- Email: [email protected]
High-sensitivity C-reactive protein in the prediction of coronary events in patients with premature coronary artery disease
Clinical Investigations
Acute Ischemic Heart Disease
High-sensitivity C-reactive protein in the prediction of coronary events in patients with premature coronary artery disease Walter S. Speidl, MD, Senta Graf, MD, Stefan Hornykewycz, MD, Mariam Nikfardjam, MD, Alexander Niessner, MD, Gerlinde Zorn, Johann Wojta, PhD, and Kurt Huber, MD Vienna, Austria
Background and Methods Inflammation plays an important role in the initiation and progression of atherosclerosis and in the pathogenesis of acute cardiovascular events. Recent studies have indicated a possible association between C-reactive protein (CRP) and the clinical outcome of coronary artery disease (CAD). We studied prospectively in a group of 125 patients with premature CAD whether plasma levels of CRP as measured with a high-sensitivity assay predict risk for future coronary events. All patients had stable CAD at time of blood sampling but had originally been seen with unstable angina or myocardial infarction. The mean follow-up time after blood collection was 54 months, and death, myocardial infarction, need for coronary revascularization, or admission to hospital with angina pectoris were defined as clinical end points.
Results Patients in the highest tertile of CRP levels had a ⬎3.8-fold risk (risk ratio 3.82, 95% CI 1.19-12.17) for death, myocardial infarction, or need for coronary revascularization compared with the patients in the first tertile. The relative risk for patients in the second tertile was 3.5-fold higher (95% CI 1.04-11.56). CRP levels in the third tertile independently predicted risk after adjustment for lipids and other clinical risk factors. Conclusion In patients with clinically stable conditions who have a positive history for acute coronary syndromes before age 50 years, plasma levels of CRP higher than 1.6 mg/L are predictors of future coronary events and therefore indicate the role of underlying chronic inflammation for the clinical course of CAD. Accordingly, reference limits for prediction of risk in CAD have to be lower in this specific patient group than in middle-aged or elderly patients. (Am Heart J 2002; 144:449-55.)
Coronary artery disease (CAD) is a major cause of death in the Western world. Meanwhile, inflammation is accepted as playing an important role in the initiation and progression of atherosclerosis.1,2 Especially, the transition from stable advanced lesions to vulnerable plaques is paralleled by local accumulation of macrophages and secretion of cytokines and other destabilizing factors. These active inflammatory processes contribute to a thinning of the fibrous cap of lipid-rich plaques and to plaque erosion or rupture.3 As a side effect of local inflammation, circulating cytokines (eg, interleukin-6, interleukin-1, tumor necrosis factor–␣, oncostatin M, or leukemia inhibitory factor) activate hepatocytes to produce acute phase proteins.4,5 One
From the Department of Cardiology, University of Vienna, Austria. Submitted June 15, 2001; accepted February 20, 2002. Reprint requests: Kurt Huber, MD, Department of Cardiology, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-8703/2002/$35.00 ⫹ 0 4/1/124353 doi:10.1067/mhj.2002.124353
of the most sensitive acute phase proteins is C-reactive protein (CRP). The plasma concentration of CRP is determined mainly by its synthesis rate, and, provided normal liver function exists, CRP is a sensitive marker for ongoing chronic inflammation that is not affected by ischemic injury.6 In recent studies, an association between CRP levels and the risk of future cardiovascular events has been discussed for healthy individuals7-9 and for individuals with an increased cardiovascular risk although apparently healthy.10 In addition, patients with stable or unstable angina pectoris11,12 and patients after myocardial infarction13 with elevated CRP levels are at high risk for future cardiovascular events. So far, trials have only investigated middle-aged and elderly patients,14 and less is known about the role of CRP levels to predict clinical course in patients with premature CAD. Therefore, the aim of this study was to investigate whether lower ranges of CRP as quantified with a highly specific commercial assay predict future risk for coronary events in patients with stable conditions with premature CAD.
American Heart Journal September 2002
450 Speidl et al
Table I. Characteristics of the study participants
No. Age (y) (mean ⫾ SD) Sex, male (%) Smoking status (%) Current smoker Former smoker Nonsmoker Systolic blood pressure (mm Hg) (mean ⫾ SD) Diastolic blood pressure (mm Hg) (mean ⫾ SD) Body mass index (kg/m2) (mean ⫾ SD) Total cholesterol (mg/dL) (mean ⫾ SD) LDL cholesterol (mg/dL) (mean ⫾ SD) HDL cholesterol (mg/dL) (mean ⫾ SD) Triglycerides (mg/dL) (mean ⫾ SD) Diabetes (%) Family history of CAD (%) Previous myocardial infarction (%) Number of coronary arteries narrowed ⬎70% (%) 0 1 2 3
Total
Without event
With event
P
119 39.3 ⫾ 5.6 76.5
71 38.7 ⫾ 5.2 74.6
48 40.1 ⫾ 6.2 79.2
.179 .569
76.5 12.6 10.9 114.6 ⫾ 16.2 75.4 ⫾ 12.2 26.9 ⫾ 4.4 209.9 ⫾ 48.1 137.4 ⫾ 41.2 35.0 ⫾ 11.0 184.5 ⫾ 113.8 5.2 62.9 78.2
70.4 15.5 14.1 113.8 ⫾ 15.7 75.6 ⫾ 11.8 26.7 ⫾ 4.2 203.3 ⫾ 50.3 130.1 ⫾ 41.1 37.0 ⫾ 11.4 167.3 ⫾ 107.8 2.9 62.3 77.5
85.4 8.3 6.3 115.6 ⫾ 17.1 75.1 ⫾ 12.9 27.3 ⫾ 4.6 219.6 ⫾ 43.3 148.2 ⫾ 39.3 32.0 ⫾ 9.9 211.3 ⫾ 118.8 8.5 63.6 79.2
17.2 45.7 25.0 12.1
20.6 48.5 20.6 10.3
12.5 41.7 31.3 14.6
.164
.591 .827 .47 .078 .018 .014 .048 .16 .888 .826 .382
CAD, Coronary artery disease.
