Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Relation Between Previous Lipid-Lowering Therapy and Infarct Size (Creatine Kinase-MB Level) in Patients Presenting With Acute Myocardial Infarction Herbert D. Aronow, MD, MPHa,*, A. Michael Lincoff, MDb, Martin J. Quinn, MD, PhDc, A. Thomas McRae, MDd, Hitinder S. Gurm, MDe, Penny L. Houghtaling, MSb, Christopher B. Granger, MDf, Robert A. Harrington, MDf, Franz Van de Werf, MD, PhDg, Eric J. Topol, MDh, and Michael S. Lauer, MDi, on behalf of the GUSTO IIb and PURSUIT Trial Investigators Animal experimental data have shown that lipid-lowering agents reduce myocardial infarct size. This association has not been well studied in humans. We compared infarct size in 10,548 patients in the GUSTO IIb and PURSUIT trials who were (n ⴝ 1,028) or were not (n ⴝ 9,520) on lipid-lowering therapy before an enrolling myocardial infarction (MI). Patients using lipid-lowering agents before their index MI had smaller infarcts than those who were not using these agents (median peak creatine kinase [CK]-MB 4.2 vs 5.2 times the upper limit of normal [ULN]; p <0.0001). Similarly, in an unadjusted model, patients on previous lipid-lowering therapy were less likely to have a peak CK-MB >3 times the ULN (620 of 1,028 [60.3%] vs 6,486 of 9,520 patients [68.1%]; p <0.001; relative risk 0.88, 95% confidence interval 0.84 to 0.93, p <0.0001). In a covariate- and propensity-adjusted multivariable model, the association between pretreatment with lipid-lowering agents and smaller infarct size persisted (relative risk for CK-MB >3 times the ULN 0.94, 95% confidence interval 0.88 to 0.99, p ⴝ 0.04). In conclusion, patients on lipid-lowering agents before an MI had significantly smaller infarcts. These findings suggest that lipid-lowering therapy may exert additional salutary effects in the setting of acute coronary syndromes. © 2008 Elsevier Inc. All rights reserved. (Am J Cardiol 2008;102:1119 –1124) We sought to determine whether the use of lipid-lowering therapy before the development of an acute myocardial infarction (MI) was associated with less myocardial cell necrosis in patients in the Global Use of Streptokinase or t-PA for Occluded Coronary Arteries (GUSTO) IIb1 and Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor a Clinical Scholars Program, Michigan Heart and Vascular Institute at St. Joseph Mercy Hospital, Ann Arbor, Michigan; bCardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio; cSt. Vincents University Hospital, Dublin, Ireland; dCentennial Heart Cardiovascular Consultants, Nashville, Tennessee; eDivision of Cardiovascular Medicine, University of Michigan Medical School, Ann Arbor, Michigan; fDuke Clinical Research Institute, Durham, North Carolina; gDepartment of Cardiology, University Hospital Gasthuisberg, Leuven, Belgium; hDivision of Cardiovascular Diseases, Scripps Clinic, La Jolla, California; and iDivision of Prevention and Population Sciences, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. Manuscript received February 27, 2008; revised manuscript received and accepted June 17, 2008. Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy (PURSUIT) was supported by COR Therapeutics, Inc., San Francisco, California, and Schering-Plough Research Institute, Kenilworth, New Jersey; and Global Use of Streptokinase or t-PA for Occluded Coronary Arteries IIb (GUSTO IIb) was supported by CibaGeigy, Inc., Summit, New Jersey; Boehringer-Mannheim, Indianapolis, Indiana; and Guidant Corp., Redwood City, California. *Corresponding author: Tel: 734-712-8000; Fax: 734-712-8010. E-mail address: [email protected] (H. Aronow). 