Urinary 8-iso-Prostaglandin F 2 as a Risk Marker in Patients With Coronary Heart Disease. A Matched Case-Control Study

Transcription

1 Urinary 8-iso-Prostaglandin F 2 as a Risk Marker in Patients With Coronary Heart Disease A Matched Case-Control Study Edzard Schwedhelm, PhD; Asja Bartling, MD; Henrike Lenzen, MD; Dimitrios Tsikas, PhD; Renke Maas, MD; Jens Brümmer, MD; Frank-Mathias Gutzki, Ing Chem; Jürgen Berger, PhD; Jürgen C. Frölich, MD, FRSM; Rainer H. Böger, MD Background Oxidative stress is involved in the pathophysiology of atherosclerosis, diabetes mellitus, hypertension, obesity, and cigarette smoking, all of these being risk factors for coronary heart disease (CHD). We tested the hypothesis that risk factors of CHD are associated with abundant systemic oxidative stress. Methods and Results We conducted a case-control study with 93 CHD patients and 93 control subjects frequencymatched by age and sex. Urinary excretion of the F 2 -isoprostane 8-iso-prostaglandin (PG) F 2 and its major urinary metabolite, 2,3-dinor-5,6-dihydro-8-iso-PGF 2, were measured by gas chromatography tandem mass spectrometry. Body mass index, systolic blood pressure, and C-reactive protein were elevated in CHD patients (P 0.01). Urinary 8-iso-PGF 2 and 2,3-dinor-5,6-dihydro-8-iso-PGF 2 also differed, from 77 (interquartile range, ) to 139 (93 231) pmol/mmol creatinine and from 120 (91 151) to 193 ( ) pmol/mmol in control subjects and case subjects, respectively (P 0.001). 8-iso-PGF 2 and its metabolite were highly correlated (Spearman s 0.664, P 0.001). HDL cholesterol was diminished in CHD patients (P 0.001). All of these characteristics predicted CHD in univariate analysis. In a multivariate model, the odds ratios were increased only for 8-iso-PGF 2 ( 131 pmol/mmol, P 0.001) and C-reactive protein ( 3 mg/l, P 0.01), ie, by 30.8 (95% CI, ) and 7.2 ( ), respectively. 8-iso-PGF 2 was found to be a novel marker in addition to known risk factors of CHD, ie, diabetes mellitus, hypercholesterolemia, hypertension, and smoking. Urinary excretion of 8-iso-PGF 2 correlated with the number of risk factors for all subjects (P for trend). Conclusions 8-iso-PGF 2 is a sensitive and independent risk marker of CHD. (Circulation. 2004;109: ) Key Words: prostaglandins coronary disease C-reactive protein isoprostanes risk factors The past decades have shown that oxidative stress is one fundamental of pathogenesis. This covers conditions as diverse as cancer, inflammation, neurodegenerative diseases, and aging. Likewise, oxidative damage is involved in the pathogenesis of atherosclerosis and cardiovascular disease. 1,2 For many years, investigations in the field of oxidative stress were limited by a confusing number of surrogate markers with clinical relevance open to question. Reliable quantitative indices of free radical induced modification of DNA, proteins, and lipids in vivo have emerged only recently. 3,4 In the cardiovascular system, lipids are in the first line of radical attack. Morrow et al 5 have identified isoprostanes as end products of lipid peroxidation. Isoprostanes are chemically stable and can be measured noninvasively in human urine by mass spectrometry. A growing body of evidence has been provided by several groups underlining the validity of the isoprostanes as in vivo markers of oxidative stress. 3,6,7 Studies have suggested isoprostanes as markers for patients at risk of atherosclerosis and coronary heart disease (CHD): elevated isoprostane formation was found in underlying diseases such as diabetes mellitus, 8 hypertension and renovascular disease, 9 hyperlipidemia, 10 obesity, 11 and cigarette smoking. 