Table II. Characteristics of the study participants according to tertiles of CRP Tertile
CRP (mg/L) (median) Range Age (y) (mean ⫾ SD) Sex, male (%) Smoking status (%) Current smoker Former smoker Nonsmoker Systolic blood pressure (mm Hg) (mean ⫾ SD) Diastolic blood pressure (mm Hg) (mean ⫾ SD) Body mass index (kg/m2) (mean ⫾ SD) Total cholesterol (mg/dL) (mean ⫾ SD) LDL cholesterol (mg/dL) (mean ⫾ SD) HDL cholesterol (mg/dL) (mean ⫾ SD) Triglycerides (mg/dL) (mean ⫾ SD) Diabetes (%) Family history of CAD (%) Previous myocardial infarction (%) Number of coronary arteries narrowed ⬎70% (%) 0 1 2 3
1
2
3
P
0.09 0.19-1.59 38.7 ⫾ 6.0 74.4
3.09 1.69-5.51 39.7 ⫾ 5.3 80.0
9.19 5.58-174.00 39.4 ⫾ 5.7 75.0
.751 .81
74.4 12.8 12.8 113.5 ⫾ 15.4 74.6 ⫾ 11.0 25.9 ⫾ 3.9 201.8 ⫾ 44.0 129.4 ⫾ 32.6 36.4 ⫾ 11.8 176.2 ⫾ 112.9 2.6 61.3 71.8
80.0 5.0 15.0 116.1 ⫾ 16.6 75.0 ⫾ 12.7 26.7 ⫾ 3.8 222.4 ⫾ 46.0 147.1 ⫾ 41.0 34.6 ⫾ 10.2 203.9 ⫾ 133.6 10.3 56.8 80.0
24.3 54.1 13.5 8.1
12.5 40.0 35.0 12.5
75.0 20.0 5.0 114.1 ⫾ 16.9 76.6 ⫾ 13.0 28.0 ⫾ 5.1 202.6 ⫾ 52.3 135.6 ⫾ 47.5 34.1 ⫾ 11.2 173.2 ⫾ 92.1 2.6 70.3 82.5 15.4 43.6 25.6 15.4
.223
.775 .761 .114 .151 .155 .628 .445 .213 .474 .485 .324
American Heart Journal Volume 144, Number 3
Methods Patients We investigated 125 consecutive patients with angiographically proven and clinically stable premature CAD who had originally attended the patient ward of the department of cardiology (University of Vienna) with signs of acute coronary syndromes (unstable angina or myocardial infarction). All patients were aged ⬍50 years (range 21-50 y, average 39.2 ⫾ 5.6 y) at the time of inclusion. Original diagnosis included the angiographic extent and severity of disease (⬎70% narrowing of the lumen of ⱖ1 major epicardial coronary artery). Exclusion criteria were any situations that might have influenced CRP plasma levels (eg, ongoing clinical instability, recent coronary interventions, or infectious diseases). The patients were followed for a maximum of 72 months, with an average follow-up time of 54.1 ⫾ 2.4 months. No follow-up data were obtainable after inclusion for 6 patients (4.8%). Primary clinical end points were death from CAD, nonfatal myocardial infarction, need for coronary revascularization, and admission to hospital in case of renewed angina pectoris. To detect and verify these study end points, most patients were seen in the outpatient ward. In rare cases, telephone and mail contact with the patients themselves or with their general practitioners was necessary. For patients who died during the follow-up period, the official national death register was reviewed.
Blood collection and assay systems At 8 to 12 weeks after the acute event, blood samples were taken from all subjects between 8 and 10 am after a 14-hour overnight fast. Venous blood was drawn directly into plastic tubes containing either ethylenediamine tetraacetic acid (5 ⫻ 10–2 mol/L final concentration) or citrate (0.11 mol/L final concentration). Tubes were centrifuged immediately at 3000g for 10 minutes at 4°C, and plasma was stored at – 40°C until use. Plasma levels of CRP were determined with the N high sensitivity CRP test (Dade Behring, Marburg, Germany). Plasma levels of triglycerides and cholesterol were determined with laboratory standard techniques.
Statistical analysis Data are represented as mean and standard deviation if not stated otherwise. Proportions or means for baseline risk factors were calculated. The significance of any differences in proportions was tested with the 2 test, and the significance of any difference in means was tested with analysis of variance. Because the distribution of CRP is rightward skewed, values derived from natural logarithm-transformed means were used as means for this variable and the significance of any differences in mean values of CRP was tested with analysis of variance of natural logarithm-transformed data. Median concentrations were computed for CRP, and the significance of any differences in median values was tested with the Wilcoxon rank-sum test. The patients were divided into tertiles according to CRP distribution for calculating the risk and for adjusted risk estimates, calculated with use of logistic-regression models, including clinical risk factors (age, sex, body mass index, smoking, diabetes, and family history for CAD) and blood lipids. All P values were 2 tailed, and CIs were calculated at the 95% level. All calculations were performed
Speidl et al 451
Figure 1
Kaplan-Meier survival plots for death, myocardial infarction, and need for revascularization procedure according to tertiles of CRP.
with a computer program (SPSS for Windows 10.0.1, Chicago, Ill).
