ATHEROSCLEROSIS IS CHARACterized

Transcription

1 ORIGINAL CONTRIBUTION Relationship Between Interleukin 6 and Mortality in Patients With Unstable Coronary Artery Disease Effects of an Early Invasive or Noninvasive Eva Lindmark, MMed Erik Diderholm, MD Lars Wallentin, MD, PhD Agneta Siegbahn, MD, PhD See also pp 2136 and Context Inflammatory activity is associated with high rates of long-term mortality in unstable coronary artery disease (CAD). Interleukin 6 (IL-6) induces C-reactive protein and fibrinogen, systemic markers of inflammation. Objectives To determine whether plasma levels of IL-6 are predictive of mortality and to evaluate the interaction of IL-6 levels with the effects of invasive vs noninvasive treatment strategies in unstable CAD patients. Design, Setting, and Patients The prospective, randomized Fragmin and Fast Revascularisation During Instability in Coronary Artery Disease II trial, conducted among 3489 patients, 3269 of whom had plasma samples analyzed for IL-6 levels, with diagnosed unstable CAD (67% male; median age, 67 years) at 58 Scandinavian hospitals between June 1996 and August Interventions Patients were randomly assigned to receive either an early invasive (n=1222) or a noninvasive treatment strategy (n=1235). The latter group, as well as 666 patients with contraindications to invasive therapy, were further randomized to 90-day treatment with low-molecular-weight heparin (dalteparin, IU twice per day; n=1140) or placebo (n=1127). Main Outcome Measure Mortality at 6 and 12 months in the medically and interventionally randomized cohorts, respectively, in relation to IL-6 levels, measured at randomization. Results Plasma levels of IL-6 that were at least 5 ng/l compared with levels lower than 5 ng/l were associated with greatly increased mortality in the noninvasive group (7.9% vs 2.3%; relative risk [RR], 3.47; 95% confidence interval [CI], ) and in the placebo-treated group (7.9% vs 2.5%; RR, 3.19; 95% CI, ). The association remained significant after adjustment for most established risk indicators. An early invasive treatment strategy strongly reduced 12-month mortality among those with elevated IL-6 levels (5.1% absolute reduction; P=.004) whereas mortality was not reduced among patients without elevated IL-6 concentrations. Those taking dalteparin with elevated IL-6 levels experienced lower 6-month mortality than those who did not take dalteparin (3.5% absolute reduction; P=.08). Conclusions Circulating IL-6 is a strong independent marker of increased mortality in unstable CAD and identifies patients who benefit most from a strategy of early invasive management. JAMA. 2001;286: ATHEROSCLEROSIS IS CHARACterized by inflammation in the vessel wall. 1 On the site of the atherosclerotic plaque, the intima is infiltrated with activated macrophages and T lymphocytes, which both produce and secrete cytokines to drive the inflammatory process on. In unstable coronary artery disease (CAD), increased levels of systemic markers of inflammation, such as the acute phase reactants C-reactive protein and fibrinogen, are common. 2-4 Also, increased plasma levels of proinflammatory cytokines, such as interleukin (IL) 1, IL-6, and IL-8, have been reported. 4-6 Large cohort studies have shown an association between elevated levels of circulating C- reactive protein or fibrinogen and increased risk for cardiac events or death, both in patients who have experienced an episode of unstable CAD 2,7,8 as well as in apparently healthy individuals Interleukin 6 is a cytokine with both proinflammatory and anti-inflammatory effects on many cell types, affecting both B-cell immunoglobulin production and T-cell cytotoxic activity. 14 Interleukin 6 also affects platelet production and reactivity as well as endothelial function. It is the only substance known to induce synthesis of all of the acute phase proteins by the liver. Large prospective studies of healthy populations have shown that IL-6 plasma levels in the upper quartile of the considered normal range are inde- Author Affiliations: Department of Medical Sciences, Laboratory for Coagulation Research, Clinical Chemistry (Ms Lindmark and Dr Siegbahn) and Cardiology (Drs Diderholm and Wallentin), Uppsala University Hospital, Uppsala, Sweden. Corresponding Author and Reprints: Agneta Siegbahn, MD, PhD, Department of Medical Sciences, Clinical Chemistry, University Hospital, S Uppsala, Sweden ( agneta.siegbahn@clm.uas.lul.se) American Medical Association. All rights reserved. (Reprinted) JAMA, November 7, 2001 Vol 286, No

