The Second Elevation of Neuron-Specific Enolase Peak after Ischemic Stroke Is Associated with Hemorrhagic Transformation Bum Joon Kim, MD,1 Yeon-Jung Kim, MD,1 Sung Ho Ahn, MD, Na Young Kim, MD, Dong-Wha Kang, MD, Jong S. Kim, MD, and Sun U. Kwon, MD, PhD
Background: Neuron-specific enolase (NSE) is a surrogate marker for the extent of brain damage after ischemic stroke and affords a good predictor of stroke prognosis. We hypothesized that the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke is dependent on stroke mechanism and is associated with hemorrhagic transformation. Methods: Acute ischemic stroke patients visiting our center within 24 hours of symptom onset were recruited into the study. NSE levels were obtained serially at various time points after stroke, and the pattern of change was categorized into no significant change, continuously increasing, continuously decreasing, with 1 peak and with 2 peaks. Clinical, laboratory, and imaging variables were compared among the patient groups. Multivariate analysis was performed to verify the independent association between the second NSE peak and hemorrhagic transformation after adjusting for potential confounders. Results: Among 83 patients, NSE levels were stationary in 22 (26.5%) of the patients, increased in 9 (10.8%), decreased in 18 (21.7%), and showed 1 peak in 17 (20.5%) and 2 peaks in 17 (20.5%) patients. The incidence of atrial fibrillation and hemorrhagic transformation was significantly elevated (P 5 .02) in patients with 2 NSE peaks. Furthermore, the second NSE peak and the initial lesion volume were associated independently with hemorrhagic transformation after we adjusted for potential confounders (odds ratio 5 6.844 and 1.024, P 5 .04 and .02, respectively). Conclusions: Serial NSE analysis during the acute period of ischemic stroke is useful for monitoring hemorrhagic transformation and the blood-brain barrier disruption status. Key Words: Blood-brain barrier—hemorrhagic transformation—ischemic stroke—neuron-specific enolase—stroke mechanism. Ó 2014 by National Stroke Association
Introduction
From the Department of Neurology, Asan Medical Center, Seoul, Korea. Received April 20, 2014; revision received May 14, 2014; accepted May 22, 2014. Supported by a grant from the Korea Healthcare Technology R&D Project, funded by the Ministry of Health and Welfare of the Republic of Korea (Grant No. HI10C2020). Address correspondence to Sun U. Kwon, MD, PhD, Stroke Center and Department of Neurology, Asan Medical Center, 388-1 Pungnapdong, Songpa-gu, Seoul 138-736, Korea. E-mail: [email protected]. 1 B.J. Kim and Y.J. Kim contributed equally to this manuscript. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05.020
Neuron-specific enolase (NSE) is a cytosolic brain enzyme that functions as a glycolytic isoenzyme.1,2 NSE is confined solely to neurons under normal conditions, and only a negligible amount of NSE is physiologically present in the peripheral blood.3 NSE is released into the systemic circulation after neurologic damage, including ischemic stroke, cerebral hemorrhage, and traumatic or hypoxic brain injury.4-7 Therefore, abrupt increases in NSE levels in the peripheral blood provide a diagnostic and prognostic surrogate marker of brain damage and blood–brain barrier (BBB) dysfunction.8 Major cerebral vessel occlusions cause brain damage. Secondary cerebral injury then occurs after reperfusion or restoration of the blood flow.9 Multiple pathologic
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2437-2443
2437
2438
processes are involved in ischemic/reperfusion injury, encompassing leukocyte infiltration, platelet and complement activation, and finally, breakdown of the BBB.10 BBB breakdown in turn results in hemorrhagic transformation and massive vasogenic edema.11 These consequences increase intracranial pressure and the risk of brain herniation and delay the appropriate use of antithrombotic agents. Therefore, the evaluation of the extent of vasogenic edema and an early diagnosis of hemorrhagic transformation are critical to a favorable stroke outcome. We hypothesized herein that the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke varies according to stroke mechanism and the degree of BBB breakdown and is associated with hemorrhagic transformation. To explore this hypothesis, we investigated changes in NSE levels at various time points after stroke and compared clinical, laboratory, and imaging factors in stroke patients according to NSE trends.
Methods Patients and Study Design Patients with ischemic stroke who were admitted to the Stroke Center of Asan Medical Center (Seoul, Korea) during July 2013 to October 2013 via the emergency department were screened for study suitability. Patients with acute ischemic stroke who presented with a high signal intensity lesion on diffusion-weighted images (DWIs) and arrived at the emergency department within 24 hours of symptom onset were admitted to the study. Exclusion criteria consisted of any of the following: (1) receipt of intravenous tissue plasminogen activator or any intraarterial revascularization procedure, (2) initial demonstration of hemorrhagic transformation, or (3) admission to the stroke center on weekends or holidays, when laboratory measurement of NSE was unavailable. Demographic and clinical variables, including the conventional risk factors for stroke (ie, hypertension, diabetes, hyperlipidemia, atrial fibrillation, and previous history of stroke or coronary artery disease) were obtained for each patient. The neurologic deficit at admission and at discharge or transfer was graded on the basis of the National Institutes of Health Stroke Scale (NIHSS) score. The subtype of ischemic stroke was categorized according to the Trial of Org 10172 in Acute Stroke Treatment classification. The results of routine serologic tests performed on the day of admission (complete blood cell counts, fasting glucose levels, lipid profiles, renal function testing, and hemoglobin A1c and C-reactive protein levels) also were obtained. The initial lesion volume of the acute ischemic stroke was manually measured on the DWI. Follow-up magnetic resonance imaging (MRI) and magnetic resonance angiography were performed in selected patients (see Evaluation of Hemorrhagic Transformation for the criteria
B.J. KIM ET AL.
for follow-up MRI) to evaluate the presence of recanalization or hemorrhagic transformation at 4-7 days after the initial MRI was performed. NSE levels in the peripheral blood were assessed serially during the first week after admission; however, if the patient was discharged before 1 week elapsed, NSE levels were instead measured until the day of discharge. This study was approved by the Institutional Review Board of Asan Medical Center; however, written informed consent was not obtained because of the retrospective design of the study.
Serial Measurement and Analysis of NSE Levels Most of the NSE levels in the peripheral blood were first measured on the day after patient admission. Every day at the same time, usually in the morning, venous blood (2 mL) was collected. The blood sample was then refrigerated at 2-8 C before laboratory analysis by immunoradiometric assay. Samples showing hemolysis were excluded from analysis. Serial changes in the NSE levels were classified into the following5 patterns: (1) no significant change, (2) continuous increase, (3) continuous decrease, (4) 1 peak, and (5) 2 peaks. A significant change was defined as a change in NSE content that was greater than 50% of the baseline NSE content or the former NSE level. If no clear difference was revealed during the study period, the pattern was described as ‘‘no significant change’’ (Fig 1A). If NSE level increased by more than 50% from baseline with no significant decrease, the pattern was described as a ‘‘continuous increase’’ (Fig 1B). On the other hand, if NSE levels decreased by more than 50% from baseline with no significant increase, the pattern was regarded as a ‘‘continuous decrease’’ (Fig 1C). If a significant decrease was observed after a significant increase, the pattern was defined as a ‘‘1-peak’’ pattern (Fig 1D), whereas if a second significant increase followed the first peak, or a significant increase followed a significant decrease, the pattern was defined as a ‘‘2-peak’’ pattern (Fig 1E). Figure 1F shows each of the 5 patterns for the NSE trends, showing the average NSE value at each time point for all of the affected patients within each category.
Evaluation of Hemorrhagic Transformation Patients presenting with (1) an initial occlusion of the major vessel, (2) signs of early neurologic deterioration (more than 2-point increase of NIHSS score during 5 days after stroke),12 (3) sudden headache or unexplained vomiting, or (4) uncertain etiology of ischemic stroke received a follow-up MRI. The follow-up MRI included DWI, gradient echo imaging, and magnetic resonance angiography. The appearance of any new dark signal on gradient echo imaging within the initial DWI lesion was regarded as hemorrhagic transformation. Hemorrhagic transformation was classified according to
NSE AND HEMORRHAGIC TRANSFORMATION
2439
Figure 1. Serial changes in NSE levels in the peripheral blood according to the classified patterns of NSE level changes. The patterns shown are as follows: (A) no significant change, (B) continuous increase, (C) continuous decrease, (D) 1 peak, and (E) 2 peaks. (F) Each of the five patterns is shown, showing average NSE values at each time point for all of the affected patients within each category. Abbreviation: NSE, neuron-specific enolase.
the (European Cooperative Acute Stroke Study I criteria as follows: hemorrhagic infarction (HI)-1, small petechiae; HI-2, larger, more confluent petechiae; parenchymal hemorrhage (PH)-1, mild space-occupying effect, comprising ,30% of infarcted area; and PH-2, significant space-occupying effect, comprising .30% of the infarcted area).13
Statistical Analysis The demographic, clinical, laboratory, and imaging variables were first compared among the patients grouped according to the pattern of NSE level changes. Student’s t-test, the chi-square test, and analysis of variance were performed as appropriate for categorical and numerical variables. The association between the trend of NSE level changes and the subtype of ischemic stroke was then analyzed. Finally, variables with potential association from the univariate analysis (P , .10) were entered into a multivariate analysis to evaluate factors independently associated with hemorrhagic transformation. Changes in NSE levels were entered as dichotomized data (ie, patients with a second NSE peak versus patients without a second NSE peak). In each case, a P-value of ,.05 was considered statistically significant. All statistical analyses were performed by using SPSS for Windows, version 17.0 (SPSS Inc., Chicago, IL).
Results Patient Characteristics One hundred twenty-two patients with acute ischemic stroke initially were admitted to the stroke center via the emergency department within 24 hours of symptom onset. Of these, 90 patients were admitted during regular working days, when NSE level measurement was available. Six patients who received thrombolysis
and 1 patient with intra-arterial thrombectomy were excluded. Finally, 83 patients were enrolled in the study. Of these patients, 22/83 (26.5%) demonstrated no significant change in NSE levels throughout the course of the study, 9 of 83 (10.8%) showed a continuous increase, 18 of 83 (21.7%) showed a continuous decrease, 17 of 83 (20.5%) showed 1 peak, and another 17 of 83 (20.5%) showed 2 peaks (Fig 1). Follow-up MRI was performed in 39 of 83 patients (47.0%) for the purpose of evaluating the status of initially occluded cerebral vessel (n 5 36; including 8 patients with multiple reasons of follow-up MRI), after early neurologic deterioration (n 5 6), after sudden headache or unexplained vomiting (n 5 3) or for the evaluation of uncertain etiology of stroke (n 5 2). Among these 39 individuals, hemorrhagic transformation was observed in 13 of 39 (33.3%).
Comparison of Demographic and Risk Factors among Patient Groups According to the Pattern of NSE Level Changes No differences were discerned in terms of demographic or risk factors among the patients with different NSE trends, except that atrial fibrillation was observed more frequently in patients with a 2-peak pattern (Table 1). In addition, the occurrence of cardioembolic stroke was relatively high in the 2-peak pattern patient group (Table 2). Although no statistically significant differences were detected between the patient groups in initial neurologic severity as assessed by the NIHSS score or the initial DWI lesion volume, the mean NSE level and its standard deviation were both greater in patients presenting with 2 NSE peaks versus the other 4 patterns during the acute stage of ischemic stroke (Table 1). Finally, hemorrhagic transformation of the ischemic lesion was most frequently observed in patients with a second NSE peak during the acute stage (Table 1).
B.J. KIM ET AL.
2440
Table 1. Comparison of patient variables according to the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke No change (n 5 22) Age, y Male Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous history History of CAD Initial NIHSS Discharge NIHSS Mean NSE, ng/mL SD NSE, ng/mL WBC, /mm3 Hb, g/dL Platelets, 3103/mm3 Glucose, mg/dL Hb A1c, % BUN, mg/dL Creatinine, mg/dL CRP, mg/dL Total cholesterol, mg/dL LDL cholesterol, mg/dL HDL cholesterol, mg/dL Initial DWI lesion volume, mm3 Hemorrhagic transformation
71.5 (9.1) 15 (68.2) 15 (68.2) 6 (27.3) 11 (50.0) 2 (9.1) 2 (9.1) 1 (4.5) 6.4 (7.6) 4.1 (7.9) 9.9 (4.1) 1.7 (1.0) 8486 (3560) 13.2 (2.4) 213 (88) 134 (51) 6.3 (.7) 19.6 (11.8) 1.1 (.8) .7 (.8) 185 (46) 116 (34) 47 (19) 8 (16) 0 of 10 (.0)
Increase (n 5 9)
Decrease (n 5 18)
One peak (n 5 17)
Two peaks (n 5 17)
P value
61.0 (17.2) 8 (88.9) 5 (55.6) 3 (33.3) 7 (77.8) 0 (.0) 1 (11.1) 0 (.0) 6.3 (4.2) 3.2 (2.4) 9.7 (2.2) 3.4 (1.4) 8800 (3431) 14.3 (1.9) 237 (38) 145 (60) 6.5 (.8) 16.8 (6.0) 1.00 (.84) 1.1 (1.3) 158 (71) 116 (26) 46 (7) 11 (15)
70.4 (10.6) 10 (55.6) 15 (83.3) 3 (16.7) 8 (44.4) 5 (27.8) 7 (38.9) 1 (5.6) 5.3 (4.8) 3.2 (3.8) 10.6 (4.4) 4.5 (4.0) 7072 (1636) 13.3 (1.6) 228 (55) 121 (34) 6.5 (1.3) 14.9 (4.9) .82 (.23) 2.2 (4.4) 176 (43) 108 (35) 49 (13) 22 (65)
65.7 (8.5) 13 (76.5) 11 (64.7) 5 (29.5) 9 (52.9) 6 (35.3) 4 (23.5) 1 (5.9) 7.2 (6.0) 4.7 (6.0) 16.7 (5.4) 6.4 (5.8) 8471 (3143) 13.6 (2.5) 206 (75) 130 (35) 6.2 (.8) 18.7 (6.5) .93 (.26) 1.9 (4.4) 169 (45) 93 (35) 57 (40) 35 (45)
67.7 (8.5) 11 (64.7) 9 (52.9) 2 (11.8) 7 (41.2) 8 (47.1) 1 (5.9) 0 (.0) 6.9 (6.9) 4.3 (6.0) 18.7 (10.6) 6.6 (4.2) 8029 (3063) 13.6 (2.2) 217 (55) 124 (39) 6.0 (.5) 19.8 (20.3) 1.38 (2.3) .4 (.9) 180 (37) 111 (37) 50 (16) 39 (55)
.24 .44 .38 .59 .47 .02 .07 .83 .91 .95 ,.001 ,.001 .53 .73 .79 .65 .46 .71 .68 .30 .59 .28 .69 .20
1 of 3 (33.3)
1 of 5 (20.0)
4 of 11 (36.4)
7 of 10 (70.0)
.02
Abbreviations: BUN, blood urea nitrogen; CAD, coronary artery disease; CRP, C-reactive protein; DWI, diffusion-weighted image; Hb, hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; NSE, neuronspecific enolase; SD, standard deviation; WBC, white blood cell. Results are given as the number of occurrences (%) or the mean (SD).
Hemorrhagic Transformation Type According to the Pattern of NSE Level Changes Hemorrhagic transformation was not observed in patients without a significant change in NSE levels during the acute stage of stroke (Fig 2). Among the 11 patients showing 1 peak pattern, 1 of 11 (9.1%), 2 of 11 (18.2%), and 1 of 11 (9.1%) demonstrated HI-1–, HI-2–, and
PH-1–type hemorrhagic transformation, respectively. Of the 10 patients with a second NSE peak, 1 of 10 patients (10%) presented with PH-1–type hemorrhagic transformation, and another 1 of 10 (10%) presented with PH-2– type hemorrhagic transformation. Finally, HI-1– and HI2–type hemorrhagic transformation were observed in 3 of 10 (30%) and 2 of 10 individuals (20%), respectively
Table 2. Association between the pattern of NSE level changes and stroke mechanism
Large artery atherosclerosis Small-vessel occlusion Cardioembolism Other determined Undetermined
No change (n 5 22)
Increase (n 5 9)
Decrease (n 5 18)
One peak (n 5 17)
Two peaks (n 5 17)
12 (42.9) 5 (29.4) 2 (7.4) 3 (30.0) 0
6 (21.4) 1 (5.9) 0 (.0) 2 (20.0) 0
4 (14.3) 8 (47.1) 6 (22.2) 0 (.0) 0
5 (17.9) 1 (5.9) 7 (25.9) 4 (40.0) 0
1 (3.6) 2 (11.8) 12 (44.4) 1 (10.0) 1 (100)
Abbreviation: NSE, neuron-specific enolase. Results are given as the number of occurrences (%).
P ,.001
NSE AND HEMORRHAGIC TRANSFORMATION
2441
Discussion
Figure 2. Hemorrhagic transformation type according to the pattern of NSE level changes during the acute period of ischemic stroke. Abbreviations: NSE, neuron-specific enolase; HI, hemorrhagic infarction; HT, hemorrhagic transformation; PH, parenchymal hemorrhage.
(Fig 2). Among the 7 patients with second NSE peak and hemorrhagic transformation, 3 patients demonstrated worsening of neurologic symptom.
Pattern of NSE Level Changes According to Stroke Mechanism The pattern of NSE changes during the acute period of ischemic stroke differed according to the stroke subtype (P , .001; Table 2). In the large artery atherosclerosis group, NSE levels showed no significant change in 12 of 28 patients (42.9%) and increased continuously in another 6 of 28 patients (21.4%). By contrast, NSE levels decreased continuously in 8 of 17 patients (47.1%) in the small vessel occlusion group. Patients with ischemic stroke caused by cardioembolism demonstrated a second peak in nearly half (12/27, 44.4%) of the affected patients (Table 2).
Independent Predictors of Hemorrhagic Transformation The results of the univariate analysis demonstrated that atrial fibrillation (odds ratio [OR] 5 4.900, 95% confidence interval [CI] 5 1.135-21.162, P 5 .03), a large initial DWI lesion volume (OR 5 1.026, 95% CI 5 1.007-1.046, P 5 .008), and the presence of a second NSE peak (OR 5 8.944, 95% CI 5 1.763-45.366, P 5 .008) were all positively associated with hemorrhagic transformation. After adjusting for potential confounders, a large initial DWI lesion volume (OR 5 1.024, 95% CI 5 1.003-1.045, P 5 .02) and a second NSE peak (OR 5 6.844, 95% CI 5 1.123-41.700, P 5 .04) were both independently associated with hemorrhagic transformation during the acute period of ischemic stroke.
The present study compared the clinical, laboratory, and imaging variables among 83 patients grouped according to serial NSE level trends during the acute phase of ischemic stroke (Table 1, Fig 1). We found that the pattern of NSE changes during the acute ischemic period differed according to the stroke mechanism (Table 2). In particular, patients with a second NSE peak presented more frequently with a greater incidence of ischemic stroke caused by cardioembolism (Table 2). The 2-peak pattern patients also showed a greater incidence of hemorrhagic transformation (Table 1). The half-life of NSE under physiological conditions is 48 hours14 and therefore, serum NSE levels continue to increase for a few days after the initial ischemic event. The maximum NSE levels usually are observed within the first 96 hours after ischemic stroke, and in some cases, the maximum NSE levels are not observed until 6 days after stroke.1 In the current investigation, NSE trends during the acute period after stroke were diverse. Because large artery disease causes an infarction with a penumbra around the damaged tissue, the infarction has the capacity to enlarge during the acute phase, continuously increasing or at least maintaining the initial level of NSE. By contrast, ischemic lesions caused by small vessel occlusion demonstrated a high incidence of continuously decreasing NSE levels. Interestingly, almost half of the patients with cardioembolic stroke showed a second NSE peak after the first peak within 1 week of stroke onset (Table 2). Because of the disposition of blood clots, an occlusion of the major vessel often spontaneously recanalizes during cardioembolic stroke; however, delayed recanalization of cardioembolic infarction is futile and is in fact an independent predictor of hemorrhagic transformation.15 Considering that peripheral blood NSE levels show good correlation with cerebrospinal fluid/serum albumin quotients and are therefore indicative of BBB permeability, the second NSE peak in cardioembolic stroke patients may be associated with reperfusion injury, including BBB breakdown.8 The absolute value of the peripheral blood NSE concentration is closely related to the extent of brain damage, as well as to stroke prognosis.16,17 Although we did not find a significant association between the initial DWI lesion volume and the pattern of NSE level changes, the incidence of hemorrhagic transformation was significantly associated with NSE trends during the acute period of ischemic stroke (Table 1). In addition, the patterns of NSE level changes were independently associated with hemorrhagic transformation after we adjusted for the initial DWI lesion volume (Table 3). Therefore, regardless of the degree of brain injury, the second NSE peak is undeniably linked with hemorrhagic transformation. Considering that hemorrhagic
B.J. KIM ET AL.
2442
Table 3. Independent predictors for hemorrhagic transformation
Age Male Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous history Initial DWI lesion volume WBC Glucose CRP Second NSE peak
OR (95% CI)
P
0.980 (.932-1.032) 0.847 (.213-3.363) 0.536 (.139-2.059) 0.606 (.104-3.527) 0.079-1.337 4.900 (1.135-21.162) .764 (.127-4.596) 1.026 (1.007-1.046) .944 (.740-1.204) 1.008 (.992-1.025) 1.131 (.829-1.543) 8.944 (1.763-45.366)
.45 .81 .36 .61 .12 .03y .77 .008 .64 .31 .44 .008
Adjusted OR (95% CI)
P*
-
-
1.024 (1.003-1.045)
.02
6.844 (1.123-41.700)
.04
Abbreviations: CI, confidence interval; CRP, C-reactive protein; DWI, diffusion-weighted image; NSE, neuron-specific enolase; OR, odds ratio; WBC, white blood cell. *The P-values represent the results of the multivariate analysis (adjusted P-value). yThe P-value of ,.10 from the univariate analysis was entered into the multivariate analysis.
transformation is the direct result of BBB breakdown, the association between hemorrhagic transformation and the second NSE peak may indicate that elevated NSE levels in the peripheral blood are at least partly influenced by NSE leakage after BBB breakdown. Our study has noteworthy limitations stemming from both its small sample size and its retrospective nature. In particular, the exact mechanism of second NSE peak formation is not clear. Although the second peak was clearly associated with hemorrhagic transformation in this study, it may have occurred secondary to ischemic/ reperfusion brain damage, as a result of BBB breakdown, or both. Second, as the follow-up MRI was only performed in selective group, a selection bias may exist. Furthermore, the predictive value of the second NSE
peak was not sufficiently shown by our observations. For example, among 7 patients with hemorrhagic transformation and a second peak, 6 individuals demonstrated the second NSE peak and hemorrhagic transformation on the follow-up MRI, which was performed on the same day (Fig 3). Finally, only 3 patients demonstrated symptomatic hemorrhagic transformation. Although all the 3 patients demonstrated a second peak of NSE, still the small sample size limits the clinical implications of the present results. A prospective study with larger number of patients may help to verify these issues. In addition, in the future prospective study, evaluating the effect of controlling the modifiable risk factors of hemorrhagic transformation in patients with second NSE peak may extend the clinical application of our preliminary results.
Figure 3. Serum NSE level, timing of hemorrhagic transformation detection and neurologic worsening. Abbreviations: NSE, neuron-specific enolase; HT, hemorrhagic transformation.
NSE AND HEMORRHAGIC TRANSFORMATION
In conclusion, the current investigation showed that the pattern of serial NSE changes during the acute stage of ischemic stroke differed according to stroke mechanism. In addition, a second peak in NSE levels during the acute stroke phase was significantly associated with hemorrhagic transformation.
References 1. Missler U, Wiesmann M, Friedrich C, et al. S-100 protein and neuron-specific enolase concentrations in blood as indicators of infarction volume and prognosis in acute ischemic stroke. Stroke 1997;28:1956-1960. 2. Laskowitz DT, Grocott H, Hsia A, et al. Serum markers of cerebral ischemia. J Stroke Cerebrovasc Dis 1998;7:234-241. 3. Schoerkhuber W, Kittler H, Sterz F, et al. Time course of serum neuron-specific enolase. A predictor of neurological outcome in patients resuscitated from cardiac arrest. Stroke 1999;30:1598-1603. 4. Grubb NR, Simpson C, Sherwood RA, et al. Prediction of cognitive dysfunction after resuscitation from out-ofhospital cardiac arrest using serum neuron-specific enolase and protein S-100. Heart 2007;93:1268-1273. 5. Oertel M, Schumacher U, McArthur DL, et al. S-100B and NSE: markers of initial impact of subarachnoid haemorrhage and their relation to vasospasm and outcome. J Clin Neurosci 2006;13:834-840. 6. Celtik C, Acunas B, Oner N, et al. Neuron-specific enolase as a marker of the severity and outcome of hypoxic ischemic encephalopathy. Brain Dev 2004;26:398-402. 7. Einav S, Kaufman N, Algur N, et al. Modeling serum biomarkers S100 beta and neuron-specific enolase as predictors of outcome after out-of-hospital cardiac arrest: an aid to clinical decision making. J Am Coll Cardiol 2012; 60:304-311.
2443 8. Correale J, Rabinowicz AL, Heck CN, et al. Status epilepticus increases CSF levels of neuron-specific enolase and alters the blood-brain barrier. Neurology 1998; 50:1388-1391. 9. Jickling GC, Liu D, Stamova B, et al. Hemorrhagic transformation after ischemic stroke in animals and humans. J Cereb Blood Flow Metab 2014;34:185-199. 10. Lyden PD. Hemorrhagic transformation during thrombolytic therapy and reperfusion: effects of age, blood pressure, and matrix metalloproteinases. J Stroke Cerebrovasc Dis 2013;22:532-538. 11. Alvarez-Sabin J, Maisterra O, Santamarina E, et al. Factors influencing haemorrhagic transformation in ischaemic stroke. Lancet Neurol 2013;12:689-705. 12. Kwan J, Hand P. Early neurological deterioration in acute stroke: clinical characteristics and impact on outcome. QJM 2006;99:625-633. 13. Fiorelli M, Bastianello S, von Kummer R, et al. Hemorrhagic transformation within 36 hours of a cerebral infarct: relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort. Stroke 1999; 30:2280-2284. 14. Ishiguro Y, Kato K, Ito T, et al. Nervous system-specific enolase in serum as a marker for neuroblastoma. Pediatrics 1983;72:696-700. 15. Molina CA, Montaner J, Abilleira S, et al. Timing of spontaneous recanalization and risk of hemorrhagic transformation in acute cardioembolic stroke. Stroke 2001; 32:1079-1084. 16. Schaarschmidt H, Prange HW, Reiber H. Neuron-specific enolase concentrations in blood as a prognostic parameter in cerebrovascular diseases. Stroke 1994;25:558-565. 17. Butterworth RJ, Wassif WS, Sherwood RA, et al. Serum neuron-specific enolase, carnosinase, and their ratio in acute stroke. An enzymatic test for predicting outcome? Stroke 1996;27:2064-2068.
Background: Neuron-specific enolase (NSE) is a surrogate marker for the extent of brain damage after ischemic stroke and affords a good predictor of stroke prognosis. We hypothesized that the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke is dependent on stroke mechanism and is associated with hemorrhagic transformation. Methods: Acute ischemic stroke patients visiting our center within 24 hours of symptom onset were recruited into the study. NSE levels were obtained serially at various time points after stroke, and the pattern of change was categorized into no significant change, continuously increasing, continuously decreasing, with 1 peak and with 2 peaks. Clinical, laboratory, and imaging variables were compared among the patient groups. Multivariate analysis was performed to verify the independent association between the second NSE peak and hemorrhagic transformation after adjusting for potential confounders. Results: Among 83 patients, NSE levels were stationary in 22 (26.5%) of the patients, increased in 9 (10.8%), decreased in 18 (21.7%), and showed 1 peak in 17 (20.5%) and 2 peaks in 17 (20.5%) patients. The incidence of atrial fibrillation and hemorrhagic transformation was significantly elevated (P 5 .02) in patients with 2 NSE peaks. Furthermore, the second NSE peak and the initial lesion volume were associated independently with hemorrhagic transformation after we adjusted for potential confounders (odds ratio 5 6.844 and 1.024, P 5 .04 and .02, respectively). Conclusions: Serial NSE analysis during the acute period of ischemic stroke is useful for monitoring hemorrhagic transformation and the blood-brain barrier disruption status. Key Words: Blood-brain barrier—hemorrhagic transformation—ischemic stroke—neuron-specific enolase—stroke mechanism. Ó 2014 by National Stroke Association
Introduction
From the Department of Neurology, Asan Medical Center, Seoul, Korea. Received April 20, 2014; revision received May 14, 2014; accepted May 22, 2014. Supported by a grant from the Korea Healthcare Technology R&D Project, funded by the Ministry of Health and Welfare of the Republic of Korea (Grant No. HI10C2020). Address correspondence to Sun U. Kwon, MD, PhD, Stroke Center and Department of Neurology, Asan Medical Center, 388-1 Pungnapdong, Songpa-gu, Seoul 138-736, Korea. E-mail: [email protected]. 1 B.J. Kim and Y.J. Kim contributed equally to this manuscript. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.05.020
Neuron-specific enolase (NSE) is a cytosolic brain enzyme that functions as a glycolytic isoenzyme.1,2 NSE is confined solely to neurons under normal conditions, and only a negligible amount of NSE is physiologically present in the peripheral blood.3 NSE is released into the systemic circulation after neurologic damage, including ischemic stroke, cerebral hemorrhage, and traumatic or hypoxic brain injury.4-7 Therefore, abrupt increases in NSE levels in the peripheral blood provide a diagnostic and prognostic surrogate marker of brain damage and blood–brain barrier (BBB) dysfunction.8 Major cerebral vessel occlusions cause brain damage. Secondary cerebral injury then occurs after reperfusion or restoration of the blood flow.9 Multiple pathologic
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 9 (October), 2014: pp 2437-2443
2437
2438
processes are involved in ischemic/reperfusion injury, encompassing leukocyte infiltration, platelet and complement activation, and finally, breakdown of the BBB.10 BBB breakdown in turn results in hemorrhagic transformation and massive vasogenic edema.11 These consequences increase intracranial pressure and the risk of brain herniation and delay the appropriate use of antithrombotic agents. Therefore, the evaluation of the extent of vasogenic edema and an early diagnosis of hemorrhagic transformation are critical to a favorable stroke outcome. We hypothesized herein that the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke varies according to stroke mechanism and the degree of BBB breakdown and is associated with hemorrhagic transformation. To explore this hypothesis, we investigated changes in NSE levels at various time points after stroke and compared clinical, laboratory, and imaging factors in stroke patients according to NSE trends.
Methods Patients and Study Design Patients with ischemic stroke who were admitted to the Stroke Center of Asan Medical Center (Seoul, Korea) during July 2013 to October 2013 via the emergency department were screened for study suitability. Patients with acute ischemic stroke who presented with a high signal intensity lesion on diffusion-weighted images (DWIs) and arrived at the emergency department within 24 hours of symptom onset were admitted to the study. Exclusion criteria consisted of any of the following: (1) receipt of intravenous tissue plasminogen activator or any intraarterial revascularization procedure, (2) initial demonstration of hemorrhagic transformation, or (3) admission to the stroke center on weekends or holidays, when laboratory measurement of NSE was unavailable. Demographic and clinical variables, including the conventional risk factors for stroke (ie, hypertension, diabetes, hyperlipidemia, atrial fibrillation, and previous history of stroke or coronary artery disease) were obtained for each patient. The neurologic deficit at admission and at discharge or transfer was graded on the basis of the National Institutes of Health Stroke Scale (NIHSS) score. The subtype of ischemic stroke was categorized according to the Trial of Org 10172 in Acute Stroke Treatment classification. The results of routine serologic tests performed on the day of admission (complete blood cell counts, fasting glucose levels, lipid profiles, renal function testing, and hemoglobin A1c and C-reactive protein levels) also were obtained. The initial lesion volume of the acute ischemic stroke was manually measured on the DWI. Follow-up magnetic resonance imaging (MRI) and magnetic resonance angiography were performed in selected patients (see Evaluation of Hemorrhagic Transformation for the criteria
B.J. KIM ET AL.
for follow-up MRI) to evaluate the presence of recanalization or hemorrhagic transformation at 4-7 days after the initial MRI was performed. NSE levels in the peripheral blood were assessed serially during the first week after admission; however, if the patient was discharged before 1 week elapsed, NSE levels were instead measured until the day of discharge. This study was approved by the Institutional Review Board of Asan Medical Center; however, written informed consent was not obtained because of the retrospective design of the study.
Serial Measurement and Analysis of NSE Levels Most of the NSE levels in the peripheral blood were first measured on the day after patient admission. Every day at the same time, usually in the morning, venous blood (2 mL) was collected. The blood sample was then refrigerated at 2-8 C before laboratory analysis by immunoradiometric assay. Samples showing hemolysis were excluded from analysis. Serial changes in the NSE levels were classified into the following5 patterns: (1) no significant change, (2) continuous increase, (3) continuous decrease, (4) 1 peak, and (5) 2 peaks. A significant change was defined as a change in NSE content that was greater than 50% of the baseline NSE content or the former NSE level. If no clear difference was revealed during the study period, the pattern was described as ‘‘no significant change’’ (Fig 1A). If NSE level increased by more than 50% from baseline with no significant decrease, the pattern was described as a ‘‘continuous increase’’ (Fig 1B). On the other hand, if NSE levels decreased by more than 50% from baseline with no significant increase, the pattern was regarded as a ‘‘continuous decrease’’ (Fig 1C). If a significant decrease was observed after a significant increase, the pattern was defined as a ‘‘1-peak’’ pattern (Fig 1D), whereas if a second significant increase followed the first peak, or a significant increase followed a significant decrease, the pattern was defined as a ‘‘2-peak’’ pattern (Fig 1E). Figure 1F shows each of the 5 patterns for the NSE trends, showing the average NSE value at each time point for all of the affected patients within each category.
Evaluation of Hemorrhagic Transformation Patients presenting with (1) an initial occlusion of the major vessel, (2) signs of early neurologic deterioration (more than 2-point increase of NIHSS score during 5 days after stroke),12 (3) sudden headache or unexplained vomiting, or (4) uncertain etiology of ischemic stroke received a follow-up MRI. The follow-up MRI included DWI, gradient echo imaging, and magnetic resonance angiography. The appearance of any new dark signal on gradient echo imaging within the initial DWI lesion was regarded as hemorrhagic transformation. Hemorrhagic transformation was classified according to
NSE AND HEMORRHAGIC TRANSFORMATION
2439
Figure 1. Serial changes in NSE levels in the peripheral blood according to the classified patterns of NSE level changes. The patterns shown are as follows: (A) no significant change, (B) continuous increase, (C) continuous decrease, (D) 1 peak, and (E) 2 peaks. (F) Each of the five patterns is shown, showing average NSE values at each time point for all of the affected patients within each category. Abbreviation: NSE, neuron-specific enolase.
the (European Cooperative Acute Stroke Study I criteria as follows: hemorrhagic infarction (HI)-1, small petechiae; HI-2, larger, more confluent petechiae; parenchymal hemorrhage (PH)-1, mild space-occupying effect, comprising ,30% of infarcted area; and PH-2, significant space-occupying effect, comprising .30% of the infarcted area).13
Statistical Analysis The demographic, clinical, laboratory, and imaging variables were first compared among the patients grouped according to the pattern of NSE level changes. Student’s t-test, the chi-square test, and analysis of variance were performed as appropriate for categorical and numerical variables. The association between the trend of NSE level changes and the subtype of ischemic stroke was then analyzed. Finally, variables with potential association from the univariate analysis (P , .10) were entered into a multivariate analysis to evaluate factors independently associated with hemorrhagic transformation. Changes in NSE levels were entered as dichotomized data (ie, patients with a second NSE peak versus patients without a second NSE peak). In each case, a P-value of ,.05 was considered statistically significant. All statistical analyses were performed by using SPSS for Windows, version 17.0 (SPSS Inc., Chicago, IL).
Results Patient Characteristics One hundred twenty-two patients with acute ischemic stroke initially were admitted to the stroke center via the emergency department within 24 hours of symptom onset. Of these, 90 patients were admitted during regular working days, when NSE level measurement was available. Six patients who received thrombolysis
and 1 patient with intra-arterial thrombectomy were excluded. Finally, 83 patients were enrolled in the study. Of these patients, 22/83 (26.5%) demonstrated no significant change in NSE levels throughout the course of the study, 9 of 83 (10.8%) showed a continuous increase, 18 of 83 (21.7%) showed a continuous decrease, 17 of 83 (20.5%) showed 1 peak, and another 17 of 83 (20.5%) showed 2 peaks (Fig 1). Follow-up MRI was performed in 39 of 83 patients (47.0%) for the purpose of evaluating the status of initially occluded cerebral vessel (n 5 36; including 8 patients with multiple reasons of follow-up MRI), after early neurologic deterioration (n 5 6), after sudden headache or unexplained vomiting (n 5 3) or for the evaluation of uncertain etiology of stroke (n 5 2). Among these 39 individuals, hemorrhagic transformation was observed in 13 of 39 (33.3%).
Comparison of Demographic and Risk Factors among Patient Groups According to the Pattern of NSE Level Changes No differences were discerned in terms of demographic or risk factors among the patients with different NSE trends, except that atrial fibrillation was observed more frequently in patients with a 2-peak pattern (Table 1). In addition, the occurrence of cardioembolic stroke was relatively high in the 2-peak pattern patient group (Table 2). Although no statistically significant differences were detected between the patient groups in initial neurologic severity as assessed by the NIHSS score or the initial DWI lesion volume, the mean NSE level and its standard deviation were both greater in patients presenting with 2 NSE peaks versus the other 4 patterns during the acute stage of ischemic stroke (Table 1). Finally, hemorrhagic transformation of the ischemic lesion was most frequently observed in patients with a second NSE peak during the acute stage (Table 1).
B.J. KIM ET AL.
2440
Table 1. Comparison of patient variables according to the pattern of NSE level changes in the peripheral blood during the acute period of ischemic stroke No change (n 5 22) Age, y Male Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous history History of CAD Initial NIHSS Discharge NIHSS Mean NSE, ng/mL SD NSE, ng/mL WBC, /mm3 Hb, g/dL Platelets, 3103/mm3 Glucose, mg/dL Hb A1c, % BUN, mg/dL Creatinine, mg/dL CRP, mg/dL Total cholesterol, mg/dL LDL cholesterol, mg/dL HDL cholesterol, mg/dL Initial DWI lesion volume, mm3 Hemorrhagic transformation
71.5 (9.1) 15 (68.2) 15 (68.2) 6 (27.3) 11 (50.0) 2 (9.1) 2 (9.1) 1 (4.5) 6.4 (7.6) 4.1 (7.9) 9.9 (4.1) 1.7 (1.0) 8486 (3560) 13.2 (2.4) 213 (88) 134 (51) 6.3 (.7) 19.6 (11.8) 1.1 (.8) .7 (.8) 185 (46) 116 (34) 47 (19) 8 (16) 0 of 10 (.0)
Increase (n 5 9)
Decrease (n 5 18)
One peak (n 5 17)
Two peaks (n 5 17)
P value
61.0 (17.2) 8 (88.9) 5 (55.6) 3 (33.3) 7 (77.8) 0 (.0) 1 (11.1) 0 (.0) 6.3 (4.2) 3.2 (2.4) 9.7 (2.2) 3.4 (1.4) 8800 (3431) 14.3 (1.9) 237 (38) 145 (60) 6.5 (.8) 16.8 (6.0) 1.00 (.84) 1.1 (1.3) 158 (71) 116 (26) 46 (7) 11 (15)
70.4 (10.6) 10 (55.6) 15 (83.3) 3 (16.7) 8 (44.4) 5 (27.8) 7 (38.9) 1 (5.6) 5.3 (4.8) 3.2 (3.8) 10.6 (4.4) 4.5 (4.0) 7072 (1636) 13.3 (1.6) 228 (55) 121 (34) 6.5 (1.3) 14.9 (4.9) .82 (.23) 2.2 (4.4) 176 (43) 108 (35) 49 (13) 22 (65)
65.7 (8.5) 13 (76.5) 11 (64.7) 5 (29.5) 9 (52.9) 6 (35.3) 4 (23.5) 1 (5.9) 7.2 (6.0) 4.7 (6.0) 16.7 (5.4) 6.4 (5.8) 8471 (3143) 13.6 (2.5) 206 (75) 130 (35) 6.2 (.8) 18.7 (6.5) .93 (.26) 1.9 (4.4) 169 (45) 93 (35) 57 (40) 35 (45)
67.7 (8.5) 11 (64.7) 9 (52.9) 2 (11.8) 7 (41.2) 8 (47.1) 1 (5.9) 0 (.0) 6.9 (6.9) 4.3 (6.0) 18.7 (10.6) 6.6 (4.2) 8029 (3063) 13.6 (2.2) 217 (55) 124 (39) 6.0 (.5) 19.8 (20.3) 1.38 (2.3) .4 (.9) 180 (37) 111 (37) 50 (16) 39 (55)
.24 .44 .38 .59 .47 .02 .07 .83 .91 .95 ,.001 ,.001 .53 .73 .79 .65 .46 .71 .68 .30 .59 .28 .69 .20
1 of 3 (33.3)
1 of 5 (20.0)
4 of 11 (36.4)
7 of 10 (70.0)
.02
Abbreviations: BUN, blood urea nitrogen; CAD, coronary artery disease; CRP, C-reactive protein; DWI, diffusion-weighted image; Hb, hemoglobin; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NIHSS, National Institutes of Health Stroke Scale; NSE, neuronspecific enolase; SD, standard deviation; WBC, white blood cell. Results are given as the number of occurrences (%) or the mean (SD).
Hemorrhagic Transformation Type According to the Pattern of NSE Level Changes Hemorrhagic transformation was not observed in patients without a significant change in NSE levels during the acute stage of stroke (Fig 2). Among the 11 patients showing 1 peak pattern, 1 of 11 (9.1%), 2 of 11 (18.2%), and 1 of 11 (9.1%) demonstrated HI-1–, HI-2–, and
PH-1–type hemorrhagic transformation, respectively. Of the 10 patients with a second NSE peak, 1 of 10 patients (10%) presented with PH-1–type hemorrhagic transformation, and another 1 of 10 (10%) presented with PH-2– type hemorrhagic transformation. Finally, HI-1– and HI2–type hemorrhagic transformation were observed in 3 of 10 (30%) and 2 of 10 individuals (20%), respectively
Table 2. Association between the pattern of NSE level changes and stroke mechanism
Large artery atherosclerosis Small-vessel occlusion Cardioembolism Other determined Undetermined
No change (n 5 22)
Increase (n 5 9)
Decrease (n 5 18)
One peak (n 5 17)
Two peaks (n 5 17)
12 (42.9) 5 (29.4) 2 (7.4) 3 (30.0) 0
6 (21.4) 1 (5.9) 0 (.0) 2 (20.0) 0
4 (14.3) 8 (47.1) 6 (22.2) 0 (.0) 0
5 (17.9) 1 (5.9) 7 (25.9) 4 (40.0) 0
1 (3.6) 2 (11.8) 12 (44.4) 1 (10.0) 1 (100)
Abbreviation: NSE, neuron-specific enolase. Results are given as the number of occurrences (%).
P ,.001
NSE AND HEMORRHAGIC TRANSFORMATION
2441
Discussion
Figure 2. Hemorrhagic transformation type according to the pattern of NSE level changes during the acute period of ischemic stroke. Abbreviations: NSE, neuron-specific enolase; HI, hemorrhagic infarction; HT, hemorrhagic transformation; PH, parenchymal hemorrhage.
(Fig 2). Among the 7 patients with second NSE peak and hemorrhagic transformation, 3 patients demonstrated worsening of neurologic symptom.
Pattern of NSE Level Changes According to Stroke Mechanism The pattern of NSE changes during the acute period of ischemic stroke differed according to the stroke subtype (P , .001; Table 2). In the large artery atherosclerosis group, NSE levels showed no significant change in 12 of 28 patients (42.9%) and increased continuously in another 6 of 28 patients (21.4%). By contrast, NSE levels decreased continuously in 8 of 17 patients (47.1%) in the small vessel occlusion group. Patients with ischemic stroke caused by cardioembolism demonstrated a second peak in nearly half (12/27, 44.4%) of the affected patients (Table 2).
Independent Predictors of Hemorrhagic Transformation The results of the univariate analysis demonstrated that atrial fibrillation (odds ratio [OR] 5 4.900, 95% confidence interval [CI] 5 1.135-21.162, P 5 .03), a large initial DWI lesion volume (OR 5 1.026, 95% CI 5 1.007-1.046, P 5 .008), and the presence of a second NSE peak (OR 5 8.944, 95% CI 5 1.763-45.366, P 5 .008) were all positively associated with hemorrhagic transformation. After adjusting for potential confounders, a large initial DWI lesion volume (OR 5 1.024, 95% CI 5 1.003-1.045, P 5 .02) and a second NSE peak (OR 5 6.844, 95% CI 5 1.123-41.700, P 5 .04) were both independently associated with hemorrhagic transformation during the acute period of ischemic stroke.
The present study compared the clinical, laboratory, and imaging variables among 83 patients grouped according to serial NSE level trends during the acute phase of ischemic stroke (Table 1, Fig 1). We found that the pattern of NSE changes during the acute ischemic period differed according to the stroke mechanism (Table 2). In particular, patients with a second NSE peak presented more frequently with a greater incidence of ischemic stroke caused by cardioembolism (Table 2). The 2-peak pattern patients also showed a greater incidence of hemorrhagic transformation (Table 1). The half-life of NSE under physiological conditions is 48 hours14 and therefore, serum NSE levels continue to increase for a few days after the initial ischemic event. The maximum NSE levels usually are observed within the first 96 hours after ischemic stroke, and in some cases, the maximum NSE levels are not observed until 6 days after stroke.1 In the current investigation, NSE trends during the acute period after stroke were diverse. Because large artery disease causes an infarction with a penumbra around the damaged tissue, the infarction has the capacity to enlarge during the acute phase, continuously increasing or at least maintaining the initial level of NSE. By contrast, ischemic lesions caused by small vessel occlusion demonstrated a high incidence of continuously decreasing NSE levels. Interestingly, almost half of the patients with cardioembolic stroke showed a second NSE peak after the first peak within 1 week of stroke onset (Table 2). Because of the disposition of blood clots, an occlusion of the major vessel often spontaneously recanalizes during cardioembolic stroke; however, delayed recanalization of cardioembolic infarction is futile and is in fact an independent predictor of hemorrhagic transformation.15 Considering that peripheral blood NSE levels show good correlation with cerebrospinal fluid/serum albumin quotients and are therefore indicative of BBB permeability, the second NSE peak in cardioembolic stroke patients may be associated with reperfusion injury, including BBB breakdown.8 The absolute value of the peripheral blood NSE concentration is closely related to the extent of brain damage, as well as to stroke prognosis.16,17 Although we did not find a significant association between the initial DWI lesion volume and the pattern of NSE level changes, the incidence of hemorrhagic transformation was significantly associated with NSE trends during the acute period of ischemic stroke (Table 1). In addition, the patterns of NSE level changes were independently associated with hemorrhagic transformation after we adjusted for the initial DWI lesion volume (Table 3). Therefore, regardless of the degree of brain injury, the second NSE peak is undeniably linked with hemorrhagic transformation. Considering that hemorrhagic
B.J. KIM ET AL.
2442
Table 3. Independent predictors for hemorrhagic transformation
Age Male Hypertension Diabetes Hyperlipidemia Atrial fibrillation Previous history Initial DWI lesion volume WBC Glucose CRP Second NSE peak
OR (95% CI)
P
0.980 (.932-1.032) 0.847 (.213-3.363) 0.536 (.139-2.059) 0.606 (.104-3.527) 0.079-1.337 4.900 (1.135-21.162) .764 (.127-4.596) 1.026 (1.007-1.046) .944 (.740-1.204) 1.008 (.992-1.025) 1.131 (.829-1.543) 8.944 (1.763-45.366)
.45 .81 .36 .61 .12 .03y .77 .008 .64 .31 .44 .008
Adjusted OR (95% CI)
P*
-
-
1.024 (1.003-1.045)
.02
6.844 (1.123-41.700)
.04
Abbreviations: CI, confidence interval; CRP, C-reactive protein; DWI, diffusion-weighted image; NSE, neuron-specific enolase; OR, odds ratio; WBC, white blood cell. *The P-values represent the results of the multivariate analysis (adjusted P-value). yThe P-value of ,.10 from the univariate analysis was entered into the multivariate analysis.
transformation is the direct result of BBB breakdown, the association between hemorrhagic transformation and the second NSE peak may indicate that elevated NSE levels in the peripheral blood are at least partly influenced by NSE leakage after BBB breakdown. Our study has noteworthy limitations stemming from both its small sample size and its retrospective nature. In particular, the exact mechanism of second NSE peak formation is not clear. Although the second peak was clearly associated with hemorrhagic transformation in this study, it may have occurred secondary to ischemic/ reperfusion brain damage, as a result of BBB breakdown, or both. Second, as the follow-up MRI was only performed in selective group, a selection bias may exist. Furthermore, the predictive value of the second NSE
peak was not sufficiently shown by our observations. For example, among 7 patients with hemorrhagic transformation and a second peak, 6 individuals demonstrated the second NSE peak and hemorrhagic transformation on the follow-up MRI, which was performed on the same day (Fig 3). Finally, only 3 patients demonstrated symptomatic hemorrhagic transformation. Although all the 3 patients demonstrated a second peak of NSE, still the small sample size limits the clinical implications of the present results. A prospective study with larger number of patients may help to verify these issues. In addition, in the future prospective study, evaluating the effect of controlling the modifiable risk factors of hemorrhagic transformation in patients with second NSE peak may extend the clinical application of our preliminary results.
Figure 3. Serum NSE level, timing of hemorrhagic transformation detection and neurologic worsening. Abbreviations: NSE, neuron-specific enolase; HT, hemorrhagic transformation.
NSE AND HEMORRHAGIC TRANSFORMATION
In conclusion, the current investigation showed that the pattern of serial NSE changes during the acute stage of ischemic stroke differed according to stroke mechanism. In addition, a second peak in NSE levels during the acute stroke phase was significantly associated with hemorrhagic transformation.
References 1. Missler U, Wiesmann M, Friedrich C, et al. S-100 protein and neuron-specific enolase concentrations in blood as indicators of infarction volume and prognosis in acute ischemic stroke. Stroke 1997;28:1956-1960. 2. Laskowitz DT, Grocott H, Hsia A, et al. Serum markers of cerebral ischemia. J Stroke Cerebrovasc Dis 1998;7:234-241. 3. Schoerkhuber W, Kittler H, Sterz F, et al. Time course of serum neuron-specific enolase. A predictor of neurological outcome in patients resuscitated from cardiac arrest. Stroke 1999;30:1598-1603. 4. Grubb NR, Simpson C, Sherwood RA, et al. Prediction of cognitive dysfunction after resuscitation from out-ofhospital cardiac arrest using serum neuron-specific enolase and protein S-100. Heart 2007;93:1268-1273. 5. Oertel M, Schumacher U, McArthur DL, et al. S-100B and NSE: markers of initial impact of subarachnoid haemorrhage and their relation to vasospasm and outcome. J Clin Neurosci 2006;13:834-840. 6. Celtik C, Acunas B, Oner N, et al. Neuron-specific enolase as a marker of the severity and outcome of hypoxic ischemic encephalopathy. Brain Dev 2004;26:398-402. 7. Einav S, Kaufman N, Algur N, et al. Modeling serum biomarkers S100 beta and neuron-specific enolase as predictors of outcome after out-of-hospital cardiac arrest: an aid to clinical decision making. J Am Coll Cardiol 2012; 60:304-311.
2443 8. Correale J, Rabinowicz AL, Heck CN, et al. Status epilepticus increases CSF levels of neuron-specific enolase and alters the blood-brain barrier. Neurology 1998; 50:1388-1391. 9. Jickling GC, Liu D, Stamova B, et al. Hemorrhagic transformation after ischemic stroke in animals and humans. J Cereb Blood Flow Metab 2014;34:185-199. 10. Lyden PD. Hemorrhagic transformation during thrombolytic therapy and reperfusion: effects of age, blood pressure, and matrix metalloproteinases. J Stroke Cerebrovasc Dis 2013;22:532-538. 11. Alvarez-Sabin J, Maisterra O, Santamarina E, et al. Factors influencing haemorrhagic transformation in ischaemic stroke. Lancet Neurol 2013;12:689-705. 12. Kwan J, Hand P. Early neurological deterioration in acute stroke: clinical characteristics and impact on outcome. QJM 2006;99:625-633. 13. Fiorelli M, Bastianello S, von Kummer R, et al. Hemorrhagic transformation within 36 hours of a cerebral infarct: relationships with early clinical deterioration and 3-month outcome in the European Cooperative Acute Stroke Study I (ECASS I) cohort. Stroke 1999; 30:2280-2284. 14. Ishiguro Y, Kato K, Ito T, et al. Nervous system-specific enolase in serum as a marker for neuroblastoma. Pediatrics 1983;72:696-700. 15. Molina CA, Montaner J, Abilleira S, et al. Timing of spontaneous recanalization and risk of hemorrhagic transformation in acute cardioembolic stroke. Stroke 2001; 32:1079-1084. 16. Schaarschmidt H, Prange HW, Reiber H. Neuron-specific enolase concentrations in blood as a prognostic parameter in cerebrovascular diseases. Stroke 1994;25:558-565. 17. Butterworth RJ, Wassif WS, Sherwood RA, et al. Serum neuron-specific enolase, carnosinase, and their ratio in acute stroke. An enzymatic test for predicting outcome? Stroke 1996;27:2064-2068.