Results During the follow-up period, 5 patients (4.2%) died from CAD, 10 patients (8.4%) had an acute myocardial infarction develop, 15 patients (12.6%) underwent coronary revascularization (coronary artery bypass grafting n ⫽ 5, percutaneous coronary intervention n ⫽ 10), and 18 patients (15.1%) were admitted to a hospital because of renewed angina pectoris. Seventy-one patients (59.7%) were completely free of coronary events during the follow-up period. The baseline characteristics of study participants as further subdivided into patients without or with further events are shown in Table I. Patients in whom a cardiovascular event developed tended to be smokers or to have diabetes, both of which were not statistically different compared with those without cardiovascular event. No differences were found for mean systolic or diastolic blood pressures, body mass index, or total cholesterol level. However, significant differences could be found for triglycerides (P ⫽ .048), low-density lipoprotein cholesterol level (P ⫽ .018), and highdensity lipoprotein cholesterol level (P ⫽ .014). No significant differences between patients without or with further events could be found in mean (2.66 ⫾ 37.7 mg/L vs 3.69 ⫾ 28.4 mg/L, P ⫽ .155) and median (2.48 mg/L vs 3.66 mg/L, P ⫽ .105) CRP levels. To calculate the relative risk, patients were divided into 3 groups according to the distribution of
American Heart Journal September 2002
452 Speidl et al
Table III. Relative risk of death, myocardial infarction, and need for revascularization procedure according to tertiles of CRP Tertile Variables adjusted for
1
2
3
Unadjusted Relative risk 1 3.472 3.816 95% CI 1.04-11.56 1.19-12.17 P .043 .024 Age Relative risk 1 3.469 3.808 95% CI 1.04-11.55 1.19-12.15 P .043 .024 Age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.019 3.810 95% CI 0.85-10.70 1.13-12.84 P .087 .031 Triglycerides Relative risk 1 3.364 3.840 95% CI 1.00-11.33 1.19-12.37 P .05 .024 Total cholesterol Relative risk 1 3.056 3.871 95% CI 0.90-10.36 1.20-12.49 P .073 .024 LDL cholesterol Relative risk 1 2.838 3.595 95% CI 0.83-9.73 1.11-11.69 P .097 .033 HDL cholesterol Relative risk 1 3.432 3.612 95% CI 1.00-11.77 1.10-11.83 P .05 .034 Total/HDL cholesterol Relative risk 1 3.090 3.772 95% CI 0.90-10.66 1.14-12.51 P .074 .03 Triglycerides and total/HDL cholesterol Relative risk 1 3.073 3.735 95% CI 0.89-10.63 1.13-12.38 P .076 .031 Triglycerides, total/HDL cholesterol, age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 2.924 3.728 95% CI 0.79-10.80 1.06-13.18 P .108 .041
P for trend
.026
.027
.032
.027
.025
.036
.039
.033
.034
.043
BMI, Body mass index.
CRP levels into tertiles. The patients in the 3 tertiles did not show any significant differences in risk factors (Table II). Patients in the tertile with the highest levels of CRP had a ⬎3.8-fold increased risk (risk ratio [RR] 3.82, P ⫽ .024) for harder end points (death, myocardial infarction, or need for coronary revascularization) compared with the patients in the first tertile (RR 1). The relative risk for patients in the second tertile was 3.5-fold increased (RR 3.47, P ⫽ .043) compared with the patients in the first tertile. The relative risk for development of any of the study end points also including admission to hospital with angina was ⬎2.7-fold
higher for patients in the third tertile (RR 2.73, P ⫽ .043) and ⬎3.6-fold elevated for patients in the second tertile (RR 3.68, P ⫽ .008). To show a relation between the time course of CAD and the plasma levels of CRP, we calculated KaplanMeier survival plots for the patients in the different tertiles (Figure 1). During the first 6 months, the event rate was similar in any CRP tertile group. Thereafter, the patients in the lowest tertile showed only a few combined primary end point events compared with the patients in the second and third CRP tertiles. This different course over time was statistically significant
American Heart Journal Volume 144, Number 3
Speidl et al 453
Table IV. Relative risk of death, myocardial infarction, need for revascularization procedure and admission to hospital with angina according to tertiles of CRP Tertile Variables adjusted for
1
2
Unadjusted Relative risk 1 3.684 95% CI 1.40-9.71 P .008 Age Relative risk 1 3.606 95% CI 1.36-9.56 P .01 Age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.238 95% CI 1.19-8.82 P .022 Triglycerides Relative risk 1 3.555 95% CI 1.33-9.52 P .012 Total cholesterol Relative risk 1 3.387 95% CI 1.27-9.03 P .015 LDL cholesterol Relative risk 1 3.205 95% CI 1.19-8.60 P .021 HDL cholesterol Relative risk 1 3.656 95% CI 1.35-9.87 P .011 Total/HDL cholesterol Relative risk 1 3.449 95% CI 1.27-9.38 P .008 Triglycerides and total/HDL cholesterol Relative risk 1 3.439 95% CI 1.26-9.35 P .016 Triglycerides, total/HDL cholesterol, age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.227 95% CI 1.14-9.11 P .027
(P ⬍ .05). To evaluate whether increased baseline CRP is an independent predictive marker, we performed logistic-regression analyses and adjusted for plasma lipids and the clinical risk factors age, sex, body mass index, hypertension, diabetes, and positive family history. CRP levels in the third tertile still remained an independent predictor of risk for death, myocardial infarction, or need for coronary revascularization (Table III) but lost independency for all study end points, including admission to hospital with angina (Table IV). To test a positive relation between the extent of CAD and CRP plasma levels, we correlated the median values of CRP with the number of main epicardial cor-
3
P for trend
2.727 1.03-7.20 .043
.051
2.684 1.01-7.13 .048
.056
2.608 0.95-7.15 .062
.055
2.849 1.06-7.65 .038
.044
2.816 1.06-7.52 .039
.043
2.630 0.98-7.05 .054
.062
2.586 0.96-6.99 .061
.073
2.751 1.01-7.53 .049
.056
2.773 1.01-7.62 .048
.054
2.703 0.94-7.75 .065
.064
onary arteries affected in the initial coronary angiography (Figure 2). A trend was seen, but no significant relationship between the extent of CAD and CRP plasma levels was found (P ⫽ .263). Similarly, for patients with a history of myocardial infarction, we found also a nonsignificant increase in CRP levels compared with those without such a history (data not shown).