12 The major risk factors for CHD are elevated blood pressure, elevated total cholesterol, LDL cholesterol, HDL cholesterol, diabetes mellitus, obesity, cigarette smoking, and Received September 24, 2003; revision received November 7, 2003; accepted November 18, From the Clinical Pharmacology Unit, Institute of Experimental and Clinical Pharmacology (E.S., R.M., R.H.B.), Institute of Clinical Chemistry (J. Brümmer), and Institute of Mathematics and Data Processing in Medicine (J. Berger), University Hospital Hamburg-Eppendorf, and the Institute of Clinical Pharmacology (A.B., H.L., D.T., F.-M.G., J.C.F.), Hannover Medical School, Germany. E. Schwedhelm was responsible for the methodology of isoprostane measurement, interpretation of data, and writing the report. A. Bartling and H. Lenzen were responsible for patient recruiting and organization of urine and plasma analysis. D. Tsikas and F.-M. Gutzki were responsible for the isoprostane measurement. R. Mass did data analysis and interpretation. J. Brümmer performed C-reactive protein analysis. J. Berger was responsible for statistical analysis and interpretation of data. J.C. Frölich and R.H. Böger designed and organized the study. Correspondence to Dr rer nat Edzard Schwedhelm, Institut für Experimentelle und Klinische Pharmakologie, Universitätsklinikum Hamburg- Eppendorf, Martinistraße 52, D Hamburg, Germany. schwedhelm@uke.uni-hamburg.de 2004 American Heart Association, Inc. Circulation is available at DOI: /01.CIR

2 844 Circulation February 24, 2004 advancing age. 13 Studies evaluating the quantitative relationships between these risks have elucidated their additive nature in predicting CHD. The relative magnitude at which markers of oxidative stress such as isoprostanes contribute to these risk factors is as yet unknown. In addition, the extent to which oxidative stress contributes independently to hypertension, diabetes mellitus, hyperlipidemia, and obesity is open to question. Finally, increasing oxidative stress from different clinical conditions would be expected to contribute to CHD in a cumulative manner. To evaluate the hypothesis that isoprostanes are independent and cumulative risk markers of CHD, we conducted a case-control study. Ninety-three patients with a confirmed diagnosis of CHD were included and frequency-matched by age and sex with control subjects (controls) from the general population. The isoprostanes 8-iso-prostaglandin (PG) F 2 and 2,3-dinor-5,6-dihydro-8-iso-PGF 2 were measured by gas chromatography tandem mass spectrometry (GC-tandem MS) along with C-reactive protein (CRP), oxidized LDL, and further biochemical parameters. Methods Patients and Study Protocol Patients between 35 and 80 years old with established coronary artery disease were recruited for this study. Patients were included if they had stable angina pectoris or a history of myocardial infarction, plus diagnostic criteria for recurrent myocardial ischemia (positive exercise ECG, evidence for myocardial ischemia in thallium scintigraphy, documentation of coronary stenosis by coronary angiography). Age- and sex-matched healthy controls were recruited from the general population via local newspaper advertisement. They were included in the study if they had no clinical or diagnostic evidence for coronary artery disease. Patients or controls suffering from any concomitant acute or chronic severe disease, from renal failure (creatinine clearance 30 ml/min) or hepatic insufficiency, severe chronic heart failure (NYHA functional class III or above), acute occlusive cardiovascular events (ie, within the previous 30 days before inclusion into the study), or 80 years old were excluded from the study. We assessed eligibility, obtained informed consent, and included 