Introduction

Acute ischemic stroke (AIS) is the main cause of adult permanent disability, second most common cause of dementia, and third most common cause of mortality worldwide [1]. It also causes high treatment costs and labor losses. Ischemic infarct volume on imaging modality and clinic National Institutes of Health Stroke Scale/Score (NIHSS) are important factors in predicting the post-stroke prognosis of patients. Therefore, current treatment strategies for AIS including anti-aggregant therapy, intravenous or intraarterial recombinant tissue plasminogen activators, and mechanical endovascular therapies aim to limit the infarct volume [2]. In addition to these well-defined traditional prognostic factors, inflammatory markers are also associated with poor prognosis in AIS and may help to detect the patients that required intensive treatment at early stage.

Inflammation has a pivotal role in initiation and progression of atherosclerosis [3]. White blood cell (WBC) count and subtypes of WBC are commonly used inflammatory markers [4]. It was demonstrated that higher WBC and neutrophil counts were correlated with larger infarct volume and increased stroke severity in patients with AIS [5]. However, recently, neutrophil to lymphocyte ratio (NLR) has emerged as a novel marker of the systemic inflammation. It is an easily obtainable parameter from complete blood count and quite inexpensive [4]. Previous studies have reported that NLR was associated with poor outcomes and predict short-term mortality in patients with AIS [6, 7].

AIS can be classified into two groups according to stroke territory: anterior circulation stroke (ACS) and posterior circulation stroke (PCS) [8]. ACS accounts for 70–75% of all ischemic stroke cases, whereas the range of PCS is approximately 20–25%. Although the association between NLR and infarct volume has been investigated previously in patients with AIS, to our knowledge, no study so far evaluated the association of NLR with stroke volume according to the stroke territory. In this study, we aimed to investigate the link between NLR and infarct volume according to the stroke territory in patients with presenting AIS. In addition, we will evaluate the prognostic value of NLR for predicting 3-month mortality in these patients.

Methods

Study population

A total of 107 patients presenting with AIS between 01 January 2017 and 31 December 2017 were included in this study. The diagnosis of AIS was based on World Health Organization’s definition [9]. Patients were etiologically classified according to the Trial of Org 10,172 in Acute Stroke Treatment (TOAST) criteria: cardioembolic, atherosclerotic, small vessel/lacunar, and cryptogenic/others [10]. Exclusion criteria for the patients were the following: (i) those patients with trauma or previous history of surgery; (ii) those with an active infection before stroke onset or within 72 h after admission; (iii) previously known hematologic disorders; (iv) severe hepatic or renal disease; (v) acute metabolic disease or intoxication; (vi) patients with cerebrovascular and/or coronary heart disease; (vii) a history of cancer or the use of steroids or immunosuppressant agent; (viii) those patients having both ACS and PCS together; and (ix) lacking of the medical, demographic, clinical, laboratory, and/or radiologic data. The study was conducted in full accordance with the Declaration of Helsinki and was approved by the local ethics committee. All patients provided written, informed consent before inclusion in this study.

Clinical data

Baseline clinical data including age, gender, and risk factors such as hypertension (HT), diabetes mellitus (DM), hyperlipidemia (HL), heart failure, and atrial fibrillation (AF) were recorded on pre-prepared standard study forms for all patients. HT was defined as systolic blood pressure (BP) ≥ 140 mmHg, and/or diastolic BP ≥ 90 mmHg, taking anti-hypertensive medications, and/or previously diagnosed hypertension. DM was defined as fasting serum glucose of ≥ 126 mg/dL (7 mmol/L), a non-fasting glucose of ≥ 200 mg/dL (11.1 mmol/L), use of anti-diabetic medications, or a previously established diagnosis. HL was diagnosed if low-density lipoproteins (LDL)-cholesterol level is ≥ 100 mg/dL or use of lipid-lowering agents after being diagnosed with HL. Atrial fibrillation was defined as AF recorded at the time of the electrocardiography or any previously known episode of AF.