0002-9149/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2008.06.032 Suppression Using Integrilin Therapy (PURSUIT)2 studies. Both randomized controlled trials included patients with acute MI; GUSTO IIb enrolled patients with both ST-elevation and non–ST-elevation MI, whereas only those with non–ST-elevation MI were eligible for participation in PURSUIT. Methods Patients: GUSTO IIb enrolled 12,142 patients with STor non–ST-elevation acute coronary syndromes from 373 hospitals in 13 countries from May 1994 to October 1995. Patients were eligible for participation if they had experienced chest discomfort and ischemic electrocardiographic changes during the previous 12 hours. PURSUIT enrolled 10,948 patients with non–ST-elevation acute coronary syndromes from 726 centers in 28 countries from November 1995 to January 1997. Eligible patients had chest pain within the previous 24 hours and ischemic electrocardiographic changes within 12 hours of chest pain. Those who experienced an MI as their enrolling event, had available creatine kinase (CK)-MB data, and survived to hospital discharge (GUSTO IIb, n ⫽ 6,414; PURSUIT, n ⫽ 4,134) were included in this analysis. Treatment: In GUSTO IIb, patients were randomly assigned to an intravenous bolus and infusion of recombinant hirudin or unfractionated heparin for 3 to 5 days. Use of fibrinolytic therapy was left to the discretion of the treating www.AJConline.org 1120 The American Journal of Cardiology (www.AJConline.org) Table 1 Patient characteristics in the overall cohort (n ⫽ 10,548) Variable Median age (yrs) Median body mass index (kg/m2) Women Hypertension Hyperlipidemia Diabetes mellitus Current smoker Peripheral vascular disease Heart failure Previous MI Previous percutaneous coronary intervention Previous coronary artery bypass Previous angina pectoris Enrolling ST-elevation MI Enrolling non–ST-elevation MI Median systolic blood pressure (mm Hg) Rales Previous aspirin Previous angiotensin-converting enzyme inhibitor Previous  blocker Thrombolytic therapy GUSTO IIb trial PURSUIT trial Median propensity score Previous Lipid-Lowering Therapy No (n ⫽ 9,520) Yes (n ⫽ 1,028) p Value 64 (54–72) 26.6 (24.2–29.4) 2,502 (26%) 4,281 (45%) 3,138 (33%) 1,689 (18%) 3,440 (36%) 692 (7%) 587 (6%) 2,146 (23%) 567 (6%) 607 (6%) 5,894 (62%) 3,105 (33%) 6,415 (67%) 130 (118–148) 1,136 (12%) 5,996 (63%) 1,339 (14%) 2,283 (24%) 2,727 (29%) 5,917 (62%) 3,603 (38%) 0.012 (0.003–0.110) 63 (55–70) 27.5 (24.9–30.4) 300 (29%) 606 (59%) 955 (93%) 296 (29%) 273 (27%) 124 (12%) 94 (9%) 477 (46%) 213 (21%) 245 (24%) 795 (78%) 201 (20%) 827 (80%) 130 (116–146) 137 (13%) 794 (77%) 249 (24%) 464 (45%) 205 (20%) 497 (48%) 531 (52%) 0.287 (0.154–0.451) 0.034 ⬍0.001 0.045 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 — 0.34 0.19 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 — ⬍0.001 Values expressed as median (interquartile range) or number (percent). physician, except for patients randomly assigned to tissue plasminogen activator or primary angioplasty in the GUSTO IIb angioplasty substudy (n ⫽ 1,012). In PURSUIT, patients were randomly allocated to an intravenous bolus and infusion of either eptifibatide or placebo for up to 96 hours. In both studies, medications used before admission, including lipid-lowering agents, were captured on the trial case report form; lipid levels were not captured. Hyperlipidemia was defined as total cholesterol ⬎240 mg/dl in PURSUIT, but not strictly defined in GUSTO IIb. Outcomes: The primary end point in our study was infarct size, gauged by peak CK-MB. Peak CK-MB is highly correlated with anatomic infarct size3 irrespective of reperfusion therapy4 and appeared to more accurately reflect infarct size than CK alone.5 In PURSUIT, CK-MB was collected within 12 and 18 hours of enrollment for enrolling MI, within 24 hours of percutaneous coronary intervention for periprocedure MI, within 24 hours of coronary artery bypass grafting for perioperative MI, and within 18 hours for other MIs. In GUSTO IIb, CK-MB was collected at enrollment and hours 6 to 8 and 12 to 16. Peak CK-MB indicated the maximum CK-MB ratio (to the upper limit of normal) available for each patient at the listed times. CK assays were performed at respective institutions and institutional upper limits of normal (ULNs) were used for determination of abnormal values. CK-MB data were available for all except 323 patients in GUSTO IIb and 8 patients in PURSUIT. In a secondary analysis of all patients, including those who did not survive until hospital discharge (n ⫽ 11,038), we examined the incidence of in-hospital adverse clinical outcomes, including death, reinfarction, and devel- opment of heart failure according to previous use of lipidlowering therapy. Statistical analysis: Continuous data appear as mean ⫾ SD when normally distributed and median with interquartile range when