Progress on the mechanism of pathological injury secondary to venous stroke and their therapeutic strategies: A narrative review

Abstract

Cerebral venous thrombosis (CVT) is a subset of stroke with a rising trend over the recent years, namely, venous stroke. Its pathogenesis is distinctly heterogeneous from that of arterial ischemic stroke. Severe CVT often results in irreversible brain damage due to concomitant cerebral venous infarction and/or hemorrhage, and the prognosis is extremely poor. Currently, venous stroke has received increasing attention and focus from the scientific community. It has been identified that blood–brain barrier disruption, impaired lymphatic drainage, immune inflammatory response, endoplasmic reticulum stress, oxidative/nitrative stress, and excitotoxicity are intimately tied to the pathophysiology secondary to CVT; however, the specific pathological molecular mechanisms remain to be elaborated in detail. Based on the latest research, combined with the current transcriptomics data, this study reviews the potential pathomechanisms linked with secondary brain injury post-CVT and summarizes the relevant therapeutic strategies, which will contribute to the grasp of the pathology of venous stroke for early clinical recognition and intervention.

Introduction

Stroke is by far the dominant cause of mortality in China. It can be triggered not only by thrombosis of the brain arteries but also by cerebral venous thrombosis (CVT).[1] Venous stroke, also known as CVT, is a distinct subset of stroke that commonly occurs in young adults and children.[2,3] The incidence of CVT has been documented to be 1.3–1.6 cases per 100,000 population in developed countries and even more in Asia.[4] Severe CVT is highly susceptible to venous cerebral infarction and/or hemorrhage due to damaged cerebral vascular endothelium, disrupted cerebrospinal fluid (CSF) uptake, and raised intracranial pressure.[5] The case fatality rate of severe patients can be as high as 34.2%,[6] and 20% to 40% of survivors are routinely left with varied grades of neurologic deficits (e.g., seizures, psychiatric injury, and disorders of consciousness), leading to descent or absence of work capacity, and inflicting a serious burden to the family and society.[7,8,9] Noteworthy, no efficacious approaches are available other than anticoagulation, thrombolysis, and endovascular therapy, especially for its secondary brain injury.

Research on the mechanism of venous cerebral infarction after CVT is a current hot topic.[9,10] Sporadic studies have revealed that blood–brain barrier (BBB) breakdown,[11] neuroinflammation,[12] oxidative/nitrative stress,[13] endoplasmic reticulum (ER) stress,[14] and excitotoxicity[13] are involved in the pathophysiologic processes secondary to CVT. In addition, transcriptomics has recently been exploited to explore the mechanisms of cerebral ischemic injury post-CVT.[10,15] However, to date, fundamental and clinical investigations on the pathomolecular mechanisms of secondary brain damage after CVT remain scarce, which restricts our insights into venous stroke and the development of relevant medications. In this article, we review the potential details of neurovascular injury secondary to CVT [Figure 1] and pertinent therapeutic options.

Figure 1.

A brief summary for the pathophysiologic mechanisms of brain injury secondary to cerebral venous thrombosis (CVT). It was discovered that blood–brain barrier disruption, impaired lymphatic drainage, brain edema formation, immunoinflammatory cascade response, oxidative and nitrative stress, endoplasmic reticulum stress, and excitotoxicity may be collectively linked to secondary brain injury post-CVT

Blood–Brain Barrier Disruption, Impaired Lymphatic Drainage, and Brain Edema

The BBB is a unique membrane-like structure that guards the central nervous system (CNS) from microorganisms and toxins in the bloodstream. In the capillary transect, BBB comprises endothelia, pericytes, basal laminas, and astrocytes in the inside-out direction.[16,17] Recently, it has been proposed that the disruption of the BBB is the initial driver of CVT pathology.[11] Specifically, obstruction of venous return after CVT increases pressure in small veins and capillaries, which in turn leads to enhancement of BBB permeability, with a corresponding elevation in arterial pressure and reduction in perfusion, giving rise to local cerebral ischemia and hypoxia, which in succession comediates vasogenic cerebral edema and cytotoxic cerebral edema.[18,19] In addition to permeability, the structural integrity of the BBB is even more considerable for vasogenic edema. Impairment of endothelial intercellular tight junction proteins (TJPs) as well as microvascular basal lamina are typical features of BBB integrity breakdown.[20] Vosko et al. initially reported that substantial damage to collagen IV-positive microvascular basilemma occurs in rats 3 h post-CVT.[21] In another rat model of CVT, Mu et al. revealed that the levels of TJPs (e.g., ZO-1 and claudin-5) were diminished in affected brain tissues, whereas their levels in serum were markedly elevated, accompanied by obvious vasogenic cerebral edema and neurological deficits.[22] A recent clinical investigation also concluded that serum concentrations of claudin-5 were notably increased in patients with venous cerebral infarction post-CVT compared with controls, hinting that BBB integrity is also disrupted post-CVT.[23] In addition, due to impaired CSF absorption and uncontrolled cerebral edema, intracranial pressure dramatically rises, thereby mediating venous cerebral infarction and/or hemorrhage.[5]

The currently known mechanisms regarding the cause of BBB destruction after CVT may involve activation of matrix metalloproteinase 9 (MMP-9),[11] vascular endothelial growth factor (VEGF),[24] and thrombotic inflammation,[25] along with accumulation of neutrophil extracellular traps (NETs).[26] MMP-9 is a hydrolase that degrades essential ingredients of the basilemma of the BBB and destroys TJPs.[27] Rashad et al. corroborated that MMP9 is the main cause of BBB disruption following CVT and is primarily generated by pericytes.[11] Furthermore, a multicenter prospective cohort study confirmed that circulating MMP-9 is a hallmark of cerebral injury in CVT patients.[28] VEGF, also referred to as angiopermeability factor, induces vasculogenesis in response to hypoxia.[29] Argaw et al. verified that VEGF mediated claudin-5 destruction and promoted BBB breakdown.[30] In a two-cerebral vein obstruction rat model, Kimura et al. initially corroborated that VEGF was expressed in the neuronal cell cytoplasm in the region of angiogenic edema, and inhibition of VEGF could reduce angiopermeability and attenuate brain edema as well as venous cerebral infarction.[24] Moreover, early after CVT, mice expressing the hematopoietic cell JAK^2V617F presented enhanced platelet and neutrophil adherence in venules near cerebral thrombosis, which mediated neutrophil infiltration and BBB disruption through a thrombotic inflammatory response; furthermore, the thrombotic inflammatory markers were markedly elevated in patients with CVT who were JAK^2V617F positive.[25] Besides, Jin et al. first revealed that plasma and thrombus from CVT patients were rich in NETs; meanwhile, in vitro experiments confirmed that NETs could disrupt the endothelial barrier and convert endothelial cells to a procoagulant phenotype, thereby exacerbating thrombosis.[31] In addition, in a CVT rat model, Mu et al. discovered that the accumulation of NETs in the cerebral cortex leads to damage of the BBB. This, conversely, facilitates the infiltration of NETs into the brain, thus constituting a vicious circle, and inhibition of the recruitment and production of NETs reverses the BBB injury following CVT.[26]

BBB disruption directly mediates the progression of encephaledema, which is the main cause of persistently high intracranial pressure and poor prognosis after CVT.[32] Except as the above factors mediating BBB destruction, other mechanisms may be implicated in post-CVT brain edema. Briefly, aquaporin 4 (AQP4) is a family member of aquaporins, comprised six helicoidal cross-membrane domains and two heavily conservative ASN-pro-ALA sequences,[33] and is mainly local to astrocyte, mediating water transport across the BBB.[34] Inhibition of AQP-4 has been found to attenuate cerebral edema in a rat model of CVT.[35] Of even greater interest, alterations in the lymphatic system after stroke may be a momentous contributor to the evolution of encephaledema.[36,37,38] The intracerebral lymphatic system consists mainly of the perivascular compartment in the parenchyma and the aquaporin AQP4 on the footplate of astrocytes.[36,39] CSF-carrying metabolic wastes commingle with interstitial fluid, along the perivascular space, and can be diverted away from the cerebral parenchyma through the meningeal lymphatics and the deep jugular lymphatic system.[40,41] The microstructures of the BBB and the lymphatic network partially overlap, as yet the exact mode of interaction between them is uncertain.[41] Preceding findings have verified that enlargement of the perivascular space in the basal nuclei is linked with higher rates of BBB seepage, hinting that the BBB breakdown is a risk factor for the dysfunction of glymphatic flow,[42] which may constitute a vicious circle of aggravated cerebral edema. Recent studies have found that NETs induce acute lymphatic endothelial lesions and thrombogenesis, leading to obstacles of lymphatic return and hydrocephalus, whereas degradation of NETs remarkably restores meningeal lymphatic drainage, which consequently improves intracranial pressure.[43] Given impaired CSF return and elevated intracranial pressure after CVT, drainage of CSF through the accelerated lymphatic system appears to be an effective strategy for reducing cerebral edema and alleviating intracranial pressure post-CVT.

