Cerebral Venous Thrombosis And Multidetector Ct Angiography- Tips And Tricks.pdf

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

RadioGraphics

CLINICAL APPLICATIONS OF VASCULAR IMAGING

S5

Cerebral Venous Thrombosis and Multidetector CT Angiography: Tips and Tricks1 CME FEATURE See accompanying test at http:// www.rsna.org /education /rg_cme.html

LEARNING OBJECTIVES FOR TEST 1 After reading this article and taking the test, the reader will be able to: 䡲 Describe the anatomy of the cerebral venous system at CT venography. 䡲 Identify the imaging features of cerebral venous thrombosis with CT venography. 䡲 Discuss the use of multidetector CT angiography, MR imaging, and conventional angiography in diagnosis of cerebral venous thrombosis.

Mathieu H. Rodallec, MD ● Alexandre Krainik, MD, PhD ● Antoine Feydy, MD, PhD ● Annick He´lias, MD ● Jean-Michel Colombani, MD Marie-Christine Julle`s, MD ● Ve´ronique Marteau, MD ● Marc Zins, MD Because of the great diversity of clinical features, its unforeseeable evolution, and a small proportion of cases that will worsen in the acute phase, cerebral venous thrombosis must be diagnosed as early as possible so that specific treatment can be started, typically transcatheter thrombolysis or systemic anticoagulation. Unenhanced computed tomography (CT) is usually the first imaging study performed on an emergency basis. Unenhanced CT allows detection of ischemic changes related to venous insufficiency and sometimes demonstrates a hyperattenuating thrombosed dural sinus or vein. Helical multidetector CT venography with bolus power injection of contrast material and combined use of two-dimensional and three-dimensional reformations (maximum intensity projection, integral display, and volume rendering) provides exquisite anatomic detail of the deep and superficial intracranial venous system and can demonstrate filling defects. However, common variants of the sinovenous system should not be mistaken for sinus thrombosis. A comprehensive diagnostic approach facilitates imaging of cerebral venous thrombosis with multidetector CT. ©

RSNA, 2006

TEACHING POINTS See last page

Abbreviations: DSA ⫽ digital subtraction angiography, MIP ⫽ maximum intensity projection, MPR ⫽ multiplanar reformatted, 3D ⫽ three-dimensional, 2D ⫽ two-dimensional RadioGraphics 2006; 26:S5–S18 ● Published online 10.1148/rg.26si065505 ● Content Codes: 1From

the Department of Radiology, Fondation Hoˆpital Saint-Joseph, 185 rue Raymond Losserand, 75674 Paris cedex 14, France (M.H.R., M.C.J., V.M., M.Z.); the Department of Neuroradiology/MRI Unit, Centre Hospitalier Universitaire de Grenoble, Grenoble, France (A.K.); the Department of Radiology B, Hoˆpital Cochin, Paris, France (A.F.); and the Department of Radiology, Hoˆpital Beaujon, Clichy, France (A.H., J.M.C.). Presented as an education exhibit at the 2005 RSNA Annual Meeting. Received February 2, 2006; revision requested March 8 and received April 6; accepted April 21. All authors have no financial relationships to disclose. Address correspondence to M.H.R. (e-mail: [email protected]).

See also the article by Leach et al (pp S19 –S41) and the commentary by Phillips (pp S42–S43) in this issue. ©

RSNA, 2006

RadioGraphics

S6

October 2006

Introduction Cerebral venous thrombosis is a disorder that is challenging to diagnose. The diagnostic difficulty results in large part from the wide variety of clinical manifestations of this uncommon disorder (1). The mode of onset is highly variable, anything from sudden to progressive, so that cerebral venous thrombosis can mimic a host of conditions (2). Given these various clinical presentations, unenhanced computed tomography (CT) is usually the first investigation performed. Unenhanced CT may show nonspecific changes and may show the spontaneously hyperattenuating thrombus. CT venography has proved to be a reliable method to investigate the structure of the cerebral veins, with a reported sensitivity of 95% with multiplanar reformatted (MPR) images when compared with digital subtraction angiography (DSA) as the standard of reference (3). Owing to its vascular detail and ease of interpretation, CT venography can provide a rapid and reliable diagnosis of cerebral venous thrombosis (4 – 6). At our institutions, unenhanced CT is performed as the initial imaging study in most cases of acute neurologic disorders. In patients with suspicion of cerebral venous thrombosis, it can be immediately followed by CT venography, thus decreasing the time to diagnosis. In this article, we start with a brief presentation of the pathophysiology, clinical diagnosis, and treatment of cerebral venous thrombosis. Technical aspects of CT venography with a focus on data acquisition and postprocessing are discussed. Normal anatomy and typical variants that should not be confused with cerebral venous thrombosis are described. We present imaging findings of cerebral venous thrombosis with CT venography and also compare the use of multisection CT to other imaging modalities in the diagnosis of cerebral venous thrombosis.

