S-100 Protein and Neuron-Specific Enolase Concentrations in Blood as Indicators of Infarction Volume and Prognosis in Acute Ischemic Stroke
Background and Purpose Better techniques are needed to monitor infarction volume and predict neurological outcome after ischemic brain infarction. We evaluated the usefulness of serial measurements of S-100 protein versus neuron-specific enolase (NSE) in blood samples from patients with acute stroke.
Methods Using nonisotopic sandwich immunoassays, we measured plasma concentrations of S-100 protein and NSE on admission and on days 3, 4, 7, and 14 after infarction in 44 patients (age range, 22 to 86 years; mean age, 65.1 years; 12 female, 32 male). Infarct volume was measured by volumetric CT on day 4 after ictus, and clinical outcome was assessed at discharge from hospital with the Activities of Daily Living Scale and 6 months after infarction with the Glasgow Outcome Scale.
Results Peak blood levels of S-100 protein were found on day 2.5±1.3, and peak levels of NSE were found on day 1.9±0.8 after infarction. Peak plasma levels of S-100 protein correlated well with infarct volume (r=.75, P<.001) and with clinical outcome assessed with the Glasgow Outcome Scale (r=.51, P<.001). Serum levels of NSE correlated with infarct volume (r=.37, P<.05) but not with clinical outcome (r=.18, P>.05).
Conclusions The results of our study indicate that measuring blood concentrations of S-100 protein periodically in the first 10 days after cerebral infarction helps to predict infarct volume and the long-term neurological outcome more accurately than periodic measurements of blood concentrations of NSE.
In recent years many techniques have been investigated for their usefulness in monitoring the patient’s neurological status and predicting the outcome of therapy for ischemic brain disease. Neurological examinations are helpful when neurological function is largely intact but are of little value in assessing infarct volume or in patients who are comatose after cerebral infarction. Modern neuroradiological imaging techniques such as CT, MRI, and ultrasound help clinicians identify the location and volume of an infarct and thus plan treatment, such as intravenous or intra-arterial administration of fibrinolytic agents and neuroprotective drugs to halt tissue damage. However, repeating neuroradiological imaging—usually daily—is impractical.
Several monitoring techniques have been developed based on measuring levels of various proteins, including neuron-specific enolase (NSE), myelin basic protein, glial fibrillary acidic protein, and S-100 protein. However, in most studies1 2 3 4 5 6 7 8 9 10 these substances were measured in CSF only, and collecting CSF samples by lumbar puncture is only indicated when the patient has a small infarction and no brain edema. Even in these cases, daily sampling of CSF is difficult and associated with high risk of complications in patients being treated with heparin. A technique to measure blood levels of protein markers of brain injury could allow frequent testing at relatively low risk and thus be useful for monitoring the course of disease.
We developed nonradioisotopic solid-phase immunoassays for quantitative determination of S-100 protein and NSE in CSF and blood.11 In the study we report here, we evaluated whether serial measurements of blood levels of S-100 protein and NSE are useful for predicting the extent of brain damage and neurological outcome after ischemic stroke.
Subjects and Methods
Patients and Control Subjects
The subjects for this study included 44 patients admitted to the hospital between February and October 1995 by faculty of the Department of Neurology of our institution for evaluation and management of acute ischemic brain infarction. The 12 female and 32 male patients ranged in age between 22 and 86 years (mean age, 65.1 years), and every patient who participated in the study underwent neurological examination and CT of the brain on admission. Patients with documented or clinical evidence of brain infarction, hemorrhage, head trauma, or central nervous system infection within the 3 months before admission or with a central nervous system tumor were excluded from this study.
This study was approved by the local Research Ethics Committee, and written consent to participate in the study was obtained from each patient or, when a patient was unconscious or confused, from the patient’s relatives.
Control subjects included 120 healthy blood donors (60 male and 60 female; age range, 18 to 65 years; mean age, 37.6±13.1 years) whose blood samples were used to determine reference values for concentrations of S-100 protein and 98 healthy blood donors (50 male and 48 female; age range, 25 to 58 years; mean age, 39.3±11.7 years) who provided blood samples for reference measurement of NSE concentration.
Measurement of Infarct Volume
Four days (mean, 4.0 ± 2.2 days) after infarction, patients in this study underwent CT scanning with measurement of infarct volume with the use of the volumetry program of the Siemens Somatom Plus S CT scanner. The accuracy of the system was tested by imaging a “phantom” consisting of two balloons filled with contrast medium (Ultravist, Schering) diluted in water and placed in a skull that was then placed under water before scanning.
Blood Sample Collection
The first blood sample was obtained from all 44 patients immediately after admission for acute stroke. Admission occurred within 24 hours after infarction in 41 patients, within 48 hours in after infarction in 2 patients, and within 72 hours after infarction in 1 patient. Therefore, the first blood sample was obtained from 41 patients on day 0, from 2 patients on day 1, and from 1 patient on day 2. Subsequent blood samples were collected on days 3, 4, 7, and 14 after acute infarction.
In 8 patients blood samples were obtained daily for 9 days after infarction. Within 6 hours of collection, all blood samples were centrifuged and aliquots of plasma were frozen and stored at −80°C until analysis for S-100 protein and NSE concentrations.
Evaluation of Neurological Outcomes
On patients’ discharge from the hospital, their neurological function was assessed with the ADL scale.12 The long-term outcome of cerebral infarction was assessed at patients’ 6-month follow-up visit with the GOS.13
S-100 Protein Analysis
The concentration of S-100 protein in blood samples was measured as described earlier.11 In brief, all measurements were set up in duplicate. Microtiter plates coated with 10 μL of anti–S-100 β chain (Sigma) in 20 mL phosphate buffer (0.05 mol/L, pH 8.6) were incubated with 200 μL per well of S-100 calibrators, controls, and samples for 120 minutes. Biotin-labeled rabbit anti–S-100 antibody (DAKO) in a Tris 0.05 mol/L, NaCl 0.15 mol/L, CaCl2 10 mmol/L, NaN3 0.15 mmol/L buffer was added, and plates were incubated for another hour. After the plates had been washed, 200 μL of streptavidin-europium in assay buffer (Tris 0.05 mol/L, NaCl 0.15 mol/L, bovine serum albumin 1 g/L, bovine gamma globulin 0.5 g/L, both from Sigma, NaN3 0.15 mmol/L) was added to each well, and the plates were incubated for 30 minutes. As a last step, 200 μL of enhancement solution (acetic acid 0.01 mol/L, tri-n-octyl phosphine oxide 38 mg/L, potassium phtalate 1.3 g/L, thenoyltrifluoroacetone 222 mg/L, Triton X-100 2 mL/L) was added to each well, and the plates were incubated for 15 minutes. The resulting fluorescence was measured with a DELFIA 1232 fluorometer (Wallac).
Microtiter plates coated with 10 μL rabbit anti-NSE (DAKO) in 20 mL carbonate buffer (0.05 mol/L, pH 9.6) were incubated with 20 μL per well of NSE calibrators, controls (both from the Hoffmann–La Roche NSE enzyme immunoassay kit), samples, and 200 μL per well assay buffer for 2 hours. After the plates had been washed, 200 μL per well of biotin-labeled rabbit anti-NSE antibody (DAKO) in assay buffer was added, and the plates were incubated for another 2 hours. The plates were then incubated with streptavidin and enhancement solution as described for the S-100 assay. Assay results were validated by testing 96 patient samples with the use of the NSE enzyme immunoassay kit available commercially from Hoffmann–La Roche.
All results are reported as mean±SD unless stated otherwise. The S-100 and NSE peak levels of each individual were used for statistical analysis. In addition, AUC values were calculated for S-100 and NSE concentrations. Two patients died within the first 4 days after infarction and were therefore excluded from the AUC calculation. Also excluded from AUC calculations were the 3 patients who were admitted later than 24 hours after infarction. Regression analyses were performed and correlations were calculated by the Spearman rank method. Significance of differences between groups was calculated by the Wilcoxon signed rank method. A value of P<.05 was considered significant.
Infarctions were infratentorial in 12 of the 44 patients and supratentorial in the remaining 32. Overall, 12% of patients died (GOS 1), and 41% recovered normal neurological function (GOS 5). Mild disability was present in 11% 6 months after cerebral infarction (GOS 4), 36% had severe disability (GOS 3), and none were in a vegetative state (GOS 2).