93 patients and 93 controls in the study. The study protocol was approved by the local Institutional Review Board for Studies in Humans. All patients and controls were invited to the study center in the morning. The patients were asked for their previous medical history and the presence of cardiovascular risk factors by use of a standardized questionnaire. A fasting blood sample was drawn, and a urine sample was collected into a vial containing 5 mmol/l 4-hydroxy- 2,2,6,6-tetramethylpiperidine 1-oxyl (4-hydroxy-TEMPO) and 5 mmol/l EDTA as antioxidants. Blood samples were centrifuged (2000g, 10 minutes, 4 C), and plasma was divided into aliquots and stored at 20 C until analysis. Urine samples were divided into aliquots and kept frozen at 80 C until analysis. Diurnal variation of creatinine-indexed isoprostane excretion was assessed in 3 subjects. No divergence was found between morning urine samples and 24-hour urine samples. 14 Therefore, we decided to collect morning urine samples instead of 24-hour collection. Biochemical Analyses Urinary concentrations of 8-iso-PGF 2 and 2,3-dinor-5,6-dihydro-8- iso-pgf 2 were determined by GC-tandem MS as described previously in detail. 15 Briefly, urine samples were thawed, and labeled internal standards of 8-iso-PGF 2 and 2,3-dinor-5,6-dihydro-8-iso- PGF 2 were added at 1 ng/ml, solid-phase-extracted on ODS, and subjected to thin-layer chromatography on silica gel. Transitions from m/z 569 to m/z 299 and from m/z 543 to m/z 273 were monitored for 8-iso-PGF 2 and 2,3-dinor-5,6-dihydro-8-iso-PGF 2, respectively, by GC-tandem MS. Validity of the method was proven as reported previously. 16 Urinary concentrations of 8-iso-PGF 2 and 2,3-dinor-5,6-dihydro-8-iso-PGF 2 were corrected by urinary creatinine concentration to account for differences in renal excretory function. Plasma concentration of oxidized LDL cholesterol was determined with a commercially available immunoassay (Mercodia SA). Plasma concentration of CRP was measured by a highsensitivity assay (N Latex Mono test) on a Behring BN II nephelometer with polystyrene microbeads coated with monoclonal mouse antibodies. 17 The detection limit of the assay was 0.2 mg/l. Plasma total cholesterol, LDL and HDL cholesterol, and triglyceride levels as well as plasma and urinary creatinine concentrations were determined by standard laboratory methods using certified assays in the local clinical laboratory. Calculations and Statistical Methods All data were tested for normal distribution with the Kolmogorov- Smirnov test. The distribution of 8-iso-PGF 2, 2,3-dinor-5,6-dihydro- 8-iso-PGF 2, and CRP was highly skewed, as reported previously. 18,19 Continuous variables were expressed as arithmetic mean SD if normally distributed or otherwise by median with 25% and 75% percentiles. Because explorative data analysis was performed, no correction of the type I error was made. Differences between groups are given as mean or median differences. The 95% CIs are given for mean or median differences. Comparisons between 2 subgroups were performed with the Mann-Whitney U test (2-sided). Bivariate correlations were analyzed by Spearman s as indicated. Continuous variables were split into tertiles for univariate and multivariate analysis. Categorized biochemical characteristics (body mass index, systolic blood pressure, CRP, HDL cholesterol, and 8-iso-PGF 2 ) and risk factors of CHD (diabetes mellitus, hypercholesterolemia, hypertension, and cigarette smoking) were included into the unconditional logistic regression models. Biochemical characteristics were included stepwise in one multivariate model and risk factors of CHD, including 8-iso-PGF 2, in a second multivariate