Background Magnetic resonance (MR) imaging studies have shown that cerebral venous thrombosis can induce varying degrees of edema without infarction and can even have no detectable effect on the brain (7–9). In sinovenous thrombosis, the mechanism for venous infarction is obstruction of venous drainage with increasing venous pressure in the affected region of brain (10). Cerebral venous thrombosis progresses to cerebral venous

RG f Volume 26

●

Special Issue

Table 1 Causes of and Predisposing Factors for Cerebral Venous Thrombosis Local conditions Brain and skull damage Intracranial and local regional infections Systemic conditions Hormonal (pregnancy or puerperium, estroprogestative and steroid therapy) Surgery, immobilization Hematologic and hypercoagulable disorders Connective tissue disease Malignancy Systemic infection Dehydration Idiopathic causes (12.5%)

Table 2 Clinical Features of Cerebral Venous Thrombosis Symptoms Headaches Obscurations of vision Altered consciousness Nausea, vomiting Signs Papilledema Focal neurologic deficit Cranial nerve palsies Seizures, coma

infarction in approximately 50% of cases (11). Cerebral venous infarction is initiated by thrombus propagation into draining cortical veins, causing significant extravasation of fluid into brain (vasogenic edema) and producing focal cerebral edema and hemorrhage (12). The lesion volume is probably influenced by the development of collateral veins (13). Thrombotic occlusion of the superior sagittal sinus or the dominant transverse sinus interferes with the absorption of cerebrospinal fluid through impaired function of the arachnoid granulations that line these sinuses (10). The latter mechanism further increases the extent of cerebral swelling. Unlike the onset of arterial strokes, the onset of cerebral venous thrombosis is very variable: subacute in 50% of cases (between 2 and 30 days), acute in 30% of cases (⬍2 days), and chronic in 20% of cases (14). Causes and clinical features are summarized in Tables 1 and 2.

●

Special Issue

Rodallec et al

S7

RadioGraphics

RG f Volume 26

Figure 1. Influence of the prescanning delay on sinus enhancement. (a) Axial contrastenhanced CT image obtained with a prescanning delay of 25 seconds shows inadequate enhancement of the sigmoid sinuses and jugular foramina. (b) Axial contrast-enhanced CT image obtained with a prescanning delay of 45 seconds clearly shows a thrombus (arrow) in the right sigmoid sinus.

Any underlying cause should be identified and treated. Anticoagulant therapy is aimed at preventing extension of the initial thrombus and allowing the fibrinolytic system to achieve dissolution of the existing thrombus. Unlike in an arterial ischemic stroke, relief of venous obstruction, even if delayed, may relieve the circulatory congestion with clinical benefit (10). Systemic anticoagulation is the first-line treatment for cerebral venous thrombosis because of its efficacy, safety, and feasibility (2). Several variables that increase the risk of a bad outcome include hemorrhage on an admission CT scan, thrombosis of the deep cerebral venous system, and central nervous system infection (1). Patients with these characteristics deserve additional close monitoring, and some may be candidates for more aggressive intervention such as local thrombolysis and reduction of intracranial pressure (1,15,16). The adjunction of local thrombolysis is indicated in the rare cases of worsening despite adequate anticoagulation and optimal symptomatic and etiologic treatments (2). The role of endovascular reopening of the dural sinuses with angioplasty and stent placement in the presence of severe intracranial venous hypertension is anecdotal, and long-term results are not known (17).

CT Venography CT venography can be defined as a fast thinsection volumetric helical CT examination performed with a time-optimized bolus of contrast

medium in order to enhance the cerebral venous system (5). To visualize the intracranial veins and sinuses, the examination includes the region from the calvarial vertex down to the first vertebral body. We include the atlas (C1) in the study to ensure incorporation of the origin of the jugular internal veins.

Data Acquisition We perform CT venography with a commercially available Light Speed Plus scanner (GE Healthcare, Milwaukee, Wis) with a 16-row detector ring, a minimum section thickness of 0.625 mm, and a pitch of 0.9. We administer 100 mL of nonionic contrast medium (iodine, 300 mg/mL) at a rate of 3 mL/ sec with a 45-second prescanning delay. A helical scan is performed by scanning caudally from the calvarial vertex to C1. A shorter prescanning delay than 30 seconds increases the risk of a nondiagnostic scan due to insufficient enhancement of the venous structures and flow-related artifacts (Fig 1).

Postprocessing Because all methods of three-dimensional (3D) imaging are subject to some loss of information, none of them can substitute for the thorough

RG f Volume 26

October 2006

●

Special Issue

RadioGraphics

S8

Figure 2. Normal sinovenous anatomy. (a, b) Axial MIP CT image (a) and 3D volume-rendered image from CT venography (oblique anterosuperior view) (b) show the internal cerebral veins (ICV), basal veins of Rosenthal (BVR), vein of Galen (VOG), and straight sinus (StrS). On the volume-rendered image, note the asymmetric appearance of the torcular herophili (TH), which is formed by the union of the superior sagittal sinus (SSS), straight sinus, and transverse sinuses (TS). The volume-rendered image also shows the sigmoid sinus (SS) and superficial middle cerebral vein (SMCV). (c) Sagittal MIP CT image shows the inferior sagittal sinus (ISS), as well as the internal cerebral vein, superior sagittal sinus, straight sinus, and vein of Galen.

Teaching Point

analysis of source images (18). Evaluation of venous structures includes multiplanar reformatting on a 3D workstation, in which the loss of information is minimized. With an isotropic volume acquisition, sagittal, coronal, and oblique planes can be reformatted with the same quality as the source image (19). Our practical analysis of CT venography is standardized and involves the following steps: First, two-dimensional (2D) MPR images are used to visualize dural venous sinuses and cerebral veins, with adequate window level and width. Windows setting are wider than those typically used for brain parenchyma (20,21). The source images are displayed with a window higher than or equal to 260 HU and a level of approximately 130 HU to clearly visualize the cerebral veins and dural sinuses as separate from the adjacent bone of the calvaria. Second, 2D maximum intensity projection (MIP) series are created and saved.

