Progression of Neuroinflammation Is Associated With Clinical Prognosis of Patients Undergoing Intravenous Thrombolysis

Abstract

BACKGROUND:

Recent research has shown that neuroinflammation progresses rapidly within a few hours after stroke; however, the relationship between its progression and clinical outcomes remains unclear. Therefore, this study aimed to investigate the effect of neuroinflammation, measured by serum GFAP (glial fibrillary acidic protein), on patient outcomes, as well as the influence of baseline peripheral inflammation on the progression of neuroinflammation.

METHODS:

This prospective cohort study enrolled patients with acute ischemic stroke who received intravenous thrombolysis (IVT) between September 2016 and April 2023 across 16 centers in China. Serum GFAP levels were measured before (baseline, within 4.5 hours of onset) and at 24 hours after IVT. GFAP changes were determined by subtracting baseline levels from those measured 24 hours post-IVT. Outcome measures included final infarct volume during hospitalization, National Institutes of Health Stroke Scale scores at 24 hours and 7 days post-IVT, early neurological deterioration within 24 hours, delayed neurological deterioration within 7 days, and 3-month modified Rankin Scale scores. A modified Rankin Scale score of ≥2 was classified as an unfavorable outcome. Peripheral inflammation indicators were measured at baseline. Binary logistic and linear regressions were used as the main statistical methods.

RESULTS:

Overall, 743 patients were included. A significant increase in GFAP levels was observed, indicating progression of neuroinflammation. Regression analyses revealed that increased GFAP after IVT was independently associated with larger infarct volume (β, 30.965 [95% CI, 19.185–42.745]; P<0.001), higher 24-hour and 7-day National Institutes of Health Stroke Scale scores (24-hour: β, 2.632 [95% CI, 1.644–3.620]; P<0.001; 7-day: β, 3.298 [95% CI, 2.179–4.417]; P<0.001), and unfavorable outcomes (odds ratio, 3.631, [95% CI, 2.159–6.106]; P<0.001). Furthermore, baseline peripheral inflammation, assessed using peripheral inflammation indicators, was significantly associated with elevated GFAP levels.

CONCLUSIONS:

The increase in GFAP levels over the first 24 hours after IVT is independently associated with clinical outcomes, with higher baseline peripheral inflammation correlating with greater GFAP elevation during that period.

Ischemic stroke is a significant global health burden characterized by high incidence rates, substantial disability, and limited treatment options.¹ Timely administration of intravenous thrombolysis (IVT) facilitates vessel recanalization and functional recovery, thereby significantly improving patient outcomes.^2,3 However, the prognosis of patients post-IVT remains uncertain, with only a subset experiencing substantial functional improvement. Furthermore, predicting the prognosis of their conditions is challenging.⁴

Neuroinflammation, induced by ischemic events, is pivotal in the pathophysiology of stroke and influences both progression and recovery.^5,6 Resident immune cells, including microglia and astrocytes, rapidly respond to ischemic stimuli and undergo polarization.^7,8 These cells recruit peripheral immune cells (including neutrophils, monocytes, dendritic cells, and lymphocytes) to the lesion site by releasing inflammatory factors.^8,9 This recruitment contributes to local secondary inflammatory damage during ischemic stroke.⁹ Extensive animal studies have suggested that increased neuroinflammation leads to more severe neural damage.⁹ However, current noninvasive methods for assessing the neuroinflammatory state of patients present significant challenges, thereby limiting clinical research.

The neuroinflammatory cascade is initiated within minutes of a stroke event.¹⁰ Recent studies have indicated that within 36 minutes of symptom onset, significant elevations in neuroinflammatory markers, such as GFAP (glial fibrillary acidic protein), are detectable in the peripheral blood of patients, escalating rapidly to >10× the normal levels within ≈2 hours.¹¹ Nevertheless, the implications of this progression on prognosis remain unclear. GFAP is an intermediate filament protein predominantly located in the astrocytes of the central nervous system¹² and exhibits elevated levels in peripheral blood during neuroinflammation.¹³ Therefore, this study used changes in GFAP levels in peripheral blood to monitor the progression of neuroinflammation and explore its association with outcomes after IVT in a large sample cohort.

In addition, to explore the potential factors underlying the changes in neuroinflammation, we used markers including leukocyte count, neutrophil count, neutrophil-to-lymphocyte ratio, systemic inflammation response index, and systemic immune-inflammation index to characterize peripheral inflammation and assess their association with neuroinflammation.

Methods

This multicenter, prospective, observational cohort study followed the STROBE guidelines (Strengthening the Reporting of Observational Studies in Epidemiology¹⁴; Supplemental Material) and was approved by the ethics committee of the First Hospital of Jilin University (2015-156). All participants provided written informed consent and had the right to withdraw from the study at any point. The data sets used and analyzed during the current study are available from the corresponding author on reasonable request.

Participants

Consecutive patients who met the inclusion and exclusion criteria between September 2016 and April 2023 at The First Hospital of Jilin University and between September 2021 and January 2022 at the other 15 hospitals in Northeast China were included (Table S1). The inclusion criteria were as follows: (1) aged >18 years, regardless of sex; (2) diagnosis of ischemic stroke; (3) having received standard alteplase treatment within 4.5 hours after stroke onset; (4) a modified Rankin Scale (mRS) score ≤1 before the stroke; and (5) agreement to participate in the study and written informed consent. The exclusion criteria were as follows: (1) lacking a complete set of pre- and post-IVT measurements for serum GFAP levels; and (2) incomplete baseline (before IVT/within 4.5 hours after stroke onset) leukocyte, neutrophil, monocyte, lymphocyte, and platelet count measurements.

Clinical Parameters

Baseline data included demographic characteristics (name, age, and sex), risk factors (cigarette smoking, hypertension, diabetes, dyslipidemia, hyperhomocysteinemia, previous ischemic stroke, and coronary heart disease^15,16), and clinical data (blood pressure, serum fasting glucose, onset-to-alteplase bolus time, and stroke severity assessed using the National Institutes of Health Stroke Scale [NIHSS],¹⁷ infarct location, stroke subtypes classified according to the TOAST [Trial of ORG 10172 in the Acute Stroke Treatment] criteria,¹⁸ and post-IVT flow grade evaluated using the Thrombolysis in Brain Ischemia score).^19,20 The Thrombolysis in Brain Ischemia grade, consisting of 6 flow levels—0, absent; 1, minimal; 2, blunted; 3, dampened; 4, stenotic; and 5, normal flow—was evaluated based on the worst flow signal determined at the presumed culprit arteries. Patients with a normal artery waveform corresponding to the stroke lesion were classified as having 5 points. Symptomatic intracranial hemorrhage (sICH) was defined as any apparently extravascular blood in the brain or within the cranium that was associated with clinical deterioration—indicated by a≥4-point increase in the NIHSS score or leading to death—when identified as the predominant cause of the neurological deterioration (ECASS III criteria [European Cooperative Acute Stroke Study]).²¹

Outcome measures included final infarct volume during hospitalization, NIHSS scores at 24 hours and 7 days post-IVT, early neurological deterioration (END) within 24 hours, delayed neurological deterioration (DND) within 7 days, and 3-month mRS scores. END was defined as an increase of ≥4 points in the NIHSS score within the first 24 hours after IVT.²² DND was similarly defined as END, but occurring between 24 hours and 7 days.²³ The mRS score at 90 days was recorded via telephone interviews with each patient or their next of kin. An mRS score ≥2 was defined as an unfavorable outcome.^24,25 The infarct volume was calculated using brain diffusion-weighted magnetic resonance imaging obtained 3 to 7 days after stroke onset, utilizing the open-source 3-dimensional Slicer software, version 5.3.0.²⁶ The mRS score and imaging parameters were assessed by examiners who were blinded to the clinical parameters and serum biomarker levels.

Neuroinflammation Biomarker Measurement

Human blood samples were collected before (baseline, within 4.5 hours of onset) and at 24 hours after IVT and subsequently stored in the Biobank Department of Jilin University’s First Hospital Clinical Research Division. GFAP was measured twice, before and at 24 hours after IVT, using an automated magnetic particle-based chemiluminescent enzyme immunoassay system (MS-Fast/Aceso 80A; Sophonix, Beijing, China) as previously reported.²⁷ Laboratory staff were blinded to the clinical status of the patients, and the assay results were not disclosed to the healthcare providers. The lower limit of quantification of GFAP was 5 pg/mL. Samples were measured in duplicate to control for batch effects, with a variation coefficient of <8.0%.

Peripheral Inflammation Biomarker Measurement

Peripheral inflammation was evaluated using baseline blood samples (before IVT/within 4.5 hours after stroke onset) through leukocyte and neutrophil counts, neutrophil-to-lymphocyte ratio, systemic inflammation response index, and systemic immune-inflammation index.²⁸ Systemic inflammation response index was calculated as (neutrophils×monocytes)/lymphocytes, and systemic immune-inflammation index was calculated as platelets×neutrophils/lymphocytes.

Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics, version 26.0 (IBM Corp, Armonk, NY). A 2-sided P<0.05 was considered statistically significant. Categorical variables were presented as frequencies and percentages and compared using the χ² test or Fisher exact test. Continuous variables were expressed as means and standard deviations or medians and interquartile ranges according to normal or skewed distributions, respectively. Independent continuous variables were compared using 1-way ANOVA or the Kruskal-Wallis test. The GFAP levels before and at 24 hours after IVT were compared using the Wilcoxon signed-rank test.

ΔGFAP was determined by subtracting baseline levels from those measured 24 hours post-IVT. ΔGFAP was then transformed into categorical variables throughout the analyses, according to the following criteria. ΔGFAP ≤0 was defined as no increase in GFAP, whereas ΔGFAP >0 was classified as a mild, moderate, or severe increase based on the tertile distribution. The baseline peripheral inflammation data were transformed into categorical variables according to the quantiles of the distribution. Sensitivity analyses of the underestimated and overestimated conditions of ΔGFAP categories were conducted due to the presence of a lower limit. For investigating the underestimation of ΔGFAP categories, concentrations of GFAP below the limit of quantification were recorded as 5 pg/mL. For patients whose both pre- and post-IVT GFAP levels were <5 pg/mL, ΔGFAP was calculated as 0 pg/mL, and these patients were classified under the nonincrease GFAP group. For investigating the overestimation of ΔGFAP categories, concentrations of GFAP below the limit of quantification were recorded as 0 pg/mL. For patients whose both pre- and post-IVT GFAP levels were <5 pg/mL, ΔGFAP was recorded as 5 pg/mL, and these patients were included in the mild-increase GFAP group. The detailed interpretation of underestimated and overestimated ΔGFAP categories is presented in Table S2. Multivariable binary logistics were used to explore the association of ΔGFAP with END, DND, and unfavorable 3-month outcome defined as mRS score ≥2; and the association of baseline peripheral inflammation with ΔGFAP. Multivariable linear regressions were used to explore the association of ΔGFAP with infarct volume and NIHSS score. Model 1 was unadjusted; model 2 was adjusted for age and sex; model 3 was adjusted for age, sex, and vascular risk factors (smoking, drinking, hypertension, diabetes, dyslipidemia, hyperhomocysteinemia, previous ischemic stroke, and coronary heart disease); model 4 was adjusted for age, sex, vascular risk factors, and clinical data (blood pressure, heart rate, blood glucose level, admission NIHSS score, onset-to-alteplase bolus time, TOAST, infarct location, and bridging therapy); and model 5 was adjusted for variables in Model 4 plus sICH and Thrombolysis in Brain Ischemia. In the analysis of the association of baseline peripheral inflammation with ΔGFAP, model 6 was applied, adjusting for the confounders of model 5 plus infarct volume to confirm the independent influence of baseline peripheral inflammation on ΔGFAP. The associations of ΔGFAP with outcomes and baseline peripheral inflammation with ΔGFAP were analyzed both in underestimation and overestimation conditions to conduct sensitivity analysis.

To evaluate whether adding ΔGFAP would further increase the predictive value of conventional risk factors (model 5) for 3-month unfavorable outcome, we compared reclassification and discriminatory power using the area under the receiver operating characteristic curve, net reclassification index, and integrated discrimination improvement after adding ΔGFAP. Comparison of the 2 receiver operating characteristic curves was based on the method of DeLong et al.²⁹ Decision curve analysis was used to quantify the net benefit.

Results

Participant Characteristics

Initially, 1580 patients who received IVT were screened for eligibility. GFAP levels were measured 2 times in 753 patients; of these, 10 patients did not complete the measurements of baseline leukocyte, neutrophil, monocyte, lymphocyte, and platelet counts. A total of 743 patients were included in the final analysis (Figure S1). Among them, 59 patients had missing magnetic resonance imaging data, and 684 patients were included in the analysis of infarct volume; 8 patients had missing 7-day NIHSS score data, and 735 patients were included in the analysis of 7-day NIHSS score. Two patients were discharged within 7 days, and 741 patients were included in the analysis of DND. All 743 patients were included in the analysis of 24-hour NIHSS score, END, and 3-month mRS score. The mean age of the patients was 62.38 years, and 537 were men (72.3%). Patient characteristics are presented in Table 1. One hundred forty-seven (19.8%) patients had GFAP values <5 pg/mL at baseline, and 32 (4.3%) had GFAP values <5 pg/mL at 24 hours after IVT. The visualization of the value distribution is shown in Figure S2. A higher proportion of patients with moderate or severe GFAP increases had END, DND, or unfavorable outcomes (nonincrease versus mild versus moderate versus severe increase: END, 4.4% versus 2.0% versus 8.9% versus 8.4%, P=0.010; DND, 0.7% versus 3.5% versus 5.4% versus 9.9%, P=0.039; unfavorable outcome, 36.5% versus 42.3 versus 58.9 versus 73.9%, P<0.001).

Table 1.

Clinical Characteristics

GFAP Changes in Different Categories of Patients

GFAP levels were significantly higher at 24 hours after IVT than before IVT in all patients, suggesting the progression of neuroinflammation within 24 hours after IVT in patients with stroke. Furthermore, we compared ΔGFAP among patients in different categories. ΔGFAP was significantly higher in patients with END, DND, or unfavorable outcomes than in those with non-END, non-DND, or favorable outcomes, of which END was defined as an NIHSS score increase of ≥4 points within the first 24 hours after IVT,²² and DND was defined in the same manner as END but occurring between 24 hours and 7 days.²³ These results indicated a more severe progression of neuroinflammation in patients with END, DND, or unfavorable outcomes (Table 2).

Table 2.

GFAP Changes Before and 24 Hours After Thrombolysis Stratified by Outcome Status

Association of GFAP Changes With Infarct Volume

Univariable linear regression showed that a severe ΔGFAP increase (>50.7 pg/mL) was significantly associated with larger infarct volume. After adjusting for confounders, a severe ΔGFAP increase was independently associated with larger infarct volume in model 5 (β, 30.965 [95% CI, 19.185–42.745]; P<0.001). The detailed results are presented in Figure 1 and Table 3.

Figure 1.

Forest plots of the association between GFAP (glial fibrillary acidic protein) changes and outcomes. Age, sex, vascular risk factors (including smoking, drinking, hypertension, diabetes, dyslipidemia, hyperhomocysteinemia, previous ischemic stroke, and coronary heart disease), and clinical data (including systolic blood pressure, diastolic blood pressure, heart rate, blood glucose, admission National Institutes of Health Stroke Scale [NIHSS] score, onset-to-alteplase bolus time, TOAST [Trial of ORG 10172 in Acute Stroke Treatment], infarct location, bridging therapy, symptomatic intracranial hemorrhage, and Thrombolysis in Brain Ischemia score) were adjusted for. OR indicates odds ratio.

Table 3.

The Association of GFAP Changes (Underestimated) With Outcomes

Association of GFAP Changes With Functional Indicators

As shown in Figure 1 and Table 3, a severe ΔGFAP increase was independently associated with higher NIHSS scores at 24 hours (β, 2.632 [95% CI, 1.644–3.620]; P<0.001) and higher NIHSS scores at 7 days in Model 5 (β, 3.298 [95% CI, 2.179–4.417]; P<0.001).

Severe ΔGFAP increase was independently associated with DND (odds ratio, 8.887 [95% CI, 1.100–71.814]; P=0.040); however, the association between ΔGFAP and END was not significant in model 5 (odds ratio, 2.723 [95% CI, 0.940–7.885]; P=0.065).

For 3-month unfavorable outcomes, both moderate and severe increases in ΔGFAP were independently associated with unfavorable outcomes in model 5 (moderate: odds ratio, 2.468 [1.521–4.004]; P<0.001; severe: odds ratio, 3.631 [2.159–6.106]; P<0.001). The distribution of mRS scores according to ΔGFAP is shown in Figure 2. The correlation between ΔGFAP categories and 3-month unfavorable outcome is shown in Figure S3. As demonstrated in Table S3 and Figure S4, the addition of ΔGFAP significantly improved the area under the receiver operating characteristic curve of model 5, which accounted for age and sex; vascular risk factors (smoking, drinking, hypertension, diabetes, dyslipidemia, hyperhomocysteinemia, previous ischemic stroke, and coronary heart disease); and clinical data (blood pressure, heart rate, blood glucose level, admission NIHSS score, onset-to-alteplase bolus time, TOAST, infarct location, bridging therapy, sICH, and Thrombolysis in Brain Ischemia). Moreover, adding ΔGFAP substantially enhanced the risk reclassification and discriminatory power (net reclassification index 49.45%, P<0.001; and integrated discrimination improvement 9.47%, P<0.001; Table S3). In the decision curve analysis, the model plus ΔGFAP showed a better clinical net benefit than Model 5 (Figure S5).

Figure 2.

Distribution of modified Rankin Scale (mRS) score according to GFAP (glial fibrillary acidic protein) changes.

Influence of Bridging Therapy and sICH on the Association of GFAP Changes With Outcomes

Thirty-one (4.2%) patients received bridging therapy in our study. No significant interaction function was observed between ΔGFAP and bridging therapy across outcomes, indicating that bridging therapy may not significantly influence the association of ΔGFAP with outcomes (Table S4). We further excluded patients with bridging therapy and found that its association with outcomes was similar to that of the entire cohort (Table S5).

Five (0.7%) patients had sICH according to ECASS III criteria. The interaction between ΔGFAP and sICH across the 7-day NIHSS score and infarct volume was significant, suggesting that sICH post-IVT may influence the association of ΔGFAP with short-term outcomes like 7-day NIHSS score and infarct volume (Table S6). After excluding patients with sICH, the association between ΔGFAP and outcomes remained unchanged with the entire cohort (Table S7). However, because the number of patients with sICH was too small to conduct regression analyses, the impact of ΔGFAP on outcomes in patients with sICH remains uncertain.

Association of Peripheral Inflammation With GFAP Changes

Because moderate and severe increases in GFAP were associated with outcomes, we further explored the association of peripheral inflammation with moderate to severe ΔGFAP increases and severe ΔGFAP increases, separately. As shown in Table S8 and Figure 3, baseline leukocyte counts, neutrophil counts, neutrophil-to-lymphocyte ratio, and systemic immune-inflammation index were associated with moderate to severe increases in ΔGFAP in models 1 to 5. We considered infarct volume as a confounder in Model 6 to exclude the influence of infarct volume on the association between peripheral inflammation and ΔGFAP. The independent correlation of peripheral inflammation with moderate to severe ΔGFAP was stable, suggesting that peripheral inflammation may contribute to the progression of neuroinflammation after stroke (Figure 3; Table S8).

Figure 3.

Forest plots of the association between baseline peripheral inflammation and GFAP (glial fibrillary acidic protein) changes. A through E, Forest plots of the association between baseline peripheral inflammation and moderate to severe increases in GFAP; (F through J) Forest plots of the association between baseline peripheral inflammation and severe increases in GFAP. Age, sex, vascular risk factors (including smoking, drinking, hypertension, diabetes, dyslipidemia, hyperhomocysteinemia, previous ischemic stroke, and coronary heart disease), and clinical data (including systolic blood pressure, diastolic blood pressure, heart rate, blood glucose, admission National Institutes of Health Stroke Scale score, onset-to-alteplase bolus time, TOAST [Trial of ORG 10172 in Acute Stroke Treatment], infarct location, bridging therapy, symptomatic intracranial hemorrhage, and Thrombolysis in Brain Ischemia score) were adjusted for. OR indicates odds ratio; SII, systemic immune-inflammation index; and SIRI, systemic inflammation response index.

Regarding the association of peripheral inflammation with a severe increase in GFAP, baseline leukocyte and neutrophil counts, neutrophil-to-lymphocyte ratio, and systemic inflammation response index were independently associated with severe ΔGFAP after considering infarct volume (Figure 3; Table S8). ΔGFAP distribution according to quartered baseline peripheral inflammation is shown in Figure 4.

Figure 4.

Distribution of GFAP (glial fibrillary acidic protein) changes according to quartiles of markers of baseline inflammation. Distribution of GFAP changes according to quartiles of baseline (A) leukocyte count, (B) neutrophil count, (C) neutrophil-to-lymphocyte ratio (NLR), and (D) systemic inflammation response index (SIRI).

The above results were obtained from underestimated ΔGFAP levels. The findings obtained from overestimated levels showed that the association of ΔGFAP with DND was not significant, whereas the association with other outcomes was similar. This indicates that the association of ΔGFAP with END, NIHSS score, infarct volume, and unfavorable outcomes was stable; however, its association with DND cannot be concluded in the present study (Tables S9 and S10). After excluding patients with bridging therapy or sICH, the association of overestimated ΔGFAP levels with outcomes remained unchanged compared with that of the entire cohort (Tables S11 through S14). The correlation between overestimated ΔGFAP categories and 3-month unfavorable outcomes was similar to that of underestimated ΔGFAP categories, as shown in Figure S3. The risk reclassification, discriminatory power, and net benefit were similarly improved after adding overestimated ΔGFAP categories (Table S3; Figures S4 and S5). In addition, the association of baseline peripheral inflammation with ΔGFAP was stable regardless of whether ΔGFAP levels were underestimated or overestimated (Table S15).

Discussion

Our study revealed that GFAP changes within 24 hours after IVT (ΔGFAP) were independently associated with clinical outcomes. In addition, we found that higher baseline peripheral inflammation was correlated with greater 24-hour increases in GFAP.

GFAP, a primary intermediate filament protein in mature astrocytes, is a crucial component of the cytoskeleton. When exposed to external stimuli, including ischemia, astrocytes undergo reactive hypertrophy and proliferation and contribute to the progression of neuroinflammation, which is accompanied by increased GFAP expression and secretion.^13,30 This protein is released into the peripheral blood via a disturbed blood–brain barrier and serves as a detectable biomarker.¹³ In studies on central nervous system diseases, including Alzheimer disease and multiple sclerosis, GFAP has been identified as being closely associated with neuroinflammation, making it a valuable biomarker for these neurodegenerative conditions.^31,32 In the field of ischemic stroke, previous research has demonstrated that the level of neuroinflammation, reflected by GFAP during hospital admission, correlates with the prognosis of patients with ischemic stroke.¹³ Animal experiments have shown that the development of neuroinflammation is dynamic, particularly within a few hours after stroke.^8,10,33 A recent clinical study found that the neuroinflammatory cascade, as indicated by serum GFAP, is activated within 36 to 133 minutes after stroke and progresses rapidly.¹¹ However, whether neuroinflammation progression influences clinical outcomes remains unclear.

Our study found that GFAP levels at 24 hours after IVT were significantly elevated compared with those at baseline, suggesting neuroinflammatory progression within 24 hours after stroke, consistent with previous findings.^11,13 Further analysis revealed that ΔGFAP within 24 hours after IVT was independently correlated with infarct volume and functional outcomes. Moreover, we verified a substantially better reclassification and discrimination power according to a significant increase in the area under the receiver operating characteristic curve, net reclassification index, and integrated discrimination improvement after adding ΔGFAP beyond traditional factors. Decision curve analysis revealed a better clinical net benefit. These findings suggest that patients exhibiting more pronounced neuroinflammation within 24 hours after IVT may have poorer prognoses. In clinical practice, within 24 hours after IVT, paying attention to and promptly intervening in neuroinflammation is crucial for improving patient prognosis. We further calculated the correlation between ΔGFAP categories and 3-month unfavorable outcome (Figure S3), which may help clinicians judge patient condition and conduct potential risk-stratification of unfavorable outcomes after IVT. Notably, although we found no significant association between ΔGFAP and END, the P values were near the threshold (underestimated: P=0.065; overestimated: P=0.055). We considered that neuroinflammatory progression within 24 hours, as reflected by ΔGFAP, might also partly contribute to END, although it did not reach the level of statistical significance. In addition, the association of underestimated and overestimated ΔGFAP with DND was inconsistent, indicating that its relationship with DND cannot be concluded in this study and requires further exploration.

The mechanism by which neuroinflammation affects prognosis may involve the following aspects: first, a previous study suggested that neuroinflammation directly inflicts secondary damage on neurons in the ischemic penumbra,³⁴ ultimately influencing the extent of neuronal damage and clinical outcomes.³³ Second, this intense neuroinflammatory activity can trigger severe glial proliferation, impeding neural repair during the recovery phase of stroke.^35,36 Furthermore, the neuroinflammatory cascade after a stroke can progress to chronic neuroinflammation, leading to persistent neurological deficits.³⁷

In addition, we investigated the potential causes of 24-hour increases in GFAP. We found that elevated peripheral inflammation before IVT was associated with neuroinflammation based on GFAP increases, potentially resulting in severe neurological damage. Previous studies have linked infarct volumes and GFAP levels³⁸; therefore, we considered infarct volume as a confounder and found a significant association between peripheral inflammation and increases in GFAP. We hypothesized that this independent influence of peripheral inflammation on central nervous system inflammation could be attributed to the infiltration of peripheral immune cells. After a stroke, initial central inflammation is activated, leading to the release of inflammatory cytokines and chemokines.⁸ These mediators facilitate the recruitment of peripheral immune cells to the central nervous system via a compromised blood–brain barrier, thereby amplifying neuroinflammation.^39,40 Accordingly, our findings highlight the importance of monitoring the degree of peripheral inflammation during the hyperacute phase of ischemic stroke and suggest that addressing peripheral inflammation in patients with acute ischemic stroke undergoing IVT is crucial for optimizing patient prognosis.

This study had some limitations. First, owing to the inherent limitations of observational studies, a definitive causal relationship between peripheral and central nervous system inflammation cannot be established. Second, GFAP was only sampled up to 24 hours, leaving the longer-term trajectory (eg, days 2–7) uncertain, which requires further investigation. Third, the history of previous ischemic stroke appears imbalanced across ΔGFAP categories. Although we explored the association between ΔGFAP and outcomes using multivariable binary logistic and linear regressions and adjusted for the history of previous ischemic stroke as a confounder, this imbalance may still influence results. Therefore, further studies with a larger sample size are warranted. Fourth, because of the presence of the lower limit of quantification, potential assay variability near this threshold could influence the assignment of ΔGFAP stratification.

Conclusions

The increase in GFAP levels over the first 24 hours after IVT is independently associated with clinical outcomes, with higher baseline peripheral inflammation correlating with greater 24-hour increases in GFAP.

Article Information

Acknowledgments

The authors thank the Department of Biobank, Division of Clinical Research, The First Hospital of Jilin University, for providing human blood samples, and The First Hospital of Jilin University Real-World Data Application Platform for providing data.

Author Contributions

Z.-N. Guo and Y. Yang drafted the initial protocol, which was reviewed with critical revisions and approval by all authors. Dr Qu and P. Zhang did the statistical analysis. Dr Zhou, C. Li, Dr Abuduxukuer, L.-J. Wang, L.-C. Yang, Dr Tan, Dr Yuan, Dr Han, Dr H, Dr Zheng, Dr Zhao, L.-J. Guo, L. Jiang, J.-F. Li, Y. Jiang, Dr Zou, D. Xu, H. Xu, X.-J. Wang, Dr Qi, Dr Hu, and Y. Zhang collected data. Drs Jin and Sun interpreted the data. Dr Qu, Dr Zhou, and Z.-N. Guo wrote the first draft of the manuscript. All authors contributed to data acquisition. All authors contributed to the critical revision of the manuscript and approved the final manuscript for submission.

Sources of Funding

This project was supported by the National Natural Science Foundation of China (U24A20686) to Y. Yang; the Fundamental Research Funds for the Central Universities (45124031D051) to Z.-N. Guo; Science and Technology Department of Jilin Province (YDZJ202302CXJD061), Jilin Provincial Key Laboratory (YDZJ202302CXJD017), and the Norman Bethune Health Science Center of Jilin University (2025JBGS02) to Y. Yang; the Talent Reserve Program of the First Hospital of Jilin University (JDYYCB-2023002) to Z.-N. Guo; the Doctor of Excellence Program of The First Hospital of Jilin University (JDYY-DEP-2024036) and the Hospital Youth Development Fund of The First Hospital of Jilin University (JDYY16202502) to Dr Qu.

Disclosures

None.

Supplemental Material

Tables S1–S15

Figures S1–S5

STROBE Checklist

Supplementary Material

str-57-697-s001.pdf ^{(1.3MB, pdf)}

str-57-697-s002.pdf ^{(142.1KB, pdf)}