model. Estimates of risk (odds ratios) and 95% CI were calculated on the basis of coefficients from the logistic regression models. For statistical analyses, SPSS version 11.0 was used. Results We enrolled 93 CHD patients and 93 controls, frequencymatched by sex and age. Characteristics of patients and controls are given in Table 1. Total cholesterol, LDL cholesterol, HDL cholesterol, and triglyceride levels were lower in patients than in controls. Systolic blood pressure and body mass index were higher in the CHD group. Oxidative stress and inflammation were evaluated by oxidized LDL, isoprostanes, and high-sensitive CRP, respectively. CRP was more abundant in patients. Oxidized LDL was not different in patients and controls. The prevalence of cardiovascular risk factors for both groups is given in Table 2. As expected, patients were more likely to have diabetes mellitus, hypercholesterolemia, or hypertension. Patients were also more often smokers. Likewise, lipid-lowering agents (64% versus 6%, patients and controls, respectively), -blockers (61% versus 6%), ACE inhibitors (47% versus 3%), calcium antagonists (17% versus 6%), nitrates (62% versus 1%), and aspirin (100% versus 4%) were used more frequently in patients. Urinary excretion of 8-iso-PGF 2 was 139 (93 231) pmol/mmol creatinine in patients and 77 (61 101) pmol/ mmol creatinine in controls (P 0.001, Figure 1). Excretion of 8-iso-PGF 2 metabolite was 193 ( ) pmol/mmol creatinine in patients and 120 (91 151) pmol/mmol creatinine in controls (P 0.001). Individual urinary excretion of the

3 Schwedhelm et al 8-iso-Prostaglandin F 2 and CHD 845 TABLE 1. Characteristics of Patients All (n 186) Controls Cases Difference Between Groups Difference* 95% CI* P Age, y 63.5 (9.1) 63.5 (7.9) 63.4 (10.3) Sex, M/F 160/26 80/13 80/13 Body mass index, kg/m (3.6) 25.7 (3.4) 27.2 (3.6) to Blood pressure, mm Hg Systolic 141 (20) 136 (16) 145 (24) to Diastolic 84 (11) 82 (10) 85 (12) to Plasma lipids, mmol/l Total cholesterol 5.1 ( ) 5.4 ( ) 4.7 ( ) to HDL cholesterol 0.9 ( ) 1.0 ( ) 0.9 ( ) to LDL cholesterol 3.2 ( ) 3.6 ( ) 2.7 ( ) to Triglycerides 1.2 ( ) 1.4 ( ) 1.0 ( ) to CRP high-sensitive, mg/l 1.6 ( ) 1.1 ( ) 3.7 ( ) to Creatinine clearance, ml/min 80 (25) 78 (22) 83 (28) to Oxidized LDL, U/L 31 (23 40) 31 (25 40) 30 (22 41) to Data are mean (SD) except for plasma lipids, CRP high-sensitive, and oxidized LDL, which are median (1st and 3rd quartiles). *Mean or median. Mann-Whitney U test (two-sided). Controls were matched for these criteria. To convert values for cholesterol to mg/dl, divide by ; to convert values for triglycerides to mg/dl, divide by metabolite was highly dependent on 8-iso-PGF 2 excretion ( 0.664, P 0.001). Neither 8-iso-PGF 2 nor its metabolite correlated with blood pressure, total cholesterol, LDL cholesterol, or HDL cholesterol. Treatment with lipid-lowering agents, -blockers, ACE inhibitors, calcium antagonists, and nitrates was not associated with reduced urinary excretion of 8-iso-PGF 2 or its metabolite in patients. There was no correlation between the 2 isoprostanes and oxidized LDL. However, a correlation was found between 8-iso-PGF 2, 2,3-dinor-5,6-dihydro-8-iso-PGF 2, and CRP ( 0.225, P 0.01, and 0.321, P 0.001, respectively). Urinary excretion of 8-iso-PGF 2 correlated with the number of cardiovascular risk factors in the 2 groups (P for trend, Figure 2). On univariate analysis, tertiles of body mass index, systolic blood pressure, CRP, and 8-iso-PGF 2 were concentrationdependent risk markers of CHD (Table 2). The odds ratio for HDL cholesterol was decreased for plasma levels 0.9 mmol/l. Diabetes mellitus, hypercholesterolemia, hypertension, cigarette smoking, and obesity (body mass index 30 kg/m 2 ) predicted CHD in univariate analysis as well. On multivariate analysis, the categorized characteristics of patients, including body mass index ( 30 kg/m 2 ), systolic blood pressure, CRP, HDL cholesterol, and 8-iso-PGF 2, were tested as independent markers of CHD (Table 3). Only urinary excretion of 8-iso-PGF 2 remained a strong risk marker in the 2 higher tertiles. The odds ratios were increased by 4.61 for 8-iso-PGF 2 levels 80 pmol/mmol creatinine and by 30.8 for 8-iso-PGF 2 levels 130 pmol/mmol creatinine. For CRP, an increased risk was found only for the highest tertile, ie, by 7.20 for CRP 3.0 mg/l. In a second model of multivariate analysis, tertiles of 8-iso-PGF 2 were included together with risk factors of CHD (Table 4). Included risk factors were diabetes mellitus, hypercholesterolemia, hypertension, and cigarette smoking. With the exception of cigarette smoking, all other risk factors remained strong and independent predictors of CHD. 8-iso- PGF 2 was found to be a strong concentration-dependent risk marker in this logistic regression model also. Discussion Cardiovascular risk is believed to be at least in part related to increased systemic oxidative stress. Established risk factors of CHD have been associated with elevated levels of markers of systemic oxidative stress such as isoprostanes However, such evidence is restricted to single risk factors. The principal result of our study is that systemic oxidative stress, evaluated as creatinine-indexed urinary isoprostanes, is related to the number of cardiovascular risk factors as well as to the prevalence of CHD. Other categorized patients characteristics, such as body mass index, systolic blood pressure, HDL cholesterol, and CRP, included in univariate analysis also predicted the prevalence of CHD (Table 2). However, in multivariate analysis, only isoprostane and CRP remained independent risk markers of CHD. Cigarette smoking, hypertension, hypercholesterolemia, diabetes mellitus, and obesity are well-known risk factors of CHD. 13 With the exception of obesity, they also predicted the prevalence of CHD in our study. Obesity missed the significance level probably because of its comparatively low prevalence and was therefore not included into the multivariate analysis. Included risk factors, ie, diabetes mellitus, arterial hypertension, and hypercholesterolemia remained strong predictors of CHD in multivariate regression analysis. 8-iso-PGF 2 divided into tertiles was found to be a strong predictor of CHD in this model as well. The odds ratio was significantly elevated for levels of

4 846 Circulation February 24, 2004 TABLE 2. Risk Factors for CHD in Univariate Analysis* All (n 186) 8-iso-PGF 2 80 pmol/mmol creatinine. The close relationship previously observed between cigarette smoking and increased lipid peroxidation may explain why cigarette smoking was not found to be an independent risk marker in Controls Cases Odds Ratios (95% CI) P Body mass index, kg/m (34) 41 (44) 23 (25) (49) 40 (43) 51 (55) 2.27 ( ) (17) 12 (13) 19 (20) 2 82 ( ) 0.02 Systolic blood pressure, mm Hg (41) 52 (56) 24 (26) (37) 29 (31) 40 (43) 2.99 ( ) (22) 12 (13) 29 (31) 5.24 ( ) CRP high-sensitive, mg/l (29) 34 (45) 9 (13) (34) 29 (38) 21 (29) 2.74 ( ) (37) 13 (17) 42 (58) 12.2 ( ) Unknown HDL cholesterol, mmol/l (40) 22 (27) 40 (55) (60) 59 (73) 33 (45) 0.31 ( ) Unknown iso-PGF 2a, pmol/mmol (33) 53 (57) 9 (10) (33) 29 (31) 33 (35) 6.70 ( ) (33) 11 (12) 51 (55) 27.3 ( ) Diabetes mellitus 28 (15) 3 (3) 25 (27) 11.0 ( ) Hypercholesterolemia 93 (50) 24 (26) 69 (74) 8.27 ( ) Hypertension 139 (75) 50 (54) 89 (96) 19.1 ( ) Cigarette smoking Never 70 (39) 45 (51) 25 (27) 1.00 Former 93 (51) 42 (47) 51 (55) 2.19 ( ) 0.02 Current 19 (10) 2 (2) 17 (18) 15.2 ( ) Unknown 4 4 Data are counts (% of known). *In 93 patients and 93 controls frequency-matched by age and sex. Reference. LDL cholesterol 4.1 mmol/l, total cholesterol 6.2 mmol/l, or statin. Blood pressure 140/90 mm Hg or antihypertensive drug. multivariate analysis. 