| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2007;115:3071-3078.)
© 2007 American Heart Association, Inc.
Coronary Heart Disease |
From the Department of Public Health and Primary Care (K.K.R.), University of Cambridge, Cambridge, United Kingdom; TIMI Study Group (D.A.M., M.S.S., A.S., C.P.C., E.B.), Cardiovascular Division and the Department of Medicine, Brigham and Womens Hospital/Harvard Medical School, Boston, Mass; and Childrens Hospital (N.R.), Boston, Mass.
Correspondence to Kausik K. Ray, MD, Department of Public Health and Primary Care, Strangeways Research Laboratory, University of Cambridge, Worts Causeway, Cambridge, CB1 8RN, United Kingdom. E-mail koshray{at}gmail.com
Received October 3, 2006; accepted April 2, 2007.
| Abstract |
|---|
|
|
|---|
Methods and Results Neopterin levels were measured on average at 7 days (in 3946 patients) and at 4 months (in 3369 patients) after ACS in the PRavastatin Or atorVastatin Evaluation Infection TherapyThrombolysis In Myocardial Infarction (PROVE ITTIMI 22) trial. We assessed the relationship between plasma neopterin levels and the risk of death and death or acute coronary events (nonfatal myocardial infarction or unstable angina) over 2 years. Seven days after an ACS event, neopterin levels
12.11 nmol/L (upper quartile, derived from a post hoc analysis) were associated with an increased risk of death and an increased risk of death or acute coronary events after adjustment for age, gender, history of diabetes mellitus, history of hypertension, history of smoking, type of ACS presentation, use of percutaneous coronary intervention for the index event, statin regimen, low-density lipoprotein cholesterol, and high-sensitivity C-reactive protein (hazard ratio, 1.86 [95% CI, 1.24 to 2.77], P=0.003; and hazard ratio, 1.33 [95% CI, 1.09 to 1.63] P=0.006, respectively). Neopterin levels
12.11 nmol/L at 4 months remained a predictor of death alone and of death or acute coronary events after multivariable adjustment that included adjustment for month 4 low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, and statin regimen (hazard ratio, 2.39 [95% CI, 1.43 to 3.99], P=0.001; and hazard ratio, 1.60 [95% CI, 1.21 to 2.11], P=0.001). High-dose atorvastatin significantly attenuated the risk among subjects with neopterin levels
12.11 nmol/L at baseline (interaction P for death or acute coronary event, 0.018).
Conclusions Increased monocyte activation detected by an elevated plasma neopterin level identifies patients at long-term risk of death or recurrent acute coronary events after ACS. Intensive statin therapy significantly attenuates the risk of recurrent events among patients with an elevated neopterin level.
Key Words: inflammation immune system statins myocardial infarction prevention prognosis
| Introduction |
|---|
|
|
|---|
Clinical Perspective p 3078
| Methods |
|---|
|
|
|---|
242 mg/dL, or if patients had been treated with chronic lipid-lowering therapy at the time of the index event,
203 mg/dL. Eligible patients were randomly assigned in a 1:1 ratio to pravastatin 40 mg/d or atorvastatin 80 mg/d and gatifloxacin versus placebo. Patients were followed up for 18 to 36 months, with an average follow-up of 24 months.
Blood Sampling and Neopterin Analysis
Baseline plasma samples were collected in EDTA (before the initiation of blinded study medication) and frozen at the study site at 20°C or lower and then shipped on dry ice to the TIMI Biomarker Core laboratory (Boston, Mass), where the specimens were subsequently stored at 70°C. Samples were also collected at 30 days, 4 months, and end of study (average 2 years). Assays for neopterin were performed with a commercially available ELISA kit (ALPCO Diagnostics, Windham, NH) by personnel blinded to treatment allocation and outcomes on a robotic high-throughput platform (TECAN Genesis RSP 200/8). The assay uses a competitive enzyme immunoassay technique with a lower detection limit of 2.0 nmol/L and the day-to-day imprecision values at concentrations of 7.91 and 89.6 nmol/L of 3% and 6%, respectively. High-sensitivity testing for C-reactive protein (CRP) and lipids was performed as described previously.12
Clinical End Points
The clinical end points for this analysis included death alone, death or nonfatal myocardial infarction (MI), and the composite of death, nonfatal MI, or unstable angina (UA) that required rehospitalization. All end points were adjudicated independently by the clinical events committee.11
Statistical Analyses
Neopterin was measured at randomization (baseline) in 3946 patients and at 4 months in 3369 patients. We also randomly identified subjects from the pravastatin arm (n=198) who were free from clinical events by end of study and who had provided us with a core laboratory blood sample at 4 time points (baseline, 30 days, 4 months, and end of study). We then identified a similar group of patients from the atorvastatin arm (n=220) after matching for age, gender, history of diabetes mellitus, history of hypertension, history of smoking, and type of ACS presentation.
