July 2013, VOLUME 201
NUMBER 1

Recommend & Share

July 2013, Volume 201, Number 1

Neuroradiology/Head and Neck Imaging

Original Research

Performance of Spin-Echo and Gradient-Echo T1-Weighted Sequences for Evaluation of Dural Venous Sinus Thrombosis and Stenosis

+ Affiliations:
1 Department of Radiology and Imaging Sciences, Emory University Hospital and Emory University School of Medicine, 1364 Clifton Rd NE, BG22, Atlanta, GA 30322.

2 Department of Biostatistics and Bioinformatics, Emory University School of Medicine, Atlanta, GA.

Citation: American Journal of Roentgenology. 2013;201: 162-169. 10.2214/AJR.12.9095

ABSTRACT
Next section

OBJECTIVE. Dural venous sinus abnormalities are clinically important but can potentially be overlooked using various MRI techniques. This study evaluates the diagnostic accuracy of spin-echo (SE) T1-weighted imaging, 3D gradient-recalled echo (GRE) T1-weighted imaging, and contrast-enhanced MR venography (MRV) for the detection of dural venous sinus thrombosis and transverse sinus (TS) stenosis.

MATERIALS AND METHODS. Seventy-three patients underwent MRI evaluation with unenhanced and contrast-enhanced axial SE T1-weighted imaging, contrast-enhanced sagittal 3D GRE T1-weighted imaging, and contrast-enhanced MRV sequences. Three neuroradiologists each evaluated these 219 total datasets in a randomized blinded fashion for the presence or absence of TS stenosis and for dural venous sinus thrombosis in each of 10 venous sinus segments (730 total segments). Diagnostic performance characteristics and kappa statistics were calculated for each technique.

RESULTS. Thirteen patients (37 segments) had suspected dural venous sinus thrombosis by one or more readers; of those 13 patients, nine (23 segments) were thought to have definite thrombosis on contrast-enhanced MRV. Compared with contrast-enhanced MRV, the positive predictive value (PPV) and negative predictive value (NPV) for thrombosis were 60% and 97%, respectively, for both unenhanced and contrast-enhanced SE T1-weighted imaging and 100% and 98% for 3D GRE T1-weighted imaging. Kappa values calculated per venous segment were as follows: 0.41 for SE T1-weighted imaging, 0.72 for 3D GRE T1-weighted imaging, and 0.95 for contrast-enhanced MRV. Thirty patients (58 segments) had TS stenosis suspected by one or more readers; of those 30 patients, TS stenosis was deemed to be definite on contrast-enhanced MRV in 25 patients (50 segments). Compared with contrast-enhanced MRV, the PPV and NPV were 75% and 80%, respectively, for SE T1-weighted imaging and 91% and 92% for 3D GRE T1-weighted imaging for the detection of stenosis. Kappa values calculated per patient were −0.038 for SE T1-weighted imaging, 0.58 for 3D GRE T1-weighted imaging, and 0.98 for contrast-enhanced MRV.

CONCLUSION. Contrast-enhanced 3D GRE T1-weighted imaging is superior to SE T1-weighted imaging for the detection of dural venous sinus thrombosis and TS stenosis but does not substitute for dedicated MRV. Hyperintensity on unenhanced SE T1-weighted imaging is unreliable for the detection of dural venous sinus thrombosis.

Keywords: idiopathic intracranial hypertension, MR venography, sensitivity and specificity, venous sinus thrombosis

Important potential causes of headache are pathologic processes involving the dural venous sinuses, the primary outflow pathway of blood from the brain. The clinical diagnosis of dural venous sinus thrombosis is often challenging because of its infrequency (annual incidence of 3–4 cases per 1 million individuals) and variable, nonspecific clinical presentation [1, 2]. Prompt diagnosis is crucial because early intervention, including anticoagulation and systemic or catheter-directed thrombolysis, is associated with favorable clinical outcomes, whereas hemorrhagic complications from dural venous sinus thrombosis can result in devastating neurologic outcomes [3, 4]. Stenosis of the dural venous sinuses—specifically, transverse sinus (TS) stenosis—has been highly associated with increased intracranial pressure (ICP) in the setting of idiopathic intracranial hypertension (IIH) [57]. Patients with IIH may present with headaches, transient visual disturbance, or pulsatile tinnitus; if IIH is left untreated, patients may develop blindness or complications of CSF leak from chronically elevated ICP. The diagnosis of IIH requires documentation of elevated opening pressure on lumbar puncture without evidence of other causes of elevated ICP including an intracranial mass lesion, hydrocephalus, and dural venous sinus thrombosis [8]. Highly effective medical and surgical treatments are available for decreasing CSF production and for diverting CSF [911], and in some patients intra-sinus stenting has been effective in reducing ICP and treating symptoms [12, 13].

