Hamostaseologie
DOI: 10.1055/a-2518-9103
Review Article

Pathophysiology and Management of Cerebral Venous Thrombosis

Mariana B. Morais
1   Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal
2   Lopes Lab Unit, Gulbenkian Institute for Molecular Medicine, Lisbon, Portugal
,
Diana Aguiar de Sousa
1   Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal
2   Lopes Lab Unit, Gulbenkian Institute for Molecular Medicine, Lisbon, Portugal
3   Neurosciences Department, Stroke Center, Centro Hospitalar Universitário Lisboa Central – ULS São José, Lisbon, Portugal
› Author Affiliations

Funding No funding sources to report.
 

Abstract

Cerebral venous thrombosis (CVT) is a less common type of stroke that can occur across all age groups but predominantly affects children and young adults. Diagnosis is often challenging due to the nonspecific and variable clinical presentation. The disease course is heterogeneous, with CVT-related parenchymal lesions developing in approximately 50 to 60% of cases. Despite some advancements, significant gaps persist in understanding the pathophysiology of CVT, including the mechanisms underlying brain injury. Anticoagulation is the cornerstone of CVT treatment, but strategies targeting secondary mechanisms of parenchymal damage are still lacking. Here, the current state of the field is briefly reviewed, with the aim to introduce a wide audience (neuroscientists and clinicians alike) to the disease and inform clinical practice and future research.


Epidemiology and Risk Factors

Cerebral venous thrombosis (CVT) is a form of stroke that affects the dural venous sinuses and/or cortical cerebral veins. It is also often classified as part of the venous thrombotic events occurring at an unusual site—the intracranial vessels. Among adults, the incidence of CVT is around 1.3 per 100,000 per year, being approximately 100 times lower than venous thromboembolism (VTE) of lower limbs and lungs.[1] Interestingly, while the incidence of VTE and arterial stroke increases with age,[2] CVT predominantly affects individuals younger than 50 years. In fact, less than 10% of patients with CVT are older than 65 years.[3] Another striking contrast with VTE is the sex ratio—women are three times more often affected by CVT than men.[4]

Similar to venous thrombosis in other anatomical locations, the etiopathogenesis of CVT can be explained by the interplay of the components of Virchow's triad—blood hypercoagulability, blood flow disturbances, and endothelial damage—which disrupt the balance between prothrombotic and thrombolytic processes.[5] Specific underlying risk factors in each patient can trigger elements of Virchow's triad. These factors are typically categorized as either transient/provocative (e.g., infections, oral contraceptives) or permanent (e.g., genetic prothrombotic disorders, neoplasms). The skewed sex distribution is primarily attributed to sex-specific risk factors, including the use of oral contraceptives, pregnancy, puerperium, or hormone replacement therapy.[4] [6] [7] These factors are reported in 64 to 79% of cases in large cohorts.[6] [8] However, in certain geographic regions, an increasing incidence in men has been observed.[9] Another common cause is thrombophilia, either hereditary or acquired. There is an association between CVT and all inherited thrombophilias.[4] [10] [11] More recent research has also linked elevated factor VIII levels to an increased risk of thrombosis, including CVT, though the precise mechanisms are not yet fully understood.[12] Additionally, a large genome-wide association study identified an association between CVT risk and polymorphisms within the 9q34.2 region, which includes the ABO gene. Individuals with non-O blood groups were found to have a threefold higher risk of developing CVT compared to those with blood type O.[13] Inflammatory diseases, such as inflammatory bowel disease,[4] [14] Behçet disease,[15] sarcoidosis,[16] [17] systemic lupus erythematosus,[18] [19] and the presence of antiphospholipid antibodies,[4] [20] are also associated with CVT. Furthermore, myeloproliferative neoplasms[4] [21] [22] [23] and other oncogenic conditions[4] are known to increase the risk of CVT.

Historically, infections, particularly head and neck bacterial infections, were a major cause of CVT,[4] [8] [24] but septic CVT is now less common, with only about 10% of CVT patients showing evidence of an underlying infection.[4] Dural arteriovenous fistulas (DAVFs) have a complex relation with CVT, as they can both result from and contribute to the condition.[25] Neurosurgery, lumbar puncture, and head trauma are associated with an increased risk of CVT.[10] [26] Additional risk factors have been identified in the last decade, including obesity,[27] anemia,[28] [29] SARS-CoV-2 infection,[31] [32] and vaccine-induced thrombotic thrombocytopenia after adenoviral vaccines for SARS-CoV-2.[30] CVT has been a rare but severe complication of the ChAdOx1 NcOV-19 and Ad26.COV2.S vaccines, presenting with thrombocytopenia, increased rates of intracerebral hemorrhage, and high mortality.[30] [33] Interestingly, while both heparin-induced thrombocytopenia and vaccine-induced thrombotic thrombocytopenia involve antiplatelet factor 4 antibodies, CVT is rare in heparin-induced thrombocytopenia, while approximately half of vaccine-induced thrombotic thrombocytopenia cases present with CVT.[34]


Anatomy and Functions of the Intracranial Venous System

The cerebral venous system is divided into a superficial and a deep venous system.[35] Unlike systemic veins, cerebral veins lack valves and do not follow the arterial territories. In addition to transporting venous blood, cerebral sinuses are essential for cerebrospinal fluid (CSF) absorption through the arachnoid villi, which are most prominently located along the superior sagittal sinus and transverse sinuses, particularly near the entry points of superficial veins.[36] The main mechanisms regulating cerebral venous blood flow include passive pressure-driven processes, transmission of pulse oscillations from the arterial side, and modulation by the autonomic nervous system.[37] Due to their morphological characteristics, cerebral venous vessels are particularly influenced by changes in intracranial pressure (ICP). Once ICP changes, cerebral arteries contract or relax via autoregulation (Starling effect),[38] ensuring relative stability of the cerebral perfusion pressure.[39] However, blood pressure in the dural sinuses remains at a constant range—no structural support exists, contributing to their direct exposition to CSF pressure and disturbance by ICP alterations.[40] Under physiological conditions, maintaining homeostasis in venous drainage is essential for normal ICP and effective cerebral blood flow dynamics.


Pathophysiology of Intracranial Venous Thrombosis

Venous thrombi form and flourish in an environment of stasis, low oxygen tension, and upregulation of proinflammatory genes.[41] A venous thrombus has two main components: an inner platelet-rich white thrombus forming the so-called lines of Zahn surrounded by an outer red cell-dense fibrin clot and fibrin and extracellular DNA complexed with histone proteins forming the outer scaffold.[42] According to the classical perspective, arterial thrombosis has long been held to be a phenomenon of platelet activation, whereas venous thrombosis is a matter of activation of the clotting system.[42] However, there is a growing body of evidence supporting that this dichotomy is likely to be an oversimplification[43] and that platelets also play a crucial role in the development of venous thrombi. Specifically, for CVT, data on the pathological mechanisms of thrombosis and thrombus composition remain limited[42] ([Fig. 1]). However, the available evidence indicates that thrombi from patients with CVT are rich in NETs, suggesting that platelet-driven NET formation may be an important step for venous thrombogenesis.[44]

Zoom
Fig. 1 Pathophysiology of cerebral venous thrombosis (CVT). CVT involves factors related to the classical Virchow's Triad. Once the primary hint (venous thrombus) occurs, a cascade of events follows, resulting in (a) decrease of the cerebrospinal fluid absorption due to dural sinus occlusion, leading to intracranial pressure elevation and (b) blocking of venous drainage causing blood flow accumulation and, therefore, increased venous and capillary pressure with blood brain barrier disruption. Overall, it is likely to exist a complex interplay between these mechanisms, all contributing to the disease progression and parenchymal brain lesion. The cornerstone of CVT treatment is anticoagulation. Venous recanalization may play a role in mitigating brain injury. CVT, cerebral venous thrombosis; NETosis, neutrophil extracellular traps; CSF, cerebrospinal fluid; ICP, intracranial pressure; CPP, cerebral perfusion pressure; BBB, blood brain barrier. [rerif]

Pathophysiology of Parenchymal Brain Injury Related to CVT

Following venous thrombus formation, partial or complete occlusion is established within the cerebral venous system (dural sinus and/or vein). The clinical manifestations that occur are heterogeneous and intrinsically related to the pathological changes after the impaired cerebral venous drainage ([Fig. 2]). To explain the symptoms and signs of CVT, two main different mechanisms can be distinguished: (1) increased intracranial hypertension, mostly due to thrombosis of the major dural sinuses, and (2) increased venous pressure, due to thrombosis of the cerebral veins and dural sinuses, with microcirculatory and tissue effects caused by venous obstruction and stasis. In most patients, these two processes occur simultaneously, adding to the complexity of CVT manifestations and contributing to its highly variable disease course. It is estimated that 50 to 60% of patients experience parenchymal changes due to CVT.[4] [45] [46] However, predicting the development, type, and severity of parenchymal lesions remains challenging, even among patients with similar thrombosis patterns. Furthermore, while parenchymal lesions may progress in some patients, they can stabilize or even regress in others. Neuroimaging markers show promise as potential predictors of brain damage development and progression in CVT.

Zoom
Fig. 2 Schematic representation of the modifications in venous pressure and CSF flow dynamics after CVT occurrence. (A) Normal brain. In physiological conditions there is a constant and balanced blood flow through arteries and veins; also, there is a constant shift of bulk water from the capillary space into the ventricle system, contributing to a small percentage of cerebrospinal fluid production. CSF absorption through arachnoid granulations is undisturbed. All these factors assure a normal cerebral perfusion pressure; (a) horizontal brain section, MRI (B) Thrombus formation (CVT). Once a thrombus is present within a dural sinus and/or a cortical vein, venous drainage is blocked; initially, the compliant venous system dilates and collaterals may also contribute to accommodate the increase in venous pressure, at a certain degree. However, increased venous pressure triggers a cascade of pathological events as the disease progresses. (b) MR venography; (C) Intracranial pressure and cerebrospinal fluid dynamics. As venous pressure and transudate volume increase, the transependymal drainage route becomes satured, and a larger percentage of CSF production is secondary to transependymal flow. Additionally, CSF absorption is decreased due to dysfunction of arachnoid granulations. Headache and papilledema are one of the most common clinical features of CVT patients, being related to the increased ICP; (c) papilledema (D) Parenchymal brain damage. The events that follow the increase in venous pressure contribute, ultimately, to secondary brain lesion. When venous pressure exceeds the integrity of the vascular wall (increase in capillary hydrostatic pressure), intraparenchymal hemorrhage may result; (d) horizontal brain section, MRI showing parenchymal lesion. CSF, cerebrospinal fluid; CPP, cerebral perfusion pressure; ICP, intracranial pressure; CVT, cerebral venous thrombosis. [rerif]
  • A. Macrovascular blood flow changes

Thrombus initiation and propagation in cerebral dural sinuses or veins will ultimately cause obstruction of venous flow, forcing venous blood back into small vessels and capillaries. When venous drainage is sufficiently obstructed, venous pressure increases. However, cerebral venous territories are less well defined than arterial territories due to anatomical variations. Also, the presence of extensive anastomoses often facilitates the development of collateral circulation, mitigating the effects of venous flow obstruction. Experimental studies suggested that venous collateral circulation can influence increases in venous pressure and subsequent brain damage.[47] However, to date, the impact of baseline intracranial venous collaterals on the type of brain damage, clinical manifestations, and vital or functional prognosis has not been demonstrated in cohorts of CVT in humans.[48] The role of intrasinusal venous hypertension in the pathophysiology of cerebral venous infarction is also supported by the observation, both in experimental models[49] [50] [51] [52] and in patients with CVT,[53] that an increase in intrasinus pressure, proximal to the thrombosis, correlates with the development and severity of parenchymal changes.

  • B. Microvascular changes and brain tissue injury

Evidence regarding the effects of venous occlusion on the cerebral microvasculature remains limited. Increased venous pressure likely plays a key role in disrupting the blood–brain barrier (BBB). This disruption leads to fluid leakage, resulting in vasogenic edema, and can also contribute to hemorrhagic events. Experimental studies provide further support for this mechanism. Sato et al[54] demonstrated that fluorescein dye extravasation through the vessel wall occurred as early as 5 minutes after sinus occlusion, indicating that BBB disruption may occur during the early stages of dural sinus occlusion. Additionally, significant microvascular basal lamina damage has been observed in a rat model within 3 hours of induced CVT.[55] Compared with models of arterial infarction, this damage was more pronounced in the context of CVT, supporting the hypothesis that BBB disruption is more severe in CVT than in arterial infarction.[55] [56] Matrix metalloproteinases (MMPs) are part of the proteolytic mechanisms involved in BBB disruption.[57] [58] Rashad et al[56] showed a rapid upregulation of MMP-9 in a rat model of CVT, correlating with BBB breakdown and suggesting that MMP-9 may serve as an initial driver of CVT pathology. In a cohort study of CVT patients, those with parenchymal brain lesions had higher circulating MMP-9 levels compared with both controls and patients with CVT without brain lesion.[59] Also, in patients who achieved early venous recanalization, MMP-9 levels tended to decline, suggesting its potential as a biomarker for dynamic changes in CVT hemodynamics.[59] While vasogenic edema is a common manifestation, magnetic resonance imaging (MRI) studies of CVT patients also frequently reveal parenchymal lesions with restricted diffusion, which is usually considered indicative of cytotoxic edema. However, the sequence of these events remains uncertain.[60] [61] Some experimental studies report that cytotoxic edema arises later, following BBB disruption and vasogenic edema, as cerebral blood flow decreases and energy metabolism fails. Others, however, suggest cytotoxic edema may occur earlier, preceding BBB disruption and vasogenic edema.[60] [62] [63] Notably, experimental studies have demonstrated the reversibility of these changes with heparin treatment,[55] [60] highlighting the hypothesis that rapid venous recanalization and normalization of venous pressure may facilitate the recovery of brain lesions. The reversibility of the lesions and recovery of functional impairment would support a penumbra-like state of metabolically compromised yet viable tissue, different from what is usually expected for lesions with diffusion restriction in arterial ischemic stroke models. Similarly, the reversal of brain lesions showing restricted diffusion at baseline can be often observed in patients with CVT.[64] Hemorrhage is another potential complication of CVT, also likely stemming from increased venous pressure.[65] Beyond its immediate effects, hemorrhage introduces blood as a source of iron, which can catalyze free radical formation, exacerbating inflammation and contributing to a proinflammatory environment.[66] [67] However, brain parenchymal damage and BBB disruption can also provoke an inflammatory response[56] and, in cohorts of patients with CVT, increased inflammatory markers are observed regardless of the presence of brain lesions.[68] Nagai et al reported that inhibiting leucocyte adhesion using a CD18 monoclonal antibody reduced CVT-induced brain edema and BBB dysfunction, further supporting the role of inflammation in CVT progression.[69] A delay between MMP9-mediated BBB breakdown and inflammatory changes has also been observed in a rat model of CVT, suggesting that the leakage of blood products through the damaged BBB can contribute to inflammation.[56] In the pathophysiology of venous infarction—Prediction of Infarction and Recanalization in CVT cohort study (PRIORITy-CVT), increased baseline neutrophil-to-lymphocyte ratio, C-reactive protein, and interleukin-6 levels were predictors of unfavorable functional outcome at 90 days.[64] However, these markers were not associated with brain lesion outcomes or early recanalization. Experimental studies have also implicated oxidative stress[70] and endoplasmic reticulum stress as potential pathways contributing to secondary brain injury in CVT models.

  • C. Protection and recovery of brain damage

Experimental evidence suggests that recanalization can be associated with the resolution of brain lesions.[60] [62] [63] Röther et al[60] observed partial resolution of hyperintensities in diffusion-weighted imaging in the parasagittal cortex after treatment with recombinant tissue plasminogen activator. Another study in a rat model of superior sagittal sinus thrombosis demonstrated that early local thrombolytic therapy reduced parenchymal histopathological changes and improved neurological outcomes when compared to anticoagulation treatment.[71] More recently, data from the PRIORITy-CVT study[64] also suggested an association between persistent venous occlusion and a higher risk of worsening of nonhemorrhagic brain lesions in patients with CVT, within the first 8 days after treatment start. Early venous recanalization of previously obstructed sinuses/veins was also associated with nonhemorrhagic parenchymal lesion regression.[64] Meta-analyses of uncontrolled observational data from CVT cohorts suggest better long-term functional outcomes in patients achieving venous recanalization.[72] [73]