Normality was assessed with the Kolmogorov-Smirnov test, and variables that significantly differed from normality were assessed with nonparametric tests. Neopterin levels during follow-up were compared against baseline with ANOVA after log transformation. The effect of intensity of statin therapy on neopterin levels was similarly assessed. The correlation between neopterin levels in the 418-patient cohort at different time points was assessed by the Spearman correlation coefficient.
Patients were divided into quartiles on the basis of baseline concentrations of neopterin. Means (or medians) and proportions of baseline variables were compared across quartiles of neopterin with the Kruskal-Wallis test for continuous variables and
2 tests for categorical variables. To evaluate its association with clinical outcomes, neopterin was considered a categorical variable based on quartiles of baseline concentration. Cox regression analysis was used to evaluate the association between neopterin at baseline and at 4 months and the risk of adverse outcomes on average over 2 years. The risk of adverse outcomes was then assessed further with neopterin as a dichotomous variable to compare patients with the highest neopterin levels (
75th percentile) with those whose values were below the 75th percentile. Stratified analyses were performed with this dichotomous cut point across patients with various baseline characteristics and among patients receiving intensive versus standard statin therapy.
For the clinical end points of interest, we constructed a Cox proportional hazards model that adjusted for age, gender, history of diabetes mellitus, hypertension, smoking, type of ACS presentation, use of percutaneous coronary intervention for the index event, statin regimen, low-density lipoprotein cholesterol (LDL-C), and high-sensitivity CRP (hsCRP) to assess the hazard over 2 years. Approximately 4% and 1% of patients had missing data for 1 or more of these variables at baseline and 4 months, respectively. Only patients for whom data on all of these clinical variables were available were entered into the multivariable model (n=3797 at baseline and n=3326 at month 4). Scaled and unscaled Schoenfeld residuals were used to test for proportionality of hazards and deviance of residuals to graphically assess the linearity of age-related effects. CRP and LDL-C were entered as continuous variables into the main models, except when specifically stated, and dichotomous cut points of 2 mg/L and 70 mg/dL were used, respectively.13 We also assessed the benefit of intensive statin therapy among patients with the highest neopterin levels (
75th percentile) using terms for the main effect and interaction between therapy and neopterin in the full multivariable model.
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
| Results |
|---|
|
|
|---|
Stability of neopterin levels within individuals was assessed in a cohort of 418 patients at 30 days, 4 months, and the end of the study. Median neopterin levels were lower during follow-up than at baseline (Table 1); however, within individuals, neopterin at baseline correlated well with levels at 4 months and the end of the study (intraclass correlation for both r=0.60, P<0.0001). Similarly, neopterin levels at 4 months were correlated with levels at the end of the study (r=0.63, P<0.0001).
|
Relationship Between Neopterin and Baseline Variables
In univariate analyses, higher baseline levels of neopterin were more likely to be associated with older age, a prior history of hypertension or diabetes, lower LDL-C levels, and higher hsCRP levels and less likely to be associated with a history of smoking or percutaneous coronary intervention for the index event (Table 2). There was no consistent relationship between quartiles of neopterin and gender or the type of index ACS presentation. Although statistically significant, the association between concentrations of neopterin and hsCRP (r=0.25) was weak.
|
Relationship Between Neopterin and Clinical Outcomes
Seven days after ACS, median neopterin levels were higher among subjects who subsequently died during follow-up (11.99 versus 9.62 nmol/L, P<0.0001) or who either died or experienced a nonfatal coronary event (10.34 versus 9.58 nmol/L, P<0.0001). Similarly, 4-month median neopterin levels were also higher among subjects who experienced a recurrent event after month 4 (death, 11.62 versus 8.76 nmol/L, P<0.0001; death or nonfatal coronary event, 9.54 versus 8.75 nmol/L, P<0.0001).
