The Relationship Between Serum Neuron-Specific Enolase Levels and Severity of Bleeding and Functional Outcomes in Patients With Nontraumatic Subarachnoid Hemorrhage BACKGROUND: The value of neuron-specific enolase (NSE) in predicting clinical outcomes has been investigated in a variety of neurological disorders. OBJECTIVE: To investigate the associations of serum NSE with severity of bleeding and functional outcomes in patients with subarachnoid hemorrhage (SAH). METHODS: We retrospectively reviewed the records of patients with SAH from June 2008 to June 2012. The severity of SAH bleeding at admission was measured radiographically with the Fisher scale and clinically with the Glasgow Coma Scale, Hunt and Hess grade, and World Federation of Neurologic Surgeons scale. Outcomes were assessed with the modified Rankin Scale at discharge. RESULTS: We identified 309 patients with nontraumatic SAH, and 71 had NSE testing. Median age was 54 years (range, 23-87 years), and 44% were male. In multivariable analysis, increased NSE was associated with a poorer Hunt and Hess grade (P = .003), World Federation of Neurologic Surgeons scale score (P , .001), and Glasgow Coma Scale score (P = .003) and worse outcomes (modified Rankin Scale at discharge; P = .001). There was no significant association between NSE level and Fisher grade (P = .81) in multivariable analysis. CONCLUSION: We found a significant association between higher NSE levels and poorer clinical presentations and worse outcomes. Although it is still early for any relevant clinical conclusions, our results suggest that NSE holds promise as a tool for screening patients at increased risk of poor outcomes after SAH. KEY WORDS: Neuron-specific enolase, Glasgow Coma Scale, Fisher grade, Hunt and Hess, Modified Rankin Scale, Subarachnoid hemorrhage, WFNS Scale Neurosurgery 78:487–491, 2016
ubarachnoid hemorrhage (SAH) affects approximately 30 000 people annually in the United States, and about one-third of those who survive aneurysmal SAH develop brain injury.1 Until recently, the tools used to assess the severity of SAH and to predict outcomes have been limited to clinical examinations and imaging studies. Biomarkers such as neuron-specific enolase (NSE), glial fibrillary
ABBREVIATIONS: BMI, body mass index; GCS, Glasgow Coma Scale; mRS, modified Rankin Scale; NSE, neuron-specific enolase; SAH, subarachnoid hemorrhage; WFNS, World Federation of Neurologic Surgeons
acidic protein, t protein, myelin basic protein, S100-b, and b amyloid oligomer have been used to measure neuronal and glial damage.2,3 With the emergence of individualized medicine, there is increased interest in the use of biomarkers as potential adjuncts to assess SAH severity and to provide insight into prognosis.4,5 NSE is a globular glycolytic isoenzyme present in neurons throughout the brain with a half-life of approximately 24 hours. Serum NSE levels are elevated in the first 3 days after aneurysmal SAH,4,6 making it a potential biomarker to assess the extent of injury and to predict outcomes. In this study, we aimed to investigate the associations of serum NSE levels with severity of bleeding and functional outcome in patients with
nontraumatic SAH and to examine the potential for using this enzyme as a biomarker of neuronal damage and cognitive or physical disability after SAH.
TABLE 1. Patient Characteristics, Subarachnoid Hemorrhage Severity, and Outcomes for All Patientsa Variable
METHODS After Institutional Review Board approval was granted, we retrospectively reviewed the medical records of all patients with SAH admitted between June 2008 and June 2012. Patients with traumatic SAH were excluded from the study. We analyzed several variables, including patient age, sex, body mass index (BMI), Fisher grade, Glasgow Coma Scale (GCS) score, Hunt and Hess score, and World Federation of Neurologic Surgeons (WFNS) scale score at admission; modified Rankin Scale (mRS) at discharge; approximate time and date of symptom onset; time and date of NSE collection; and concentration of NSE (ng/mL) in blood. For consistency, only the first NSE laboratory performed was used in the statistical analysis. The approximate time of hemorrhage was recorded on the basis of the onset time of severe headaches in every patient. Continuous variables were summarized with the sample median, minimum, 25th percentile, 75th percentile, and maximum. Normal levels of NSE were defined as #15 ng/mL from established laboratory reference numbers. In association analysis, we considered NSE as a continuous and as a categorical variable (normal vs intermediate/elevated). As a categorical variable, a normal NSE level was defined as #15 ng/mL, intermediate NSE was defined as 15 to 30 ng/mL, and elevated NSE was defined as .30 ng/mL. Patients with intermediate and elevated NSE levels were collapsed into 1 category in association analysis because of the small number of patients with an elevated NSE level. Baseline characteristics (age, sex, BMI, length of time between SAH and serum collection for NSE) were compared between patients with a normal (#15 ng/mL) and an intermediate/elevated (.15 ng/mL) NSE level with the use of a Wilcoxon rank-sum test or Fisher exact test. In single-variable analysis without adjustment for potentially confounding variables, associations of NSE level with SAH bleeding severity measures (Hunt and Hess grade, Fisher grade, WFNS grade, initial GCS score) and functional outcome (mRS at discharge) were evaluated with the Spearman test of correlation (considering NSE as a continuous variable) and a Wilcoxon rank-sum test (considering NSE as a dichotomous categorical variable). The Spearman correlation coefficient (r) and corresponding 95% confidence interval (CI) were estimated. In multivariable analysis, associations of NSE with SAH severity and mRS score at discharge were evaluated with the use of proportional odds logistic regression models7 adjusted for age, sex, BMI, and length of time for drawing blood for NSE. Values of P #.05 were considered statistically significant. All statistical analyses were performed with SAS (version 9.3, SAS Institute Inc, Cary, North Carolina).
RESULTS A summary of characteristics, SAH severity, and outcomes for the 71 study patients is provided in Table 1. The median age was 54 years (range, 23-87 years), and 31 patients (44%) were male. Median BMI was 28.1 kg/m2 (range, 17.0-51.0 kg/m2). The median NSE level was 11.0 ng/mL (range, 2.5-51.0 ng/mL). A total of 56 patients (79%) had a normal NSE level (#15 ng/mL), 11 patients (15%) had an intermediate NSE level (.15 and #30 ng/mL), and 4 patients (6%) had an elevated NSE level (.30 ng/mL). The median length of time from onset
488 | VOLUME 78 | NUMBER 4 | APRIL 2016
Summary (n = 71)
Age, y 54 (23, 46, 67, 87) Male sex, n (%) 31 (43.7) 28.1 (17.0, 23.5, 32.7, 51.0) BMI, kg/m2 NSE, ng/mL 11.0 (2.5, 7.8, 14.0, 51.0) Normal (#15 ng/mL), n (%) 56 (78.9) Intermediate (.15 and #30 ng/mL), 11 (15.5) n (%) Elevated (.30 ng/mL), n (%) 4 (5.6) Length of time NSE was in serum, h 25.2 (0.8, 12.2, 39.4, 131.5) Patients with aneurysmal hemorrhage, n 56 Hospital duration, d 17 (1, 11, 22, 53) Hunt and Hess grade, n (%) 1 10 (14.1) 2 20 (28.2) 3 15 (21.1) 4 8 (11.3) 5 18 (25.4) Fisher grade, n (%) 1 2 (2.8) 2 3 (4.2) 3 19 (26.8) 4 47 (66.2) WFNS grade, n (%) 1 26 (36.6) 2 12 (16.9) 3 2 (2.8) 4 12 (16.9) 5 19 (26.8) Initial GCS score, n (%) 0-3 10 (14.1) 4-6 9 (12.7) 7-9 8 (11.3) 10-12 4 (5.6) 13-15 40 (56.3) mRS score, n (%) 0 3 (4.2) 1 11 (15.5) 2 11 (15.5) 3 19 (26.8) 4 12 (16.9) 5 3 (4.2) 6 12 (16.9) a
GCS, Glasgow Coma Scale; mRS, modified Rankin Scale; NSE, neuron-specific enolase; WFNS, World Federation of Neurological Surgeons. Continuous variables are summarized with the sample median (minimum, 25th percentile, 75th percentile, maximum).
of hemorrhage until NSE blood draw was 25.2 hours (range, 0.8-131.5 hours). A comparison of patient age, sex, BMI, and length of time that NSE was detected in the serum was made between patients with normal NSE (n = 56) and those with elevated NSE (n = 15), as shown in Table 2. No significant differences were noted (all P $ .38).
