Association of Inflammatory Markers With Symptomatic Intracranial Hemorrhage, Malignant Cerebral Edema, and Poor Functional Recovery in Patients With Acute Ischemic Stroke Undergoing Endovascular Thrombectomy

Abstract

Background

To investigate the association of inflammatory markers with the risks of clinical outcomes in patients with acute ischemic stroke and large‐vessel occlusion who underwent endovascular thrombectomy treatment.

Methods

A total of 586 patients with acute ischemic stroke and large‐vessel occlusion treated with endovascular thrombectomy between May 2017 and February 2025 were enrolled. Patients were stratified into quartiles according to inflammatory markers measured within 24 hours after endovascular thrombectomy, including neutrophil count, neutrophil‐to‐lymphocyte ratio (NLR), systemic inflammation response index, systemic immune‐inflammation index, inflammatory prognostic index, and aggregate index of systemic inflammation. Logistic regression was used to assess associations between these markers and poor functional outcome (modified Rankin Scale score 3–6) at 90 days, symptomatic intracranial hemorrhage within 24 hours, and malignant cerebral edema within 5 days. Receiver operating characteristic curves evaluated predictive performance.

Results

After adjusting for confounders, participants in the highest quartile of each inflammatory marker had significantly higher odds of poor functional recovery (all P for trend≤0.011), symptomatic intracranial hemorrhage (all P for trend≤0.010), and malignant cerebral edema (all P for trend≤0.005), compared with those in the lowest quartile. Receiver operating characteristic curves showed the NLR had the highest discriminative ability to predict poor functional recovery (area under the curve for NLR, 0.692; P<0.001), symptomatic intracranial hemorrhage (area under the curve for NLR, 0.740; P<0.001) and malignant cerebral edema (area under the curve for NLR, 0.739; P<0.001) compared with other inflammatory markers.

Conclusions

All inflammatory markers, particularly the NLR, were independently associated with an increased risk of poor functional recovery, symptomatic intracranial hemorrhage, and malignant cerebral edema in patients with acute ischemic stroke and large‐vessel occlusion after endovascular thrombectomy.

Nonstandard Abbreviations and Acronyms

AIS

acute ischemic stroke

AISI

aggregate index of systemic inflammation

EVT

endovascular thrombectomy

IPI

inflammatory prognostic index

LVO

large‐vessel occlusion

NIHSS

National Institutes of Health Stroke Scale

NLR

neutrophil‐to‐lymphocyte ratio

sICH

symptomatic intracerebral hemorrhage

SIRI

systemic inflammation response index

SII

systemic immune‐inflammation index

TOAST

Trial of ORG 10172 in Acute Stroke Treatment

Clinical Perspective.

What Is New?

• This study is the first to comprehensively evaluate and demonstrate that systemic inflammatory markers (systemic inflammation response index, inflammatory prognostic index, systemic immune‐inflammation index, neutrophil‐to‐lymphocyte ratio, neutrophil count, and aggregate index of systemic inflammation), particularly neutrophil‐to‐lymphocyte ratio, are significantly associated with an increased risk of poor functional recovery, symptomatic intracranial hemorrhage, and malignant cerebral edema in patients with acute ischemic stroke due to large‐vessel occlusion who underwent endovascular thrombectomy.

What Are the Clinical Implications?

• Systemic inflammatory markers, particularly the neutrophil‐to‐lymphocyte ratio, may help identify patients with acute ischemic stroke at higher risk of complications and poor outcomes after endovascular thrombectomy, thereby supporting early risk stratification and clinical decision making. The results highlight the potential utility of routine inflammatory marker assessment in both clinical practice and future research to improve prognosis prediction and personalize management strategies for patients with acute ischemic stroke undergoing endovascular thrombectomy.

Endovascular thrombectomy (EVT) is the recommended and most effective treatment for a broad range of patients with acute ischemic stroke (AIS) due to large‐vessel occlusion (LVO) within 24 hours of symptom onset. ¹ Although successful reperfusion can be achieved after EVT, 50% to 60% of patients with AIS and LVO experience disability or death at 90 days, partly due to symptomatic intracranial hemorrhage (sICH) and malignant cerebral edema. ² , ³ , ⁴ Therefore, early identification of risk factors associated with sICH, malignant cerebral edema, and poor functional recovery after EVT is extremely important for further interventions.