2 Figure 1. Overview of the Main Fragmin and Fast Revascularization During Instability in Coronary Artery Disease II Trial 1140 Assigned to Receive Dalteparin 666 With Contraindications to Early Revascularization 366 Late Enrollment 2267 Randomized in Medical Study 3489 Patients Admitted to Hospital With Chest Pain and Signs of Ischemia pendently predictive of an increased risk of premature death or future myocardial infarction (MI), even after accounting for C-reactive protein level. 13,15 This indicates a possible role for IL-6 in the progression of CAD. We have previously shown a strong relationship between increased levels of C-reactive protein, fibrinogen or troponin T and increased risk for long-term mortality in patients with unstable CAD. 8 Since IL-6 induces both C-reactive protein and fibrinogen, we sought to determine its predictive value for long-term risk of death or MI in patients included in the Fragmin and Fast Revascularisation During Instability in Coronary Artery Disease II (FRISC II) trial. We also investigated the effects of an invasive vs a noninvasive treatment strategy and prolonged treatment with low-molecularweight heparin (dalteparin) in relation to IL-6 levels at admission. METHODS Patients In all, 3489 patients were recruited at 58 Scandinavian hospitals between June 1996 and August Patients were eligible for study inclusion if they had 1235 Assigned to Noninvasive 1127 Assigned to Receive Placebo 2457 No Contraindications to Early Revascularization 2457 Randomized in Interventional Study 1222 Assigned to Invasive and Dalteparin or Placebo Three hundred sixty-six patients enrolled after the end of recruitment for the intervention trial. symptoms of ischemia that were increasing or occurring at rest or that warranted the suspicion of acute MI, with the last episode within 48 hours before initiating dalteparin or heparin treatment. Myocardial ischemia had to be verified by electrocardiography (STsegment depression 0.1 mv or T- wave inversion 0.1 mv) or by raised biochemical markers (creatine kinase (CK)-MB 6 µg/l, troponin T 0.10 µg/l, positive qualitative troponin-t test, or catalytic activity of CK, CK-B, or CK-MB higher than the diagnostic limit for MI). Exclusion criteria were increased risk for bleeding episodes, anemia, indication for treatment in the past 24 hours with thrombolysis, angioplasty in the previous 6 months, being on a waiting list for a coronary revascularization procedure, other acute or severe cardiac disease, renal or hepatic insufficiency, known clinically relevant osteoporosis, other severe illness, hypersensitivity to randomized drugs, anticipated difficulties with cooperation, or previous participation in this or another clinical trial. Patients with previous open heart surgery, advanced age ( 75 years), or other disorders that made randomization to early revascularization inappropriate were excluded from randomized intervention strategies but were still assigned to receive dalteparin or placebo. Study Design The FRISC II study was a prospective, randomized, multicenter trial with parallel groups. The details of the FRISC II medical and interventional studies have been described previously. 16,17 A simplified overview of the study design and the subgroups evaluated in this article are shown in FIGURE 1. All patients were initially treated with subcutaneous dalteparin or standard heparin by intravenous infusion. Randomization into the different treatment strategies was completed within 72 hours of start of open-label dalteparin or standard heparin. All patients received dalteparin subcutaneously (120 IU/kg every 12 hours) for at least 5 days until they underwent an exercise test or revascularization. Thereafter, they entered the double-blind treatment with twicedaily subcutaneous injections of dalteparin (5000 IU per dose for men weighing 70 kg and women 80 kg; 7500 IU per dose for those who exceeded these weight limits) or placebo until 90 days after entry. The direct invasive strategy required coronary angiography within a few days of enrollment, aiming for revascularization within 7 days of the start of open-label treatment. Revascularization was recommended in all patients with an obstruction of at least 70% of the diameter of any artery supplying a substantial proportion of the myocardium. Percutaneous coronary intervention was recommended if there were 1 or 2 target lesions, and coronary artery bypass surgery was preferred in patients with 3-vessel or left main artery disease. In the noninvasive strategy, coronary angiography and, if appropriate, revascularization were recommended in patients with refractory or recurrent symptoms despite maximum medical treatment or severe ischemia on an exercise electrocardiography test before discharge. During follow-up, invasive procedures were considered, irrespec JAMA, November 7, 2001 Vol 286, No. 17 (Reprinted) 2001 American Medical Association. All rights reserved.