Discussion Our data for the first time show a relationship between high-sensitivity CRP plasma levels and the risk
American Heart Journal September 2002
454 Speidl et al
Figure 2
Median values of CRP in mg/L according to number of affected coronary vessels (⬎70% narrowing of lumen).
of secondary coronary events in patients with stable conditions with premature CAD. In opposition to the higher CRP concentrations generated in older patients with stable or unstable angina11 or with a history of myocardial infarction,13 in our hands, CRP levels of ⬎1.6 mg/L were associated with an increased cardiovascular event rate. The fact that patients with combined hard end points (death, myocardial infarction, or need for coronary revascularization) showed an increased event rate with increasing CRP levels but this significant linear trend was lost when renewed angina was added may be explained by the fact that angina is a more subjective parameter, being related also to noncardiac reasons.15 For our patients, we can exclude an influence of acute phase reactions on the CRP levels because patients with acute infections, recent coronary interventions, or myocardial infarctions, or other conditions that may have altered CRP levels, were excluded from the study. Furthermore, blood was collected 8 to 12 weeks after an acute event and CRP plasma levels are known to decline to baseline within a few days.16 Patients showed a positive nonsignificant association between CRP plasma levels and the number of severely affected coronary arteries at time of angiographic control (8-12 weeks before the blood collection for CRP determination). A similar association has been described elsewhere for middle-aged individuals.17 Because some of the patients underwent immediate percutaneous coronary intervention or diagnostic angiography and because of the relatively small num-
ber of patients in our study, the association may have been insignificant. Our results of the positive role of CRP as predictor of further cardiovascular events in premature CAD open the question of whether different cut-off levels for CRP in CAD should be used for the prediction of coronary events because the risk for a future coronary event increases rapidly in young patients with CAD at lower CRP levels (CRP ⬎ 1.6 mg/L) compared with middle-aged and elderly patients.2 One possible explanation for this increase in risk at lower CRP levels in patients aged ⬍50 years may be explained by the fact that CRP levels are positively correlated with age.18 Several limitations of our data merit consideration. The relatively small number of patients involved may be causally related with the tendentious but nonsignificant results in several calculations. Furthermore, we only measured baseline CRP plasma levels once and have no evidence for alterations during the course of follow-up, which may be more accurate in terms of prediction of clinical course. On the other hand, CRP level are in most cases relatively constant over longer periods in stable diseases.18 In conclusion, in patients with stable premature CAD, CRP levels in the presently accepted upper normal range or higher are accompanied with an increased long-term risk for development of cardiovascular events, such as cardiac death, acute myocardial infarction, or need for revascularization. The high-sensitivity assay for exact determination of relatively low CRP levels is reliable, can be carried out easily, and might, therefore, represent a new and noninvasive tool to help to identify an increased risk for future coronary events in younger patients with CAD.
References 1. Ross R. Atherosclerosis an inflammatory disease. N Engl J Med 1999;340:115-26. 2. Blake GJ, Ridker PM. High sensitivity C-reactive protein for predicting cardiovascular disease: an inflammatory hypothesis. Eur Heart J 2001;22:349-52. 3. Gutstein DE, Fuster V. Pathophysiology and clinical significance of atherosclerotic plaque rupture. Cardiovasc Res 1999;41:323-33. 4. Gabay C, Kushner I. Acute phase proteins and other systemic responses to inflammation. N Engl J Med 1999;340:448-54. 5. Caligiuri G, Liuzzo G, Biasucci LM, et al. Immune system activation follows inflammation in unstable angina: pathogenetic implications. J Am Coll Cardiol 1998;32:1295-304. 6. Liuzzo G, Biasucci LM, Rebuzzi AG, et al. Plasma protein acute phase response in unstable angina is not induced by ischemic injury. Circulation 1996;94:2373-80. 7. Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973-9. 8. 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
American Heart Journal Volume 144, Number 3
9.
10.
11.
12.
13.
MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation 1999;99:237-42. Ridker PM, Hennekens CH, Buring JE, et al. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342:836-43. Kuller LH, Tracy RP, Shaten J, et al. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case control study. Multiple Risk Factor Intervention Trial. Am J Epidemiol 1996;144:537-47. Haverkate F, Thompson SG, Pyke SD, et al. Production of C reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet 1997;349:462-6. Liuzzo G, Biasucci LM, Gallimore JR, et al. The prognostic value of C reactive protein and serum amyloid a protein in severe unstable angina. N Engl J Med 1994;331:417-24. Ridker PM, Rifai N, Pfeffer MA, et al. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in pa-
Speidl et al 455
14.
15.
16.
17.
18.
tients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation 1998;98:89-44. Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly: results from the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol 1997;17:1121-7. Kinne G, Droste C, Fahrenberg J, et al. Symptomatic myocardial ischemia and everyday life: implications for clinical use of interactive monitoring. J Psychosom Res 1999;46:369-77. Beranek JT. C-reactive protein and complement in myocardial infarction and postinfarction heart failure. Eur Heart J 1997;18:184-6. Heinrich J, Schulte H, Schonfeld R, et al. Association of variables of coagulation, fibrinolysis and acute phase with atherosclerosis in coronary and peripheral arteries and those arteries supplying the brain. Thromb Haemost 1995;73:374-9. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of Creactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem 1997;43:52-8.