These first two steps require approximately 5 minutes and are sufficient in cases of superficial sinus and deep cerebral venous thrombosis. Optional reformations include 3D MIP and volume rendering display algorithms, which typically require less than 5 minutes. Further postprocessing with a 3D integral display algorithm is performed in cases of cortical venous thrombosis and requires 10 –15 minutes.

Model Preparation for 3D Display Algorithms Postprocessing is performed with a 3D workstation (Advantage Windows; GE Healthcare). For reformation, a graded subtraction is used to remove bone without deleting venous structures (5). This method is best performed by a radiologist or a highly skilled technologist with an excellent knowledge of the cerebrovascular anatomy, to avoid accidental subtraction of dural sinuses and cortical veins during the procedure. The high

●

Special Issue

RadioGraphics

RG f Volume 26

Figure 3. Normal sinovenous anatomy. Threedimensional integral image from CT venography (lateral view) shows the anastomotic vein of Trolard (AVOT) draining into the superior sagittal sinus (SSS), the anastomotic vein of Labbe´ (AVOL) draining into the transverse sinus (TS), and the superficial middle cerebral vein (SMCV).

attenuation of cortical bone of the skull is isolated first (the mask) without including any veins. This is achieved by setting a threshold above the upper limit of attenuation for enhancing veins (4,5,22). The bone model (mask) is then subtracted from the main data set to create a first-phase vascular model. This model contains residual bone, which is similar in attenuation to the vessels. The residual bone can be added to the mask through a process called dilation (4,5,22). Dilation adds a specified number of pixels around each preexisting pixel in a model in an attenuation-independent manner (4). Graded subtractions of the dilated bone mask from the vascular model allow complete removal of the skull, typically after 2-pixel dilation of the mask, without sacrificing vascular and soft-tissue detail for the integral display algorithm. The skin and residual calvaria are no longer in contact with the brain and can then be easily discarded for the integral display algorithm (3). Complete subtraction of the thin bone layers at the skull base is not always possible without a cutoff of sinus structures because of an overlap in attenuation values (3).

Three-Dimensional Display Algorithms The MIP algorithm projects intensity on the viewing screen that is the brightest intensity in the 3D model volume along a ray perpendicular to

Rodallec et al

S9

the viewing screen (6,18). Thus, from a given direction, it is possible to get some information about the brightness of an object, but the depth information is lost (18). This display technique enables one to visualize the high-attenuation vessels through the low-attenuation brain (6). A small intraluminal thrombus can be obscured if surrounded by higher-attenuation contrast-enhanced blood on a CT scan (22). The integral algorithm provides the average intensity value of the first 5 voxels deep to the model surface that is nearest to the viewer. The integral image displays low-attenuation surface features, such as the sulci, and offers improved visualization of the superficial venous system (22). It provides the best 3D CT venographic demonstration of filling defects in superficial dural sinuses and cortical veins. Volume rendering requires the user to define thresholds for the selection of voxels on the basis of their attenuation (23). The definition of the thresholds is performed interactively by the user and significantly influences the appearance of the vascular structures. Meaningful 3D representations of even overlying structures like intracranial vessels within the skull are possible, depending on the selected opacity (high opacity produces low transparency and vice versa). Both brightness and depth information are presented (18).

Anatomy of the Normal Intracranial Venous System at CT Venography The intracranial venous system is often asymmetric and considerably more variable than the arterial anatomy. Knowledge of the most frequent variations of the sinovenous system is useful for accurate diagnosis of cerebral venous thrombosis with CT venography. The deep venous system is concerned with centripetal venous drainage of deep cerebral white matter and basal ganglia, with two levels operating in a hemodynamic balance: (a) the internal cerebral veins, basal veins of Rosenthal, and vein of Galen (Fig 2); and (b) the transcerebral venous system, which typically is not visualized because of the small caliber of the veins that drain the cerebral hemispheric white matter (24). The superficial cerebral veins course over the surface of the brain, draining the cortex and a portion of the subjacent white matter (Fig 3) (25).

RadioGraphics

S10

Teaching Point

October 2006

The infratentorial veins have three major drainage systems: (a) superiorly into the vein of Galen, (b) anteriorly into the petrosal sinuses, and (c) posteriorly into dural sinuses (25). The dural sinuses collect blood from cerebral veins into the internal jugular veins. There are two groups of dural venous sinuses. The superior group drains the majority of the brain and skull and includes the superior and inferior sagittal sinus and straight, occipital, transverse, and sigmoid sinuses (26). The superior sagittal sinus collects the superficial cerebral veins draining the cerebral convexities. Its luminal surface is triangular in shape and traversed by septa, which play an important role in maintaining laminar flow and in thwarting venous reflux into cortical veins (27). The confluence of the sinuses (torcular herophili) is formed by the union of the superior sagittal sinus, straight sinus, and transverse sinuses and is often asymmetric in appearance (26). The transverse sinuses are commonly asymmetric, with the right transverse sinus being dominant in the majority of cases (Fig 4) (28). A unilateral atretic posteromedial segment of the left transverse sinus is also common (26). The small variable occipital sinus lies in the midline at the attachment of the falx cerebelli and extends from the foramen magnum to drain upward into the confluence of the sinuses (25). Alternatively, with variant anatomy, the occipital sinus may drain toward the foramen magnum or into the jugular fossa or suboccipital veins and may replace an aplastic transverse sinus (25,26). The inferior group of dural venous sinuses drains the superficial cerebral veins, the basal and medial parts of the undersurface of the brain, the orbits, and the sphenoparietal sinus and collects at the cavernous sinus (26).

