Combination of White Matter Hyperintensity and Neutrophil-to-Lymphocyte Ratio Predicts Short-Term Prognosis of Acute Ischemic Stroke Patients

Introduction

Stroke is a severe brain injury disease, and acute ischemic stroke (AIS) is the most common subtype accounting for about 85% of all subtypes.¹ Stroke occupies the second leading cause of deaths and disability globally,² and its prevalence and incidence in China have been escalating over the past decade.³ China has the largest amount of stroke patients in the world, leading to a huge burden on families of patients and a major drain on public health-care funding.⁴ Therefore, prognosis prediction of AIS is useful and preoccupying to families and society.

White matter hyperintensity (WMH) is a common neuroimaging manifestation involved in the aging process and reflects the presence of cerebral small-vessel disease.^5,6 It is related to loss of microstructural white matter integrity and increased blood–brain barrier permeability,⁷ and can affect the remodeling of the myelin sheath after stroke.⁸ WMH has been demonstrated the association with poor functional outcomes after ischemic stroke,⁹ and can be used to predict post-stroke function and cognition impairment.^7,10,11

Neuroinflammatory responses can aggravate ischemic brain damage and neurological dysfunction, playing a critical role in the pathophysiology of ischemic stroke.^12–14 Neutrophil-to-lymphocyte ratio (NLR), as an inexpensive biomarker in systematic inflammatory process, has been confirmed to possess diagnostic and prognostic capabilities in many diseases.^15–17 In AIS patients, NLR is associated with severity of stroke, recurrence of cerebral infarction, short-term mortality and functional prognosis.^18,19 However, the value of WMH combined with NLR in predicting short-term prognosis of AIS patients has still not been investigated. In this study, combination of the neuroimaging marker (WMH) and the inflammatory biomarker (NLR) was first applied to predict short-term prognosis of AIS patients. The aim is to provide a more accurate tool for predicting short-term prognosis of AIS patients.

Patients and Methods

Patients

This study was a prospective observational research, enrolling a consecutive cohort of AIS patients in Nanjing Drum Tower Hospital Group Suqian Hospital between January 2023 and March 2024. This study was permitted by the Ethical Committee of Nanjing Drum Tower Hospital Group Suqian Hospital (SQ2021016045) and conducted strictly according to the guidelines of the Declaration of Helsinki. Written informed consents were provided by the participants before the study commencement.

Inclusion criteria included ① acute, first-ever, ischemic stroke confirmed by brain magnetic resonance imaging (MRI) scan; ② age ≥18 years old; ③ admitted within the first 24 h after the onset of symptoms; ④ complete clinical data including the magnetic resonance images, neutrophil and lymphocyte on admission. Exclusion criteria included ① transient ischemic attack; ② pre-stroke modified Rankin scale (mRS) score ≥2; ③ MRI demonstrating conditions other than cerebral infarction, such as encephalitis, cerebral tumors and hemorrhage; ④ severe kidney or liver dysfunction, or cardiac impairment, or coexistence of other severe systemic diseases; ⑤ other autoimmune conditions or neurodegenerative disorders affecting the nervous systems.

Data Collection

We collected the data of all participants, including demographics (sex, age, weight and height), relevant vascular-based risk factors (smoking, drinking, coronary heart disease, diabetes mellitus, hypertension, hyperlipidemia and atrial fibrillation), clinical data on admission (systolic blood pressure, diastolic blood pressure, NIHSS score, heart rate and random blood glucose), Trial of Org 10172 in acute stroke treatment (TOAST) subtypes, therapeutic methods, laboratory test parameters [white blood cell count (WBC), neutrophil count, lymphocyte count, HbA1c, monocyte (mmol/L), fibrinogen (g/L), homocysteine (µmol/L), triglyceride (mmol/L), total cholesterol (mmol/L), high-density lipoprotein (HDL)-cholesterol (mmol/L), low-density lipoprotein (LDL)-cholesterol (mmol/L)], and magnetic resonance images. The fasting venous blood samples were collected within 24 h after admission, and the magnetic resonance images were analyzed by two experienced neurologists blinded to clinical characteristics, laboratory test results and functional outcomes. Fazekas rating scale was used to assess WMH via deep WMH and periventricular WMH (P-WMH) rating separately. Total Fazekas score, the sum of deep WMH and P-WMH was computed and ranged from 0 to 6.