DISCUSSION TABLE 2. Comparison of Age, Sex, Body Mass Index, and Length of Time Neuron-Specific Enolase Was in Blood Serum According to Neuron-Specific Enolase Concentrationa
NSE # 15 ng/mL (n = 56)
NSE . 15 ng/mL (n = 15)
Age, y 53.5 (23.0, 83.0) 56.0 Male sex, n (%) 25 (44.6) 6 28.3 (17.0, 51.0) 27.3 BMI, kg/m2 Length of time NSE 25.1 (0.8, 131.5) 25.3 was in blood serum, h
(35.0, 87.0) (40.0) (19.8, 49.3) (0.8, 116.0)
P Value .45 .78 .38 .88
BMI, body mass index; NSE, neuron-specific enolase. Continuous variables are summarized with sample median (minimum, maximum). P values result from a Wilcoxon rank-sum test or Fisher exact test.
An evaluation of the association of serum NSE concentration with SAH severity and functional outcome in single-variable analysis without adjustment for potential confounding variables is shown in Table 3. In this unadjusted analysis, there was evidence of an association of NSE level with SAH severity as measured by Hunt and Hess grade (r = 0.40; 95% CI, 0.18-0.58; P = .001), WFNS grade (r = 0.48; 95% CI, 0.28-0.64; P , .001), initial GCS score (r = 20.44, 95% CI, 20.61 to 20.24; P , .001), and functional outcome as measured by the mRS at discharge (r = 0.40; 95% CI, 0.18-0.58; P = .001). Slightly weaker but still significant associations were observed when these measures were compared between patients with normal and patients with elevated NSE levels (all P # .04; Table 3). The only measure of SAH severity that was not significantly associated with NSE level was Fisher grade (P $ .43). In multivariable analysis adjusted for age, sex, BMI, and length of time for NSE collection, the significant associations of NSE (as a continuous variable) with Hunt and Hess grade (P = .003), WFNS grade (P , .001), initial GCS score (P = .003), and mRS score at discharge (P = .001) remained consistent, and the lack of association with Fisher grade also was similar (P = .81). Multivariable adjustment for the aforementioned variables did not noticeably alter results when NSE was considered as a categorical variable (#15 vs .15 ng/mL) and associations with Hunt and Hess grade (P = .06), WNFS grade (P = .01), initial GCS score (P = .03), mRS score at discharge (P , .001), and Fisher grade (P = .95) were assessed. In terms of the association between NSE and mRS at discharge, although it is unclear whether it would be reasonable to adjust our multivariable analysis for SAH severity measures as a result of the potential for NSE itself to be a measure of SAH severity, we examined the sensitivity of our results to such adjustments. When multivariable analysis were additionally adjusted individually for Hunt and Hess grade, Fisher grade, WFNS grade, and initial GCS score, the association between NSE and mRS at discharge remained significant when NSE was considered as a continuous variable (all P # .03) and as a categorical variable (all P # .03).
Our study found an association between NSE levels and SAH severity as measured by Hunt and Hess grade (P = .001), WFNS grade (P , .001), and initial GCS score (P , .001) and with functional outcome as measured by mRS at discharge (P = .001). When these measures were compared between patients with normal and patients with elevated NSE levels, slightly weaker but still significant associations were observed (all P # .04; Table 3). The only measure of SAH severity that was not significantly associated with NSE level was Fisher grade (P $ .43). In other studies, the use of biomarkers to determine prognosis in nontraumatic SAH has been limited to outcome prediction based on clinical and radiological markers of unfavorable outcome.8 While individualized medicine continues to gain momentum, we believe that integration of biomarkers that are indicative of the amount of neuronal injury will hold great promise for improving the accuracy of outcome prediction and will help clinicians in therapeutic decisions.9 For example, cerebral microdialysis levels of t, a microtubule-associated protein, are associated with 12month outcomes, and total t may be an important biomarker for predicting long-term outcomes in patients with severe SAH.10 In another study, serum glial fibrillary acidic protein concentrations were used as a marker for brain damage with potential to define the severity and prognosis in aneurysmal SAH.11 Additionally, patients with secondary events had a significantly different serum glial fibrillary acidic protein profile (later and higher maximum) compared with those without secondary events. In comparison, NSE is a glycolytic enzyme that can be detected in the serum after damage of the neuronal cell membrane and its release from the intracellular compartment. Although the role of NSE in the assessment of functional outcome has not been established, preliminary reports have emerged in ischemic stroke and SAH literature.12-14 Release kinetics of NSE has been associated with SAH patients’ clinical deficits and infarct volume, and this enzyme has been suggested as an additional predictor of the early course and functional outcome.13 Although these preliminary data hold great promise, additional work is needed to establish sensitivity and specificity because serum concentrations have been detected in severe internal organ damage.15,16 In our study, we found a significant correlation between serum NSE concentrations and functional outcomes in patients with nontraumatic SAH. A similar study used stringent inclusion criteria for acute brain hemorrhage and tested NSE as a biomarker using GCS between 3 and 4, Fisher grade, and mRS, whereas in this investigation, the entire range of the grading scales were included, allowing a more practical analysis of NSE as an accurate and representative biomarker for predicting functional outcome.4 This is the first study in which several neurological diagnostic grading scales were used to determine the relation between serum NSE concentrations and patient functional outcome after SAH. Serum NSE concentrations .15 ng/mL were considered relatively high and were associated with poor Hunt and Hess grade, WFNS grade, and initial GCS score. There was also a significant correlation
TABLE 3. Association of Serum Neuron-Specific Enolase Concentration With Severity of Subarachnoid Hemorrhage and Functional Outcome in Single Variable Analysis Without Adjustment for Potential Confounding Variablesa P Value
GCS, Glasgow Coma Scale; mRS, modified Rankin Scale; NSE, neuron-specific enolase; WFNS, World Federation of Neurological Surgeons. P values for general tests of correlation result from the Spearman test of correlation. P values comparing NSE #15 and .15 ng/mL groups result from a Wilcoxon rank-sum test.
between NSE .15 ng/mL and mortality (defined as a score of 6 on the mRS). Although functional outcome depends on a variety of factors apart from brain lesion size and location, because of the significant correlation between serum NSE concentrations and functional outcome, we believe there is potential for using NSE as a biomarker to predict functional outcome. Our data strongly suggest a high degree of neuronal injury in poor-grade SAH patients, and as the focus on functional outcome gains momentum, it will be useful to determine factors such as specificity, sensitivity, and positive predictive value to see whether an elevated concentration of NSE forecasts a poor outcome. Finally, this analysis is not without limitations. Because of its retrospective design and relatively small sample size, future studies on larger cohorts will be important. NSE levels were captured only intermittently for patients admitted with SAH because the test was ordered when the senior author was covering service, and NSE was
490 | VOLUME 78 | NUMBER 4 | APRIL 2016
collected in both comatose and conscious subjects, as evidenced by GCS ranges. In addition, the study did not account for patients who required placement of an external ventricular drain, which has been shown to increase NSE in the cerebrospinal fluid.17,18 Fisher grade was not found to be significantly different in the 2 groups.19 Thus, NSE might not be helpful in predicting the risk of vasospasm. Additionally, because of the design of our observational study, we were unable to prove a causal relationship between elevated NSE and specific causes of brain injury such as hypoxic, ischemic, and metabolic. However, as research into biomarkers continues to evolve, additional markers will continue to emerge, and in line with previous studies, we propose that assessment of cognitive functioning should be an integral part of neurological evaluation after SAH. We also hope to see the addition of a panel of biomarkers and cognitive neuropsychological testing for all patients with cerebrovascular
disorders to further our ability to predict detailed outcome measures and potentially to forecast imminent complications.
CONCLUSION This analysis shows a significant association between increased serum NSE levels and poor functional outcome. Although patients who presented with worse grades (Hunt and Hess, GCS, and WFNS scores) had higher levels of NSE, there was no association between the severity of Fisher grade and NSE levels. Further studies are needed to formulate a representative temporal profile of serum NSE elevation after SAH, to define precise cutoffs to determine prognosis, and to better understand the potential relationship between serum NSE levels after SAH and functional outcome. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Roger VI, Go AS, Lloyd-James DM, et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics— 2011 update: a report from the American Heart Association. Circulation. 2011;123 (4):e18-e209. 2. Ahmad O, Wardlaw J, Whitely WN. Correlation of levels of neuronal and glial markers with radiological measures of infarct volume in ischemic stroke: a systematic review. Cerebrovasc Dis. 2012;33(1):47-54. 3. Gatson JW, Warren V, Abdelfattah K, et al. Detection of b-amyloid oligomers as a predictor of neurological outcome after brain injury. J Neurosurg. 2013;118(6): 1336-1342. 4. Siman R, Giovannone N, Toraskar N, et al. Evidence that a panel of neurodegeneration biomarkers predicts vasospasm, infarction, and outcome in aneurysmal subarachnoid hemorrhage. PLoS One. 2011;6(12):e28938. 5. Jung CS, Lange B, Zimmermann M, Seifert V. CSF and serum biomarkers focusing on cerebral vasospasm and ischemia after subarachnoid hemorrhage. Stroke Res Treat. 2013;2013:560305. 6. Kacira T, Kemerdere R, Atukeren P, et al. Detection of caspase-3, neuron specific enolase, and high-sensitivity C-reactive protein levels in both cerebrospinal fluid and serum of patients after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2007;60(4):674-679. 7. Agrestri A. Logic models for multinomial responses. In: Categorical Data Analysis. New York, NY: Wiley-Interscience; 2002:274-281. 8. Bederson JB, Connolly ES Jr, Batjer HH, et al. Guidelines for the management of aneurysmal SAH: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke. 2009;40(3):994-1025. 9. Chou SH, Robertson CS; Participants in the International Multi-Disciplinary Consensus Conference on the Multimodality Monitoring. Monitoring biomarkers of cellular injury and death in acute brain injury. Neurocrit Care. 2014;21(suppl 2):S187-S214. 10. Helbok R, Schiefecker A, Delazer M, et al. Cerebral tau is elevated after aneurysmal subarachnoid haemorrhage and associated with brain metabolic distress and poor functional and cognitive long-term outcome. J Neurol Neurosurg Psychiatry 2015;86(1):79-86. 11. Nylen K, Ludwig ZC, Ost O, et al. Serum glial fibrillary acidic protein in related to focal brain injury and outcome after aneurysmal subarachnoid hemorrhage. Stroke. 2007;38(5):1489-1494. 12. Brea D, Sobrino T, Blanco M, et al. Temporal profile and clinical significance of serum neuron-specific enolase and S100 in ischemic and hemorrhagic stroke. Clin Chem Lab Med. 2009;47(12):1513-1518. 13. Oertel M, Schumacher U, McArthur DL, Kastner S, Boker D. S-100b and NSE: markers of initial impact of subarachnoid hemorrhage and their relation to vasospasm and outcome. J Clin Neurosci. 2006;13(8):834-840. 14. Bharosay A, Bharosay VV, Varma M, Saxena K, Sodani A, Saxena R. Correlation of brain biomarker neuron specific enolase (NSE) with degree of disability and neurological worsening in cerebrovascular stroke. Indian J Clin Biochem. 2012;27(2):186-190.
15. Hullin DA, Brown K, Kynoch P, Smith C, Thompson RJ. Purification, radioimmunoassay, and distribution of human brain 14-3-2 protein (nervoussystem enolase) in human tissues. Biochim Biophys Acta. 1980;628(1):98-108. 16. Pelinka LE, Hertz H, Mauritz W, et al. Nonspecific increase of systemic neuron-specific enolase after trauma: clinical and experimental findings. Shock. 2005;24(2):119-123. 17. Brandner S, Thaler C, Buchfelder M, Kleindienst A. Brain-derived protein concentrations in the cerebrospinal fluid: contribution of trauma resulting from ventricular drain insertion. J Neurotrauma. 2013;30(13):1205-1210. 18. Beems T, Simons KS, Van Geel WJ, De Reus HP, Vos PE, Verbeek MM. Serum and CSF concentrations of brain specific proteins in hydrocephalus. Acta Neurochir (Wien). 2003;145(1):37-43. 19. Fisher C, Kistler J, Davis J. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 1980;6(1):1-9.
he authors report a retrospective review of 71 patients with nontraumatic subarachnoid hemorrhage who had neuron-specific enolase (NSE) tested at their institution. Severity of bleeding was assessed with the Fisher scale, Glasgow Coma Scale, Hunt and Hess grade, and the World Federation of Neurologic Surgeons Scale, whereas outcomes were assessed at discharge with the modified Rankin Scale. NSE was analyzed in both a continuous fashion and categorically with the following definitions: normal, #15 ng/ mL; intermediate, 15 to 30 ng/mL; and elevated, .30 ng/mL. In multivariable analysis, the authors reported that increased NSE was associated with poorer Hunt and Hess, World Federation of Neurologic Surgeons, and Glasgow Coma Scale grades, as well as worse outcomes at discharge. There was no significant association between NSE level and Fisher grade. Although a number of previous studies have demonstrated that NSE is increased after subarachnoid hemorrhage, the correlation to clinical grade and outcome is less clear.1-4 The present study is a valuable addition to the literature because it clearly demonstrates an association between NSE and bleeding severity and discharge outcome. Further investigation of both serum and cerebrospinal fluid NSE is warranted, however, because we already know that there is a strong association between clinical grade and functional outcome, and a serum biomarker does not add significantly to this. Potentially interesting correlations that were not explored in the present study include the relationship of NSE to external ventricular drain placement, aneurysm treatment type, vasospasm incidence, and long-term clinical outcome. It would be useful to know, for instance, if a lower NSE level in a poor-grade patient portends an improved outcome at 1 year. Perhaps this would be the type of poor-grade patient to continue maximal medical and surgical management despite a seemingly dismal prognosis. Justin Mascitelli J Mocco New York, New York
1. Mabe H, Suzuki S, Mase M, Umemura A, Nagai H. Serum neuron-specific enolase levels after subarachnoid hemorrhage. Surg Neurol. 1991;36(3):170-174. 2. Vos PE, van Gils M, Beems T, Zimmerman C, Verbeek MM. Increased GFAP and S100beta but not NSE serum levels after subarachnoid haemorrhage are associated with clinical severity. Eur J Neurol. 2006;13(6):632-638. 3. Kacira T, Kemerdere R, Atukeren P, et al. Detection of caspase-3, neuron specific enolase, and high-sensitivity C-reactive protein levels in both cerebrospinal fluid and serum of patients after aneurysmal subarachnoid hemorrhage. Neurosurgery. 2007; 60(4):674-679; discussion 679-680. 4. Moritz S, Warnat J, Bele S, Graf BM, Woertgen C. The prognostic value of NSE and S100B from serum and cerebrospinal fluid in patients with spontaneous subarachnoid hemorrhage. J Neurosurg Anesthesiol. 2010;22(1):21-31.
Biochemical mediators alter cerebral perfusion and have been implicated in delayed cerebral ischemia (DCI) and poor outcomes after aneurysmal subarachnoid hemorrhage (aSAH). Estrogens (estrone [E1] and estradiol [E2]) are mediators with neuroprotecti
Preclinical studies show that iron plays a key role in mediating neuronal injury. This study was performed in order to identify the relationship between the serum level of ferritin and severity of the brain injury which occur after an Intracerebral h
Aneurysmal subarachnoid hemorrhage (SAH) is one of the most severe forms of stroke, which results from the rupture of a cerebral aneurysm. SAH is the only type of stroke with a female predominance, suggesting that reproductive factors may play a sign
To determine whether serum procalcitonin (PCT) levels at admission were associated with short-term functional outcome after acute ischemic stroke (AIS) in a cohort Chinese sample. We prospectively studied 378 patients with AIS who were admitted withi
The use of biomarkers to predict stroke prognosis is gaining particular attention nowadays. Neuron specific enolase (NSE), which is a dimeric isoenzyme of the glycolytic enzyme enolase and is found mainly in the neurons is one such biomarker.
In recent years, the relationship between schizophrenia and environmental factors has come into prominence. This study investigated the relationship between vitamin D levels and the positive and negative symptoms of schizophrenia by comparing vitamin