RG f Volume 26

●

Special Issue

Figure 4. Normal sinovenous anatomy. Axial MIP CT image shows asymmetric transverse sinuses (TS). The sigmoid sinuses (SS) begin where the transverse sinuses leave the tentorial margin. The right cavernous sinus (CS) is also demonstrated.

Normal Appearance of Arachnoid Granulations of Pacchioni

Teaching Point

Arachnoid granulations of Pacchioni play a major role in the resorption of cerebrospinal fluid. They are most commonly found within the lacunae laterales of the superior sagittal sinus, producing calvarial impressions between 13 and 15 mm lateral to midline in the region (Fig 5) (29). Arachnoid granulations can also protrude directly into the sinus lumen, adjacent to venous entrance sites, and should not be mistaken for sinus thrombosis. Arachnoid granulations are present in the superior sagittal sinus, transverse sinus, cavernous sinus, superior petrosal sinus, and straight sinus in decreasing order of frequency (Fig 6).

Figure 5. Arachnoid granulation of Pacchioni in the parasagittal region. Axial contrast-enhanced CT image shows the calvarial impression produced by a parasagittal arachnoid granulation.

●

Special Issue

Rodallec et al

S11

RadioGraphics

RG f Volume 26

Figure 6. Arachnoid granulations of Pacchioni in the venous sinuses. (a) Sagittal 2D MIP image from CT venography show arachnoid granulations (arrows) in the superior sagittal sinus and straight sinus. (b) Axial contrast-enhanced CT image shows a well-limited lobulated defect (arrow) in the right transverse sinus.

They increase in number and conspicuity with age, their prevalence being approximately 66% (21). Arachnoid granulations produce well-defined focal filling defects within the dural venous sinuses and measure 2–9 mm in diameter (21,30). They are isoattenuating (one-third) or hypoattenuating (two-thirds) relative to brain parenchyma (21).

CT Venographic Findings in Cerebral Venous Thrombosis Imaging findings of cerebral venous thrombosis can be categorized as direct, as when there is visualization of cortical or dural sinus thrombus, or indirect, as when there are ischemic or vascular changes related to the venous outflow disturbance (17). On the basis of the results of the largest cohort ever published (624 patients), collected over a short period of time, cerebral venous thrombosis involves the following vessels in decreasing frequency: the superior sagittal sinus (62%), left and right transverse sinus (respectively 44.7% and 41.2%), straight sinus (18%), cortical veins (17.1%), deep venous system (10.9%), cavernous sinus (1.3%), and cerebellar veins (0.3%) (1).

Most data about CT of cerebral venous thrombosis concern older conventional CT scanners. Technical limitations are related mainly to suboptimal enhancement and to the analysis of axial images with a narrow window level (31–34).

Indirect Signs Secondary to Veno-occlusive Disease Indirect signs are not specific, but they should draw attention and prompt a search for direct visualization of a cortical or sinus thrombus. Early changes are often subtle, with brain edema and swelling of the gyri. Venous infarction manifests as a low-attenuation lesion (17) with or without subcortical hemorrhage (35). Brain lesions are related to a venous distribution. Bilateral parasagittal hemispheric lesions are suggestive of superior sagittal sinus thrombosis (Fig 7). Ipsilateral temporo-occipital and cerebellar lobe lesions can be found in transverse sinus thrombosis (Fig 8). In bilateral thalamic lesions, deep cerebral venous thrombosis should be suspected (Fig 9). Additional cerebral structures adjacent to the thalami such as the basal ganglia and

Teaching Point

October 2006

RG f Volume 26

●

Special Issue

RadioGraphics

S12

Figure 7. Thrombosis of the superior sagittal sinus in a 20-year-old woman. (a, b) Axial unenhanced CT images obtained at different levels show subarachnoid hemorrhage with a “dense triangle” sign (arrowhead) and a cord sign (arrow), findings suggestive of sinus and cortical vein thrombosis. (c) Three-dimensional integral image from CT venography (posterosuperior view) shows filling defects in the superior sagittal sinus and parieto-occipital veins, an appearance indicative of thrombosis.

Figure 8. Thrombosis of the left transverse sinus in a 42-year-old woman. (a, b) Axial unenhanced CT images show left cerebellar and temporal hematoma with increased attenuation in the left transverse sinus (cord sign) (* in a). (c) On a 3D MIP image from CT venography, the left transverse sinus is not visible.

the mesencephalon are affected in one-third of patients with deep cerebral venous thrombosis (36). Thrombosis of a deep cerebral vein can very rarely manifest unilaterally (36). After dural sinus thrombosis, transcortical medullary veins can become sources of collateral blood drainage. Subsequent dilatation of these vessels causes them to be visible at contrast-enhanced CT (27). Tentorial enhancement is caused by prominence of dural collaterals (31). Small subdural hematomas or effusion are seen occasionally. Cerebral venous thrombosis rarely

manifests as isolated subarachnoid hemorrhage of the convexity (Fig 10) (7,9,37,38).

Direct Visualization of Thrombus At unenhanced CT, a recently thrombosed dural sinus or cortical vein may be identified as an elongated hyperattenuating lesion (39). The so-called dense triangle or cord sign represents an intravascular acute blood clot (17). This sign is reported in 20% of patients (27) and takes approximately 1–2 weeks to disappear (Fig 11) (31). However, similarly increased attenuation of the cerebral venous sinuses may also represent polycythemia (40), and nonmyelinated brain in neonates makes sinuses appear unusually attenuating (32).

