Systemic immune-inflammation index upon admission correlates to post-stroke cognitive impairment in patients with acute ischemic stroke

Background: The purpose of this prospective study was to evaluate the association of systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI), with PSCI in patients with acute ischemic stroke (AIS). Methods: First-onset AIS patients were consecutively included from January 1, 2022 to March 1, 2023. The baseline information was collected at admission. Fasting blood was drawn the next morning. Cognitive function was assessed by the Montreal Cognitive Assessment (MoCA) 3 months after onset. Logistic regression analysis was performed to explore the correlation between SII, SIRI, and PSCI. Receiver operating characteristic (ROC) was conducted to evaluate the predictive ability of SII. Results: 332 participants were recruited, and 193 developed PSCI. Compared with patients without PSCI, the patients with PSCI had higher SII (587.75 (337.42, 988.95) vs. 345.66 (248.44, 572.89), P<0.001) and SIRI (1.59 (0.95, 2.84) vs. 1.02 (0.63, 1.55), P=0.007). SII and SIRI negatively correlated with MoCA scores (both P<0.05). The multivariable logistic regression analysis indicated that SII was independently associated with PSCI (P<0.001), while SIRI was not. The optimal cutoff for SII to predict PSCI was 676.83×109/L. Conclusions: A higher level of SII upon admission was independently correlated to PSCI three months later in AIS patients.


INTRODUCTION
According to the 2019 Global Burden of Disease (GBD) study, stroke remains the second leading cause of death worldwide [1].As a prevalent non-motor complication of AIS, PSCI has been confirmed to be related to poor outcomes [2].Early identification and intervention of PSCI can avoid the progressive deterioration of cognitive function and effectively improve the prognosis of patients.
Recent studies have suggested potential mechanistic links between inflammation, stroke and dementia [3,4].Neutrophils, lymphocytes, platelets, and monocytes are essential immune system elements [5].The balance between innate and adaptive immunity can be better indicated by the systemic inflammation response index (SIRI) and the systemic immune-inflammation index (SII), which are calculated from the counts of neutrophils, platelets, monocytes, and lymphocytes [6].The diagnostic and predictive efficacy of cardiovascular diseases, tumors, and inflammatory diseases has been confirmed by previous research [7][8][9][10].Besides, another research has shown a link between the SII and a high incidence of dementia in the general public [3].Recent findings have also confirmed the correlation between SII and hemorrhagic transformation as well as poor prognosis in AIS patients [6].Meanwhile, the relationship between SII or SIRI and PSCI remains uncertain.Therefore, we designed this prospective cohort study to explore the association of SII and SIRI with PSCI and further evaluate their predictive value for PSCI.

Relationship between SII, SIRI and PSCI
The findings from the logistic regression models with SII and SIRI as continuous variables are presented in Table 2. Results from the univariable logistic regression analysis demonstrated a significant association of PSCI with age, education level, history of atrial fibrillation,    were proved to be independently associated with PSCI in the multivariable regression analysis.Furthermore, SII and SIRI were then entered into the multivariable regression model as tertiles.The findings demonstrated that when the first tertile was taken as a reference, the second and third tertile of SII were both independent risk factors for PSCI (OR=2.355,P=0.021 and OR=10.369,P<0.001, respectively).However, no significant correlation between SIRI and PSCI was found (Table 3).

ROC analysis of SII for predicting PSCI
The diagnostic utility of SII in predicting PSCI was assessed using ROC analysis, with the AUC of 0.659 (P<0.001).(Figure 6).The optimal cutoff value was ≥676.83×10 9 /L, and the sensitivity and specificity were 44.6% and 82.0%, respectively.

DISCUSSION
The inflammation response has been reported to be crucial in stroke and PSCI pathobiology [11,12].Previous studies have revealed that systemic inflammation processes are closely related to endothelial dysfunction, cell death, blood-brain barrier (BBB) disruption, cerebral blood flow autoregulation disorder and platelet aggregation [5,13].Studies consecutively showed that PSCI was associated with some inflammatory biomarkers and cytokines [14,15].Otherwise, as important primary immune mediators that can release inflammatory signals, infiltrating leukocytes, including neutrophils, monocytes, and lymphocytes, have been reported to be related to stroke and dementia [16,17].Previous studies have shown that peripheral neutrophils and neutrophil to lymphocyte ratio (NLR) correlate to poor prognosis and hemorrhagic transformation of ischemic stroke, as well as cerebral small vessel disease and dementia [18][19][20].
The SII and SIRI derived from different blood cells can better reflect the inflammation or immune status than one cell alone.Hu et al. created SII and reported that SII was an effective predictor of poor outcomes of patients after an operation for hepatocellular carcinoma [9].Meanwhile, SII was also considered to be correlated to unfavorable outcomes of various tumors such as cholangiocarcinoma, lung cancer, gliomas, etc. [10,21,22].Subsequently, there has been an established link between SII and the occurrence and prognosis of chronic heart failure and coronary heart disease [23,24].A recent system review showed that elevated SII could significantly increase the risk of vascular disease, including ischemic stroke, hemorrhagic stroke, myocardial infarction, and peripheral arterial disease [7].According to a recent large-scale general population study, elevated SII and SIRI could increase the incidence of stroke and all-cause death [25].In addition, SII on admission was reported to be positively associated with symptomatic intracranial hemorrhage after endovascular treatment in AIS patients with large vessel occlusion [24].Recently, several research have confirmed the potential relationship between SII and cognitive impairment.According to a retrospective study, elevated SII was closely related to the occurrence AGING of postoperative cognitive decline [26].Another research also showed a strong correlation between SII and cerebral small vessel disease (CSVD) and cognitive impairment [27].Therefore, we speculate that SII and PSCI might have a potential relationship.However, there are few reports on the correlation between them.In this research, we discovered that SII was independently correlated to PSCI and might be used as a valid predictor.
The specific mechanisms for the association between SII and PSCI are not yet well understood.Nevertheless, it is hypothesized that blood-brain barrier disruption, endothelial dysfunction, CVSD, and neuroinflammation could have significant implications [11,28].Many studies have shown that neutrophils, platelets, and lymphocytes, essential components of SII, were related to endothelial dysfunction and bloodbrain barrier disruption [17,7].Neutrophils play a negative role by releasing reactive oxides, synthesizing cytokines, intercellular adhesion molecules, and other inflammatory mediators, while platelets secrete proinflammatory cytokines and growth factors.[29][30][31].Furthermore, accumulating evidence has shown that increased SII levels are related to more severe CSVD, which plays a crucial role in cognitive impairment [32].A community-based population study has demonstrated that individuals with a higher SII had an increased risk of moderate-to-severe enlarged perivascular space (EPVS) and modified white matter hyperintensity (WMH) burden [19].In addition, peripheral inflammation can penetrate the BBB and induce central neuroinflammation, ultimately contributing causally to cognitive impairment [33].Emerging evidence has suggested that neuroinflammation plays an active role rather than being passive activation in the pathogenesis of cognitive impairment [13,34].
Our study confirmed the potential association of SII with PSCI.Nevertheless, some limitations should be noted.First, a few patients with severe aphasia, dysarthria or disturbance of consciousness were not included, which could lead to bias.Second, although we excluded patients with previous stroke, it is difficult to guarantee that the cognitive impairment was exclusively stroke-related because we could not accurately assess pre-stroke cognitive function.Third, we only measured SII and SIRI levels at admission and MoCA scores three months after stroke.The lack of serial detections for levels of SII, SIRI, and cognitive performance may have obscured any potential impact of treatment interventions on the noted correlation.Fourth, we could not rule out the possible impact of some potential risk factors we did not measure, such as ApoE status.Finally, the single center and restricted sample size limit the generalization of the results of our study, and the predictive effect of SII for PSCI in this study was statistically significant but not strong enough.

CONCLUSIONS
The potential correlation between SII and PSCI was confirmed by our study.A high level of SII at admission might be an effective predictor of PSCI.Further exploration of the potential mechanism might provide new targets for PSCI treatment.

Patient enrollment
This was a prospective cohort study conducted in the First People's Hospital of Yancheng.The participants were consecutively screened from inpatient department between January 1, 2022, and March 1, 2023.Patients were enrolled if they met the following criteria: (1) ≥ 18 years old, (2) met the World Health Organization diagnostic criteria for AIS confirmed by neuroimaging, (3) admission within seven days of symptom onset, (4) first stroke.Exclusion criteria were as follows: (1) with pre-existing cognitive disorder from diverse diseases, (2) with neurological dysfunctions that may affect cognitive evaluation, such as hearing impairment, aphasia or dysarthria, (3) with diseases which might affect inflammatory conditions, such as blood disease, acute infection, malignant tumors, or trauma, (4) intake of antibiotics, psychotropic or nootropics medications within three months.AGING

Baseline clinical and laboratory data
Clinical and laboratory data were collected in a manner similar to that described in our other article [35].On admission, we used a standard questionnaire to collect clinical data and to assess the presence of pre-existing cognitive disorder.Baseline demographics (gender, age, education, body mass index (BMI), history of smoking and drinking, as well as medical history (diabetes, hypertension, atrial fibrillation, and coronary artery disease) were collected.We classified the patients into three education levels: those with less than one year of education were classified as illiterate, those with one to six years of education were classified as primary school, and those with more than six years were classified as secondary school or above.The etiology and severity of stroke were determined according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria and National Institutes of Health Stroke Scale (NIHSS), respectively.Fasting blood samples were obtained from all patients the morning after admission and processed and recorded by a single laboratory physician.The laboratory data included glycosylated hemoglobin A1 (HbA1c), fasting plasma glucose (FPG), uric acid (UA), peripheral blood cell counts, triglyceride (TG), total cholesterol (TC), low-density lipoprotein (LDL), highdensity lipoprotein (HDL), and homocysteine (Hcy).We calculated SII versus SIRI using the following formula: platelet count × neutrophil count/lymphocyte count, neutrophil count × monocyte count / lymphocyte count, respectively [9,25].All blood laboratory assessments were conducted in the hospital's clinical laboratory.An auto-analyzer (XN-1000, Sysmex, Kobe, Japan) analyzed all blood cell counts, and other profiles were assessed with an automated biochemical analyzer.
All participants underwent brain MRI scans within 72 hours after admission, and imaging data were collected and analyzed by a doctor from the imaging department.The severity of white matter hyperintensities was assessed using the Fezakas scores, which ranges from 0 to 6.The infarct volume was calculated by multiplying the infarct area of each slice by the slice thickness on the DWI sequence and then summing [36].The carotid plaque and stenosis were evaluated by carotid ultrasound or CTA.

Assessments of cognitive function
Cognitive function was assessed at 3 months after stroke by two trained neurologists using the MoCA scale.The total score was 0-30 points, and a score < 26 points was defined as PSCI [37].One point was added to the total score if the patient had less than 12 years of education as MoCA is closely associated with educational level.

Figure 3 .
Figure 3. Violin plot about the distribution of SII and SIRI in the PSCI and nonPSCI subgroups.

Figure 6 .
Figure 6.Receiver operating characteristic (ROC) curve for SII as a predictor of PSCI.
Statistical analyses were conducted by SPSS version 23.0 (IBM, New York, NY, USA) and GraphPad Prism version 8.0.2 (GraphPad Software, San Diego, CA, USA).Continuous variables were presented as the mean±standard deviation or median (interquartile range[IQR]) and categorical variables were presented as numbers (percentages [%]).We compared all characteristics between the PSCI and nonPSCI subgroups, as well as among the SII and SIRI tertiles.The Chi-square test or Fisher's exact test was used for categorical variables (such as sex and medical history), and one-way ANOVA, analysis of variance, the Mann-Whitney U test or Kruskal-Wallis test was used for continuous variables (such as age).The association between SII, SIRI and MoCA score was analyzed by Spearman's correlation.Univariable binary regression analysis was conducted to investigate the association of baseline characteristics with PSCI, and all variables with P < 0.1 were entered into the subsequent multivariable logistic regression model.Odds ratio (OR) or adjusted OR combined with 95% confidence intervals (CIs) demonstrate associations.Subsequently, we evaluated the potential predictive effect of SII on PSCI using ROC curve.All statistical analyses were defined as statistically significant with a two-sided P < 0.05.editing: S.X, S.S, Y.C; funding acquisition: D.S.All authors have read and agreed to the published version of the manuscript.