Resuscitation 85 (2014) 1654–1661
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Post-cardiac arrest serum levels of glial fibrillary acidic protein for predicting neurological outcome Ing-Marie Larsson a,∗ , Ewa Wallin a , Marja-Leena Kristofferzon b,c , Marion Niessner d , Henrik Zetterberg e,f , Sten Rubertsson a a
Department of Surgical Sciences–Anaesthesiology & Intensive Care, Uppsala University, SE- 751 85 Uppsala, Sweden Faculty of Health and Occupational Studies, Department of Health and Caring Sciences, University of Gävle, SE- 801 76 Gävle, Sweden c Department of Public Health and Caring Sciences, Uppsala University, SE- 751 22 Uppsala, Sweden d Roche Diagnostics GmbH, Penzberg, Germany e Department of Psychiatry and Neurochemistry, the Sahlgrenska Academy at the University of Gothenburg, Institute of Neuroscience and Physiology, Mölndal, Sweden f UCL Institute of Neurology, Queen Square, London WC1N 3BG, United Kingdom b
a r t i c l e
i n f o
Article history: Received 5 June 2014 Received in revised form 6 August 2014 Accepted 9 September 2014 Keywords: Cardiac arrest Therapeutic hypothermia Prognostication Biomarkers
a b s t r a c t Aim of the study: To investigate serum levels of glial fibrillary acidic protein (GFAP) for evaluation of neurological outcome in cardiac arrest (CA) patients and compare GFAP sensitivity and specificity to that of more studied biomarkers neuron-specific enolas (NSE) and S100B. Method: A prospective observational study was performed in three hospitals in Sweden during 20082012. The participants were 125 CA patients treated with therapeutic hypothermia (TH) to 32-34 ◦ C for 24 hours. Samples were collected from peripheral blood (n = 125) and the jugular bulb (n = 47) up to 108 hours post-CA. GFAP serum levels were quantified using a novel, fully automated immunochemical method. Other biomarkers investigated were NSE and S100B. Neurological outcome was assessed using the Cerebral Performance Categories scale (CPC) and dichotomized into good and poor outcome. Results: GFAP predicted poor neurological outcome with 100% specificity and 14-23% sensitivity at 24, 48 and 72 hours post-CA. The corresponding values for NSE were 27-50% sensitivity and for S100B 2130% sensitivity when specificity was set to 100%. A logistic regression with stepwise combination of the investigated biomarkers, GFAP, did not increase the ability to predict neurological outcome. No differences were found in GFAP, NSE and S100B levels when peripheral and jugular bulb blood samples were compared. Conclusion: Serum GFAP increase in patients with poor outcome but did not show sufficient sensitivity to predict neurological outcome after CA. Both NSE and S100B were shown to be better predictors. The ability to predict neurological outcome did not increased when combining the three biomarkers. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Cardiac arrest (CA) is associated with high mortality rate and among those who survive, there is a major risk of neurological sequelae.1–3 During CA, the brain is exposed to hypoxia, which may cause anoxic brain injury and associated long-term dysfunction.2,4 Treating comatose survivors after CA with therapeutic hypothermia (TH) to 32-34 ◦ C for 12-24 hours has been shown to improve neurological outcome 5,6 and survival,5 and TH
∗ Corresponding author. Department of Surgical Sciences–Anaesthesiology & Intensive Care, Uppsala University, SE-751 85 Uppsala, Sweden. E-mail address: [email protected] (I.-M. Larsson). http://dx.doi.org/10.1016/j.resuscitation.2014.09.007 0300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.
is now recommended as treatment after CA.7,8 A recently published study showed no difference in neurological outcome and survival rate when using TH at 33 ◦ C or 36 ◦ C.9 Prognostication and prediction of neurological outcome are important aspects of care of CA patients. It is recommended that prognostication be postponed until 72 hours after normothermia.10,11 Several prognostication methods have been reported: clinical examination, neurophysiological examination (electroencephalogram (EEG), somatosensory evoked potentials (SSEP)), imaging (magnetic resonance imaging (MRI), computed tomography (CT)) and biomarkers.10–12 Biomarkers can be proteins or peptides released from the brain across the blood-brain barrier (BBB) into the systemic circulation.13 They may provide information about the severity of brain damage.