Laboratory methods

The complete blood count and venous blood samples including C-reactive protein (CRP) and creatinine were taken from the patients on admission to the emergency department. In addition, venous blood samples for cholesterol levels were obtained within 24 h after admission and after a 12-h overnight fasting period. All samples were collected in EDTA tubes, and were sent to the laboratory for analysis within 2 h after collection. Biochemical tests such as total cholesterol, LDL-cholesterol, high-density lipoproteins-cholesterol, triglycerides, and fasting glucose were measured spectrophotometrically. The mean platelet volume (MPV), red cell distribution width (RDV), and WBC counts were measured using a Cell-Dyne counter of Abbott Laboratories (Abbott Laboratories, IL). The NLR was obtained by dividing the neutrophil count to the lymphocyte count.

Magnetic resonance image acquisition and analysis

Magnetic resonance images (MRI) were performed with scanners of 1.5 T (Magnetom Avanto; Siemens, Erlangen, Germany) by using dedicated head coil. Patients were placed in the supine position. The scanning protocol included diffusion-weighted imaging (DWI) and apparent diffusion coefficient sequence with 5-mm section. All the MRI images were evaluated on a workstation (Leonardo; Siemens, Erlangen, Germany) by a neurologist. The infarct volume was calculated based on the DWI images. First, in DWI images with infarct, the infarct areas in every image were calculated separately and added to each other. Second, whole infarct areas were multiplied by section thickness. Whole infarct volumes were obtained as millimeter. Volume calculation on diffusion MRI was performed by a neurologist with 10-year experience. The infarct volume was calculated using the previously described formula [11]. Treatments were managed for all patients according to the previously described guidelines.

Stroke territory and severity

Stroke territory was classified according to MRI. PCS was defined as ischemia in the vascular territory of basilar, vertebral, or posterior cerebral arteries. ACS was defined as ischemia in the vascular territory internal carotid, middle, or anterior cerebral arteries [8]. Our study population was divided into two groups according to this classification: ACS (n = 84) and PCS (n = 23). Stroke severity was assessed based on the NIHSS [12].

Medications and follow-up

The study population was evaluated in terms of the therapy of stroke including anti-aggregant therapy, intravenous or intraarterial recombinant tissue plasminogen activator, and mechanical endovascular therapy. The hospital’s electronic record system was assessed for evaluation of in-hospital and 3-month mortality after AIS and contacted by the telephone and national mortality record.

Statistical analysis

Statistical analyses were performed using the SPSS Statistics 24.0 software (SPSS Inc., Chicago, IL). Continuous data were presented as the mean value ± standard deviation or median with interquartile range (IQR; 25th and 75th) according to distribution normality using the Kolmogorov–Smirnov test. Student’s t test was used for comparisons of parametric variables, and Mann–Whitney U test was used for comparisons of non-parametric variables. Categorical data were expressed as percentage, and compared using the chi-squared test. Pearson’s and Spearman’s rank correlation coefficients were used for correlation analysis. Variables which had unadjusted p value of < 0.1 in univariate analysis were considered as a potential risk predictors and included in the full multivariate model. Multivariate logistic regression analysis was performed to determine the independent predictors of 3-month mortality. Receiver operating characteristic (ROC) curve analysis was used to obtain the sensitivity and specificity of the NLR for predicting 3-month mortality. A p value < 0.05 was defined as statistically significant.

Results

A total of 107 patients were enrolled in this study and followed up 3 months in terms of mortality. The mean age of the study population was 67 (55–74) years, and 53.3% of study population was male. Eighty-four patients (78.5%) had ACS, whereas 23 patients (21.5%) had PCS. Baseline characteristics of the study groups were shown in Table 1. There were no significant differences between two groups regarding the age, sex, and co-morbid diseases including DM, HT, HL, heart failure, and AF. The ACS group was treated mostly with endovascular/thrombolytic, while PCS was treated mostly with anti-aggregant therapy (p = 0.023). In addition, infarct volume was significantly higher in patients with ACS group compared to PCS group (p = 0.005).

