a b s t r a c t

Objective: Migraine and subarachnoid hemorrhage patients present to emergency departments with the similar symptoms as headache, nausea, and vomiting.

This study investigated whether the neutrophil-lymphocyte ratio (NLR) could distinguish patients with SAH from those with migraine.

Methods: This retrospective study was performed after research ethics committee approval. Data were gathered from the ED and neurology clinics of a university hospital between January 2015 and January 2016, from patients with symptoms of headache (primarily), nausea and vomiting. One hundred and twenty one with SAH, 123 pa- tients with migraine and 987 with other primary headache syndromes were considered. Neutrophil-lymphocyte ratios (NLR-1) were compared between groups on admission. In SAH patients NLR taken on the 24th-30th hour of admission (NLR-2) was compared to admission NLR. Results: NLR values, showed that the median NLR values of SAH patients were significantly higher than migraine and other headaches group values (p b 0.001; p b 0.001). There was no statistically significant difference between the NLR values of the migraine and control groups (p N 0.05). An NLR cut-off value of 4.02 produced 85.95% sen- sitivity, 97.46% specificity, a 33.79 positive likelihood ratio (LR+), and a 0.14 negative likelihood ratio . A sta- tistically significant increase was observed in median NLR-2 values compared to median NLR-1 values in SAH patients (p b 0.001).

Conclusion: In this retrospective analysis, NLR distinguished patients with SAH from those with migraine. Pres- ence of SAH should be evaluated from discharged and readmitted patients (with headache symptoms) when an increase in NLR between initial and readmission levels is observed.

(C) 2017

Introduction

Headache is one of the most common reasons for admission to the emergency department (ED), and one that affects all age groups. No life-threatening pathology is often found in patients admitted to the emergency department due to headache. However, it can be fatal when secondary causes of headache are not diagnosed.

Subarachnoid hemorrhage (SAH), a mortal condition, is seen in 1% of patients presenting with headache in the ED. Generally, 5%-12% of these patients are misdiagnosed at first admission. Prognosis of SAH patients is poor if diagnosis is delayed [1,2].

Patients with migraine may present to the ED with symptoms such as headache, nausea and vomiting, and can thus mimic SAH [3]. Further examination of a patient with migraine is important in terms of

* Corresponding author.

E-mail address: [email protected] (U. Eryigit).

radiation exposure.related to computed tomography performed to ex- clude SAH, cost effectiveness and patient turnover.

It is crucially important to be able to distinguish SAH from migraine using an easy and practical method. Computed tomography has a high sensitivity to diagnose SAH in early stages but sensitivity decreases by the days [4]. But it is not practical to perform a computed tomography for all headache patients and has high dose radiation exposure. It would therefore be beneficial to use several biochemical markers in di- agnostic steps. Some studies have reported that migraine and SAH are related to Inflammatory processes [5,6]. white blood cell count is an inflammatory marker used widely in clinical practice. Even if WBC is within normal limits, an increased “neutrophil-to-lymphocyte ratio” (NLR) may reflect inflammation and inflammation-related pa- thologies [7].

To best of our knowledge, there is no available information about using the NLR to distinguish SAH patients from migraineurs. In this study, we analyzed the diagnostic value of the NLR which has been

http://dx.doi.org/10.1016/j.ajem.2017.03.063

0735-6757/(C) 2017

Laboratory analysis“>shown to be a reliable marker of inflammation in distinguishing SAH and migraine, since these diseases may present with the similar symp- toms as headache, nausea and vomiting.