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
However, there are limitations regarding reliability, such as different cut-off values for predicting poor outcomes, influence of hemolysis in the sample,10,11,13 and the fact that levels may differ across laboratories.14 Of the candidate blood biomarkers, most data are available for neuron-specific enolase (NSE) and S-100B, but other biomarkers have also been investigated.11,13,15–19 NSE is a dimeric enzyme, mainly expressed in neurons, and its half-life is estimated to 24-30 h.13 Red blood cells contain NSE, and therefore, hemolysis affects NSE levels.11,13 Increased NSE concentrations after CA are associated with poor outcome in comatose patients, but there is as yet no established cut-off value for this marker.11,13,20–23 S100B is an intracellular calcium-binding protein with particularly high expression in astroglial cells in the white matter of the brain. S100B has a short half-life, about 2 hours, with its highest levels during the first 24 hours after anoxic brain injury.13 Its serum levels may also be influenced by release from fat and skeletal tissues.24 Different cut-off values for predicting poor neurological outcome after CA have been reported.11,13,15,20,25,26 Glial fibrillary acidic protein (GFAP) is a filament exclusively expressed by mature astrocytes of the central nervous system (CNS).27 GFAP has been shown to be a biomarker of traumatic brain injury (TBI), and GFAP levels are related to the severity of the TBI.28,29 Moreover, elevated serum GFAP levels have been reported in stroke patients.30 The levels of GFAP in CA patients tend to increase in patients with poor neurological outcome.15,31 Since GFAP is a relatively new biomarker in the field of prognostication after CA, the aim of this study was to examine the association of serum GFAP levels, determined using a novel, fully automated immunochemical method, at different time points post-CA with outcome and compare its sensitivity and specificity to NSE and S100B.
1655
2. Materials and methods 2.1. Study population The present investigation was a prospective observational study conducted at one university hospital and two general county hospitals in Sweden. Inclusion criteria were CA treated with TH and age > 18 years. To start TH, successful restoration of spontaneous circulation (ROSC) with systolic arterial blood pressure ≥ 80 mmHg > 5 min and unconscious with a Glasgow Coma Scale (GSC)32 score < 8 had to be achieved. During the study period, May, 2008 to May, 2012, 242 CA patients were admitted to the ICUs, 209 underwent TH, and of them, 125 were included in the study. Sixty-three patients survived until hospital discharge (Fig. 1). 2.2. Hypothermia treatment The decision to start TH was taken for all patients, irrespective of the first registered ECG rhythm or whether the CA occurred in or out of hospital. The patients were cooled to 32-34 ◦ C for 24 hours. Hypothermia treatment was induced by infusion of 4 ◦ C saline, in a planned volume of 30-40 ml/kg. To maintain cooling, all hospitals used external cooling, either ice packs or cooling suits depending on hospital-specific routines. Rewarming at a rate of 0.5 ◦ C/h was used, and 36 ◦ C was considered the normal core temperature. Before cooling, all patients were sedated, intubated and mechanically ventilated according to local ICU guidelines for severely ill patients adapted for TH after CA. 2.3. Data collection Blood samples for biomarkers were collected as soon as possible upon arrival to the ICU and at 24, 48, 72, 96 and 108 h after
Fig. 1. Flow chart over CA patients admitted to ICU.
1656
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
CA from a peripheral artery or vein. In patients who were provided with a catheter in the jugular bulb, blood samples were collected at the same time points. The samples were stored in a -70 ◦ C freezer and analysed at the same time after the study period. Medical background variables, information about the CA and data on temperature management were retrieved from the medical chart. Results of the biomarkers were not available to the caregivers and therefore could not affect treatment or evaluation of the patient. 2.4. Analysis of biomarkers Serum levels of NSE and S-100B were measured using a Cobas e601 instrument and NSE and S-100B reagent kits as described by the manufacturer (Roche Diagnostics, Penzberg, Germany). For detection of hemolysis, serum measurement index of hemolysis (Sindh) was used. Serum GFAP was measured using the non-commercial Elecsys® GFAP prototype test on a Cobas e411 instrument (Roche, Penzberg, Germany). In the first step, biotinand ruthenium-labeled monoclonal anti-GFAP antibodies are combined with 50 l of sample and incubated for 9 minutes. In the second step, streptavidin-coated magnetic microparticles are added and the mixture is incubated for further 9 minutes. After the second incubation step, the reaction mixture is transferred to the measuring cell where the beads are captured on the surface of an electrode by a magnet. Unbound label is removed by washing the measuring cell. In the last step, voltage is applied to the electrode in the presence of a tri-propylamine (TPA)-containing buffer, the resulting electrochemiluminescent signal is recorded by a photomultiplier and the GFAP concentration derived from a calibration curve. Because no acknowledged reference method is available at present, the method has been standardized by weighing pure human GFAP in analyte-free serum matrix. Three internal samples based on human serum matrix spiked with human GFAP low (< 5 ng/ml), medium (20–30 ng/ml) and high (> 50 ng/ml) were used for quality control during the assay. The within- and between-run precisions were 1.1-1.9% and 2.7-4.2%, respectively. 2.5. Outcome assessment Functional outcome was assessed using the Pittsburgh Cerebral Performance Categories (CPC). The CPC scale ranges from 1 to 5, representing: 1 - good recovery, 2 - moderate disability, 3 - severe disability, 4 - vegetative state and 5 - death.33 Data were collected at discharge from the intensive care unit (ICU), discharge from the medical ward, and at one and six months post-CA. A CPC of 1 or 2 was considered as good outcome and a CPC of 3-5 a poor outcome.11 The functional outcome was counted at the score at 6 months postCA, except for two patients who died after one month, and before 6 months, the best CPC score was used.34 2.6. Statistical analysis The scores were dichotomized into good (CPC 1-2) and poor (CPC 3-5) neurological outcome. The continuous data were not normally distributed, and therefore non-parametric statistics were used. Descriptive statistics were used to present participants’ demographic and medical characteristics. Mann Whitney U-test (continuous variables) and Chi-squared test (categorical variables) were used to compare demographic and medical characteristics between the good and poor group. To test differences between the good and poor group for GFAP, NSE and S100B at each time point, Mann Whitney U-test was used. The incremental
Fig. 2. Boxplots showing levels of glial fibrillary acidic protein (GFAP) (A), neuron-specific enolas (NSE) (B) and S100B (C) at different sampling times plotted on logarithmic scale. The line in the box = median, the outer limits of the box = interquartile range, the whiskers = non-outlier range, outliers are marked by and . Extreme outliers, not visual in the figure, were found in the poor outcome group in GFAP; acute in 2 patients 1.95-436.0 ng/ml, 24 hours in 12 patients 1.35-445.0 ng/ml, 48 hours in 10 patients 1.63-49.0 ng/ml, 72 hours in 4 patients 10.27-51.6 ng/ml, 96 hours in 3 patients 1.23-26.0 ng/ml and at 108 hours 3 patients 1.78-26.2 ng/ml. White represents good outcome and grey poor outcome. *p ≤ 0.05, **p ≤ 0.01. Calculated from peripheral blood samples.
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
1657
Table 1 Demographic, medical characteristics and outcome. All patients n = 125 66 ±14 (18-85)
Age, mean, SD (min-max) Gender, number (%) Male Female Out of hospital, number (%) Witnessed CA, number (%) Bystander CPR, number (%) First registered rhythm, number (%) VF/VT Asystoli PEA Unknown Angiography, number (%) PCI, number (%) Minutes to ROSC, mean, SD (min-max) Minutes from CA to 34 ◦ C, mean, SD (min-max) Time spent in ICU days, mean, (min-max) Medical history, number (%) No previous illness Ischemic heart disease Heart failure Hypertension Lung disease Diabetes Stroke Malignancy CPC at 6 months, number (%) CPC 1 CPC 2 CPC 3 CPC 4 CPC 5
Good outcome n = 57 (46%)
Poor outcome n = 68 (54%)
63 ±14 (23-85)
69 ±12 (18-85)
83 (66) 42 (34) 83 (66) 108 (86)
38 (67) 19 (33) 38 (67) 52 (91)
45 (66) 23 (34) 45 (66) 56 (82)
73 (58)
34 (60)
39 (57)
57 (46) 43 (34) 7 (6) 18 (14) 63 (50) 36 (30) 22 ±15 (5-90) 360 ±185 (30-920) 7 (1-93)
35 (61) 12 (21) 2 (4) 8 (14) 33 (58) 21 (37) 20 ±14 (5-90) 367 ±189 (30-920) 6 (2-23) 16 (28) 18 (32) 14 (25) 24 (42) 10 (18) 7 (12) 1 (2) 1 (2)
6 (9) 21 (31) 16 (24) 38 (56) 15 (22) 22 (32) 12 (18) 6 (9)
48 (38) 8 (6) 5 (4) 64 (51)
48 (84) 8 (14) 1 (2)
5 (7) 63 (93)
p = 0.005 NS
NS NS NS p = 0.009
22 (32) 31 (46) 5 (7) 10 (15) 30 (44) 15 (22) 24 ±15 (5-75) 354 ±182 (56-850) 7 (1-93)
22 (18) 39 (31) 30 (24) 62 (50) 25 (20) 29 (23) 13 (10) 7 (6)
p-value
NS NS NS NS NS p = 0.006 NS NS NS NS p = 0.007 p = 0.004 NS
CA = cardiac arrest, CPR = cardio pulmonary resuscitation, VF = ventricular fibrillation, VT = ventricular tachycardia, PEA = pulseless electric activity, PCI = percutan coronary intervention, ROSC = return of spontaneous circulation, ICU = intensive care unit, CPC = cerebral performance category.
predictive value of GFAP, S100B and NSE was evaluated using logistic regression models and presented with area under the receiver operating characteristic (ROC) curve (AUC). Sensitivity and specificity for specific cut-off values were derived using ROC. The Wilcoxon signed rank test was used for analysis of the difference
between blood samples from peripheral artery or vein and samples from the jugular bulb. P-values
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Clinical Paper
Post-cardiac arrest serum levels of glial fibrillary acidic protein for predicting neurological outcome Ing-Marie Larsson a,∗ , Ewa Wallin a , Marja-Leena Kristofferzon b,c , Marion Niessner d , Henrik Zetterberg e,f , Sten Rubertsson a a
Department of Surgical Sciences–Anaesthesiology & Intensive Care, Uppsala University, SE- 751 85 Uppsala, Sweden Faculty of Health and Occupational Studies, Department of Health and Caring Sciences, University of Gävle, SE- 801 76 Gävle, Sweden c Department of Public Health and Caring Sciences, Uppsala University, SE- 751 22 Uppsala, Sweden d Roche Diagnostics GmbH, Penzberg, Germany e Department of Psychiatry and Neurochemistry, the Sahlgrenska Academy at the University of Gothenburg, Institute of Neuroscience and Physiology, Mölndal, Sweden f UCL Institute of Neurology, Queen Square, London WC1N 3BG, United Kingdom b
a r t i c l e
i n f o
Article history: Received 5 June 2014 Received in revised form 6 August 2014 Accepted 9 September 2014 Keywords: Cardiac arrest Therapeutic hypothermia Prognostication Biomarkers
a b s t r a c t Aim of the study: To investigate serum levels of glial fibrillary acidic protein (GFAP) for evaluation of neurological outcome in cardiac arrest (CA) patients and compare GFAP sensitivity and specificity to that of more studied biomarkers neuron-specific enolas (NSE) and S100B. Method: A prospective observational study was performed in three hospitals in Sweden during 20082012. The participants were 125 CA patients treated with therapeutic hypothermia (TH) to 32-34 ◦ C for 24 hours. Samples were collected from peripheral blood (n = 125) and the jugular bulb (n = 47) up to 108 hours post-CA. GFAP serum levels were quantified using a novel, fully automated immunochemical method. Other biomarkers investigated were NSE and S100B. Neurological outcome was assessed using the Cerebral Performance Categories scale (CPC) and dichotomized into good and poor outcome. Results: GFAP predicted poor neurological outcome with 100% specificity and 14-23% sensitivity at 24, 48 and 72 hours post-CA. The corresponding values for NSE were 27-50% sensitivity and for S100B 2130% sensitivity when specificity was set to 100%. A logistic regression with stepwise combination of the investigated biomarkers, GFAP, did not increase the ability to predict neurological outcome. No differences were found in GFAP, NSE and S100B levels when peripheral and jugular bulb blood samples were compared. Conclusion: Serum GFAP increase in patients with poor outcome but did not show sufficient sensitivity to predict neurological outcome after CA. Both NSE and S100B were shown to be better predictors. The ability to predict neurological outcome did not increased when combining the three biomarkers. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Cardiac arrest (CA) is associated with high mortality rate and among those who survive, there is a major risk of neurological sequelae.1–3 During CA, the brain is exposed to hypoxia, which may cause anoxic brain injury and associated long-term dysfunction.2,4 Treating comatose survivors after CA with therapeutic hypothermia (TH) to 32-34 ◦ C for 12-24 hours has been shown to improve neurological outcome 5,6 and survival,5 and TH
∗ Corresponding author. Department of Surgical Sciences–Anaesthesiology & Intensive Care, Uppsala University, SE-751 85 Uppsala, Sweden. E-mail address: [email protected] (I.-M. Larsson). http://dx.doi.org/10.1016/j.resuscitation.2014.09.007 0300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.
is now recommended as treatment after CA.7,8 A recently published study showed no difference in neurological outcome and survival rate when using TH at 33 ◦ C or 36 ◦ C.9 Prognostication and prediction of neurological outcome are important aspects of care of CA patients. It is recommended that prognostication be postponed until 72 hours after normothermia.10,11 Several prognostication methods have been reported: clinical examination, neurophysiological examination (electroencephalogram (EEG), somatosensory evoked potentials (SSEP)), imaging (magnetic resonance imaging (MRI), computed tomography (CT)) and biomarkers.10–12 Biomarkers can be proteins or peptides released from the brain across the blood-brain barrier (BBB) into the systemic circulation.13 They may provide information about the severity of brain damage.
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
However, there are limitations regarding reliability, such as different cut-off values for predicting poor outcomes, influence of hemolysis in the sample,10,11,13 and the fact that levels may differ across laboratories.14 Of the candidate blood biomarkers, most data are available for neuron-specific enolase (NSE) and S-100B, but other biomarkers have also been investigated.11,13,15–19 NSE is a dimeric enzyme, mainly expressed in neurons, and its half-life is estimated to 24-30 h.13 Red blood cells contain NSE, and therefore, hemolysis affects NSE levels.11,13 Increased NSE concentrations after CA are associated with poor outcome in comatose patients, but there is as yet no established cut-off value for this marker.11,13,20–23 S100B is an intracellular calcium-binding protein with particularly high expression in astroglial cells in the white matter of the brain. S100B has a short half-life, about 2 hours, with its highest levels during the first 24 hours after anoxic brain injury.13 Its serum levels may also be influenced by release from fat and skeletal tissues.24 Different cut-off values for predicting poor neurological outcome after CA have been reported.11,13,15,20,25,26 Glial fibrillary acidic protein (GFAP) is a filament exclusively expressed by mature astrocytes of the central nervous system (CNS).27 GFAP has been shown to be a biomarker of traumatic brain injury (TBI), and GFAP levels are related to the severity of the TBI.28,29 Moreover, elevated serum GFAP levels have been reported in stroke patients.30 The levels of GFAP in CA patients tend to increase in patients with poor neurological outcome.15,31 Since GFAP is a relatively new biomarker in the field of prognostication after CA, the aim of this study was to examine the association of serum GFAP levels, determined using a novel, fully automated immunochemical method, at different time points post-CA with outcome and compare its sensitivity and specificity to NSE and S100B.
