Cerebral Venous Thrombosis

Elio Agostoni; Angelo Aliprandi; Marco Longoni


Expert Rev Neurother. 2009;9(4):553-564. 

In This Article

Clinical Features & Diagnosis of CVT

Clinical Features & Presenting Syndromes

The clinical presentation of CVT is highly variable, ranging from cases with mild symptoms and normal neurological examination to others who present with coma and have a severe neurological dysfunction.[21] This variability represents an engaging diagnostic challenge for clinical neurologists and is one of the main difficulties in the field of CVT. However, four main clinical syndromes can be clearly identified, although individual patients may shift from one group to another during the course of the disease.

Patients with isolated intracranial hypertension

Patients with isolated intracranial hypertension have the mildest symptoms and the best prognosis; 22.9% of patients in the ISCVT and 38% in the USA study were in this group. Most patients in this group only have a headache,[22] but diplopia secondary to increased intracranial pressure may also be present. When papilledema is also present (29% of patients; Table 2 ), peripheral visual loss can also be part of the syndrome. However, since papilledema is not found in many patients, its absence does not rule out a diagnosis of CVT.

Patients with focal neurological signs
Patients with focal neurological signs may be both negative (deficits), as a result of a focal cortical damage secondary to a venous stroke (over 40% of patients; Table 3 ), or positive, reflecting a focal epileptic activity. Focal seizures were present in 17% of the patients recruited in the two larger studies available ( Table 3 ), while seizures with generalization (primary or secondary), were present in 29% of patients. Overall, seizures were the presenting symptoms in 40% of patients, a much higher rate than that observed in arterial strokes. Therefore, the presence of seizures in a patient with an ischemic brain lesion should always suggest the possibility of venous thrombosis. Patients who present with seizure should be divided into those with and those without a parenchymal brain lesion, as the former have a higher risk of recurrence. The presence of bilateral neurological signs or the rapid development of controlateral signs, although rare, is highly suggestive of an occlusion of the superior sagittal sinus (3% of patients in the ISCVT and ultrasound studies had this syndrome).

Patients who present with coma or behavioral alterations
Patients who present with coma or behavioral alterations, such as delirium, amnesia or mutism, are associated with a negative outcome, as it is often the consequence of a thrombotic occlusion of the deep or posterior circulation or of occlusion of multiple sinuses. Also, patients with a large supratentorial ischemic or hemorrhagic lesion causing transtentorial herniation may present with this syndrome. In the ISCVT, 14% of patients were in this group (Figure 1).

Figure 1.

Signal hyperintensity on T1-weighted MRI of the right transverse sinus.

Patients presenting with headache, diplopia secondary to the paresis of a cranial nerve, periorbital swelling & chemosis
A few patients (1.3% in ISCVT) present with headache, diplopia secondary to the paresis of a cranial nerve, periorbital swelling and chemosis. This syndrome is nearly pathognomonic of cavernous sinus occlusion. In addition, the timing of onset, as with the presenting clinical features, is variable. In the ISCVT the mode of onset was acute in 37.2% of cases, subacute in 55% and chronic in 7.2%.[11] One of the most frequent symptoms of CVT is headache, which is present in over 80% of patients ( Table 1 ) and is the presenting symptom in nearly 70%.[23] Although the headache is associated with other neurologic signs in the majority of patients, it can also by isolated (20% in the ISCVT). Therefore, the identification of headache features associated with CVT could have a high clinical relevance and may be helpful to identify the patients that need to be studied with imaging techniques. For this reason, we studied a group of 49 patients with a diagnosis of CVT and headache and compared them with 90 control patients with headache of another cause.[14] Unfortunately, headache associated with CVT has no specific features, as it can have an acute, subacute or chronic onset, be localized or diffuse and have any severity. The pain can be associated with nausea and vomiting, and can be both continuous or pulsating ( Table 4 ). When pain is lateralized and pulsating, an erroneous diagnosis of migraine is often made.[24] Moreover, it is worth noting that in as many as 7% of our patients the pain had a hyperacute onset, mimicking subarachnoid hemorrhage.[25,26] We found a positive correlation between CVT, acute headache onset (p < 0.005) and severe headache (p = 0.024). These data were confirmed in a small prospective study on 27 subjects, as we found a positive correlation between CVT, acute headache onset (p = 0.001) and severe headache (p = 0.004). Therefore, in patients with acute or subacute headache onset of severe intensity CVT should always be considered. However, since headache has no specific features, all cases of unexplained headache in patients at high risk, such as subjects with known states of hypercoagulation or women in pregnancy-puerperium or who take OCs, should raise the clinical suspect of cerebral venous occlusion.

Imaging of CVT

The diagnosis of CVT is based on neuroimaging, particularly on MRI and MRI venography techniques that represent the gold standard and might be used as first-line diagnostic tools in cases of high clinical suspicion. However, unenhanced CT scan is still the first examination in the emergency room setting. It may show nonspecific lesions, such as hemorrhages, infarcts or edema in isolation or in combination, but which can be normal in up to 25% of patients.[1] Direct signs of venous thrombosis are seen in only a third of cases. Direct visualization of thrombosis in dural sinus may give a 'dense clot sign' or a 'triangle sign' if the sinus involved in the thrombosis is the superior sagittal sinus. The 'cord sign', although with a low specificity, represents direct visualization of a thrombosed cortical vein that is seen as a linear hyperdensity. The empty delta sign, which may be seen 5 days to 2 months from onset,[27] is the most frequent direct sign of CVT (30% of patients) and can be seen only on enhanced CT scan. It represents a filling defect (thrombus) in the dural sinus, with peripheral enhancement, possibly secondary to the development of collaterals. An empty delta sign may be mimicked by intrasinus septa or by a split or fenestrated dural sinus, which may manifest as false-positive filling defects.[3,4]

More often, a CT scan only shows the indirect signs of CVT. Diffuse or localized brain edema is seen in 20-50% of cases. The edema might appear as hypodensity of the brain, decreased ventricular size and cortical sulcal effacement. An infarction not conforming to a major arterial vascular territory, such as the presence of multiple isolated lesions, involvement of a subcortical region with sparing of the cortex and extension over more than one arterial distribution, is highly suspicious for a venous cause. The location of the infarction may give a clue to the venous structure involved. Thrombosis in the sagittal sinus often leads to impaired venous drainage and, therefore, parenchymal change in the parasagittal region. Thrombosis in Labbé's vein should lead to infarction in the temporal lobe. Bilateral or unilateral infarction in the thalami, basal ganglia and internal capsule is typically seen in thrombosis of the deep venous system. A hemorrhagic infarction is very frequent, while subarachnoid or subdural hemorrhages are rare manifestations of CVT. Finally, indirect evidence of CVT may be seen as contrast enhancement of the falx and tentorium secondary to venous stasis and hyperemia of the dura mater, which is seen in approximately 20% of cases.

The key to diagnosis is the imaging of the venous system itself, which may show the occluded vessel or the intravascular thrombus. The current gold standard for CVT diagnosis is the combination of MRI to visualize the thrombosed vessel and magnetic resonance venography to detect the nonvisualization of the same vessel.[28,29] The abnormal signal intensity of thrombus follows the signal characteristics of intracranial hemorrhage and may evolve through the stages of oxyhemoglobin, deoxyhemoglobin, methemoglobin and hemosiderin.[30,31] Usually, the thrombotic material appears isointense in T1 and hypointense in T2 during the first 5 days, subsequently it becomes hyperintense in both the sequences until 1 month (Figure 1 & 2). Eventually, the thrombus becomes isointense in both sequences and the persisting vessel occlusion might be documented with angio-MRI techniques. Even with the combination of MRI and magnetic resonance venography, the diagnosis can still be difficult, particularly in isolated cortical vein thrombosis, which requires conventional angiography. Considering the potential risk of this invasive procedure compared with the other techniques, the use of conventional angiography has been restricted, but it is still necessary for cases of isolated cortical thrombosis, for CVT associated with subarachnoid hemorrhage to rule out other causes and for its potential therapeutic approach in selected cases (e.g., failure of anticoagulation therapy).

Figure 2.

T2-weighted MRI showing signal hyperintensity suggestive of cortical vein thrombosis.

Some authors have described that on T2-weighted gradient-echo images, exaggerated signal loss is often seen in the first days of CVT, and they concluded that the sequence is useful in isolated cortical venous thrombosis and during the very early days of acute CVT when T1 and T2 lack sensitivity.[32,33] Also, diffusion-weighted imaging may represent a useful tool for CVT diagnosis. In particular, diffusion-weighted imaging examination is able to differentiate vasogenic edema, which is more typical of CVT, from cytotoxic edema, which is more typical of arterial ischemia.[34,35] Moreover, early high-intensity diffusion-weighted imaging signal of the thrombosed vein is associated with a low rate of spontaneous recanalization.[36]

Magnetic resonance (MR) venography is important to visualize the venous system and the occluded vessel with the possible presence of collateral flow; MR venography is not useful for cortical vein thrombosis. MR venography may be performed without the use of a contrast agent using the time-of-flight (TOF) technique. Unfortunately, this technique is subject to flow-related image artifacts. Contrast-enhanced MR venography takes advantage of luminal filling, by contrast, and recently it has been shown to be superior to TOF MR venography. For this reason, the gadolinium-enhanced MR venography might be preferred to the TOF technique, particularly in cases with MRI not exhaustive for CVT diagnosis. Another recent tool that can be used to evaluate CVT is computed tomographic (CT) venography. CT venography allows direct visualization of thrombus as filling defects and is now emerging as a competing technique. It has been shown to be at least comparable to MR venography and, in some situations, to generate more diagnostic information.[37] These two techniques probably provide comparable performance and preference will be dictated by the experience and resources of the individual institutions.

Finally, for the diagnosis of CVT some authors have reported a possible role of the D-dimer test. D-dimer values are usually elevated in patients with neurological deficits; by contrast, a quarter of patients with headache as the only CVT symptom have a normal D-dimer, underlying a poor diagnostic value of this parameter in this subgroup of patients.[38,39]

Prognosis of CVT & Indicators of Outcome

Before imaging techniques became available, CVT was thought to be a condition associated with a high mortality and disability;[40,41] early reports based on autoptic series even lead to the belief that it was invariably fatal. This perception has dramatically changed after the publication of more recent case series that have consistently shown that most patients have a favorable outcome. One of the first reports based on modern diagnostic methods was that of Peters, Ameri and Bousser on a cohort of 77 adult patients who were followed for a mean duration of 77 months.[13] The majority of patients (86%) had no neurologic sequelae, while the remaining patients survived with neurological deficits. Nine patients out of the 11 who had deficits had focal neurological deficits as the clinical presentation. Recurrence of thrombosis was observed in 20% of patients and was always within 1 year from the first event. The ISCVT study confirmed these data in a large group and with a prospective enrollment:[11] after a median follow up of 16 months, 57.1% of patients were asymptomatic and had no signs, 22% had some neurological sign but no disability (modified Rankin score = 1), 7.5% had mild impairment, 2.9% moderate impairment, 2.2% were severely handicapped and 8.3% had died. Predictors of a negative outcome found with multivariate analysis were: male sex, coma, mental status disorder, hemorrhage at admission, thrombosis of the deep cerebral venous system, CNS infection and cancer. Recurrence of sinus thrombosis was very low (2.2%). Wasay et al. reported a higher mortality (13%) and dependency (28% of patients were bedridden);[12] the best predictors of a negative outcome were coma as the presenting symptom and hemorrhage.

Overall, the available evidence indicates that CVT often has a benign course, as most of the patients recover fully or have only mild sequalae. However, a considerable minority, ranging from 10 to 40%, have a negative outcome despite the therapy. The patients most likely to have a negative clinical evolution are those with reduced awareness, hemorrhagic complications or evidence of thrombosis of the deep venous system. These cases could be considered to be candidates for more aggressive therapies, such as local or systemic thrombolysis or mechanical procedures (see 'Five-year view' section).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.