Materials and methods

Study population

This retrospective study was performed after local research ethics committee approval (Approval no: 2016-34). Data were gathered from the ED and neurology clinics of a university hospital between Jan- uary 2015 and January 2016, from patients with symptoms of headache (primarily), nausea and vomiting. Patients with ICD-10 codes for head- ache (R51.X), migraine (G43.X) and subarachnoid hemorrhage (I60.X) were identified from the ED database. Supporting information was ob- tained from the neurology clinic database. Patient flow diagram for the study protocol were shown in Fig. 1. One hundred and twenty one of the patients with one of these codes (with CT results compatible with SAH) were included in the “SAH group”. One hundred and fifty three patients with R51.X and G43.X codes and diagnosed with mi- graine in the emergency department and followed by neurology depart- ment (according to ICHD migraine criteria) were included in the “migraine patient group”. Migraine was defined as having at least five attacks fulfilling criteria of 1- Headache attacks lasting 4-72 h (untreat- ed or unsuccessfully treated); 2- Headache has at least two of the fol- lowing four characteristics: a. unilateral location b. pulsating quality c. moderate or severe pain intensity d. aggravation by or causing avoid- ance of routine physical activity (e.g. walking or climbing stairs); 3- During headache at least one of the following: a. nausea and/or vomiting b. photophobia and phonophobia and 4. Not better accounted for by another ICHD-3 diagnosis of headache [8]. Nine hundred and eighty seven patients with one of these codes and with no diagnosis of

SAH/migraine constituted the “other primary headaches except migraine”.

Patients with trauma, infectious disease (urinary tract infection, si- nusitis, cellulitis, Otitis media), malignancy, heart failure, coronary ar- tery disease, pulmonary disease or chronic inflammatory disease were excluded from the study, since these might have affected the results. Also patients diagnosed as migraine in the emergency department and not admitted to neurology clinics for follow-up was excluded from this study.

Laboratory analysis

All values in the study were calculated from patients’ initial com- plete blood count (CBC) analysis. NLR values for all patients were calcu- lated from their initial CBC results as dividing the number of neutrophils by number of lymphocytes. Routine electronic CBC device was used (Cell-Dyne 3700, Abbott, Abbott Park, IL, USA) for this purpose.

Statistical analysis

The difference between median NLR values for the groups was ana- lyzed. SAH patients’ NLR values were calculated from initial CBC results. NLR2 values from blood samples taken 24-30 h after admission, and in- flammation and time-dependent changes in NLR in SAH patients’ values were analyzed. After then SAH patients were grouped according to Hunt Hess scale. Patients in grade 1 (asymptomatic or minimal headache and any neurologic deficit) in Hunt Hess scale (n = 52) was compared with the migraine and other headache syndromes. spearman correlation analysis was performed between NLR and age/sex to determine wheth- er age and gender was a confounder.

SPSS 13 for Windows (Statistical Package for Social Sciences for Win- dows v13.0; SPSS, Chicago, IL) and MedCalc for Windows v15.0 (MedCalc Software, Ostend, Belgium) were used for data analysis.

Fig. 1. Patient flow diagram for the study protocol. Abbreviations: ICD, The International Classification of Diseases; SAH, Subarachnoid hemorrhage; NLR2, NLR (neutrophil lymphocyte ratio) values of SAH patients after 24-30th hours of admission.

Table 1

Demographic and laboratory values of patients on admission.

Variable	Subarachnoid hemorrhage (n = 121)	Migraine (n = 153)	Other headache (n = 987)	p
Age	57.0(46.50-67.50)a,b	37.0 (26.50-46.0)a,c	39.0 (27-53)b,c	b0.001
(years)
Gender	60 (49.6)	89 (58.2)	582 582 (59.0)	N 0.05
Female n (%) WBC	12,400(9500-14,940)a,b	6400(5425-7960)a	6750(5600-8110)b	b0.001
(/uL) Neutrophil	10,830 (8200-13,785)a,b	4160 (3215-5280)a	4250 (3370-5310)b	b0.001
(/uL) Lymphocyte	1230 (910-1900)a,b	2280 (1900-2785)a	2430 (1940-2990)b	b0.001
(/uL) NLR	8.68 (5.60-13.18)a,b	1.86 (1.40-2.40)a	1.73 (1.32-2.29)b	b0.001

Values are reported as median (25%-75%), NLR values was calculated as Neutrophil count divided by lymphocyte count. ap b 0.001, b p b 0.001 for WBC;

ap b 0.001, b p b 0.001, c p b 0.05 for age; a p b 0.001, b p b 0.001 for neutrophil; ap b 0.001, b p b 0.001 for lymphocyte; a p b 0.001, b p b 0.001 for NLR;

WBC, White Blood Cell; NLR, Neutrophil lymphocyte ratio; PDW, Platelet distribution width; RDW, red cell distribution width; MPV, mean platelet volume.