Subjects in the highest quartile of neopterin values at baseline (
12.11 nmol/L) had the highest rate of death (5.30%) compared with those in quartiles 1 to 3 (1.42%, 2.07%, and 2.43%, respectively, P<0.0001) at 2 years. Similarly, baseline neopterin levels
12.11 nmol/L were associated with the highest rate of death or MI (13.66%) compared with those in quartiles 1 to 3 (6.94%, 7.85%, and 8.10%, respectively, P<0.0001) and with the highest rate of death/MI or UA (16.98% versus 10.52%, 11.40%, and 12.01% for quartiles 1 to 3, respectively, P<0.0001; Figure 1). Neopterin levels
12.11 nmol/L at baseline tended to be associated with increased risk of death/MI or UA (Figure 2) compared with subjects with neopterin levels <12.11 nmol/L across a range of baseline characteristics. Similarly, subjects with neopterin levels
12.11 nmol/L at 4 months (which corresponds to the 82nd percentile) had higher event rates than those with neopterin levels <12.11 nmol/L (death 4.6% versus 1.2%, death or MI 10.2% versus 4.5%, and death/MI or UA 14% versus 6.7%, P<0.0001 for each).
|
|
Neopterin levels
12.11 nmol/L were associated with an increased risk of death and of death or acute coronary events after adjustment for age, gender, history of diabetes mellitus, history of hypertension, history of smoking, type of index ACS event, percutaneous coronary intervention for the index event, statin regimen, baseline LDL-C, and hsCRP (Table 3). At 4 months after the index ACS event, neopterin levels
12.11 nmol/L were again significantly associated with death and death or acute coronary events after adjustment for the baseline characteristics described previously and month 4 LDL-C and month 4 hsCRP (Table 3). In addition to assessing a dichotomized cut point, we also assessed neopterin as a continuous variable. After multivariable adjustment, a 1-nmol/L higher neopterin level at baseline was associated with an increased risk of death (hazard ratio [HR], 1.03 [95% CI, 1.01 to 1.06]), death or MI (HR, 1.03 [95% CI, 1.01 to 1.04]), and death or nonfatal coronary events (HR, 1.02 [95% CI 1.01 to 1.04]). The corresponding risk at 4 months for a 1-nmol/L higher neopterin level were HR of 1.02 (95% CI, 0.99 to 1.04) for death; HR, 1.02 (95% CI, 1.00 to 1.03) for death or MI; and HR, 1.02 (95% CI, 1.00 to 1.04) for death or nonfatal coronary events.
|
Comparison of Elevated Neopterin at Month 4 With Other Risk Factors
After multivariable adjustment, the long-term hazard of death, MI, or UA related to a neopterin level
12.11 nmol/L (HR, 1.52 [95% CI, 1.15 to 2.0]) was similar in magnitude to that of other risk factors, including hsCRP
2 mg/L (HR, 1.49 [95% CI, 1.16 to 1.9]; Figure 3). Even when a more extreme hsCRP cut point of 5.12 mg/L, which corresponds to the 82nd percentile, was used, the corresponding hazards for neopterin and CRP remained comparable (HR, 1.53 [95% CI, 1.16 to 2.02]; and HR, 1.47 [95% CI, 1.12 to 1.94], respectively). As at baseline, the correlation between neopterin and hsCRP remained weak but significant (r=0.23, P<0.001). Importantly, among subjects who achieved the dual goals of LDL <70 mg/dL and CRP <2 mg/L,14 a neopterin level
12.11 nmol/L identified a subset of patients at increased long-term risk of death or acute coronary events (Figure 4). Patients who did not achieve dual goals and additionally had a higher neopterin level were at greatest risk.
|
|
Interaction Between Statin Regimen, Neopterin Levels, and Clinical Outcomes
There was no statistical difference in absolute neopterin levels at 30 days, 4 months, or end of study between standard (pravastatin 40 mg/d) and intensive (atorvastatin 80 mg/d) statin therapy, although a nominally greater relative reduction was observed at 4 months with atorvastatin 80 mg/d (Table 1). We also assessed the effect of gatifloxacin on neopterin levels versus placebo and did not find any significant absolute or relative differences in levels at any of the time points assessed.