●

Special Issue

Rodallec et al

S13

RadioGraphics

RG f Volume 26

Figure 9. Deep cerebral venous thrombosis in a 41-year-old woman. (a) Axial unenhanced CT image shows an area of low attenuation in both thalami (*) associated with increased attenuation in the straight sinus (arrow) and internal cerebral veins (arrowheads). Note the hemorrhage in the left choroid plexus. (b) Axial contrast-enhanced CT image shows filling defects in the straight sinus (arrow) and internal cerebral veins (white arrowheads), an appearance consistent with thrombosis. Compare this appearance with the normally enhancing superior sagittal sinus (black arrowhead).

Figure 10. Thrombosis of the left transverse and sigmoid sinuses and temporooccipital veins in a 45-year-old woman. (a, b) Axial unenhanced CT images show parieto-occipital subarachnoid hemorrhage (arrowhead in b) with superficial temporooccipital areas of high attenuation (arrow in a), findings suggestive of thrombosed cortical veins (cord sign). (c) Axial contrast-enhanced CT image shows thrombosis of the left transverse and sigmoid sinuses (arrows). (d) Threedimensional integral image from CT venography shows filling defects in the left transverse sinus and temporo-occipital veins, an appearance consistent with thrombosis.

RadioGraphics

S14

RG f Volume 26

October 2006

●

Special Issue

Figure 11. Chronic thrombosis of the left sigmoid sinus and jugular foramen in a 78-yearold woman with a history of breast carcinoma. (a) Axial unenhanced CT image shows low attenuation in the left sigmoid sinus (arrow) and jugular foramen (arrowhead). (b) Axial contrast-enhanced CT image shows thrombi in the left sigmoid sinus (arrow) and jugular foramen (arrowhead).

Figure 12. Effect of window settings on the detection of sinus and venous thrombosis. (a) Axial contrast-enhanced CT image obtained with wide window settings (width ⫽ 260 HU, level ⫽ 130 HU) shows thrombi in the superior sagittal sinus (arrowhead) and a cortical vein (arrow). (b) The thrombi are not visible on an axial contrast-enhanced CT image obtained with parenchymal window settings (width ⫽ 90 HU, level ⫽ 50 HU).

At contrast-enhanced CT, the thrombus does not enhance but the dura enhances (17). When the thrombus is located in the superior sagittal and transverse sinuses, a triangular defect (empty delta sign) can be demonstrated on postcontrast images by using multiplanar reformations and window settings wider than those typically used for brain parenchyma (Fig 12). The empty delta sign is seen in 25%–75% of cases in the literature (27,31,34). Because of volume averaging, con-

ventional axial CT scans can fail to demonstrate a hyperattenuating sinus or an empty delta sign in the horizontal segment of the superior sagittal sinus or transverse sinus. The empty delta sign can disappear in chronic stages with enhancement of organized clot (31,34). The possibility of multiplanar reformations with CT venography is very helpful in detecting sinus and cortical venous thrombosis. CT venography with the integral display algorithm can provide excellent demonstration of filling defects in the superficial venous sinuses and cortical veins

RadioGraphics

RG f Volume 26

●

Special Issue

Rodallec et al

S15

Figure 13. Thrombosis of superficial cerebral veins in a 44-year-old man with Behc¸et disease. (a) Axial unenhanced CT image shows parietal subcortical hematoma with the cord sign (arrowhead) in a cortical vein (presumably the anastomotic vein of Trolard). (b, c) Axial (b) and coronal reformatted (c) unenhanced CT images show the cord sign in the right superficial middle cerebral vein (arrowhead) along the lateral fissure. (d–f) Axial (d, e) and coronal reformatted (f ) contrast-enhanced CT images show enlargement and vascular defects of the right superficial middle cerebral vein (arrowhead in f ) and of a cortical vein (arrowhead in d) (presumably the anastomotic vein of Trolard), with extension of the thrombus to the superior sagittal sinus. (g, h) Posterosuperior (g) and lateral (h) 3D integral images from CT venography show thrombosis of the right superficial middle cerebral vein (short arrow) and of the anastomotic vein of Trolard (arrowhead) with extension to the superior sagittal sinus (long arrow in g) and swelling of the gyri.

(Fig 13) (6,22). Volume rendering display can be helpful to demonstrate collateral pathways in cortical vein thrombosis (Fig 14). The most reliable criterion with which to establish the diagnosis of

cavernous sinus thrombosis is the presence of a large filling defect of non–fat attenuation with sinus expansion (41).

October 2006

RG f Volume 26

●

Special Issue

RadioGraphics

S16

Figure 14. Thrombosis of the superior sagittal sinus in a 31-year-old woman. (a) Coronal reformatted 2D MIP CT image shows the empty delta sign in the superior sagittal sinus with enlargement and a vascular defect of the adjacent cortical vein. Note the small collateral vein (arrow) under the enlarged thrombosed cortical vein. (b) Three-dimensional integral image from CT venography (superior view) shows thrombosis of the anterior part of the superior sagittal sinus with extension to the left frontoparietal cortical veins. (c) Three-dimensional volume-rendered image from CT venography with an inferior cut (same projection as in b) shows the collateral pathways along the superior sagittal sinus and under the enlarged thrombosed cortical vein (arrow).

