Advertisement

Imaging of Cerebral Venous and Sinus Thrombosis

  • Pedro VilelaEmail author
Living reference work entry

Abstract

Cerebral venous and sinus thrombosis (CVST) is an important cause of stroke, especially in young patients and children, resulting from clot formation and propagation leading to the occlusion of intracranial venous system components, including the dural sinuses, the cortical veins, and the proximal segment of the jugular veins.

Imaging techniques are essential to establish the diagnosis and prognosis and to monitor treatment efficacy. Endovascular interventional neuroradiological procedures may also be useful for venous recanalization in patients’ refractory to conventional treatments.

Cerebral venous thrombosis may be isolated or associated with a wide spectrum of other vascular disorders with direct impact on their natural history and is characterized by a variability of clinical and imaging presentation. Headache is the most common symptom. Focal neurological syndrome (deficits and seizures), intracranial hypertension syndrome, and encephalopathy are also common clinical manifestations. The imaging diagnosis of CVST includes the demonstration of the thrombus/venous occlusion (direct signs) and the presence of intracranial lesions (indirect signs), such as brain swelling, brain edema, venous infarction, intracranial hemorrhage, and decreased CSF absorption. Imaging also permits the understanding of the main pathophysiological hemodynamic changes induced by the venous thrombosis and to predict the outcome.

Clinical neuroradiology plays a central role in the diagnosis and monitoring of CVST and, by means of endovascular procedures, has a potential role for treatment of the most severe and refractory cases.

Keywords

Cerebral venous thrombosis Venous thrombosis Dural sinus thrombosis Cortical vein thrombosis Cerebral edema CT venography MR venography 

Abbreviations

ADC

Apparent diffusion coefficient

BBB

Blood-brain barrier

CBF

Cerebral blood flow

CE

Contrast enhancement

CPP

Capillary perfusion pressure

CSF

Cerebrospinal fluid

CT

Computed tomography

CTV

Computed tomography venography

CVST

Cerebral venous and sinus thrombosis

dAVF

Dural arteriovenous fistula

DSA

Digital subtraction angiography

DWI

Diffusion-weighted imaging

FLAIR

Fluid-attenuated inversion recovery

GRE

Gradient-recalled echo

HU

Hounsfield unit

ICA

Internal carotid artery(ies)

MPR

Multiplanar reconstructions

MPRAGE

Magnetization-prepared rapid acquisition gradient-echo

MR

Magnetic resonance

MRI

Magnetic resonance imaging

MRV

Magnetic resonance venography

PC

Phase-contrast

SS PC

Single-slice phase-contrast

TOF

Time-of-flight

VA

Vertebral artery(ies)

WI

Weighted Imaging

Definition of Entity and Clinical Highlights

Cerebral venous and sinus thrombosis (CVST) is a disease resulting from clot formation and propagation leading to the occlusion of intracranial venous system components, including the dural sinuses, the cortical veins, and the proximal segment of the jugular veins.

In 1825, the French physician Ribes made the first description of dural sinuses thrombosis in a patient presenting with severe headaches and seizures. Since then, several advances have developed in the diagnosis and treatment of this disease.

To date, the major multicenter studies evaluating CVST were the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT), consisting of a registry with approximately 624 patients, the Cerebral Vein Thrombosis International Study (CEVTIS) corresponding to a registry with approximately 706 patients, the Canadian Pediatric Ischemic Stroke Registry (CPIS) that is a registry with approximately 160 pediatric patients, and the recent TO-ACT trial that evaluated the efficacy of different endovascular revascularization techniques for the treatment of worst prognosis cases (DeVeber et al. 2001).

CVST is an important but not frequent cause of stroke, accounting for only 1 to 2% of them, with a strong female predominance and a peak age incidence far younger than in the arterial stroke, between 30 and 40 years. Cerebral venous and sinus thrombosis may involve single or multiple (intracranial and/or extracranial) venous structures, such as dural sinus, and deep or superficial veins.

Opposite to arterial stroke, CVST’s clinical presentation and imaging findings are remarkably variable, making CVST a diagnostic challenge, for both clinicians and neuroradiologists.

Also, the prognosis uncertainty and the absence of an optimized treatment make this disease challenging. CVST has also direct impact on the natural history of the other venous diseases, such as dural arteriovenous fistula and brain arteriovenous malformations as discussed in other chapters.

Basic Epidemiology/Demographics/Pathophysiology

Epidemiology and Demographics

The overall adult incidence ranges from 0.3 to 1.5 per 100,000, representing about 1–2% of all strokes (Stam 2005). There is a higher incidence among females due to a peak incidence during peri- and postpartum periods and in association with oral contraception.

In children the incidence is also higher, especially during the first year of age (approximately 40/100,000), with a peak incidence during the first 3 months of life (DeVeber et al. 2001). Other high-risk patients for CVST include patients with hematological disorders, such as hypercoagulable states, and with malignancies (Stam 2005).

There has been a tendency for an increase in CVST incidence, due to higher clinical awareness and improving imaging diagnosis, resulting from the widespread use of noninvasive imaging modalities, which increased the diagnosis of the less severe cases. On the other hand, the CVST mortality rate is decreasing, mainly due to the increased diagnosis of less severe cases, the better supportive clinical care, and the shift of the underlying etiology, with lesser number of infectious cases (Coutinho et al. 2014b).

Pathophysiology

There are an extremely large number of causes for CVST that can be gathered into three major pathophysiology mechanisms:
  1. 1.

    Vessel wall disease, associated with infective/inflammatory phlebitis, malignant infiltration, venopathy, and trauma

     
  2. 2.

    Venous blood flow constraint with venous obstruction and stasis and/or increased venous blood viscosity

     
  3. 3.

    Congenital or acquired hypercoagulable state

     

In more than 85% of cases, the underlying cause(s) can be identified, and in more than 1/3 of cases, multiple causes concur to the development of CVST (Stam 2005).

The pathophysiology of CSVT remains to be completely understood, but venous hypertension is accepted as the most important mechanism underlying all the brain abnormalities associated with this disease (Fig. 1a, b).
Fig. 1

(a)–(b) Schematic summary of the pathophysiological changes associated with intracranial venous thrombosis and hypertension. CVST cerebral venous and sinus thrombosis, CBF cerebral blood flow, BBB blood-brain barrier

The intracranial venous system that accommodates 70–80% of all intracranial blood has several particularities, different from the arterial system, which are important for the comprehension of the CVST pathophysiology. The venous anatomy, in terms of number, disposition, and size of veins and sinuses, is extremely variable. This variation is greater for the superficial venous system than for the deep venous system (with the exception of the basal vein of Rosenthal), which may explain the smaller variability of the clinical presentation in deep venous thrombosis. There is a remarkable potential for venous anastomosis, between intracranial and extracranial, superficial and deep, and infra- and supratentorial venous systems. The intracranial veins have bidirectional flow, since they possess no valves preventing the back flow of venous blood. This allows the development of new venous outlets for venous drainage in the presence of a venous occlusion, providing tremendous possibilities for venous collateral circulation development.

Acute venous occlusion is responsible for an increase venous pressure. The extent of CVST-related intracranial abnormalities is dependent on both the extent of the brain drained by the venous obstructed outlet and the brain ability to have alternative venous drainage routes (collaterals). These factors are dependent on the venous anatomy distribution and venous anastomosis, which are highly variable among individuals.

Since the venous anatomy is highly variable, the prognosis of CVST may be very different among subjects with similar venous thrombosis, regarding location and extension. The venous collateral status and other venous compensatory mechanisms are also highly variable between subjects and may thus explain the different outcomes in patients with CVST.

Venous compensatory mechanisms, such as venous and capillary dilatation and/or recruitment, may accommodate the excess of blood, and the venous collateral circulation regulation, with distal venous flow reversion, recruitment of adjacent venous territories, and venous outlets, may be sufficient to overcome the venous obstruction without causing brain lesions.

If such compensation does not occur, venous hypertension is transmitted backward, increasing venous and capillary pressure and reducing capillary perfusion pressure. This causes the slowing of the venous flow and the extension of the venous thrombosis, promoting a positive feedback mechanism for venous thrombosis propagation.

With progressively increasing venous hypertension, a pathophysiological cascade is initiated, such as cerebral blood volume increase, intracranial pressure increase, decreased reuptake of extracellular fluids, brain extracellular edema, reduced capillary perfusion pressure and cerebral blood flow, blood-brain barrier (BBB) disruption, venous infarct, and hemorrhage.

The increase of the capillary hydrostatic pressure and/or the BBB disruption associated with venous thrombosis leads to the leakage of fluid into the extracellular space, causing extracellular edema, which is reversible if the underlying caused is eliminated and the capillary flow is sufficient to maintain brain tissue metabolism.

Brain edema may cause the compression of capillaries and venules worsening the venous outflow and promoting the progression of the venous thrombosis.

The decrease of the cerebral blood flow, which seems to be caused mainly by the decrease of the capillary perfusion pressure and increase of the venous pressure, may lead to ischemia and cytotoxic edema. The increased venous and capillary pressure and corresponding reduction of the capillary perfusion pressure and cerebral blood flow (CBF) are the main causes underlying venous infarcts.

The venous compartment (venules and veins) is the most frequent site for BBB disruption. This disruption is more likely a consequence of increases in venular than arterial pressure. The venous thrombosis BBB breakdown is related with both increased venular pressure and endothelial dysfunction and correlates with the presence of plasma biomarkers of BBB disruption.

Hemorrhage is the result of small veins’ rupture due to the venous hypertension that may be potentiated by concomitant ischemic capillary necrosis. The intracranial hemorrhage may occur in any brain compartment. Hemorrhage also results from the coalescence of small petechial hemorrhages to present with larger parenchymal hematomas. Subarachnoid hemorrhage associated with cortical venous thrombosis also results from the venous pressure increase in these veins.

Dural sinus thrombosis may also cause increased intracranial pressure and decrease cerebrospinal fluid (CSF) absorption by the venous hypertension or by the obstruction of the arachnoid’s villi in the superior sagittal sinus thrombosis.

It is not uncommon to find different types of lesions caused by the several abovementioned pathophysiological mechanisms in the same patient.

Clinical Scenario and Indications for Imaging

Clinical Scenario

The CVST clinical presentation is remarkably variable.

Subacute presentation (ranging from 7 days to 1 month) is the most common mode of onset accounting for more than half of cases, followed by acute presentation in one third of cases and, less commonly, chronic presentation. The latency period, between the venous thrombosis development and the clinical symptoms onset is generally unknown, which makes the imaging findings interpretation more difficult, as will be explained later (Stam 2005).

The clinical presentation depends on the veins or sinus thrombosed, on the rate of the thrombosis progression, on the individual venous collateralization, and on the rate of recanalization. Clinical presentation also varies accordingly to the age of patient and presence of an underlying disease.

There are three major clinical presentations, namely, focal neurological syndrome, intracranial hypertension syndrome, and encephalopathy (Stam 2005; Saposnik et al. 2011).

Headache is the most common single symptom, being present in more than 85% of cases, followed by focal neurological signs and seizures. Headache (without any other clinical symptoms or signs) may be the only clinical manifestation. The presence of headache and focal neurological signs should raise the suspicion of CVST.

Focal neurological syndrome includes seizures, which occur frequently in the pediatric population, and diverse deficits depending on venous thrombosis site. Sinus thrombosis typically causes symptoms of both cerebral hemispheres.

Encephalopathy presentation with multifocal signs and mental status change, which can range from slight decreased alertness to coma, is more frequent in the elderly population and in deep venous thrombosis.

Intracranial hypertension may manifest with headache, nausea and vomiting, papilledema, and visual disturbances such as diplopia due to VI nerve palsy, among others.

Other clinical presentations may be present depending on the CVST location, such as a cavernous sinus syndrome.

In pediatric population, seizures are the most common clinical presentation in neonates (approximately 70% of cases) and are present in up to 50% of older infants and children. Increased intracranial pressure is the major form of clinical presentation in older infants and children, manifesting with headache and/or papilledema. In neonates, the presence of tense fontanel, splaying of cranial sutures, and dilated scalp veins are alarm signs for intracranial hypertension (DeVeber et al. 2001).

It should be emphasized that the severity of clinical symptoms is not necessarily correlated with the extension of the thrombosis and that vessel recanalization does not necessarily parallel the clinical recovery. Clinical improvement may occur much earlier than the imaging signs of complete recanalization of the affected veins or sinuses.

Indications for Imaging

Imaging is essential for the diagnosis of CVST, as there is no laboratory test that can make the diagnosis (Saposnik et al. 2011; Ferro et al. 2017).

For the initial evaluation, imaging should be able to:
  1. 1.

    Make the diagnosis of CVST by showing the venous occlusion (direct imaging signs).

     
  2. 2.

    Depict the presence of parenchymal and other intracranial lesions (indirect imaging signs).

     
  3. 3.

    Identify the venous thrombosis cause, such as the presence of infection or malignancy, requiring specific treatment; establish the prognosis; and identify the presence of risk factors that predict a poor outcome.

     

For follow-up, imaging is used to evaluate the recanalization of the venous thrombosis, to guide the treatment strategy, and to depict chronic complications and/or recurrence.

Imaging Technique, Findings and Pitfalls, and Recommended Protocol

Imaging Technique

Vascular and parenchymal imaging is decisive to make the diagnosis of CVST and to evaluate the presence of intracranial lesions. The combination of CT/CT venography (CTV) or MRI/MR venography (MRV) is highly accurate for the evaluation of CVST.

MRI and MRV have the advantage of better evaluating parenchymal lesions, not exposing the patient to radiation and most of the times not requiring the administration of contrast agent, but require patient collaboration due to the long imaging acquisition time.

Imaging Findings

The CVST imaging findings may be divided into direct signs (of the thrombosis) and indirect signs (intracranial complications).

The direct signs of CVST rely on the visualization of the thrombus and/or absence of venous flow on the CT or MR venograms and venous phase of digital subtraction angiography (DSA).

CT is generally the first-line study in the emergency setting, but non-contrast-enhanced CT has a low sensitivity, being positive on only one third of cases.

On non-enhanced CT, there may be hyperdense thrombus inside the dural sinus (dense triangle sign) or cortical vein (cord sign) associated with an enlargement of the cortical vein and/or sinus, the latter showing a round shape with convex posterior wall. The thrombus inside the venous structure has a similar density evolution to a parenchymal hemorrhage, with progressive hyperdensity decrease during the first 7 to 30 days (Fig. 2).
Fig. 2

Non-contrast brain CT. Axial plane. (a) Observe the spontaneous hyperdensity of the right transverse sinus, (dense triangle sign) and cortical vein draining into the transverse sinus (cord sign); (c) and (b) Same patient at 6 and 30 days illustrating the temporal evolution of venous thrombosis with normalization of the initial findings

CT multiplanar reconstruction (MPR) views improve the CT accuracy for depicting CVST, especially those affecting the cortical venous thrombosis. The CT false-negative cases include subacute and chronic thrombus or partially recanalized thrombus.

There are several conditions that may mimic the hyperdense appearance of venous thrombosis, such as higher hematocrit in hematological disorders (like in polycythemia) or in dehydration states, and in normal pediatric subjects that generally have higher hematocrit, which in association with the normal low density of the brain white matter may give the false (qualitative) impression of venous thrombosis.

The quantitative Hounsfield unit (HU) evaluation of the venous structures, such as a close comparison with the arterial attenuation, or quantitative evaluation of venous attenuation or the ratio between the venous attenuation and the hematocrit values, may increase the diagnostic accuracy of CT. Venous HU levels equal or above 62–70 and the ratio hematocrit/dural sinus (H:H) above 1.4 have higher sensitivity (above 80%) to depict CVST (Buyck et al. 2013). However, several factors may limit this evaluation, such as the variation of the vessel attenuation with gender and age; perfused/thrombosed vessel; physiological variables (hematocrit); type of vessel, artery/vein; the vessel attenuation heterogeneity; or technical issues like region of interest size and placement, among others.

After contrast administration there may be a venous non-enhancing central filling defect, hypodense relative to the intense surrounding dural enhancement, due to dilated dural venous channels (empty delta sign). Since the thrombus may be spontaneously hyperdense, it is important to obtain a non-contrast CT to compare with the contrast-enhanced (CE) CT images (Fig. 3a–c).
Fig. 3

Contrast-enhanced CT/CTV. Axial plane. (a)–(c) show empty delta sign, corresponding to the venous non-enhancing central filling defect representing the thrombus inside the superior sagittal and straight sinuses; (d) demonstrates the asymmetric timing of sigmoid sinus venous filling with contrast, with a physiological left delay without the presence of venous thrombosis

On MRI the thrombus signal intensity is time dependent, varying accordingly to the hemoglobin degradation products that are present. In the same patient, thrombi in different evolution phases are commonly seen. The initial thrombosis can be deduced from this MRI information.

There are four major thrombus states:
  • Hyperacute thrombus (Fig. 4). This is seldom seen since most cases have a subacute presentation. In this phase the thrombus is generally isointense on T1WI and hypointense on T2WI due to the presence of oxyhemoglobin.
    Fig. 4

    Coronal T1 (a) and T2 (b) WI MRI show venous thrombus in different phases: superior sagittal sinus has thrombus in the early and late subacute phases. Hyperacute thrombus in a cortical vein in the right lateral brain convexity

  • Acute thrombus (Fig. 5a–d). This phase corresponds to the first week (between days 1 and 5–7), in which the sinuses are generally enlarged and the thrombus is isointense on T1WI and hypointense on T2WI, due to the presence of deoxyhemoglobin.
    Fig. 5

    MRI in different evolution phases. Acute phase thrombus: isointense on T1WI (a, b), and hypointense on T2WI (c/d); early subacute phase thrombus, showing the hyperintensity centripetal progression, from the periphery to the center of the vessel, due to the conversion of deoxyhemoglobin into methemoglobin (e/f); subacute phase thrombus, hyperintense on T1WI and T2WI (g, h)

  • Subacute thrombus. This phase corresponds to the period between the first and the third/fourth weeks. Immediately, after the acute phase, there is an intermediate phase, corresponding to the late acute/early subacute phase, in which the thrombus becomes hyperintense, firstly on T1WI, (high signal on T1WI and low signal on T2WI) corresponding to intracellular methemoglobin. In large vessels the changes proceed in a centripetal fashion, from the periphery to the center of the vessel, due to the conversion of deoxyhemoglobin into methemoglobin (Fig. 5e–f). In subacute phase, corresponding to the second week of thrombosis, the thrombus becomes completely hyperintense on both T1 and T2WI in relation to the presence of extracellular methemoglobin (Fig. 5g–h). After gadolinium administration, there is also an empty delta analogue with intense peripheral dural enhancement and no central filling.

  • Chronic thrombus. In this phase, which occurs after 3 weeks, the affected sinus reduces its size, and there is a wider variability of the thrombosed sinuses MRI signal intensity. Generally, the thrombus becomes isointense on T1WI and hyperintense on T2WI and has intense enhancement due to the organized and vascularized thrombus, but minimal or absence of enhancement may also occur. Furthermore, flow voids inside the thrombus may be seen, especially on the T2WI, representing recanalized channels.

In acute thrombosis, due to the T2 “blackout” effect of deoxyhemoglobin, the thrombus will be both hypointense on DWI and ADC maps (Fig. 6a). DWI may show present restricted diffusivity inside the thrombus, with high signal on DWI and low signal on the corresponding ADC maps (Fig. 6b). The presence of restricted diffusivity is associated with lower short-term recanalization rate (Favrole 2003).
Fig. 6

DWI MRI. (a) Axial ADC map and DWI show the low signal of the right transverse sinus thrombus; (b) axial ADC map and DWI show the low signal on ADC map and high signal on DWI of the superior sagittal sinus (restricted diffusivity)

On gradient-recalled echo sequences (GRE; T2*), the thrombus may exhibit magnetic susceptibility effect with an extended hypointense signal (blooming effect) (Fig. 7). The blooming effect is more pronounced in red thrombus that predominates in venous thrombosis. T2*WI and T2-FLAIR generally have opposite signal intensities in patent sinus but may show similar signal intensity change in the presence of thrombosis. T2*WI sequence is especially important for the detection of cortical venous thrombosis, and coronal images might be more beneficial, with a sensitivity above 90% (Altinkaya et al. 2015). On susceptibility-weighted imaging (SWI), the thrombus may appear markedly hypointense due to the higher deoxyhemoglobin content. SWI also shows the extension of the venous stasis and collateral slow flow (Fig. 7).
Fig. 7

Right transverse sinus thrombosis. Axial T2 WI (a), T2*WI (b), and SWI (c) demonstrate the marked low signal on T2* and SWI due to the blooming effect

Other MRI sequences have been advocated for the diagnosis of CVST, such as contrast-enhanced 3D T1WI and black blood acquisitions. The 3D MPRAGE (magnetization-prepared rapid acquisition gradient-echo) contrast-enhanced T1 WI may allow the direct identification of the venous filling defects. To avoid false-negative findings resulting from thrombus enhancement, the image acquisition should be initiated immediately after contrast injection, and the comparison with non-contrast-enhanced T1 images to rule out hyperintense (subacute) thrombus is also required. Black blood MR sequences improve the blood signal intensity suppression and reduce the flow artifacts (hyperintensity on T1WI due to slow flow) increasing the accuracy of MRI to depict venous thrombosis.

Venous angiograms, both by CT, MR, and less frequently by DSA, complement the evaluation of venous thrombosis.

CT venography (CTV) is a fast and reliable imaging modality to demonstrate intracranial venous thrombosis, especially dural sinus thrombosis, and has the additional advantages of being widely and immediately available (Table 1).

CTV is not sensitive to flow artifacts or flow direction; its interpretation is straightforward and based on the presence or absence of contrast inside the venous structure (e.g., the empty delta sign) (Rodallec et al. 2006). It has a high sensitivity and specificity both above 90%. The image acquisition should be delayed in 45–60 s after contrast injection or with manual start. The manual acquisition starts after the filling of the thinner common jugular vein distal to the confluence with the external jugular vein provides sufficient time for full intracranial venous filling with contrast (Fig. 3d).

A close comparison between non-contrast-enhanced and contrast-enhanced images is recommended to confirm the absence of venous enhancement, especially whenever the thrombosed vein/sinuses are spontaneously hyperdense.

The evaluation of the vessel structure, in this case of dural sinus, with its bone impression/foramen (which is proportional to its original size) allows the recognition of a hypoplastic sinus.

New CT units allow the acquisition of dynamic 4D CTA/V, imaging the different vascular phases: arterial, capillary, and venous. This is a major step forward, as it reduces the possibility of error, such as those cases of non-thrombosed sinuses with slow flow; it allows the evaluation of the venous collateral circulation and/or the recanalization/partial thrombosis; and it depicts the secondary dural arteriovenous fistulae development.

There are three major MRV techniques commonly used for the diagnosis of CVST, namely, phase-contrast (PC MRV), time-of-flight (TOF MRV), and 3D contrast-enhanced (3D CE), the latter having the possibility to be dynamic (4D CE) (Table 1; Fig. 8).
Fig. 8

MR venography of sinus thrombosis. (a) Sagittal T1 SE and 2D PC (VENC 10) showing the absence of straight sinus vascular signal; (b)–(c) Sagittal T1 SE and 3D PC (VENC 10) demonstrating extensive superior sagittal and straight sinuses and bilateral asymmetric transverse sinuses thrombosis; (d)–(e) Sagittal T1 SE and 3D TOF illustrating an extensive superior sagittal and straight sinuses and bilateral asymmetric transverse sinuses thrombosis; (f)–(g) 3D CE MRV and coronal source images exhibiting superior sagittal sinus (partial) thrombosis and right (complete) transverse sinuses thrombosis

Table 1

Venous vascular imaging techniques

Imaging technique

Advantages

Disadvantages

Pitfalls

CTV

Widely available

Rapid acquisition

Indicated for unstable/no cooperative patients

Images not related to flow artifacts

Evaluation of the skull base (foramina size) to distinguish hypoplastic sinus from partial thrombosis

Lower cost

Invasive (contrast administration)

Exposure to ionized radiation

Nonimmediately repeatable (if needed)

Need to compare with NC CT images (high attenuating thrombus may mimic contrast filling)

 CTV

Widely available (any type of CT scanner can obtain a diagnostic CTV)

Need for correct timing acquisition to obtain the venous phase (slow flow may mimic thrombosis)

Monitor the contrast bolus arrival to the common jugular vein on the side of the suspected thrombosis and/or smaller size common jugular vein

 4D CTV

Obtains a precontrast phase to depict high attenuating venous thrombus before contrast arrival

Avoids the need for proper acquisition time – slow-flowing vessels will be identified

Only available with recent CT scanners

Exposure to a higher ionized radiation dose

May need to look into the different phase of acquisition to depict partial thrombosis

MRV

Combination with MRI allows better depiction of brain lesions (indirect signs) and cortical venous thrombosis

No exposure to iodine radiation

Noninvasive in case of NCE MRV (no need for contrast administration)

Immediately repeatable (if needed) in case of NCE MRV

Longer acquisitions – more prone to motion artifacts

Less suitable for uncooperative patients

Higher cost

Pitfalls are mainly related to the MR sequence used

 NCE MRV

TOF

2D

High contrast between vessel and background (stationary) issue

Shorter scan times

Minimal saturation effects

Sensitive to slow flow

No need for choosing the adequate venous velocity

Low SNR (signal-to-noise ratio)

Low spatial resolution (thick slices)

Poor in-plane flow sensitivity (in-plane saturation effects)

Prone to flow artifacts that may originate flow gaps (false positive cases)

Poor background suppression

Short T1 (high T1 signal; postgadolinium) lesions may be mistaken by vessels

Double oblique plane (sagittal and coronal) acquisition reduces the vascular signal loss occurring when the acquisition plane is parallel to the venous flow

Subacute thrombus (high T1 signal) may give false negative results (need to compare with the T1 signal intensity)

3D

High SNR (signal-to-noise ratio)

High spatial resolution (thinner slices <0.4 mm)

Blood signal is easily saturated by slow flow (distal vessels; stenosis)

Poor background suppression

Short T1 (high T1 signal; postgadolinium) lesions may be mistaken by vessels

Longer TR acquisitions with thinner increase the sensitivity to slow flow

Double oblique plane (sagittal and coronal) acquisition reduces the vascular signal loss occurring when the acquisition plane is parallel to the venous flow

Subacute thrombus (high T1 signal) may give false negative results (need to compare with the T1 signal intensity)

PC

2D/SS

Very fast sequence (acquisition times lesser than 1 min)

May be acquired in different orientation planes: sagittal, coronal, and axial planes to increase the accuracy for the superior sagittal and straight sinus, torcula and transverse sinus, and sigmoid sinus

Helpful to choose the correct velocity (VENC) for individual venous evaluation before the acquisition of the 3D PC MRV

Good background suppression (not affected by thrombus high T1 signal)

Low spatial resolution (single thick slice)

Vessel overlap artifact (may miss partial thrombosis)

Lower sensitivity to turbulent flow that may originate flow gaps (false positive cases)

Absence of sinus visualization does not necessary means thrombosis could be the result of turbulence or intravoxel dephasing artifacts

Due to the low spatial resolution, false negative cases of partial thrombosis and partial recanalization may occur

3D

Excellent background suppression (not affected by thrombus high T1 signal)

High SNR (signal-to-noise ratio)

Higher spatial resolution

Higher sensitivity to slow flow

Variable velocity sensitivity

Allows qualitative and quantitative flow velocity and direction analysis

Longer acquisition time

Need to choose the correct VENC (velocity encoding)

PC MRV obtained after contrast administration have a higher accuracy to show small/slow flow vessels

 CE MRV

3D

Shorter acquisition times (compared with 3D PC or 2D TOF)

Signal results from the T1 shortening due to the intravascular contrast (gadolinium)

Signal is not influenced by slow, turbulent flow, acquisition plane orientation or the presence of hypoplastic sinuses

Better distinction between slow or turbulent flow and thrombosis

Higher special resolution

Sensitive to small vessels and/or slow flow

Need for contrast administration

No need for special correct bolus timing and acquisition synchronization

No temporal resolution (hemodynamic information)

 

4D/Dynamic

Hemodynamic information: better evaluation of recanalization degree and the exclusion of dural arteriovenous fistula (dAVF)

Need for contrast administration

No need for special correct bolus timing and acquisition synchronization

Lower spatial resolution

 

CTV computed tomography venography, NCE noncontrast-enhanced, CE contrast enhancement, TOF time of flight, PC phase contrast

Single-slice phase-contrast (SS PC) MRV is a fast sequence, with acquisition times of less than 1 minute. It may be acquired in sagittal, coronal, and axial planes to increase the accuracy for detecting the superior sagittal and straight sinus, torcula and transverse sinus, and sigmoid sinus, respectively. The absence of sinus visualization does not necessarily mean thrombosis, as in this sequence turbulence or intravoxel dephasing generates signal loss (greater than with a TOF MRV sequence), and complementary MRV is needed. Depiction of the dural sinus rules out complete dural sinus thrombosis. However, one should be aware that the spatial resolution of this sequence is extremely low, and as a consequence partial thrombosis and partial recanalization may be missed. It may be used to choose the correct velocity (VENC) for individual venous evaluation before the acquisition of the 3D PC MRV.

The 3D PC MRV provides excellent background suppression, not being affected by the high T1 signal of the thrombus, and has high spatial resolution and sensitivity to slow flow present in the venous system, at the expenses of longer acquisition times.

Time-of-flight (TOF) sequences have good spatial resolution. The major disadvantage is being less sensitive to slow flow, such as flow in hypoplastic sinuses; to in-plane flow, flow saturation when the slices are parallel to the vessel; and to turbulent flow patterns that may give signal loss. In normal subjects, flow gaps may be encountered in up to 1/3 of cases when associated with hypoplastic sinus. These artifacts are more pronounced in the 3D acquisition than in the 2D acquisition, which is recommended. Parameters may be optimized taking into consideration that the sensitivity to slow flow increases with longer TR acquisitions and thinner slices. Since there is vascular signal loss if the acquisition plane is parallel to the venous flow, double oblique plane (sagittal and coronal) acquisition is recommended. On TOF sequences, the T1 shortening of a subacute thrombus may give false-negative results, mimicking blood flow. The use of contrast followed by a TOF venogram increases the visibility of small venous structures, but the enhancement of a normal or pathological structure may also mimic blood flow in the sinus. 2D TOF MRV with a double oblique plane (sagittal and coronal) acquisition is recommended for venous evaluation.

Contrast-enhanced MRV has shown to better demonstrate the venous structures compared to any other type of MRV. This is the standard MRV reference for CVST diagnosis (Saposnik et al. 2011).

Contrast-enhanced MRV is accurate for the diagnosis of venous thrombosis in all phases, including in the chronic phase, as can depict the partially recanalized channels. This sequence is not influenced by slow, turbulent flow, acquisition plane orientation or the presence of hypoplastic sinuses. The use of a sagittal orientation reduces the size of the section and the acquisition time. It can be used in combination with TOF MRV for a more confident diagnosis. The enhancement of the chronic thrombus considered by mistake as venous flow is not a current problem for CE MRV, since imaging acquisition duration is very fast and acquired at the peak of the venous phase. Some other pitfalls to take into consideration are the limited spatial resolution, the intracranial enhancement of structures, and lesions, such as the meninges or intracranial tumors, which may be in close topographic relationship with dural sinuses/cortical veins.

Dynamic 4D MRA of arterial and venous phases may be very effective for the diagnosis of venous thrombosis. It provides images at different times allowing the distinction between slow flow and thrombosis and the evaluation of recanalization degree and the exclusion of dural arteriovenous fistula (dAVF). It also dispenses the need for correct timing of the acquisition for depicting mainly the venous phase, since several time phases are acquired. The major drawback is that the higher temporal resolution chosen, the lower spatial resolution is obtained (and vice versa).

Nowadays, digital subtraction angiography (DSA) is reserved for patients requiring endovascular treatment. The abovementioned imaging signs are also valid for DSA. Since DSA is often performed with selective arterial injections, it is mandatory to have absence of venous flow in all intracranial vessel injections (both ICA and VA) to confirm intracranial venous thrombosis.

Isolated cortical venous thrombosis represents the major challenge for CVST diagnosis (Coutinho, Gerritsma, et al. 2014a). On CT, the diagnosis is made by the demonstration of the cord sign. Generally, MPR views are needed to depict the vein on its superficial trajectory. MRI is the best noninvasive method for the diagnosis. Cortical venous thrombosis is best depicted by the combined information of the T1 WI, T2 WI, the source images of MRV, and T2*WI (Fig. 9). The latter, is the most accurate sequence for this diagnosis. Coronal acquisitions are preferable to depict the cortical venous thrombosis near the superior sagittal or transverse sinuses.
Fig. 9

Cortical vein thrombosis. Axial non-contrast CT (a, b) shows a small brain hemorrhage with focal cortical subarachnoid hemorrhage and a hyperdense cortical vein (arrows). Coronal T1 WI (c), T2 WI (d), T2*WI (f); TOF SI (g) MRI demonstrate the concordance of acute thrombus signal intensity on the different sequences

The most common intracranial complications of venous thrombosis are brain edema with swelling, decreased CSF absorption with hydrocephalus, brain and subarachnoid hemorrhage, and venous infarction (Fig. 1).

Imaging studies, especially MRI, show the wide spectrum of brain abnormalities and these different types of intracranial complications, which may coexist simultaneously in the same patient. Since the physiopathology of venous lesions is distinct from the arterial stroke, the prognostic value of similar abnormalities, such as DWI and perfusion changes, should not be applied, since there is a high degree of reversibility in venous brain lesions. From this perspective, the term “venous infarction” should be used with caution. Arterial infarction means irreversible brain tissue injury. “Venous infarction” has been used interchangeably with venous or congestive edema that can disappear completely with the restoration of venous drainage (Fig. 10).
Fig. 10

Brain MRI in a patient with straight sinus thrombosis. (a) Sagittal T1WI shows the hyperintense thrombus; (b) axial T2-FLAIR demonstrates bilateral asymmetric edematous brain lesions in the deep venous territory; (c) axial ADC/DWI reveal the presence of concurrent vasogenic/extracellular (high ADC) and cytotoxic (low ADC) edema; (d) follow-up axial T2-FLAIR demonstrates the complete reversibility of the lesions

Distinct types of edema may be present in venous thrombosis: cytotoxic (decreased ADC values) and congestive (extracellular edema with increased ADC values). Extracellular edema, resulting from BBB dysfunction, predominates over cytotoxic edema. Both types of edema may coexist simultaneously in the same patient, and extracellular edema may be present from the beginning. The edema timing is distinct from arterial strokes (since cytotoxic and extracellular edema are both present from the beginning of the venous thrombosis) and does not have the same prognostic value as seen in arterial stroke (areas of decreased ADC values are frequently reversible) (Fig. 10).

BBB disruption is also responsible for contrast enhancement seen in the different types of brain lesions associated with venous thrombosis and may also be depicted in initial imaging studies.

Perfusion MRI may show different patterns of hemodynamic abnormalities, the most common being the increase in circulation times (mean transit time, MTT; time to peak, TTP) and the decreased or normal cerebral blood flow. Cerebral blood volume may be decreased, normal, or increased. These perfusion abnormalities do not have the same prognostic value as in arterial stroke, since most of them are reversible (Fig. 11).
Fig. 11

CT perfusion in a patient with right frontal cortical venous thrombosis. (a) Axial non-contrast CT shows the hyperdense thrombus in a cortical frontal vein (arrows); (b) CT perfusion parametric maps illustrate the increase of MTT (left image), reduction of CBF (middle image), and absence of significant change on CBV (right image) over the cortical vein territory. (c) Axial T2-FLAIR demonstrates the absence of brain lesions on follow-up images

Venous infarcts are the result of increased venous and capillary pressure, reducing the cerebral perfusion pressure and CBF. They have a venous territory distribution; do not respect the topography of arterial territories, being mostly subcortical and frequently hemorrhagic (in up to 2/3 of cases); have often a combination of extracellular and cytotoxic edema; may exhibit contrast enhancement due to the BBB breakdown; and may be reversible if the blood finds a venous collateral drainage (Figs. 10 and 12).
Fig. 12

MRI of extensive superior sagittal sinus thrombosis and venous “infarct.” (a) Sagittal T1WI, (b) coronal T2WI and T1WI, and (c) axial T2WI and DWI demonstrate a mostly subcortical lesion location with hemorrhagic components and a combination of vasogenic and cytotoxic edema

Brain hemorrhage associated with venous thrombosis is caused by the elevation of venous pressure inside the cortical or deep venous system veins causing the dilatation and rupture of thin-walled venules and consequently brain or subarachnoid hemorrhage. BBB disruption (with diapedesis) and reperfusion of areas with venous infarct are other mechanisms causing hemorrhage. CVST-associated parenchymal hemorrhages may have different appearances but in the presence of certain types of hemorrhages CVST should be suspected.

The juxtacortical hemorrhages, located at the gray-white matter junction with a concave morphology (“cashew nut sign”) caused by the shape of the adjacent sulcus, are typical. Other types of hemorrhages include bilateral cerebral hemorrhagic lesions in cases of superior sagittal sinus thrombosis, supra- and infratentorial hemorrhages in cases of transverse sinus hemorrhage, and intraventricular and thalamic hemorrhages in cases of deep venous thrombosis. (Fig. 13).
Fig. 13

Brain MRI. (a) Sagittal T1 WI and (b) axial T2 WI demonstrate the presence of straight sinus thrombosis and bilateral thalamic hematomas

Cortical subarachnoid hemorrhage is present in up to 10% of CVST, being the only finding in up to 6% of cases, and is usually caused by cortical venous thrombosis (Fig. 14).
Fig. 14

Brain MRI in a patient with cortical vein thrombosis and focal hemorrhage. (a) Coronal T2*WI shows small hematoma with focal subarachnoid hemorrhage; (b) 3D CE MRV demonstrate permeability of the sagittal; and (c) axial contrast-enhanced T1WI depicts right frontal isolated cortical vein thrombosis (arrows)

Imaging Pitfalls

There are some pitfalls to consider when evaluating a patient with CVST, namely, anatomical variants, different signal intensities of thrombus on MRI, and limitations of some venograms.

The presence of normal intrasinus bands or septae and arachnoid granulations may produce filling defects, mimicking thrombosis.

Individual asymmetry of the dural sinuses sizes is common. It is useful in these cases to assess the size of the jugular foramen, which should be asymmetric.

The venous phase and timing of venous filling may vary among individuals especially in transverse-sigmoid sinuses, which also may mimic a thrombosis.

Additionally, one should retain that a fresh thrombus may be hypointense on T2WI mimicking flow void and that an old organized thrombus may enhance, mimicking normal venous flow.

On CE-CT and CTV there is no defined empty delta sign if the thrombus is hyperdense. The comparison of these studies with CT images obtained before contrast administration is essential. On the other hand, CE-CT and/or CTV false-positive cases may occur in patients with non-pathological slow flow in which the acquisition is obtained before the contrast passage.

MRV studies have several pitfalls. Partial thrombosis (or partially recanalized sinuses) is only depicted with the MRV source images and may not be detected with low spatial resolution (with large slice thickness) MRV. The high signal intensity of the thrombus on T1WI is not suppressed on TOF MRV and may mimic the presence of normal venous flow.

Recommended Protocol

The combination of different noninvasive imaging modalities is generally sufficient to make the diagnosis of CVST, and the choice of the modality depends on the local availability and operator experience. The imaging protocol should include vascular (venous) and parenchymal imaging. Both the combination of CT/CTV and MRI/MRV have high accuracy for the evaluation of CVST. MRI has the advantage of a more accurate parenchymal evaluation, but as described before, non-contrast MRV has more interpretation pitfalls than CTV (Saposnik et al. 2011; Ferro et al. 2017).

Venous thrombosis recommended imaging protocol:
  • MRI (T2*WI are obligatory, and T1WI on two orthogonal planes, such as axial and coronal, are highly recommended to evaluate the vessel flow/thrombosis) and MRV (preferable 3D CE MRV or 4D/dynamic CE MRA/V)

  • If MR is not immediately available or in case of noncooperative patients: CT/CTV (obtained with manual triggering or 45–60 seconds acquisition delay; it is highly recommended to have non-contrast CT images for comparison)

Interpretation Checklist and Structured Reporting

  • Acute phase:

    Diagnose the CVST:
    • Topography and extension of the CVST

    • Deep venous system involvement

    Evaluate parenchymal and other intracranial lesions:
    • Brain swelling

    • Vasogenic edema

    • Venous “infarction”

    • Brain hemorrhage

    • Subarachnoid hemorrhage

    • Others

    Define the cause of CVST (exclude underlying pathology specially infection).

    Establish the prognosis:
    • Mass effect/brain herniation

    • Hydrocephalus

    • Deep venous thrombosis

    • Cortical vein thrombosis

    • Intracranial hemorrhage

    • Posterior fossa lesions

  • Follow-up (subacute/chronic phase)
    • Evaluate the recanalization (total/partial).

    • Depict chronic complications (dural arteriovenous fistula and/or CVST recurrence).

Treatment Monitoring: Follow-Up Scheme and Findings/Pitfalls

In most cases, CVST have a favorable outcome with a complete recovery in up to 78% of adults and 54% of children. There is still a significant mortality rate, which has been decreasing over the years, mainly due to the increase in the diagnosis of less severe cases (as a consequence of higher clinical awareness and better imaging) and an etiology shift with lesser number of infectious cases. The mortality rate is estimated in the range of 5–10%.

There are clinical and imaging predictors of worse outcome, namely, the presence of underlying malignancy or infection and the imaging prognosis factors such as the presence of posterior fossa lesions, deep venous thrombosis, and intracranial hemorrhage (Saposnik et al. 2011; Ferro et al. 2017).

There is no optimized specific treatment for CVST. The initial approach is to stabilize the patient, to treat the underlying cause (especially in the presence of infections and dehydration), and to prevent the major cause of death: brain herniation. For the latter, surgical decompression (craniectomy) and/or hematoma evacuation may be warranted (Saposnik et al. 2011; Ferro et al. 2017).

Intravenous systemic heparin (or subcutaneous low molecular weighted heparin) is the first-line treatment for CVST, aiming the prevention of thrombosis progression. The presence of hemorrhage does not preclude this treatment. This indication is based on the limited evidence available, resulting from three small trials performed years ago, showing that anticoagulant treatment was safe and associated with a potentially reduction in the risk of death or dependency (Saposnik et al. 2011; Ferro et al. 2017).

Endovascular treatment has been reserved for those CVST patients refractory to heparin treatment. There have been several techniques reported for CVST endovascular treatment, including selective chemical thrombolysis and thrombectomy with clot disruption, aspiration, and/or retrieval (Saposnik et al. 2011; Ferro et al. 2017).

A recent trial (Thrombolysis or Anticoagulation for Cerebral Venous Thrombosis Trial [TO-ACT]), including a subgroup of patients with severe forms of CVT and a high chance of incomplete recovery, defined by the presence of one or more risk factors, such as intracerebral hemorrhage, mental status disorder, coma (Glasgow Coma Scale <9), and/or thrombosis of the deep cerebral venous system, did not provide significant clinical benefit compared with the conventional treatment.

Imaging is important during follow-up evaluation of CVST patients, with the aim of evaluating veins/sinus recanalization, recurrence, and development of dural arteriovenous fistula (dAVF9). The recanalization rate of the thrombosed sinuses is high, reaching 72–82% at 4 months, almost half the cases recanalized during the first month, and until 12 months some recanalization may occur.

Recurrent CVST is estimated to occur in 12–13% of patients and is generally associated with an underlying disease predisposing to thrombosis, such as hematological prothrombotic diseases (DeVeber et al. 2001). CVST predisposes to the development of dAVF, and the opposite is also true, since dAVF natural history may be associated with spontaneous venous thrombosis that may worsen the venous drainage with increasing risk and severity of brain hemorrhage.

CVST Clinical Case

Clinical History Summary

36-year-old female admitted at emergency due to increasing headache during the last 5 days without any type relieve with analgesic therapy (acetaminophen).

The headache characteristics were the following: right-side pulsatile with photophobia, without any relief factors, and different from all the previous headaches that she had experienced before. No focal neurological and focal meningeal signs were depicted at the neurological examination. No abnormalities were found at the ocular examination. The physical examination was unremarkable.

The patient has past history of migraine without aura. The only medication that the patient is taken is oral contraception.

The initial blood laboratory exams were normal.

Admission Head NCE CT Report (Fig. 15_case report)

Clinical information: Subacute headache.
Fig. 15

Admission Head NCE CT Report

Comparison: None.

Technique: Axial 3.6 mm planes (soft tissue and bone algorithms).

Report:

No acute hemorrhage is present. No brain parenchyma lesions are depicted.

The extra-axial spaces, namely, the ventricular system, sulci, and cisternal spaces have regular size and morphology. There is no hydrocephalus and/or middle line shift.

Spontaneous transverse and sigmoid sinuses hyperattenuation.

The middle ears, mastoids, and paranasal sinuses are clean.

Impression:

Suspicion of dural sinus thrombosis (transverse and sigmoid sinuses). MRI/MRV is recommended.

Admission Head MRI/MRV Report (Fig. 16_case report)

Clinical information: Subacute headache. CVST suspicion.
Fig. 16

Admission Head MRI/MRV Report

Comparison: head NCE CT.

Technique: Sagittal T1WI; axial T2WI; axial T2-FLAIR, axial T2*; coronal T2 WI; axial DWI; 4D contrast-enhanced MRV.

Report:

Signs of dural sinus thrombosis, with a right transverse and sigmoid sinus subacute phase (methemoglobin) thrombus. The thrombus exhibits T2* blooming and restricted diffusibility.

The MRV confirms the occlusion (absence of filling) of the transverse and sigmoid sinuses. No deep venous or cortical vein thrombosis is depicted.

No acute hemorrhage is present. No brain parenchyma lesions are depicted.

The extra-axial spaces, namely, the ventricular system, sulci, and cisternal spaces have regular size and morphology.

There is no hydrocephalus and/or middle line shift.

The middle ears, mastoids, and paranasal sinuses are clean.

Impression:

Transverse and sigmoid sinuses subacute dural sinus thrombosis. No brain or other intracranial lesions are depicted.

Follow-Up Imaging Studies Head MRI/MRV and DSA

Technique: 4D contrast-enhanced MRV (Fig. 17).
Fig. 17

Technique: 4D contrast-enhanced MRV

The dynamic (4D) MRV shows the early filling of the right transverse sinus at the arterial phase raising the suspicion of the presence of a dAVF.

Technique: Follow-up DSA (Fig. 18).
Fig. 18

Technique: Follow-up DSA

The right external carotid injection confirms the presence of Borden type I dAVF of the right transverse sinus, fed by dural branches from the middle meningeal and occipital arteries. No cortical venous reflux is identified.

References

  1. Altinkaya N, et al. Diagnostic value of T2*-weighted gradient-echo MRI for segmental evaluation in cerebral venous sinus thrombosis. Clin Imaging. 2015;39(1):15–9.CrossRefGoogle Scholar
  2. Buyck PJ, et al. CT density measurement and H:H ratio are useful in diagnosing acute cerebral venous sinus thrombosis. AJNR Am J Neuroradiol. 2013;34(8):1568–72.CrossRefGoogle Scholar
  3. Coutinho JM, Gerritsma JJ, et al. Isolated cortical vein thrombosis: systematic review of case reports and case series. Stroke. 2014a;45(6):1836–8.CrossRefGoogle Scholar
  4. Coutinho JM, Zuurbier SM, Stam J. Declining mortality in cerebral venous thrombosis: a systematic review. Stroke. 2014b;45(5):1338–41.CrossRefGoogle Scholar
  5. DeVeber G, et al. Cerebral sinovenous thrombosis in children. N Engl J Med. 2001;345(6):417.CrossRefGoogle Scholar
  6. Favrole P. Diffusion-weighted imaging of intravascular clots in cerebral venous thrombosis. Stroke. 2003;35(1):99–103.CrossRefGoogle Scholar
  7. Ferro JM, et al. European stroke organization guideline for the diagnosis and treatment of cerebral venous thrombosis – endorsed by the European Academy of Neurology. Eur J Neurol. 2017;24(10):1203–13.CrossRefGoogle Scholar
  8. Rodallec MH, et al. Cerebral venous thrombosis and multidetector CT angiography: tips and tricks. Radiographics. 2006;26(Suppl 1):S5–18.. discussion S42–3.CrossRefGoogle Scholar
  9. Saposnik G, et al. Diagnosis and management of cerebral venous thrombosis: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2011;42(4):1158–92.CrossRefGoogle Scholar
  10. Stam J. Thrombosis of the cerebral veins and sinuses. N Engl J Med. 2005;352(17):1791.CrossRefGoogle Scholar

Further Reading

  1. Boukobza M, et al. Radiological findings in cerebral venous thrombosis presenting as subarachnoid hemorrhage: a series of 22 cases. Neuroradiology. 2015;58(1):11–6.CrossRefGoogle Scholar
  2. Coutinho JM. Cerebral venous thrombosis. J Thromb Haemost. 2015;13(Suppl. 1):S238–44.CrossRefGoogle Scholar
  3. Dentali F, et al. Long-term outcomes of patients with cerebral vein thrombosis: a multicenter study. J Thromb Haemost. 2012;10(7):1297–302.CrossRefGoogle Scholar
  4. Ferro JM. Prognosis of cerebral vein and dural sinus thrombosis: results of the international study on cerebral vein and dural sinus thrombosis (ISCVT). Stroke. 2004;35(3):664–70.CrossRefGoogle Scholar
  5. Leach JL, Fortuna RB, Jones BV, Gaskill-Shipley MF. Imaging of cerebral venous thrombosis: current techniques, spectrum of findings, and diagnostic pitfalls. Radiographics. 2006;26(Suppl 1):S19–41. discussion S42–3.CrossRefGoogle Scholar
  6. Patel D, et al. Diagnostic performance of routine brain MRI sequences for dural venous sinus thrombosis. AJNR Am J Neuroradiol. 2016;37(11):2026–32.CrossRefGoogle Scholar
  7. Poon CS, et al. Radiologic diagnosis of cerebral venous thrombosis: pictorial review. AJR Am J Roentgenol. 2007;189(6_supplement):S64–75.CrossRefGoogle Scholar
  8. Sadigh G, Mullins ME, Saindane AM. Diagnostic performance of MRI sequences for evaluation of dural venous sinus thrombosis. AJR Am J Roentgenol. 2016;206(6):1298–306.CrossRefGoogle Scholar
  9. Siddiqui FM, et al. Mechanical thrombectomy in cerebral venous thrombosis: systematic review of 185 cases. Stroke. 2015;46(5):1263–8.CrossRefGoogle Scholar
  10. Yuh WTC, et al. Venous sinus occlusive disease: MR findings. Am J Neuroradiol AJNR. 1994;15:309–16.PubMedGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Head of NeuroradiologyHospital Beatriz ÂngeloLoures – LisbonPortugal
  2. 2.Diagnostic and Interventional Neuroradiology Graduate ConsultantHospital da LuzLisbonPortugal
  3. 3.Hospital Garcia de OrtaAlmadaPortugal

Section editors and affiliations

  • Rüdiger von Kummer
    • 1
  1. 1.Institut und Poliklinik für NeuroradiologieUniversitätsklinikum Carl Gustav Carus, DresdenDresdenGermany

Personalised recommendations