non-Gaussian in distribution. Unpaired t tests and Mann-Whitney rank sum tests were used for bivariate analyses of normally and non-normally distributed continuous data, respectively. Categorical data were given as frequencies and percentages, and bivariate analyses of these data were performed using chi-square or Fisher’s exact tests, when appropriate. Prespecified subgroup analyses were performed in patients with and without age ⬎65 years, female gender, diabetes, hyperlipidemia, current smoking, ST elevation at presentation, and use of fibrinolytic therapy and by clinical trial. In observational studies, unlike in randomized controlled trials, treatment allocation is often biased by external factors. Before standard multivariable modeling of our primary outcome to adjust for measurable confounding, a propensity analysis was performed6 to minimize bias associated with the decision to initiate lipid-lowering therapy before hospital admission. The intent of propensity analyses is to mimic the conditions of a randomized trial such that patients are similar in every measurable respect except for treatment allocation.7 Propensity analysis was conducted as follows. First, the probability that each patient would be using lipidlowering therapy before hospital admission was predicted using a multivariable logistic regression model that included available demographic, clinical, and treatment characteristics (this probability is the propensity score). The ability of the regression model to discriminate between patients who Coronary Artery Disease/Lipid-Lowering Therapy and Infarct Size 1121 Figure 1. Relation between lipid-lowering therapy and myocardial infarct size (n ⫽ 10,548). LLA ⫽ lipid-lowering agent. were and were not receiving lipid-lowering therapy before admission was estimated using the c statistic. Second, the entire cohort was divided into propensity score deciles, and within each decile, patients were stratified according to whether they were or were not receiving lipidlowering therapy before admission. There are a number of available propensity score methods for reducing imbalances in measured characteristics between treatment groups, including matching and stratification. Whereas matching may result in greater variance, stratification can culminate in more bias.8 Stratification by decile was used to minimize loss of patients from the much smaller lipid-lowering therapy group. Propensity scores and clinical characteristics were then compared across treatment groups by propensity score decile. There were too few patients in propensity score deciles 1 to 4 to permit statistical comparisons of measured characteristics across treatment groups. Consequently, patients in these deciles were excluded from the propensityrestricted analysis. In deciles 5 to 10, differences in patient characteristics across treatment groups observed in the overall cohort were no longer statistically significant. Analyses were subsequently performed in the overall (i.e., all deciles included) and propensity-restricted (i.e., only the most closely matched deciles included) cohorts. Third, propensity scores were entered along with other potential confounders into multivariable logistic regression models predicting the independent effect of previous lipidlowering agent use on infarct size (defined as peak CK-MB ⬎3 vs ⱕ3 times the ULN). Statistical analyses were performed using the SAS system, version 8.02 (SAS Inc., Cary, North Carolina). Results Baseline characteristics: There were 10,548 patients with an enrolling MI and available CK-MB data in the GUSTO IIb and PURSUIT trials (6,414 and 4,134 patients, respectively). Baseline characteristics according to treatment group are listed in Table 1. Lipid-lowering therapy and infarct size: Patients using lipid-lowering therapy before their index MI had smaller infarcts (median peak CK-MB 4.2 vs 5.2 times the ULN; p ⬍0.0001) and were more likely to be enrolled with non– ST- than ST-elevation MI in GUSTO IIb (PURSUIT included only patients with non–ST-elevation MI). When infarct size was categorized according to the range of Figure 2. Subgroup analysis: Previous lipid-lowering therapy and risk of CK-MB ⬎3 times the ULN (n ⫽ 10,548). Abbreviation as in Figure 1. CK-MB