When subjects were stratified by statin regimen, intensive statin therapy attenuated the risk of death, MI, or UA in subjects within the highest quartile of neopterin at baseline (Figure 5). For subjects with neopterin levels <12.11 nmol/L, the hazard ratio of death or nonfatal coronary events was 0.89 (95% CI, 0.72 to 1.11) for atorvastatin 80 mg versus pravastatin 40 mg; by comparison, subjects with neopterin levels
12.11 nmol/L had a hazard ratio of 0.57 (95% CI, 0.42 to 0.78; interaction P=0.022). After adjustment for age, gender, history of diabetes, history of hypertension, history of smoking, type of ACS presentation, use of percutaneous coronary intervention for the index event, statin regimen, baseline LDL-C, and hsCRP, the interaction between higher neopterin levels (
12.11 nmol/L) and attenuation of clinical risk with high-dose atorvastatin remained significant (interaction P for death, 0.27; for death or MI, 0.046; and for death/MI or UA, 0.018).
|
| Discussion |
|---|
|
|
|---|
Enhanced T-cell activity that results in increased production of interferon-
is implicated in the pathogenesis of ACS,15,16 and neopterin production by monocytes and macrophages is primarily in response to stimulation by interferon-
released by activated T lymphocytes.1,7 Therefore, circulating neopterin levels may reflect the overall level of cell-mediated immunity within individuals. Previous small studies have demonstrated that neopterin levels are higher in patients with ACS than in patients with stable coronary artery disease or apparently normal subjects.1719 Additionally, neopterin is not correlated with markers of myonecrosis, which suggests that immune activation may contribute to ACS rather than reflect an immune response to myocardial injury.19 In particular, among patients with ACS, neopterin levels are higher in patients with complex vulnerable coronary stenoses and correlate with the total number of complex stenoses present.2022 Patients with stable coronary artery disease who are more likely to develop rapidly progressive coronary artery disease also appear to have higher neopterin levels than patients in whom coronary artery disease does not progress rapidly.23 Taken together, these findings suggest that elevated neopterin levels may reflect atherosclerotic disease activity and vulnerability to the development of ACS.
Recently, the potential prognostic value of an elevated neopterin level was observed in 297 patients with stable coronary artery disease during 1 year of follow-up.24 In the study by Avanzas et al,24 the relationship between neopterin and risk of cardiac death or MI/UA was not linear across tertiles and was more suggestive of a "threshold effect" (cumulative rates, 12.1%, 14.1%, and 25.3% across tertiles 1 to 3, respectively), which has also been observed with other inflammatory biomarkers in the context of ACS.2527 For instance, our group has previously shown that monocyte chemoattractant protein-1 levels above the 75th percentile in ACS patients (which corresponds to greater than the 90th percentile in healthy volunteers) is an independent predictor of death or MI at 10 months, but there is little gradation in risk between quartiles 1 to 3.25 Large-scale data on neopterin cut points in ACS are lacking, and the largest reported study on 210 patients with nonST-elevation MI from the FLORIDA (FLuvastatin On RIsk Diminishing after Acute myocardial infarction) trial used a threshold of 9.7 nmol/L, which represented the lower limit of the upper tertile, but observed qualitatively similar results to the present findings over a 1-year period of follow-up.28 We also assessed neopterin as a continuous variable and observed that for each 1-nmol/L increase in neopterin, clinical risk increased by 2% to 3%. Taken together, these data suggest a continuous relationship between neopterin levels and clinical risk, which is perhaps more marked at higher levels. The present study therefore extends previous observations to nearly 4000 patients with ACS in a contemporary setting at 2 time points over a 2-year period of follow-up.
