Diabetes mellitus (DM) increases mortality in acute ST-segment elevation myocardial infarction (STEMI) but the responsible mechanism is not fully elucidated. We compared the rate of successful myocardial reperfusion measured by tissue myocardial perfusion grade (TMPG) and outcomes in patients with and without DM undergoing primary percutaneous coronary intervention (PCI) for STEMI. Patients enrolled in the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS AMI) trial were analyzed according to presence of DM with respect to TMPG after PCI and outcomes at 30 days and 3 years. Multivariable logistic regression was performed to identify the independent contribution to mortality of DM and TMPG and the interaction between the 2 was assessed. Complete data were available for 3,265 patients, of whom 533 (16.3%) had DM. Diabetic patients were significantly older and heavier and had more risk factors for coronary disease and more previous MI, revascularization, and heart failure. There were no differences in rates of Thrombolysis In Myocardial Infarction grade 3 flow after PCI in the infarct artery or TMPG 2/3 between patients with and without DM. Compared to nondiabetics, mortality was significantly higher at 30 days and at 3 years in the DM group (1.8% vs 4.5%, p = 0.0002 and 5.4% vs 11.0%, p <0.0001, respectively). DM and TMPG were significantly associated with 3-year mortality, but there was no statistical interaction between DM and TMPG (p = 0.70). In conclusion, DM is associated with a significantly higher risk of death but this association is not mediated by impaired epicardial or myocardial reperfusion.
Diabetic patients with acute ST-segment elevation myocardial infarction (STEMI) have at a least a twofold greater risk of early and late death compared to nondiabetic patients. Excess mortality is typically attributed to more extensive coronary artery disease, chronic kidney disease, and other co-morbidities. Reperfusion therapy with fibrinolysis or primary percutaneous coronary intervention (PCI) improves the outcome of patients with STEMI by decreasing the amount of lost myocardium, preserving antegrade flow in the infarct-related artery, and improving remodeling after MI. The quality of reperfusion can be judged by the briskness of flow in the infarct-related artery (Thrombolysis In Myocardial Infarction [TMI] grade) and by the tissue myocardial perfusion grade (TMPG), characterized by myocardial blush after reperfusion, which have been validated as a useful predictor of short- and long-term outcomes in randomized clinical trials and registries. It is not clear whether diabetes mellitus (DM) exerts its deleterious effect in patients with STEMI by jeopardizing reperfusion as judged by TIMI flow and TMPG or by other independent mechanisms.
Methods
The Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS AMI) trial has been described in detail. In brief, 3,602 patients presenting with STEMI within 12 hours of symptom onset were randomized before angiography to bivalirudin or heparin and a glycoprotein IIb/IIIa inhibitor (1:1 ratio). Subsequently, 3,006 patients suitable for stenting were randomized again to a drug-eluting stent or an identical bare-metal stent (3:1 ratio). DM was identified based on patient-supplied information and its management during hospitalization was left to the discretion of the site investigator. Independent adjudication of clinical events and angiographic core laboratory analysis were performed at the Cardiovascular Research Foundation (New York, New York). End points for the present study included 30-day and 3-year rates of death (all-cause and cardiac), reinfarction, stroke, and ischemia-driven target vessel revascularization. We also analyzed the incidence of stent thrombosis according to Academic Research Consortium definitions. TIMI flow rates in the infarct-related artery and dynamic TMPG were analyzed using standard definitions. TMPGs 0/1 and 2/3 were considered suboptimal and optimal myocardial reperfusion, respectively. Outcomes were then analyzed according to presence of DM and optimal TMPG.
Continuous variables are presented as median with interquartile range and were compared using the Wilcoxon rank-sum test. Categorical variables are presented as proportions and were compared using chi-square or Fischer’s exact test. A multivariable model was developed to identify independent predictors of 3-year all-cause death. Candidate variables (based on the overall HORIZONS AMI trial model for mortality) were age, gender, DM, baseline creatinine clearance, baseline hemoglobin, baseline white blood cell count, Killip classes II to IV (vs I), left anterior descending coronary artery stenosis >50%, time from symptom onset to PCI, left anterior descending coronary artery infarct-related artery, baseline TIMI flow grade 0/1, bivalirudin allocation, and final TMPG 2/3. The interaction between DM and TMPG was tested for its significance. Significance level was set at 0.05. All analyses were performed with SAS 9.0 (SAS Institute, Cary, North Carolina).
Results
Baseline characteristics of the 3,265 patients with complete angiographic data according to DM status are listed in Table 1 . The 533 diabetic patients (16.3%) were significantly older and heavier and had significantly more risk factors for coronary artery disease except for a lower incidence of smoking. They also had significantly more previous MI, revascularization, and heart failure than nondiabetic patients.
| Variable | DM | p Value | |
|---|---|---|---|
| Yes (n = 533) | No (n = 2,732) | ||
| Age (years) | 64.1 (55.9–71.8) | 59.4 (51.9–68.9) | <0.0001 |
| Age ≥65 years | 46.7% (249/533) | 34.1% (932/2,732) | <0.0001 |
| Men | 73.5% (392/533) | 77.9% (2,127/2,732) | 0.030 |
| White | 91.0% (485/533) | 94.7% (2,587/2,732) | 0.001 |
| Height (cm) | 172.0 (166.0–178.0) | 172.0 (166.0–178.0) | 0.75 |
| Subjects | 530 | 2,715 | |
| Weight (kg) | 85.0 (75.0–97.0) | 80.0 (70.0–90.0) | <0.0001 |
| Subjects | 533 | 2,727 | |
| Body mass index (kg/m 2 ) | 29.1 (25.8–32.0) | 26.8 (24.4–29.6) | <0.0001 |
| Subjects | 530 | 2,714 | |
| Cardiac medical history | |||
| Previous hypertension | 71.1% (379/533) | 48.6% (1,329/2,732) | <0.0001 |
| Previous hyperlipidemia | 59.7% (318/533) | 39.6% (1,082/2,732) | <0.0001 |
| Previous smoking | 57.3% (304/531) | 66.0% (1,795/2,719) | <0.0001 |
| Current | 36.5% (194/531) | 49.1% (1,334/2,719) | <0.0001 |
| Former | 20.7% (110/531) | 16.9% (460/2,719) | 0.04 |
| Previous diabetes mellitus | 100.0% (533/533) | 0.0% (0/2,732) | <0.0001 |
| Diet only | 19.1% (102/533) | 0.0% (0/2,732) | <0.0001 |
| Insulin | 27.2% (145/533) | 0.0% (0/2,732) | <0.0001 |
| Any oral treatment | 62.5% (333/533) | 0.0% (0/2,732) | <0.0001 |
| Oral treatment only | 52.9% (282/533) | 0.0% (0/2,732) | <0.0001 |
| Previous myocardial infarction | 16.1% (86/533) | 9.5% (260/2,732) | <0.0001 |
| Previous angioplasty | 15.4% (82/532) | 9.7% (265/2,732) | <0.0001 |
| Previous bypass surgery | 5.1% (27/533) | 2.1% (57/2,732) | <0.0001 |
| Family premature coronary disease | 25.3% (135/533) | 30.6% (836/2,732) | 0.02 |
| Previous angina pectoris | 25.0% (133/532) | 21.2% (580/2,732) | 0.05 |
| Previous heart failure | 4.7% (25/533) | 2.3% (62/2,732) | 0.002 |
| Killip class | |||
| I | 87.8% (466/531) | 92.2% (2,517/2,731) | 0.001 |
| II | 9.4% (50/531) | 6.4% (176/2,731) | 0.014 |
| III | 1.7% (9/531) | 0.7% (19/2,731) | 0.04 |
| IV | 1.1% (6/531) | 0.7% (19/2,731) | 0.28 |
| Previous renal insufficiency | 7.3% (39/533) | 2.0% (55/2,731) | <0.0001 |
| Current dialysis | 0.6% (3/533) | 0.1% (2/2,731) | 0.03 |
TIMI flow and TMPG before and after PCI in the 2 groups are presented in Table 2 . It is noteworthy that patients with DM had higher rates of a patent infarct-related artery and optimal TMPG before PCI than patients without DM and achieved similar final TIMI flows and TMPGs compared to nondiabetics despite a longer symptom to balloon time. Approximately 85% of patients had TIMI grade 3 flow in the infarct-related artery and 3/4 of all patients had an optimal TMPG after PCI.