increase, patients on lipid-lowering therapy were significantly more likely to have infarcts of smaller size (1 to 3 times the ULN), significantly less likely to have larger infarcts (⬎10 times the ULN), and equally likely to have infarcts of moderate size (3 to 5 and 5 to 10 times the ULN; Figure 1). When the population was dichotomized into those with CK-MB increases ⱕ3 vs ⬎3 times the ULN, bivariate predictors of larger infarcts included younger age, male gender, current smoking, absence of diabetes, and ST-elevation MI. Patients with larger infarcts were also less likely to have a history of previous angina, MI, or coronary revascularization (percutaneous transluminal coronary angioplasty or coronary artery bypass grafting surgery) or be using angiotensin-converting enzyme inhibitors,  blockers, or calcium channel blockers before admission. Patients who received lipid-lowering agents before their index MI were less likely to have infarcts with a peak CK-MB ⬎3 times the ULN (620 of 1,028 [60.3%] vs 6,486 of 9,520 patients [68.1%]; p ⬍0.001; relative risk 0.88, 95% confidence interval 0.84 to 0.93, p ⬍0.0001). The association between previous lipid-lowering therapy and infarct size persisted across nearly all predefined subgroups, although it appeared to be more robust in the GUSTO IIb than PURSUIT population (Figure 2). In the 41% of patients with information available for left ventricular function, average left ventricular ejection fraction was not statistically different between those on and not on lipid-lowering therapy before admission (53% vs 52%, p ⫽ 0.13). Propensity analysis: The multivariable logistic regression model of lipid-lowering therapy use before admission included the significant positive independent predictors of older age; white race/ethnicity; hyperlipidemia; diabetes; absence of current smoking; family history of premature coronary disease; previous MI; previous percutaneous transluminal coronary angioplasty or coronary artery bypass graft surgery; concomitant use of  blockers, angiotensinconverting enzyme inhibitors, or calcium channel blockers; enrollment in the PURSUIT trial; and enrollment in North America, Western Europe, or Australia/New Zealand (compared with Eastern Europe or Latin America). The c statistic was 0.89, indicating that the propensity model discriminated 1122 The American Journal of Cardiology (www.AJConline.org) Table 2 Infarct size models Patients All All (with propensity score only) All (with propensity score and covariates) Deciles 5–10 Deciles 5–10 (with propensity score only) Deciles 5–10 (with propensity score and covariates) RR (95% confidence interval) p Value 0.88 (0.84–0.93) 0.93 (0.88–0.98) 0.94 (0.88–0.99) ⬍0.0001 0.0095 0.0145 0.93 (0.88–0.98) 0.93 (0.87–0.99) 0.0067 0.0145 0.93 (0.87–0.99) 0.0153 RR comparing infarct size (CK-MB ⬎3 times the ULN) in patients using versus not using lipid-lowering therapy before admission. Figure 3. Select independent predictors of myocardial infarct size from propensity- and covariate-adjusted models (n ⫽ 10,548). SBP ⫽ systolic blood pressure. well between patients who did and did not receive lipidlowering therapy before admission. Multivariable analysis of lipid-lowering therapy and infarct size: Multivariable models of the association between previous lipid-lowering therapy and infarct size in overall and propensity-restricted cohorts are listed in Table 2. When propensity score was added to the unadjusted model described, the resulting RR estimate was larger, suggesting that the propensity score method avoided potential overestimation of the true relative risk reduction. RR estimates from a model adjusted for both propensity score and covariates in this cohort were similar. When the cohort was restricted to propensity score deciles in which patients were most closely matched (propensity score deciles 5 to 10, n ⫽ 6,371), RR estimates were similar in unadjusted, propensity-adjusted, or covariate- and propensity-adjusted models. Select independent predictors of infarct size from the propensity- and covariate-adjusted model in the overall cohort are shown in Figure 3. Lipid-lowering therapy and in-hospital clinical events: In-hospital adverse clinical event rates were analyzed for all