MRI is frequently the first imaging examination performed in the setting of chronic headaches or may be performed in the acute setting after a CT study of the head is unrevealing. Alternatively, evaluation of the dural venous sinuses with MRI may be performed specifically for clinically suspected dural venous sinus thrombosis or IIH. It is desirable for MRI protocols to be able to evaluate the dural venous sinuses for these relatively infrequent but clinically important and treatable conditions; however, there is considerable variability in routine brain MRI protocols used for patients presenting with headaches. Various T1-weighted pulse sequences including unenhanced and contrast-enhanced spin-echo (SE)–based T1-weighted sequences and contrast-enhanced 3D gradient-recalled echo (GRE) T1-weighted imaging sequences may be used for routine brain MRI. SE T1-weighted imaging may show intrinsically hyperin-tense thrombus and contrast-enhanced filling defects, although the former depends on imaging thrombus in the subacute phase and the latter may be confounded by flow-related artifacts and enhancement of thrombus in the chronic phase [1417]. The contrast-enhanced 3D GRE T1-weighted imaging sequence has become a standard part of the routine brain MRI protocol at many centers and can provide a volumetric assessment of the venous structures without significant flow artifact [18, 19].

When clinical attention is directed at the dural venous sinuses, phase contrast, time-of-flight (TOF), or contrast-enhanced MRV may be added to a brain MRI protocol. Contrast-enhanced MRV has become the gold standard noninvasive technique for diagnosing dural venous sinus thrombosis because it has shown superior sensitivity to TOF and phase contrast techniques, improved visualization of smaller venous structures, and shorter imaging times [2028]. Other techniques such as diffusion-weighted imaging and T2*-weighted GRE and other susceptibility-weighted sequences are useful adjuncts in the diagnosis of thrombosis [2931].

Given the variability in techniques used for routine brain MRI and for specific evaluation of the dural venous sinuses, it is important to understand the reliability and limitations of these techniques when evaluating for dural venous sinus thrombosis and TS stenosis. The purpose of this study was to evaluate the diagnostic accuracy and interobserver variability of un-enhanced SE T1-weighted imaging, contrast-enhanced SE T1-weighted imaging, and 3D GRE T1-weighted imaging for the detection of dural venous sinus thrombosis and TS stenosis in comparison with contrast-enhanced MRV.

Materials and Methods
Previous sectionNext section
Patients

Institutional review board approval for this study was obtained. Informed patient consent was not required for the retrospective review of medical records and imaging studies. The radiology information system at our institution was searched for patients who had undergone MRI of the brain both without and with IV gadolinium-based contrast material and dedicated contrast-enhanced MRV during the same imaging session over a 14-month period from the inception of contrast-enhanced MRV into clinical practice at our institution. Patients were included in the study if all of the following sequences had been performed and were available for review: axial unenhanced SE T1-weighted imaging, axial contrast-enhanced SE T1-weighted imaging, contrast-enhanced 3D GRE T1-weighted imaging, and contrast-enhanced MRV including axial contrast-enhanced and subtracted axial source images and rotating maximum-intensity-projection (MIP) reconstructions. Patients were excluded if all the sequences were not acquired in the same imaging session or were not available for review. This search of records over a 14-month period yielded 73 consecutive patients who met the inclusion criteria for evaluation.

MRI

MRI was performed at either 3 T (Trio, Siemens Healthcare) or 1.5 T (Avanto, Siemens Healthcare; or Signa, GE Healthcare) using a standard head coil. A standard dose (0.1 mmol/kg) of gadobenate dimeglumine (MultiHance, Bracco Diagnostics) was injected at 2–3 mL/s for all patients using a standard length of IV tubing. Unenhanced and contrast-enhanced SE T1-weighted sequences were performed with a TR of 400–500 ms and TE of 8–9 ms with a slice thickness of 5 mm. Three-dimensional GRE T1-weighted imaging sequences (magnetization prepared rapid acquisition gradient-echo) were performed in the sagittal plane with a TR of 1900 ms, TE of 2–3 ms, and flip angle of 9–12° with a slice thickness of 1.0–1.4 mm. Un-enhanced and contrast-enhanced MRV sequences were performed in the axial plane with a TR of 4–6 ms, TE of 1–2 ms, and flip angle of 22–30° with a slice thickness of 0.8–1.4 mm.

The relative order of the sequences was as follows: unenhanced SE T1-weighted imaging, un-enhanced MRV mask, contrast injection with a 60-second pause, contrast-enhanced MRV, contrast-enhanced SE T1-weighted imaging, and contrast-enhanced 3D GRE T1-weighted imaging. The total scanner time for the unenhanced MRV sequence, delay, and contrast-enhanced sequence was approximately 3 minutes.