Outcome Evaluation

All participants were evaluated at 3-month after the onset of AIS using modified Rankin Scale (mRS) score. An mRS score of 0–2 was defined as a favourable outcome, while an mRS score of 3–6 was defined as an unfavourable outcome.

Statistical Analysis

All statistical analysis was performed using the Statistical Package for the Social Sciences version 22.0 (SPSS Inc., USA). The normality of continuous variables was ascertained by Kolmogorov–Smirnov test, and the variables with a normal distribution were described with mean ± standard deviation (SD) and compared for the differences between AIS patients with favourable and unfavourable outcome via Student’s t test, and the variables with a non-normal distribution were described with median (M) and interquartile range (IQR) and compared for the intergroup differences via Mann–Whitney U-test. Categorical variables were described with percentages/ratios (%) and compared for the intergroup differences via chi-square test. In order to determine the independent associations of WMH and NLR with unfavourable outcome of AIS, multivariate analysis was then performed for the variables with two-sided P<0.10 in univariate analysis via binary logistic regression model. Receiver operating characteristic (ROC) curves were used to evaluate the predictive values of WMH, NLR and their combination for unfavourable outcome of AIS, and Z test was used to compare the area under curve (AUC). A two-sided P value of <0.05 was used to compute the statistical significance.

Results

General Information

According to the inclusion and exclusion criteria, a total of 322 AIS patients were enrolled during the study period. Thirteen patients were not evaluated for the 3-month outcome. Finally, 309 AIS patients were included in this analysis. These 309 patients consisted of 191 males (61.8%) and 118 females (38.2%) with a mean age of 65.67 ± 12.05 years and a median admission NIHSS of 6 (IQR 3–13). Among them, 201 (65.0%) had a favorable 3-month outcome, while 108 (35.0%) had an unfavorable outcome.

Associations of WMH and NLR with Unfavourable Outcome of AIS

In order to determine the independent associations of WMH and NLR with unfavourable outcome of AIS, univariate analysis was first carried out between AIS patients with favourable and unfavourable outcome. As shown in Table 1, WMH, NLR, diabetes mellitus, atrial fibrillation, TOAST subtype and on-admission NIHSS score were statistically different (P < 0.05), while the remaining variables were not statistically different (P > 0.05). However, the P value of LDL-cholesterol was <0.10.

Table 1 Results of Univariate Analysis Between AIS Patients with Favorable and Unfavorable Outcome at 3-Month After the Onset of Symptoms

Multivariate analysis was then performed for WMH, NLR, diabetes mellitus, atrial fibrillation, TOAST subtype, on-admission NIHSS score and LDL-cholesterol via binary logistic regression model. As shown in Table 2, WMH, NLR and admission NIHSS score were independently associated with unfavourable outcome of AIS after adjusting for diabetes mellitus, atrial fibrillation, TOAST subtype and LDL-cholesterol.

Table 2 Results of Multivariate Analysis Results Between AIS Patients with Favorable and Unfavorable Outcome at 3-Month After the Onset of Symptoms

Predictive Values

ROC curves (Figure 1) showed that the AUCs of WMH and NLR for predicting short-term prognosis of AIS patients were 0.760 [standard error (SE): 0.029, 95% confidence interval (CI): 0.703–0.817, P<0.001] and 0.717 (SE: 0.030, 95% CI: 0.661–0.774, P<0.001), respectively. In order to elevate the predictive value, a combination of WMH and NLR was employed. The result (Figure 1) showed that the AUC of combination prediction was 0.906 (SE: 0.019, 95% CI: 0.868–0.944, P < 0.001), significantly higher than those of individual predictions (0.906 vs 0.760, Z=4.211, P < 0.001; 0.906 vs 0.717, Z=5.322, P < 0.001).

Figure 1 ROC curves for WMH, NLR and their combination in predicting short-term prognosis of AIS patients.