patients, including those who did not survive their index hospitalization (n ⫽ 11,038). Comparing patients who did and did not receive lipid-lowering therapy before hospitalization, there were no statistically significant differences in incidences of in-hospital death (56 of 1,084 [5.2%] vs 434 of 9,954 patients [4.4%], p ⫽ 0.22), nonfatal reinfarction (63 of 1,084 [6.1%] vs 503 of 9,954 patients [5.4%], p ⫽ 0.31), or heart failure (62 of 1,084 [5.7%] vs 650 of 9,954 patients [6.5%], p ⫽ 0.30), respectively. Discussion Experimental data from animal models have shown that statins, when given at least several hours before the induction of ischemia and reperfusion injury, limit myocardial infarct size.9,10 However, very limited clinical data have addressed this issue. In 1 single-center study (n ⫽ 264), patients who received statins before or shortly after admission for acute MI had smaller infarcts, gauged by median peak CK (416 vs 699 U/L; p ⫽ 0.02), an effect that was reportedly more pronounced in patients using these agents before admission.11 In another single-center study (n ⫽ 253) of consecutive patients presenting for primary percutaneous coronary intervention, those using statins for ⱖ1 week before the MI had better myocardial perfusion, gauged by greater angiographic myocardial blush grade than control patients.12 The present study extends this observation to a much larger cohort of patients with both ST- and non–STelevation MI hospitalized in hundreds of medical centers worldwide. In contrast to a study from the National Registry of Myocardial Infarction,13 we did not observe a significant decrease in in-hospital clinical events in patients on lipidlowering therapy before admission. Reasons for this disparity are unclear, but may relate to the smaller number of events in our study or the use of nonstatin lipid-lowering agents by some of our patients. Nevertheless, the early, intermediate, and late benefits of lipid-lowering therapy after MI have borne out in numerous randomized controlled14,15 and observational16 studies. There are both cholesterol-independent and -dependent mechanisms that might account for statin-mediated reductions in infarct size. Statins ameliorate reperfusion injury by improving endothelial function independent of their lipidlowering capacity. They appear to do so in part by increasing myocardial endothelial nitric oxide synthase messenger RNA expression and endothelial nitric oxide synthase activity, and their infarct size–limiting effect was abolished when they were co-administered with the nitric oxide synthase inhibitor nitro L-arginine methyl ester (L-NAME) 10 or administered to nitric oxide synthase– deficient mice.9 Furthermore, in normocholesterolemic mice, statin withdrawal led to suppression of endothelial nitric oxide production.17 Statins may also improve endogenous fibrinolysis.18 Whether statins can effect smaller infarcts by modifying lipid levels is unclear. Hyperlipidemia (both acute and chronic) is associated with larger myocardial infarcts in animal experimental models.19,20 However, some animal data suggest that the myocardial infarct–sparing effect of statin therapy is exerted independently of lipid lowering.21 Although speculative, it is possible that any or all of these mechanisms might account for some of the reductions in myocardial infarct size observed in patients on lipid-lowering therapy. Irrespective of the responsible mechanism(s), if antecedent lipid-lowering therapy were associated with smaller infarcts, this finding would have important clinical implica- Coronary Artery Disease/Lipid-Lowering Therapy and Infarct Size tions. First, although we did not observe such a relation in our study, lipid-lowering therapy might positively influence the clinical course after MI. Data from the National Registry of Myocardial Infarction support this premise13; National Registry of Myocardial Infarction investigators observed that compared with discontinuation, continuation of statin therapy led to fewer in-hospital complications and deaths. In a study of statin users who presented within 24 hours of an acute ischemic stroke, those randomly assigned to statin therapy withdrawal had larger infarct volume and greater risk of death or dependency