Table 1 Comparison of clinical and demographic features of patients with anterior and posterior circulation stroke
Full size table

The laboratory parameters of the study group were listed in Table 2. Biochemical and laboratory parameters including neutrophil, lymphocyte, NLR, and CRP were similar between ACS and PCS groups (p > 0.05 for all).

Table 2 Comparison of laboratory parameters of patients with anterior and posterior circulation stroke
Full size table

In correlation analysis, infarct volume was positively correlated with NIHSS (r = 0.609, p < 0.001), ASPECT (r = 0.742, p < 0.001), CRP (r = 0.350, p = 0.001), and NLR (r = 482, p ≤ 0.001) in ACS group, but it was only correlated with NIHSS (r = 0.593, p = 0.003) in PCS group (Table 3). In contrary to ACS group, there was no significant correlation between infarct volume and NLR in PCS group. In addition, NLR was positively correlated with NIHSS score in ACS group (r = 0.326, p = 0.002), while no significant correlation was detected between these two parameters in PCS group.

Table 3 The correlation of anterior and posterior infarct volumes with clinical scores and laboratory parameters
Full size table

The 3-month mortality rate was 20.6% in study population. Patients in dead group were significantly older (p = 0.024) and had higher CRP level (p = 0.004), higher NLR level (p < 0.001), and higher frequency of AF (p = 0.002) compared to patients in surviving group (Table 4). Univariate and multivariate logistic regression analyses were performed to determine the independent predictors of 3-month mortality. In univariate analysis, age, the presence of AF, NIHSS, CRP, and NLR were found to be associated with increased risk of 3-month mortality. Multivariate logistic regression analysis demonstrated that NLR was the only independent predictor of 3-month mortality. ROC curve analysis was used to detect the sensitivity and specificity of the NLR for predicting 3-month mortality (Table 5). NLR ≥ 4.7 predicted 3-month mortality with a specificity of 76.5%, sensitivity of 63.6%, positive predictive value of 41.2%, and negative predictive value of 89% (Fig. 1).

Table 4 Comparison of clinical, demographic, and laboratory features in surviving and dead patients
Full size table
Table 5 Univariate and multivariate logistic regression analyses showing the independent predictors of 3-month mortality
Full size table
Fig. 1
figure 1

ROC curve of neutrophil to lymphocyte ratio ≥ 4.7 for predicting 3-month mortality. PPV positive predictive value, NPV negative predictive value

Full size image

Discussion

This study revealed that NLR was significantly related to ACS infarct volume, but not related to PCS infarct volume. Also, NLR was the independent predictor of 3-month mortality in patients with AIS. The results point out that a higher NLR can indicate stroke severity on admission, and can help to make the decision of the necessity of endovascular intervention particularly in ACS.

It has been well recognized that inflammatory response is implicated in all stages of AIS [13]. Inflammatory mediators are expressed at low levels in normal brain tissue. However, an ischemic condition could provide the release of proinflammatory cytokines and recruitment of immune cells, which represent an important mechanism of secondary progression of brain lesion [14]. Studies showed that elevated WBC count is associated with poor outcomes in patients with AIS [15]. However, every WBC subtypes have different prognostic role [16]. Neutrophils are the first WBC subtype recruited to the damaged brain area [17]. Activated neutrophils cause the release of proteolysis enzymes such as acid phosphatase or reactive oxygen products, and exacerbate ischemic brain injury. Early neutrophilia has been reported to be associated with larger infarct volumes [5]. In addition, it has been demonstrated that infarct volume is significantly reduced when neutrophil infiltration is inhibited [18]. In contrast to neutrophils, the role of lymphocytes in AIS is controversial. Nevertheless, it has been demonstrated that lymphocyte counts decreased after AIS, and this reduction in lymphocytes is associated with poor prognosis [16, 19]. Therefore, recent studies have demonstrated that NLR which reflects the balance between neutrophils and lymphocytes is a more useful marker than other WBC subtypes for risk prediction.

Increased NLR has been generally reported to be associated with poor prognosis and predict short-term mortality in patients with AIS [6, 7, 20,21,22]. Similar to previous studies, we also showed that NLR was an independent predictor of 3-month mortality in AIS patients and NLR ≥ 4.7 predicted 3-month mortality with a specificity of 76.5% and sensitivity of 63.6%. We think that the possible underlying mechanism of the relationship between higher NLR and mortality in patients may be explained as follows: NLR may increase as a result of inflammatory response in AIS or be a marker indicating the severity/extent of necrotic tissue and its associated biochemical pathways. Previous studies have shown that increased neutrophils release cytokins, chemokines, free oxygen radicals, and other inflammatory mediators, which result in blood-brain barrier disruption, neuronal cell death, and hemorrhagic transformation after AIS. All these occurred changes are associated with poor outcome and may contribute to the mortality of AIS [20]. Our results confirm that elevated NLR on the hospital admission is an easy and strong predictor of poor outcomes in AIS patients. Therefore, we can conclude that a higher NLR should be taken into consideration in patients with AIS and these patients should be followed more closely in terms of poor outcomes. On the other hands, RDW and MPV were also found to be associated with AIS in the previous studies [23, 24]. However, we found no association among RDW, MCV, and mortality. This may be due to our small patient population.

Infarct volume is a good predictor of the severity of neurological deficits, functional outcomes, and short-term mortality after stroke [6, 25]. The NIHSS is a neurologic severity scale that is a valid, reliable, and reproducible scale, and it is correlated with infarct volume [12]. It is also showed that NLR is positively correlated with infarct volume and NIHSS [7]. However, these studies included all stroke patients and did not classify the patients according to the stroke territory. In this study, we divided patients into two groups as ACS and PCS according to the stroke territory. Consistent with previous findings in literature, we found that NIHHS was positively correlated with both ACS and PCS infarct volumes. We also detected that NLR was significantly correlated with ACS infarct volume. However, there was no significant correlation between NLR and PCS infarct volume. This unexpected situation between the NLR and PCS infarct volume may possibly be explained as follows. A very small infarct in posterior circulation can be fatal such as pontine infarct which causes severe inflammatory response involving cardio-respiratory system. Thus, the association between NLR and mortality may be based on different mechanisms in the ACS and PCS groups (i.e., extent of tissue necrosis, autonomic derangement, or another mechanism). Our results suggest that both infarct volume and autonomic distress may be related to mortality in ACS, whereas only autonomic distress due to severe deficits, impairment of consciousness, and direct damage to autonomic pathways may be associated with mortality in PCS. To our knowledge, this is the first study that investigates the association of NLR with ACS and PCS infarct volumes. We think that our results will need to be validated in larger prospective studies.

Early risk stratification in patients with AIS is important to determine the appropriate treatment and improve outcomes. Therefore, risk assessment should be performed as soon as possible after patient’s admission. NLR is a simple marker and easily obtain from complete blood count. A higher NLR can indicate more severe form of AIS and be a clinically beneficial tool for risk stratification. Therefore, we suggest that NLR can identify the higher-risk patients that required endovascular intervention particularly in ACS patients.

This study has several limitations. First, the study conducted with limited time and sample size was relatively small. It could be interesting to evaluate long-term prognostic value of NLR with a larger sample size in AIS patients. Second, although we excluded patients with active infection, many diseases and factors that might affect inflammatory markers may not be taken into consideration. Larger studies with long-term follow-up are required to better elucidate the prognostic role of NLR in AIS patients according to stroke territory.

In conclusion, NLR is significantly correlated with ACS infarct volume, whereas not correlated with PCS infarct volume. In addition, it is an independent predictor of 3-month mortality in AIS patients. Therefore, it may guide the physician deciding the necessity of endovascular intervention particularly in ACS patients.