The Korea Acute Myocardial Infarction Registry-National Institutes of Health (KAMIR-NIH) was designed to evaluate real-world practice and long-term clinical outcomes in patients with AMI between November 2011 and December 2015. The 20 tertiary university hospitals having facilities for PCI and onsite cardiac surgery consecutively enrolled patients. Data about baseline characteristics, laboratory findings, and cardiovascular outcomes were collected online by a clinical research coordinator. Out of 13,707 patients enrolled in the KAMIR-NIH, 6,366 consecutive patients diagnosed with STEMI within 48 hours of symptom onset were identified. Patients who had been treated with thrombolysis, had a symptom-to-door time (S2DT) > 48 hours, did not receive primary PCI, had multivessel disease, or were lost to follow-up at 6 months were excluded (Fig. 1). Consequently, a total of 2,920 patients with STEMI (60.8 ± 12.8 years; men, 80.1%) were analyzed and categorized into preprocedural TIMI flow grade 0 (n = 1,966), preprocedural TIMI flow grade 1-2 (n = 583), or preprocedural TIMI flow grade 3 (n = 371). PCI was performed according to standard guidelines. All patients received loading doses of aspirin (300 mg) and a P2Y₁₂ inhibitor (clopidogrel, 600 mg; ticagrelor, 180 mg; prasugrel, 60 mg), followed by corresponding maintenance doses for at least 1 year. Adjunctive drugs to support PCI, thrombus aspiration, and the use of intravascular imaging were left to the physician’s discretion. Medications such as statins, beta-blockers, and renin–angiotensin–aldosterone system blockers were prescribed as per the guidelines. The completeness of follow-up at 3 years ( ≥ 1,005 days) was 93.3% (2,725/2,920). The study protocol was approved by the ethics committee of each participating center. The present study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants.

The co-primary outcomes were all-cause mortality and major adverse cardio-cerebrovascular events (MACCE; a composite of all-cause mortality, nonfatal myocardial infarction, coronary revascularization by PCI or coronary artery bypass grafting, and nonfatal stroke) at 3 years. The secondary outcomes were individual components of MACCE at 3 years. All clinical events were identified via interview, chart review, or phone call by trained clinical research coordinators and confirmed by the principal investigator of each hospital. All data were collected in a web-based case report form. STEMI was diagnosed based on the presence of a new ST-segment elevation ≥ 0.1 mV in ≥ 2 contiguous leads ( ≥ 0.2 mV in V2 to V3 leads) or a new left bundle branch block with a concomitant elevated cardiac marker. Primary PCI was defined as when the procedure was performed as soon as possible after the diagnosis of STEMI in the index hospitalization. Off-hour admission was defined as arrival at the hospital during weekends, holidays, and night shifts (6 PM to 8 AM) on weekdays. The PCI was considered successful when residual stenosis was < 30% with final TIMI flow grade 2 or 3. Potent P2Y₁₂ inhibitors included ticagrelor and prasugrel.

Analyses were performed using R software version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Categorical and continuous variables are presented as the number of cases (percentages) and mean ± standard deviation or median (interquartile range), respectively. The categorical variables were compared using the chi-square test or Fisher’s exact test and continuous variables were compared using one-way analysis of variance or the Kruskal‒Wallis test. Kaplan‒Meier analysis of the outcomes according to the groups was performed using the log-rank test. Correlates of clinical outcomes were assessed using multivariable Cox regression analysis. Baseline variables with P < 0.1 in the univariable analysis and any other baseline variables judged to be of clinical relevance from previously published studies were included in the multivariable analysis after considering the assumption of proportionality and linearity of the Cox proportional hazards model. Specifically, these variables comprised preprocedural TIMI flow grade, age, sex, body weight, symptom-do-door time, off-hour presentation, Killip class, heart rate, systolic blood pressure, diabetes mellitus, hypertension, angina, previous myocardial infarction or revascularization, family history of premature coronary artery disease, current smoker status, previous cerebrovascular accident, anemia, creatinine clearance, total cholesterol, white blood cell count, and left anterior descending artery stenosis. The following variables with missing values were included in the multivariable analysis: systolic blood pressure (n = 6), heart rate (n = 6), body weight (n = 73), hemoglobin (n = 4), creatinine clearance (n = 4), and total cholesterol (n = 92). Missing values were imputed using multiple imputation method. The assumptions of proportionality and linearity were assessed using the log-minus-log plots and the cumulative sum of martingale-based residuals, respectively. Collinearity diagnostics were assessed using the variance inflation factor and eigensystem analysis among variables included in the multivariable Cox regression analysis. A two-sided P < 0.05 was considered to indicate statistical significance.