An increased inflammatory response is a key pathophysiological mechanism after cerebral ischemia and is significantly associated with a higher risk of disability and death in patients with AIS. ⁵ , ⁶ , ⁷ , ⁸ Several simple and combined inflammatory biomarkers, such as neutrophil count, neutrophil‐to‐lymphocyte ratio (NLR), systemic inflammation response index (SIRI), systemic immune‐inflammation index (SII), inflammatory prognostic index (IPI), and aggregate index of systemic inflammation (AISI), have been reported to be associated with an increased risk of poor functional outcomes after AIS. ⁵ , ⁶ , ⁷ , ⁸

However, there is a lack of evidence comprehensively evaluating the predictive effect of these inflammatory biomarkers on poor functional recovery in patients with AIS and LVO undergoing EVT. Moreover, whether all these inflammatory biomarkers are also associated with sICH and malignant cerebral edema remains unclear. In addition, the comparative predictive value of these inflammatory biomarkers for poor functional recovery, sICH, and malignant cerebral edema have not been studied.

Herein, our study aimed to evaluate and compare the predictive capability of these inflammatory biomarkers for sICH, malignant cerebral edema, and 90‐day functional recovery in patients with AIS treated with EVT.

METHODS

Data Availability Statement

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

Study Design and Participants

This study retrospectively analyzed the clinical data of patients with AIS due to LVO who underwent EVT in the Second Affiliated Hospital of Soochow University from May 2017 to February 2025. In brief, the criteria for our study were as follows: (1) aged >18 years old; (2) LVO in the anterior or posterior circulation confirmed by computed tomographic (CT) angiography or digital subtraction angiography; (3) time from symptom onset to treatment within 24 hours, with patients presenting with wake‐up stroke or within 6 to 24 hours from last known well required to demonstrate salvageable tissue mismatch on CT angiography and CT perfusion; and (4) treated with mechanical thrombectomy. Exclusion criteria included (1) lack of 90‐day follow‐up modified Rankin Scale data; (2) incomplete inflammatory data that prevented calculation of combined inflammatory biomarkers, including SIRI, IPI, SII, and AISI.

This retrospective study was approved by the Institutional Review Board of the Second Affiliated Hospital of Soochow University (JD‐HG‐2025‐073). Given the retrospective design and use of deidentified data, the requirement for informed consent was waived by the Institutional Review Board.

Data Collection

The following variables were systematically collected and reported: baseline demographics (age, sex), medical history (hypertension, diabetes, atrial fibrillation, stroke), lifestyle factors (smoking, alcohol consumption), clinical characteristics (TOAST [Trial of ORG 10172 in Acute Stroke Treatment] classification, National Institutes of Health Stroke Scale [NIHSS] score, Alberta Stroke Program Early CT Score), laboratory parameters (neutrophil, monocyte, lymphocyte, and platelet counts; albumin; glucose; lipids; creatinine; CRP [C‐reactive protein]), interventional details (thrombolysis administration, occlusion site, onset‐to‐puncture time, onset‐to‐recanalization time, successful reperfusion is defined as achieving a modified Thrombolysis in Cerebral Infarction score of 2b–3 and recanalization rate), and clinical outcomes (sICH, malignant cerebral edema, and 90‐day functional outcome).

Finger‐stick blood glucose and blood pressure were measured before the EVT procedure. Blood samples were collected and measured within 24 hours after EVT. Neutrophil, monocyte, lymphocyte, and platelet counts were measured using the XN‐3000 (Sysmex) or BC‐6800 (Mindray) automated hematology analyzer. CRP, albumin, and other blood sample levels were determined using the cobas c502 or c702 analyzer (Roche). All data were rigorously assessed by trained clinical specialists to ensure accuracy and consistency.

Inflammatory Indicators

Absolute neutrophil counts were obtained from routine laboratory tests. Other inflammatory markers or scores (NLR, SIRI, IPI, SII, AISI) were calculated from admission laboratory data using standard formulas: NLR=neutrophil count/lymphocyte count, ⁹ SIRI=neutrophil count×(monocyte count/lymphocyte count), ¹⁰ IPI=CRP×NLR/albumin, ¹¹ SII=platelet count×NLR, ¹² AISI=neutrophil count×platelet count×(monocyte count/lymphocyte count). ¹³

Study Outcomes

The primary outcome was poor functional recovery, defined as a modified Rankin Scale score of 3 to 6 at 90 days. The secondary outcomes include (1) sICH, defined according to European Cooperative Acute Stroke Study III criteria as the presence of extravascular blood in the cranium that was associated with an increase in the NIHSS score of ≥4 points within 24 hours after EVT ¹⁴ ; (2) malignant cerebral edema, defined as a midline shift of ≥5 mm in the plane of the septum pellucidum or pineal gland on follow‐up brain CT in anterior circulation within 5 days after EVT. ¹⁵

Statistical Analysis

Continuous variables were expressed as mean±SD or median (interquartile range) and were compared using the analysis of independent Student's t test or Wilcoxon rank‐sum test. Categorical variables were expressed as frequency (percentage) and were compared using the χ² test.

Patients were divided into 4 groups according to quartiles of each inflammatory marker. Receiver operating characteristic analyses were used to assess the predictive performance of inflammatory biomarkers (neutrophil count, NLR, SIRI, SII, IPI, and AISI) for poor functional recovery at 90 days, sICH, and malignant cerebral edema, and the area under the curve (AUC) analyses were compared using the Delong test. The associations between inflammatory biomarkers and poor functional recovery at 90 days, sICH, and malignant cerebral edema were assessed using logistic regression models. The covariates included in the multivariable models were selected on the basis of clinical relevance and prior literature including age, sex, hypertension, diabetes, atrial fibrillation, smoking, alcohol consumption, TOAST classification, prestroke modified Rankin Scale score, baseline NIHSS score, preoperative glucose, systolic blood pressure, thrombolysis, history of stroke, Alberta Stroke Program Early CT Score, successful reperfusion, intracranial hemorrhage (no intracranial hemorrhage, asymptomatic intracranial hemorrhage, and sICH) and pneumonia. For the logistic regression model with sICH as the outcome, intracranial hemorrhage and pneumonia were not included as covariates. For the model with malignant cerebral edema as the outcome, pneumonia was not included in the adjustment. Odds ratios (ORs) and 95% CIs for inflammatory biomarker quartiles and per 1 SD increment were calculated. Tests for linear trends in ORs across inflammatory biomarker quartiles were conducted using the median value of each quartile as the predictor. Furthermore, to assess the robustness of the associations between each inflammatory marker and clinical outcomes, sensitivity analyses were performed by categorizing patients into 2 groups on the basis of the median value of each marker.

We further assessed the predictive performance of inflammatory biomarkers for clinical outcomes. Receiver operating characteristic curve analyses were performed, and the AUC was compared between the baseline model (including conventional risk factors) and models with the addition of each inflammatory biomarker using the DeLong test. In addition, net reclassification index and integrated discrimination improvement were used to evaluate improvements in risk reclassification of models with inflammatory biomarkers added to conventional risk factors. All P values were 2‐tailed, and a significance level of 0.05 was used. Statistical analysis was conducted using R statistical software version 4.4.3 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Clinical Characteristics of the Study Cohort

Among 687 patients with AIS due to LVO who underwent EVT, 68 were excluded due to missing inflammatory marker data, and 33 were excluded due to lack of 90‐day follow‐up data. Ultimately, 586 patients with AIS (mean age, 65.7±13.5 years; 214 [37.0%] women) who had available data were included in the analyses of functional recovery and sICH, and 499 patients with anterior circulation AIS were included in the analysis of malignant cerebral edema (patient flowchart in Figure S1). Of the 586 patients with AIS, 209 (35.7%) achieved good functional recovery, while 377 (64.3%) had poor functional recovery at 90 days. Table 1 outlines the baseline characteristics of participants with good and poor functional recovery at 90 days. In comparison with participants with good functional recovery, those with poor functional recovery tended to be older and women and had more severe stroke (higher NIHSS score); more comorbidities including hypertension, diabetes, atrial fibrillation, prior stroke, higher systolic blood pressure, higher glucose levels; and elevated inflammatory markers including CRP, SIRI, IPI, SII, NLR, and AISI. Additionally, they were less likely to be smokers, consume alcohol, or receive thrombolysis treatment. Patients with poor functional recovery also had lower Alberta Stroke Program Early CT Score scores, were less likely to achieve good reperfusion after EVT, and were more likely to develop sICH and pneumonia.

Table 1.