CT Venography versus Other Imaging Modalities Unlike primary intracerebral hemorrhage and arterial cerebral infarction, which typically manifest as acute neurologic deficits (strokes) and whose diagnosis relies on demonstrating the parenchymal lesions, the diagnosis of cerebral venous thrombosis relies on demonstrating thrombi in the cerebral veins and/or sinuses (14). DSA has been the key procedure in the investigation of intracranial venous disease for many years (39) but is an invasive procedure with wellknown associated risks (42,43). The partial or complete lack of filling of veins or sinuses is the best angiographic sign of cerebral venous thrombosis. However, this sign is difficult to interpret in locations such as the anterior third of the superior sagittal sinus, the left transverse sinus, and the cortical veins because of common variants (28,39). With the advent of CT venography and MR imaging, DSA is now rarely required for diagnosis. However, in patients with subarachnoid hemorrhage and cerebral venous thrombosis, DSA should be considered to rule out other causes of rebleeding, such as distal aneurysm and dural arteriovenous fistula, before anticoagulant therapy is administered (7,9,37,38). CT venography has several advantages over DSA. It is less invasive and less expensive, and the time to diagnosis in the initial work-up of a patient is shorter (3). In a comparative study of CT venography with DSA in imaging cerebral

venous anatomy and pathologic conditions, MPR images from CT venography were superior to DSA images in showing the cavernous sinus, inferior sagittal sinus, and basal vein of Rosenthal (3). Reported advantages of CT venography compared with MR imaging techniques are rapid image acquisition (reduction of motion-related artifacts), no contraindication to pacemaker and ferromagnetic devices, increased imaging resolution, and fewer equivocal imaging findings. No flowrelated artifacts have been reported with use of a contrast material bolus and acquisition in the venous phase (4 – 6,22). CT venography more commonly shows sinuses or smaller cerebral veins with low flow as compared with MR venography (2D phase contrast, 3D phase contrast, and 2D time of flight) (6). Limitations of CT venography include exposure to ionizing radiation, adverse reactions to iodinated contrast medium, and limited visualization of skull base structures in 3D display. The use of MR imaging in the diagnosis of cerebral venous thrombosis demands knowledge of the different stages of thrombus evolution and pitfalls (44). The main sign of cerebral venous thrombosis with a standard MR imaging protocol is the lack of expected signal flow void on standard spin-echo T1- and T2-weighted images (45). Diagnosis of cerebral venous thrombosis at the acute stage can be challenging because the hypointense signal of acute intraluminal thrombus mimics a normal flow void on T2-weighted images (8,39,44,45). The absence of flow void on T1-weighted images at this early stage must be carefully sought because thrombus is isointense to brain tissue (31,45). Then, time-dependent in-

Teaching Point

RadioGraphics

RG f Volume 26

●

Special Issue

creasing hyperintensity of intraluminal thrombus seen on spin-echo images corresponds to the release of extracellular methemoglobin. After 30 days, the thrombus gradually becomes isointense on T1-weighted images. Pitfalls of MR imaging in the diagnosis of cerebral venous thrombosis include flow-related enhancement (entry section phenomenon) and refocusing of slow flow (even echo rephasing or use of flow-compensation gradients), which may mimic intraluminal thrombus, or paramagnetic blood products (intracellular deoxyhemoglobin or methemoglobin) mimicking a normal signal void on long repetition time images (8,31). A thrombus shows identical signal appearance in all section orientations. T2*-weighted imaging is able to demonstrate areas of hypointensity in thrombosed veins and sinuses (46), but the exact sensitivity of T2*-weighted imaging remains unknown. Moreover, the susceptibility effect of T2*-weighted imaging does not always indicate intravascular thrombosis or blood products, since arterial flow voids, calcifications, and the bone surfaces of the skull commonly result in susceptibility artifacts. Diffusion-weighted imaging of intravascular clots might be of value in prediction of the risk of persistent venous occlusion (47). MR imaging is more sensitive than CT in demonstrating the parenchymal lesions. In MR venography, thrombosis is suggested by absence of normal flow signal in a sinus or a vein. However, this finding may also be attributed to artifacts (26,48). In 2D time-of-flight imaging, flow gaps result from slow intravascular blood flow, in plane flow, and complex blood flow patterns. Most of the artifactual loss of vascular signal seen with the use of 2D MR venography occurs in nondominant transverse sinuses. Depiction of flowing blood with time-of-flight imaging is limited when short T1 substances such as methemoglobin are present (49). In 2D phasecontrast imaging, if blood flow is sufficiently slow, the motion-induced phase shifts may be inadequate to distinguish the flow from stationary tissue. Postcontrast 3D magnetization-prepared rapid gradient-echo T1-weighted images are not affected by the angle between vessels and imaging slab or flow velocity. Three-dimensional surface algorithms can also be used with postcontrast 3D MR imaging to display the brain surface and the superficial venous system (22).

Conclusions CT venography should be strongly recommended when cerebral venous thrombosis is suspected, particularly in situations in which MR imaging has inconclusive results, is not available, or is contraindicated. In patients with unenhanced CT findings suggestive of venous thrombosis, CT

Rodallec et al

S17

venography can be performed without delay to confirm the diagnosis and to start appropriate therapy immediately.