In small studies, statin use has been associated with lower neopterin levels,28,29 and statins have been shown to reduce interferon-
stimulated release of neopterin from monocytes.30 In the present study, intensive statin therapy (80 mg of atorvastatin daily) resulted in only a nominally greater relative reduction in neopterin levels at 4 months (9.3% versus 8.7%) than with a standard dose of statin (40 mg of pravastatin daily) but was not associated with significant differences in absolute levels. However, use of intensive therapy was associated with a more marked attenuation in clinical risk among subjects with higher neopterin levels at baseline, which suggests that the clinical risk associated with an elevated neopterin level may be modifiable. The present findings are therefore analogous to the benefits of aspirin in reducing cardiac risk among subjects with high hsCRP.31 Similar findings have emerged recently in other diseases; for instance, in renal allograft recipients, an elevated neopterin level identified a group of patients who appeared to benefit from specific immunosuppressive therapy.5
Some markers of inflammation, such as interleukin-6 and serum amyloid A, correlate strongly with hsCRP and therefore may not offer significant additional information. Additionally many of these acute-phase reactants are elevated in response to myocardial injury early after ACS. Although neopterin could be considered an inflammatory marker, it was only weakly correlated with hsCRP both at 7 days and at 4 months after ACS, and it was relatively stable 7 days after ACS, in contrast to hsCRP.13 The measurement of neopterin may therefore provide information beyond that provided by hsCRP. This is supported by our observation that a high neopterin level appeared equally predictive of death or acute coronary events among patients with high or low baseline levels of hsCRP. Even after adjustment for hsCRP, a high neopterin level measured at 2 different time points remained an independent risk predictor for long-term events, which suggests that it provides information that is supplementary to and complementary to that provided by hsCRP. Recently, the clinical relevance of achieving the dual goal of CRP <2 mg/L and LDL <70 mg/dL in ACS patients has been demonstrated.13,14,32 Significantly, the present data demonstrate that a high neopterin level provides additional prognostic information both in subjects who achieve and those who fail to reach dual goals.
In nonvascular disease such as infections, connective tissue disease, allograft rejection, and malignancy, which may all have an autoimmune component, levels of neopterin in body fluids not only provide insight into the state of the cell-mediated immune response but have also allowed monitoring of treatment efficacy and disease progression.2,3,5,33 For example, in Austria, additional testing of neopterin among blood donors was made mandatory since 1986, such that between 1986 and 2002, nearly 900 000 voluntary blood donations were screened as part of routine practice with commercially available assays that used a cut point of 10 nmol/L to define an abnormal result.34 Importantly, in the present large cohort, the intraclass correlation between neopterin at 7 days after ACS and at 2 years was 0.6, which was comparable to the long-term intraindividual variance in both hsCRP and LDL-C.13 This suggests that neopterin levels are relatively stable in any given individual over time. When the same cut point of 12.11 nmol/L, initially assessed at baseline, was used, which corresponded to the 82nd percentile at 4 months, a high neopterin level at time points distant from the presenting syndrome remained a strong predictor of adverse clinical events independent of hsCRP. The present findings support a putative role of the monocyte-macrophage system in plaque instability by providing large-scale clinical evidence for a potentially pathogenic role of monocyte activation in plaque vulnerability.
Study Limitations
We studied a cohort of ACS patients selected for participation in a clinical trial. Although the qualitative relationships are likely to be similar, the quantitative association merits evaluation in unselected patients presenting with ACS. In consecutive unselected patients with stable coronary artery disease, "threshold effects" have been observed with neopterin levels of 8.8 and 7 nmol/L that identify patients more likely to develop progressive angiographic disease or to experience adverse cardiac events,23,24 and a cut point of 9.7 nmol/L has been evaluated in selected nonST-elevation MI patients.28 The neopterin cut point of
12.11 nmol/L used to define a higher level in the present study was not specified a priori and was based on an assessment of quartiles of neopterin at study entry. However, a cut point of 12 nmol/L has been used to define an elevated level of neopterin in a study of 20 000 healthy blood donors.35 Nevertheless, validation of this cut point and the interaction with statin therapy will be valuable in assessing the candidacy of neopterin for use in the routine clinical evaluation of patients after ACS. We also observed a significant relationship between neopterin levels and outcomes when assessed as a continuous variable.
Conclusion
Elevated neopterin levels at hospital discharge and at 4 months after an ACS event identified a group of patients at increased long-term risk of death and nonfatal coronary events.
| Acknowledgments |
|---|
Sources of Funding
The PROVE IT trial was funded by Bristol-Myers Squibb and Sankyo Co Ltd. The TIMI Study Group and Brigham and Womens Hospital received an unrestricted research grant from Pfizer Inc. Funding sources had no access to the biomarker database and did not make any contribution to the intellectual content of this manuscript.
Disclosures
The TIMI Study Group has received significant research grant support from Accumetrics, Amgen, AstraZeneca, Bayer Healthcare, Beckman Coulter, Biosite, Bristol-Myers Squibb, CV Therapeutics, Eli Lilly and Co, GlaxoSmithKline, Inotek Pharmaceuticals, Integrated Therapeutics, Merck and Company, Merck-Schering-Plough Joint Venture, Millennium Pharmaceuticals, Novartis Pharmaceuticals, Nuvelo, Ortho-Clinical Diagnostics, Pfizer, Roche Diagnostics, Sanofi-Aventis, Sanofi-Synthelabo, and Schering-Plough. Dr Ray has received honoraria for educational presentations from Pfizer, Novartis, Sanofi-Aventis, Bristol-Myers Squibb, and AstraZeneca. He has served as a consultant for Pfizer and Novartis. Dr Morrow has received honoraria for educational presentations from Bayer Diagnostics, Beckman-Coulter, Dade-Behring, Sanofi-Aventis, and Roche Diagnostics. He has served as a consultant for GlaxoSmithKline and Sanofi-Aventis and on advisory boards for Critical Diagnostics, Genentech, Ortho-Clinical Diagnostics, and Beckman-Coulter. Dr Sabatine has received honoraria for educational presentations from Amgen, Bristol-Myers Squibb, Novartis, and Sanofi-Aventis. He has served as a consultant to Daiichi Sankyo and Millenium Pharmaceuticals and on advisory boards for AstraZeneca, Amgen, Bristol-Myers Squibb, and Sanofi-Aventis. Dr Rifai has received research support from Merck Research Laboratories, is a consultant for Sanofi-Aventis, and has received an honorarium for a lecture from Ortho Diagnostics. Dr Cannon currently serves as a consultant on scientific/advisory boards for AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Merck, Pfizer, Sanofi-Aventis, and Schering-Plough. He has also received honoraria for continuing medical educationapproved lectures or preparation of educational material for Accumetrics, AstraZeneca, Bristol-Myers Squibb, Merck, Pfizer, Sanofi-Aventis, Schering-Plough, BGB New York, Discovery Institute of Medical Education, and the National Council on Measurement in Education. Dr Braunwald reports to have participated occasionally (maximum of 2 to 3 times per year) in symposia/advisory board meetings/consultancies for the following companies for which he has received honoraria: AstraZeneca Pharmaceuticals, Bayer AG, Daiichi Sankyo, Merck & Co, Pfizer, and Schering-Plough.
| References |
|---|
|
|
|---|
12.11 nmol/L) at hospital discharge were significantly associated with an increased risk of death or of either death or nonfatal coronary events after adjustment for confounders including high-sensitivity C-reactive protein (hazard ratio, 1.86 [95% CI, 1.24 to 2.77] and hazard ratio, 1.33 [95% CI, 1.09 to 1.63], respectively). Similarly, 4 months after ACS, neopterin levels
12.11 nmol/L remained a powerful predictor of risk (hazard ratio for death, 2.39 [95% CI, 1.43 to 3.99], and hazard ratio for death or nonfatal events, 1.60 [95% CI, 1.21 to 2.11]). Significantly, elevated neopterin identified high-risk subgroups even in populations with both 4-month low-density lipoprotein cholesterol <70 mg/dL and C-reactive protein <2 mg/L (61% increased risk versus low-neopterin group). Among subjects with an elevated neopterin level, high-dose statin therapy significantly attenuated the risk of adverse events. We conclude that neopterin appears to be a stable novel marker of immune activation that provides important prognostic information over and above traditional and novel risk factors and that may be clinically useful in identifying subjects who would benefit from intensive statin therapy.
| Footnotes |
|---|
Clinical trial registration informationURL: http://www.clinicaltrials.gov. Unique identifier: NCT00382460.
Related Article:
Circulation 2007 115: 3039.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2007 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |