Abstract
Background
Data on the differential impact of interventions on subsequent myocardial infarction (MI) type are limited.
Objectives
This post-hoc analysis was done to evaluate the 30-day rate of subsequent MI by type (ie, type 1 and 2) among patients enrolled in the MINT (Myocardial Ischemia and Transfusion; NCT02981407) trial.
Methods
Subdistribution HRs and cumulative incidences of subsequent MI types were computed using Fine-Gray subdistribution models that accounted for the competing risk of death and other MI types, if applicable. Effect modification of treatment strategy by index MI type was tested using log-binomial regression models.
Results
Among 3,504 MINT trial patients, 275 (7.8%) had a 30-day subsequent MI, of which 118 (43%) were type 2 MI, 79 (28%) were uncertain MI type, 40 (15%) were type 4 MI, and 38 (14%) were type 1 MI. The rate of subsequent type 2 MI in patients randomized to the restrictive vs liberal transfusion was 3.5% (n = 61) vs 3.2% (n = 57) (HR: 1.07; 95% CI: 0.74-1.53) as compared with a subsequent type 1 MI rate of 1.3% (n = 23) vs 0.9% (n = 15) (HR: 1.53; 95% CI: 0.80-2.94).
Conclusions
Among patients with MI and anemia, subsequent MI occurred within 30 days in 7.8% of patients, with type 2 MI occurring 3 times more often than type 1 MI. A differential effect of the restrictive vs liberal transfusion strategy on the type of subsequent MI (eg, type 1 vs type 2) was not observed.
Key words: acute coronary syndrome, blood transfusion, critical care cardiology, myocardial infarction, type 1 myocardial infarction, type 2 myocardial infarction
Central Illustration
Reflecting the greater understanding of the diverse causes of myocardial infarction (MI), the universal definition of MI defines 5 etiologically distinct types of acute MI.1,2 The 2 most common are type 1 MI, resulting from a coronary thrombus overlying a disrupted atherosclerotic plaque (eg, rupture or erosion) and type 2 MI, resulting from a mismatch in myocardial oxygen supply and demand unrelated to atherothrombosis. Despite differences in underlying pathophysiology between type 1 and type 2 MI, most clinical trials have either selectively enrolled type 1 MI or not differentiated the type of MI enrolled or classified the type of MI as an outcome variable. Therefore, the evaluation of the effectiveness of therapies on subsequent MI type is limited.3
The MINT (Myocardial Ischemia and Transfusion) trial4 randomized patients with acute MI (including type 1 and type 2) and a hemoglobin concentration <10 g/dL to a restrictive or a liberal transfusion strategy. The primary outcome was a composite of all-cause death or MI up to 30 days following randomization4 and subsequent MIs were adjudicated, including MI type, using Universal Definition of MI criteria and taxonomy.1
The primary results of the MINT trial demonstrated that acute MI patients randomized to a restrictive vs liberal transfusion strategy had a relative risk of death or subsequent MI of 1.15 (95% CI: 0.99 to 1.34; P = 0.07), suggesting potential harm with a restrictive strategy.4 A prespecified subgroup analysis found the concern for harm with a restrictive transfusion strategy to be more apparent in patients with index type 1 than type 2 MI.5
In the current post hoc analysis, we evaluated the types of first subsequent MI (type 1 vs 2) that occurred among patients with MI and anemia. We compared whether a restrictive vs liberal transfusion strategy at the time of index MI differentially impacted subsequent MI type. Given the high rate of noncardiac comorbidities among patients with MI and anemia, we hypothesized that subsequent type 2 MI would be more common than type 1 in patients with index MI and anemia. Furthermore, based on MI pathophysiology in the setting of anemia, we hypothesized that the benefits of a liberal transfusion strategy would be most evident in the prevention of subsequent type 2 MI vs other MI types.
Methods
The methods and main findings of the MINT trial (NCT2981407) have been previously published.4,6 In summary, patients experiencing acute MI accompanied by anemia were randomly assigned to receive either a restrictive or a liberal red blood cell transfusion strategy (Central Illustration). In the restrictive transfusion group, transfusions were allowed, but not mandated, when hemoglobin levels dropped below 8 g/dL, and were strongly encouraged if levels fell below 7 g/dL. In contrast, the liberal group received transfusions to maintain hemoglobin concentrations at or above 10 g/dL. The assigned transfusion strategy was implemented until either discharge from the index hospitalization or 30 days postrandomization, whichever came first. The assigned transfusion strategy was not blinded to patients or clinical staff.
Central Illustration.
Subsequent Myocardial Infarction Type Among Patients Treated With Different Transfusion Strategies
Hb = hemoglobin; MI = myocardial infarction.
The trial was carried out at 144 clinical sites across the United States, Canada, France, Brazil, New Zealand, and Australia from April 2017 through April 2023. Institutional review board or ethics committee approval was obtained at each participating center and all patients provided informed consent before enrollment. An independent data and safety monitoring board conducted unblinded safety reviews biannually and performed protocol-defined interim efficacy analyses annually.
