- Email: [email protected]
Plasma high-mobility group box 1 levels and prediction of outcome in patients with traumatic brain injury
Clinica Chimica Acta 413 (2012) 1737–1741
Contents lists available at SciVerse ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Plasma high-mobility group box 1 levels and prediction of outcome in patients with traumatic brain injury Ke-Yi Wang c, Guo-Feng Yu b, Zu-Yong Zhang a, Qiang Huang b, Xiao-Qiao Dong a,⁎ a b c
Department of Neurosurgery, The First Hangzhou Municipal People's Hospital, Nanjing Medical University, 261 Huansha Road, Hangzhou 310006, China Department of Neurosurgery, Quzhou People's Hospital, 2 Zhongloudi Road, Quzhou 324100, China Central Laboratory, The First Hangzhou Municipal People's Hospital, Nanjing Medical University, 261 Huansha Road, Hangzhou 310006, China
a r t i c l e
i n f o
Article history: Received 19 May 2012 Received in revised form 29 June 2012 Accepted 2 July 2012 Available online 10 July 2012 Keywords: High-mobility group box 1 Traumatic brain injury Functional outcome Mortality
a b s t r a c t Background: High-mobility group box 1 (HMGB1), a marker of inflammation, has been associated with poor outcome of critical illness. The present study was undertaken to investigate the plasma HMGB1 concentrations in patients with traumatic brain injury (TBI) and to analyze the correlation of HMGB1 with TBI outcome. Methods: We performed an observational, clinical study. Plasma HMGB1 concentration of 106 healthy subjects and 106 patients with severe TBI was measured by ELISA. The correlation with 1-y mortality and unfavorable outcome (Glasgow Outcome Scale score of 1–3) was analyzed. Results: Thirty-one patients (29.2%) died and 48 patients (45.3%) had an unfavorable outcome at 1 y after TBI. Plasma HMGB1 level was substantially higher in patients than in healthy controls. A multivariate analysis selected plasma HMGB1 level as an independent predictor for 1-y mortality and unfavorable outcome of patients. A receiver operating characteristic curve analysis showed plasma HMGB1 level statistically significantly predicted 1-y mortality and unfavorable outcome. The prognostic value of HMGB1 was similar to that of Glasgow Coma Scale score for 1-y clinical outcomes. Conclusions: Plasma HMGB1 concentration emerges as a novel biomarker for predicting 1-y clinical outcomes of TBI. © 2012 Elsevier B.V. All rights reserved.
1. Introduction High-mobility group box 1 (HMGB1), a nonhistone DNA binding protein, has been implicated in the stabilization of nucleosomal structure and the facilitation of gene transcription [1,2], and further identified as a cytokine-like mediator of delayed endotoxin lethality and acute lung injury [3,4]. HMGB1, actively secreted by immune cells or released by necrotic cells into the extracellular milieu, might be involved in the triggering of inflammation [5,6]. Recombinant HMGB1 can induce acute inflammation in animal models of lung injury and endotoxemia [3,4], and anti-HMGB1 antibody can attenuate endotoxininduced lethality [7]. In addition, high serum concentrations of HMGB1 in patients with sepsis or hemorrhagic shock have been reported to be associated with increased mortality and disease severity [3,8]. HMGB1 is widely expressed in the brain [9–11]. Moreover, in the brain, HMGB1 has been reported to be released after cytokine stimulation and to be involved in the inflammatory process after it was administered into the cerebral ventricles [12,13]. HMGB1 can be released from damaged neurons in mouse models with middle cerebral artery occlusion [14]. Moreover, HMGB1 is increased in the cerebrospinal fluid from meningitis patients [15] and in the serum of ⁎ Corresponding author. Tel.: +86 571 87065701. E-mail address: [email protected] (X.-Q. Dong). 0009-8981/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2012.07.002
cerebral ischemia patients [16,17]. Recent data have identified HMGB1 in the cerebrospinal fluid of subarachnoid hemorrhage [18,19] and in the serum of intracerebral hemorrhage [20] as a potential biomarker of neurological outcome, suggesting HMGB1 may represent a marker of neurological injury. Thus, we furthermore investigated the change of HMGB1 in the peripheral blood and its association with disease outcome in traumatic brain injury. 2. Subjects and methods 2.1. Study population The inclusion period was from January 2008 to April 2010. Inclusion criteria were isolated head trauma and post resuscitation Glasgow Coma Scale (GCS) score of ≤ 8. Exclusion criteria were b 18 y, admission time > 6 h, previous head trauma, neurological disease including ischemic or hemorrhagic stroke, use of antiplatelet or anticoagulant medication, and presence of other prior systemic diseases including uremia, liver cirrhosis, malignancy, chronic heart or lung disease, diabetes mellitus and hypertension. Subjects were initially evaluated if they presented to our hospital and had blood collected and brain magnetic resonance imaging completed as parts of medical examination. Finally, a control group consisted of 106 healthy ageand sex-matched subjects with normal results on brain magnetic
1738
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
resonance imaging and without vascular risk factors. Informed consent to participate in the study was obtained from them or their relatives. This protocol was approved by the Ethics Committee of Quzhou People's Hospital before implementation. 2.2. Clinical and radiological assessment At admission, clinical severity was assessed using initial post resuscitation GCS score. Shock was defined as systolic blood pressure b90 mm Hg, and hypoxia was defined as blood oxygen saturation b85% [21]. Hyperglycemia was defined as blood glucose >11.1 mmol/l [22]. Hypoglycemia was defined as blood glucose b2.2 mmol/l [23]. Neurology deterioration was defined as occurring in patients who manifested clinically identified episodes of ≥1 of the following: 1) a spontaneous decrease in GCS motor scores of ≥2 points from the previous examination; 2) a further loss of pupillary reactivity; 3) development of pupillary asymmetry greater than 1 mm; or 4) deterioration in neurological status sufficient to warrant immediate medical or surgical intervention [21]. All computerized tomography (CT) scans were performed according to the neuroradiology department protocol. Investigators who read them were blinded to clinical information. CT classification was performed using Traumatic Coma Data Bank criteria on initial post resuscitation CT scan according to Marshall et al. [24]. 2.3. Patient management The treatments included surgical therapy, ventilatory support, arterial pressure maintenance, glycemic control, intravenous fluids, hyperosmolar agents, H2 blockers, early nutritional support, and physical therapy. The decision to intubate and use mechanical ventilation was based on the individuals' concentration of consciousness, ability to protect their airway and arterial blood gas concentrations [25]. Adequate intravascular volume was pursued aggressively, and vasopressors are used only after volume expansion. When clinical and radiological examinations provide an estimate of elevation of intracranial pressure, osmotherapy in the form of intravenous mannitol was administered, if available, deepening sedation, and hyperventilation. Hyperglycemia and hypoglycemia were strictly avoided. Intracranial mass lesions associated with midline displacement greater than 5 mm were surgically removed when necessary. If intracranial pressure remained high despite maximal medical therapy or after intracranial mass lesion was removed, decompressive craniectomy was performed as soon as possible.
scale (GOS) score. GOS was defined as follows: 1 = death; 2 = persistent vegetative state; 3 = severe disability; 4 = moderate disability; and 5 = good recovery [26]. GOS scores were dichotomized in favorable and unfavorable outcomes (GOS of 4–5 vs. GOS of 1–3). For follow-up, we used structure telephone interviews performed by 1 doctor, blinded to clinical information and HMGB1 concentrations.
2.6. Statistical analysis Statistical analysis was performed with SPSS 10.0 (SPSS Inc., Chicago, IL) and MedCalc 9.6.4.0. (MedCalc Software, Mariakerke, Belgium). All values are expressed as mean ± standard deviation or counts (percentage) unless otherwise specified. Comparisons were made by using (1) χ 2 test or Fisher exact test for categorical data, and (2) unpaired Student t test for continuous distributed variables. The relation of HMGB1 to 1-yr outcome was assessed in a logistic-regression model with odds ratio and 95% confidence interval. For multivariate analysis, we included the significantly different outcome predictors as assessed in univariate analysis. The receiver operating characteristic curves were used to determine the best threshold for on admission values of HMGB1 to predict poor outcome. Assessment of the predictive performance of on admission values of HMGB1 was analyzed by calculating the sensitivity and specificity. In a combined logistic-regression model, we estimated the additive benefit of HMGB1 to GCS score. A P b 0.05 was considered statistically significant.
3. Results 3.1. Patient characteristics During the study period, 128 patients were admitted to our intensive care unit with an isolated head trauma diagnosis. Of these, 22 were excluded for the following reasons shown in Fig. 1, and 106 were finally included in the analysis. Table 1 shows the main characteristics of this population. The plasma HMGB1 concentration in patients was substantially higher than that in healthy controls (10.6 ± 4.1 ng/ml vs.1.4 ± 0.5 ng/ml; P b 0.0001). Moreover, the plasma HMGB1 concentration was highly associated with GCS score (r = − 0.653, P b 0.001) and C-reactive protein concentration (r = 0.594, P b 0.001).
2.4. Determination of HMGB1 in plasma 3.2. Functional outcome prediction The informed consents were obtained from healthy subjects of the control group or family members of the patients before the blood were collected. In the control group, venous blood was drawn at study entry. In the patients, venous blood was drawn on admission. The blood samples were immediately placed into sterile EDTA test tubes and centrifuged at 1500 ×g for 20 min at 4 °C to collect plasma. Plasma was stored at −70 °C until assayed. The concentration of HMGB1, C-reactive protein, fibrinogen and D-dimer in plasma was analyzed by enzyme-linked immunosorbent assay (ELISA) using commercial kits (HMGB1 ELISA kits from Shino-Test Corp., Kanagawa, Japan; C-reactive protein, fibrinogen and D-dimer ELISA kits from Huijia Biotechnology, Xiamen, Fujian Province, China) in accordance with the manufacturer's instructions. The blood samples were run in duplicate. Researchers running ELISAs were blinded to all patient details. 2.5. End point Participants were followed up until death or completion of one year after head trauma. The end points were unfavorable outcome and death after 1 y. The functional outcome was defined by Glasgow outcome
Forty-eight patients (45.3%) suffered from the unfavorable outcome in 1 y after head trauma. Higher plasma HMGB1 concentration was associated with 1-y unfavorable outcome, as well as other variables shown in Table 2. When the above variables found to be significant in the univariate analysis were introduced into the logistic model, a multivariate analysis selected GCS (OR 0.346, 95% CI 0.237–0.849, P = 0.001) and plasma HMGB1 concentration (OR 1.397, 95% CI 1.113–2.407, P = 0.002) as the independent predictors for 1-y unfavorable outcome of patients. A receiver operating characteristic curve identified that a baseline plasma HMGB1 concentration >10.8 ng/ml predicted 1-y unfavorable outcome of patients with 81.2% sensitivity and 84.5.0% specificity (area under curve, 0.880; 95% confidence interval, 0.802–0.935) (Fig. 2A). The predictive value of the HMGB1 concentration was thus similar to that of GCS scores (area under curve, 0.920; 95% confidence interval, 0.850–0.963) (P = 0.346). In a combined logistic-regression model, HMGB1 improved the area under curve of GCS scores to 0.953 (95% confidence interval, 0.893–0.984), but the differences were not significant (P = NS).
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
1739
Fig. 1. Graph documenting patients' entry into the study from screening.
3.3. Mortality prediction Thirty-one patients (29.2%) died from head trauma in 1 year. Higher plasma HMGB1 concentration was associated with 1-y mortality, as well as other variables shown in Table 3. When the above variables found to be significant in the univariate analysis were introduced into the logistic model, a multivariate analysis selected GCS (OR 0.298, 95% CI 0.195–0.766, P = 0.001) and plasma HMGB1 concentration (OR 1.641, 95% CI 1.128–2.726, P = 0.001) as the independent predictors for 1-y mortality of patients.
A receiver operating characteristic curve identified that a baseline plasma HMGB1 concentration > 11.7 ng/ml predicted 1-y mortality of patients with 90.3% sensitivity and 73.3% specificity (area under curve, 0.883; 95% confidence interval, 0.806–0.937) (Fig. 2B). The predictive value of the HMGB1 concentration was thus similar to that of GCS scores (area under curve, 0.933; 95% confidence interval, 0.867–0.972) (P = 0.234). In a combined logistic-regression model, HMGB1 improved the area under curve of GCS scores to 0.955 (95% confidence interval, 0.894–0.986), but the differences were not significant (P = 0.275). 4. Discussion
Table 1 The characteristics for 106 patients. Characteristics Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Presence of traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
76/30 45.4 ± 18.4 5.3 ± 1.9 21 (19.8%) 26 (24.5%) 5 (4.7%) 13 (12.3%) 57 (53.8%) 55 (51.9%) 49 (46.2%) 51 (48.1%) 58 (54.7%) 24 (22.6%) 93 (87.7%) 46 (43.4%) 2.1 ± 1.2 3.1 ± 1.3 11 (10.4%) 119.3 ± 31.5 71.8 ± 20.5 87.6 ± 23.2 84.3 ± 21.3 36.4 ± 0.8 18.7 ± 4.1 86.9 ± 6.9 9.8 ± 3.6 7.5 ± 2.4 4.4 ± 2.0 2.4 ± 1.1 10.6 ± 4.1
Numerical variables were presented as mean ± SD. Categorical variables were expressed as counts (percentage). GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
The current study demonstrated that plasma HMGB1 concentrations in the patients were significantly higher than those in healthy controls on admission; and in patients who died or suffered from unfavorable outcome in a year, the HMGB1 concentrations on admission were significantly higher compared with concentrations in survivors or those with favorable outcome. Previous reports have shown that elevation of HMGB1 in serum correlates with severity of acute intracerebral hemorrhage [20], and this study also found that a significant correlation emerged between plasma HMGB1 concentration and severity of acute traumatic brain injury (reflected by GCS score). Furthermore, in multivariate logistic regression models of predictors of death and unfavorable outcome, the HMGB1 concentrations on admission were an independent predictor. Importantly, the prognostic values of HMGB1 were similar to those of GCS scores, substantiating its potential as a new prognostic biomarker in traumatic brain injury. HMGB1 is a highly conserved nonhistone nuclear protein that contributes to the architecture of chromatin DNA [27]. Recently, HMGB1 was also recognized as a late mediator in septic shock as well as a proinflammatory factor [3,4]. The cytokine profile of HMGB1 has shed new light on the role of nuclear proteins and promoted the studies on roles of this unique factor in different disease conditions that are accompanied by a variety of inflammatory responses [28–31]. Extracellular HMGB1 is known to induce complex cascades of signaling via binding to its receptors, including receptor for advanced glycation end product or toll-like receptors 2 and 4 [32–34]. After stroke onset, HMGB1 released from rapidly dying cells may bind onto constitutively expressed toll-like receptor 4 in adjacent brain, thus upregulating matrix metalloproteinase-9 and expanding neurovascular damage and ischemic brain injury [35]. Recently, Muhammad et al. have demonstrated, using a mouse middle cerebral artery occlusion model, that
1740
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
Table 2 The factors associated with 1-year unfavorable outcome.
Number Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
Unfavorable outcome
Favorable outcome
P value
48 (45.3%) 37/11 46.3 ± 17.7 3.8 ± 0.8 14 (29.2%) 19 (39.6%) 3 (6.2%) 9 (18.8%) 43 (89.6%) 31 (64.6%) 35 (72.9%)
58 (54.7%) 39/19 44.6 ± 19.1 6.5 ± 1.5 7 (12.1%) 7 (12.1%) 2 (3.5%) 4 (6.9%) 14 (24.1%) 24 (41.4%) 14 (24.1%)
NS 0.626 b0.001 0.028 0.001 NS NS b0.001 0.017 b0.001
33 (68.8%)
18 (31.0%)
b0.001
32 (66.7%) 13 (27.1%) 48 (100.0%) 23 (47.9%) 2.0 ± 0.8 3.0 ± 0.9 6 (12.5%) 113.9 ± 35.3
26 (44.8%) 11 (19.0%) 45 (77.6%) 23 (39.7%) 2.2 ± 1.4 3.1 ± 1.6 5 (8.6%) 123.7 ± 27.5
0.025 NS b0.001 NS NS NS NS NS
68.8 ± 23.5
74.3 ± 17.4
NS
83.8 ± 26.2 80.9 ± 19.4 36.4 ± 0.8 18.2 ± 4.5
90.8 ± 20.0 87.1 ± 22.6 36.3 ± 0.8 19.1 ± 3.8
NS NS NS NS
84.5 ± 6.4
87.8 ± 7.8
NS
10.9 ± 4.2 8.2 ± 2.9
8.9 ± 2.8 6.8 ± 1.6
0.005 0.004
4.9 ± 1.9 2.7 ± 1.2 13.5 ± 3.6
3.9 ± 2.0 2.1 ± 1.0 8.3 ± 2.8
0.010 0.009 b0.001
examination of the patient. Obviously, measurement of circulating HMGB1 may currently not add much to the armamentariums of the clinician. However, our findings might have some potential clinical applications such as the use of HMGB1 concentration as a surrogate marker in future trials testing hemostatic agents or neuroprotective drugs.
5. Conclusions In this study, increased plasma HMGB1 concentrations are in association with 1-y functional outcome and mortality after traumatic brain injury.
Numerical variables were presented as mean±SD and analyzed by unpaired student t test. Categorical variables were expressed as counts (percentage) and analyzed by χ2 test or Fisher exact test. GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
HMGB1 engagement of receptor for advanced glycation end product triggers inflammation and mediates ischemic brain damage [36]. This study found that plasma HMGB1 concentrations were highly correlated with C-reactive protein. Connected with previous studies, our results suggest HMGB1 may act in concert to promote brain inflammation following traumatic brain injury. Severe traumatic brain injury is associated with high rates of morbidity and mortality. Early prognostication of the risk of death or of a poor long-term outcome would enable optimized care and improved allocation of health-care resources. GCS score is well known to be associated with mortality and poor clinical outcomes of severe traumatic brain injury. However, a readily measurable predictive marker predicting clinical outcomes in patients with severe traumatic brain injury would be helpful for early prognostication and risk stratification. However, it will be useful to demonstrate a varying concentration of HMGB-1 concentrations with varying severity of head injury in the moderate and mild traumatic brain injury. This is a limitation of this study and warrants further investigation. In addition, the results demonstrate that plasma HMGB1 concentration is an important outcome predictor; however, currently, the average turnaround time for laboratory results in the emergency departments is between 30 and 40 min. On average, head computed tomographic images are available to be viewed within 10 min, and the GCS score is immediately apparent upon physical
Fig. 2. Graph showing receiver operating characteristic (ROC) curve analysis of plasma high mobility group box‐1 (HMGB1) concentration for one-year unfavorable outcome (2A) and one-year mortality (2B). ROC curves were constructed based on the sensitivity and specificity of the plasma HMGB1 concentration for identifying 1-y unfavorable outcome and mortality. The area under curve (AUC) was calculated based on the ROC curves and expressed as 95% confidence interval (CI). AUC ranges from 0.5 to 1.0. An AUC closer to 1 indicates a higher predictive power.
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741 Table 3 The factors associated with 1-year mortality.
Number Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
Nonsurvival group
Survival group
P value
31 (29.2%) 23/8 44.7 ± 18.3 4.0 ± 0.9 11 (35.5%) 14 (45.2%) 3 (9.7%) 6 (19.4%) 26 (83.9%) 22 (71.0%) 26 (83.9%) 28 (90.3%)
75 (70.8%) 53/22 45.6 ± 18.6 5.8 ± 2.0 10 (13.3%) 12 (16.0%) 2 (2.7%) 7 (9.3%) 31 (41.3%) 33 (44.0%) 23 (30.7%) 23 (30.7%)
NS NS b0.001 0.009 0.002 NS NS b0.001 0.011 b0.001 b0.001
22 (71.0%) 10 (32.3%) 31 (100.0%) 16 (51.6%) 2.0 ± 0.9 3.1 ± 0.9 4 (12.9%) 121.8 ± 34.6 73.9 ± 21.0 89.8 ± 23.7 82.5 ± 19.9 36.3 ± 0.8 19.1 ± 4.8 84.7 ± 9.6
36 (48.0%) 14 (18.7%) 62 (82.7%) 30 (40.0%) 2.1 ± 1.3 3.1 ± 1.4 7 (9.3%) 118.2 ± 30.3 71.0 ± 20.3 86.7 ± 23.0 85.0 ± 22.0 36.4 ± 0.8 18.5 ± 3.9 87.9 ± 5.2
0.031 NS 0.013 NS NS NS NS NS NS NS NS NS NS NS
11.2 ± 4.5 8.6 ± 2.8
9.2 ± 3.1 7.0 ± 2.0
0.008 0.001
5.0 ± 1.9 2.8 ± 1.4 14.6 ± 3.8
4.1 ± 2.0 2.2 ± 0.9 9.0 ± 2.9
0.028 0.017 b0.001
Numerical variables were presented as mean ± SD analyzed by unpaired student t test. Categorical variables were expressed as counts (percentage) and analyzed by χ2 test or Fisher exact test. GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
Abbreviations GCS GOS HMGB1 TBI
Glasgow Coma Scale Glasgow outcome scale High-mobility group box 1 traumatic brain injury
Acknowledgment This work was financially supported by the Zhejiang Provincial Program (No. 2012KYB148). References [1] Landsman D, Bustin MA. Signature for the HMG-1 box DNA binding proteins. Bioessays 1993;15:539–46. [2] Bustin M. Regulation of DNA-dependent activities by the functional motifs of the high-mobility-group chromosomal proteins. Mol Cell Biol 1999;19:5237–46. [3] Wang H, Bloom O, Zhang M, et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science 1999;285:248–51. [4] Abraham E, Arcaroli J, Carmody A, Wang H, Tracey KJ. HMG-1 as a mediator of acute lung inflammation. J Immunol 2000;165:2950–4. [5] Lotze MT, Tracey KJ. High-mobility group box 1 protein (hmgb1): nuclear weapon in the immune arsenal. Nat Rev Immunol 2005;5:331–42.