Immunoinflammatory Cascade

The dynamics of inflammation and immunity have a tremendous impact on the pathologic process and prognosis of cerebrovascular disorders,[44] including CVT.[45] Increasing evidence supports that the immune inflammatory response occurs within minutes after CVT and can last for 7–14 days. The invasion of immune cells (e.g., neutrophils and T-cells) into the brain from the peripheral blood post-CVT and the activation of in situ immunoinflammatory cells (e.g., microglia) in the brain, with concomitant emission of inflamed factors and activation of inflamed pathways, conjointly mediate disruption of the BBB and cerebral edema, which in turn worsens parenchymal brain lesions and exacerbates neurological damage.[10,46] Furthermore, both fundamental and clinical investigations revealed that anti-inflammatory strategies may be promising for the management of CVT.[12] However, the inflammatory mechanisms that underlie the pathology of cerebral injury secondary to CVT are still obscure.

Immune cells

Neutrophils, the pioneer cells that invade the brain poststroke, are critical culprits in stroke-associated inflammation.[47,48] Neutrophil activation not only releases lysozyme and secretes inflammatory mediators[49] but also produces NETs[50] that release histone-containing extracellular DNA fibers, which in turn exacerbate CNS damage through a variety of mechanisms, containing activation of microglial cytoplasmic dsDNA receptor.[51] Research in different animal models has found that eosinophilic necrotic foci and dispersed neutrophils are often seen in affected brain regions 1–2 day after CVT.[19,52] Importantly, antineutrophilic serum may attenuate post-CVT BBB dysfunction and brain edema by lowering the number of neutrophils.[46]

Microglia, the main immunoinflammatory cells in the mind, are essential for maintaining CNS homoeostasis.[53,54] It can not only aggravate brain tissue damage by generating inflamed factors, chemokines, and morbid substances but also promote neuritogenesis, vasculogenesis, and synapse reshaping, thus accelerating the restoration of neurological function.[55] Activation of microglia is mainly divided into M1 and M2 types with opposite roles. The M1 type is distinguished by the generation of proinflamed factors, including interleukin (IL)-1β, IL-6, and so on, which exacerbate inflammatory responses. In contrast, the M2 type is distinguished by the generation of factors containing IL-10, transforming growth factor-β, etc., which suppress inflammation and boost tissue restoration.[54] One study reported widespread microglia invasion in the affected cerebral cortex and upper sagittal sinus as long ago as 1-day post-CVT, peaking at 3 days, with concomitant extension into the subarachnoid space. Meantime, microglia showed a proinflammatory morphology with an amoeboid shape and presented a clear colocalization with IL-1β. In contrast, at 7d of CVT, microglia in and around the infarct zone began to show branching changes and a noninflammatory morphology.[11] Therefore, activation of microglia may exert a crucial adjective influence in the pathologic course after CVT, and modulation of microglia inflammatory phenotypic transformation may be a practical therapeutic tactic for targeting brain injury secondary to CVT.