Discussion

The mechanisms of the adverse impact of WMH on the prognosis of AIS patients remain poorly understood. First, WMH is related to integrity impairment of microstructural white matter and the blood–brain barrier (BBB),⁷ and integrity impairment of the BBB increases its permeability, frequently leading to blood extravasation. Second, WMH can affect the remodeling of the myelin sheath after stroke.⁸ Third, cerebrovascular reactivity and cerebral blood flow are decreased in WMH regions.²⁰ Fourth, WMH is confirmed to be a risk factor for cerebral infarction, and its severity may be correlated with the progression and increase in infarct volume.⁵ Finally, WMH is involved in the progression from normal to mild cognitive impairment and is independently associated with dementia.^11,21

Griessenauer et al reported that the WMH burden had an adverse effect on 3-month functional outcome in patients with large vessel occlusion (LVO) stroke or non-LVO stroke;²² and Derraz et al further reported that the WMH burden was related to poor 3-month functional outcome in LVO stroke patients undergoing endovascular thrombectomy, but it was not related to the occurrence of symptomatic intracranial hemorrhage or 90-day mortality.²³ Zhu et al reported that the higher WMH burden was related to increased risk of major disability or death at 14 days or hospital discharge and 3 months in AIS patients.²⁴ Park et al found that higher WMH burden was related to increased risk of recurrent stroke and a composite of stroke/myocardial infarction/vascular death during a 2-year follow-up in Asian patients with ischemic stroke,²⁵ and Ryu et al also found a significant association between higher WMH burden and recurrent stroke through a 1-year follow-up in ischemic stroke patients.²⁶ In addition, Molad et al found that higher WMH burden was correlated with poor cognitive performance after stroke.²⁷ Our study confirmed the independent association of the WMH burden with poor short-term prognosis of AIS patients and moreover demonstrated its moderate predictive value.

In AIS, post-ischemic inflammation was resulted from activation of intravascular leukocytes, stagnant blood flow, and release of pro-inflammatory mediators from brain parenchyma, platelet granules and ischemic endothelium.¹² Activated neuroinflammatory responses can aggravate ischemic brain damage and neurological dysfunction.^12–14 Neutrophils have been shown to be able to aggravate oxidative stress, release pro-inflammatory mediators and increase the permeability of the BBB, thus leading to the aggravation of brain damage.^28,29 The lymphocyte count is an indicator for general health and is affected by acute physiologic stress. Studies have confirmed its active involvement in a protective mechanism in the ischemic brain.^30,31 Its decrease is correlated with the deterioration of immunological function and the early neuroinflammation, therefore resulting in poor clinical outcomes.³²

NLR reflects the balance between neutrophils and lymphocytes in the peripheral blood, and its increase has been demonstrated to be involved in multiple atherosclerotic events including ischemic stroke, peripheral arterial occlusive disease, coronary artery disease and so on.^33–35 Especially, NLR is associated with severity of stroke, recurrence of cerebral infarction, short-term mortality and functional prognosis in AIS patients, demonstrating the potential of serving as a prognostic predictor.^18,19 Chen et al reported that NLR was independently associated with 3-month functional outcomes in AIS patients, possessing moderate predictive value with an AUC of 0.776.³⁶ Xu et al found that NLR was an independent prognostic indicator for stroke progression and a poor 3-month functional status.³⁷ Ying et al showed that increased NLR at admission was independently correlated with the initial stroke severity and poor 14-day prognosis in AIS patients with intracranial atherosclerotic stenosis.³⁸ Wu et al reported that NLR and its dynamic changes were related to 3-month outcome and mortality in AIS patients receiving intravenous thrombolysis with good predictive values.³⁹ Lee et al demonstrated that increased NLR was correlated with failed reperfusion in AIS patients undergoing endovascular treatment.⁴⁰ Gong et al showed that NLR was related to post-thrombolysis early neurological improvement and early neurological deterioration in AIS patients receiving intravenous thrombolysis.⁴¹ Our study showed that NLR was independently associated with poor short-term prognosis in AIS patients, but its predictive value was relatively low.

In order to elevate the predictive value for short-term prognosis of AIS patients, a combination of WMH and NLR was employed in this study. The result showed that the AUC of combination prediction was significantly higher than those of individual predictions, reaching more than 0.900. Therefore, the combination of WMH and NLR could be applied in predicting short-term prognosis of AIS patients, having a high predictive value.

Conclusion

WMH and NLR were independently associated with short-term prognosis of AIS patients, and the combination of WMH and NLR could be applied in predicting short-term prognosis of AIS patients, having a high predictive value.