| Variable | DM | p Value | |
|---|---|---|---|
| Yes | No | ||
| (n = 533) | (n = 2,732) | ||
| Symptom to balloon time (min) | 256.5 (175.5–382.0) | 217.0 (158.0–325.0) | <0.0001 |
| Preprocedure epicardial flow ⁎ | 0.07 | ||
| Grade 0/1 | 60.0% (319/532) | 64.3% (1,768/2,750) | |
| Grade 2 | 13.7% (73/532) | 13.9% (383/2,750) | |
| Grade 3 | 26.3% (140/532) | 21.8% (599/2,750) | |
| Preprocedure dynamic blush | 0.01 | ||
| Blush 0/1 | 78.6% (419/533) | 83.1% (2,271/2,732) | |
| Blush 2 | 14.6% (78/533) | 11.5% (313/2,732) | |
| Blush 3 | 6.8% (36/533) | 5.4% (148/2,732) | |
| Blush 2/3 | 21.4% (114/533) | 16.9% (461/2,732) | |
| Postprocedure final epicardial flow ⁎ | 0.91 | ||
| Grade 0/1 | 3.2% (17/530) | 3.0% (83/2,748) | |
| Grade 2 | 11.5% (61/530) | 11.0% (302/2,748) | |
| Grade 3 | 85.3% (452/530) | 86.0% (2,363/2,748) | |
| Postprocedure dynamic blush | 0.19 | ||
| Blush 0/1 | 22.5% (120/533) | 20.0% (547/2,732) | |
| Blush 2 | 20.3% (108/533) | 20.4% (556/2,732) | |
| Blush 3 | 57.2% (305/533) | 59.6% (1,629/2,732) | |
| Blush 2/3 | 77.5% (413/533) | 80.0% (2,185/2,732) | |
⁎ Epicardial flow includes worst grade in all vessels treated during the index procedure.
Compared to nondiabetics, patients with DM had a higher rate of death in-hospital (1.8% vs 3.4%, p = 0.018), at 30 days (1.8% vs 4.5%, p = 0.0002), at 1 year (3.3% vs 6.2%, p = 0.0009), at 2 years (4.3% vs 8.6%, p <0.0001), and at 3 years (5.4% vs 11.0%, p <0.0001). Cardiac and noncardiac deaths were more common in the DM group (p <0.0001 and p = 0.03, respectively). The 30-day and 3-year outcomes according to DM status and dynamic TMPG after PCI are presented in Tables 3 and 4 , respectively, and in Figure 1 . Rates of death, stroke, stent thrombosis, and major bleeding were higher in diabetic patients than in nondiabetic patients by a factor of 2 to 3 regardless of reperfusion status. There was no difference in rates of reinfarction.
| Outcome | DM (n = 533) | No DM (n = 2,732) | ||||
|---|---|---|---|---|---|---|
| Blush 0/1 (n = 120) | Blush 2/3 (n = 413) | p Value | Blush 0/1 (n = 547) | Blush 2/3 (n = 2,185) | p Value | |
| Death | 13.3% | 1.9% | <0.0001 | 5.5% | 0.9% | <0.0001 |
| Reinfarction | 1.7% | 2.7% | 0.59 | 2.1% | 1.7% | 0.58 |
| Death or reinfarction | 14.2% | 4.6% | 0.0002 | 6.8% | 2.4% | <0.0001 |
| Stroke | 2.6% | 0.5% | 0.034 | 0.7% | 0.4% | 0.23 |
| Target vessel revascularization | 2.6% | 3.4% | 0.70 | 3.0% | 2.2% | 0.26 |
| Stent thrombosis | 3.9% | 3.0% | 0.65 | 2.8% | 2.1% | 0.31 |
| Major bleeding | 15.6% | 8.7% | 0.030 | 11.2% | 5.9% | <0.0001 |
| Outcome | DM (n = 533) | No DM (n = 2,732) | ||||
|---|---|---|---|---|---|---|
| Blush 0/1 (n = 120) | Blush 2/3 (n = 413) | p Value | Blush 0/1 (n = 547) | Blush 2/3 (n = 2,185) | p Value | |
| Death | 19.3% | 8.6% | 0.0003 | 9.4% | 4.4% | <0.0001 |
| Reinfarction | 6.7% | 9.6% | 0.41 | 5.9% | 7.2% | 0.35 |
| Death or reinfarction | 23.5% | 16.3% | 0.033 | 14.2% | 10.9% | 0.01 |
| Stroke | 5.6% | 2.6% | 0.093 | 1.6% | 1.3% | 0.61 |
| Target vessel revascularization | 18.7% | 17.6% | 0.86 | 13.1% | 14.0% | 0.75 |
| Stent thrombosis | 7.2% | 6.2% | 0.73 | 4.3% | 4.9% | 0.65 |
| Major bleeding | 17.5% | 10.9% | 0.044 | 11.6% | 7.4% | 0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