12 However, our results confirm previous studies showing increased isoprostane formation in diabetes mellitus, renal hypertension, hypercholesterolemia, and cigarette smoking. 8 10,12 Furthermore, they are in line with Figure 1. Creatinine-indexed urinary excretion of 8-iso-PGF 2 and its urinary metabolite 2,3-dinor-5,6-dihydro-8-iso- PGF 2 in 93 patients with CHD and 93 age- and sex-matched controls. Median, interquartile range, outliers, and extremes of urinary excretion are given; *P

5 Schwedhelm et al 8-iso-Prostaglandin F 2 and CHD 847 TABLE 4. Odds Ratios From a Multivariate Logistic Regression Model Including Risk Factors of CHD* Figure 2. Creatinine-indexed urinary excretion of 8-iso-PGF 2 and risk factors of CHD. Risk factors for CHD are obesity, diabetes mellitus, hypercholesterolemia, hypertension, and smoking. Median, interquartile range, outliers, and extremes of urinary excretion are given; P for trend. recent findings from the Framingham study showing an association between urinary excretion of 8-iso-PGF 2 and history of cardiovascular disease, smoking, and body mass index. 18 Moreover, our data show for the first time that systemic oxidative stress is an additive risk marker of CHD in addition to traditional risk factors (Table 4). In line with this result, urinary excretion of 8-iso-PGF 2 continuously increased with the number of traditional CHD risk factors (Figure 2). To confirm the validity of our data, we correlated 8-iso- PGF 2 with its major urinary metabolite, 2,3-dinor-5,6- dihydro-8-iso-pgf 2. We found a very good correlation (Spearman s 0.66) analyzing both parameters with GCtandem MS. Both were highly different between case subjects TABLE 3. Multivariate Model* Predicting CHD Odds Ratio (95% CI) P Body mass index, kg/m ( ) 0.61 Systolic blood pressure, mm Hg ( ) ( ) 0.13 CRP high-sensitive, mg/l ( ) ( ) 0.01 HDL cholesterol, mmol/l ( ) iso-PGF 2a, pmol/mmol ( ) ( ) *A stepwise regression model including categorized biochemical characteristics in 93 patients and 93 controls frequency-matched by age and sex. Adjusted for variables. Reference. Odds Ratio (95% CI) P Diabetes mellitus 7.8 ( ) 0.01 Hypercholesterolemia 12.1 ( ) Hypertension 20.6 ( ) Cigarette smoking Never 1.00 Former 2.5 ( ) 0.06 Current 2.8 ( ) iso-PGF 2a, pmol/mmol ( ) ( ) *A stepwise model including risk factors for CHD in 93 patients and 93 controls frequency-matched by age and sex. Adjusted for variables. Reference. (cases) and controls (P 0.001). Further evidence for the suitability of 8-iso-PGF 2 as a novel risk marker comes from analysis of possible confounders. Not surprisingly, in our cohort of cases, we found decreased levels of total cholesterol, LDL cholesterol, and triglycerides, reflecting the widespread use of lipid-lowering therapy in CHD patients (64%). Urinary excretion of 8-iso-PGF 2 and its metabolite were not confounded by either lipid-lowering therapy or by total cholesterol, LDL cholesterol, or triglyceride levels. This finding is of particular interest, because it has been demonstrated that intervention with statins in mildly hypercholesterolemic patients reduced oxidant burden, ie, 8-iso-PGF 2 and 3-nitrotyrosine formation. 20,21 However, our trial was designed as a case-control study and therefore was not intended to detect such effects. Recently, a difference in urinary 8-iso-PGF 2 in CHD patients with 1-vessel and multivessel disease was found. 