Image Processing and Review

In contrast to 3D GRE T1-weighted imaging, contrast-enhanced MRV utilizes a subtraction of the unenhanced and contrast-enhanced source images by the technologist at the scanner console. From this subtracted dataset, multiple oblique MIP slabs are generated per protocol by the technologist with rotation around the craniocaudal axis (“spin”) or the transverse axis (“nod”) at 6° increments. The postprocessing time for the subtraction and generation of MIPs is less than 1 minute of technologist time.

For each of the 73 patients, the imaging sequences were separated into three datasets. The first data-set consisted of the axial unenhanced SE T1-weighted imaging and contrast-enhanced SE T1-weighted imaging examinations. The second consisted of the sagittal 3D GRE T1-weighted imaging examination. The third consisted of the contrast-enhanced MRV examination including the axial contrast-enhanced and subtracted axial source images and rotating MIP reconstructions. Each group of sequences was separately stripped of all patient-identifying information and was assigned a random case number so that each could be interpreted independently in a blinded randomized fashion. This resulted in a total of 219 datasets for review. Each dataset was then reviewed independently on the same workstation by each of three experienced subspecialty-certified neuroradiologists blinded to patient information and diagnosis and without knowledge of the findings on other imaging sequences. For the 3D datasets, the workstation could be used to generate multiplanar displays.

Each dataset was scored for the presence or absence of thrombus in each of the following 10 venous segments: intracranial right and left internal jugular veins (IJVs), sigmoid sinuses, TSs, cavernous sinuses, superior sagittal sinus, and straight sinus. During evaluation of the SE T1-weighted imaging dataset, readers assessed whether hyper-intense thrombus was visible on unenhanced images in any location. Additionally, the presence or absence of tapered narrowing involving the right or left TS was assessed.

Data Analysis

Contrast-enhanced MRV was used as the reference standard to calculate diagnostic accuracy of the other sequences. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and false-positive (FP) and false-negative (FN) rates for dural venous sinus thrombosis were calculated per venous segment on SE T1-weighted imaging and 3D GRE T1-weighted imaging datasets compared with contrast-enhanced MRV. A segment was considered positive for thrombosis if it had been scored as positive by at least two of the three readers on contrast-enhanced MRV. The same calculations were also performed for dural venous sinus thrombosis per patient (if at least one segment scored positive by at least two of three readers on contrast-enhanced MRV) using the presence of intrinsic hyperintense thrombus on unenhanced SE T1-weighted imaging and the presence of a filling defect on contrast-enhanced SE T1-weighted imaging and 3D GRE T1-weighted imaging datasets. Additionally, an interobserver concordance rate (kappa statistic) was calculated for each sequence group (i.e., unenhanced SE T1-weighted imaging, contrast-enhanced SE T1-weighted imaging, contrast-enhanced 3D GRE T1-weighted imaging, and contrast-enhanced MRV).

Sensitivity, specificity, PPV, NPV, FP rate, and FN rate for TS stenosis were calculated per TS stenosis and per patient on contrast-enhanced SE T1-weighted imaging and 3D GRE T1-weighted imaging datasets compared with contrast-enhanced MRV. MRV findings were considered positive for stenosis if scored positive by at least two of the three readers. Finally, an interobserver concordance rate (kappa statistic) was calculated for each sequence group (i.e., SE T1-weighted imaging, 3D GRE T1-weighted imaging, and contrast-enhanced MRV).

Results
Previous sectionNext section
Dural Venous Sinus Thrombosis

Of the 73 patients and 730 venous segments evaluated in this study, there were a total of 13 patients (37 venous segments) with suspected venous sinus thrombosis as scored by at least one reader on at least one of the three datasets. Of those 13 patients, nine patients (23 segments) were considered to have definite dural venous sinus thrombosis on contrast-enhanced MRV, defined as two or three of three readers agreeing there was thrombosis of the segment. The segments that were determined to be thrombosed on contrast-enhanced MRV were the right IJV (n = 4), right sigmoid sinus (n = 5), right TS (n = 3), left TS (n = 3), left sigmoid sinus (n = 2), left IJV (n = 1), superior sagittal sinus (n = 4), and left cavernous sinus (n = 1). Of the nine patients determined to have dural venous sinus thrombosis in any segment, one had only one positive segment, three had two positive segments, four had three positive segments, and one had four positive segments.

The sensitivity, specificity, PPV, NPV, FP rate, and FN rate for dural venous sinus thrombosis per segment and per patient are listed in Table 1. Overall, both the presence of intrinsic hyperintensity on unenhanced SE T1-weighted imaging and a filling defect on contrast-enhanced SE T1-weighted imaging showed a PPV of 60% and a NPV of 97% for the presence of dural venous sinus thrombosis per patient in comparison with contrast-enhanced MRV. Three-dimensional GRE T1-weighted imaging had a PPV of 100% and NPV of 98% for dural venous sinus thrombosis per patient. Kappa statistics for intrinsic hyperintensity on SE T1-weighted imaging and a filling defect on SE T1-weighted imaging, 3D GRE T1-weighted imaging, and contrast-enhanced MRV per patient and per segment are listed in Table 2.