1655
2. Materials and methods 2.1. Study population The present investigation was a prospective observational study conducted at one university hospital and two general county hospitals in Sweden. Inclusion criteria were CA treated with TH and age > 18 years. To start TH, successful restoration of spontaneous circulation (ROSC) with systolic arterial blood pressure ≥ 80 mmHg > 5 min and unconscious with a Glasgow Coma Scale (GSC)32 score < 8 had to be achieved. During the study period, May, 2008 to May, 2012, 242 CA patients were admitted to the ICUs, 209 underwent TH, and of them, 125 were included in the study. Sixty-three patients survived until hospital discharge (Fig. 1). 2.2. Hypothermia treatment The decision to start TH was taken for all patients, irrespective of the first registered ECG rhythm or whether the CA occurred in or out of hospital. The patients were cooled to 32-34 ◦ C for 24 hours. Hypothermia treatment was induced by infusion of 4 ◦ C saline, in a planned volume of 30-40 ml/kg. To maintain cooling, all hospitals used external cooling, either ice packs or cooling suits depending on hospital-specific routines. Rewarming at a rate of 0.5 ◦ C/h was used, and 36 ◦ C was considered the normal core temperature. Before cooling, all patients were sedated, intubated and mechanically ventilated according to local ICU guidelines for severely ill patients adapted for TH after CA. 2.3. Data collection Blood samples for biomarkers were collected as soon as possible upon arrival to the ICU and at 24, 48, 72, 96 and 108 h after
Fig. 1. Flow chart over CA patients admitted to ICU.
1656
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
CA from a peripheral artery or vein. In patients who were provided with a catheter in the jugular bulb, blood samples were collected at the same time points. The samples were stored in a -70 ◦ C freezer and analysed at the same time after the study period. Medical background variables, information about the CA and data on temperature management were retrieved from the medical chart. Results of the biomarkers were not available to the caregivers and therefore could not affect treatment or evaluation of the patient. 2.4. Analysis of biomarkers Serum levels of NSE and S-100B were measured using a Cobas e601 instrument and NSE and S-100B reagent kits as described by the manufacturer (Roche Diagnostics, Penzberg, Germany). For detection of hemolysis, serum measurement index of hemolysis (Sindh) was used. Serum GFAP was measured using the non-commercial Elecsys® GFAP prototype test on a Cobas e411 instrument (Roche, Penzberg, Germany). In the first step, biotinand ruthenium-labeled monoclonal anti-GFAP antibodies are combined with 50 l of sample and incubated for 9 minutes. In the second step, streptavidin-coated magnetic microparticles are added and the mixture is incubated for further 9 minutes. After the second incubation step, the reaction mixture is transferred to the measuring cell where the beads are captured on the surface of an electrode by a magnet. Unbound label is removed by washing the measuring cell. In the last step, voltage is applied to the electrode in the presence of a tri-propylamine (TPA)-containing buffer, the resulting electrochemiluminescent signal is recorded by a photomultiplier and the GFAP concentration derived from a calibration curve. Because no acknowledged reference method is available at present, the method has been standardized by weighing pure human GFAP in analyte-free serum matrix. Three internal samples based on human serum matrix spiked with human GFAP low (< 5 ng/ml), medium (20–30 ng/ml) and high (> 50 ng/ml) were used for quality control during the assay. The within- and between-run precisions were 1.1-1.9% and 2.7-4.2%, respectively. 2.5. Outcome assessment Functional outcome was assessed using the Pittsburgh Cerebral Performance Categories (CPC). The CPC scale ranges from 1 to 5, representing: 1 - good recovery, 2 - moderate disability, 3 - severe disability, 4 - vegetative state and 5 - death.33 Data were collected at discharge from the intensive care unit (ICU), discharge from the medical ward, and at one and six months post-CA. A CPC of 1 or 2 was considered as good outcome and a CPC of 3-5 a poor outcome.11 The functional outcome was counted at the score at 6 months postCA, except for two patients who died after one month, and before 6 months, the best CPC score was used.34 2.6. Statistical analysis The scores were dichotomized into good (CPC 1-2) and poor (CPC 3-5) neurological outcome. The continuous data were not normally distributed, and therefore non-parametric statistics were used. Descriptive statistics were used to present participants’ demographic and medical characteristics. Mann Whitney U-test (continuous variables) and Chi-squared test (categorical variables) were used to compare demographic and medical characteristics between the good and poor group. To test differences between the good and poor group for GFAP, NSE and S100B at each time point, Mann Whitney U-test was used. The incremental
Fig. 2. Boxplots showing levels of glial fibrillary acidic protein (GFAP) (A), neuron-specific enolas (NSE) (B) and S100B (C) at different sampling times plotted on logarithmic scale. The line in the box = median, the outer limits of the box = interquartile range, the whiskers = non-outlier range, outliers are marked by and . Extreme outliers, not visual in the figure, were found in the poor outcome group in GFAP; acute in 2 patients 1.95-436.0 ng/ml, 24 hours in 12 patients 1.35-445.0 ng/ml, 48 hours in 10 patients 1.63-49.0 ng/ml, 72 hours in 4 patients 10.27-51.6 ng/ml, 96 hours in 3 patients 1.23-26.0 ng/ml and at 108 hours 3 patients 1.78-26.2 ng/ml. White represents good outcome and grey poor outcome. *p ≤ 0.05, **p ≤ 0.01. Calculated from peripheral blood samples.
I.-M. Larsson et al. / Resuscitation 85 (2014) 1654–1661
1657
Table 1 Demographic, medical characteristics and outcome. All patients n = 125 66 ±14 (18-85)
Age, mean, SD (min-max) Gender, number (%) Male Female Out of hospital, number (%) Witnessed CA, number (%) Bystander CPR, number (%) First registered rhythm, number (%) VF/VT Asystoli PEA Unknown Angiography, number (%) PCI, number (%) Minutes to ROSC, mean, SD (min-max) Minutes from CA to 34 ◦ C, mean, SD (min-max) Time spent in ICU days, mean, (min-max) Medical history, number (%) No previous illness Ischemic heart disease Heart failure Hypertension Lung disease Diabetes Stroke Malignancy CPC at 6 months, number (%) CPC 1 CPC 2 CPC 3 CPC 4 CPC 5
Good outcome n = 57 (46%)
Poor outcome n = 68 (54%)
63 ±14 (23-85)
69 ±12 (18-85)
83 (66) 42 (34) 83 (66) 108 (86)
38 (67) 19 (33) 38 (67) 52 (91)
45 (66) 23 (34) 45 (66) 56 (82)
73 (58)
34 (60)
39 (57)
57 (46) 43 (34) 7 (6) 18 (14) 63 (50) 36 (30) 22 ±15 (5-90) 360 ±185 (30-920) 7 (1-93)
35 (61) 12 (21) 2 (4) 8 (14) 33 (58) 21 (37) 20 ±14 (5-90) 367 ±189 (30-920) 6 (2-23) 16 (28) 18 (32) 14 (25) 24 (42) 10 (18) 7 (12) 1 (2) 1 (2)
6 (9) 21 (31) 16 (24) 38 (56) 15 (22) 22 (32) 12 (18) 6 (9)
48 (38) 8 (6) 5 (4) 64 (51)
48 (84) 8 (14) 1 (2)
5 (7) 63 (93)
p = 0.005 NS
NS NS NS p = 0.009
22 (32) 31 (46) 5 (7) 10 (15) 30 (44) 15 (22) 24 ±15 (5-75) 354 ±182 (56-850) 7 (1-93)
22 (18) 39 (31) 30 (24) 62 (50) 25 (20) 29 (23) 13 (10) 7 (6)
p-value
NS NS NS NS NS p = 0.006 NS NS NS NS p = 0.007 p = 0.004 NS
CA = cardiac arrest, CPR = cardio pulmonary resuscitation, VF = ventricular fibrillation, VT = ventricular tachycardia, PEA = pulseless electric activity, PCI = percutan coronary intervention, ROSC = return of spontaneous circulation, ICU = intensive care unit, CPC = cerebral performance category.
predictive value of GFAP, S100B and NSE was evaluated using logistic regression models and presented with area under the receiver operating characteristic (ROC) curve (AUC). Sensitivity and specificity for specific cut-off values were derived using ROC. The Wilcoxon signed rank test was used for analysis of the difference
between blood samples from peripheral artery or vein and samples from the jugular bulb. P-values