Demographics and Clinical Characteristics of Patients With AIS by 90‐Day Functional Recovery

Characteristic	Overall (N=586)	Good functional recovery (N=209)	Poor functional recovery (N=377)	P value
Age, y	65.7±13.5	61.5±14.5	68.0±12.2	<0.001
Female sex	214 (37)	63 (30)	151 (40)	0.022
Medical history
Hypertension	403 (69)	132 (63)	271 (72)	0.037
Diabetes	127 (22)	29 (14)	98 (26)	<0.001
Atrial fibrillation	237 (40)	72 (34)	165 (44)	0.035
Stroke	88 (15)	20 (10)	68 (18)	0.009
Smoking	211 (36)	91 (44)	120 (32)	0.006
Alcohol consumption	159 (27)	71 (34)	88 (23)	0.007
Clinical features
TOAST pathogenesis				0.008
Large‐artery atherosclerosis	288 (49)	99 (47)	189 (50)
Cardioembolism	263 (45)	89 (43)	174 (46)
Others	35 (6)	21 (10)	14 (4)
Prestroke modified Rankin Scale score				0.5
0	518 (88)	191 (91)	327 (87)
1	47 (8)	12 (6)	35 (9)
Baseline NIHSS	17.0 (13.0–20.0)	14.0 (11.0–18.0)	18.0 (15.0–22.0)	<0.001
ASPECTS	7.0 (7.0–8.0)	8.0 (7.0–8.0)	7.0 (6.0–8.0)	<0.001
Glucose, mmol/L	7.8 (6.7–9.4)	7.2 (6.3–8.4)	8.0 (6.8–9.9)	<0.001
Systolic blood pressure, mm Hg	148.7±23.7	144.5±22.6	151.1±24.0	<0.001
Diastolic blood pressure, mm Hg	86.9±16.3	85.1±15.2	88.6±28.3	0.039
Total cholesterol, mmol/L	4.2 (3.5–4.9)	4.1 (3.5–4.8)	4.3 (3.6–5.0)	0.073
Triglycerides, mmol/L	1.0 (0.8–1.4)	1.0 (0.8–1.5)	1.0 (0.7–1.4)	0.078
CRP, mg/L	5.7 (3.4–13.8)	5.4 (2.5–11.0)	6.0 (3.9–17.6)	0.001
Creatinine, μmol/L	66.0 (56.0–80.0)	66.0 (56.0–80.0)	66.0 (56.0–80.0)	0.7
Inflammatory indicators
SIRI	3.7 (2.2–6.8)	2.9 (1.8–4.8)	4.4 (2.5–7.8)	<0.001
IPI	1.3 (0.6–3.8)	0.8 (0.4–1.9)	1.7 (0.7–4.5)	<0.001
SII	1618.2 (966.6–2614.1)	1197.7 (844.6–1931.3)	1913.5 (1132.2–3195.5)	<0.001
NLR	8.6 (5.4–13.8)	6.3 (4.1–10.3)	9.8 (6.6–16.2)	<0.001
Neutrophil count, ×109/L	8.3 (6.3–10.8)	7.2 (5.6–9.0)	9.0 (6.8–11.8)	<0.001
AISI	703.7 (364.8–1374.5)	536.2 (309.3–975.0)	854.0 (436.5–1660.7)	<0.001
Procedure related
Thrombolysis	194 (33)	83 (40)	111 (29)	0.015
Occlusion site				0.2
Internal carotid artery	138 (24)	42 (20)	96 (25)
Middle cerebral artery	330 (56)	129 (62)	201 (53)
Posterior circulation	87 (15)	27 (13)	60 (16)
Others	31 (5)	11 (5)	20 (5)
Onset‐to‐puncture time, min	304.0 (230.0–412.0)	305.0 (230.0–410.0)	299.0 (230.0–414.0)	0.9
Onset‐to‐recanalization time, min	375.0 (295.0–508.0)	365.0 (287.0–470.0)	382.5 (299.5–520.0)	0.2
mTICI ≥2b	538 (92)	201 (96)	337 (89)	0.007
Complications
sICH	102 (17)	4 (2)	98 (26)	<0.001
Pneumonia	315 (54)	65 (31)	250 (66)	<0.001
Malignant cerebral edema*	121 (24)	12 (7)	109 (34)	<0.001

Age, systolic blood pressure, and diastolic blood pressure are presented as mean±SD; other continuous variables are presented as median with interquartile range; categorical variables are presented as number (percentage). AIS indicates acute ischemic stroke; AISI, aggregate index of systemic inflammation; ASPECTS, Alberta Stroke Program Early CT Score; CRP, C‐reactive protein; IPI, inflammatory prognostic index; mTICI, modified Thrombolysis in Cerebral Infarction; NIHSS, National Institutes of Health Stroke Scale; NLR, neutrophil‐to‐lymphocyte ratio; sICH, symptomatic intracranial hemorrhage; SII, systemic immune‐inflammation index; SIRI, systemic inflammation response index; and TOAST, Trial of ORG 10172 in Acute Stroke Treatment.

^*Proportion of malignant cerebral edema in patients with anterior circulation infarction.