at 3 months.22 Our findings also may have implications regarding the underuse23 and cessation13 of lipid-lowering therapy in eligible patients with established atherosclerotic vascular disease. Patient24 and physician25 perceptions of lipid-lowering agent efficacy are important determinants of drug adherence. Nearly 1 in 5 patients who discontinued lipidlowering therapy did so because of a perceived lack of efficacy.25 Our findings suggest that even when lipid-lowering agents fail to prevent nonfatal adverse cardiovascular events, they may still confer important clinical benefits. These data add to the growing body of evidence supporting the role of lipid-lowering therapy in the setting of acute coronary syndromes.14-16,26 A number of important limitations warranted mention. First, our study was observational. Accordingly, our multivariable models may have incompletely accounted for confounding; some potential confounders (e.g., renal function) were not captured. Second, “healthy-user bias,” whereby patients using preventative therapies have better outcomes because of unmeasured variables that relate to a generally improved state of health, may in part explain the improved outcomes associated with a particular therapy (e.g., estrogen replacement therapy and vitamin E) in observational studies. Propensity scoring may not have completely eliminated this bias from our observed associations. Third, we were unable to discriminate between statin and nonstatin lipidlowering agent use on the trial case report forms and did not have information about statin type, dose, or duration of therapy. Nevertheless, most patients on ”lipid-lowering therapy” during trial enrollment likely were receiving statins,27,28 and nonstatin lipid-lowering agents have also been associated with decreases in MI.29 Finally, patients who did not survive the index hospitalization were excluded from the primary analysis (we did so because peak CK-MB data were not available in many of these patients). Nevertheless, it is unlikely that this exclusion influenced our results because in-hospital death rates in patients receiving lipid-lowering agents were similar to those in patients not receiving these agents before hospital admission. 1. The GUSTO IIb Investigators. A comparison of recombinant hirudin with heparin for the treatment of acute coronary syndromes. The Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) IIb investigators. N Engl J Med 1996;335:775–782. 2. The PURSUIT Trial Investigators. Inhibition of platelet glycoprotein IIb/IIIa with eptifibatide in patients with acute coronary syndromes. Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy. N Engl J Med 1998;339: 436 – 443. 3. Hackel DB, Reimer KA, Ideker RE, Mikat EM, Hartwell TD, Parker CB, Braunwald EB, Buja M, Gold HK, Jaffe AS, et al. Comparison of 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 1123 enzymatic and anatomic estimates of myocardial infarct size in man. Circulation 1984;70:824 – 835. Licka M, Zimmermann R, Zehelein J, Dengler TJ, Katus HA, Kubler W. Troponin T concentrations 72 hours after myocardial infarction as a serological estimate of infarct size. Heart 2002;87:520 –524. Roberts R, Ambos HD, Sobel BE. Estimation of infarct size with MB rather than total CK. Int J Cardiol 1983;2:479 – 492. Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997;127:757–763. Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol 1999;150:327–333. Austin PC, Mamdani MM. A comparison of propensity score methods: a case-study estimating the effectiveness of post-AMI statin use. Stat Med 2006;25:2084 –2106. Jones SP, Gibson MF, Rimmer DM III, Gibson TM, Sharp BR, Lefer DJ. Direct vascular and cardioprotective effects of rosuvastatin, a new HMG-CoA reductase inhibitor. J Am Coll Cardiol 2002;40:1172– 1178. Wolfrum S, Dendorfer A, Schutt M, Weidtmann B, Heep A, Tempel K, Klein HH, Dominiak P, Richardt G. Simvastatin acutely reduces myocardial reperfusion injury in vivo by activating the phosphatidylinositide 3-kinase/Akt pathway. J Cardiovasc Pharmacol 2004;44: 348 –355. Bybee KA, Kopecky SL, Williams BA, Murphy JG, Scott Wright R. Reduced creatine kinase release with statin use at the time of myocardial infarction. Int J Cardiol 2004;96:461– 466. Hoffmann R, Haager P, Suliman H, Christott P, Radke P, Blindt R, Kelm M. Effect of statin therapy before Q-wave myocardial infarction on myocardial perfusion. Am J Cardiol 2008;101:139 –143. Spencer FA, Fonarow GC, Frederick PD, Wright RS, Every N, Goldberg RJ, Gore JM, Dong W, Becker RC, French W. Early withdrawal of statin therapy in patients with non-ST-segment elevation myocardial infarction: National Registry of Myocardial Infarction. Arch Intern Med 2004;164:2162–2168. Schwartz GG, Olsson AG, Ezekowitz MD, Ganz P, Oliver MF, Waters D, Zeiher A, Chaitman BR, Leslie S, Stern T. Effects of atorvastatin on early recurrent ischemic events in acute coronary syndromes: the MIRACL study: a randomized controlled trial. JAMA 2001;285:1711– 1718. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R, Joyal SV, Hill KA, Pfeffer MA, Skene AM. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med 2004;350:1495–1504. Aronow HD, Topol EJ, Roe MT, Houghtaling PL, Wolski KE, Lincoff AM, Harrington RA, Califf RM, Ohman EM, Kleiman NS, et al. Effect of lipid-lowering therapy on early mortality after acute coronary syndromes: an observational study. Lancet 2001;357: 1063–1068. Laufs U, Endres M, Custodis F, Gertz K, Nickenig G, Liao JK, Bohm M. Suppression of endothelial nitric oxide production after withdrawal of statin treatment is mediated by negative feedback regulation of rho GTPase gene transcription. Circulation 2000;102: 3104 –3110. Seljeflot I, Tonstad S, Hjermann I, Arnesen H. Improved fibrinolysis after 1-year treatment with HMG CoA reductase inhibitors in patients with coronary heart disease. Thromb Res 2002;105:285–290. Hoshida S, Nishida M, Yamashita N, Igarashi J, Hori M, Kamada T, Kuzuya T, Tada M. Amelioration of severity of myocardial injury by a nitric oxide donor in rabbits fed a cholesterol-rich diet. J Am Coll Cardiol 1996;27:902–909. Hoshida S, Yamashita N, Otsu K, Kuzuya T, Hori M. Cholesterol feeding exacerbates myocardial injury in Zucker diabetic fatty rats. Am J Physiol Heart Circ Physiol 2000;278:H256 –H262. Bulhak AA, Gourine AV, Gonon AT, Sjoquist PO, Valen G, Pernow J. Oral pre-treatment with rosuvastatin protects porcine myocardium from ischaemia/reperfusion injury via a mechanism related to nitric oxide but not to serum cholesterol level. Acta Physiol Scand 2005; 183:151–159. Blanco M, Nombela F, Castellanos M, Rodriguez-Yanez M, GarciaGil M, Leira R, Lizasoain I, Serena J, Vivancos J, Moro MA, Davalos A, Castillo J. Statin treatment withdrawal in ischemic stroke: a controlled randomized study. Neurology 2007;69:904 –910. Spencer FA, Goldberg RJ, Gore JM, Fox KA, Avezum A, Agnelli G, Kritharides L, Anderson FA, Goodman SG, FitzGerald G, Allegrone J, 1124 The American Journal of Cardiology (www.AJConline.org) Brieger D. Comparison of utilization of statin therapy at hospital discharge and six-month outcomes in patients with an acute coronary syndrome and serum low-density lipoprotein ⬎ or ⫽ 100 mg/dl versus ⬍100 mg/dl. Am J Cardiol 2007;100:913–918. 24. Kiortsis DN, Giral P, Bruckert E, Turpin G. Factors associated with low compliance with lipid-lowering drugs in hyperlipidemic patients. J Clin Pharm Ther 2000;25:445– 451. 25. Simons LA, Levis G, Simons J. Apparent discontinuation rates in patients prescribed lipid-lowering drugs. Med J Aust 1996;164:208 –211. 26. de Lemos JA, Blazing MA, Wiviott SD, Lewis EF, Fox KA, White HD, Rouleau JL, Pedersen TR, Gardner LH, Mukherjee R, et al. Early intensive vs a delayed conservative simvastatin strategy in patients with acute coronary syndromes: phase Z of the A to Z trial. JAMA 2004;292:1307–1316. 27. Siegel D, Lopez J, Meier J. Use of cholesterol-lowering medications in the United States from 1991 to 1997. Am J Med 2000;108: 496 – 499. 28. EUROASPIRE Study Group. EUROASPIRE. A European Society of Cardiology survey of secondary prevention of coronary heart disease: principal results. European Action on Secondary Prevention through Intervention to Reduce Events. Eur Heart J 1997;18: 1569 –1582. 29. Rubins HB, Robins SJ, Collins D, Fye CL, Anderson JW, Elam MB, Faas FH, Linares E, Schaefer EJ, Schectman G, Wilt TJ, Wittes J. Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial Study Group. N Engl J Med 1999;341:410 – 418.