighbor matching; following matching, a logistic regression model was used to generate the propensity score, including assessed demographic and clinical variables. We then performed multivariable adjusted logistic regression to assess risk of subsequent composite cardiovascular event following index AMI. Finally, we performed Kaplan-Meier Event-free Survival Analysis and Cox Proportional Hazards Models to assess for differences in time-to-event between patients who did vs did not receive a colchicine prescription. As all patients were required to have follow up in our health system, the last follow-up time was considered the time of first event or the end of the study period (12/5/2024), whichever occurred first. Analyses were then repeated for sensitivity testing of patients newly prescribed colchicine after their index AMI. All statistical testing was conducted using Stata version 18 (StataCorp, LLC, College Station, TX). This study was approved by the Cedars-Sinai IRB with a waiver for informed consent.
Results
A total of 1,796 patients met inclusion criteria with a mean age of 66.0 ± 14.2 years ( Table 1 ). The majority were male (70.4%) and non-Hispanic White (55.4%). Approximately a third of AMI presentations were for ST-elevation myocardial infarction, while the remainder were for non-ST-elevation myocardial infarction. The most common comorbid condition was hypertension (70.1%) followed by hyperlipidemia (64.5%) and diabetes mellitus (31.7%). Most patients were admitted to our hospital for their index AMI (n = 1,460, 81.3%) while the index AMI code was placed in the outpatient setting for the remainder (n = 336, 18.7%). Those included through the hospital pathway were slightly older (66.7 ± 14.1 vs 63.1 ± 14.3, p < 0.001), more frequently male (71.8% vs 64.0%, p = 0.004), and had higher rates of hypertension, diabetes mellitus, chronic kidney disease, prior stroke, prior PCI, and prior CABG compared to those identified through the outpatient pathway ( Table 1 ).
Table 1
Demographic and clinical characteristics of patients with a new diagnosis of acute myocardial infarction during the study period, overall and stratified by if their diagnosis was documented during an acute hospitalization or from an outpatient encounter
| Total (n=1,796) | Hospitalized (n=1,460) | Outpatient Code (n=336) | P-Value | |
|---|---|---|---|---|
| Age, mean (SD) | 66.0 (14.2) | 66.7 (14.1) | 63.1 (14.3) | <0.001 |
| Male Sex, n (%) | 1,264 (70.4%) | 1,049 (71.8%) | 215 (64.0%) | 0.004 |
| Race/Ethnicity, n (%) | 0.012 | |||
| Non-Hispanic White | 995 (55.4%) | 780 (53.4%) | 215 (64.0%) | |
| Non-Hispanic Black | 250 (13.9%) | 213 (14.6%) | 37 (11.0%) | |
| Asian | 169 (9.4%) | 142 (9.7%) | 27 (8.0%) | |
| Other/Unknown | 164 (9.1%) | 142 (9.7%) | 22 (6.5%) | |
| Hispanic | 218 (12.1%) | 183 (12.5%) | 35 (10.4%) | |
| AMI Type, n (%) | 0.27 | |||
| STEMI | 569 (31.7%) | 471 (32.3%) | 98 (29.2%) | |
| NSTEMI | 1,227 (68.3%) | 989 (67.7%) | 238 (70.8%) | |
| Cr Prior to Discharge, mean in mg/dL (SD) | 1.17 (0.90) | 1.18 (0.91) | 1.05 (0.75) | 0.1 |
| Statin Allergy, n (%) | 85 (4.7%) | 71 (4.9%) | 14 (4.2%) | 0.59 |
| Comorbidities, n (%) | ||||
| Diabetes Mellitus | 570 (31.7%) | 495 (33.9%) | 75 (22.3%) | <0.001 |
| Chronic Kidney Disease | 355 (19.8%) | 302 (20.7%) | 53 (15.8%) | 0.042 |
| Hyperlipidemia | 1,158 (64.5%) | 928 (63.6%) | 230 (68.5%) | 0.091 |
| Hypertension | 1,269 (70.7%) | 1,066 (73.0%) | 203 (60.4%) | <0.001 |
| Stroke | 319 (17.8%) | 273 (18.7%) | 46 (13.7%) | 0.03 |
| History PCI | 638 (35.5%) | 568 (38.9%) | 70 (20.8%) | <0.001 |
| History CABG | 19 (1.1%) | 16 (1.1%) | 3 (0.9%) | 0.74 |
| Pericarditis | 83 (4.6%) | 66 (4.5%) | 17 (5.1%) | 0.67 |
| Dressler’s Syndrome | 22 (1.2%) | 21 (1.4%) | 1 (0.3%) | 0.087 |
| Gout | 120 (6.7%) | 93 (6.4%) | 27 (8.0%) | 0.27 |
| Familial Mediterranean Fever | 1 (0.1%) | 1 (0.1%) | 0 (0.0%) | 0.63 |
| Paget’s Disease | 1 (0.1%) | 1 (0.1%) | 0 (0.0%) | 0.63 |
| Chronic Kidney Disease | 355 (19.8%) | 302 (20.7%) | 53 (15.8%) | 0.042 |
| CKD Stage | 0.14 | |||
| No CKD | 1,441 (80.2%) | 1,158 (79.3%) | 283 (84.2%) | |
| Stage 1 | 17 (0.9%) | 14 (1.0%) | 3 (0.9%) | |
| Stage 2 | 51 (2.8%) | 42 (2.9%) | 9 (2.7%) | |
| Stage 3 | 182 (10.1%) | 149 (10.2%) | 33 (9.8%) | |
| Stage 4 | 30 (1.7%) | 28 (1.9%) | 2 (0.6%) | |
| Stage 5 | 30 (1.7%) | 27 (1.8%) | 3 (0.9%) | |
| Unspecified CKD | 45 (2.5%) | 42 (2.9%) | 3 (0.9%) |
Use of colchicine after AMI ranged from 4.3% to 9.3% each calendar year during the study period ( Figure 1 ). A total of n = 126 (7.0%) of patients received colchicine across the entire study period, of whom n = 33 were already on colchicine prior to their index AMI. The median time to colchicine prescription after AMI was 7 [IQR-2 to 158] days overall and 29 [0 to 308] days among those newly initiated on colchicine following their index AMI. There was a higher proportion of male patients who received colchicine compared to those who did not (78.6% vs 69.8%, p = 0.037). Patients with a diagnosis of inflammatory conditions including pericarditis, Dressler’s Syndrome, and gout were also more frequently prescribed colchicine. There were no other significant differences in demographic or clinical characteristics between participants who were or were not prescribed colchicine ( Table 2 ).
Trending count of new acute myocardial infarction diagnoses and prescription of Colchicine by calendar year. Note 2024 extends until 12/5/2024. AMI = acute myocardial infarction.
Table 2
Characteristics of patients with a new diagnosis of acute myocardial infarction during the study period stratified by prescription of colchicine
| No colchicine | Colchicine | p-value | |
|---|---|---|---|
| (n = 1,670) | (n = 126) | ||
| Age, mean (SD) | 66.17 (14.15) | 63.94 (14.49) | 0.088 |
| Male sex, n (%) | 1,165 (69.8%) | 99 (78.6%) | 0.037 |
| Race/Ethnicity, n (%) | 0.7 | ||
| Non-Hispanic White | 925 (55.4%) | 70 (55.6%) | |
| Non-Hispanic Black | 229 (13.7%) | 21 (16.7%) | |
| Asian | 156 (9.3%) | 13 (10.3%) | |
| Other/Unknown | 153 (9.2%) | 11 (8.7%) | |
| Hispanic | 207 (12.4%) | 11 (8.7%) | |
| AMI type, n (%) | 0.44 | ||
| STEMI | 533 (31.9%) | 36 (28.6%) | |
| NSTEMI | 1,137 (68.1%) | 90 (71.4%) | |
| AMI group, n (%) | 0.57 | ||
| Hospitalized | 1,360 (81.4%) | 100 (79.4%) | |
| Outpatient code | 310 (18.6%) | 26 (20.6%) | |
| Cr prior to discharge, mean in mg/dL (SD) | 1.17 (0.92) | 1.15 (0.67) | 0.82 |
| Statin allergy, n (%) | 79 (4.7%) | 6 (4.8%) | 0.99 |
| Comorbidities, n (%) | |||
| Diabetes mellitus | 538 (32.2%) | 32 (25.4%) | 0.11 |
| Hyperlipidemia | 1,070 (64.1%) | 88 (69.8%) | 0.19 |
| Hypertension | 1,182 (70.8%) | 87 (69.0%) | 0.68 |
| Stroke | 298 (17.8%) | 21 (16.7%) | 0.74 |
| History PCI | 590 (35.3%) | 48 (38.1%) | 0.53 |
| History CABG | 19 (1.1%) | 0 (0.0%) | 0.23 |
| Pericarditis | 50 (3.0%) | 33 (26.2%) | < 0.001 |
| Dressler’s syndrome | 14 (0.8%) | 8 (6.3%) | < 0.001 |
| Gout | 88 (5.3%) | 32 (25.4%) | < 0.001 |
| Familial mediterranean fever | 0 (0.0%) | 1 (0.8%) | < 0.001 |
| Paget’s disease | 1 (0.1%) | 0 (0.0%) | 0.78 |
| Chronic kidney disease | 325 (19.5%) | 30 (23.8%) | 0.24 |
| CKD stage: | |||
| No CKD | 1,345 (80.5%) | 96 (76.2%) | 0.17 |
| Stage 1 | 16 (1.0%) | 1 (0.8%) | |
| Stage 2 | 44 (2.6%) | 7 (5.6%) | |
| Stage 3 | 171 (10.2%) | 11 (8.7%) | |
| Stage 4 | 28 (1.7%) | 2 (1.6%) | |
| Stage 5 | 28 (1.7%) | 2 (1.6%) | |
| Unspecified CKD | 38 (2.3%) | 7 (5.6%) | |
| Clinic outcomes/events, n (%) | 666 (39.9%) | 56 (44.4%) | 0.31 |
| AMI | 273 (16.3%) | 21 (16.7%) | 0.93 |
| PCI | 230 (13.8%) | 30 (23.8%) | 0.002 |
| CABG | 20 (1.2%) | 5 (4.0%) | 0.01 |
| Stroke | 195 (11.7%) | 13 (10.3%) | 0.65 |
| Death | 126 (7.5%) | 5 (4.0%) | 0.14 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