In the CNS, macrophages are mainly composed of activated brain parenchymal microglia and infiltrating peripheral monocytes.[56] The origins and functions of different macrophage subpopulations vary.[57] An immune infiltration analysis using transcriptomics by Kui et al. found that macrophages also infiltrated into the brain 2d after CVT.[10] In addition, Rashad et al. also detected the intensive accumulation of macrophages surrounding the upper sagittal sinus and in the injured cortex 3 days after CVT in rats and considered that it might be mainly originated from cerebral microglial cell activation in situ.[11]

T cells included multiple subpopulations expressing varying clusters of differentiation (CD) markers and thus function in distinct properties to regulate immune and inflamed homoeostasis.[58] As long ago as 24 h following brain ischemia, T cells were found in peri-infarct tissues and were identified to exert a damaging function in the early stages of ischemic stroke.[59] At 3 days post-CVT, Rashad et al. were the first to observe CD3⁺ T cells in the upper sagittal sinus and ambient parenchyma of rats, indicative of a role for CD3⁺ T cells in CVT-induced inflammation.[11] In addition, by transcriptome sequencing and immune infiltration analysis, Kui et al. have also recently demonstrated a significant increase in T-cell infiltration, especially T helper cells (namely CD4⁺ T-cells), after CVT.[10] It was found that infiltrating CD3⁺ CD4⁻ CD8⁻ T cells (namely, double-negative T cells) could promote microglia-mediated neuroinflammation and exacerbate ischemic brain injury.[60] Meanwhile, CD4⁺ T lymphocytes are implicated in the generation of multiple inflamed cytokines (like ILs), whereas CD8⁺ T lymphocytes are involved in neuronal damage by emitting neurotoxic proleases.[61] However, their specific pathomechanism in venous stroke is rarely covered.

NK cells, as an essential cell type of the immunologic response, can mediate the inflammatory response through the emission of multiple pathogenic factors, catalyzing nerve cell death and worsening cerebral infarction.[62,63] Recently, Rashad et al.[11] identified a pronounced reduction in the quantity of infiltrating NK cells in the brain 2d following CVT, which may be related to feedback inhibitory signals from ischemic neurons;[64,65] however, the dynamics of NK cells in brain-damaged regions following CVT and the specific mechanism of action are not clear.

Inflammatory factors and inflammation-related signaling pathways

Along with the in situ activation or invasion of immune inflammatory cells, plenty of inflammatory factors are subsequently released. In animal models of CVT, the levels of pro-inflammatory parameters such as IL-1β, IL-18, tumor necrosis factor-α, and monocyte chemotactic protein-1 were elevated to varying degrees in damaged brain tissues, which may be attributed to the classical nuclear factor kappa B (NF-κb) signaling pathway or the activation of NLRP3 inflammasome, etc.;[11,46,52] whereas the level of the anti-inflammatory factor IL-10 was reduced.[46] Except for IL-1β, which may be mainly originated from microglia/macrophages,[11,52] the cellular origin of the other cytokines is unknown. In addition, our previous work[15] by transcriptome sequencing of the cortex of CVT mice also confirmed that substantial inflammatory mediators were engaged in the pathological change of CVT, among which the levels of IL-1β, Tlr4, Itgam, Tlr2, and CCl-2 were remarkably elevated. Meanwhile, according to the enrichment analysis, we found that the immunoinflammatory pathways closely related to the pathological damage of CVT mainly included complement and coagulation responses, cytokine-cytokine receptor interaction, NF-κb signaling pathway, NLR signaling pathway, TLR signaling pathway, and NETs generation. These results are in general agreement with those recently reported in a rat CVT model by the analogous method.[10]

Besides, some classical signaling axes such as cGAS-STING (a cytoplasmic aberrant DNA sensing pathway)[52] and HMGB1-RAGE (a powerful inflammatory signaling connected with a variety of damages)[66,67] have been revealed to exert a decisive effect on the inflammatory pathology of CVT, and suppression of their activation can curb the levels of pro-inflammatory mediators to varying degrees, which in turn salvages nerve injury and neurological deficits after CVT. Therefore, the modulation of the inflammatory microenvironment and inhibition of the immune inflammatory cascade may be an effective strategy for targeting secondary lesions of CVT, thereby ameliorating the prognosis of patients.