Categorical variables were expressed as numbers and percentages, and numerical variables were expressed as median (inter- quartile range, IQR). Kruskal-Wallis analysis of variance test was used to compare me- dian NLR values between groups. NLR values between two groups were analyzed using Mann-Whitney U test with bonferroni correction. NLR-1 (initial values) and NLR-2 (24-30 h after admission) median values for SAH patients were assessed using the Wilcoxon test for paired samples in order to determine any statistical significance. Receiver operating characteristic (ROC) analysis was performed to determine the diagnos- tic value of the NLR in SAH patients. p b 0.05 were considered statistical- ly significant.

Results

Patients’ socio-demographic characteristics and CBC results are sum- marized in Table 1. Median age of patients in the SAH, migraine and other headache syndromes groups were 57.0 (46.50-67.50), 37.0

(26.50-46.0) and 39.0 (27-53), respectively. Age was statistically signif- icantly higher in SAH patients than migraine and other headache groups (p b 0.001, p b 0.001). About 49.6% of SAH patients were female, Female gender was found about 58.2% and 59.0% in migraine and other head- ache syndromes group, respectively. Female gender ratio was lower in SAH patients than migraine and other headache groups. But this differ- ence was not statistically significant (p N 0.05, p N 0.05) Median NLR values in the SAH, migraine and other headache syndromes groups were 8.68 (5.60-13.18), 1.86 (1.40-2.40) and 1.73 (1.32-2.29), respec-

tively. Mann-Whitney U test with bonferroni correction, used to deter- mine differences between median initial NLR values, showed that the median NLR values of SAH patients were significantly higher than migraine and other headache syndromes group values (p b 0.001; p b 0.001). There was no statistically significant difference between the NLR values of the migraine and other headache syndromes groups (p N 0.05). Median values for the three groups are shown on a box plot graphy in Fig. 2. A subgroup analysis of SAH patients in the Hunt

Fig. 2. Box plot graphy showing median (25%-75%) values of groups. Box plot explanation: upper horizontal line of box, 75th percentile; lower horizontal line of box, 25th percentile; horizontal bar within box, median; The whiskers extend out to the most extreme data point that is at most 1.5 times the interquartile range above the third quartile or below the first quartile. Circle (?) represent outliers, diamond (?) represent extreme values. SAH, Subarachnoid hemorrhage.

Fig. 3. Receiver Operating Characteristic curve analyses of the diagnostic capability of NLR for SAH. AUC, Area under the curve; NLR, Neutrophil lymphocyte ratio; SAH, Subarachnoid hemorrhage.

Hess grade 1 patients was evaluated. Median NLR values in Hunt Hess grade 1 patients were 8.05 (4.69-13.57). NLR values of SAH patients in Hunt Hess grade 1 were found to be significantly higher than mi- graine and other headache syndromes group (p b 0.001; p b 0.001). Spearman correlation analysis showed that there was a positive very weak correlation between NLR and age (p = 0.014, r = 0.069), and a very weak negative correlation between NLR and female gender (p = 0.034, r = -0.060). Median NLR values analyzed to consider inflamma- tory process in SAH patients after the 24-30 h of admission (NLR-2) was

14.00 (9.74-20.75). Using the Wilcoxon test for paired samples, a statis- tically significant increase was observed in median NLR-2 values com- pared to median NLR-1 values in SAH patients (p b 0.001). An area under the curve (AUC) value of 0.946 investigated using ROC analysis to determine the diagnostic value of the NLR in SAH patients was mea- sured (Fig. 3). An NLR cut-off value of 4.02 produced 85.95% sensitivity (95% confidence interval; 78.5%-91.6%), 97.46% specificity (95% confi- dence interval; 96.4%-98.3%), a 33.79 positive likelihood ratio (LR+), and a 0.14 negative likelihood ratio (LR-). Sensitivity, specificity, (LR+) and (LR-) for different cut-off values are shown in Table 2. Cut off points to show maximum specificity and sensitivity for different criterions are chosen in Table 2.

