Acute ischemic stroke (AIS) is the predominant cause of death and disability among adults in China, with high rates of morbidity, mortality, and long-term disability. It accounts for approximately 75%−85% of all stroke cases (1). Mechanical thrombectomy (MT) is the preferred treatment for recanalization in patients with AIS caused by major vascular occlusion (2, 3). Although MT is highly effective in achieving brain tissue reperfusion, 20%−30% of patients with AIS still experience insufficient blood flow following MT in clinical settings (4, 5). Therefore, when efficient reperfusion flow cannot be maintained following thrombus removal, intracranial remedial stenting (RS)—as a remedial therapy—offers a significant therapeutic alternative for patients with AIS.

Recent research highlights the efficacy of thrombus composition in the success of MT. Thrombi rich in erythrocytes are easier to remove because they adhere more readily to the retrieval stent (6). In contrast, fibrin-rich thrombi present a greater challenge due to their rigidity and friction, making them more difficult to dislodge and fragment (7). The physical properties of these thrombi can also cause damage to the arterial endothelium during removal, triggering an inflammatory reaction and promoting the expression of pertinent biological components. Therefore, research into the role of inflammation-related biological markers in regarding RS has garnered significant attention. These factors may indirectly affect the outcomes of RS by altering the intravascular milieu and thrombus composition. Research shows a strong correlation between thrombosis, stroke prognosis, and inflammatory markers such as von Willebrand factor (VWF), CD147, CD163, C-reactive protein (CRP), neutrophil extracellular trap (NET), and Actin (8–13). These markers can influence thrombus composition and affect reperfusion effectiveness by intensifying the local inflammatory response.

We hypothesize that inflammatory marker expression in thrombo-pathology varies among patients who undergo RS after MT, depending on the difficulty of MT in removing different thrombus components. Therefore, this study aims to investigate the relationship between thrombo-pathology and prognosis in patients with RS to reveal how inflammation-related biological factors influence thrombo-pathology characteristics and provide a theoretical basis for optimizing RS techniques. The findings could enhance reperfusion efficacy to improve the prognosis of patients with AIS.

Our findings are as follows: (1) VWF, CD147, Actin, white blood cells, and F/P content showed similar expression in the thrombi of both patient groups. However, thrombi from patients with RS exhibited higher levels of CD163, CRP, and NET than those from patients with NRS. (2) In patients with RS, the thrombo-pathological factors VWF and CD163 independently predict a poor prognosis at 90 days (mRS: 3–6).

The evidence suggests that RS occurs following thrombus extraction in patients with AIS, primarily owing to thrombus diversity (including stiffness and friction) and atherosclerosis-induced stenosis, among other factors (16). These variables can lead to endothelial injury and residual thrombus, causing hemodynamic anomalies, inflammatory reactions, and impaired blood flow post-thrombus removal. In contrast to patients with NRS, thrombi in patients with RS can be further exacerbated by hemodynamic alterations, inflammatory responses, and associated biological variables. Our study shows that thrombi from patients with RS exhibit higher NETs, CD163, and CRP levels, along with greater cellular content, than those of patients with NRS. Additionally, CD163 and VWF were poor prognostic markers in patients with RS. These findings suggest that in patients undergoing RS following thrombus ablation, biomarkers such as inflammatory variables may be important in thrombosis and reperfusion processes.

This study reveals that thrombus NET expression is higher in RS patients compared to NRS patients. NETs are crucial components of thrombi in ischemic stroke and contribute to the pathogenesis of AIS. Research indicates that NETs in the thrombus of patients with AIS resist thrombolysis and prevent the recanalization of blocked arteries (17). First, NETs, with their specialized mesh fiber structure, trap and encapsulate red blood cells, platelets, and other components (18). Furthermore, NET DNA and other components alter the fibrin structure, enhancing the physical properties of the thrombus and limiting its total removal (19). A study found a dense fibrin network containing VWF and NETs in the thrombus of AIS patients, which provides an adhesive scaffold for red blood cells, platelets, fibrin, and coagulation factors, making the thrombus difficult to dissolve. The researcher suggested targeting multiple thrombus components in thrombolytic therapy to benefit patients clinically (20). Residual NETs in the thrombus can trigger platelet activation and local inflammatory response, increasing the risk of re-thrombosis and impairing blood flow continuity and stability following thrombus removal. Second, a high neutrophil count correlates closely with disease severity and is a significant predictor of AIS (21). Neutrophils in patients with AIS release substantial amounts of NETs, which can directly cause endothelial dysfunction by activating and damaging endothelial cells, attracting inflammatory cells involved in atherosclerotic plaque formation, and leading to localized cerebral artery stenosis. Other studies have shown that NETs are involved in the composition of all AIS thrombi, especially in their outer layers. Quantitative measurement of thrombus NETs content was not associated with clinical outcomes or AIS pathogenesis, but was significantly correlated with the procedure length of endovascular therapy and the number of device passes (22). Consequently, we examine how these elements interact. NETs not only hinder blood flow from achieving effective reperfusion after thrombus retrieval but also enhance the physical properties of the thrombus and promote inflammation and vasculopathy.

Studies show that CRP is a significant potential biomarker for assessing the prognosis of patients with AIS (23, 24). CRP is commonly used to assess the pathologic inflammatory response and has been extensively studied in relation to thrombosis linked to atherosclerosis. This study suggests a direct correlation between high CRP levels in the thrombi of patients with RS and their impaired ability to restore blood flow after thrombus removal. CRP contributes to oxidative stress and the generation of reactive oxygen species (ROS), among other roles, in the atherosclerotic process (25). These ROS can oxidize low-density lipoprotein cholesterol, producing oxidized low-density lipoprotein, an essential mediator in the development of atherosclerotic plaques (26). Recurrent thrombus removal may lead to endothelial damage and plaque rupture in patients with AIS, initiating a vicious cycle that exacerbates endothelial injury, increases vascular permeability, increases CRP production, and activates inflammatory responses. This cycle enhances the possibility of blood clot formation and adhesion to injured arterial walls, thereby obstructing blood flow. Additionally, CRP activates the complement system, which promotes thrombus formation (25). Activation of the complement system promotes platelet aggregation, supports the formation of a stable thrombus structure, and converts fibrinogen to fibrin. CRP also significantly stimulates the migration and proliferation of vascular smooth muscle cells. These cells serve as bridges during coagulation, enhancing thrombus durability and reducing the likelihood of complete dissolution. Based on the cited studies, CRP expression in thrombi is likely higher in patients with RS than in those with NRS.

Research indicates that CD163 can serve as a biomarker for various diseases, including infections, autoimmune diseases, cancers, and nervous system abnormalities (27). CD163 is a significant molecular marker on the surface of macrophages, binding to erythrocytes and their metabolites to facilitate their removal. Its elevated expression in tissues often indicates a specific adaption mechanism to the inflammatory response (28). In patients with AIS, CD163 is necessary for the local immune response involving brain microglia and circulating monocytes (29). The pathogenesis of AIS is correlated with the role of CD163 in stimulating the anti-inflammatory response of monocytes (30). Therefore, we hypothesize that in patients with AIS undergoing RS, CD163 is predominantly expressed in monocytes and macrophages. These cells can adhere to the arterial wall during thrombosis and release inflammatory mediators such as interleukin-6 and tumor necrosis factor-α. These inflammatory agents exacerbate endothelial damage and contribute to the atherosclerotic process, which narrows the arterial lumen and thickens the arterial wall, thereby impeding effective blood circulation and reperfusion following thrombus removal.

In this study, where we examined thrombi in patients with RS and NRS, minimal variation was observed in the expression of key biomarkers and their components, such as VWF, CD147, actin, RBC, and F/P, between the two groups. The reasons for selecting above biomarkers for detection are as follows. High levels of actin in thrombi enhance the red blood cell-fibrin scaffold, making it difficult to completely remove the thrombus during the first thrombectomy, which may require stent rescue to improve treatment outcomes (4); CD147 upregulates MMP-9, disrupting the blood-brain barrier and amplifying local inflammation, and has been shown to be closely associated with post-operative reperfusion injury and hemorrhagic transformation (8); VWF, released in large quantities after endothelial denudation, rapidly mediates platelet adhesion, triggering early reocclusion (31). The above indicators are crucial for the success rate of stent implantation, thrombosis formation, and vascular stability. This study shows that the proportion of LAA is higher in the RS group, while the proportion of CE is higher in the NRS group. Previous studies have indicated that there is no significant difference in RBC and F/P expression in thrombi from CE-type and LAA-type stroke patients (32). The infiltration of neutrophils and the stimulation of oxidized low-density lipoprotein within atherosclerotic plaques can induce the formation of NETs. Repeated mechanical traction or residual stenosis during thrombectomy may also cause persistent endothelial damage, triggering neutrophils to release large amounts of NETs. NETs, in turn, can enlarge the thrombus volume by capturing red blood cells and platelets, but they do not alter the relative proportions of F/P and RBC. This could be the potential reason for the lack of significant differences in F/P and RBC ratios between the two groups, while there is a significant difference in NETs content. The specific mechanisms involved in this process need to be further investigated in future studies.

These preliminary findings suggest that endothelial damage and inflammation are crucial in AIS development, with high CD163 expression levels potentially indicating a poorer prognosis. Jiang et al. (27) found a strong connection between CD163 levels and poor outcomes in patients with AIS based on embolism studies involving 25 patients with LAA-type strokes and 42 CE-type strokes. Greco et al. (33) categorized human monocytes into CD16+ and CD14+ subtypes, revealing a significant correlation between stroke severity and disability and CD163 levels in CD16+ monocytes in patients with ischemic stroke. This is in line with the findings of our study in patients with AIS undergoing RS treatment, where we hypothesize that repeated thrombus removal contributes to endothelial damage. This damage increases vascular permeability, which is stimulated by CD163, leading to inflammatory mediator release and activation of VEGF-A/VEGFFR2 signaling by macrophages through hemoglobin clearance (34). Such changes compromise the integrity of the blood-brain barrier, increasing the risk of hemorrhagic transformation and malignant cerebral edema, thereby worsening patient prognosis.

This study revealed a correlation between elevated levels of VWF and both poor prognosis and stroke severity (35). VWF, a glycoprotein produced by vascular cells, plays a crucial role in blood coagulation and platelet adhesion. Abnormally high VWF levels are associated with thrombosis, inflammation, and endothelial cell damage (24). Given that patients with AIS undergoing RS can experience stenosis after thrombus removal, high shear rates in such cases may cause VWF to stimulate inflammation by binding to GPIb on platelets, thereby forming a VWF-GPIbα axis (36). This leads to leukocyte recruitment and edema, triggering an allergic cascade and reducing the efficiency of intravascular blood flow perfusion in the cerebral infarction region. Moreover, thrombus removal damages endothelial cells, causing a significant spillover of VWF. This initiates the thrombo-inflammatory cycle, which exacerbates clinical impairment in patients with AIS and is connected to ischemia, reperfusion (I/R) injury, and the post-ischemic inflammatory response (37). Therefore, we conclude that these factors collectively represent a positive association between VWF and both stroke severity and poor prognosis in patients with AIS.

However, analyzing the thrombo-pathological characteristics of patients with AIS who received RS vs. NRS showed significant differences in the expression of biomarkers associated with AIS and thrombosis, such as CD163, NET, and CRP, between the two patient groups. While VWF levels were similar across both groups, poor prognosis in patients with RS was strongly associated with elevated VWF and CD163 levels. This study has some limitations. It is a retrospective, single-center study, which introduces potential selection bias. The RS group only included 39 cases, with 15 cases showing poor prognosis. The events-per-variable (EPV) ratio was far below the commonly used EPV ≥10 rule for logistic regression. A low EPV can lead to instability in maximum likelihood estimates, manifesting as extreme point estimates (with OR often >1,000) and confidence intervals spanning multiple orders of magnitude, we therefore, attempted the following statistical methods to validate the reliability of the results. Due to the inclusion of too many variables, we attempted variable selection using Lasso regression. Based on Supplementary Table S1, we found that the selected variables still retained VWF, CD163, and NET, which were reported in the “Results” section. This suggests that despite the small EPV, our results are meaningful. Wide confidence intervals do not indicate that the “true effect is large,” but rather reflect the statistical uncertainty due to the small sample size. Therefore, the study may carry an elevated risk of both type I (false-positive) and type II (false-negative) errors, and all ORs should be considered as “directional indicators” rather than precise clinical effects. Here, we used the logistf package (38) to perform Firth penalized regression to attempt to eliminate extreme estimates caused by small sample sizes. After shrinkage, the OR for VWF, CD163, and NET are much smaller than the OR reported in the manuscript, but it still remains statistically significant. Thus, although the OR is large, our results remain meaningful after shrinkage (Supplementary Figure S1). We also used bootstrap (1,000 resamplings) to validate the stability of the OR results from multivariate logistic regression. The results showed that the bootstrap OR trends for VWF (1.15 × 10¹¹), CD163 (1.17 × 10¹⁰), and NET (2.79 × 104) were consistent with those reported in the results, and all fell within the confidence intervals (Supplementary Table S2), suggesting that the trends identified in our analysis are relatively reliable. We then supplemented with receiver operating characteristic (ROC) - area under the curve (AUC) to make the predictions more robust. It is important to note that based on the above “directional results,” the ROC curve showed that both VWF (AUC 0.7256) and CD163 (AUC 0.7639) fall within the 0.7–0.8 range, indicating moderate individual predictive ability, thus VWF/CD163 can serve as a “red flag” indicator, alerting the surgeon to develop a rigorous post-operative monitoring or enhanced antiplatelet strategy in advance, rather than directly deciding “whether to place a stent.” There are also some other limitations in this study. In the multivariate logistic regression analysis, only variables with P < 0.05 in univariate screening were included; confounding factors such as age, sex, and comorbidities were not included in the analysis, which may leave residual confounding. In addition, timely blood specimens and detailed clinical data, such as the number and duration of procedures, were not obtained, and data from only three views of a single thrombus section were used in this study, which may not fully represent the thrombus. Moreover, owing to the scarcity of residual thrombus material, dedicated sections for isotype-matched negative controls could not be allocated. Given the small sample size and potential biases in data analysis, further multicenter studies with larger sample sizes and stratification are needed to identify and validate the independent predictive value of more biomarkers, while minimizing confounding factors as much as possible.

Patients receiving NRS did not exhibit a statistically significant difference in Actin, VWF, CD147, RBC, and F/P levels in their thrombi compared to those receiving RS treatment. In contrast, thrombi from patients undergoing RS exhibited significantly higher levels of CD163, CRP, and NET. Furthermore, elevated VWF and CD163 levels in thrombi from patients with RS were linked to poor prognosis for 90 days. These findings suggest that relevant bleeding biomarkers may be valuable targets for future treatments and offer significant prognostic insights for managing RS in patients with AIS.