Acute Ischemic Heart Disease
High-sensitivity C-reactive protein in the prediction of coronary events in patients with premature coronary artery disease Walter S. Speidl, MD, Senta Graf, MD, Stefan Hornykewycz, MD, Mariam Nikfardjam, MD, Alexander Niessner, MD, Gerlinde Zorn, Johann Wojta, PhD, and Kurt Huber, MD Vienna, Austria
Background and Methods Inflammation plays an important role in the initiation and progression of atherosclerosis and in the pathogenesis of acute cardiovascular events. Recent studies have indicated a possible association between C-reactive protein (CRP) and the clinical outcome of coronary artery disease (CAD). We studied prospectively in a group of 125 patients with premature CAD whether plasma levels of CRP as measured with a high-sensitivity assay predict risk for future coronary events. All patients had stable CAD at time of blood sampling but had originally been seen with unstable angina or myocardial infarction. The mean follow-up time after blood collection was 54 months, and death, myocardial infarction, need for coronary revascularization, or admission to hospital with angina pectoris were defined as clinical end points.
Results Patients in the highest tertile of CRP levels had a ⬎3.8-fold risk (risk ratio 3.82, 95% CI 1.19-12.17) for death, myocardial infarction, or need for coronary revascularization compared with the patients in the first tertile. The relative risk for patients in the second tertile was 3.5-fold higher (95% CI 1.04-11.56). CRP levels in the third tertile independently predicted risk after adjustment for lipids and other clinical risk factors. Conclusion In patients with clinically stable conditions who have a positive history for acute coronary syndromes before age 50 years, plasma levels of CRP higher than 1.6 mg/L are predictors of future coronary events and therefore indicate the role of underlying chronic inflammation for the clinical course of CAD. Accordingly, reference limits for prediction of risk in CAD have to be lower in this specific patient group than in middle-aged or elderly patients. (Am Heart J 2002; 144:449-55.)
Coronary artery disease (CAD) is a major cause of death in the Western world. Meanwhile, inflammation is accepted as playing an important role in the initiation and progression of atherosclerosis.1,2 Especially, the transition from stable advanced lesions to vulnerable plaques is paralleled by local accumulation of macrophages and secretion of cytokines and other destabilizing factors. These active inflammatory processes contribute to a thinning of the fibrous cap of lipid-rich plaques and to plaque erosion or rupture.3 As a side effect of local inflammation, circulating cytokines (eg, interleukin-6, interleukin-1, tumor necrosis factor–␣, oncostatin M, or leukemia inhibitory factor) activate hepatocytes to produce acute phase proteins.4,5 One
From the Department of Cardiology, University of Vienna, Austria. Submitted June 15, 2001; accepted February 20, 2002. Reprint requests: Kurt Huber, MD, Department of Cardiology, University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria. E-mail: [email protected] © 2002, Mosby, Inc. All rights reserved. 0002-8703/2002/$35.00 ⫹ 0 4/1/124353 doi:10.1067/mhj.2002.124353
of the most sensitive acute phase proteins is C-reactive protein (CRP). The plasma concentration of CRP is determined mainly by its synthesis rate, and, provided normal liver function exists, CRP is a sensitive marker for ongoing chronic inflammation that is not affected by ischemic injury.6 In recent studies, an association between CRP levels and the risk of future cardiovascular events has been discussed for healthy individuals7-9 and for individuals with an increased cardiovascular risk although apparently healthy.10 In addition, patients with stable or unstable angina pectoris11,12 and patients after myocardial infarction13 with elevated CRP levels are at high risk for future cardiovascular events. So far, trials have only investigated middle-aged and elderly patients,14 and less is known about the role of CRP levels to predict clinical course in patients with premature CAD. Therefore, the aim of this study was to investigate whether lower ranges of CRP as quantified with a highly specific commercial assay predict future risk for coronary events in patients with stable conditions with premature CAD.
American Heart Journal September 2002
450 Speidl et al
Table I. Characteristics of the study participants
No. Age (y) (mean ⫾ SD) Sex, male (%) Smoking status (%) Current smoker Former smoker Nonsmoker Systolic blood pressure (mm Hg) (mean ⫾ SD) Diastolic blood pressure (mm Hg) (mean ⫾ SD) Body mass index (kg/m2) (mean ⫾ SD) Total cholesterol (mg/dL) (mean ⫾ SD) LDL cholesterol (mg/dL) (mean ⫾ SD) HDL cholesterol (mg/dL) (mean ⫾ SD) Triglycerides (mg/dL) (mean ⫾ SD) Diabetes (%) Family history of CAD (%) Previous myocardial infarction (%) Number of coronary arteries narrowed ⬎70% (%) 0 1 2 3
Total
Without event
With event
P
119 39.3 ⫾ 5.6 76.5
71 38.7 ⫾ 5.2 74.6
48 40.1 ⫾ 6.2 79.2
.179 .569
76.5 12.6 10.9 114.6 ⫾ 16.2 75.4 ⫾ 12.2 26.9 ⫾ 4.4 209.9 ⫾ 48.1 137.4 ⫾ 41.2 35.0 ⫾ 11.0 184.5 ⫾ 113.8 5.2 62.9 78.2
70.4 15.5 14.1 113.8 ⫾ 15.7 75.6 ⫾ 11.8 26.7 ⫾ 4.2 203.3 ⫾ 50.3 130.1 ⫾ 41.1 37.0 ⫾ 11.4 167.3 ⫾ 107.8 2.9 62.3 77.5
85.4 8.3 6.3 115.6 ⫾ 17.1 75.1 ⫾ 12.9 27.3 ⫾ 4.6 219.6 ⫾ 43.3 148.2 ⫾ 39.3 32.0 ⫾ 9.9 211.3 ⫾ 118.8 8.5 63.6 79.2
17.2 45.7 25.0 12.1
20.6 48.5 20.6 10.3
12.5 41.7 31.3 14.6
.164
.591 .827 .47 .078 .018 .014 .048 .16 .888 .826 .382
CAD, Coronary artery disease.