3 tive of randomized strategy, for all patients with incapacitating symptoms, recurrence of instability, or MI. Aspirin was given to all patients on admission (initial dose, mg; maintenance dose, mg once daily), and -blockers were given unless contraindicated. Statins for lowering cholesterol, angiotensin-converting enzyme inhibitors, and aggressive antidiabetic treatment were given at the discretion of the treating physicians. End Points The end points investigated in this study were all-cause death and the composite of death or nonfatal MI, which was the primary end point of the FRISC II trial. The medical study had a follow-up for 6 months. Concerning the interventional trial, longer follow-up was preplanned and, hence, information on death or MI was also available at 12 months. Follow-up was unavailable for 14 patients in the dalteparin group and for 18 patients in the placebo group. In the interventional study, vital status was not available for 1 patient who requested withdrawal from the study. Definitions of the various predefined end points in FRISC IIhavebeenpreviouslydescribed. 16,17 The study complied with the Declaration of Helsinki, and all local ethics committees approved the protocol. Informed consent was obtained from all included patients. Analyses Venous blood samples in tubes containing EDTA (Vacutainer, Becton- Dickinson, Plymouth, England) were taken from all patients at randomization. The plasma was separated by centrifugation (2000 g for 20 minutes) within 30 minutes of blood sampling, aliquoted, and stored at 70 C until analysis. Troponin T levels were measured with a third-generation assay kit on an Elecsys 2010 instrument (Roche- Boehringer Mannheim, Mannheim, Germany). Concentrations of C- reactive protein were analyzed using the Immulite Automated Analyzer and assay kit (DPC, Diagnostic Products Corp, Los Angeles, Calif). The IL-6 antigen levels were measured in plasma from 3269 patients by a sequential immunometric assay, also using the Immulite analyzer and kit. The lower detection limit of this system is 5 ng/l, which was therefore used as the cutoff for the statistical analyses. Levels higher than the cutoff are reported herein as higher than 5 ng/l, high, raised, orelevated. Statistical Analyses The levels of IL-6, below or above the cutoff level were used to test the association between IL-6 and outcome events. All statistical comparisons between randomized treatments were performed according to the intention-totreat principle. In the medical part of the study, events were recorded from the initiation of the open-label dalteparin treatment until the 6-month follow-up. In the interventional study, recording of events terminated at the 12-month follow-up. The efficacy analyses of the 6-month and 12-month follow-ups were point estimates including only patients with an adjudicated event or with recorded absence of the specific event until at least 170 days and 335 days of the respective follow-up period. Pearson 2 analysis was used to test significance of the overall degree of association. Graphs of the Kaplan-Meier estimate of the survival function were used without statistical tests. Forward stepwise logistic regression analysis was used to adjust for established risk indicators regarding mortality and for evaluating factors that could contribute to elevated IL-6 levels. All P values are 2-tailed, and values lower than.05 were considered statistically significant. Relative risks (RRs) and odds ratios (ORs) are expressed with (95% confidence intervals [CIs]). Data processing and statistical analyses were performed using SPSS version 10.0 software (SPSS, Chicago, Ill). RESULTS Patient Characteristics and Procedures As shown in Figure 1, 2267 patients were included in the medical part of the FRISC II trial while 2457 were enrolled in the interventional trial. Within 7 days, 96% of patients in the invasive group and only 7% of the patients in the noninvasive group underwent coronary angiography. Within the first 10 days, 71% in the invasive group and 9% in the noninvasive group underwent revascularization procedures. Within 12 months, 78% in the invasive and 43% of the patients in the noninvasive groups had undergone revascularization. The details of treatments, followup, and outcome on these materials have previously been reported The 3269 analyzed plasma samples were taken at a median of 39 hours after onset of the last episode of chest pain (interquartile range, hours). Plasma IL-6 levels were distributed equally between the randomized patient groups (data not shown). There were no significant differences in other baseline characteristics between the groups. 