Author + information
- Received March 18, 2005
- Revision received June 15, 2005
- Accepted August 1, 2005
- Published online December 6, 2005.
- Thor Ueland, PhD⁎,§,⁎ (, )
- John Kjekshus, MD, PhD†,
- Stig S. Frøland, MD, PhD‡,§,
- Torbjørn Omland, MD, PhD∥,
- Iain B. Squire, MD#,
- Lars Gullestad, MD, PhD†,
- Kenneth Dickstein, MD, PhD¶ and
- Pål Aukrust, MD, PhD‡,§
- ↵⁎Reprint requests and correspondence:
Dr. Thor Ueland, Section of Endocrinology, Medical Department, National University Hospital, N-0027 Oslo, Norway
Objectives We sought to determine the relationship between circulating cytokine levels and clinical outcomes in patients with heart failure (HF) following acute myocardial infarction (AMI).
Background Persistent inflammation plays a role in the development of HF, and various inflammatory cytokines predict cardiovascular events in acute coronary syndromes.
Methods We measured plasma levels of interleukin (IL)-6, monocyte chemotractant protein 1, IL-10, and soluble tumor necrosis factor receptor type 1 (sTNFR1) during longitudinal testing over a period of two years in 234 patients with HF following AMI recruited for participation in the OPTIMAAL trial, focusing on the possible prognostic value of circulating cytokine levels in these patients.
Results Measurement of sTNFR1 at baseline predicted all-cause mortality and cardiovascular death in patients with post-MI HF after adjustment for other biomarkers that have been shown to give prognostic information in HF patients, such as N-terminal B-type natriuretic peptide.
Conclusions Assessment of sTNFR1 levels might provide important prognostic information in patients who develop HF during the acute phase following AMI.
Persistent inflammation plays a pathogenic role in heart failure (HF) (1,2). In particular, tumor necrosis factor (TNF)-alpha has been implicated as a mediator in this process. Although several studies have focused on the possible role of TNF-alpha, the individual cytokine exists as part of a cytokine network, and the effect of various inflammatory cytokines is influenced by other inflammatory as well as anti-inflammatory mediators, illustrating the complexity of this network (3). Thus, we and others have suggested a pathogenic role for interleukin (IL)-6 and chemokines such as monocyte chemoattractant peptide (MCP-1) in the development of myocardial failure (3). Moreover, anti-inflammatory cytokines such as IL-10 could potentially have beneficial effects on cardiac remodeling (1,2).
Several studies have examined cytokine levels in HF during cross-sectional testing, but few have reported these mediators during longitudinal follow-up, and the ability of these parameters to predict forthcoming clinical events is far from clear. To further elucidate the possible role of cytokines in HF, we therefore measured plasma levels of IL-6, MCP-1, IL-10, and soluble TNF receptor type 1 (sTNFR1) as stable markers of activity in the TNF system (4), during longitudinal testing in patients with evidence of HF or left ventricular (LV) dysfunction following acute myocardial infarction (AMI), particularly focusing on the possible prognostic value of these cytokines.
The design and main results of the OPtimal Trial In Myocardial infarction with Angiotensin II Antagonist Losartan (OPTIMAAL) have previously been reported in detail (5). Briefly, 5,477 patients with AMI complicated with HF during the acute phase were randomly assigned and titrated to a target dose of losartan (50 mg daily) or captopril (50 mg three times daily) as tolerated, and the patients were followed for a median 2.7 years for mortality and morbidity. The present study was a prospectively designed sub-study of the main OPTIMAAL trial comprising 234 consecutive patients from six centers that was designed to analyze plasma/serum levels of inflammatory/anti-inflammatory mediators (6). For comparison, cytokine levels were also measured in 20 age- and gender-matched healthy controls.
Blood samples were obtained at baseline and after one month, one year , and two years as previously described (6). Plasma concentrations of IL-6, MCP-1, IL-10, and sTNFR1 were quantified by enzyme immunoassays obtained from R&D Systems (Minneapolis, Minnesota). N-terminal B-type natriuretic peptide (N-BNP) and high-sensitivity C-reactive protein (hsCRP) were analyzed as previously described (6).
Differences between groups were compared by the Mann-Whitney Urank-sum test. Within groups, differences were analyzed by the Wilcoxon rank-sum test. Relationships between variables were analyzed by the Spearman rank test. Associations between cytokine concentrations and cardiovascular events (morbidity and mortality) were analyzed by univariate analysis a priori and then by forced or forward conditional multivariate Cox regression adjusting for predefined conventional predictors (smoking, previous MI, diabetes, medication, age, gender, creatinine clearance, NYHA functional class, N-BNP, and hsCRP). Cardiovascular events (end points) were cardiovascular death, total mortality, re-infarction, and angina. Kaplan-Meier curves were generated, and the log rank test was used to compare event rates in relation to cytokine levels when appropriate. Probability values are two sided, with p < 0.05 being considered statistically significant.
Circulating cytokine levels in post-MI HF patients
Baseline levels of IL-6, IL-10, MCP-1, and sTNFR1 were all higher in our patients compared with healthy controls (Fig. 1).Although these mediators markedly decreased during follow-up, circulating levels remained significantly increased compared with controls throughout the study. There was no significant difference in cytokines levels between the two treatment groups, and the cytokine levels are presented as pooled data from both groups (Fig. 1).
Associations between cytokine levels and baseline clinical and biochemical variables
Several associations between baseline variables (Table 1)and cytokine levels were revealed. First, high sTNF-R1 (p < 0.01) and IL-6 (p < 0.05) levels were observed in patients with clinical signs of HF (NYHA functional class >II). Second, patients taking diuretics had significantly higher levels of all cytokines, patients taking beta-blockers had lower levels of IL-10 (p < 0.05) and sTNFR1 (p < 0.001), and low sTNFR1 levels were also seen in subjects using statins (p < 0.001) or aspirin (p < 0.05). Third, a low creatinine clearance was associated with high sTNFR1 levels (p < 0.001) and both sTNFR1 (r = 0.35; p < 0.001) and IL-6 (r = 0.29; p < 0.001) correlated with age. Fourth, patients with a history of MI had lower IL-10 levels (p < 0.01). Finally, significant positive relationships were found between N-BNP, hsCRP, and all cytokines (except MCP-1 for hsCRP) (Fig. 2).In contrast to these correlations, there were no significant relationships between cytokine levels and LV ejection fraction.
Association between cytokine levels and clinical events
All patients, regardless of treatment (losartan or captopril), were divided into two groups based on the median cytokine level at baseline and at one month. Figure 3shows the unadjusted risk estimates in these groups in relation to clinical events. All-cause mortality was associated with high sTNFR1 levels at baseline (>1,981 pg/ml, risk ratio [RR] 7.1 [95% confidence interval (CI) 2.6 to 19.3]) and after one month (>1,387 pg/ml, RR 6.2 [95% CI 1.7 to 21.8]) and high IL-6 levels at one month (>3.0 pg/ml, RR 4.4 [95% CI 1.4 to 13.8]). Similarly, subjects dying from cause-specific cardiovascular events had high sTNFR1 levels at baseline (RR 5.9 [95% CI 2.1 to 16.0]) and one month (RR 4.4 [95% CI 1.2 to 16.2]). Recurrent nonfatal AMI during follow-up was associated with high IL-10 levels at one month (>0.35 pg/ml, RR 2.7 [95% CI 1.2 to 6.0]). Finally, the composite end point was associated with high levels of sTNFR1 at baseline (RR 2.9 [95% CI 1.6 to 5.3]) and one month (RR 2.5 [95% CI 1.3 to 5.1]) and high IL-10 levels at one month (RR 2.0 [95% CI 1.0 to 3.9]).
In a multivariate model, adjusting for potential confounders (see Statistical methods), baseline, but not one-month, sTNFR1 levels were still associated with long-term all-cause mortality and cardiovascular death (Table 2).N-terminal B-type natriuretic peptide and hsCRP were also included in the multivariate model, but in contrast to sTNFR1 survival curves, the model showed no significant relationship with total mortality for baseline levels of either hsCRP or N-BNP (Fig. 4).Interleukin-10 levels measured at one month remained a significant predictor of recurrent nonfatal AMI both in the forced (p = 0.006; hazard ratio 3.3 [95% CI 1.4 to 7.5]) and stepwise (p = 0.013; hazard ratio 2.7 [95% CI 1.2 to 5.8]) model. Similar results were obtained for all calculations when patients with diabetes were excluded.
sTNFR1 levels during longitudinal follow-up in relation to cardiovascular events
When comparing sTNFR1 in survivors with non-survivors, we found that the latter group had persistently higher levels at all time points during follow-up (Fig. 5).Furthermore, the prognostic value of plasma sTNFR1 measurements in relation to all-cause mortality remained significant at both one year (p = 0.004; RR 2.1 [95% CI 1.8 to 2.5]) and two years (p = 0.044; RR 2.0 [95% CI 1.7 to 2.4]).
We have shown that high plasma levels of sTNFR1 during the acute phase are associated with long-term all-cause mortality and cardiovascular death in HF following AMI. Previously, Rauchhaus et al. (7) reported that sTNF-R1 emerged among all cytokine parameters as the strongest and most accurate prognostic marker in chronic HF. Recently, Valgimigli et al. (8) reported that among several inflammatory markers, sTNF-R1 remained the only independent predictor of death and HF after MI. The present study confirms and extends these findings, evaluating the prognostic significance of raised plasma sTNF-R1 in HF both in the acute phase and during follow-up. In fact, we show that non-survivors have raised plasma levels of sTNFR1 compared with survivors not only at baseline but at all time points during the observation period. Importantly, we show that sTNF-R1 has predictive power independent of other well-established risk markers such as N-BNP and hsCRP (9), possibly reflecting that we are focusing on a high-risk subgroup of AMI patients that develop HF during the acute phase.
The lack of a relationship between sTNF-R1 and re-infarction in our study may be somewhat unexpected. However, TNF-alpha and sTNF-Rs may be released from the failing myocardium, not only from inflamed tissue, and it is possible that the high baseline levels of sTNFR1 in these patients with post-MI HF during the acute phase may reflect the degree of myocardial damage and HF rather than the degree of inflammation. Moreover, the marked decline in cytokine levels after one month could reflect a more steady state when the “cytokine storm” has settled. However, although there was an initial fall in sTNFR1 levels, these post-MI HF patients had markedly raised sTNFR1 levels compared with controls throughout the study, reflecting an ongoing inflammation in these patients.
In the present study we show that plasma levels of sTNFR1 during the acute phase represent a strong and independent predictor of all-cause and cardiovascular mortality in post-MI HF patients. Although relatively few patients were studied and although we had no placebo group, our findings suggest that assessment of sTNFR1 during hospitalization may be useful in identifying patients who are at high risk of death following complicated MI.
Supported by Merck, Sharp, and Dohme Research Laboratories. Dr. Squire has received honoraria from Merck, Sharp, and Dohme Research Laboratories for speaking at symposia and acting as an adviser.
- Abbreviations and Acronyms
- acute myocardial infarction
- heart failure
- high-sensitivity C-reactive protein
- left ventricular
- monocyte chemoattractant peptide
- N-terminal B-type natriuretic peptide
- soluble tumor necrosis factor receptor type 1
- Received March 18, 2005.
- Revision received June 15, 2005.
- Accepted August 1, 2005.
- American College of Cardiology Foundation
- Mann D.L.
- Dibbs Z.,
- Thornby J.,
- White B.G.,
- Mann D.L.
- Dickstein K.,
- Kjekshus J.
- Ueland T.,
- Jemtland R.,
- Godang K.,
- et al.
- Rauchhaus M.,
- Doehner W.,
- Francis D.P.,
- et al.
- Valgimigli M.,
- Ceconi C.,
- Malagutti P.,
- et al.
- James S.K.,
- Lindahl B.,
- Siegbahn A.,
- et al.