Pathophysiology of Increased Intracranial Pressure in CVT

At physiological conditions, CSF is transported from the cerebral ventricles through the subarachnoid spaces to the arachnoid villi, where it is absorbed and drained into the superior sagittal sinus. As mentioned, thrombosis of sinuses leads to impaired absorption of CSF, and thus, elevation in ICP. Moreover, the increased blood volume stranded in the brain may worsen the ICP itself.[38] [40] Some of the most common CVT clinical manifestations are mostly due to the high ICP, represented by headache and visual impairment. Headache is reported in about 90% of patients[4] and although it often presents with typical features suggestive of intracranial hypertension including cranial nerve palsy, there is no specific pattern. Optic papilledema, visual impairment, and cranial nerve palsy are also manifestations associated with this presenting syndrome. In the ISCVT study, at baseline, 28.3% of patients had papilledema, 13.5% had diplopia, and 13.2% had significant visual loss.[4]


Diagnosis—Laboratory Workup and Neuroimaging

Once a clinical suspicion is raised, CVT diagnosis should be confirmed or excluded. A D-dimer assay is often included at the admission biochemical workup. However, although a positive D-dimer supports the diagnosis, a normal result does not exclude the hypothesis of CVT, as up to 10% of patients have a normal value.[74] Categorically, the diagnosis of CVT depends on the demonstration of the thrombi in the cerebral venous system (sinus or vein), requiring brain and vascular imaging for that purpose.[75] [76] The diagnosis is usually based on CT venography or MRI. These image modalities provide two types of findings, reflecting the primary event (venous thrombus) and the secondary parenchymal brain lesions, signs include direct visualization of the thrombus, particularly using MRI, and visualization of the thrombotic material as a filling defect on CT venography and MR venography.[77] [78] [79] Although CT venography can be a reasonable informative method, MRI combined with magnetic resonance venography (MRV) has improved sensitivity,[28] [76] especially for assessing cortical vein thrombosis. Digital subtraction angiography (DSA) has long been considered the gold standard for the diagnosis of CVT, although is currently used almost exclusively to allow interventional radiological procedures, being reserved to confirm the diagnosis only when data from noninvasive imaging examinations are inconclusive or when a DAVF is suspected.[80] However, DSA is still superior to CT-MR venography in terms of dynamic information and can yield important additional data, particularly concerning collateral venous drainage. Venous edema, venous infarction, parenchymal hemorrhage, subarachnoid hemorrhage, and subdural hematoma are all potential manifestations of CVT. An area for potential improvement in CVT diagnosis is the development of tailored frameworks to optimize the evaluation of imaging studies in acute patients with suspected CVT.[81] A recently developed multisequence multitask deep learning algorithm has shown high sensitivity and specificity in the diagnosis of CVT using routine brains.[82]


Management and Therapeutics

We categorized treatments according to the main pathological processes they target, adopting a mechanistic approach.

Interventions Addressing Venous Thrombus and Secondary Prevention of Thrombotic Events

Once the diagnosis of CVT has been confirmed, therapy should be started as soon as possible. The goal of initial treatment is to implement an antithrombotic strategy, aiming the inhibition of ongoing thrombosis, promoting recanalization of the thrombosed vessel, and preventing recurrent venous thrombotic events.[75] [83] [84] Anticoagulation is the mainstay therapy for CVT. Current evidence suggests that low-molecular-weight heparin (LMWH) may be superior to unfractionated heparin in the treatment of CVT.[76] [85] [86] [87] Parenteral anticoagulation is usually followed by oral anticoagulation.[76] Vitamin K antagonists (VKAs) have been used for decades, but more recently, a new body of evidence has been emerging regarding the use of direct oral anticoagulants (DOACs) in the CVT clinical setting. The first study to evaluate the use of DOACs in CVT was the RE-SPECT CVT trial (Efficacy and Safety of Dabigatran Etexilate vs Warfarin in CVT),[88] followed by others, including the SECRET-CVT trial (Study of Rivaroxaban for Cerebral Venous Thrombosis),[89] the CHOICE-CVT trial (Dabigatran Etexilate versus Warfarin in Cerebral Venous Thrombosis in Chinese Patients),[90] and the observational studies ACTION-CVT (Direct Oral Anticoagulants vs Warfarin in the Treatment of CVT)[91] and DOAC-CVT (Direct Oral Anticoagulants for the treatment of Cerebral Venous Thrombosis).[92] Although some results are still pending, the existing evidence suggests that DOACs may be a viable alternative to VKAs for managing CVT patients.

Another potential therapy targeting the venous thrombus is endovascular treatment (EVT), which includes both local thrombolysis and/or mechanical thrombectomy (MT), utilizing various techniques.[93] [94] However, the efficacy and specific selection criteria for these interventional therapies in CVT are still uncertain. The multicenter, randomized clinical trial TO-ACT (Thrombolysis or Anticoagulation for Cerebral Venous Thrombosis) compared EVT combined with the best medical treatment (BMT, including therapeutic anticoagulation) to BMT alone in patients with severe CVT.[95] The trial was neutral for its primary outcome, the modified Rankin scale (mRS) 0 to 1. Despite this, endovascular techniques are continuously improving, and the patient selection criteria in TO-ACT were relatively broad, focusing mainly on predictors of poor prognosis. As a result, EVT is currently reserved for highly selected cases, typically as a rescue intervention when there is persistent clinical deterioration[127] despite standard medical therapy, and when immediate venous recanalization is thought to be potentially beneficial.

The optimal duration of oral anticoagulant treatment after the acute phase is unknown. Recurrent sinus thrombosis occurs in about 2% of patients,[96] [97] and about 4% have an extracranial thrombotic event within 1 year.[7] In making decisions around the duration of anticoagulant treatment, individuals with CVT can be stratified regarding the presence or absence of transient and permanent risk factors for thrombosis. Usually, anticoagulation is maintained for 3 to 12 months after the first episode of CVT, unless there is a previous history of thrombotic events or a condition associated with high thrombotic risk that requires long-term anticoagulation.[75] [76] However, there is a lack of evidence on the optimal duration of oral anticoagulation for the prevention of recurrent CVT and other venous thrombotic events.[76] The results of the ongoing cluster randomized EXCOA trial (EXtending Oral Anticoagulant Treatment after acute Cerebral Venous Thrombosis)[98] should increase the evidence around this issue.

Several studies showed that OCs carry an increased risk of CVT. The association with hormonal factors is stronger for CVT than for other VTE events.[99] Current guidelines recommend that all female CVT patients should be informed about the risks of combined hormonal contraceptives and advised against their use.[75] [76]

As for pregnant women with diagnosis of CVT, due to the increased risk of VTE during the course of pregnancy and early postpartum phase, anticoagulation is recommended until at least 6 weeks postpartum.[76] Observational data collected from pregnant women with history of previous CVT suggest a reduced risk of thrombotic events and abortion in women receiving prophylaxis with LMWH during pregnancy and puerperium, although the relative risk is still higher when compared to the general population of pregnant women.[76] [100] [101]


Interventions Addressing Intracranial Hypertension

In the event of symptoms secondary to increased ICP, causing headache, and threatening vision, carbonic anhydrase inhibitor acetazolamide may be used to decrease CSF production.[76] [102] The objective of the treatment is to lower the ICP, to relieve headache, and to reduce papilledema. However, there are no randomized controlled trials on the effect of acetazolamide in CVT. According to guidelines,[80] there is no specific recommendation regarding therapy with therapeutic lumbar puncture to improve clinical outcome in patients with CVT and signs of intracranial hypertension. However, clinical practice suggests that therapeutic lumbar puncture may be transiently considered in patients with CVT and signs of intracranial hypertension, because of a potential beneficial effect on visual loss, whenever its safety profile is acceptable. In addition, in case the patient experiences progressive loss of visual acuity despite BMT, an interventive approach might be prompted, namely, ventriculoperitoneal or lumboperitoneal shunt[103] or optic nerve sheath fenestration.[104]


Interventions Addressing Other Clinical Signs and Symptoms

Seizures

Acute symptomatic seizures occur in approximately one in three patients with CVT.[105] Once a patient develops seizures, proper management with antiseizure medications[75] [106] is usually required. The choice of antiseizure medication is dependent on individual factors including comorbidities and interactions with other treatments including anticoagulation. Predictors for acute seizure development in CVT include intracerebral hemorrhage, cerebral edema/infarction, cortical vein thrombosis, superior sagittal thrombosis, focal neurological deficit, and sulcal subarachnoid hemorrhage.[105] However, primary prophylactic antiseizure medication treatment is not recommended.


Hydrocephalus and Impending Brain Herniation

About 4% of patients develop supratentorial parenchymal lesions and cerebral edema severe enough to cause brain herniation.[107] Adding to the previous retrospective data,[108] the results from the large international prospective cohort study DECOMPRESS-2 study (Decompressive Surgery for Patients with Cerebral Venous Thrombosis)[109] reinforce the benefit of decompressive surgery in CVT, demonstrating that two-thirds of patients with severe CVT were alive and more than one-third was independent 1 year after decompressive surgery. Among survivors, surgery was considered worthwhile by four out of five patients and caregivers. These findings support the current recommendations from both the AHA/ASA and ESO guidelines, which indicate that decompressive surgery should be performed as early as possible for patients showing clinical and radiological signs of herniation due to large or expanding hemorrhagic or edematous infarcts, in order to prevent death.[75] [76] Hydrocephalus is rare in CVT, but it can occur, especially in patients with deep CVT and basal ganglia edema, in which case neurosurgical consultation for CSF diversion should be considered.[83] [85]




Prognosis, Outcome, and Long-Term Comorbidities

Clinical outcome is often favorable,[110] with around 80% of the patients achieving mRS 0 or 1,[4] [74] [75] and an in-hospital mortality rate of around 4%.[85] The cause of death is generally due to transtentorial herniation, status epilepticus, and/or medical complications.[4] Cohort studies have described indicators of poor outcome[85] including older age,[111] [112] altered mental status,[4] rapid deterioration of consciousness,[76] coma and ICH,[113] venous infarcts,[114] central nervous system infection,[115] [116] malignancy,[21] [117] thrombosis of the deep venous system,[118] and hyperglycemia at admission.[119] Long-term mortality is approximately 8 to 10%[120] and 6 to 10% of CVT surviving patients have mRS 3 to 5.[4] [8] [118] Also, in recent years, attention has increasingly focused on the long-term consequences of CVT in patients who appear to have fully recovered. While approximately 80% of CVT survivors are reported to have no physical impairment, a significant number continue to experience chronic residual symptoms. Headaches and neuropsychiatric difficulties are noted in more than 50% of cases[4] [120] and it is estimated that approximately 20 to 40% of patients are unable to return to their prior working life.[118] [121] In addition, 10% of CVT patients experience seizures during follow-up.[122]


Bridging Bench and Bedside

Effective bench-to-bedside translation in CVT research depends on the careful selection of appropriate animal models. While rats and mice are the most commonly used species, others, such as rabbits,[123] pigs,[124] dogs,[49] and cats,[50] have also been employed. Various techniques exist to replicate CVT in animals, including (1) photochemical induction using ferric chloride or rose Bengal dye; (2) stasis-induced CVST with occlusive thrombi in the superior sagittal sinus (SSS); (3) thrombogenic substance-induced CVST via in situ thrombus injection; and (4) thromboembolic models using endovascular techniques or clot injections into the SSS. Each method has unique strengths and limitations, and no single model fully captures all stages of CVT pathology. An ideal model should be reproducible and representative of the range of injury patterns observed in human cases. The choice of model should align with the specific research question, considering factors such as species-specific venous anatomy, relevant time points, endpoints, available equipment, and technical expertise.

Given the thrombotic nature of CVT, preclinical studies have largely focused on the evaluation of treatments targeting the thrombus, including anticoagulation, local and systemic thrombolysis, and GP IIb/IIIa inhibitors.[71] [125] [126] While these studies have provided valuable insights, the progression of the brain lesions and the mechanisms of brain damage, including neuroinflammation and blood–brain barrier dysfunction, have been less explored.[56] [69] Similarly, clinical CVT research has also centered on thrombus-targeted therapies. While important, this narrow focus overlooks the complex cascade of events triggered by thrombus formation within the intracranial venous system. Parenchymal brain lesions, which play a significant role in determining patient outcomes, remain poorly understood, and the factors predicting their development are still unclear.


Future Directions

Despite recent advances, our understanding of the pathophysiology of CVT and its associated manifestations remains incomplete. Gaining deeper insights into the pathways driving disease progression and brain damage could reveal novel therapeutic targets and refine prediction of prognosis and strategies for patient selection in future clinical trials. Moreover, the development of more nuanced predictors and outcome measures, extending beyond the conventional mRS, is essential for capturing the long-term and subtle effects of CVT on recovery and quality of life.



Conflict of Interest

MM has no conflict of interest to declare. DAS reports advisory board participation for Daiichi-Sankyo, Bayer and Johnson & Johnson, speaker fees from Astrazeneca and Bial, and grants from Fundação para a Ciência e Tecnologia, MSD, and European Society of Radiology.

Data Availability Statement

This manuscript is a narrative review—original data are not implicated.


Statement of Ethics

Ethical approval is not required for this study in accordance with the local and/or national guidelines and journal policy.


Authors' Contributions

M.B.M.: conceptualization and writing.


D.A.d.S.: conceptualization, editing, and review.



Address for correspondence

Diana Aguiar de Sousa, MD, PhD
Neurosciences Department, Stroke Center, Centro Hospitalar Universitário Lisboa Central – ULS São José
Rua José António Serrano, PT-1150-99 Lisbon
Portugal   

Publication History

Received: 02 November 2024

Accepted: 03 February 2025

Article published online:
08 April 2025

© 2025. Thieme. All rights reserved.

Georg Thieme Verlag KG
Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany


Zoom
Fig. 1 Pathophysiology of cerebral venous thrombosis (CVT). CVT involves factors related to the classical Virchow's Triad. Once the primary hint (venous thrombus) occurs, a cascade of events follows, resulting in (a) decrease of the cerebrospinal fluid absorption due to dural sinus occlusion, leading to intracranial pressure elevation and (b) blocking of venous drainage causing blood flow accumulation and, therefore, increased venous and capillary pressure with blood brain barrier disruption. Overall, it is likely to exist a complex interplay between these mechanisms, all contributing to the disease progression and parenchymal brain lesion. The cornerstone of CVT treatment is anticoagulation. Venous recanalization may play a role in mitigating brain injury. CVT, cerebral venous thrombosis; NETosis, neutrophil extracellular traps; CSF, cerebrospinal fluid; ICP, intracranial pressure; CPP, cerebral perfusion pressure; BBB, blood brain barrier. [rerif]
Zoom
Fig. 2 Schematic representation of the modifications in venous pressure and CSF flow dynamics after CVT occurrence. (A) Normal brain. In physiological conditions there is a constant and balanced blood flow through arteries and veins; also, there is a constant shift of bulk water from the capillary space into the ventricle system, contributing to a small percentage of cerebrospinal fluid production. CSF absorption through arachnoid granulations is undisturbed. All these factors assure a normal cerebral perfusion pressure; (a) horizontal brain section, MRI (B) Thrombus formation (CVT). Once a thrombus is present within a dural sinus and/or a cortical vein, venous drainage is blocked; initially, the compliant venous system dilates and collaterals may also contribute to accommodate the increase in venous pressure, at a certain degree. However, increased venous pressure triggers a cascade of pathological events as the disease progresses. (b) MR venography; (C) Intracranial pressure and cerebrospinal fluid dynamics. As venous pressure and transudate volume increase, the transependymal drainage route becomes satured, and a larger percentage of CSF production is secondary to transependymal flow. Additionally, CSF absorption is decreased due to dysfunction of arachnoid granulations. Headache and papilledema are one of the most common clinical features of CVT patients, being related to the increased ICP; (c) papilledema (D) Parenchymal brain damage. The events that follow the increase in venous pressure contribute, ultimately, to secondary brain lesion. When venous pressure exceeds the integrity of the vascular wall (increase in capillary hydrostatic pressure), intraparenchymal hemorrhage may result; (d) horizontal brain section, MRI showing parenchymal lesion. CSF, cerebrospinal fluid; CPP, cerebral perfusion pressure; ICP, intracranial pressure; CVT, cerebral venous thrombosis. [rerif]