Oxidative and Nitrative Stress

Oxidative stress is a disorder induced by an unbalance between oxygen radical generation and antioxidant enzyme systems in the body, resulting in damage to a vast array of molecules (e.g., lipids, proteins, and DNA).[68,69] Nitrative stress, on the other hand, is mainly a nitrosative modification of proteins, caused by reactive nitrogen species (containing nitric oxide [NO], peroxynitrite, etc.), which in turn leads to cellular damage.[70] The two arise in tandem and interact with each other and are predominant pathophysiological processes in multiple CNS injuries.[71,72] In addition, evidence suggests that a principal generator of free radicals that give rise to oxidative/nitrative injury is mitochondrion.[73]

Reactive oxygen spices (ROS) and peroxynitrite (generated by diffusion-controlled responses between superoxide radicals and NO), as typical hallmarks of oxidative and nitrative stress, respectively, increased uncontrollably in the early post-CVT period.[74,75] These data were corroborated by an experimental CVT research, which found that the radicals NO, ROS, and peroxynitrite were significantly raised in isolated forebrain neurons on day 1 and/or day 2.[76] Moreover, in an alternative CVT rat model, the levels of the lipid peroxidation marker malondialdehyde were elevated, whereas the antioxidant agents Mn-SOD and GPx were extensively minimized, further validating the burst of oxidative stress in the incipient phase of CVT, which might be of great value in its etiopathogenesis.[13] Besides, our previous study also confirmed that the 3-nitrotyrosine, a peroxynitrite-mediated label for protein tyrosine nitration, was visibly increased after CVT in rats and mainly observed in the neurons and microglia, indicating a role for nitrative stress in CVT pathology. Thioredoxin-interacting protein (TXNIP), alternatively termed thioredoxin (TRX)-binding protein 2, potentiates oxidative stress by interacting directly with the key antioxidant protein, TRX, and restraining its antioxidant functionality.[77] Meanwhile, credible evidence shows that both ROS and peroxynitrite could enhance TXNIP expression or activation, whereas repression or elimination of ROS and peroxynitrite could evidently reverse the corresponding actions.[78,79] It was further revealed by immunoprecipitation that both peroxynitrite and TXNP may be engaged in the mobilization of NLRP3 inflammasome, intensely hinting that oxidative/nitrosative stress may act a pivotal function in neuroinflammation post-CVT.[14] A similar elevation of oxidative/nitrosative stress markers was also confirmed by us in a CVT mouse model. In contrast, mitigation of oxidative/nitrosative stress by drugs facilitated the improvement of CVT neuropathology.[52]

Of interest, the latest transcriptome sequencing discovered that the levels of hemoglobin family-linked genes Hba and Hbb were reduced in CVT rats compared to controls, revealing that the oxygen-carrying capacity or oxygen-binding capacity of the affected brain tissues was decreased after CVT.[10] Furthermore, it has been found that oxidative stress can lead to a decrease in hemoglobin oxygen-binding capacity, and antioxidants can reverse the above dilemma,[80,81] again implying the detrimental behavior of oxidative stress in CVT ischemic pathology. Therefore, stabilizing mitochondrial oxidative/nitrosation stress may be a promising therapeutic tactic against brain injury secondary to CVT.

Endoplasmic Reticulum Stress and Excitotoxicity

The ER is a major organizer for protein synthesis, transport, and maintenance of intracellular Ca²⁺ homoeostasis. ER stress arises from the sustained destruction of ER homoeostasis in neurologic illnesses, triggering a variety of pro-apoptotic mechanisms via the mobilization of the unfolded protein reaction (UPR), which plays a deleterious role in neuronal cells.[82,83,84] Priming of the UPR is contingent on the activation of three type-I membrane spaning proteins or sensors, i.e., protein kinase RNA-like ER kinase (PERK), inositol-required protein 1 (IRE1), and activating transcription factor-6.[82] Tiwari et al. were the first to report that pivotal elements (e.g., CHOP and caspase-12) in ER stress-mediated apoptotic pathway were markedly elevated on day 1 and continued to enhance on day 2 of CVT, and their levels were prominently connected with neurobehavioral alterations and infarct volume at the incipient phase of CVT.[85] Furthermore, our previous research discovered that early upregulation of the ER sensors p-IRE 1α and p-PERK was pretty evident and mostly located in neurons after the onset of CVT. Conclusive evidence indicates that p-IRE 1α and p-PERK can induce or modify TXNIP, at transcriptional or posttranscriptional levels,[86,87] whereas restraint of their phosphorylation notably lowers TXNIP expression.[88] Moreover, obvious colocalization between p-IRE 1α/p-PERK and TXNIP was visible in our previous research, which further verifies the upregulation of TXNIP in the incipient stage post-CVT.[14] Inspiringly, recent findings revealed that ER stress was also associated with MMP-9 activation and BBB disruption, which could be partly countered by the specific inhibitors.[89] In light of the aforementioned findings, therefore, we can broadly deduce that ER stress may contribute to the initial neurovascular injury post-CVT through multiple mechanisms.

Excitotoxicity, which is a distinct form of neurotoxicity triggered by glutamate, may be the inevitable step between hypoxic brain injury and neuronal death, and blocking the signal that gives rise to excitotoxicity may guard against stroke lesions.[90] Tiwari et al. initially reported that glutamate levels in frontal cortex were markedly elevated in CVT rats relative to controls on days 1 and 2, with a nonsignificant difference in glutamate levels on day 7. The opposite trend was observed with decreased expression of N-methyl-D-aspartate glutamate receptor subunits NR1, NR2A, and NR2B.[13] Recently, based on cerebral microdialysis, Simonin et al.[91] found that endovascular mechanical thrombectomy reduced glutamate levels in damaged brain regions and intracranial pressure of patients with severe CVT, which subsequently promotes amelioration of neuropathology and function. Importantly, it has been proposed that glutamate-triggered excitotoxicity actively induces the destruction of the ER microenvironment, giving rise to ER stress and consequently UPR.[92] Similarly, Sokka et al. revealed that ER response is an essential part of excitotoxicity mediated by glutamate receptor activation, and that restraint of ER stress shields against excitatory neuronal damage in rat brain.[93] Therefore, we hypothesized that targeting ER stress and/or glutamate excitotoxicity may ameliorate CVT secondary neuropathological injury.

Therapeutic Strategies Targeting Secondary Pathology in Cerebral Venous Thrombosis

The current treatment of cerebral ischemic injury secondary to CVT, i.e., venous infarction and/or bleeding, is mainly referenced to arterial ischemic stroke[90] and intracerebral hemorrhage,[94] as they share similar molecular mechanisms, although with heterogeneity, as reported above. Effective strategies against the secondary pathologic mechanisms of these disorders, such as edaravone[95] and hypothermia[96,97] therapy, have been confirmed in clinical trials and are widely used in the clinic. Besides, Chen et al.[98] have also previously reported that endovascular therapy (including thrombectomy and thrombolysis) coupled with therapeutic hypothermia and decompressive craniectomy is secure and efficacious for malignant CVT, with good functional outcomes even in patients with brain herniation. However, Stam et al. also revealed no benefit of endovascular therapy in CVT patients with large infarcts and impending brain herniation due to complications such as hemorrhage.[99] Currently, caution is needed to individualize its use in clinical practice.[100] Of interest, recently, Hu et al.[101] indicated that short-term steroid use may be an effective and safe adjunctive treatment for acute/subacute severe CVT; however, these findings require further validation in multicenter randomized controlled trials.

Limitations and Conclusion

Venous stroke can lead to venous cerebral infarction and/or hemorrhage with a complex secondary pathological process. However, due to the constraints of stable animal model establishment and the lack of clinical longitudinal data, fewer mechanisms of brain injury and therapies for venous stroke have been identified. In addition, the lack of systematic methodology in this study should be of concern.

In conclusion, this review summarizes the active pathological processes or molecular mechanisms of secondary brain injury in venous stroke, along with the existing therapeutic approaches, and comprehensively elucidates the possible therapeutic targets in anticipation of deepening the understanding of the cerebral injury in CVT and exploring new therapeutic options.

Author contributions

Guangrong Meng and Rui Ding conceived and designed the article. Shanshan Wei and Yan Li wrote the manuscript and drew the mechanism diagram. Guangrong Meng and Xiaohong Yu revised the manuscript. All authors reviewed and approved the manuscript.

Ethical policy and institutional review board statement

Not applicable.

Data availability statement

Data sharing is not applicable to this article as no datasets were generated and/or analyzed during the current study.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

This work was supported by the Youth Foundation of the National Natural Science Foundation of China (Grant number: 82301536).