Current Smoking and Prognosis After Acute ST-Segment Elevation Myocardial Infarction

Objectives The aim of this study was to mechanistically investigate associations among cigarette smoking, microvascular pathology, and longer term health outcomes in patients with acute ST-segment elevation myocardial infarction (MI). Background The pathophysiology of myocardial reperfusion injury and prognosis in smokers with acute ST-segment elevation MI is incompletely understood. Methods Patients were prospectively enrolled during emergency percutaneous coronary intervention. Microvascular function in the culprit artery was measured invasively. Contrast-enhanced magnetic resonance imaging (1.5-T) was performed 2 days and 6 months post-MI. Infarct size and microvascular obstruction were assessed using late gadolinium enhancement imaging. Myocardial hemorrhage was assessed with T2* mapping. Pre-specified endpoints included: 1) all-cause death or first heart failure hospitalization; and 2) cardiac death, nonfatal MI, or urgent coronary revascularization (major adverse cardiovascular events). Binary logistic regression (odds ratio [OR] with 95% confidence interval [CI]) with smoking status was used. Results In total, 324 patients with ST-segment elevation MI were enrolled (mean age 59 years, 73% men, 60% current smokers). Current smokers were younger (age 55 ± 11 years vs. 65 ± 10 years, p < 0.001), with fewer patients with hypertension (52 ± 27% vs. 53 ± 41%, p = 0.007). Smokers had better TIMI (Thrombolysis In Myocardial Infarction) flow grade (≥2 vs. ≤1, p = 0.024) and ST-segment resolution (none vs. partial vs. complete, p = 0.010) post–percutaneous coronary intervention. On day 1, smokers had higher circulating C-reactive protein, neutrophil, and monocyte levels. Two days post-MI, smoking independently predicted infarct zone hemorrhage (OR: 2.76; 95% CI: 1.42 to 5.37; p = 0.003). After a median follow-up period of 4 years, smoking independently predicted all-cause death or heart failure events (OR: 2.20; 95% CI: 1.07 to 4.54) and major adverse cardiovascular events (OR: 2.79; 95% CI: 2.30 to 5.99). Conclusions Smoking is associated with enhanced inflammation acutely, infarct-zone hemorrhage subsequently, and longer term adverse cardiac outcomes. Inflammation and irreversible myocardial hemorrhage post-MI represent mechanistic drivers for adverse long-term prognosis in smokers. (Detection and Significance of Heart Injury in ST Elevation Myocardial Infarction. [BHF MR-MI]; NCT02072850)

The pathophysiology of myocardial reperfusion injury in smokers with acute ST-segment elevation MI (STEMI) is incompletely understood. Higher epicardial coronary flow rates after primary percutaneous coronary intervention (PCI) have been reported in current smokers versus nonsmokers (10). Studies of smoking and microvascular pathology, using cardiac magnetic resonance (CMR), have reported conflicting results (12,16), specifically pertaining to myocardial hemorrhage (an independent predictor of adverse outcome post-MI) (17,18). The MRI method in previous studies was not specific for detecting myocardial hemorrhage, and information on health outcomes was limited to 12month follow-up (12) or was unavailable (16). T2* mapping has emerged as a specific technique for detecting myocardial hemorrhage post-MI (18), with potential to resolve this controversy. Further insights into the effect of smoking on myocardial reperfusion injury could be provided by invasive microcirculatory assessment in the culprit coronary artery, using index of microcirculatory resistance (IMR), which is associated with infarct size, left ventricular (LV) pathology, and health outcomes (19)(20)(21)(22)(23)(24).
We aimed to prospectively investigate associations between smoking, microvascular pathology and longer term health outcomes in patients with acute STEMI.
We hypothesized that current smoking in patients with acute STEMI would be associated with acute microvascular dysfunction revealed by IMR and progressive hemorrhagic transformation within the infarct zone revealed by T2* MRI 2 days post-MI. We also hypothesized that smoking would be an independent predictor of adverse longer term health outcomes.

METHODS STUDY POPULATION.
We performed a prospective cohort study at a regional cardiac center from July  (27) and frame count (28) were assessed at initial angiography and at the end of the procedure.
TIMI myocardial perfusion grade (29) was assessed at the end of the procedure (Supplemental Appendix).        Values are mean AE SD, n (%), or median (interquartile range). The p values were obtained from Student t-tests (t), Mann-Whitney U tests (MW), or Fisher exact tests. TIMI flow grades pre-and post-PCI were grouped as 0/1 versus 2/3 for this analysis. *Diabetes mellitus was defined as a history of diet-controlled or treated diabetes.
†Killip classification of heart failure after acute myocardial infarction: class I, no heart failure; class II, pulmonary rales or crepitations, a third heart sound, and elevated jugular venous pressure; class III, acute pulmonary edema; class IV, cardiogenic shock. ‡Multivessel coronary artery disease was defined according to the number of stenoses of at least 50% of the reference vessel diameter, by visual assessment, and whether or not there was left main stem involvement. §C-reactive protein was available in 316 subjects, and troponin T was available in 313 subjects. IMR was available in 283 subjects.

DISCUSSION
We have undertaken a large prospective study of smoking status, infarct pathophysiology, and longterm prognosis in patients with acute STEMI. We found that current smoking is associated with a more favorable cardiovascular risk profile at initial presentation (e.g. younger age, fewer patients with hypertension, and higher coronary flow grades at the end of primary PCI), reflecting better proce-   Values are mean AE SD, n (%), or median (interquartile range). Area at risk was measured using T2 mapping. The p values were obtained using Student t-tests (t), Mann-Whitney U tests (MW), or Fisher exact tests. LV ejection fraction was missing in 29 subjects at follow-up. LVEDV at follow-up was missing in 16 men and 8 women. T2* imaging for myocardial hemorrhage was available in 245 subjects.
Despite their younger age, the longer-term prognosis of smokers is worse than that of nonsmokers, including for cardiac events. Our findings should dispel the false notion of any favorable associations between smoking and prognosis after acute STEMI. Consistent with prior studies (4,7-12,16), we found that smoking was crudely associated with more favorable presenting characteristics (e.g., younger age) and procedure outcomes. In contrast to prior reports (7)(8)(9)(10)(11)(12), smoking was independently associated with increased risk for adverse longer term health outcomes.
Our results indicate distinct phases in the early course of MI in smokers ( Figure 3). We observed a reverse paradox in that smoking was associated with less reperfusion injury acutely, as revealed by angiography, electrocardiography, and invasive microcirculatory measurements using IMR, but 2 days later, myocardial hemorrhage was more pronounced (Figure 1), even after adjustment for confounding covariates.
We postulate the following explanations for these findings. First, smokers were younger, they had fewer risk factors for microvascular dysfunction (4,7-12), and they presented with anterior MI less often. These factors most likely explain why smokers had less reperfusion injury acutely. Given the harmful effects of smoking on vascular health (4), the reperfused microvessels in smokers may have reduced repair potential and thus greater susceptibility to progressive degradation within the infarct core in the days after reperfusion. Second, in a serial imaging study, we found that myocardial hemorrhage was preceded by MVO (18), as MVO is an upstream event that may resolve. Our results indicate that the progression to infarct zone hemorrhage, rather than recovery without hemorrhage, was more likely in smokers than nonsmokers. Even after accounting for infarct size, the association between smoking and infarct zone hemorrhage persisted, unlike for MVO. This observation is prognostically relevant because myocardial hemorrhage reflects irreversible tissue damage (17,18). A recent study that described the independent prognostic importance of MVO post-MI did not include information on myocardial hemorrhage (42).
Using contemporary, multiparametric cardiac magnetic resonance using T2* mapping, we have found that myocardial hemorrhage is a much stronger determinant of adverse prognosis than MVO (18).
The lack of an association between smoking and myocardial hemorrhage in prior studies (12) may have been related to the use of dark-blood T2-weighted