Baseline Inflammatory Markers and Poor Functional Recovery at 90 Days After EVT

Figure 1 shows that the risk of poor functional recovery at 90 days increased across quartiles of all inflammatory markers, including SIRI, IPI, SII, NLR, neutrophil count, and AISI. After adjusting for age, sex, prestroke modified Rankin Scale score, baseline NIHSS score, pneumonia, sICH, and other potential confounders, participants in the highest quartile of each inflammatory marker had significantly higher odds of poor functional recovery compared with those in the lowest quartile: SIRI (OR, 2.82 [95% CI, 1.32–6.20]), IPI (OR, 2.50 [95% CI, 1.19–5.35]), SII (OR, 2.81 [95% CI, 1.36–5.97]), NLR (OR, 3.81 [95% CI, 1.80–8.27]), neutrophil count (OR, 2.87 [95% CI, 1.36–6.22]), and AISI (OR, 3.33 [95% CI, 1.57–7.29]), respectively. Moreover, each inflammatory marker was associated with a dose‐dependent increase in the risk of poor functional recovery. For every SD increase in SIRI, IPI, SII, NLR, neutrophil count, and AISI, the risk of poor functional recovery increased by 1.87‐fold, 1.46‐fold, 2.17‐fold, 2.04‐fold, 1.44‐fold, and 2.07‐fold, respectively (Table S1). In addition, when participants were categorized into 2 groups according to the median value of each inflammatory marker, we found that higher levels of these markers were still associated with an increased risk of poor functional recovery compared with lower levels (Table S2).

Figure 1. Baseline inflammatory markers and poor functional recovery after EVT in patients with AIS.

Models adjusted for age, sex, hypertension, diabetes, atrial fibrillation, smoking, alcohol consumption, TOAST classification, prestroke modified Rankin Scale score, baseline NIHSS score, preoperative glucose, systolic blood pressure, thrombolysis, history of stroke, ASPECTS, successful reperfusion, intracranial hemorrhage, and pneumonia. AIS indicates acute ischemic stroke; AISI, aggregate index of systemic inflammation; ASPECTS, Alberta Stroke Program Early Computed Tomography Score; EVT, endovascular thrombectomy; IPI, inflammatory prognostic index; NIHSS, National Institutes of Health Stroke Scale; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; SII, systemic immune‐inflammation index; SIRI, systemic inflammation response index; and TOAST, Trial of ORG 10172 in Acute Stroke Treatment.

Baseline Inflammatory Markers and sICH After EVT

Overall, 102 patients (17.4%) experienced sICH within 24 hours after EVT. Figure 2 shows that the rates of sICH increased across quartiles of all inflammatory markers. Participants in the highest quartile of each inflammatory marker had significantly higher risk of sICH compared with those in the lowest quartile: SIRI (OR, 3.93 [95% CI, 1.82–9.07]), IPI (OR, 2.22 [95% CI, 1.10–4.65]), SII (OR, 6.30 [95% CI, 2.83–15.25]), NLR (OR, 7.30 [95% CI, 3.21–18.46]), neutrophil count (OR, 4.59 [95% CI, 2.11–10.79]), and AISI (OR, 2.97 [95% CI, 1.43–6.44]), respectively. Additionally, the association between elevated inflammatory markers and the risk of sICH remained consistent in sensitivity analyses in which each marker was analyzed per SD increase (Table S3) and when participants were categorized into 2 groups on the basis of the median value of each inflammatory marker (Table S4).

Figure 2. Baseline inflammatory markers and sICH after EVT in patients with AIS.

Models adjusted for age, sex, hypertension, diabetes, atrial fibrillation, smoking, alcohol consumption, TOAST classification, prestroke modified Rankin Scale score, baseline NIHSS score, preoperative glucose, systolic blood pressure, thrombolysis, history of stroke, ASPECTS, and successful reperfusion. AIS indicates acute ischemic stroke; AISI, aggregate index of systemic inflammation; ASPECTS, Alberta Stroke Program Early Computed Tomography Score; EVT, endovascular thrombectomy; IPI, inflammatory prognostic index; NIHSS, National Institutes of Health Stroke Scale; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; sICH, symptomatic intracranial hemorrhage; SII, systemic immune‐inflammation index; SIRI, systemic inflammation response index; and TOAST, Trial of ORG 10172 in Acute Stroke Treatment.