Table 2

Sensitivity, specificity and likelihood ratios in some points of NLR values.

Criterion	Sensitivity	95% CI	Specificity	95% CI	LR+	LR-
N 1.18	99.17	95.5-100.0	16.93	14.8-19.2	1.19	0.049
N 1.87	95.04	89.5-98.2	57.1	54.2-60.0	2.22	0.087
N 4.02a	85.95	78.5-91.6	97.46	96.4-98.3	33.79	0.14
N 4.96	80.17	71.9-86.9	99.04	98.3-99.5	83.08	0.20
N 14.28	16.53	10.4-24.4	99.91	99.5-100.00	188.43	0.84

Abbreviations: CI, Confidence interval; +LR, Positive likelihood ratio; -LR, negative like- lihood ratio; NLR, Neutrophil lymphocyte ratio.

^a The best cut off point.

Discussion

The first conclusion from this study is that patients with SAH have significantly higher admission NLR values than migraine and other pri- mary headache syndrome patients. The second is that NLR values are significantly higher in the 24th-30th hours of admission than the time of presentation in SAH patients. NLR values in patients with SAH whose complete blood counts are investigated at time of presentation and in whom diagnosis is missed will subsequently increase. With the second finding, NLR values may then be useful in the diagnosis of SAH especially in the late presenting patients.

A number of studies have investigated the diagnostic and therapeu- tic role of NLR in several inflammatory diseases [9,10].

In addition, several studies have stated that SAH involves inflamma- tory processes and have proposed the use of inflammatory markers for the diagnosis and prognosis of the condition. Proinflammatory and neu- roprotective markers have been shown to increase in SAH and following brain damage. Hollig et al. identified neuroinflammatory mechanisms in brain damage occurring after SAH, an increase in inflammatory param- eters such as CRP, ICAM-1, MMP9, selectin and WBC, and early elevation of IL-6 levels as indicating poor prognosis after discharge [6,11]. Frijins et al. identified elevation of ED1-fibronectin (ED-1fn) and von Willebrand factor (vWF) levels in endothelial injury after SAH, and Handa Y. et al. reported an increase in ICAM-1 expression [12,13]. As shown in these studies, leukocyte migration and inflammation process- es begin after cytokine elevation in the injury area. In our opinion, NLR elevation after SAH is associated with initiation of neuroprotective mechanisms and inflammatory processes. And NLR increases by the time as seen in our results.

There are studies showing the inflammatory processes in the mi- graine patients in the literature. Ceylan et al. showed higher levels of fi- brinogen and PTX-3 levels in the acute attack of migraine patients than healthy control patients. But in the same study there was no significant difference in CRP levels between migraine patients and the other head- ache syndromes group [14]. Karabulut et al. showed that NLR increases as a sign of inflammation during migraine attacks than healthy Control subjects [15]. Our study did not show any statistically significant differ- ence in NLR values between migraine and other primary headache syn- drome patients.

This study demonstrates the role of inflammatory processes in SAH etiopathogenesis and shows that the NLR can safely be used as an in- flammation marker. SAH patients had significantly higher NLR values than migraine and other headache syndromes group patients. As re- ported in literature, the same inflammatory markers are involved in the pathogenesis of migraine, but this exhibits significantly higher NLR values than other Headache types [5]. In contrast, the NLR values of mi- graine patients in our study were not higher statistically significant than those of the other headache syndromes group as previous studies. In our opinion, this may be due to migraine patients presenting to hospital at the early onset of headache. Inflammatory processes and leukocyte mi- gration may thus not have sufficient time to progress significantly.

In conclusion, The NLR can be used in the differential diagnosis of

SAH and migraine.