Infarct Volume and Correlation With Clinical Outcome
All measurements of infarct volumes were performed by the same neuroradiologist. Six infarctions had a volume of less than 5 mL, 12 had volumes between 5 and 20 mL, 6 had volumes between 21 and 100 mL, and 11 had volumes greater than 100 mL.
In the procedure for validation of CT volumetry results, the smaller of the two contrast medium–filled balloons had a volume of 34.6 mL and the larger had a volume of 107.5 mL (measured after CT volumetry by weighing the balloon). Balloon volumes determined by CT volumetry were 35.6 mL (+2.6%) and 118.6 mL (+10.3%). Intrarater variation for CT volumetry measurements, determined by averaging the results of 10 measurements of the volume of each balloon, was 2.2% for the small balloon and 0.54% for the large balloon (CV) (n=10).
Greater infarct volume correlated with worse clinical outcome both at discharge (infarct volume/ADL score: r=.56, P<.001) and at long-term (6-month) follow-up (infarct volume/GOS score: r=.66, P<.001).
The lower detection limit of the S-100 assay was 0.015 μg/L (0+3 SD, n=24) of S-100 protein. The intra-assay (within-run) variability (CV) was 3.2% at 0.51 μg/L, 2.1% at 5.97 μg/L, and 2.3% at 11.4 μg/L of S-100 protein (n=20), and the between-day variability (CV) was 11.5% at 0.45 μg/L, 7.9% at 4.79 μg/L, and 7.8% at 15.45 μg/L of S-100 protein (n=21). The assay predominantly detected S-100b (detection ratio for S-100b/S-100a=14:1). The reference value for S-100b in blood was 0.069±0.058 μg/L.
The lower detection limit of the NSE assay was 1 μg/L (0+3 SD, n=21). The intra-assay (within-run) variability (CV) was 2.99% at 10.3 μg/L (n=16), 4.7% at 83.7 μg/L (n=21), and 2.0% at 184 μg/L (n=20) NSE. The reference value for NSE in blood was 11.1±4.7 μg/L (n=98). The results obtained were almost identical to the values measured using the Hoffmann–La Roche NSE enzyme immunoassay kit (r=.99, P<.001, n=96).
Analysis of S-100 protein and NSE levels in blood samples from healthy control subjects showed no relationship of blood levels to age or sex.
Times to Peak S-100 Protein and NSE Levels After Infarction
In 8 patients we measured S-100 protein level daily for 9 days, and the mean concentrations in this group over time are shown in Fig 1A⇓. Peak S-100 protein plasma levels in these patients were measured 2.5±1.3 days after infarction. In the majority of patients plasma levels returned to normal within 9 days after the event.
Maximal serum concentrations of NSE were observed 1.9±0.8 days after infarction (Fig 1B⇑). However, there were no statistically significant differences between NSE levels on days 1, 2, and 3 (P>.05).
Mean Levels of S-100 and NSE After Cerebral Infarction
Blood levels of S-100 protein and NSE were significantly higher in patients who had suffered cerebral infarction than in healthy control subjects. Data are shown in the Table⇓.
Correlation Between S-100 Protein or NSE and Infarct Volume
The highest concentration of S-100 protein recorded in this study (4.1 μg/L) was observed in a patient with a supratentorial infarct that had a volume of 328 mL. Higher peak S-100 protein levels correlated positively with larger infarct volume (r=.75, P<.001) (Fig 2⇓). NSE serum levels were less clearly correlated with infarct volume (r=.37, P<.05).
Analysis of S-100 protein and NSE AUC calculations showed correlations of r=.67 (P<.001) (for S-100 AUC calculation/infarct volume) and r=.48 (P<.01) (for NSE AUC calculation/infarct volume).
Correlation Between S-100 Protein or NSE and Clinical Outcome
Higher peak S-100 plasma levels correlated with worse clinical outcome both at discharge (S-100 peak concentration/ADL: r=.44, P<.01) and at long-term follow-up (S-100 peak concentration/GOS: r=.49, P<.01). However, we did not find significant correlation between peak NSE serum levels and clinical outcome (NSE peak concentration/ADL: r=.24, P>.05; NSE peak concentration/GOS: r=.18, P>.05). Calculations of AUC concentrations showed correlations of r=.36 (P<.05) for S-100 AUC calculation/ADL; r=.39 (P<.05) for S-100 AUC calculation/GOS; r=.13 (P>.05) for NSE AUC calculation/ADL; and r=.11 (P>.05) for NSE AUC calculation/GOS.
The term “S-100” refers to a mixture of dimeric proteins consisting of two subunits of Mr 10 500 termed α and β. Three isoforms are known. S-100a (αβ) is found in glial cells and melanocytes, and S-100b (ββ) is present in high concentrations in glial cells and Schwann cells of the central and peripheral nervous systems as well as in Langerhans cells and cells of the anterior pituitary. S-100a0 (αα), which represents 5% of the S-100 protein in the brain, is found outside the nervous system in slow-twitch muscle, heart, and kidney. The S-100 protein family constitutes a subgroup of Ca2+-binding proteins of the EF-hand type. Both intracellular and extracellular mechanisms of action have been proposed for S-100 protein, although its biological functions are not yet understood in detail.14 15 16
NSE is the γγ-isoenzyme of the glycolytic enzyme enolase. NSE is predominantly present in neurons and neuroendocrine cells.17 It has, however, also been found in several other tissues and serves as a tumor marker of oat cell carcinoma of the lung.18
Relationship of Protein Markers to Infarct Volume
Clinically, increased levels of NSE have been found soon after ischemic stroke in CSF and blood,2 3 4 5 6 7 8 10 19 20 21 22 and elevated concentrations of S-100 protein have been reported in CSF after ischemic stroke.1 4 9 21 The results of several animal studies indicate a semiquantitative relationship between elevated NSE and S-100 protein levels in CSF and larger infarct volume.4 5 6 10 There are also reports in the literature that higher NSE levels in serum correlate with larger infarct volume.3 8 19 20
Persson et al21 were the first to report studying S-100 protein in the blood of patients with stroke, noting elevated levels in two patients. Kim et al23 detected S-100 protein in the blood of only 7 of 19 patients with cerebral infarction and in none of their healthy control subjects. In contrast, we detected S-100 protein in 43 of 44 patients and in 119 of 120 healthy control subjects. These different results may be attributable to different lower detection limits of the analytical methods used to measure S-100 protein in blood.
Peak Blood Levels of Brain-Specific Proteins
Compared with control subjects, patients who had suffered acute cerebral infarction had clearly elevated blood levels of S-100 and NSE. Because our analysis of samples confirmed our literature review findings of no relationship between blood values of these proteins and subject age or sex, control subjects have not been matched with stroke patients. Our review of the literature as well as our own study results also provided no evidence that other risk factors for stroke such as diabetes mellitus or hypertension might influence the blood levels of S-100 or NSE, although at present this cannot be ruled out completely.
In most reported cases, peak levels of NSE in serum were found within the first 96 hours after cerebral infarction, although in some cases the peak levels occurred as late as day 6.3 19 20 22 The half-life of NSE in serum has been reported to be about 48 hours,24 and therefore blood levels of NSE would be expected to rise as long as infarct damage continued and NSE was washing out of brain tissue. In fact, Cunningham et al19 found that the peak level of NSE occurs later with increasing infarct size and concluded that the peak level of NSE best reflects final infarct volume. The time to peak serum level of NSE in our study was 1.9±0.8 days after infarction, which compares well with the 48-hour average reported in the literature, and we did note a trend for the peak level to occur later with increasing infarct size.
Little is known about how levels of S-100 protein in blood change over time after cerebral infarction. Kim et al23 measured S-100 protein in serum on days 1, 3, and 7 after infarction and found peak values after 3 days. The finding in our study that the concentration of S-100 protein in blood peaks 2.5±1.3 days after infarction is consistent with these results.
Correlation of Blood Concentrations of Proteins and Infarct Volume
Cunningham et al19 found a correlation between peak NSE serum levels and infarct volume (r=.41, P<.05), and our results for NSE were similar (r=.37, P<.05).
However, our data show that the blood level of S-100 protein, when measured by a sensitive method, is a better indicator of the extent of cerebral infarction (r=.75, P<.001). The results were similar for the two methods of evaluation (measurement of peak concentration and calculation of AUC concentration).
Correlation of Blood Concentrations of Proteins and Neurological Outcome
Butterworth et al22 studied the ratio of NSE to carnosinase in 124 patients with ischemic or hemorrhagic stroke and found a statistically significant correlation between this ratio and the outcome after ischemic stroke as assessed with the modified Barthel Index (r=.34, P<.001) and the modified Rankin scale (r=.30, P<.01). However, we could not find any reports in the literature of a correlation between NSE level alone and clinical outcome of cerebral infarction, and serum levels of NSE did not correlate with clinical outcome in our study.
In contrast, we found good correlation between peak S-100 protein levels in plasma and clinical outcome after 6 months as assessed with the GOS (r=.51, P<.001). To our knowledge this is the first study to report an association between a single biochemical marker measured from blood in the acute phase of stroke and functional outcome of infarction.
In conclusion, our data suggest that the concentration of S-100 protein in blood during acute stroke is a useful marker of infarct size and of long-term clinical outcome. Because this biochemical marker is not specific for cerebral infarction but indicates any cell damage in the central nervous system, elevated levels are not diagnostic of acute cerebral infarction. This marker might also be useful in charting the clinical course of other diseases of the central nervous system if the time course of changes in S-100 protein concentration can be correlated significantly with clinical outcomes. It should be considered, however, that the time course of these changes after acute cerebral infarction may differ from the time courses of changes after other brain insults. Measurements of S-100 protein concentration in blood might also be useful in monitoring the effects of new therapies for acute cerebral diseases.
Selected Abbreviations and Acronyms
|ADL||=||activities of daily living|
|AUC||=||area under the curve|
|CV||=||coefficient of variation|
|GOS||=||Glasgow Outcome Scale|
We thank Norman Kock for excellent technical assistance and the physicians and nurses of the Neurologic Department of Lübeck Medical University for their support.
- Received June 9, 1997.
- Revision received July 17, 1997.
- Accepted July 18, 1997.
- Copyright © 1997 by American Heart Association
Aurell A, Rosengren LE, Karlsson B, Olsson JE, Zbornikova V, Haglid KG. Determination of S-100 and glial fibrillary acidic protein concentrations in cerebrospinal fluid after brain infarction. Stroke. 1991;22:1254-1258.
Hårdemark HG, Persson L, Bolander HG, Hillered L, Olsson Y, Påhlman S. Neuron-specific enolase is a marker of cerebral ischemia and infarct size in rat cerebrospinal fluid. Stroke. 1988;19:1140-1144.
Hay E, Royds JA, Davies-Jones GAB, Lewtas NA, Timperley WR, Taylor CB. Cerebrospinal fluid enolase in stroke. J Neurol Neurosurg Psychiatry. 1984;47:724-729.
Kawasaki H, Wakayama Y, Okayasu H, Takahashi H, Shibuya S. Levels of serum and cerebrospinal fluid enolase in patients with cerebral vascular disease and other neurological diseases. Stroke. 1988;19:313-318.
Millikan CH. A classification and outline of cerebrovascular disease. Stroke. 1975;6:565-616.
Baudier J, Glasser N, Gérard D. Ions binding to S100 proteins. J Biol Chem.. 1986;261:8192-8203.
Schaarschmidt H, Prange HW, Reiber HO. Neuron-specific enolase concentrations in blood as a prognostic parameter in cerebrovascular diseases. Stroke. 1994;25:558-565.
Persson L, Hårdemark HG, Gustafsson J, Rundström G, Mendel-Hartvig I, Esscher T, Påhlman S. S-100 protein and neuron-specific enolase in cerebrospinal fluid and serum: markers of cell damage in human central nervous system. Stroke. 1987;18:911-918.
Butterworth RJ, Wassif WS, Sherwood RA, Gerges A, Poyser KH, Garthwaite J, Peters TJ, Bath PMW. Serum neuron-specific enolase, carnosinase, and their ratio in acute stroke: an enzymatic test for predicting outcome? Stroke. 1996;27:2064-2068.
Kim JS, Yoon SS, Kim YH, Ryu JS. Serial measurement of interleukin-6, transforming growth factor-β, and S-100 protein in patients with acute stroke. Stroke. 1996;27:1553-1557.
Ishiguro Y, Kato K, Ito T, Nagaya M, Yamada N, Sugito T. Nervous system-specific enolase in serum as a marker for neuroblastoma. Pediatrics. 1983;72:696-700.