The MINT trial enrolled patients 18 years or older who presented with an acute MI classified as type 1, 2, 4b, or 4c, according to the Third Universal Definition of MI.1 Eligibility also required the presence of anemia, defined as a hemoglobin concentration below 10 g/dL within 24 hours before randomization. In accordance with the study protocol, site investigators determined and classified the index MI using the Universal Definition criteria. Patients were excluded if they had active uncontrolled bleeding, were receiving palliative care only, were scheduled for cardiac surgery during the index admission, refused blood transfusion, had been previously enrolled in the MINT trial, or would not be able to complete 30 days of follow-up.
Patients were contacted at 30 days to assess vital status, readmission, emergency room visits, and clinical events. Study staff also reviewed medical records from the index hospitalization and all subsequent hospital or emergency room admissions to identify potential clinical events.
The primary endpoint of this analysis was the occurrence of a first subsequent MI within 30 days of randomization. A clinical events committee, blinded to treatment allocation, systematically reviewed all protocol-mandated (baseline, 12 hours, day 1, day 2, and day 3) and clinically obtained troponin measurements. The committee surveilled the troponin values to identify potential events. All surveillance detected and site-reported subsequent MI prompted the collection of relevant hospital records for centralized review. Each subsequent MI was adjudicated by the committee according to criteria from the third Universal Definition of MI, encompassing types 1 through 5, silent Q-wave MI, indeterminate type (nonprocedural), and MI of uncertain classification.
Statistical analysis
The MINT trial statistical plan has been previously described.4 For this post hoc evaluation, distributional assumptions were verified before analyses. For comparing descriptive quantities, summaries included mean ± SD, median (25th, 75th percentiles) which were tested with t-tests, Wilcoxon rank, chi-square or Fisher exact test, as appropriate. To determine if the effect of restrictive vs liberal strategies varied by index MI type, the assigned treatment, index MI type and their interaction were included in log-binomial models and appropriate contrasts were computed. Interactions were tested in log-binomial models to provide stable and interpretable effects of potential treatment heterogeneity on the relative risk scale.
Subdistribution HRs and cumulative incidences of subsequent MI types were computed using Fine-Gray subdistribution models that accounted for the competing risk of death and other MI types, if applicable. The proportional hazards assumption for each endpoint was tested by assessing a time-interaction with treatment arm in each model (P > 0.05). For patients with incomplete outcome data (n = 55, see Supplemental Figure 1, due to withdrawals or lost to follow-up after the index hospitalization), we assumed there were no unobserved events and censored patients at date of withdrawal or loss to follow-up, as specified in the study protocol. Interaction terms of index MI type and treatment assignment were specified in Fine-Gray models as a sensitivity analysis.
We use alpha = 0.05 and no multiple testing adjustments according to our statistical analysis plan. The MINT protocol and statistical analysis plan are available at NCT2981407. Analyses were performed using SAS (9.4, SAS Institute Inc) and R (version 4.3.1). GraphPad Prism (version 10.1.2) was used to generate figures (GraphPad Software, Inc).
Results
The MINT trial enrolled 3,504 patients; of whom, 1,460 (42%) were classified as type 1 MI, 1,955 (56%) were classified as type 2 MI, 22 (0.6%) type 4b, 12 (0.3%) type 4c, and 55 (1.6%) unknown index MI. The median follow-up time was 30 days (quartile 1: 30 days, quartile 3: 30 days). A total of 275 (7.8%) patients had subsequent MIs at 30 days. Compared with patients who did not have a subsequent MI (any type) at 30 days, those with a subsequent MI had a higher prevalence of a prior history of MI (before index MI), diabetes, and hypercholesterolemia (Table 1). We did not observe significant differences in index hospitalization characteristics, baseline hemoglobin, or creatinine values. Those who suffered a subsequent MI within 30 days were more likely to be discharged from the index hospitalization on a P2Y12 inhibitor and both aspirin and a P2Y12 inhibitor, but less likely to be discharged on an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker (Table 1).
Table 1.
Baseline Characteristics of Participants With and Without a Subsequent Myocardial Infarction at 30 Days of Follow-Up
| No Subsequent MI (n = 3,229) |
Subsequent MI (n = 275) |
P Value | |
|---|---|---|---|
| Age (y), mean (SD) | 72.1 (11.6) | 72.9 (11.2) | 0.261 |
| Sex, n (%) | |||
| Female | 1,467 (45.4%) | 126 (45.8%) | 0.902 |
| Male | 1,762 (54.6%) | 149 (54.2%) | |
| Enrollment location, n (%) | |||
| United States | 1,965 (60.9%) | 192 (69.8%) | 0.006 |
| Canada | 823 (25.5%) | 62 (22.5%) | |
| Europe | 313 (9.7%) | 10 (3.6%) | |
| South America | 96 (3.0%) | 9 (3.3%) | |
| New Zealand/Australia | 32 (1.0%) | 2 (0.7%) | |
| Race availablea, n (%) | 2,799 (86.7%) | 254 (92.4%) | 0.007 |
| Hispanicb, n (%) | 129 (4.6%) | 7 (2.8%) | 0.004 |
| Raceb, n (%) | |||
| White | 2,184 (78.0%) | 216 (85.0%) | 0.019 |
| Black | 391 (14.0%) | 28 (11.0%) | |
| Other | 224 (8.0%) | 10 (3.9%) | |
| BMI, kg/m2, mean (SD) | 28.6 (7.1) | 28.4 (6.2) | 0.698 |
| Tobacco smoker, n (%) | |||
| Never | 1,243 (38.5%) | 85 (30.9%) | 0.089 |
| Former | 1,300 (40.3%) | 128 (46.5%) | |
| Current | 503 (15.6%) | 45 (16.4%) | |
| Missing | 183 (5.7%) | 17 (6.2%) | |
| Medical history (before index hospitalization) | |||
| Myocardial infarction, n (%) | 1,031 (31.9%) | 107 (38.9%) | 0.018 |
| Percutaneous coronary intervention, n (%) | 1,088 (33.7%) | 112 (40.7%) | 0.055 |
| Coronary artery bypass grafting, n (%) | 690 (21.4%) | 72 (26.2%) | 0.063 |
| Heart failure, n (%) | 969 (30.0%) | 97 (35.3%) | 0.069 |
| Angiogram results available, n (%) | 1,595 (49.4%) | 143 (52.0%) | 0.461 |
| Number of vessels with >50% obstruction, n (%) | |||
| None | 120 (3.7%) | 3 (1.1%) | 0.033 |
| 1 | 418 (12.9%) | 35 (12.7%) | |
| 2 | 401 (12.4%) | 34 (12.4%) | |
| 3 | 600 (18.6%) | 68 (24.7%) | |
| Missing | 1,690 (52.3%) | 135 (49.1%) | |
| Left main disease, n (%) | |||
| Yes | 420 (13.0%) | 49 (17.8%) | 0.077 |
| No | 1,104 (34.2%) | 91 (33.1%) | |
| Missing | 1,705 (52.8%) | 135 (49.1%) | |
| LV ejection fraction available, n (%) | 2,348 (72.7%) | 210 (76.4%) | 0.191 |
| Most recent LV ejection fraction (%) within the past year, mean (SD) | 47.4 (13.5) | 47.4 (13.5) | 0.965 |
| LV ejection fraction: severity level, n (%) | |||
| 55% or greater (normal) | 1,105 (34.2%) | 84 (30.5%) | 0.082 |
| 45%-<55% (mild) | 578 (17.9%) | 63 (22.9%) | |
| 30%-<45% (moderate) | 747 (23.1%) | 60 (21.8%) | |
| <30% (severe) | 262 (8.1%) | 30 (10.9%) | |
| Missing | 537 (16.6%) | 38 (13.8%) | |
| Transient ischemic attack, n (%) | 562 (17.4%) | 56 (20.4%) | 0.217 |
| Atrial fibrillation, n (%) | 816 (25.3%) | 84 (30.5%) | 0.055 |
| Peripheral artery disease, n (%) | 651 (20.2%) | 69 (25.1%) | 0.052 |
| eGFR, n (%) | |||
| <30 | 727 (22.5%) | 75 (27.3%) | 0.213 |
| 30-59 | 915 (28.3%) | 83 (30.2%) | |
| ≥60 | 1,192 (36.9%) | 87 (31.6%) | |
| On dialysis at baseline | 385 (11.9%) | 30 (10.9%) | |
| Missing | 10 (0.3%) | 0 (0.00%) | |
| Diabetes, n (%) | 1,730 (53.6%) | 166 (60.4%) | 0.030 |
| Hypertension, n (%) | 2,734 (84.7%) | 242 (88.0%) | 0.138 |
| Hypercholesterolemia/hyperlipidemia, n (%) | 2,076 (64.3%) | 194 (70.5%) | 0.037 |
| COPD/asthma, n (%) | 764 (23.7%) | 76 (27.6%) | 0.138 |
| Cancer, n (%) | 715 (22.1%) | 54 (19.6%) | 0.335 |
| Anemia, n (%) | |||
| Acute | 987 (30.6%) | 91 (33.1%) | 0.633 |
| Chronic | 1,676 (51.9%) | 140 (50.9%) | |
| Missing | 566 (17.5%) | 44 (16.0%) | |
| Index MI characteristics | |||
| Type of MI, n (%) | |||
| Type 1 | 1,339 (41.5%) | 121 (44.0%) | 0.705 |
| Type 2 | 1,808 (56.0%) | 147 (53.5%) | |
| Type 4b | 21 (0.7%) | 1 (0.4%) | |
| Type 4c | 12 (0.4%) | 0 (0.00%) | |
| Unknown | 49 (1.5%) | 6 (2.2%) | |
| ST-segment elevation, n (%) | |||
| STEMI | 619 (19.2%) | 37 (13.5%) | 0.020 |
| NSTEMI | 2,610 (80.8%) | 238 (86.6%) | |
| Index hospitalization, prerandomization | |||
| Number of days between index MI and randomization, median (IQR) | 2 (1, 4) | 2 (1, 4) | 0.638 |
| Diagnosis of MI at time of randomization confirmed at time of discharge, n (%) | |||
| Yes | 3,072 (95.1%) | 267 (97.1%) | 0.297 |
| No | 149 (4.6%) | 8 (2.9%) | |
| Missing | 8 (0.2%) | 0 (0.00%) | |
| Percutaneous coronary intervention, n (%) | 984 (30.5%) | 72 (26.2%) | 0.136 |
| Coronary artery bypass grafting, n (%) | 11 (0.3%) | 0 (0.00%) | - |
| Heart failure, n (%) | 727 (22.5%) | 53 (19.3%) | 0.215 |
| Intubated on ventilator, n (%) | 441 (13.7%) | 40 (14.5%) | 0.681 |
| Active bleedingc, n (%) | 424 (13.1%) | 35 (12.7%) | 0.849 |
| Received red blood cell transfusion, n (%) | 1,131 (35.0%) | 106 (38.5%) | 0.241 |
| Number of units, median (IQR) | 2 (1, 3) | 2 (1, 2) | 0.762 |
| In ICU/coronary care unit at randomization, n (%) | 1,537 (47.6%) | 142 (51.6%) | 0.198 |
| Laboratory values | |||
| Baseline hemoglobin, g/dL, mean (SD) | 8.6 (0.8) | 8.6 (0.9) | 0.289 |
| Most recent creatinine prior to randomization, mg/dL, median (IQR) | 1.4 (0.9, 2.5) | 1.6 (1.0, 2.6) | 0.135 |
| Discharge medicationsd | n = 2,956 | n = 233 | |
| P2Y12 inhibitor, n (%) | 1,751 (59.2%) | 162 (69.5%) | 0.002 |
| Aspirin, n (%) | 2,270 (76.8%) | 190 (81.6%) | 0.096 |
| Dual antiplatelet therapye, n (%) | 1,513 (51.2%) | 140 (60.1%) | 0.009 |
| Oral anticoagulant, n (%) | 748 (25.3%) | 63 (27.0%) | 0.558 |
| Beta blocker, n (%) | 2,301 (77.8%) | 190 (81.6%) | 0.188 |
| Statin, n (%) | 2,525 (85.4%) | 206 (88.4%) | 0.210 |
| ACE inhibitor/ARB | 1,371 (46.4%) | 92 (39.5%) | 0.042 |
ACE = angiotensin-converting enzyme; ARB = angiotensin II receptor blocker; BMI = body mass index; COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; ICU = intensive care unit; LV = left ventricular; MI = myocardial infarction; NSTEMI = non–ST-segment elevation myocardial infarction; STEMI = ST-segment elevation myocardial infarction.
Reported among participants in the US, Canada, and New Zealand/Australia who self-reported race/ethnicity.
Among participants with race/ethnicity data available.
Sites reported clinical bleeding at the following possible locations: GI, genitourinary, intraocular, intramuscular with compartment syndrome, hematoma, intracranial, pericardial, retroperitoneal, intraspinal, intra-articular, and other.
Among those discharged.
Discharged with both aspirin and a P2Y12 inhibitor.
Classification of subsequent MI type revealed that 43% of all first subsequent MIs were type 2 MI (118/275), 29% were unknown MI type (79/275), 15% were type 4 MI (40/275), and 14% were type 1 MI (48/275) (Figure 1, Table 2). Baseline characteristics of patients with these types of subsequent MI differed and are detailed in Supplemental Table 1.
Figure 1.
30-Day Event Rate of Subsequent Myocardial Infarction Type
MI = myocardial infarction.
Table 2.
Rate of Subsequent MI Type Overall and by Arm
| Total |
Restrictive |
Liberal |
||||
|---|---|---|---|---|---|---|
| Per Participant (n = 3,504) |
Proportion of MI Events (n = 275) |
Per Participant (n = 1,749) |
Proportion of MI Events (n = 149) |
Per Participant (n = 1,755) |
Proportion of MI Events (n = 126) |
|
| Any MI | 275 (7.8) | - | 149 (8.5) | - | 126 (7.2) | - |
| Type 1 | 38 (1.1) | 14% | 23 (1.3) | 15% | 15 (0.9) | 12% |
| Type 2 | 118 (3.4) | 43% | 61 (3.5) | 41% | 57 (3.2) | 45% |
| Type 4 | 40 (1.1) | 15% | 21 (1.2) | 14% | 19 (1.1) | 15% |
| Unknown | 79 (2.3) | 29% | 44 (2.5) | 30% | 35 (2.0) | 28% |
Values are n (%) unless otherwise indicated. Type 4 MI combines type 4A (n = 28), 4 B/C (n = 12).
Nonprocedural MI with undetermined type 1 or 2 classified as unknown (n = 3).
Abbreviation as in Table 1.
Subsequent MI type by transfusion strategy
The proportion of patients with a subsequent MI among patients randomized to a restrictive vs liberal transfusion strategy was 8.5% vs 7.2% (HR: 1.19; 95% CI: 0.94-1.50) (Tables 2 and 3, Figure 2). The rate of subsequent type 2 MI was similar between patients randomized to the restrictive vs liberal transfusion strategy: 3.5% and 3.2% (HR: 1.07; 95% CI: 0.74-1.53) (Tables 2 and 3, Figure 2). Subsequent type 1 MI occurred in 1.3% vs 0.9% of patients randomized to a restrictive vs liberal transfusion strategy (HR: 1.53; 95% CI: 0.80-2.94) (Tables 2 and 3, Figure 2). Tests for a differential effect of transfusion strategy on subsequent MI type were not statistically significant (Pinteractions >0.05) (Table 2; Central Illustration). Results were similar when interaction terms were included in competing risk models.
Table 3.
Subdistribution HRs Comparing Restrictive vs Liberal Transfusion Strategies on Rate of Subsequent MI Type
| All |
Index Type 1 |
Index Type 2 |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Number of Events | Number of Competing Events | Subdistribution HR (95% CI) Restrictive vs Liberal |
Number of Events | Number of Competing Events | Subdistribution HR (95% CI) Restrictive vs Liberal |
Number of Events | Number of Competing Events | Subdistribution HR (95% CI) Restrictive vs Liberal |
|
| Any MI | 275 | 275 | 1.19 (0.94-1.50) | 121 | 113 | 1.30 (0.91-1.86) | 147 | 155 | 1.05 (0.76-1.45) |
| Type 1 | 38 | 512 | 1.53 (0.80-2.94) | 14 | 220 | 2.51 (0.79-8.00) | 24 | 278 | 1.20 (0.54-2.67) |
| Type 2 | 118 | 432 | 1.07 (0.74-1.53) | 44 | 190 | 1.10 (0.61-1.98) | 69 | 233 | 0.93 (0.58-1.49) |
| Type 4 | 40 | 510 | 1.10 (0.59-2.05) | 27 | 207 | 1.25 (0.58-2.66) | 13 | 289 | 0.87 (0.29-2.58) |
| Unknown Type | 79 | 471 | 1.25 (0.80-1.95) | 36 | 198 | 1.25 (0.65-2.40) | 41 | 261 | 1.29 (0.70-2.39) |
Competing risks include death and other outcome MI type events.
Abbreviation as in Table 1.
Figure 2.
Cumulative Incidence of Subsequent Myocardial Infarction by Randomized Transfusion Strategy
Abbreviation as in Figure 1.
Outcomes by MI type at enrollment
The rate of any subsequent MI was 8.3% (n = 121) among patients with an index type 1 MI vs 7.5% (n = 147) for index type 2 MI (Table 4, Figure 3). Subsequent type 2 MI was the most common subsequent MI type among patients enrolled with an index type 1 (3.0%) or type 2 (3.5%) MI. The proportion of patients with a subsequent type 1 MI was 1.0% and 1.2% among patients enrolled with an index type 1 vs type 2 MI (Table 4, Figure 3).
Table 4.
Rate of Subsequent MI Type by Randomized Transfusion Strategy and Index MI Type
| Index Type 1 MI (N = 1,460) |
Index Type 2 MI (N = 1,955) |
||||||
|---|---|---|---|---|---|---|---|
| Total | Restrictive (n = 730) |
Liberal (n = 730) |
Total | Restrictive (n = 967) |
Liberal (n = 988) |
Interaction P Valuea | |
| Any MI | 121 (8.3) | 68 (9.3) | 53 (7.3) | 147 (7.5) | 75 (7.8) | 72 (7.3) | 0.39 |
| Type 1 | 14 (1.0) | 10 (1.4) | 4 (0.5) | 24 (1.2) | 13 (1.3) | 11 (1.1) | 0.30 |
| Type 2 | 44 (3.0) | 23 (3.2) | 21 (2.9) | 69 (3.5) | 33 (3.4) | 36 (3.6) | 0.68 |
| Type 4 | 27 (1.8) | 15 (2.1) | 12 (1.6) | 13 (0.7) | 6 (0.6) | 7 (0.7) | 0.60 |
| Unknown | 36 (2.5) | 20 (2.7) | 16 (2.2) | 41 (2.1) | 23 (2.4) | 18 (1.8) | 0.92 |
Values are n (%). Type 4 MI combines type 4a, 4b, and 4c.
Nonprocedural MI with undetermined type 1 or 2 classified as unknown (n = 3).
Abbreviation as in Table 1.
Log-binomial regression model with interaction term for arm and index MI type.
Figure 3.
Subsequent Myocardial Infarction Type Stratified by Index Myocardial Infarction Type
Abbreviation as in Figure 1.
Impact of transfusion strategy by index MI type at enrollment on subsequent MI type
Among patients with an index type 1 MI randomized to a restrictive vs liberal transfusion strategy, the rate of subsequent type 1 MI was 1.4% vs 0.5% (HR: 2.51; 95% CI: 0.79-8.00) as compared to patients with an index type 2 MI, 1.3% vs 1.1% (HR: 1.20; 95% CI: 0.54-2.67) (Tables 3 and 4, Figure 4). Among patients with an index type 1 MI randomized to a restrictive vs liberal transfusion strategy, the rate of subsequent type 2 MI was 3.2% vs 2.9% (HR: 1.10; 95% CI: 0.61-1.97) as compared with patients with an index type 2 MI, 3.4% vs 3.6% (HR: 0.93; 95% CI: 0.58-1.49) (Tables 3 and 4, Figure 4).
Figure 4.
Subsequent Myocardial Infarction Type Stratified by Index Myocardial Infarction Type
Abbreviation as in Figure 1.
Discussion
Among patients with acute MI and anemia, 7.8% of patients experienced a 30-day subsequent MI. The rate of subsequent type 2 MI was ∼3-fold greater than that of type 1 MI at 30 days among patients with an index type 1 or type 2 MI and anemia, irrespective of randomization to restrictive or liberal transfusion strategy. Therefore, the potential benefit of a liberal vs restrictive transfusion strategy observed in the MINT trial4,7 appears to impact subsequent MI type (type 1 vs type 2) similarly, regardless of index MI type.
The predominance of subsequent type 2 vs type 1 MI in MINT is higher than 3 randomized control trials that classified outcome MI type. The FOURIER trial randomized patients with clinically evident cardiovascular disease to evolocumab or placebo; 68% of outcome MIs were type 1 and 15% were type 2 over 2.2 years of follow-up.8 In the ODYSSEY OUTCOMES trial, among patients with a recent acute coronary syndrome randomized to alirocumab or placebo, 66% of subsequent MIs were type 1 and only 21% were type 2 over 2.8 years of follow-up.9 In TRITON–TIMI 38, among patients with acute coronary syndrome randomized to prasugrel vs clopidogrel, type 4 (63%), followed by type 1 (33%), was the most common, with only 3.5% of outcome MIs classified as type 2 MI.10 MINT enrolled patients with both type 1 and type 2 MI and anemia as compared to the predominantly chronic cardiovascular disease population in FOURIER, the acute coronary syndrome patient population in TRITON–TIMI 38 and the ODYSSEY OUTCOMES trial. The inclusion criteria of MI with anemia in MINT are selective for a population with noncardiac comorbidities (eg, renal failure). Thus, the different study populations may have contributed to the differing proportions of subsequent type 1 and 2 MI observed in these trials.
Both FOURIER and TRITON–TIMI 38 observed a difference in the rate of subsequent MI type by the intervention being tested with both trial (lipid-lowering and antiplatelet) therapies reducing subsequent type 1 but not type 2 MIs. In contrast, the ODYSSEY OUTCOMES trial found no difference in the impact of lipid lowering therapy (alirocumab) on a reduction in subsequent type 1 MIs (HR: 0.87; 95% CI: 0.77-0.99; P = 0.032) and type 2 MIs (HR: 0.77; 0.61-0.97; P = 0.25).9 Therefore, the MINT trial was in contrast to FOURIER and TRITON–TIMI 38 but consistent with the ODYSSEY OUTCOMES trial with respect to a differential effect of a therapy on subsequent type 1 vs type 2 MI. Importantly, the patient populations, therapeutic interventions being tested, and duration of follow-up differed among these trials. MINT is unique in enrolling both type 1 and type 2 MI in a prospective therapeutic intervention trial. A larger numerical event rate reduction in subsequent type 1 MI was observed for patients with an index type 1 MI randomized to a liberal vs restrictive transfusion strategy as compared to subsequent type 2 MI; however, such a difference was not observed for patients with index type 2 MI for either subsequent MI type in the MINT trial.
These findings highlighted the importance of understanding outcome MI type in relation to the patient population being studied and the expected therapeutic effect of the intervention being tested. Therapeutics that target atherothrombosis would only be expected to have benefit in patients with MI from an atherothrombotic etiology (type 1) and prevent subsequent atherothrombotic (type 1) MIs. In contrast, therapeutics that target ischemia arising from any etiology may be expected to benefit patients with type 1 or type 2 MI and prevent subsequent type 1 and type 2 MI. Understanding for which MI type a therapy is effective and which subsequent MI type a therapy is potentially preventing will allow for the most judicious application of that therapeutic strategy.3 The MINT trial highlights the feasibility of studying the impact of therapeutics on subsequent type 1 and type 2 MI in multicenter, randomized controlled clinical trials.
For patients with acute coronary syndrome and MI, a liberal transfusion strategy targeting a hemoglobin level of ≥10 g/dL is a Class 2b recommendation in the 2025 American College of Cardiology (ACC), American Heart Association (AHA), American College of Emergency Physicians (ACEP), National Association of EMS Physicians (NAEMSP), and Society for Cardiovascular Angiography and Interventions (SCAI) consensus guideline.11 We have previously published data suggesting a greater absolute benefit of a liberal transfusion strategy in type 1 MI vs type 2 patients with anemia.5 We now present data to suggest that this benefit (liberal transfusion strategy) appears to result in effects comparable to the risk for both subsequent type 1 and type 2 MI.
Study limitations
The MINT trial began enrollment when the third, as opposed to the fourth, Universal Definition of MI was available.1,2 However, with the exception of the rare occurrence of spontaneous coronary artery dissection, the third and fourth Universal Definitions define type 1 and type 2 MI similarly. Therefore, the application of either definition is unlikely to change the classification of MI type in the MINT trial. The categorization of type 1 vs type 2 MI in a clinical trial is novel and data on precision are limited. Although some prior studies have shown limitations of classification of MI type by physicians when accessed from submitted billing codes,12 others have shown reassuring agreement in a clinical research setting similar to that of the MINT trial.13 The large proportion of subsequent MIs categorized as unknown type in MINT is consistent with an approach that aims to limit misclassification and prior work noting the difficulty in classifying MI types.14 Although the event committee reviewed hospital records, including imaging and procedure reports, primary imaging data were not visualized. Directionally, the relationship between randomization to restrictive vs liberal transfusion strategy and rate of first subsequent MI did not differ between unknown MI type and all other MI types examined (types 1, 2, and 4).
The MINT trial was powered for the primary outcome analysis of death or MI. Therefore, any subgroup analysis, including by type of MI, is likely underpowered. MI type 4 is further divided into type 4a (periprocedure MI), type 4b (acute stent thrombosis), and type 4c (stent restenosis), but the small number of recurrent MIs categorized as type 4 did not allow for analysis by type 4 subtype. Certain outcomes were rare within specific subgroups, such as subsequent type 1 MIs among participants with an index type 1 MI.
Conclusions
Among patients with an MI (type 1 or type 2) and anemia, the risk of a subsequent MI at 30 days was 3-fold more likely to be a type 2 vs type 1 MI. The potential benefit of a liberal transfusion strategy observed in the MINT trial4,7 does not appear to differ in its impact on subsequent MI types (eg, type 1 and type 2 MI).
Funding support and author disclosures
This trial is supported by National Heart, Lung, and Blood Institute (NHLBI) (U01 HL133817, U01HL132853) and the Canadian Blood Services and Canadian Institutes of Health Research (CIHR) Institute of Circulatory and Respiratory Health (grant 342193); For France, this work was in part supported by the RHU iVASC grant ‘#ANR-16-RHUS-00010’ from the French National Research Agency (ANR) as part of the “Investissements d’Avenir” program. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Dr DeFilippis reports research grants from NIH and Ionis Pharmaceutical; consulting for Velkor Inc, Tigermed, and Novo Nordisk. Dr Abbott reports consulting for Abbott, Medtronic, and Penumbra; research for Boston Scientific, Recor, Med Alliance, Shockwave, and Microport. Dr White reports grant support paid to the institution for the ODYSSEY OUTCOMES trial from Sanofi and Regeneron Pharmaceuticals, for the STRENGTH Trial from Omthera Pharmaceuticals, for the HEART-FID Study from American Regent, for the Dal GenE study from DalCor Pharma Uk Inc, for the AEGIS II Study from CSL Behring, for the Clear Outcomes Study from Esperion Therapeutics, for the SOLIST-WHF and SCORED studies from Sanofi Aventis Australia Pty Ltd, for the Librexia AF and ACS studies from Janssen, and for ISCHEMIA and the MINT studies from the National Institutes of Health; he also received personal fees as a steering committee member from DalCor Pharma UK, CSL Behring, Sanofi Australia Pty Ltd, Janessen and Esperion Therapeutics; and he was on advisory boards for CSL Behring and Genentech. Dr Lopes reports research grants or contracts from Amgen, Bristol-Myers Squibb, GlaxoSmithKline, Medtronic, Pfizer, andSanofi-Aventis; funding for educational activities or lectures from Pfizer, Daiichi Sankyo, and Novo Nordisk; and funding for consulting or other services from Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, and Novo Nordisk. Dr Alexander reports research grants to Duke University from Artivion/CryoLife, Bayer, Bristol-Myers Squibb, CSL Behring, Ferring, the U.S. FDA, Humacyte, and the U.S. NIH and advisory board or consulting payments from AbbVie, Artivion/CryoLife, AtriCure, Bayer, Bristol-Myers Squibb, Eli Lilly, Ferring, GlaxoSmithKline, Janssen, Novostia, Pfizer, Portola, Theravance, and Veralox. Dr Carson and Dr Hébert reports serving in Data Safety Monitoring Board for Cerus. Dr Goodman reports research grant support (eg, steering committee or data and safety monitoring committee) and/or speaker/consulting honoraria (eg, advisory boards) from Amgen, Anthos Therapeutics, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol Myers Squibb, CSL Behring, CYTE Ltd., Daiichi-Sankyo/American Regent, Eli Lilly, Esperion, Ferring Pharmaceuticals, HLS Therapeutics, Idorsia, JAMP Pharma, Merck, Novartis, Novo Nordisk A/C, Pendopharm/Pharmascience, Pfizer, Regeneron, Sanofi, Servier, Tolmar Pharmaceuticals, and Valeo Pharma; and salary support/honoraria from the Canadian Heart Failure Society, Canadian Heart Research Centre and MD Primer, Canadian VIGOUR Centre, Cleveland Clinic Coordinating Centre for Clinical Research, Duke Clinical Research Institute, New York University Clinical Coordinating Centre, PERFUSE Research Institute, Peter Munk Cardiac Centre Clinical Trials and Translation Unit, and TIMI Study Group (Brigham Health). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Dr Hébert states that the views in this article represent his own and not those of the Canadian Institutes of Health Research.