Table II. Characteristics of the study participants according to tertiles of CRP Tertile
CRP (mg/L) (median) Range Age (y) (mean ⫾ SD) Sex, male (%) Smoking status (%) Current smoker Former smoker Nonsmoker Systolic blood pressure (mm Hg) (mean ⫾ SD) Diastolic blood pressure (mm Hg) (mean ⫾ SD) Body mass index (kg/m2) (mean ⫾ SD) Total cholesterol (mg/dL) (mean ⫾ SD) LDL cholesterol (mg/dL) (mean ⫾ SD) HDL cholesterol (mg/dL) (mean ⫾ SD) Triglycerides (mg/dL) (mean ⫾ SD) Diabetes (%) Family history of CAD (%) Previous myocardial infarction (%) Number of coronary arteries narrowed ⬎70% (%) 0 1 2 3
1
2
3
P
0.09 0.19-1.59 38.7 ⫾ 6.0 74.4
3.09 1.69-5.51 39.7 ⫾ 5.3 80.0
9.19 5.58-174.00 39.4 ⫾ 5.7 75.0
.751 .81
74.4 12.8 12.8 113.5 ⫾ 15.4 74.6 ⫾ 11.0 25.9 ⫾ 3.9 201.8 ⫾ 44.0 129.4 ⫾ 32.6 36.4 ⫾ 11.8 176.2 ⫾ 112.9 2.6 61.3 71.8
80.0 5.0 15.0 116.1 ⫾ 16.6 75.0 ⫾ 12.7 26.7 ⫾ 3.8 222.4 ⫾ 46.0 147.1 ⫾ 41.0 34.6 ⫾ 10.2 203.9 ⫾ 133.6 10.3 56.8 80.0
24.3 54.1 13.5 8.1
12.5 40.0 35.0 12.5
75.0 20.0 5.0 114.1 ⫾ 16.9 76.6 ⫾ 13.0 28.0 ⫾ 5.1 202.6 ⫾ 52.3 135.6 ⫾ 47.5 34.1 ⫾ 11.2 173.2 ⫾ 92.1 2.6 70.3 82.5 15.4 43.6 25.6 15.4
.223
.775 .761 .114 .151 .155 .628 .445 .213 .474 .485 .324
American Heart Journal Volume 144, Number 3
Methods Patients We investigated 125 consecutive patients with angiographically proven and clinically stable premature CAD who had originally attended the patient ward of the department of cardiology (University of Vienna) with signs of acute coronary syndromes (unstable angina or myocardial infarction). All patients were aged ⬍50 years (range 21-50 y, average 39.2 ⫾ 5.6 y) at the time of inclusion. Original diagnosis included the angiographic extent and severity of disease (⬎70% narrowing of the lumen of ⱖ1 major epicardial coronary artery). Exclusion criteria were any situations that might have influenced CRP plasma levels (eg, ongoing clinical instability, recent coronary interventions, or infectious diseases). The patients were followed for a maximum of 72 months, with an average follow-up time of 54.1 ⫾ 2.4 months. No follow-up data were obtainable after inclusion for 6 patients (4.8%). Primary clinical end points were death from CAD, nonfatal myocardial infarction, need for coronary revascularization, and admission to hospital in case of renewed angina pectoris. To detect and verify these study end points, most patients were seen in the outpatient ward. In rare cases, telephone and mail contact with the patients themselves or with their general practitioners was necessary. For patients who died during the follow-up period, the official national death register was reviewed.
Blood collection and assay systems At 8 to 12 weeks after the acute event, blood samples were taken from all subjects between 8 and 10 am after a 14-hour overnight fast. Venous blood was drawn directly into plastic tubes containing either ethylenediamine tetraacetic acid (5 ⫻ 10–2 mol/L final concentration) or citrate (0.11 mol/L final concentration). Tubes were centrifuged immediately at 3000g for 10 minutes at 4°C, and plasma was stored at – 40°C until use. Plasma levels of CRP were determined with the N high sensitivity CRP test (Dade Behring, Marburg, Germany). Plasma levels of triglycerides and cholesterol were determined with laboratory standard techniques.
Statistical analysis Data are represented as mean and standard deviation if not stated otherwise. Proportions or means for baseline risk factors were calculated. The significance of any differences in proportions was tested with the 2 test, and the significance of any difference in means was tested with analysis of variance. Because the distribution of CRP is rightward skewed, values derived from natural logarithm-transformed means were used as means for this variable and the significance of any differences in mean values of CRP was tested with analysis of variance of natural logarithm-transformed data. Median concentrations were computed for CRP, and the significance of any differences in median values was tested with the Wilcoxon rank-sum test. The patients were divided into tertiles according to CRP distribution for calculating the risk and for adjusted risk estimates, calculated with use of logistic-regression models, including clinical risk factors (age, sex, body mass index, smoking, diabetes, and family history for CAD) and blood lipids. All P values were 2 tailed, and CIs were calculated at the 95% level. All calculations were performed
Speidl et al 451
Figure 1
Kaplan-Meier survival plots for death, myocardial infarction, and need for revascularization procedure according to tertiles of CRP.
with a computer program (SPSS for Windows 10.0.1, Chicago, Ill).
Results During the follow-up period, 5 patients (4.2%) died from CAD, 10 patients (8.4%) had an acute myocardial infarction develop, 15 patients (12.6%) underwent coronary revascularization (coronary artery bypass grafting n ⫽ 5, percutaneous coronary intervention n ⫽ 10), and 18 patients (15.1%) were admitted to a hospital because of renewed angina pectoris. Seventy-one patients (59.7%) were completely free of coronary events during the follow-up period. The baseline characteristics of study participants as further subdivided into patients without or with further events are shown in Table I. Patients in whom a cardiovascular event developed tended to be smokers or to have diabetes, both of which were not statistically different compared with those without cardiovascular event. No differences were found for mean systolic or diastolic blood pressures, body mass index, or total cholesterol level. However, significant differences could be found for triglycerides (P ⫽ .048), low-density lipoprotein cholesterol level (P ⫽ .018), and highdensity lipoprotein cholesterol level (P ⫽ .014). No significant differences between patients without or with further events could be found in mean (2.66 ⫾ 37.7 mg/L vs 3.69 ⫾ 28.4 mg/L, P ⫽ .155) and median (2.48 mg/L vs 3.66 mg/L, P ⫽ .105) CRP levels. To calculate the relative risk, patients were divided into 3 groups according to the distribution of
American Heart Journal September 2002
452 Speidl et al
Table III. Relative risk of death, myocardial infarction, and need for revascularization procedure according to tertiles of CRP Tertile Variables adjusted for
1
2
3
Unadjusted Relative risk 1 3.472 3.816 95% CI 1.04-11.56 1.19-12.17 P .043 .024 Age Relative risk 1 3.469 3.808 95% CI 1.04-11.55 1.19-12.15 P .043 .024 Age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.019 3.810 95% CI 0.85-10.70 1.13-12.84 P .087 .031 Triglycerides Relative risk 1 3.364 3.840 95% CI 1.00-11.33 1.19-12.37 P .05 .024 Total cholesterol Relative risk 1 3.056 3.871 95% CI 0.90-10.36 1.20-12.49 P .073 .024 LDL cholesterol Relative risk 1 2.838 3.595 95% CI 0.83-9.73 1.11-11.69 P .097 .033 HDL cholesterol Relative risk 1 3.432 3.612 95% CI 1.00-11.77 1.10-11.83 P .05 .034 Total/HDL cholesterol Relative risk 1 3.090 3.772 95% CI 0.90-10.66 1.14-12.51 P .074 .03 Triglycerides and total/HDL cholesterol Relative risk 1 3.073 3.735 95% CI 0.89-10.63 1.13-12.38 P .076 .031 Triglycerides, total/HDL cholesterol, age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 2.924 3.728 95% CI 0.79-10.80 1.06-13.18 P .108 .041
P for trend
.026
.027
.032
.027
.025
.036
.039
.033
.034
.043
BMI, Body mass index.
CRP levels into tertiles. The patients in the 3 tertiles did not show any significant differences in risk factors (Table II). Patients in the tertile with the highest levels of CRP had a ⬎3.8-fold increased risk (risk ratio [RR] 3.82, P ⫽ .024) for harder end points (death, myocardial infarction, or need for coronary revascularization) compared with the patients in the first tertile (RR 1). The relative risk for patients in the second tertile was 3.5-fold increased (RR 3.47, P ⫽ .043) compared with the patients in the first tertile. The relative risk for development of any of the study end points also including admission to hospital with angina was ⬎2.7-fold
higher for patients in the third tertile (RR 2.73, P ⫽ .043) and ⬎3.6-fold elevated for patients in the second tertile (RR 3.68, P ⫽ .008). To show a relation between the time course of CAD and the plasma levels of CRP, we calculated KaplanMeier survival plots for the patients in the different tertiles (Figure 1). During the first 6 months, the event rate was similar in any CRP tertile group. Thereafter, the patients in the lowest tertile showed only a few combined primary end point events compared with the patients in the second and third CRP tertiles. This different course over time was statistically significant
American Heart Journal Volume 144, Number 3
Speidl et al 453
Table IV. Relative risk of death, myocardial infarction, need for revascularization procedure and admission to hospital with angina according to tertiles of CRP Tertile Variables adjusted for
1
2
Unadjusted Relative risk 1 3.684 95% CI 1.40-9.71 P .008 Age Relative risk 1 3.606 95% CI 1.36-9.56 P .01 Age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.238 95% CI 1.19-8.82 P .022 Triglycerides Relative risk 1 3.555 95% CI 1.33-9.52 P .012 Total cholesterol Relative risk 1 3.387 95% CI 1.27-9.03 P .015 LDL cholesterol Relative risk 1 3.205 95% CI 1.19-8.60 P .021 HDL cholesterol Relative risk 1 3.656 95% CI 1.35-9.87 P .011 Total/HDL cholesterol Relative risk 1 3.449 95% CI 1.27-9.38 P .008 Triglycerides and total/HDL cholesterol Relative risk 1 3.439 95% CI 1.26-9.35 P .016 Triglycerides, total/HDL cholesterol, age, sex, BMI, smoking, hypertension, diabetes and family history of CAD Relative risk 1 3.227 95% CI 1.14-9.11 P .027
(P ⬍ .05). To evaluate whether increased baseline CRP is an independent predictive marker, we performed logistic-regression analyses and adjusted for plasma lipids and the clinical risk factors age, sex, body mass index, hypertension, diabetes, and positive family history. CRP levels in the third tertile still remained an independent predictor of risk for death, myocardial infarction, or need for coronary revascularization (Table III) but lost independency for all study end points, including admission to hospital with angina (Table IV). To test a positive relation between the extent of CAD and CRP plasma levels, we correlated the median values of CRP with the number of main epicardial cor-
3
P for trend
2.727 1.03-7.20 .043
.051
2.684 1.01-7.13 .048
.056
2.608 0.95-7.15 .062
.055
2.849 1.06-7.65 .038
.044
2.816 1.06-7.52 .039
.043
2.630 0.98-7.05 .054
.062
2.586 0.96-6.99 .061
.073
2.751 1.01-7.53 .049
.056
2.773 1.01-7.62 .048
.054
2.703 0.94-7.75 .065
.064
onary arteries affected in the initial coronary angiography (Figure 2). A trend was seen, but no significant relationship between the extent of CAD and CRP plasma levels was found (P ⫽ .263). Similarly, for patients with a history of myocardial infarction, we found also a nonsignificant increase in CRP levels compared with those without such a history (data not shown).