References 1. Ferro JM, Canhao P, Stam J, Bousser MG, Barinagarrementeria F; ISCVT Investigators. 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 – 670. 2. Bousser MG. Cerebral venous thrombosis: nothing, heparin, or local thrombolysis? Stroke 1999; 30(3):481– 483. 3. Wetzel SG, Kirsch E, Stock KW, Kolbe M, Kaim A, Radue EW. Cerebral veins: comparative study of CT venography with intraarterial digital subtraction angiography. AJNR Am J Neuroradiol 1999;20(2):249 –255. 4. Alberico RA, Barnes P, Robertson RL, Burrows PE. Helical CT angiography: dynamic cerebrovascular imaging in children. AJNR Am J Neuroradiol 1999;20(2):328 –334. 5. Casey SO, Alberico RA, Patel M, et al. Cerebral CT venography. Radiology 1996;198(1):163–170. 6. Ozsvath RR, Casey SO, Lustrin ES, Alberico RA, Hassankhani A, Patel M. Cerebral venography: comparison of CT and MR projection venography. AJR Am J Roentgenol 1997;169(6):1699 – 1707. 7. Oppenheim C, Domigo V, Gauvrit JY, et al. Subarachnoid hemorrhage as the initial presentation of dural sinus thrombosis. AJNR Am J Neuroradiol 2005;26(3):614 – 617. 8. Guermazi A, Miaux Y, Williams M, Turki C, Frija J. Dural sinus thrombosis: CT and MR imaging of different stages. J Belge Radiol 1997;80(4):167– 169. 9. Chang R, Friedman DP. Isolated cortical venous thrombosis presenting as subarachnoid hemorrhage: a report of three cases. AJNR Am J Neuroradiol 2004;25(10):1676 –1679. 10. Shroff M, deVeber G. Sinovenous thrombosis in children. Neuroimaging Clin N Am 2003;13(1): 115–138. 11. Tsai FY, Wang AM, Matovich VB, et al. MR staging of acute dural sinus thrombosis: correlation with venous pressure measurements and implications for treatment and prognosis. AJNR Am J Neuroradiol 1995;16(5):1021–1029. 12. Fries G, Wallenfang T, Hennen J, et al. Occlusion of the pig superior sagittal sinus, bridging and cortical veins: multistep evolution of sinus-vein thrombosis. J Neurosurg 1992;77(1):127–133. 13. Rottger C, Trittmacher S, Gerriets T, Blaes F, Kaps M, Stolz E. Reversible MR imaging abnormalities following cerebral venous thrombosis. AJNR Am J Neuroradiol 2005;26(3):607– 613. 14. Crassard I, Bousser MG. Cerebral venous thrombosis [in French]. Rev Med Interne 2006;27(2): 117–124. 15. Ferro JM, Canhao P, Bousser MG, Stam J, Barinagarrementeria F; ISCVT Investigators. Cerebral vein and dural sinus thrombosis in elderly patients. Stroke 2005;36(9):1927–1932. 16. Curtin KR, Shaibani A, Resnick SA, Russell EJ, Simuni T. Rheolytic catheter thrombectomy, balloon angioplasty, and direct recombinant tissue

RadioGraphics

S18

17.

18.

19. 20.

21.

22.

23.

24.

25. 26. 27.

28. 29. 30.

31. 32.

October 2006 plasminogen activator thrombolysis of dural sinus thrombosis with preexisting hemorrhagic infarctions. AJNR Am J Neuroradiol 2004;25(10): 1807–1811. Lee SK, terBrugge KG. Cerebral venous thrombosis in adults: the role of imaging evaluation and management. Neuroimaging Clin N Am 2003; 13(1):139 –152. Tomandl BF, Hastreiter P, Rezk-Salama C, et al. Local and remote visualization techniques for interactive direct volume rendering in neuroradiology. RadioGraphics 2001;21(6):1561–1572. Blank M, Kalender WA. Medical volume exploration: gaining insights virtually. Eur J Radiol 2000; 33(3):161–169. McMurdo SK Jr, Brant-Zawadzki M, Bradley WG Jr, Chang GY, Berg BO. Dural sinus thrombosis: study using intermediate field strength MR imaging. Radiology 1986;161(1):83– 86. Leach JL, Jones BV, Tomsick TA, Stewart CA, Balko MG. Normal appearance of arachnoid granulations on contrast-enhanced CT and MR of the brain: differentiation from dural sinus disease. AJNR Am J Neuroradiol 1996;17(8):1523–1532. Casey SO, Rubinstein D, Lillehei KO, et al. Integral and shell-MIP display algorithms in MR and CT three-dimensional models of the brain surface. AJNR Am J Neuroradiol 1998;19(8):1513–1521. Tomandl BF, Kostner NC, Schempershofe M, et al. CT angiography of intracranial aneurysms: a focus on postprocessing. RadioGraphics 2004; 24(3):637– 655. Lasjaunias P, Berenstein A, terBrugge KG, et al. Intracranial venous system. In: Lasjaunias P, Berenstein A, terBrugge KG, eds. Surgical neuroangiography. 2nd ed. Berlin, Germany: Springer Verlag, 2001; 631– 695. Scott JN, Farb RI. Imaging and anatomy of the normal intracranial venous system. Neuroimaging Clin N Am 2003;13(1):1–12. Cure JK, Van Tassel P, Smith MT. Normal and variant anatomy of the dural venous sinuses. Semin Ultrasound CT MR 1994;15(6):499 –519. Virapongse C, Cazenave C, Quisling R, Sarwar M, Hunter S. The empty delta sign: frequency and significance in 76 cases of dural sinus thrombosis. Radiology 1987;162(3):779 –785. Zouaoui A, Hidden G. Cerebral venous sinuses: anatomical variants or thrombosis? Acta Anat (Basel) 1988;133(4):318 –324. Grossman CB, Potts DG. Arachnoid granulations: radiology and anatomy. Radiology 1974;113(1): 95–100. Casey SO, Ozsvath R, Choi JS. Prevalence of arachnoid granulations as detected with CT venography of the dural sinuses. AJNR Am J Neuroradiol 1997;18(5):993–994. Cure JK, Van Tassel P. Congenital and acquired abnormalities of the dural venous sinuses. Semin Ultrasound CT MR 1994;15(6):520 –539. Davies RP, Slavotinek JP. Incidence of the empty delta sign in computed tomography in the paediat-

RG f Volume 26

●

Special Issue

ric age group. Australas Radiol 1994;38(1):17–19. 33. Provenzale JM, Joseph GJ, Barboriak DP. Dural sinus thrombosis: findings on CT and MR imaging and diagnostic pitfalls. AJR Am J Roentgenol 1998;170(3):777–783. 34. Shinohara Y, Yoshitoshi M, Yoshii F. Appearance and disappearance of empty delta sign in superior sagittal sinus thrombosis. Stroke 1986;17(6): 1282–1284. 35. Keiper MD, Ng SE, Atlas SW, Grossman RI. Subcortical hemorrhage: marker for radiographically occult cerebral vein thrombosis on CT. J Comput Assist Tomogr 1995;19(4):527–531. 36. Herrmann KA, Sporer B, Yousry TA. Thrombosis of the internal cerebral vein associated with transient unilateral thalamic edema: a case report and review of the literature. AJNR Am J Neuroradiol 2004;25(8):1351–1355. 37. Sztajzel R, Coeytaux A, Dehdashti AR, Delavelle J, Sinnreich M. Subarachnoid hemorrhage: a rare presentation of cerebral venous thrombosis. Headache 2001;41(9):889 – 892. 38. Spitzer C, Mull M, Rohde V, Kosinski CM. Nontraumatic cortical subarachnoid haemorrhage: diagnostic work-up and aetiological background. Neuroradiology 2005;47(7):525–531. 39. Ameri A, Bousser MG. Cerebral venous thrombosis. Neurol Clin 1992;10(1):87–111. 40. Healy JF, Nichols C. Polycythemia mimicking venous sinus thrombosis. AJNR Am J Neuroradiol 2002;23(8):1402–1403. 41. Schuknecht B, Simmen D, Yuksel C, Valavanis A. Tributary venosinus occlusion and septic cavernous sinus thrombosis: CT and MR findings. AJNR Am J Neuroradiol 1998;19(4):617– 626. 42. Spies JB, Berlin L. Complications of femoral artery puncture. AJR Am J Roentgenol 1998;170(1): 9 –11. 43. Heiserman JE, Dean BL, Hodak JA, et al. Neurologic complications of cerebral angiography. AJNR Am J Neuroradiol 1994;15(8):1401–1407; discussion 1408 –1411. 44. Isensee C, Reul J, Thron A. Magnetic resonance imaging of thrombosed dural sinuses. Stroke 1994;25(1):29 –34. 45. Hinman JM, Provenzale JM. Hypointense thrombus on T2-weighted MR imaging: a potential pitfall in the diagnosis of dural sinus thrombosis. Eur J Radiol 2002;41(2):147–152. 46. Selim M, Fink J, Linfante I, Kumar S, Schlaug G, Caplan LR. Diagnosis of cerebral venous thrombosis with echo-planar T2*-weighted magnetic resonance imaging. Arch Neurol 2002;59(6): 1021–1026. 47. Favrole P, Guichard JP, Crassard I, Bousser MG, Chabriat H. Diffusion-weighted imaging of intravascular clots in cerebral venous thrombosis. Stroke 2004;35(1):99 –103. 48. Tsuruda J, Saloner D, Norman D. Artifacts associated with MR neuroangiography. AJNR Am J Neuroradiol 1992;13(5):1411–1422. 49. Huston J 3rd, Ehman RL. Comparison of time-offlight and phase-contrast MR neuroangiographic techniques. RadioGraphics 1993;13(1):5–19.

This article meets the criteria for 1.0 credit hour in category 1 of the AMA Physician’s Recognition Award. To obtain credit, see accompanying test at http://www.rsna.org/education/rg_cme.html.

RadioGraphics

RG

Volume 26 • Special Issue • October 2006

Rodallec et al

Cerebral Venous Thrombosis and Multidetector CT Angiography: Tips and Tricks Mathieu H. Rodallec, MD, et al RadioGraphics 2006; 26:S5–S18 ● Published online 10.1148/rg.26si065505 ● Content Codes:

Page S8 First, two-dimensional (2D) MPR images are used to visualize dural venous sinuses and cerebral veins, with adequate window level and width. Windows setting are wider than those typically used for brain parenchyma (20,21). The source images are displayed with a window higher than or equal to 260 HU and a level of approximately 130 HU to clearly visualize the cerebral veins and dural sinuses as separate from the adjacent bone of the calvaria. Page S10 The transverse sinuses are commonly asymmetric, with the right transverse sinus being dominant in the majority of cases (Fig 4) (28). A unilateral atretic posteromedial segment of the left transverse sinus is also common (26). Page S10 Arachnoid granulations can also protrude directly into the sinus lumen, adjacent to venous entrance sites, and should not be mistaken for sinus thrombosis. Arachnoid granulations are present in the superior sagittal sinus, transverse sinus, cavernous sinus, superior petrosal sinus, and straight sinus in decreasing order of frequency (Fig 6). Page S11 Indirect signs are not specific, but they should draw attention and prompt a search for direct visualization of a cortical or sinus thrombus. Early changes are often subtle, with brain edema and swelling of the gyri. Venous infarction manifests as a low-attenuation lesion (17) with or without subcortical hemorrhage (35). Brain lesions are related to a venous distribution. Pages S16 Diagnosis of cerebral venous thrombosis at the acute stage can be challenging because the hypointense signal of acute intraluminal thrombus mimics a normal flow void on T2-weighted images (8,39,44,45). The absence of flow void on T1-weighted images at this early stage must be carefully sought because thrombus is isointense to brain tissue (31,45).