1741
[6] Scaffidi P, Misteli T, Bianchi ME. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 2002;418:191–5. [7] Yang H, Ochani M, Li J, et al. Reversing established sepsis with antagonists of endogenous high-mobility group box 1. Proc Natl Acad Sci U S A 2004;101:296–301. [8] Ombrellino M, Wang H, Ajemian MS, et al. Increased serum concentrations of high-mobility group protein 1 in haemorrhagic shock. Lancet 1999;354:1446–7. [9] Murakami K, Koide M, Dumont TM, Russell SR, Tranmer BI, Wellman GC. Subarachnoid hemorrhage induces gliosis and increased expression of the pro-inflammatory cytokine high mobility group box 1 protein. Transl Stroke Res 2011;2:72–9. [10] Watanabe M, Miyajima M, Nakajima M, et al. Expression analysis of high mobility group box-1 protein (HMGB-1) in the cerebral cortex, hippocampus, and cerebellum of the congenital hydrocephalus (H-Tx) rat. Acta Neurochir Suppl 2012;113: 91–6. [11] Guazzi S, Strangio A, Franzi AT, Bianchi ME. HMGB1, an architectural chromatin protein and an extracellular signaling factor, has a spatially and temporally restricted expression pattern in mouse brain. Gene Expr Patterns 2003;3:29–33. [12] Agnello D, Wang H, Yang H, Tracey KJ, Ghezzi P. HMGB-1, a DNA-binding protein with cytokine activity, induces brain TNF and IL-6 production, and mediates anorexia and taste aversion. Cytokine 2002;18:231–6. [13] Wang H, Vishnubhakat JM, Bloom O, et al. Proinflammatory cytokines (tumor necrosis factor and interleukin 1) stimulate release of high mobility group protein-1 by pituicytes. Surgery 1999;126:389–92. [14] Qiu J, Nishimura M, Wang Y, et al. Early release of hmgb-1 from neurons after the onset of brain ischemia. J Cereb Blood Flow Metab 2008;28:927–38. [15] Tang D, Kang R, Cao L, et al. A pilot study to detect high mobility group box 1 and heat shock protein 72 in cerebrospinal fluid of pediatric patients with meningitis. Crit Care Med 2008;36:291–5. [16] Vogelgesang A, May VE, Grunwald U, et al. Functional status of peripheral blood T-cells in ischemic stroke patients. PLoS One 2010;5:e8718. [17] Goldstein RS, Gallowitsch-Puerta M, Yang L, et al. Elevated high-mobility group box 1 concentrations in patients with cerebral and myocardial ischemia. Shock 2006;25:571–4. [18] King MD, Laird MD, Ramesh SS, et al. Elucidating novel mechanisms of brain injury following subarachnoid hemorrhage: an emerging role for neuroproteomics. Neurosurg Focus 2010;28:E10. [19] Nakahara T, Tsuruta R, Kaneko T, et al. High-mobility group box 1 protein in CSF of patients with subarachnoid hemorrhage. Neurocrit Care 2009;11:362–8. [20] Zhou Y, Xiong KL, Lin S, et al. Elevation of high-mobility group protein box-1 in serum correlates with severity of acute intracerebral hemorrhage. Mediators Inflamm 2010 pii: 142458. [21] Juul N, Morris GF, Marshall SB, Marshall LF. Intracranial hypertension and cerebral perfusion pressure: influence on neurological deterioration and outcome in severe head injury. The Executive Committee of the International Selfotel Trial. J Neurosurg 2000;92:1–6. [22] Laird AM, Miller PR, Kilgo PD, Meredith JW, Chang MC. Relationship of early hyperglycemia to mortality in trauma patients. J Trauma 2004;56:1058–62. [23] Van den Berghe G, Schoonheydt K, Becx P, Bruyninckx F, Wouters PJ. Insulin therapy protects the central and peripheral nervous system of intensive care patients. Neurology 2005;64:1348–53. [24] Marshall LF, Marshall SB, Klauber MR, Clark MV. A new classification of head injury based on computerized tomography. J Neurosurg 1991;75:S14–20. [25] Moppett IK. Traumatic brain injury. assessment, resuscitation and early management. Br J Anaesth 2007;99:18–31. [26] Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975;1:480–4. [27] Thomas JO, Travers AA. HMG1 and 2, and related ‘architectural’ DNA-binding proteins. Trends Biochem Sci 2001;26:167–74. [28] Kokkola R, Li J, Sundberg E, et al. Successful treatment of collagen-induced arthritis in mice and rats by targeting extracellular high mobility group box chromosomal protein 1 activity. Arthritis Rheum 2003;48:2052–8. [29] Taniguchi N, Kawahara K, Yone K, et al. High mobility group box chromosomal protein 1 plays a role in the pathogenesis of rheumatoid arthritis as a novel cytokine. Arthritis Rheum 2003;48:971–81. [30] Ueno H, Matsuda T, Hashimoto S, et al. Contributions of high mobility group box protein in experimental and clinical acute lung injury. Am J Respir Crit Care Med 2004;170:1310–6. [31] Tsung A, Sahai R, Tanaka H, et al. The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia–reperfusion. J Exp Med 2005;201:1135–43. [32] Dumitriu IE, Baruah P, Manfredi AA, Bianchi ME, Rovere-Querini P. HMGB1: guiding immunity from within. Trends Immunol 2005;26:381–7. [33] Bianchi ME, Manfredi AA. High-mobility group box 1 (HMGB1) protein at the crossroads between innate and adaptive immunity. Immunol Rev 2007;220: 35–46. [34] Klune JR, Dhupar R, Cardinal J, Billiar TR, Tsung A. HMGB1: endogenous danger signaling. Mol Med 2008;14:476–84. [35] Qiu J, Xu J, Zheng Y, et al. High-mobility group box 1 promotes metalloproteinase-9 upregulation through Toll-like receptor 4 after cerebral ischemia. Stroke 2010;41: 2077–82. [36] Muhammad S, Barakat W, Stoyanov S, et al. The HMGB1 receptor RAGE mediates ischemic brain damage. J Neurosci 2008;28:12023–31.
Contents lists available at SciVerse ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Plasma high-mobility group box 1 levels and prediction of outcome in patients with traumatic brain injury Ke-Yi Wang c, Guo-Feng Yu b, Zu-Yong Zhang a, Qiang Huang b, Xiao-Qiao Dong a,⁎ a b c
Department of Neurosurgery, The First Hangzhou Municipal People's Hospital, Nanjing Medical University, 261 Huansha Road, Hangzhou 310006, China Department of Neurosurgery, Quzhou People's Hospital, 2 Zhongloudi Road, Quzhou 324100, China Central Laboratory, The First Hangzhou Municipal People's Hospital, Nanjing Medical University, 261 Huansha Road, Hangzhou 310006, China
a r t i c l e
i n f o
Article history: Received 19 May 2012 Received in revised form 29 June 2012 Accepted 2 July 2012 Available online 10 July 2012 Keywords: High-mobility group box 1 Traumatic brain injury Functional outcome Mortality
a b s t r a c t Background: High-mobility group box 1 (HMGB1), a marker of inflammation, has been associated with poor outcome of critical illness. The present study was undertaken to investigate the plasma HMGB1 concentrations in patients with traumatic brain injury (TBI) and to analyze the correlation of HMGB1 with TBI outcome. Methods: We performed an observational, clinical study. Plasma HMGB1 concentration of 106 healthy subjects and 106 patients with severe TBI was measured by ELISA. The correlation with 1-y mortality and unfavorable outcome (Glasgow Outcome Scale score of 1–3) was analyzed. Results: Thirty-one patients (29.2%) died and 48 patients (45.3%) had an unfavorable outcome at 1 y after TBI. Plasma HMGB1 level was substantially higher in patients than in healthy controls. A multivariate analysis selected plasma HMGB1 level as an independent predictor for 1-y mortality and unfavorable outcome of patients. A receiver operating characteristic curve analysis showed plasma HMGB1 level statistically significantly predicted 1-y mortality and unfavorable outcome. The prognostic value of HMGB1 was similar to that of Glasgow Coma Scale score for 1-y clinical outcomes. Conclusions: Plasma HMGB1 concentration emerges as a novel biomarker for predicting 1-y clinical outcomes of TBI. © 2012 Elsevier B.V. All rights reserved.
1. Introduction High-mobility group box 1 (HMGB1), a nonhistone DNA binding protein, has been implicated in the stabilization of nucleosomal structure and the facilitation of gene transcription [1,2], and further identified as a cytokine-like mediator of delayed endotoxin lethality and acute lung injury [3,4]. HMGB1, actively secreted by immune cells or released by necrotic cells into the extracellular milieu, might be involved in the triggering of inflammation [5,6]. Recombinant HMGB1 can induce acute inflammation in animal models of lung injury and endotoxemia [3,4], and anti-HMGB1 antibody can attenuate endotoxininduced lethality [7]. In addition, high serum concentrations of HMGB1 in patients with sepsis or hemorrhagic shock have been reported to be associated with increased mortality and disease severity [3,8]. HMGB1 is widely expressed in the brain [9–11]. Moreover, in the brain, HMGB1 has been reported to be released after cytokine stimulation and to be involved in the inflammatory process after it was administered into the cerebral ventricles [12,13]. HMGB1 can be released from damaged neurons in mouse models with middle cerebral artery occlusion [14]. Moreover, HMGB1 is increased in the cerebrospinal fluid from meningitis patients [15] and in the serum of ⁎ Corresponding author. Tel.: +86 571 87065701. E-mail address: [email protected] (X.-Q. Dong). 0009-8981/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2012.07.002
cerebral ischemia patients [16,17]. Recent data have identified HMGB1 in the cerebrospinal fluid of subarachnoid hemorrhage [18,19] and in the serum of intracerebral hemorrhage [20] as a potential biomarker of neurological outcome, suggesting HMGB1 may represent a marker of neurological injury. Thus, we furthermore investigated the change of HMGB1 in the peripheral blood and its association with disease outcome in traumatic brain injury. 2. Subjects and methods 2.1. Study population The inclusion period was from January 2008 to April 2010. Inclusion criteria were isolated head trauma and post resuscitation Glasgow Coma Scale (GCS) score of ≤ 8. Exclusion criteria were b 18 y, admission time > 6 h, previous head trauma, neurological disease including ischemic or hemorrhagic stroke, use of antiplatelet or anticoagulant medication, and presence of other prior systemic diseases including uremia, liver cirrhosis, malignancy, chronic heart or lung disease, diabetes mellitus and hypertension. Subjects were initially evaluated if they presented to our hospital and had blood collected and brain magnetic resonance imaging completed as parts of medical examination. Finally, a control group consisted of 106 healthy ageand sex-matched subjects with normal results on brain magnetic
1738
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
resonance imaging and without vascular risk factors. Informed consent to participate in the study was obtained from them or their relatives. This protocol was approved by the Ethics Committee of Quzhou People's Hospital before implementation. 2.2. Clinical and radiological assessment At admission, clinical severity was assessed using initial post resuscitation GCS score. Shock was defined as systolic blood pressure b90 mm Hg, and hypoxia was defined as blood oxygen saturation b85% [21]. Hyperglycemia was defined as blood glucose >11.1 mmol/l [22]. Hypoglycemia was defined as blood glucose b2.2 mmol/l [23]. Neurology deterioration was defined as occurring in patients who manifested clinically identified episodes of ≥1 of the following: 1) a spontaneous decrease in GCS motor scores of ≥2 points from the previous examination; 2) a further loss of pupillary reactivity; 3) development of pupillary asymmetry greater than 1 mm; or 4) deterioration in neurological status sufficient to warrant immediate medical or surgical intervention [21]. All computerized tomography (CT) scans were performed according to the neuroradiology department protocol. Investigators who read them were blinded to clinical information. CT classification was performed using Traumatic Coma Data Bank criteria on initial post resuscitation CT scan according to Marshall et al. [24]. 2.3. Patient management The treatments included surgical therapy, ventilatory support, arterial pressure maintenance, glycemic control, intravenous fluids, hyperosmolar agents, H2 blockers, early nutritional support, and physical therapy. The decision to intubate and use mechanical ventilation was based on the individuals' concentration of consciousness, ability to protect their airway and arterial blood gas concentrations [25]. Adequate intravascular volume was pursued aggressively, and vasopressors are used only after volume expansion. When clinical and radiological examinations provide an estimate of elevation of intracranial pressure, osmotherapy in the form of intravenous mannitol was administered, if available, deepening sedation, and hyperventilation. Hyperglycemia and hypoglycemia were strictly avoided. Intracranial mass lesions associated with midline displacement greater than 5 mm were surgically removed when necessary. If intracranial pressure remained high despite maximal medical therapy or after intracranial mass lesion was removed, decompressive craniectomy was performed as soon as possible.
scale (GOS) score. GOS was defined as follows: 1 = death; 2 = persistent vegetative state; 3 = severe disability; 4 = moderate disability; and 5 = good recovery [26]. GOS scores were dichotomized in favorable and unfavorable outcomes (GOS of 4–5 vs. GOS of 1–3). For follow-up, we used structure telephone interviews performed by 1 doctor, blinded to clinical information and HMGB1 concentrations.
2.6. Statistical analysis Statistical analysis was performed with SPSS 10.0 (SPSS Inc., Chicago, IL) and MedCalc 9.6.4.0. (MedCalc Software, Mariakerke, Belgium). All values are expressed as mean ± standard deviation or counts (percentage) unless otherwise specified. Comparisons were made by using (1) χ 2 test or Fisher exact test for categorical data, and (2) unpaired Student t test for continuous distributed variables. The relation of HMGB1 to 1-yr outcome was assessed in a logistic-regression model with odds ratio and 95% confidence interval. For multivariate analysis, we included the significantly different outcome predictors as assessed in univariate analysis. The receiver operating characteristic curves were used to determine the best threshold for on admission values of HMGB1 to predict poor outcome. Assessment of the predictive performance of on admission values of HMGB1 was analyzed by calculating the sensitivity and specificity. In a combined logistic-regression model, we estimated the additive benefit of HMGB1 to GCS score. A P b 0.05 was considered statistically significant.
3. Results 3.1. Patient characteristics During the study period, 128 patients were admitted to our intensive care unit with an isolated head trauma diagnosis. Of these, 22 were excluded for the following reasons shown in Fig. 1, and 106 were finally included in the analysis. Table 1 shows the main characteristics of this population. The plasma HMGB1 concentration in patients was substantially higher than that in healthy controls (10.6 ± 4.1 ng/ml vs.1.4 ± 0.5 ng/ml; P b 0.0001). Moreover, the plasma HMGB1 concentration was highly associated with GCS score (r = − 0.653, P b 0.001) and C-reactive protein concentration (r = 0.594, P b 0.001).
2.4. Determination of HMGB1 in plasma 3.2. Functional outcome prediction The informed consents were obtained from healthy subjects of the control group or family members of the patients before the blood were collected. In the control group, venous blood was drawn at study entry. In the patients, venous blood was drawn on admission. The blood samples were immediately placed into sterile EDTA test tubes and centrifuged at 1500 ×g for 20 min at 4 °C to collect plasma. Plasma was stored at −70 °C until assayed. The concentration of HMGB1, C-reactive protein, fibrinogen and D-dimer in plasma was analyzed by enzyme-linked immunosorbent assay (ELISA) using commercial kits (HMGB1 ELISA kits from Shino-Test Corp., Kanagawa, Japan; C-reactive protein, fibrinogen and D-dimer ELISA kits from Huijia Biotechnology, Xiamen, Fujian Province, China) in accordance with the manufacturer's instructions. The blood samples were run in duplicate. Researchers running ELISAs were blinded to all patient details. 2.5. End point Participants were followed up until death or completion of one year after head trauma. The end points were unfavorable outcome and death after 1 y. The functional outcome was defined by Glasgow outcome
Forty-eight patients (45.3%) suffered from the unfavorable outcome in 1 y after head trauma. Higher plasma HMGB1 concentration was associated with 1-y unfavorable outcome, as well as other variables shown in Table 2. When the above variables found to be significant in the univariate analysis were introduced into the logistic model, a multivariate analysis selected GCS (OR 0.346, 95% CI 0.237–0.849, P = 0.001) and plasma HMGB1 concentration (OR 1.397, 95% CI 1.113–2.407, P = 0.002) as the independent predictors for 1-y unfavorable outcome of patients. A receiver operating characteristic curve identified that a baseline plasma HMGB1 concentration >10.8 ng/ml predicted 1-y unfavorable outcome of patients with 81.2% sensitivity and 84.5.0% specificity (area under curve, 0.880; 95% confidence interval, 0.802–0.935) (Fig. 2A). The predictive value of the HMGB1 concentration was thus similar to that of GCS scores (area under curve, 0.920; 95% confidence interval, 0.850–0.963) (P = 0.346). In a combined logistic-regression model, HMGB1 improved the area under curve of GCS scores to 0.953 (95% confidence interval, 0.893–0.984), but the differences were not significant (P = NS).
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
1739
Fig. 1. Graph documenting patients' entry into the study from screening.
3.3. Mortality prediction Thirty-one patients (29.2%) died from head trauma in 1 year. Higher plasma HMGB1 concentration was associated with 1-y mortality, as well as other variables shown in Table 3. When the above variables found to be significant in the univariate analysis were introduced into the logistic model, a multivariate analysis selected GCS (OR 0.298, 95% CI 0.195–0.766, P = 0.001) and plasma HMGB1 concentration (OR 1.641, 95% CI 1.128–2.726, P = 0.001) as the independent predictors for 1-y mortality of patients.
A receiver operating characteristic curve identified that a baseline plasma HMGB1 concentration > 11.7 ng/ml predicted 1-y mortality of patients with 90.3% sensitivity and 73.3% specificity (area under curve, 0.883; 95% confidence interval, 0.806–0.937) (Fig. 2B). The predictive value of the HMGB1 concentration was thus similar to that of GCS scores (area under curve, 0.933; 95% confidence interval, 0.867–0.972) (P = 0.234). In a combined logistic-regression model, HMGB1 improved the area under curve of GCS scores to 0.955 (95% confidence interval, 0.894–0.986), but the differences were not significant (P = 0.275). 4. Discussion
Table 1 The characteristics for 106 patients. Characteristics Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Presence of traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
76/30 45.4 ± 18.4 5.3 ± 1.9 21 (19.8%) 26 (24.5%) 5 (4.7%) 13 (12.3%) 57 (53.8%) 55 (51.9%) 49 (46.2%) 51 (48.1%) 58 (54.7%) 24 (22.6%) 93 (87.7%) 46 (43.4%) 2.1 ± 1.2 3.1 ± 1.3 11 (10.4%) 119.3 ± 31.5 71.8 ± 20.5 87.6 ± 23.2 84.3 ± 21.3 36.4 ± 0.8 18.7 ± 4.1 86.9 ± 6.9 9.8 ± 3.6 7.5 ± 2.4 4.4 ± 2.0 2.4 ± 1.1 10.6 ± 4.1
Numerical variables were presented as mean ± SD. Categorical variables were expressed as counts (percentage). GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
The current study demonstrated that plasma HMGB1 concentrations in the patients were significantly higher than those in healthy controls on admission; and in patients who died or suffered from unfavorable outcome in a year, the HMGB1 concentrations on admission were significantly higher compared with concentrations in survivors or those with favorable outcome. Previous reports have shown that elevation of HMGB1 in serum correlates with severity of acute intracerebral hemorrhage [20], and this study also found that a significant correlation emerged between plasma HMGB1 concentration and severity of acute traumatic brain injury (reflected by GCS score). Furthermore, in multivariate logistic regression models of predictors of death and unfavorable outcome, the HMGB1 concentrations on admission were an independent predictor. Importantly, the prognostic values of HMGB1 were similar to those of GCS scores, substantiating its potential as a new prognostic biomarker in traumatic brain injury. HMGB1 is a highly conserved nonhistone nuclear protein that contributes to the architecture of chromatin DNA [27]. Recently, HMGB1 was also recognized as a late mediator in septic shock as well as a proinflammatory factor [3,4]. The cytokine profile of HMGB1 has shed new light on the role of nuclear proteins and promoted the studies on roles of this unique factor in different disease conditions that are accompanied by a variety of inflammatory responses [28–31]. Extracellular HMGB1 is known to induce complex cascades of signaling via binding to its receptors, including receptor for advanced glycation end product or toll-like receptors 2 and 4 [32–34]. After stroke onset, HMGB1 released from rapidly dying cells may bind onto constitutively expressed toll-like receptor 4 in adjacent brain, thus upregulating matrix metalloproteinase-9 and expanding neurovascular damage and ischemic brain injury [35]. Recently, Muhammad et al. have demonstrated, using a mouse middle cerebral artery occlusion model, that
1740
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741
Table 2 The factors associated with 1-year unfavorable outcome.
Number Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
Unfavorable outcome
Favorable outcome
P value
48 (45.3%) 37/11 46.3 ± 17.7 3.8 ± 0.8 14 (29.2%) 19 (39.6%) 3 (6.2%) 9 (18.8%) 43 (89.6%) 31 (64.6%) 35 (72.9%)
58 (54.7%) 39/19 44.6 ± 19.1 6.5 ± 1.5 7 (12.1%) 7 (12.1%) 2 (3.5%) 4 (6.9%) 14 (24.1%) 24 (41.4%) 14 (24.1%)
NS 0.626 b0.001 0.028 0.001 NS NS b0.001 0.017 b0.001
33 (68.8%)
18 (31.0%)
b0.001
32 (66.7%) 13 (27.1%) 48 (100.0%) 23 (47.9%) 2.0 ± 0.8 3.0 ± 0.9 6 (12.5%) 113.9 ± 35.3
26 (44.8%) 11 (19.0%) 45 (77.6%) 23 (39.7%) 2.2 ± 1.4 3.1 ± 1.6 5 (8.6%) 123.7 ± 27.5
0.025 NS b0.001 NS NS NS NS NS
68.8 ± 23.5
74.3 ± 17.4
NS
83.8 ± 26.2 80.9 ± 19.4 36.4 ± 0.8 18.2 ± 4.5
90.8 ± 20.0 87.1 ± 22.6 36.3 ± 0.8 19.1 ± 3.8
NS NS NS NS
84.5 ± 6.4
87.8 ± 7.8
NS
10.9 ± 4.2 8.2 ± 2.9
8.9 ± 2.8 6.8 ± 1.6
0.005 0.004
4.9 ± 1.9 2.7 ± 1.2 13.5 ± 3.6
3.9 ± 2.0 2.1 ± 1.0 8.3 ± 2.8
0.010 0.009 b0.001
examination of the patient. Obviously, measurement of circulating HMGB1 may currently not add much to the armamentariums of the clinician. However, our findings might have some potential clinical applications such as the use of HMGB1 concentration as a surrogate marker in future trials testing hemostatic agents or neuroprotective drugs.
5. Conclusions In this study, increased plasma HMGB1 concentrations are in association with 1-y functional outcome and mortality after traumatic brain injury.
Numerical variables were presented as mean±SD and analyzed by unpaired student t test. Categorical variables were expressed as counts (percentage) and analyzed by χ2 test or Fisher exact test. GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
HMGB1 engagement of receptor for advanced glycation end product triggers inflammation and mediates ischemic brain damage [36]. This study found that plasma HMGB1 concentrations were highly correlated with C-reactive protein. Connected with previous studies, our results suggest HMGB1 may act in concert to promote brain inflammation following traumatic brain injury. Severe traumatic brain injury is associated with high rates of morbidity and mortality. Early prognostication of the risk of death or of a poor long-term outcome would enable optimized care and improved allocation of health-care resources. GCS score is well known to be associated with mortality and poor clinical outcomes of severe traumatic brain injury. However, a readily measurable predictive marker predicting clinical outcomes in patients with severe traumatic brain injury would be helpful for early prognostication and risk stratification. However, it will be useful to demonstrate a varying concentration of HMGB-1 concentrations with varying severity of head injury in the moderate and mild traumatic brain injury. This is a limitation of this study and warrants further investigation. In addition, the results demonstrate that plasma HMGB1 concentration is an important outcome predictor; however, currently, the average turnaround time for laboratory results in the emergency departments is between 30 and 40 min. On average, head computed tomographic images are available to be viewed within 10 min, and the GCS score is immediately apparent upon physical
Fig. 2. Graph showing receiver operating characteristic (ROC) curve analysis of plasma high mobility group box‐1 (HMGB1) concentration for one-year unfavorable outcome (2A) and one-year mortality (2B). ROC curves were constructed based on the sensitivity and specificity of the plasma HMGB1 concentration for identifying 1-y unfavorable outcome and mortality. The area under curve (AUC) was calculated based on the ROC curves and expressed as 95% confidence interval (CI). AUC ranges from 0.5 to 1.0. An AUC closer to 1 indicates a higher predictive power.
K.-Y. Wang et al. / Clinica Chimica Acta 413 (2012) 1737–1741 Table 3 The factors associated with 1-year mortality.
Number Sex (male/female) Age (y) GCS score on admission Shock on admission Hyperglycemia on admission Hypoglycemia on admission Hypoxia on admission Pupils unreactive on admission CT classification 5 or 6 Abnormal cisterns on initial CT scan Midline shift >5 mm on initial CT scan Traumatic SAH on initial CT scan Neurological deterioration Mechanical ventilation Intracranial surgery in 1st 24 h Admission time (h) Plasma-sampling time (h) Seizure Systolic arterial pressure (mm Hg) Diastolic arterial pressure (mm Hg) Mean arterial pressure (mm Hg) Heart rate (beats/min) Body temperature (°C) Respiratory rate (respirations/min) Blood oxygen saturation (percentage) Blood glucose level (mmol/l) Plasma C-reactive protein level (mg/l) Plasma fibrinogen level (g/l) Plasma D-dimer level (mg/l) Plasma HMGB1 level (ng/ml)
Nonsurvival group
Survival group
P value
31 (29.2%) 23/8 44.7 ± 18.3 4.0 ± 0.9 11 (35.5%) 14 (45.2%) 3 (9.7%) 6 (19.4%) 26 (83.9%) 22 (71.0%) 26 (83.9%) 28 (90.3%)
75 (70.8%) 53/22 45.6 ± 18.6 5.8 ± 2.0 10 (13.3%) 12 (16.0%) 2 (2.7%) 7 (9.3%) 31 (41.3%) 33 (44.0%) 23 (30.7%) 23 (30.7%)
NS NS b0.001 0.009 0.002 NS NS b0.001 0.011 b0.001 b0.001
22 (71.0%) 10 (32.3%) 31 (100.0%) 16 (51.6%) 2.0 ± 0.9 3.1 ± 0.9 4 (12.9%) 121.8 ± 34.6 73.9 ± 21.0 89.8 ± 23.7 82.5 ± 19.9 36.3 ± 0.8 19.1 ± 4.8 84.7 ± 9.6
36 (48.0%) 14 (18.7%) 62 (82.7%) 30 (40.0%) 2.1 ± 1.3 3.1 ± 1.4 7 (9.3%) 118.2 ± 30.3 71.0 ± 20.3 86.7 ± 23.0 85.0 ± 22.0 36.4 ± 0.8 18.5 ± 3.9 87.9 ± 5.2
0.031 NS 0.013 NS NS NS NS NS NS NS NS NS NS NS
11.2 ± 4.5 8.6 ± 2.8
9.2 ± 3.1 7.0 ± 2.0
0.008 0.001
5.0 ± 1.9 2.8 ± 1.4 14.6 ± 3.8
4.1 ± 2.0 2.2 ± 0.9 9.0 ± 2.9
0.028 0.017 b0.001
Numerical variables were presented as mean ± SD analyzed by unpaired student t test. Categorical variables were expressed as counts (percentage) and analyzed by χ2 test or Fisher exact test. GCS indicates Glasgow Coma Scale; CT, computerized tomography; SAH, subarachnoid hemorrhage; HMGB1, high mobility group box-1.
Abbreviations GCS GOS HMGB1 TBI
Glasgow Coma Scale Glasgow outcome scale High-mobility group box 1 traumatic brain injury
Acknowledgment This work was financially supported by the Zhejiang Provincial Program (No. 2012KYB148). References [1] Landsman D, Bustin MA. Signature for the HMG-1 box DNA binding proteins. Bioessays 1993;15:539–46. [2] Bustin M. Regulation of DNA-dependent activities by the functional motifs of the high-mobility-group chromosomal proteins. Mol Cell Biol 1999;19:5237–46. [3] Wang H, Bloom O, Zhang M, et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science 1999;285:248–51. [4] Abraham E, Arcaroli J, Carmody A, Wang H, Tracey KJ. HMG-1 as a mediator of acute lung inflammation. J Immunol 2000;165:2950–4. [5] Lotze MT, Tracey KJ. High-mobility group box 1 protein (hmgb1): nuclear weapon in the immune arsenal. Nat Rev Immunol 2005;5:331–42.
1741
[6] Scaffidi P, Misteli T, Bianchi ME. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 2002;418:191–5. [7] Yang H, Ochani M, Li J, et al. Reversing established sepsis with antagonists of endogenous high-mobility group box 1. Proc Natl Acad Sci U S A 2004;101:296–301. [8] Ombrellino M, Wang H, Ajemian MS, et al. Increased serum concentrations of high-mobility group protein 1 in haemorrhagic shock. Lancet 1999;354:1446–7. [9] Murakami K, Koide M, Dumont TM, Russell SR, Tranmer BI, Wellman GC. Subarachnoid hemorrhage induces gliosis and increased expression of the pro-inflammatory cytokine high mobility group box 1 protein. Transl Stroke Res 2011;2:72–9. [10] Watanabe M, Miyajima M, Nakajima M, et al. Expression analysis of high mobility group box-1 protein (HMGB-1) in the cerebral cortex, hippocampus, and cerebellum of the congenital hydrocephalus (H-Tx) rat. Acta Neurochir Suppl 2012;113: 91–6. [11] Guazzi S, Strangio A, Franzi AT, Bianchi ME. HMGB1, an architectural chromatin protein and an extracellular signaling factor, has a spatially and temporally restricted expression pattern in mouse brain. Gene Expr Patterns 2003;3:29–33. [12] Agnello D, Wang H, Yang H, Tracey KJ, Ghezzi P. HMGB-1, a DNA-binding protein with cytokine activity, induces brain TNF and IL-6 production, and mediates anorexia and taste aversion. Cytokine 2002;18:231–6. [13] Wang H, Vishnubhakat JM, Bloom O, et al. Proinflammatory cytokines (tumor necrosis factor and interleukin 1) stimulate release of high mobility group protein-1 by pituicytes. Surgery 1999;126:389–92. [14] Qiu J, Nishimura M, Wang Y, et al. Early release of hmgb-1 from neurons after the onset of brain ischemia. J Cereb Blood Flow Metab 2008;28:927–38. [15] Tang D, Kang R, Cao L, et al. A pilot study to detect high mobility group box 1 and heat shock protein 72 in cerebrospinal fluid of pediatric patients with meningitis. Crit Care Med 2008;36:291–5. [16] Vogelgesang A, May VE, Grunwald U, et al. Functional status of peripheral blood T-cells in ischemic stroke patients. PLoS One 2010;5:e8718. [17] Goldstein RS, Gallowitsch-Puerta M, Yang L, et al. Elevated high-mobility group box 1 concentrations in patients with cerebral and myocardial ischemia. Shock 2006;25:571–4. [18] King MD, Laird MD, Ramesh SS, et al. Elucidating novel mechanisms of brain injury following subarachnoid hemorrhage: an emerging role for neuroproteomics. Neurosurg Focus 2010;28:E10. [19] Nakahara T, Tsuruta R, Kaneko T, et al. High-mobility group box 1 protein in CSF of patients with subarachnoid hemorrhage. Neurocrit Care 2009;11:362–8. [20] Zhou Y, Xiong KL, Lin S, et al. Elevation of high-mobility group protein box-1 in serum correlates with severity of acute intracerebral hemorrhage. Mediators Inflamm 2010 pii: 142458. [21] Juul N, Morris GF, Marshall SB, Marshall LF. Intracranial hypertension and cerebral perfusion pressure: influence on neurological deterioration and outcome in severe head injury. The Executive Committee of the International Selfotel Trial. J Neurosurg 2000;92:1–6. [22] Laird AM, Miller PR, Kilgo PD, Meredith JW, Chang MC. Relationship of early hyperglycemia to mortality in trauma patients. J Trauma 2004;56:1058–62. [23] Van den Berghe G, Schoonheydt K, Becx P, Bruyninckx F, Wouters PJ. Insulin therapy protects the central and peripheral nervous system of intensive care patients. Neurology 2005;64:1348–53. [24] Marshall LF, Marshall SB, Klauber MR, Clark MV. A new classification of head injury based on computerized tomography. J Neurosurg 1991;75:S14–20. [25] Moppett IK. Traumatic brain injury. assessment, resuscitation and early management. Br J Anaesth 2007;99:18–31. [26] Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975;1:480–4. [27] Thomas JO, Travers AA. HMG1 and 2, and related ‘architectural’ DNA-binding proteins. Trends Biochem Sci 2001;26:167–74. [28] Kokkola R, Li J, Sundberg E, et al. Successful treatment of collagen-induced arthritis in mice and rats by targeting extracellular high mobility group box chromosomal protein 1 activity. Arthritis Rheum 2003;48:2052–8. [29] Taniguchi N, Kawahara K, Yone K, et al. High mobility group box chromosomal protein 1 plays a role in the pathogenesis of rheumatoid arthritis as a novel cytokine. Arthritis Rheum 2003;48:971–81. [30] Ueno H, Matsuda T, Hashimoto S, et al. Contributions of high mobility group box protein in experimental and clinical acute lung injury. Am J Respir Crit Care Med 2004;170:1310–6. [31] Tsung A, Sahai R, Tanaka H, et al. The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia–reperfusion. J Exp Med 2005;201:1135–43. [32] Dumitriu IE, Baruah P, Manfredi AA, Bianchi ME, Rovere-Querini P. HMGB1: guiding immunity from within. Trends Immunol 2005;26:381–7. [33] Bianchi ME, Manfredi AA. High-mobility group box 1 (HMGB1) protein at the crossroads between innate and adaptive immunity. Immunol Rev 2007;220: 35–46. [34] Klune JR, Dhupar R, Cardinal J, Billiar TR, Tsung A. HMGB1: endogenous danger signaling. Mol Med 2008;14:476–84. [35] Qiu J, Xu J, Zheng Y, et al. High-mobility group box 1 promotes metalloproteinase-9 upregulation through Toll-like receptor 4 after cerebral ischemia. Stroke 2010;41: 2077–82. [36] Muhammad S, Barakat W, Stoyanov S, et al. The HMGB1 receptor RAGE mediates ischemic brain damage. J Neurosci 2008;28:12023–31.