Presence of SAH should be evaluated from discharged and readmitted patients (with headache symptoms) when an increase in NLR between initial and readmission levels is observed. And NLR levels may be used to identify especially the late presenting sub arachnoid hemorrhage patients. Because of the sensitivity of computed tomogra- phy decreases by the days clinicians may misinterpret and may misdi- agnose late presenting SAH patients [16]. And NLR values can be used to distinguish late presenting SAH patients than migraineurs.

Limitations

Although our center is a reference hospital for SAH patients in the re- gion, patients diagnosed with SAH from other centers and especially

those with onset of headache N 6 h previously were excluded. But we do not know the certain time when the NLR increases rapidly and that there is no clear reference for a cut-off of 6 h. It is our clinical observa- tion. Inflammatory markers may therefore have increased as SAH progressed. Second, since NLR values were measured from only one blood sample, the results may have been affected by Laboratory errors, even though the devices were calibrated. Migraine was defined accord- ing to ICHD criteria and some of migraine patients may be misdiagnosed as other headache groups. Age and gender was correlated very weak with NLR values and these variables may be confounders. Further pro- spective studies, including more patient groups, are required to support these results.

Conflict of interest

None declared.

References

1. Vermeulen MJ, Schull MJ. Missed diagnosis of subarachnoid hemorrhage in the emergency department. Stroke 2007;38(4):1216-21.
2. Kowalski RG, et al. Initial misdiagnosis and outcome after subarachnoid hemorrhage.

   JAMA 2004;291(7):866-9.

   Tabatabai RR, Swadron SP. Headache in the emergency department: avoiding misdi- agnosis of dangerous secondary causes. Emerg Med Clin North Am 2016;34(4): 695-716.

3. Dubosh NM, et al. Sensitivity of early brain computed tomography to exclude aneu- rysmal subarachnoid hemorrhage: a systematic review and meta-analysis. Stroke 2016;47(3):750-5.
4. Acar E, et al. Neutrophil-lymphocyte ratio can distinguish migraine patients from other patients with nonspecific headache in the emergency department. Acta Medica Mediterranea 2015;31(4):829-34.
5. Hollig A, et al. Association of early inflammatory parameters after subarachnoid hemorrhage with functional outcome: a prospective cohort study. Clin Neurol Neurosurg 2015;138:177-83.
6. Rudiger A, et al. The relative lymphocyte count on hospital admission is a risk factor for long-term mortality in patients with acute heart failure. Am J Emerg Med 2006; 24(4):451-4.
7. Headache Classification Committee of the International Headache, S. The interna- tional classification of headache disorders, 3rd edition (beta version). Cephalalgia 2013;33(9):629-808.
8. Yalcin M, et al. Neutrophil-lymphocyte ratio may predict Left atrial thrombus in pa- tients with nonvalvular atrial fibrillation. Clin Appl Thromb Hemost 2015;21(2): 166-71.
9. Demir S, et al. Neutrophil-lymphocyte ratio in patients with major depressive disor- der undergoing no pharmacological therapy. Neuropsychiatr Dis Treat 2015;11: 2253-8.
10. Badjatia N, et al. Inflammation, negative nitrogen balance, and outcome after aneu- rysmal subarachnoid hemorrhage. Neurology 2015;84(7):680-7.
11. Frijns CJ, et al. Endothelial cell activation markers and delayed cerebral ischaemia in patients with subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 2006; 77(7):863-7.
12. Handa Y, et al. Expression of intercellular adhesion molecule 1 (ICAM-1) on the ce- rebral artery following subarachnoid haemorrhage in rats. Acta Neurochir 1995; 132(1-3):92-7.
13. Ceylan M, et al. Serum levels of pentraxin-3 and other inflammatory biomarkers in migraine: association with migraine characteristics. Cephalalgia 2016;36(6): 518-25.
14. Karabulut KU, et al. The change of neutrophils/lymphocytes ratio in migraine at- tacks: a case-controlled study. Ann Med Surg (Lond) 2016;10:52-6.
15. Backes D, et al. Time-dependent test characteristics of head computed tomography in patients suspected of nontraumatic subarachnoid hemorrhage. Stroke 2012; 43(8):2115-9.