Baseline Inflammatory Markers and Malignant Cerebral Edema After EVT

Among 499 patients with anterior circulation AIS, 121 (24.2%) developed malignant cerebral edema within 5 days after EVT. Figure 3 indicates that the frequency of malignant cerebral edema increased across quartiles of all inflammatory markers. Patients in the highest quartile of each inflammatory marker had significantly higher odds of malignant cerebral edema compared with those in the lowest quartile: SIRI (OR, 5.58 [95% CI, 2.35–14.01]), IPI (OR, 2.70 [95% CI, 1.18–6.31]), SII (OR, 4.79 [95% CI, 1.98–12.43]), NLR (OR, 4.31 [95% CI, 1.71–11.65]), neutrophil count (OR, 5.94 [95% CI, 2.42–15.73]), and AISI (OR, 8.01 [95% CI, 3.28–21.07]), respectively. Furthermore, each inflammatory marker showed a significant association with increased risk of malignant cerebral edema when analyzed as a continuous variable (per SD increase) (Table S5), and when dichotomized at the median (Table S6).

Figure 3. Baseline inflammatory markers and malignant cerebral edema after EVT in patients with AIS.

Models adjusted for age, sex, hypertension, diabetes, atrial fibrillation, smoking, alcohol consumption, TOAST classification, prestroke modified Rankin Scale score, baseline NIHSS score, preoperative glucose, systolic blood pressure, thrombolysis, history of stroke, ASPECTS, successful reperfusion, and intracranial hemorrhage. AIS indicates acute ischemic stroke; AISI, aggregate index of systemic inflammation; ASPECTS, Alberta Stroke Program Early Computed Tomography Score; EVT, endovascular thrombectomy; IPI, inflammatory prognostic index; NIHSS, National Institutes of Health Stroke Scale; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; SII, systemic immune‐inflammation index; SIRI, systemic inflammation response index; and TOAST, Trial of ORG 10172 in Acute Stroke Treatment.

Compare the Predicted Effect of Inflammatory Markers on Clinical Outcomes

Receiver operating characteristic curves and AUC showed the NLR had the highest discriminative ability to predict poor functional recovery at 90 days compared with SII, neutrophil count, IPI, SIRI, and AISI (AUC, 0.692 versus 0.668, 0.667, 0.655, 0.643, and 0.623, respectively; P<0.001) (Figure 4A). Moreover, the discriminative ability of NLR to predict sICH (AUC 0.740 versus 0.693, 0.689, 0.669, 0.629, and 0.612, respectively; P<0.001) (Figure 4B) as well as malignant cerebral edema (AUC, 0.739 versus 0.735, 0.705, 0.700, 0.672, and 0.628, respectively; P<0.001) (Figure 4C) was superior to that of other inflammatory markers.

Figure 4. Receiver operating characteristic curves of inflammatory indicators for predicting clinical outcomes following EVT in patients with AIS.

A, Receiver operating characteristic curves of inflammatory indicators for predicting poor functional recovery at 90 days. B, Receiver operating characteristic curves of inflammatory indicators for predicting symptomatic intracranial hemorrhage. C, Receiver operating characteristic curves of inflammatory indicators for predicting malignant cerebral edema. AISI indicates aggregate index of systemic inflammation; AUC, area under the curve; IPI, inflammatory prognostic index; NLR, neutrophil‐to‐lymphocyte ratio; SII, systemic immune‐inflammation index; and SIRI, systemic inflammation response index.

Incremental Prognostic Value of Inflammatory Markers

We found that the addition of each inflammatory marker to the basic model containing conventional risk factors significantly improved risk reclassification for poor functional recovery at 90 days (Table 2), sICH (Table S7), and malignant cerebral edema (Table S8). In addition, compared with the baseline model, incorporating individual inflammatory biomarkers yielded varying degrees of improvement in AUC for predicting poor functional recovery, sICH, and malignant cerebral edema. Significant improvements were observed with NLR and SII for poor functional recovery, whereas nearly all biomarkers significantly improved prediction for sICH and malignant cerebral edema (Figure S2 and Table S9).

Table 2.

The Net Reclassification Index and Integrated Discrimination Improvement Estimate of Inflammatory Indicators for Poor Functional Recovery

Variables	Net Reclassification Index		Integrated Discrimination Improvement
	Estimate (95% CI), %	P value	Estimate (95% CI), %	P value
Basic model	Reference		Reference
Basic model+SIRI	33.57 (15.14–52.01)	<0.001	3.03 (1.57–4.49)	<0.001
Basic model+IPI	26.49 (9.99–42.98)	0.002	1.25 (0.45–2.06)	0.002
Basic model+SII	56.45 (38.61–74.28)	<0.001	5.10 (3.22–6.98)	<0.001
Basic model+NLR	44.08 (25.70–62.46)	<0.001	4.43 (2.59–6.28)	<0.001
Basic model+neutrophil count	41.97 (23.00–60.93)	<0.001	4.38 (2.54–6.22)	<0.001
Basic model+AISI	41.24 (22.85–59.62)	<0.001	3.62 (2.04–5.21)	<0.001

The basic model refers to the multivariable logistic regression model adjusted for age, gender, hypertension, diabetes, atrial fibrillation, smoking, alcohol consumption, TOAST classification, prestroke modified Rankin Scale score, baseline NIHSS score, preoperative glucose, systolic blood pressure, thrombolysis, history of stroke, ASPECTS, successful reperfusion, intracranial hemorrhage and pneumonia. AISI indicates aggregate index of systemic inflammation; ASPECTS, Alberta Stroke Program Early Computed Tomography Score; IPI, inflammatory prognostic index; NIHSS, National Institutes of Health Stroke Scale; NLR, neutrophil‐to‐lymphocyte ratio; SII, systemic immune‐inflammation index; and SIRI, systemic inflammation response index.

DISCUSSION

In this study of 586 patients with AIS due to LVO who underwent EVT, we found that all inflammatory markers, including NLR, SIRI, IPI, SII, neutrophil count, and AISI, were significantly associated with an increased risk of poor functional recovery at 90 days, sICH, and malignant cerebral edema. Of these, NLR demonstrated the strongest discriminative ability for predicting poor functional recovery at 90 days, sICH, and malignant cerebral edema, outperforming all other inflammatory markers. Moreover, the inclusion of each inflammatory marker in a model with established risk factors significantly improved risk discrimination and reclassification of the study outcomes.

Given the critical role of inflammation after acute stroke, numerous inflammatory markers, such as neutrophil count, NLR, SII, SIRI, IPI, and AISI, have been investigated as potential predictors of poor functional recovery after stroke. ⁶ , ⁷ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² A study of 190 patients with AIS who underwent intravenous thrombolysis treatment indicated that higher levels of NLR, SII, SIRI, and IPI at baseline were significantly associated with poor functional recovery at 90 days. ⁶ Given the heightened inflammatory response observed in patients with AIS and LVO, a growing body of evidence has reported that ≥1 inflammatory markers are independently associated with death or disability in patients undergoing EVT. ⁵ , ²³ , ²⁴ , ²⁵ However, there remains a lack of comprehensive evidence evaluating and comparing the predictive value of these inflammatory markers for functional recovery after EVT. In the present analysis, we found that elevated baseline levels of neutrophil count, NLR, SII, SIRI, IPI, and AISI were all significantly associated with increased odds of poor functional recovery at 90 days among patients with AIS undergoing EVT. Moreover, NLR demonstrated the strongest discriminative ability among these markers in predicting unfavorable outcomes. Compared with the lowest quartile of NLR, patients in the highest quartile had nearly 4‐fold higher risk of poor functional recovery at 90 days. These findings, which are consistent with prior studies, further support the association between heightened inflammation and poor recovery after EVT and suggest that NLR may be the most reliable inflammatory marker for predicting post‐EVT functional outcomes.

Malignant cerebral edema and sICH are severe and potentially fatal complications in patients with AIS following EVT, each independently associated with poor functional outcomes. Several studies have investigated the relationship between inflammatory markers and the risk of sICH among patients with AIS. ⁸ , ²⁶ , ²⁷ , ²⁸ , ²⁹ For example, a recent cohort study conducted in Canada reported that elevated baseline neutrophil counts were independently associated with post‐EVT hemorrhage. ²⁶ Other studies have similarly demonstrated that higher levels of NLR, SII, and SIRI are independently associated with the occurrence of sICH after AIS. ⁸ , ²⁷ , ²⁸ , ²⁹ Some studies have identified elevated NLR and SII as significant predictors of malignant cerebral edema in patients with AIS. ²⁴ , ²⁵ , ³⁰ However, the associations between other inflammatory markers and the risk of sICH or malignant cerebral edema remain unclear, and their predictive performance for these outcomes has not been systematically compared. In our analysis, all assessed inflammatory markers (SIRI, IPI, SII, NLR, neutrophil count, and AISI) were independently associated with an increased risk of sICH and malignant cerebral edema among patients with AIS after EVT. Receiver operating characteristic curve analysis further demonstrated that NLR had the strongest predictive performance for both outcomes compared with the other markers. Compared with the lowest quartile of NLR, patients in the highest quartile had a 7.30‐fold higher risk of sICH and a 4.31‐fold higher risk of malignant cerebral edema. Together with previous studies, our findings provide stronger evidence supporting the link between inflammation and the development of malignant cerebral edema and sICH after EVT. In addition, our results suggest that NLR may be the most effective inflammatory marker for predicting these complications in patients with AIS following EVT.

In our study, NLR showed the strongest discriminative ability among inflammatory markers for all clinical outcomes. The AUCs of NLR for predicting 90‐day poor functional outcome, sICH, and malignant cerebral edema were 0.692, 0.740, and 0.739, respectively. These values are comparable with or slightly higher than those reported in previous studies. A Portuguese study of 553 patients with AIS treated with intravenous thrombolysis or EVT reported AUCs of 0.70 for 90‐day poor outcome and 0.71 for cerebral edema, ²⁴ while another study of 187 patients with AIS and LVO undergoing EVT showed an AUC of 0.67 for baseline NLR in predicting ICH. ⁵ Together, these findings further support the predictive value of NLR in EVT‐treated patients.

Inflammation plays a critical role in secondary brain injury after ischemic stroke. ³¹ , ³² Excessive poststroke inflammatory responses, characterized by the release of proinflammatory mediators and infiltration of activated immune cells into the brain following ischemia–reperfusion, may amplify tissue damage. ³² , ³³ Elevated inflammatory markers may reflect enhanced neutrophil activation and relative lymphocyte suppression, which have been implicated in blood–brain barrier disruption, microvascular injury, and secondary brain damage. ³² , ³³ , ³⁴ In addition, activation of microglia and recruited neutrophils can further increase blood–brain barrier permeability, exacerbate excitotoxicity and oxidative stress, induce neuronal death, and promote infarct expansion. ³⁵ , ³⁶ , ³⁷ These processes may aggravate endothelial dysfunction and vascular permeability, thereby predisposing patients to sICH and malignant cerebral edema. ³⁸ Although these mechanisms are biologically plausible and supported by experimental evidence, they were not directly assessed in our study and should therefore be considered hypothesis generating.

The strengths of our study include a comprehensive assessment of multiple inflammatory markers in relation to functional recovery, sICH, and malignant cerebral edema in patients with AIS who underwent EVT. Moreover, we compared the predictive performance of these markers and found that baseline NLR demonstrated the strongest and most consistent predictive value across all outcomes, outperforming the other inflammatory markers. With regard to clinical applicability, the inflammatory markers evaluated, particularly NLR, are derived from routine blood tests that are inexpensive, rapidly available, and easy to calculate. Therefore, they can be readily incorporated into early risk stratification. Given its strong discriminative ability and incremental predictive value beyond established risk factors, NLR may serve as a simple and practical tool to help identify high‐risk patients after EVT and guide closer monitoring and individualized management.

However, our study has several limitations. First, this was a single‐center design study, and our findings require validation in multicenter cohorts. Second, the retrospective observational design may introduce selection bias and limit the accuracy and generalizability of the results. Third, the relatively small sample size and quartile stratification may have led to wider CIs. However, the results remained consistent in sensitivity analyses using both per‐SD increases and median‐based categorization of inflammatory markers. In addition, patients receiving more aggressive management may survive long enough for malignant cerebral edema to be detected and documented, which could introduce selection bias toward poorer prognosis. Finally, we lacked serial measurement of biomarkers, which prevented us from evaluating the dynamic changes in inflammatory responses and their association with clinical outcomes.

CONCLUSIONS

In summary, we found that all inflammatory markers, particularly the NLR, were independently associated with an increased risk of poor functional recovery, sICH, and malignant cerebral edema in patients with AIS and LVO following EVT. These findings reinforce the link between inflammation and both poor outcomes and severe complications after EVT. Our results suggest that NLR may serve as a valuable inflammatory marker for predicting clinical outcomes in this population. Further large, multicenter studies are needed to validate and enhance the generalizability of our findings.

Sources of Funding

This work was supported by the National Natural Science Foundation of China (82471226), Discipline Construction Program of the Second Affiliated Hospital of Soochow University (XKTJ‐XK202401 and XKTJ‐RC202412), Jiangsu Provincial Medical Key Discipline (ZDXK202217), the Sixth Jiangsu Province 333 High Level Talents Training Project, and the Suzhou Major Disease Multicenter Clinical Research Project (DZXYJ202403).

Disclosures

None.

Supporting information

Tables S1–S9

Figures S1–S2

STROBE Checklist