16,17 Baseline characteristics according to IL-6 levels are summarized in TABLE 1. Of note, patients with increased IL-6 levels also were more likely to have increased troponin T and C-reactive protein levels but were only slightly more likely to have ST-segment depression. Outcome by Invasive and Noninvasive Cohorts Because there was no influence of longterm dalteparin in the comparisons between the invasive and noninvasive groups, analyses of these cohorts were performed disregarding medical assignment (Figure 1). For patients randomized to a noninvasive strategy, IL-6 levels 5 ng/l or higher at inclusion were associated with a 3.5-fold increase in probability of death at 12 months; 7.9% compared with 2.3% in patients with IL-6 levels less than 5 ng/l (P.001; TABLE 2). In patients with high IL-6 levels, an early invasive strategy led to a 5.1% absolute or 65% relative reduction in 12- month mortality (Table 2, FIGURE 2). At lower IL-6 levels, there was no significant difference in 12-month mortality between treatment strategies. For the composite end point of death or MI, elevated plasma IL-6 levels were not associated with any significantly 2001 American Medical Association. All rights reserved. (Reprinted) JAMA, November 7, 2001 Vol 286, No

4 larger event proportion in the noninvasive group (Table 2). Accordingly, an invasive strategy improved the end point outcome irrespective of IL-6 levels. Outcome in Noninvasive, Medically Randomized Cohorts In the noninvasive placebo-treated group, patients with IL-6 levels of 5 ng/l or higher had a 6-month mortality rate of 7.9% vs 2.5% in patients with levels lower than 5 ng/l (P=.001, Table 2). At elevated IL-6 levels, assignment to dalteparin tended to reduce the risk to 4.4% (P=.08; Table 2). In patients with Table 1. Baseline Characteristics by Interleukin 6 (IL-6) Levels* IL-6 lower than 5 ng/l, dalteparin treatment did not influence 6-month mortality. As for the combined end point, high plasma IL-6 levels were not significantly associated with risk (RR, 1.16; 95% CI, in the placebo group; Table 2). Assignment to dalteparin treatment did not reduce the composite of mortality or occurrence of MI at 6 months regardless of IL-6 levels. However, it significantly lowered the incidence of these events during the first 60 days of treatment in patients with elevated plasma IL-6 (P= at 30, Variable (n = 2372) (n = 897) P Value Age (range), y 67 (40-92) 68 (38-87).02 Sex, % Men Women Coronary risk factors Body mass index, median (range), kg/m (16-47) 26.4 (16-50).07 Hypertension 760 (32) 307 (34).24 Cholesterol mg/dl 1344 (56) 422 (47).001 Current smoker 625 (26) 241 (27).79 Diabetes mellitus 285 (12) 143 (16).004 Cardiac disease Previous myocardial infarction 668 (28) 222 (25).05 Previous PCA 101 (4) 25 (3).05 Previous CABG surgery 227 (10) 66 (7).05 Angina 48 h 1736 (73) 571 (64).001 Chest pain at rest 1918 (81) 750 (84).07 ST-segment depression at entry 1091 (46) 447 (50).04 Medication at admission Aspirin 1040 (44) 340 (38).002 -Blocker 878 (37) 292 (33).02 ACE inhibitor 325 (14) 121 (14).91 Calcium antagonist 484 (20) 181 (20).92 Long-acting nitrate 605 (26) 163 (18).001 Statins 313 (13) 77 (9).001 Diuretics 441 (19) 179 (20).40 Biochemical markers Troponin T 0.1 µg/l 1183 (50) 709 (80).001 C-reactive protein 10 mg/l 557 (24) 630 (74).001 Sample time, median (interquartile range), h 40 (27-56) 37 (27-52).001 Quartile (range), h 1( 27.1) 550 (24) 226 (26).34 2 ( ) 539 (24) 238 (28).02 3 ( ) 568 (25) 224 (26).77 4( 55.1) 599 (27) 174 (20).001 *Data are presented as number (percentage) unless otherwise indicated. PCI indicates percutaneous coronary intervention; CABG, coronary artery bypass graft; and ACE, angiotensin-converting enzyme. To convert cholesterol from mg/dl to mmol/l, multiply by 59. Blood samples for central troponin T analysis obtained at randomization in 2360 and 889 patients in the respective groups. Records of sample time were available in 3118 patients. 