22 Unfortunately, we were not able to confirm this observation in our study (data not shown). Interestingly, a correlation between 8-iso-PGF 2, its metabolite, and CRP existed ( 0.23 and 0.32, respectively; P 0.01). This observation may result in part from the involvement of inflammatory and oxidative processes in atherosclerosis. 1,23,24 Moderately elevated CRP is associated with a high risk of coronary events, possibly because of inflammation of the vascular bed. 25 Although inflammation works hand in hand with oxidative modifications, oxidative stress may occur independently of inflammation. In support of this, we found 8-iso-PGF 2 and high-sensitive CRP to be independent predictors of CHD (Table 3). Thus, the evaluation of both 8-iso-PGF 2 and CRP should be superior to the measurement of either parameter alone. Another parameter of lipid peroxidation, oxidized LDL, failed to show a difference in our case-control design. There is substantial evidence that oxidized LDL is present in vivo within atherosclerotic blood vessels. 26,27 Probably, circulating oxidized LDL derives, at least in part, from rupture of atherosclerotic plaques. Thus,

6 848 Circulation February 24, 2004 elevated levels of oxidized LDL showed a positive correlation with the severity of acute coronary events. 28 Because acute coronary syndrome and related conditions were excluded in our study, this could explain the similar levels of oxidized LDL in both groups. Interestingly, Cipollone et al 29 previously reported highly elevated urinary 8-iso-PGF 2 measured in patients with unstable angina, supporting the important role of oxidative stress in a worse scenario than investigated in our study. In conclusion, our findings introduce 8-iso-PGF 2 as a new risk marker in the field of cardiovascular disease. 8-iso-PGF 2 was identified as an independent and cumulative risk marker of CHD together with diabetes mellitus, arterial hypertension, hypercholesterolemia, and elevated CRP. The case-control design used in our study does not allow us to show the value of 8-iso-PGF 2 for predicting future coronary events. Therefore, we initiated a cohort study to evaluate 8-iso-PGF 2 as a risk factor for coronary events. Acknowledgments We thank A. Steenpaß, T. Suchy, and B. Schubert for excellent technical assistance and the patients and healthy volunteers who participated in this study. References 1. Harrison D, Griendling KK, Landmesser U, et al. Role of oxidative stress in atherosclerosis. Am J Cardiol. 2003;91:7A 11A. 2. Dhalla NS, Temsah RM, Netticadan T. Role of oxidative stress in cardiovascular disease. J Hypertens. 2000;18: Cracowski JL, Durand T, Bessard G. Isoprostanes as a biomarker of lipid peroxidation in humans: physiology, pharmacology and clinical implications. Trends Pharmacol Sci. 2002;23: Thérond P, Bonnefont-Rousselot D, Davit-Spraul A, et al. Biomarkers of oxidative stress: an analytical approach. Curr Opin Clin Nutr Metab Care. 2000;3: Morrow JD, Hill KE, Burk RF, et al. A series of prostaglandin F 2 -like compounds are produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci U S A. 1990;87: Morrow JD. The isoprostanes: their quantification as an index of oxidant stress status in vivo. Drug Metab Rev. 2000;32: Lawson JA, Rokach J, FitzGerald GA. Isoprostanes: formation, analysis and use as indices of lipid peroxidation. J Biol Chem. 1999;274: Davi G, Ciabattoni G, Consoli A, et al. In vivo formation of 8-iso-prostaglandin F 2 and platelet activation in diabetes mellitus: effects of improved metabolic control and vitamin E supplementation. Circulation. 