Compared with patients with preprocedural TIMI flow grade 3, those with preprocedural TIMI flow grade 0 were more likely to have longer S2DT, high body weight, high Killip class at admission, low left ventricular ejection fraction, high white blood cell count, high peak troponin I, and high low-density lipoprotein cholesterol levels (Table 1). Patients with preprocedural TIMI flow grade 0 were more likely to have shorter door-to-balloon time, right coronary artery culprit lesion, American College of Cardiology/American Heart Association lesion type B2/C, thrombus aspiration, and glycoprotein IIb/IIIa inhibitor use; they were less likely to have stenting, drug-eluting stent implantation, and postprocedural TIMI flow grade 3 (Table 2). There were no significant differences in medications at discharge among the groups.

Clinical outcomes were evaluated at 3 years (1,095 days; interquartile range: 1,057−1,095). Table 3 shows clinical outcomes up to 3 years according to preprocedural TIMI flow. There were no significant differences among the three groups regarding primary and secondary outcomes during the 3-year follow-up: 3.2% in preprocedural TIMI flow grade 0 vs. 2.9% in preprocedural TIMI flow grade 1 or 2 vs. 2.7% in preprocedural TIMI flow grade 3 for in-hospital mortality, P = 0.876; 8.1% vs. 7.9% vs. 5.9% for all-cause mortality at 3 years, P = 0.346, log-rank P = 0.161; and 14.1% vs. 15.1% vs. 13.7% for MACCE at 3 years, P = 0.795, log-rank P = 0.931. The cumulative time-to-event curves for all-cause mortality and MACCE among the three groups based on preprocedural TIMI flow grades over 3 years are shown in Fig. 2. For sensitivity analysis, Kaplan–Meier curves were generated for all-cause mortality and MACCE according to preprocedural TIMI flow grades in patients undergoing stenting or having postprocedural TIMI flow grade 3, which revealed no significant results (Online Fig. 1).

In a Cox time-to-event multivariable regression analysis with preprocedural TIMI flow grade 3 as the reference, preprocedural TIMI flow grade 0 was not independently associated with increased incidence of all-cause mortality (adjusted HR: 1.00; 95% CI: 0.64 to 1.57; P = 0.996) or MACCE (adjusted HR: 0.84; 95% CI: 0.62 to 1.14; P = 0.271) at 3 years (Fig. 3). The results of the stepwise Cox proportional hazard models for correlates of all-cause mortality and MACCE are presented in Online Tables 1 and 2, respectively. We performed several subgroup analyses to ensure the robustness of the results, which are based on the TIMI risk score, S2DT < 4 hours or ≥ 4 hours, infarct related artery, and another high-risk factors (Online Figs 2 to 5). There were no significant differences in all-cause mortality or MACCE at 3 years in the subgroup analyses.

Our study has several limitations. First, the KAMIR-NIH participating centers are all tertiary university hospitals with a higher patient volume, thus limiting the generalizability of the present findings. Second, we only included patients with STEMI and single obstructive coronary artery. It is also one of the strengths of our study to avoid bias associated with multivessel coronary disease and related procedures, which might not be addressed appropriately in the registry. Third, adjudication of the TIMI flow grades was performed by investigators at each participating centers without core laboratory confirmation.