45, and 60 days), but not in those with lower levels (FIGURE 3). Multivariable Analysis The independence of IL-6 levels as a predictor of mortality was assessed by forward stepwise logistic regression analysis, for which established risk indicators as well as randomized treatments and interaction terms were evaluated. Only 7 covariates remained in the interventional part of the trial and 5 covariates in the medical part as independent predictors (TABLE 3). Plasma IL-6 levels remained significantly associated with increased mortality in both the interventional study (adjusted odds ratio [OR], 2.08; 95% CI, ; P =.006) and the medical trial (adjusted OR, 2.09; 95% CI, ; P=.002). COMMENT In this study, we found that plasma IL-6 level is an independent marker for identifying patients with unstable CAD with increased risk of death over 6 to 12 months. Interleukin 6 was predictive independent of other risk indicators, including the biochemical markers troponin T and C-reactive protein. Two studies of healthy adults have shown an association between elevated IL-6 levels and total and cardiovascular mortality 13 and future MI. 15 In a study much smaller than ours, Biasucci et al 5 observed that, in patients with CAD, increased levels of IL-6 were predictive of short-term coronary events. Based on these observations, as well as the documented prognostic values of C-reactive protein and fibrinogen, 2,7-13 IL-6 seems to settle well as a predictor of long-term mortality of CAD patients having experienced 1 or more episodes of instability. The FRISC II interventional trial was the first study to show that an early invasive treatment strategy reduced mortality and occurrence of MI in patients with unstable CAD. The greatest beneficial effect was seen in patients presenting with indicators of higher risk at entry, such as elevated troponin T levels or ST-segment depression on 2110 JAMA, November 7, 2001 Vol 286, No. 17 (Reprinted) 2001 American Medical Association. All rights reserved.

5 Table 2. Outcome in Relation to Interleuken 6 (IL-6) Levels* Death Death or Myocardial Infarction No. Affected (%)/ No. in Group Comparison of and No. Affected (%)/ No. in Group Comparison of and RR (95% CI) P Value RR (95% CI) P Value Interventional Study, 12-Month End Point Noninvasive 25 (7.9)/ (2.3)/ ( ) (16.7)/ (13.4)/ ( ).15 Invasive 9 (2.8)/ (1.9)/ ( ) (11.4)/ (9.8)/ ( ).42 Comparison of invasive vs noninvasive RR (95% CI) 0.35 ( ) 0.85 ( ) 0.69 ( ) 0.73 ( ) P value Medical Study, 6-Month End Point Placebo 24 (7.9)/ (2.5)/ ( ) (14.3)/ (12.4)/ ( ).40 Dalteparin 12 (4.4)/ (3.2)/ ( ) (11.8)/ (13.4)/ ( ).50 Comparison placebo vs dalteparin RR (95% CI) 0.55 ( ) 1.27 ( ) 0.82 ( ) 1.08 ( ) P value *RR indicates relative risk; CI, confidence interval. electrocardiography. 17,18 To predict which patients may benefit from an early invasive strategy, additional markers are needed that provide greater specificity. We now demonstrate that elevated IL-6 levels, independently of other well-known risk indicators, identify patients whose risk of death can be considerably reduced by an early invasive approach. It thus seems that an invasive strategy is of choice despite the increased inflammatory activity in these patients. Assignment to prolonged treatment with subcutaneous dalteparin also tended to reduce the risk of death for patients with IL-6 levels of 5 ng/l or higher, thus identifying those who would benefit from this kind of treatment. This is noteworthy since not all patients at high risk are eligible for invasive treatment, so medical alternatives such as long-term anticoagulant therapy might be considered. Elevated plasma IL-6 levels did not correlate with any increased risk of the composite end point of death or MI at 6 to 12 months. In the medical study, however, it identified patients for whom assignment to receive dalteparin had a beneficial effect during the first 60 days. This pattern was also seen in the main FRISC II medical study for patients with troponin T levels higher than 0.1 µg/l. 16 Figure 2. Mortality in the Interventional Study Related to Interleukin 6 (IL-6) levels Cumulative Probability of Death IL-6 <5 ng/l A Days From Start of Open-Label Dalteparin B IL-6 5 ng/l Noninvasive Invasive Twelve-month probability of death in the invasive and noninvasive cohorts categorized by IL-6 plasma levels. Vital status was unavailable for 1 patient who requested to be withdrawn from the follow-up. In the patients with low IL-6 levels, there were 16 events among the 826 patients at baseline in the invasive group and 19 events among the 383 patients in the noninvasive group at 12 months. In the patients with high IL-6 levels, there were 9 events among the 324 patients in the invasive group and 25 events among the 318 patients in the noninvasive group at 12 months. Therefore, an elevated IL-6 level might be useful for identification of a highrisk subgroup of patients who are protected by low-molecular-weight heparin treatment while waiting for invasive treatment. The fact that increased IL-6 levels were not predictive for the combined end point of death or MI is worthy of comment. Considering that the majority of deaths within a year from a severe unstable episode are likely due to cardiac causes, it seems that increased IL-6 levels indicate an inflammatory condition that may result in a higher risk of death from the index MI as well as from a subsequent MI. We found that patients presenting with elevated IL-6 levels were older and were characterized by a short angina history, no statin treatment on admission, low cholesterol levels at admission, myocardial 2001 American Medical Association. All rights reserved. (Reprinted) JAMA, November 7, 2001 Vol 286, No

6 Figure 3. Death or Myocardial Infarction (MI) Among Those in Medical Study Cumulative Probability of Death or Myocardial Infarction A IL-6 <5 ng/l Days From Start of Open-Label Dalteparin B IL-6 5 ng/l Placebo Dalteparin Six-month probability of death or MI in the medical study, categorized by interleukin 6 (IL-6) plasma levels in relation to randomized treatment. Follow-up data were unavailable for 14 patients in the dalteparin group and 18 patients in the placebo group. In patients with low IL-6 levels, there were 105 events among the 786 patients at baseline in the dalteparin group and 94 events among the 760 patients in the placebo group at 6 months. In patients with high IL-6 levels, there were 32 events among the 272 patients in the dalteparin group and 43 events among the 301 patients in the placebo group at 6 months. Table 3. Multivariable Analysis of Factors Influencing Mortality, Final Models* Interventional Study, Death at 12 mo (n = 2257) Medical Study, Death 6 at mo (n = 2119) Covariates Odds Ratio (95% CI) P Value Odds Ratio (95% CI) P Value Age 1.07 ( ) ( ).001 Diabetes mellitus 4.07 ( ) ( ).001 Previous MI 1.83 ( ) ( ).02 ST-segment depression 1.89 ( ) ACE inhibitors 3.15 ( ) ( ).02 Invasive strategy 0.53 ( ) ( ) ( ).002 *Models obtained by forward stepwise logistic regression analysis. Covariates evaluated include age, sex, diabetes mellitus, hypertension, smoking, previous myocardial infarction (MI), ST-segment depression at admission; use of statins, angiotensin-converting enzyme (ACE) inhibitors and -blockers; cholesterol levels higher than mg/dl (5.5 mmol/l), troponin T levels higher than 0.1 µg/l, C-reactive protein levels higher than 10 mg/l, and interleukin 6 (IL-6) levels higher than 5 ng/l at randomization; randomized treatment (invasive or noninvasive and dalteparin or placebo, respectively); and the interaction term IL-6 levels randomized treatment. CI indicates confidence interval; ellipses, not applicable because the interventional trial model contained 7 variables and the medical trial model entry contained 5 variables. damage as indicated by increased troponin T levels, and inflammatory activity reflected by elevated levels of C- reactive protein. These observations suggest that IL-6, which is present in the atheroma 19 and secreted by endothelial cells, smooth muscle cells, macrophages, and T cells, 14,19 may reflect a greater atherosclerotic burden as well as increased inflammatory activity in the plaques. These would subsequently be more vulnerable and prone to deeper fissuring, causing more severe thrombotic episodes and myocardial damage. The ischemic or necrotic myocardium could also be a source of cytokines We saw a moderate but significant correlation between levels of troponin T and IL-6 (Spearman r correlation coefficient=0.38, P.001), which may be explained by this reasoning. However, IL-6 levels as a predictor of death was additive to but independent of troponin T. Thus, our results further the understanding that unstable CAD is an inflammatory disorder. One may also regard the issue from the opposite point of view: IL-6 reflects an ongoing low-grade inflammation other than the atherosclerotic disease and from there contributes to the progression of CAD. In 2 recent review articles, collected evidence and indices for a central role of IL-6 in the development of coronary heart disease are presented, taking into account the pleiotropicity of this cytokine and its wide range of actions, including effects on platelets, endothelium, factors of metabolism, and coagulation. 23,24 The method of analysis that we used to measure plasma levels of IL-6 can be considered as rather insensitive. ELISA-based methods can detect levels of as little as 0.1 ng/l of IL-6, whereas we had a limit of detection already at 5 ng/l. Possibly, a more sensitive IL-6 assay method would have revealed more detailed prognostic information. We also relied on a single blood sample per patient, taken at varying times after the last episode of chest pain. This could act as a confounder given the relatively short half-life of IL-6 in plasma (4 hours). Still, our results very clearly showed a pronounced difference in mortality using 5 ng/l as a cutoff level and the effect of sample time in predicting high IL-6 levels was moderate (data not shown). Only the latest sample time quartile had a slightly lower frequency of elevated levels. In conclusion, we showed that circulating IL-6 is a strong independent marker of increased risk for mortality in patients with unstable CAD and that while patients with high plasma levels of IL-6 have the highest mortality rates, they also benefit most from a strategy of early revascularization. Prolonged treatment with subcutaneous dalteparin reduces the risk of death or MI during the first 60 days of treatment in patients with elevated levels of IL-6 and could be used while these patients await invasive treatment. Thus, circulating IL-6 provides important additional information when added to established indicators for risk stratification in patients with unstable CAD. Author Contributions: Dr Siegbahn, as the principal investigator of this project, had full access to all of the data and attests to the validity of the analysis. All coauthors had access to data. Study concept and design: Diderholm, Wallentin, Siegbahn JAMA, November 7, 2001 Vol 286, No. 17 (Reprinted) 2001 American Medical Association. All rights reserved.

7 Acquisition of data: Lindmark, Diderholm, Wallentin, Siegbahn. Analysis and interpretation of data: Lindmark, Wallentin, Siegbahn. Drafting of the manuscript: Lindmark, Siegbahn. Critical revision of the manuscript for important intellectual content: Lindmark, Diderholm, Wallentin, Siegbahn. Statistical expertise: Wallentin, Siegbahn. Obtained funding: Wallentin, Siegbahn. Administrative, technical, or material support: Lindmark, Diderholm, Wallentin, Siegbahn. Study supervision: Wallentin, Siegbahn. Funding/Support: The FRISC II study was supported by and performed in collaboration with the Pharmacia and Upjohn Co. The project organization within the research group was also supported by the Swedish Heart- Lung Foundation (Drs Wallentin and Siegbahn), by grant k gx from the Swedish Medical Research Council (Dr Siegbahn), and by the Åke Wiberg Foundation (Drs Lindmark and Siegbahn). Acknowledgment: We thank Birgitta Fahlström for excellent technical assistance and gratefully acknowledge the dedicated work of the research nurses, laboratory assistants, monitors, investigators, coordinators, end point committee, and data safety and monitoring board. REFERENCES 1. Ross R. Atherosclerosis: an inflammatory disease. N Engl J Med. 1999;340: Toss H, Lindahl B, Siegbahn A, Wallentin L, for the FRISC study group. Prognostic influence of increased fibrinogen and C-reactive protein levels in unstable coronary artery disease. Circulation. 1997;96: Berk BC, Weintraub WS, Alexander RW. Elevation of C-reactive protein in active coronary artery disease. Am J Cardiol. 1990;65: Biasucci LM, Vitelli A, Liuzzo G, et al. Elevated levels of interleukin-6 in unstable angina. Circulation. 1996;94: Biasucci LM, Liuzzo G, Fantuzzi G, et al. Increasing levels of interleukin (IL)-1Ra and IL-6 during the first 2 days of hospitalization in unstable angina are associated with increased risk of in-hospital coronary events. Circulation. 1999;99: Simon AD, Yazdani S, Wang W, Schwartz A, Rabbani LE. Circulating levels of IL-1,aprothrombotic cytokine, are elevated in unstable angina vs stable angina. J Thromb Thrombolysis. 2000;9: Verheggen PW, de Maat MP, Cats VM, et al. Inflammatory status as a main determinant of outcome in patients with unstable angina, independent of coagulation activation and endothelial cell function. Eur Heart J. 1999;20: Lindahl B, Toss H, Siegbahn A, Venge P, Wallentin L, for the FRISC study group. Markers of myocardial damage and inflammation in relation to longterm mortality in unstable coronary artery disease. N Engl J Med. 2000;343: Mendall MA, Strachan DP, Butland BK, et al. C- reactive protein: relation to total mortality, cardiovascular mortality and cardiovascular risk factors in men. Eur Heart J. 2000;21: Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men [published correction in N Engl J Med. 1997;337:356]. N Engl J Med. 1997;336: 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: Ridker PM, Hennekens CH, Buring JE, Rifai N. C- reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000;342: Harris TB, Ferrucci L, Tracy RP, et al. Associations of elevated interleukin-6 and C-reactive protein levels with mortality in the elderly. Am J Med. 1999;106: Barton BE. The biological effects of interleukin 6. Med Res Rev. 1996;16: Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation. 2000;101: FRISC II Investigators. Long-term low-molecularmass heparin in unstable coronary-artery disease: FRISC II prospective randomised multicentre study [published correction appears in Lancet. 1999;354:1478]. Lancet. 1999;354: FRISC II investigators. Invasive compared with noninvasive treatment in unstable coronary-artery disease: FRISC II prospective randomised multicentre study. Lancet. 1999;354: Wallentin L, Lagerqvist B, Husted S, Kontny F, Stahle E, Swahn E, for the FRISC II Investigators. Outcome at 1 year after an invasive compared with a noninvasive strategy in unstable coronary-artery disease. Lancet. 2000;356: Schieffer B, Schieffer E, Hilfiker-Kleiner D, et al. Expression of angiotensin II and interleukin 6 in human coronary atherosclerotic plaques: potential implications for inflammation and plaque instability. Circulation. 2000;101: Rader DJ. Inflammatory markers of coronary risk. N Engl J Med. 2000;343: Torre-Amione G, Kapadia S, Benedict C, Oral H, Young JB, Mann DL. Proinflammatory cytokine levels in patients with depressed left ventricular ejection fraction: a report from the Studies of Left Ventricular Dysfunction (SOLVD). J Am Coll Cardiol. 1996;27: Prabhu SD, Chandrasekar B, Murray DR, Freeman GL. -Adrenergic blockade in developing heart failure: effects on myocardial inflammatory cytokines, nitric oxide, and remodeling [published correction appears in Circulation. 2000;102:141]. Circulation. 2000;101: Woods A, Brull DJ, Humphries SE, Montgomery HE. Genetics of inflammation and risk of coronary artery disease: the central role of interleukin-6. Eur Heart J. 2000;21: Yudkin JS, Kumari M, Humphries SE, Mohamed- Ali V. Inflammation, obesity, stress and coronary heart disease: is interleukin-6 the link? Atherosclerosis. 2000; 148: American Medical Association. All rights reserved. (Reprinted) JAMA, November 7, 2001 Vol 286, No