1999;99: Minuz P, Patrignani P, Gaino S, et al. Increased oxidative stress and platelet activation in patients with hypertension and renovascular disease. Circulation. 2002;106: Davi G, Alessandrini P, Mezzetti A, et al. In vivo formation of 8-isoprostaglandin F 2 is increased in hypercholesterolemia. Arterioscler Thromb Vasc Biol. 1997;17: Davi G, Guagnano MT, Ciabattoni G, et al. Platelet activation in obese women: role of inflammation and oxidant stress. JAMA. 2002;288: Morrow JD, Frei B, Longmire AW, et al. Increase in circulating products of lipid peroxidation (F 2 -isoprostanes) in smokers: smoking as a cause of oxidative damage. N Engl J Med. 1995;332: Grundy SM, Pasternak R, Greenland P, et al. Assessment of cardiovascular risk by use of multiple-risk-factor assessment equations: a statement for healthcare professionals from the American Heart Association and the American College of Cardiology. Circulation. 1999;100: Helmersson J, Basu S. F 2 -isoprostane excretion rate and diurnal variation in human urine. Prostaglandins Leukot Essent Fatty Acids. 1999;61: Schwedhelm E, Tsikas D, Durand T, et al. Tandem mass spectrometric quantification of 8-iso-prostaglandin F 2 and its metabolite 2,3-dinor-5,6- dihydro-8-iso-prostaglandin F 2 in human urine. J Chromatogr B. 2000; 744: Tsikas D, Schwedhelm E, Fauler J, et al. Specific and rapid quantification of 8-iso-prostaglandin F 2 in urine of healthy humans and patients with Zellweger syndrome by gas chromatography-tandem mass spectrometry. J Chromatogr B. 1998;716: Ledue TB, Weiner DL, Sipe JD, et al. Analytical evaluation of a particleenhanced immunonephelometric assay for C-reactive protein, serum amyloid A and mannose-binding protein in human serum. Ann Clin Biochem. 1998;35: Keaney JF Jr, Larson MG, Vasan RS, et al. Obesity and systemic oxidative stress: clinical correlates of oxidative stress in the Framingham study. Arterioscler Thromb Vasc Biol. 2003;23: Koenig W, Sund M, Frohlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to Circulation. 1999;99: Shishehbor MH, Aviles RJ, Brennan ML, et al. Association of nitrotyrosine levels with cardiovascular disease and modulation by statin therapy. JAMA. 2003;289: De Caterina R, Cipollone F, Filardo FP, et al. Low-density lipoprotein level reduction by the 3-hydroxy-3-methylglutaryl coenzyme-a inhibitor simvastatin is accompanied by a related reduction of F 2 -isoprostane formation in hypercholesterolemic subjects: no further effect of vitamin E. Circulation. 2002;106: Vassalle C, Botto N, Andreassi MG, et al. Evidence for enhanced 8-isoprostane plasma levels, as index of oxidative stress in vivo, in patients with coronary artery disease. Coron Artery Dis. 2003;14: Ross R. Atherosclerosis: an inflammatory disease. N Engl J Med. 1999; 340: Diaz MN, Frei B, Vita JA, et al. Mechanism of disease: antioxidants and atherosclerotic heart disease. N Engl J Med. 1997;337: Ridker PM, Rifai N, Rose L, et al. Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events. N Engl J Med. 2002;347: Kuhn H, Heydeck D, Hugou I, et al. In vivo action of 15-lipoxygenase in early stages of human atherogenesis. J Clin Invest. 1997;99: Watson AD, Leitinger N, Navab M, et al. Structural identification by mass spectrometry of oxidized phospholipids in minimally oxidized low density lipoprotein that induce monocyte/endothelial interactions and evidence for their presence in vivo. J Biol Chem. 1997;272: Ehara S, Ueda M, Naruko T, et al. Elevated levels of oxidized low density lipoprotein show a positive relationship with the severity of acute coronary syndromes. Circulation. 2001;103: Cipollone F, Ciabattoni G, Patrignani P, et al. Oxidant stress and aspirininsensitive thromboxane biosynthesis in severe unstable angina. Circulation. 2000;102: