SEARCH

SEARCH BY CITATION

Keywords:

  • acute coronary syndrome;
  • HMGB1 protein;
  • magnetic resonance imaging

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Abstract.  Andrassy M, Volz HC, Riedle N, Gitsioudis G, Seidel C, Laohachewin D, Zankl AR, Kaya Z, Bierhaus A, Giannitsis E, Katus HA, Korosoglou G (University of Heidelberg, Heidelberg, Germany). HMGB1 as a predictor of infarct transmurality and functional recovery in patients with myocardial infarction. J Intern Med 2011; 270: 245–253.

Objectives.  High-mobility group box 1 (HMGB1) protein is an innate danger signal for the initiation of host defence and tissue repair. The aim of this study was to analyse serum HMGB1 concentration and its correlation with infarct transmurality and functional recovery in patients with ST-elevation (STEMI) and non-ST-elevation myocardial infarction (NSTEMI).

Design.  We prospectively examined patients with first-time STEMI (= 46) or NSTEMI (= 49), treated according to current guidelines. Contrast-enhanced cardiac magnetic resonance imaging was performed 2–4 days after infarction for the estimation of infarct transmurality and was repeated after 6 months for the estimation of residual left ventricular function. HMGB1 was measured 2–4 days after infarction.

Results.  High-mobility group box 1 concentration was related to infarct size and to residual ejection fraction in patients with STEMI (r2 = 0.81 and r2 = 0.40, respectively, P < 0.001 for both) and NSTEMI (r2 = 0.74 and r2 = 0.25, respectively, P < 0.001 for both). Receiver operating characteristic (ROC) curve-derived cut-off values of 6.2 and 5.9 ng mL−1 for patients with STEMI and NSTEMI, respectively, were predictive of infarct transmurality greater than 75% (STEMI: area under the curve (AUC) = 0.93, standard error (SE) = 0.04, 95% confidence interval (CI) = 0.81–0.98; NSTEMI: AUC = 0.96, SE = 0.04, 95% CI = 0.86–0.99). HMGB1 cut-off values of 7.2 and 6.4 ng mL−1 for patients with STEMI and NSTEMI, respectively, were predictive of residual ejection fraction 6 months after myocardial infarction (MI) (STEMI: AUC = 0.81, SE = 0.07, 95% CI = 0.66–0.91; NSTEMI: AUC = 0.81, SE = 0.09, 95% CI = 0.68–0.91).

Conclusion.  High-mobility group box 1 serum levels represent a highly valuable surrogate marker for infarct transmurality and for the prediction of residual left ventricular function after MI.


Abbreviations:
CMR

cardiac magnetic resonance

ECG

electrocardiogram

LV

left ventricular

MI

myocardial infarction

NSTEMI

non-ST-elevation myocardial infarction

STEMI

ST-elevation myocardial infarction

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Since the introduction of percutaneous coronary interventions, the clinical outcomes in patients with ST-elevation (STEMI) and non-ST-elevation myocardial infarction (NSTEMI) have considerably improved [1, 2]. However, despite aggressive primary therapy, prognosis remains poor in patients with large infarctions and severe left ventricular (LV) dysfunction. There are a multitude of causes of chronic heart failure, but myocardial inflammation has emerged as one fundamental mechanism contributing to remodelling and disease progression in many different aetiologies of heart failure [3].

An influx of inflammatory cells into the infarct area is thought to be an essential component of the very early wound healing process. However, inflammation may persist beyond the initial repair phase and later extend into the noninfarcted remote myocardium, playing a role in long-term adverse cardiac remodelling [3–5]. Several studies have demonstrated the prognostic importance of inflammatory mediators in the setting of postinfarction remodelling and found that these mediators are expressed early in the development of heart failure compared to the classic neurohormones [6–8]. As the proinflammatory state is a universal predictor of clinical outcomes [9–11], methods that can provide objective assessment of disease activity, infarct size and clinical outcome early during the development of acute infarction may be of great potential clinical utility.

High-mobility group box 1 (HMGB1, formerly known as amphoterin) is a ubiquitous nuclear protein, constitutively expressed in quiescent cells. It is involved in several cellular functions within these cells, including determination of nucleosomal structure as well as stability and binding of transcription factors to DNA sequences [12]. Recently, we have identified HMGB1 as a critical mediator of inflammatory processes in the initiation of myocardial infarction (MI) and subsequent myocardial remodelling because of its active release by mononuclear cells and passive release from necrotic or damaged cells [13]. Increased levels of serum HMGB1 have been demonstrated in patients with acute coronary syndrome [14] and have been associated with impaired cardiopulmonary and echocardiographic findings [15]. However, the value of HMGB1, a surrogate marker of the activation of the innate immune system, for the estimation of infarct size and functional outcome has not been investigated to date.

The aim of this study was to investigate the ability of HMGB1 (i) to estimate infarct transmurality and (ii) to predict recovery of LV function in patients with STEMI and NSTEMI. All patients included in our study were treated with percutaneous angioplasty and stent placement and underwent cardiac magnetic resonance (CMR) imaging, which is a noninvasive clinical standard technique for the assessment of infarct size and LV function.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Patient population

We prospectively enrolled patients with first-time STEMI (= 46) and NSTEMI (= 49) who were admitted to the Chest Pain Unit of the University of Heidelberg because of first-time acute coronary syndrome. We included patients who had typical chest pain lasting >20 min within 24 h before the time-point of presentation and ST-elevation of >0.2 mV in at least two contiguous electrocardiogram (ECG) leads (STEMI), or elevated troponin T levels (>0.03 μg L−1) at presentation or at 4–8 h after presentation and without the presence of ST-elevation on ECG (NSTEMI). Exclusion criteria were ECG signs or history of previous infarction or a history of chronic inflammatory disease. All patients underwent cardiac catheterization and were treated with primary angioplasty and stent placement. All procedures complied with the Declaration of Helsinki and were approved by our local ethic committee, and all patients gave written informed consent.

Biochemical markers

Troponin T and creatine kinase (CK) levels were measured first on presentation and after 72–96 h to estimate the extent of myocardial damage [16]. Troponin T concentration was determined by a commercially available enzyme-linked immunosorbent assay (ELISA, Cardiac Reader; Roche, Mannheim, Germany), and CK activity was determined by photokinetic measurement (Advia® 2400 Chemistry System; Roche). Serum samples were analysed for HMGB1 expression 2–4 days after admission, using an ELISA (Shino-Test Corp., Kanagawa, Japan, distributed by IBL, Hamburg, Germany) according to the manufacturer’s instructions.

CMR imaging

Patients were examined in a clinical 1.5-T whole-body magnetic resonance imaging Achieva system (Philips Medical Systems, Best, the Netherlands) using a five-element cardiac phased-array receiver coil. A standardized protocol was followed 2–4 days after STEMI/NSTEMI, with the aim of (i) assessing baseline LV parameters (diameter, wall thickness and ejection fraction) using cine imaging and (ii) quantifying infarct size and transmurality using late-enhancement CMR imaging. A second CMR imaging examination was performed after 6 months of follow-up to assess residual LV ejection fraction.

Cine imaging

A steady-state free-precession sequence was used to obtain the cine images of the three long-axis and 7–9 short-axis views, 8 mm in thickness with a 2-mm interslice gap, to achieve full LV coverage from base to apex. Typical parameters were as follows: field of view (FOV) = 350 × 350 mm2; matrix size = 160 × 160; flip angle = 60°; repetition time/echo time (TR/TE) = 2.8/1.4 ms; and acquired voxel size = 2.2 × 2.2 × 8 mm3. The temporal resolution was 21–28 ms, and the total scan duration was 7–12 s. Planimetry of short-axis slices from the apex to the base was assessed using View Forum software (Philips Medical Systems) to determine end-systolic and end-diastolic volumes (mL) and LV ejection fraction (%).

Late-enhancement CMR imaging

For assessment of late enhancement, 10–15 min after a dose of 0.2 mmol gadolinium per kilogram body weight (Magnevist, Schering, Germany), a 3-dimensional sequence with inversion time (TI) scout was used to select the TI, which was typically 180–240 ms. Typical parameters for the late-enhancement sequence were as follows: FOV = 360 × 360 mm2; matrix size = 200 × 178, flip angle = 15°; TR/TE = 3.1/1.1 ms; acquired voxel size = 1.8 × 2.0 × 10 mm3; and total scan time = 13 s including one breath hold. For quantitative infarct size assessment, the View Forum software was used, and infarct size was defined as the area of hyperenhancement on short-axis views. Infarct size was determined visually and drawn manually by delineation of hyperenhanced versus normally saturated dark myocardium. Regions of microvascular obstruction (MVO) were defined as subendocardial hypoenhanced areas within a transmural hyperenhanced myocardial scar and were described as infarcted myocardium. Volume integration of hyperenhanced areas on short-axis slices was used, and infarct size was calculated as the amount of hyperenhanced myocardium related to the total LV mass. Furthermore, infarct transmurality was assessed semiquantitatively based on a six-grade scale (0 = no hyperenhancement, 1 = 1–25%, 2 = 26–50%, 3 = 51–75%, 4 = 76–100% transmurality without MVO and 5 = 76–100% transmurality with MVO), and the presence of infarct transmurality ≥75% (i.e. grades 4 and 5) was considered predictive of functional recovery at 6 months [17].

Definition of study end-points

Infarct transmurality ≥75% and follow-up ejection fraction <55% were chosen as the primary end-points of our study because both are important parameters that predict mortality in patients with acute coronary syndromes. Furthermore, clinical end-points including the incidence of cardiac death, nonfatal recurrent MI and new revascularization were assessed within the 6-month follow-up period in all patients.

Statistical analysis

Statistical analysis was performed using commercially available software (MedCalc8.2; MedCalc software, Mariakerke, Belgium), and data are presented as mean ± standard deviation. Receiver operating characteristic (ROC) curves were used to determine the diagnostic value of HMGB1 for the prediction of infarct size and follow-up ejection fraction. Cut-off values were selected for each parameter, to provide an optimal balance between sensitivity and specificity, and pairwise comparison of area under the curve (AUC) of the ROC curves was made [18]. Differences were considered statistically significant at P < 0.05.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Baseline characteristics

Demographic characteristics of patients with STEMI (= 46) or NSTEMI (= 49) are shown in Table 1. A trend towards older age was seen in patients with NSTEMI (P = 0.06), whereas coronary risk factors were similarly distributed between the two groups. Patients with STEMI had a 2- to 3-fold larger infarct size compared to patients with NSTEMI, as determined by both cardiac necrosis marker (peak cardiac troponin T level 4.7 ± 6.8 vs. 1.5 ± 2.3 μg L−1 for STEMI and NSTEMI, respectively, P < 0.001) and late enhancement on CMR images (15.7 ± 9.9% vs. 7.2 ± 6.0% for STEMI and NSTEMI, respectively, P < 0.001).

Table 1.   Baseline characteristics
ParametersSTEMI (= 46)NSTEMI (= 49) P-value
  1. Data presented as number of patients and percentages or as mean ± standard deviation.

  2. PCI, percutaneous coronary intervention; MVO, microvascular obstruction; NSTEMI, non-ST-elevation myocardial infarction; STEMI, ST-elevation myocardial infarction.

Demographics
 Age (years)57 ± 1262 ± 120.062
 Male gender33 (72%)36 (73%)0.852
Coronary risk factors
 Arterial hypertension25 (55%)35 (71%)0.086
 Hypercholesterolaemia21 (47%)28 (57%)0.315
 Diabetes mellitus12 (26%)15 (31%)0.884
 Family history20 (44%)23 (47%)0.738
 Smoker32 (70%)30 (61%)0.399
Biochemical markers
 Peak troponin T (μg L−1) within 96 h4.66 ± 6.761.46 ± 2.33<0.0001
 Peak creatine kinase (U L−1) within 96 h1892 ± 1812505 ± 685<0.0001
 Creatinine on admission (mg dL−1)0.80 ± 0.160.85 ± 0.190.09
Coronary angiography
 Delay from symptom-onset to PCI (min)738 ± 4781037 ± 5250.004
 Number of stents1.53 ± 0.841.70 ± 0.900.5
 Anterior wall infarction (%)41.338.80.8
Baseline MR parameters
 Delay from symptom-onset to MRI (days)4.2 ± 1.24.1 ± 1.60.687
 Left ventricular ejection fraction (%)54.3 ± 10.060.9 ± 8.00.002
 Late enhancement (%)15.7 ± 9.87.2 ± 6.00.0001
 Infarct transmurality (0 = 0% to 5 = 100%+MVO)3.59 ± 1.032.39 ± 1.23<0.0001
 Microvascular obstruction11 (24%)1 (2%)0.001
 End-diastolic volume (mL)169 ± 38158 ± 360.163
 End-systolic volume (mL)80 ± 2969 ± 430.119
 Stroke volume (mL)92 ± 2294 ± 190.666
 Cardiac output (l min−1)6.3 ± 1.66.1 ± 1.40.540

Relation between HMGB1 and infarct size

High-mobility group box 1 expression was related to infarct size both in patients with STEMI and in those with NSTEMI (r2 = 0.81 vs. r2 = 0.74, respectively, P < 0.001, Fig. 1a). HMGB1 concentrations above the ROC curve-derived cut-off values of 6.2 and 5.9 ng mL−1 for STEMI and NSTEMI, respectively, were highly predictive of infarct transmurality ≥75% (STEMI: AUC = 0.93, standard error (SE) = 0.04, 95% confidence interval (CI) = 0.85–0.98; NSTEMI: AUC = 0.96, SE = 0.04, 95% CI = 0.86–0.99; P < 0.001, Fig. 1b–d).

image

Figure 1.  High-mobility group box 1 (HMGB1) expression was related to infarct size in patients with ST-elevation myocardial infarction (STEMI) or non-ST-elevation myocardial infarction (NSTEMI) (a). Cut-off values for HMGB1 of 6.2 and 5.9 ng mL−1 for patients with STEMI and NSTEMI, respectively, were highly predictive of infarct transmurality ≥75% (b–d).

Download figure to PowerPoint

Prediction of functional recovery

Systemic HMGB1 expression was inversely related to residual ejection fraction in patients with STEMI and NSTEMI (r2 = 0.40 and r2 = 0.25, respectively, P < 0.001 for both, Fig. 2a). HMGB1 concentrations above the ROC curve-derived cut-off values of 7.2 and 6.4 ng mL−1 for STEMI and NSTEMI, respectively, were predictive of residual ejection fraction <55% at 6 months of follow-up (STEMI: AUC = 0.81, SE = 0.07, 95% CI = 0.66–0.91; NSTEMI: AUC = 0.81, SE = 0.09, 95% CI = 0.68–0.91; Fig. 2b–d). It is interesting that the predictive value of HMGB1 expression for the estimation of residual ejection was similar to that provided by infarct transmurality ≥75% in patients with STEMI and NSTEMI (Fig. 2b–d, P = NS for both by pairwise comparison of AUCs). Diagnostic characteristics of HMGB1 for the prediction of infarct transmurality and follow-up ejection fraction are summarized in Table 2.

image

Figure 2. High-mobility group box 1 (HMGB1) expression was related to residual ejection fraction in patients with ST-elevation myocardial infarction (STEMI) and non-ST-elevation myocardial infarction (NSTEMI) (a). Cut-off values for HMGB1 of 7.2 and 6.4 ng mL−1 for patients with STEMI and NSTEMI, respectively, were predictive of residual ejection fraction <55% (b–d). Of interest, the predictive value of HMGB1 expression for the estimation of residual ejection was similar to that provided by infarct transmurality ≥75% in patients with STEMI and NSTEMI (P = NS for both by pairwise comparison of area under the curves).

Download figure to PowerPoint

Table 2.   Prediction by HMGB1 level of infarct transmurality ≥75% and follow-up ejection fraction <55% in patients with STEMI and NSTEMI. For comparison, the values of infarct transmurality 75% for the prediction of follow-up ejection fraction are provided
  Cut-off valuesSensitivity (%)Specificity (%)PPV (%)NPV (%)AUC
  1. Data are presented as percentages.

  2. PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve; HMGB1, High-mobility group box 1; NSTEMI, non-ST-elevation myocardial infarction; STEMI, ST-elevation myocardial infarction.

Prediction of infarct transmurality ≥75% by HMGB1
 STEMI and NSTEMI6.2859285920.94
 STEMI6.2918283900.93
 NSTEMI5.9739789930.96
Prediction of follow-up ejection fraction <55% by HMGB1
 STEMI and NSTEMI6.4827187630.84
 STEMI7.2935691630.81
 NSTEMI6.4767983710.81
Prediction of follow-up ejection fraction <55% by infarct transmurality
 STEMI and NSTEMITransmurality ≥75%817688640.88
 STEMI728092550.84
 NSTEMI876792550.88

Subsection analysis in patients with type 2 diabetes mellitus

Patients with type 2 diabetes mellitus (T2DM) (= 27 of 95) had significantly higher HMGB1 levels (9.0 ± 4.4 vs. 4.4 ± 3.6% ng mL−1, P < 0.001) as well as increased infarct sizes (16.6 ± 9.7 vs. 9.1 ± 7.9%, P = 0.001) compared to those without T2DM. Logistic regression analysis identified HMGB1 levels and infarct size as strong predictors of LV functional recovery at 6 months, whereas a trend was observed for T2DM (Table 3).

Table 3.   Variables for the prediction of residual ejection fraction <55% using logistic regression analysis
VariablesHazard ratio95% CITotal χ2 P-value
  1. Hazard ratio indicates the relative risk with the corresponding 95% confidence interval (CI).

  2. HMGB1, High-mobility group box 1.

Age(years)1.00.9–1.11.40.23
Male gender0.40.4–1.32.50.13
T2DM0.40.2–1.03.50.06
Infarct size determined by late gadolinium enhancement (%)0.80.8–0.935.3<0.001
HMGB1 level (ng mL−1)0.70.6–0.829.6<0.001

Clinical outcomes

There were no cardiac deaths during the follow-up period. Two new MIs were observed in patients with STEMI, but none in patients with NSTEMI. Furthermore, both groups showed similar rates of recurrent revascularizations (18% vs. 21% for STEMI and NSTEMI, respectively).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

Recent studies have identified an association between ongoing postinfarction inflammatory response with consecutive LV remodelling and adverse outcome [19–21]. Thus, although inflammatory responses early after infarction promote myocardial healing processes, excessive and persistent inflammation after infarction has been associated with the development of LV aneurysm, cardiac rupture or chronic LV dilatation [20, 22, 23]. However, the mechanisms that initiate and control these processes are not fully understood.

Various markers of vascular inflammation and novel metabolic risk factors have been identified as predictors of adverse clinical outcome in patients with chronic heart failure [24, 25]. The present study was designed to investigate the correlation between systemic concentration of HMGB1, which has been identified as a critical mediator of inflammatory processes and cardiac remodelling after MI [13], and clinical and functional outcome in patients with MI. As HMGB1 is released passively from necrotic cells and actively secreted by inflammatory cells, thus being a major contributor of the inflammatory response triggered by ischaemia/reperfusion injury [13, 26], we anticipated that serum HMGB1 concentration could be predictive of infarction size and cardiac recovery following MI.

In the present study, myocardial viability and regional function were assessed using CMR imaging as a versatile noninvasive clinical tool [27–29]. However, despite the high reproducibility of CMR analyses, this technique is associated with high costs and limited availability. Therefore, a simple biochemical marker that could be used as a surrogate measure of cardiac integrity may be preferable.

Kohno et al. [21] recently reported an association between elevated HMGB1 levels and adverse clinical outcomes such as congestive heart failure, myocardial rupture and in-hospital death in patients with STEMI. Furthermore, Giallauria et al. [30] showed a close association in postinfarction patients between HMGB1 levels and autonomic dysfunction, which is regarded as a powerful predictor of mortality in those with coronary artery disease. Here, we demonstrated for the first time that elevated HMGB1 levels are closely related to infarct size in patients with acute MI. Of importance, HMGB1 offered similar diagnostic value for the prediction of functional recovery after MI as infarct transmurality, which is a well-established marker of myocardial viability in this setting.

However, the pathophysiological role of HMGB1 in MI is not fully understood. We have previously shown that systemic administration of HMGB1 causes increased inflammatory responses, impairs cardiac function and leads to adverse LV remodelling following MI [13]. Furthermore, we identified HMGB1 as the driving force in the development of diabetic cardiomyopathy in a postinfarction model of type 1 diabetes which could be partially reversed by HMGB1-specific antagonism [26, 31]. In the same context, HMGB1 was recently identified as a novel myocardial depressant that contributed to a sustained and excessive proinflammatory response in vitro [32] and as an innate immune mediator in acute allograft rejection in a murine cardiac transplantation model [33].

In addition, several lines of evidence have suggested that HMGB1 might also play a role in restoration of cardiac function after MI, probably by promoting stem cell recruitment and/or stimulating angiogenesis [34, 35]. Therefore, local administration of exogenous HMGB1 in the peri-infarcted LV might have therapeutic potential as it led to the recruitment of progenitor cells thus attenuating LV remodelling in a permanent MI model [34, 35]. Conflicting results regarding the role of HMGB1 in ischaemic heart disease might be because of diverse experimental settings and the use of different doses of recombinant HMGB1, as dose has been shown to be critical in the determination of HMGB1-mediated effects [36].

Xu et al. [37] extended the generally accepted notion that an alarmin is only actively secreted by immune cells and passively released by necrotic cells to propose the active release of the alarmin HMGB1 by ‘stressed’ but still viable cardiomyocytes, leading to reduced cardiac function. This novel concept of the function of HMGB1 expression implies an immunomodulatory role by nonimmune cells such as cardiomyocytes acting in a paracrine manner on receptors such as the receptor for advanced glycation end products and the Toll-like receptor 4 [38] expressed on adjacent myocytes, to decrease cardiac contractility. Thus, with regard to the postinfarction inflammatory response, HMGB1 might have bidirectional effects on LV remodelling depending on the area, extent and timing of HMGB1 modulation [39]. In particular, HMGB1 secreted from chronically infarcted, necrotic myocardium may lead to only a moderate local increase in HMGB1 concentration promoting stem cell recruitment and wound healing [35], whereas in the acute ischaemia/reperfusion injury setting, recruitment of inflammatory cells might enhance HMGB1 levels with consequent detrimental effects [39].

In conclusion, the results presented here further emphasize the pathophysiological importance of sustained expression of proinflammatory mediators in preclinical and clinical heart failure models in which increased HMGB1 serum levels are strongly related to infarction size and to residual LV function in patients with STEMI and NSTEMI. Thus, HMGB1 may represent a useful new marker for risk stratification and for tailoring therapeutic strategies in patients early after acute MI. Furthermore, HMGB1 can provide better monitoring of disease activity and may lead to a more stringent control of several potential anti-inflammatory therapeutic approaches. Ultimately, HMGB1 may contribute to the pathophysiological understanding of adverse remodelling leading to heart failure by unmasking different proinflammatory pathways that may be involved in ischaemic heart disease.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References

This study was supported by the Carl Baresel Stiftung (MA), the German Research Foundation (MA, AN 403/2-1), the German Heart Foundation (MA, F/36/08) and the European Foundation for the Study of Diabetes (MA).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conflict of interest statement
  8. Acknowledgements
  9. References