Acknowledgments
The authors thank the patients and study personnel for their contributions to this trial.
Footnotes
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the Author Center.
Appendix
For a supplemental table, please see the online version of this paper.
Appendix
References
- 1.Thygesen K., Alpert J.S., Jaffe A.S., et al. Third universal definition of myocardial infarction. Circulation. 2012;126:2020–2035. doi: 10.1161/CIR.0b013e31826e1058. [DOI] [PubMed] [Google Scholar]
- 2.Thygesen K., Alpert J.S., Jaffe A.S., et al. Fourth universal definition of myocardial infarction (2018) Circulation. 2018;138:e618–e651. doi: 10.1161/CIR.0000000000000617. [DOI] [PubMed] [Google Scholar]
- 3.DeFilippis A.P., Nasir K., Blaha M.J. Myocardial infarction as a clinical end point in research. Circ Res. 2019;124:1701–1703. doi: 10.1161/CIRCRESAHA.119.315101. [DOI] [PubMed] [Google Scholar]
- 4.Carson J.L., Brooks M.M., Hebert P.C., et al. Restrictive or liberal transfusion strategy in myocardial infarction and anemia. N Engl J Med. 2023;389:2446–2456. doi: 10.1056/NEJMoa2307983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.DeFilippis A.P., Abbott J.D., Herbert B.M., et al. Restrictive versus liberal transfusion in patients with type 1 or type 2 myocardial infarction: a prespecified analysis of the MINT trial. Circulation. 2024;150:1826–1836. doi: 10.1161/CIRCULATIONAHA.124.071208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Carson J.L., Brooks M.M., Chaitman B.R., et al. Rationale and design for the myocardial ischemia and transfusion (MINT) randomized clinical trial. Am Heart J. 2023;257:120–129. doi: 10.1016/j.ahj.2022.11.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Khan M.S., Spertus J.A., Chan P.S. Transfusion strategy in myocardial infarction and anemia. N Engl J Med. 2024;390:960–961. doi: 10.1056/NEJMc2400982. [DOI] [PubMed] [Google Scholar]
- 8.Wiviott S.D., Giugliano R.P., Morrow D.A., et al. Effect of evolocumab on type and size of subsequent myocardial infarction: a prespecified analysis of the FOURIER randomized clinical trial. JAMA Cardiol. 2020;5:787–793. doi: 10.1001/jamacardio.2020.0764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.White H.D., Steg P.G., Szarek M., et al. Effects of alirocumab on types of myocardial infarction: insights from the ODYSSEY OUTCOMES trial. Eur Heart J. 2019;40:2801–2809. doi: 10.1093/eurheartj/ehz299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Morrow D.A., Wiviott S.D., White H.D., et al. Effect of the novel thienopyridine prasugrel compared with clopidogrel on spontaneous and procedural myocardial infarction in the Trial to Assess Improvement in Therapeutic Outcomes by optimizing Platelet Inhibition with Prasugrel-Thrombolysis in Myocardial Infarction 38: an application of the classification system from the universal definition of myocardial infarction. Circulation. 2009;119:2758–2764. doi: 10.1161/CIRCULATIONAHA.108.833665. [DOI] [PubMed] [Google Scholar]
- 11.Rao S.V., O'Donoghue M.L., Ruel M., et al. 2025 ACC/AHA/ACEP/NAEMSP/SCAI Guideline for the management of patients with acute coronary syndromes: a report of the American College of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. J Am Coll Cardiol. 2025;85(22):2135–2237. doi: 10.1016/j.jacc.2024.11.009. [DOI] [PubMed] [Google Scholar]
- 12.McCarthy C., Murphy S., Cohen J.A., et al. Misclassification of myocardial injury as myocardial infarction: implications for assessing outcomes in value-based programs. JAMA Cardiol. 2019;4:460–464. doi: 10.1001/jamacardio.2019.0716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shah A.S., McAllister D.A., Mills R., et al. Sensitive troponin assay and the classification of myocardial infarction. Am J Med. 2015;128:493–501.e3. doi: 10.1016/j.amjmed.2014.10.056. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.DeFilippis A.P., Chapman A.R., Mills N.L., et al. Assessment and treatment of patients with type 2 myocardial infarction and acute nonischemic myocardial injury. Circulation. 2019;140:1661–1678. doi: 10.1161/CIRCULATIONAHA.119.040631. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.