Effects of diabetes mellitus complicated by admission hyperglycemia on clot histological composition and ultrastructure in patients with acute ischemic stroke

Acute ischemic stroke (AIS) is the most common cause of mortality and long-term disability worldwide [1]. The risk of AIS is more than two-fold higher and more severe in patients with type 2 diabetes mellitus (T2DM) [2]. It is associated with poorer functional outcomes and higher mortality risk [3]. About 40 to 60% of patients with AIS present with admission hyperglycemia either due to acute stress response or diabetes [4]. In both diabetics and non-diabetics AIS patients, hyperglycemia at the time of admission has been associated with negative outcomes [5]. It is related to the stress response of AIS patients as a result of excessive secretion of steroid hormones, adrenaline, glucagon and free fatty acids [6]. Furthermore, diabetes mellitus and acute hyperglycemia could enhance oxidative stress and inflammation response, impair cerebrovascular reactivity in the microvasculature, provoke a prothrombotic state, and cause cerebral injury [7].

The development of mechanical thrombectomy (MT) has enabled investigations of the composition and structure of human cerebral thrombi [8]. Cerebral thrombi consist of four major components: red blood cells (RBCs), fibrin, platelets, and white blood cells (WBCs) [9]. The evaluation of retrieved clots from patients with AIS may improve our knowledge of stroke pathology and predict treatment response. RBC-rich thrombus might be easier to recanalize in patients with AIS, while fibrin-rich clots are more refractory [10]. The detailed examination of the thrombi can help determine the effectiveness of various treatment approaches for patient selection.

Depending on the clot ultrastructure, most cerebral clots undergo intravital thrombus contraction (retraction), which may be clinically significant. In blood clots, activated platelets produce contractile forces transferred via the fibrin network [11], creating a platelet-fibrin meshwork that accumulates at the periphery of the clots and compresses RBCs into the center of the clot [12]. RBCs are one of the most abundant components of cerebral thrombi. Clot contraction leads to a reduction in the thrombus volume and deformation of the RBCs, including polyhedrocytes and polyhedral RBCs, which comprise the majority of RBCs. Polyhedrocytes provide an impermeable seal because of minimal interstitial space, promoting fibrinolysis resistance [12, 13]. Much research has shown the hyper-reactivity of platelets from diabetic patients, as evidenced by increased fibrinogen binding and enhanced aggregation [14]. In addition, acute hyperglycemia in T2DM can promote further platelet activation [15]. However, the effects of ahDM on forces generated by clot contraction on RBCs have not been investigated.

Although the negative effects of T2DM on cerebrovascular reactivity and reperfusion damage are well established, the effects of DM on the composition and ultrastructure of thrombi in AIS remain unclear. A previous study showed that clots in patients with DM had more fibrin and fewer RBC components than in nonDM patients, while hyperglycemia on admission did not show an association with clot composition [16]. The present study aimed to evaluate the association between ahDM and the composition and ultrastructure of clots in patients with AIS.

Consecutive patients with AIS who underwent MT at Beijing Hospital between June 2017 and May 2021 were enrolled. Inclusion criteria were as follows: (1) AIS caused by an occlusive intracranial clot (both anterior and posterior circulation); (2) availability of data about preoperative computed tomography (CT) (non-computed CT [NCCT] and CT angiography [CTA]) evaluation; and (3) suitable clots retrieved from patients who had undergone MT for histopathological and ultrastructural analysis. Diabetes mellitus complicated by admission hyperglycemia (ahDM) was defined as a history of physician-diagnosed T2DM with a plasma glucose level > 7.80 mmol/L in a random state when admitted to hospital caused by AIS [17]. Non-DM AIS patients were defined as patients without DM and with normal plasma glucose (≤ 7.80 mmol/L) on admission. All study participants provided informed consent. The Beijing Hospital Ethics Committee (2019BJYYEC-130-01) approved this study as it met national and international guidelines for research on humans.

Analysis of the clinical characteristics and thrombi of 55 patients with AIS revealed that ahDM affected ischemic stroke severity and was associated with poorer functional outcomes. Clots of patients with ahDM had more fibrin, fewer RBCs, higher WBC counts, and an equivalent fraction of platelets compared to nonDM patients. Additionally, the proportion of polyhedrocytes in clots was higher, and that of pervious clots was lower in patients with ahDM than in patients without DM.

In the study, patients with ahDM presented with severe ischemic stroke. Patients with ahDM had poorer functional outcomes and higher 90-day mortality rates than nonDM patients. Previous studies have compared stroke severity between patients with T2DM and without T2DM, but conflicting results have been reported [22‐25]. One study reported that stroke was more severe in patients with T2DM, which is consistent with our findings [22]. Furthermore, T2DM has been reported to independently predict more unfavorable functional outcomes at hospital discharge, whereby AIS patients with diabetes exhibit a three-fold higher mortality rate than patients without diabetes [26]. However, other studies have reported no association between T2DM and stroke severity or that patients with T2DM have a mild stroke on hospital admission [23‐25]. Similarly, a previous study did not identify a significant difference in stroke severity between patients with T2DM and without T2DM [16]. Additionally, admission hyperglycemia of acute ischemic stroke causes increased ischemic injury via endothelial dysfunction, oxidative stress, and impaired fibrinolysis [27]. The patients with T2DM included in this study had admission hyperglycemia, partly explaining the discrepancy.

This study demonstrated that clots in patients with AIS and ahDM had fewer RBCs, more fibrin, equivalent platelets, and higher WBC counts than those with AIS and nonDM. Diabetes is characterized by hyperglycemia and insulin resistance, enhanced oxidative stress, inflammatory responses, activation of coagulation and platelets, and endothelial cell dysfunction. Hyperglycemia and insulin resistance can lead to elevated expression and secretion of plasminogen activator inhibitor-1 (PAI-1) [28]. PAI-1 inhibits fibrinolysis in thrombi predominantly by inhibiting plasminogen activator, which promotes fibrin degradation in thrombi. Moreover, glycosylated plasminogen in diabetes directly affects fibrinolysis by reducing plasmin generation and impairing functional protein activity, resulting in impaired fibrinolysis [29]. Increased plasma PAI-1 and glycation of plasminogen may be a potential mechanism underlying elevated fibrin content in the thrombi of patients with ahDM.

This study showed that platelet fraction was comparable between patients with and without DM and may be related to the effect of diabetes on platelets, which is centered on platelet activity [30]. P-selectin and GPIb/CD41 levels are elevated in patients with DM, indicative of platelet activation [31]. Patients with DM who experience myocardial infarction exhibit increased thrombin production and platelet activation [32]. Further, patients with diabetes are characterized by accelerated platelet consumption/production and a resultant increase in immature platelets [33]. Results demonstrated that WBC counts were higher in patients with ahDM than in nonDM patients. It is associated with stimulating oxidative stress and inflammation caused by ahDM. WBCs and platelets from patients with diabetes have been reported to be hyperreactive and express more adhesion molecules [34]. Additionally, activated platelets induce increased formation of circulating platelet-leukocyte aggregates [35].

The history of ahDM may provide clues regarding thrombus composition and facilitate decision-making to develop strategies for MT. A previous study reported that thrombolysis was less effective in thrombi with a high fibrin content than RBC-rich thrombi. In contrast, thrombi with a high RBC count were associated with successful reperfusion [36]. RBC-rich clots are easier to recanalize, whereas fibrin-rich clots are more difficult to recanalize in patients with AIS [10]. Thrombi have higher fibrin content, which increases friction with the vessel wall and makes it more difficult to remove the clot [37]. Therefore, recanalization of thrombi may be more difficult in patients with ahDM. However, we did not observe differences in revascularization outcomes between patients with ahDM and nonDM due to limited sample size and the need for MT equipment and techniques improvements.

Polyhedrocytes cells result from the tightening of blood clots driven by platelet contraction accompanied by compaction of RBCs, gradually changing their shape from biconcave to polyhedral [12]. Platelet activation is necessary for clot contraction [38]. It requires platelet cytoskeletal motility proteins and fibrin as the substrate for the contraction of bridging platelets to generate the necessary forces to segregate platelets/fibrin from RBCs and compress these cells into a tightly packed array [12]. Activated platelets may underpin the higher polyhedrocyte content in clots among patients with AIS with a history of ahDM.

Clot permeability (also referred to as perviousness) is the degree to which blood can flow through clot structures. Clot perviousness is considered a key predictor of treatment responsiveness. When treated with intravenous thrombolysis [39] and mechanical thrombectomy [40], pervious clots are correlated with better recanalization outcomes. Further, thrombus perviousness correlates with histologic composition. A recent study by Benson et al. using MSB staining to differentiate platelets from fibrin revealed a higher RBC component and lower fibrin fraction in pervious thrombi than in impervious clots [41]. This study demonstrated that clots in patients with AIS and ahDM were less permeable and had more fibrin and fewer RBC components, consistent with Benson et al. The characteristics of activated platelets and polyhedrocytes in patients with ahDM permit minimization of the space between cells, resulting in more compact and stable clots, and less deformable and permeable. A previous study demonstrated that the objective of clot contraction was to produce a good hemostatic seal and restore blood flow [42]. In contrast, this thrombus characteristic can negatively affect patients with AIS. Relatively porous clots may allow residual arterial flow and retain a degree of oxygenation to downstream tissues [43]. Here, clot contraction may have adverse effects, such as affecting local blood flow and thrombotic permeability of fibrinolytic enzymes, thereby reducing the internal fibrinolysis rate.

This study has several limitations. Laboratory evaluations of hemoglobin A1c and oral glucose tests were not performed. The absence of HbA1c could put some patients in another group. Further, information on medications used by patients with diabetes for blood glucose control was not collected in detail. In this regard, insulin sensitizers (such as pioglitazone and metformin) may help to reduce PAI-1 levels or platelet activity by improving insulin sensitivity. In addition, patients with undiagnosed T2DM may have been classified as non-DM patients, resulting in selection bias. The patients already spontaneously (or through thrombolysis) recanalized were excluded from the study, limiting the patient’s collective and strength of conclusions.

In conclusion, we demonstrated the effects of ahDM on thrombus composition and contraction-induced RBC deformation. Ischemic stroke severity was affected by ahDM and was associated with poorer functional outcomes. Further, ahDM affected the composition and ultrastructure of clots, and clots from patients with AIS and ahDM exhibited impervious characteristics. Knowledge of the composition and contraction of cerebral thrombi may help improve and predict the effectiveness of thrombectomy or thrombolytic recanalization of occluded vessels and facilitate the development of novel treatment approaches.