Discussion Our data for the first time show a relationship between high-sensitivity CRP plasma levels and the risk
American Heart Journal September 2002
454 Speidl et al
Figure 2
Median values of CRP in mg/L according to number of affected coronary vessels (⬎70% narrowing of lumen).
of secondary coronary events in patients with stable conditions with premature CAD. In opposition to the higher CRP concentrations generated in older patients with stable or unstable angina11 or with a history of myocardial infarction,13 in our hands, CRP levels of ⬎1.6 mg/L were associated with an increased cardiovascular event rate. The fact that patients with combined hard end points (death, myocardial infarction, or need for coronary revascularization) showed an increased event rate with increasing CRP levels but this significant linear trend was lost when renewed angina was added may be explained by the fact that angina is a more subjective parameter, being related also to noncardiac reasons.15 For our patients, we can exclude an influence of acute phase reactions on the CRP levels because patients with acute infections, recent coronary interventions, or myocardial infarctions, or other conditions that may have altered CRP levels, were excluded from the study. Furthermore, blood was collected 8 to 12 weeks after an acute event and CRP plasma levels are known to decline to baseline within a few days.16 Patients showed a positive nonsignificant association between CRP plasma levels and the number of severely affected coronary arteries at time of angiographic control (8-12 weeks before the blood collection for CRP determination). A similar association has been described elsewhere for middle-aged individuals.17 Because some of the patients underwent immediate percutaneous coronary intervention or diagnostic angiography and because of the relatively small num-
ber of patients in our study, the association may have been insignificant. Our results of the positive role of CRP as predictor of further cardiovascular events in premature CAD open the question of whether different cut-off levels for CRP in CAD should be used for the prediction of coronary events because the risk for a future coronary event increases rapidly in young patients with CAD at lower CRP levels (CRP ⬎ 1.6 mg/L) compared with middle-aged and elderly patients.2 One possible explanation for this increase in risk at lower CRP levels in patients aged ⬍50 years may be explained by the fact that CRP levels are positively correlated with age.18 Several limitations of our data merit consideration. The relatively small number of patients involved may be causally related with the tendentious but nonsignificant results in several calculations. Furthermore, we only measured baseline CRP plasma levels once and have no evidence for alterations during the course of follow-up, which may be more accurate in terms of prediction of clinical course. On the other hand, CRP level are in most cases relatively constant over longer periods in stable diseases.18 In conclusion, in patients with stable premature CAD, CRP levels in the presently accepted upper normal range or higher are accompanied with an increased long-term risk for development of cardiovascular events, such as cardiac death, acute myocardial infarction, or need for revascularization. The high-sensitivity assay for exact determination of relatively low CRP levels is reliable, can be carried out easily, and might, therefore, represent a new and noninvasive tool to help to identify an increased risk for future coronary events in younger patients with CAD.
References 1. Ross R. Atherosclerosis an inflammatory disease. N Engl J Med 1999;340:115-26. 2. Blake GJ, Ridker PM. High sensitivity C-reactive protein for predicting cardiovascular disease: an inflammatory hypothesis. Eur Heart J 2001;22:349-52. 3. Gutstein DE, Fuster V. Pathophysiology and clinical significance of atherosclerotic plaque rupture. Cardiovasc Res 1999;41:323-33. 4. Gabay C, Kushner I. Acute phase proteins and other systemic responses to inflammation. N Engl J Med 1999;340:448-54. 5. Caligiuri G, Liuzzo G, Biasucci LM, et al. Immune system activation follows inflammation in unstable angina: pathogenetic implications. J Am Coll Cardiol 1998;32:1295-304. 6. Liuzzo G, Biasucci LM, Rebuzzi AG, et al. Plasma protein acute phase response in unstable angina is not induced by ischemic injury. Circulation 1996;94:2373-80. 7. Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973-9. 8. 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
American Heart Journal Volume 144, Number 3
9.
10.
11.
12.
13.
MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation 1999;99:237-42. Ridker PM, Hennekens CH, Buring JE, et al. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342:836-43. Kuller LH, Tracy RP, Shaten J, et al. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case control study. Multiple Risk Factor Intervention Trial. Am J Epidemiol 1996;144:537-47. Haverkate F, Thompson SG, Pyke SD, et al. Production of C reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet 1997;349:462-6. Liuzzo G, Biasucci LM, Gallimore JR, et al. The prognostic value of C reactive protein and serum amyloid a protein in severe unstable angina. N Engl J Med 1994;331:417-24. Ridker PM, Rifai N, Pfeffer MA, et al. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in pa-
Speidl et al 455
14.
15.
16.
17.
18.
tients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation 1998;98:89-44. Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly: results from the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol 1997;17:1121-7. Kinne G, Droste C, Fahrenberg J, et al. Symptomatic myocardial ischemia and everyday life: implications for clinical use of interactive monitoring. J Psychosom Res 1999;46:369-77. Beranek JT. C-reactive protein and complement in myocardial infarction and postinfarction heart failure. Eur Heart J 1997;18:184-6. Heinrich J, Schulte H, Schonfeld R, et al. Association of variables of coagulation, fibrinolysis and acute phase with atherosclerosis in coronary and peripheral arteries and those arteries supplying the brain. Thromb Haemost 1995;73:374-9. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of Creactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem 1997;43:52-8.