TABLE 1: Performance of Intrinsic Hyperintensity on Unenhanced Spin-Echo (SE) T1-Weighted Imaging and Filling Defect on Contrast-Enhanced SE T1-Weighted Imaging and 3D Gradient-Recalled Echo (GRE) T1-Weighted Imaging for the Detection of Dural Venous Sinus Thrombosis in Comparison With Contrast-Enhanced MR Venography (MRV)
TABLE 2: Interreader Concordance Rates for Dural Venous Sinus Thrombosis by Intrinsic T1 Hyperintensity on Unenhanced Spin-Echo (SE) T1-Weighted Imaging and Filling Defect on Contrast-Enhanced SE T1-Weighted Imaging, 3D Gradient-Recalled Echo (GRE) T1-Weighted Imaging, and Contrast-Enhanced MR Venography (MRV)

Figures 13 show representative cases of concordance and discordance of the techniques in patients with dural venous sinus thrombosis on contrast-enhanced MRV.

figure
View larger version (112K)

Fig. 1A —90-year-old woman who presented with altered mental status.

A, Axial contrast-enhanced image from MR venography (MRV) subtracted dataset shows large filling defect (arrow) in right transverse sinus and sigmoid sinus consistent with thrombosis.

figure
View larger version (159K)

Fig. 1B —90-year-old woman who presented with altered mental status.

B, Filling defect (arrow) is also apparent on sagittal contrast-enhanced 3D gradient-recalled echo (GRE) T1-weighted image.

figure
View larger version (152K)

Fig. 1C —90-year-old woman who presented with altered mental status.

C, Axial contrast-enhanced spin-echo (SE) T1-weighted image shows filling defect (arrow).

figure
View larger version (136K)

Fig. 1D —90-year-old woman who presented with altered mental status.

D, Axial unenhanced SE T1-weighted image shows intrinsic hyperintensity of thrombus (arrow). All findings were considered positive and concordant for three of three readers on contrast-enhanced MRV, 3D GRE T1-weighted imaging, SE T1-weighted imaging, and unenhanced SE T1-weighted imaging.

figure
View larger version (111K)

Fig. 2A —20-year-old woman who presented with progressive headache, nausea, vomiting, and obscured vision.

A, Axial contrast-enhanced MR venographic image shows central filling defect (arrow) in proximal left transverse sinus consistent with thrombosis.

figure
View larger version (158K)

Fig. 2B —20-year-old woman who presented with progressive headache, nausea, vomiting, and obscured vision.

B, Filling defect (arrow) is also apparent on sagittal contrast-enhanced 3D gradient-recalled echo (GRE) T1-weighted image.

figure
View larger version (138K)

Fig. 2C —20-year-old woman who presented with progressive headache, nausea, vomiting, and obscured vision.

C, Contrast-enhanced spin-echo (SE) T1-weighted image shows some heterogeneous signal in this region that is not convincing for thrombosis; this image was read as negative for thrombosis by all three readers.

figure
View larger version (131K)

Fig. 2D —20-year-old woman who presented with progressive headache, nausea, vomiting, and obscured vision.

D, There is no intrinsic hyperintense signal on unenhanced SE T1-weighted image, which was read as negative for thrombosis by all three readers.

figure
View larger version (151K)

Fig. 3A —35-year-old woman who presented with progressive headache.

A, Axial contrast-enhanced MR venographic image shows central filling defect (arrow) in proximal right transverse sinus consistent with thrombosis. Image was read as positive for thrombosis by all three blinded readers.

figure
View larger version (163K)

Fig. 3B —35-year-old woman who presented with progressive headache.

B, Sagittal contrast-enhanced 3D gradient-recalled echo T1-weighted image through same region as A does not show convincing evidence of thrombosis and was read as negative for thrombosis by all three blinded readers.

figure
View larger version (153K)

Fig. 3C —35-year-old woman who presented with progressive headache.

C, There is no evidence of filling defect on contrast-enhanced spin-echo (SE) T1-weighted image.

figure
View larger version (149K)

Fig. 3D —35-year-old woman who presented with progressive headache.

D, There is no intrinsic hyperintense signal on unenhanced SE T1-weighted image.

Transverse Sinus Stenosis

There were 58 TS or sigmoid sinus stenoses in 30 patients suspected by at least one reader on at least one technique, of which 50 stenoses in 25 patients (all bilateral) were determined to be definite on contrast-enhanced MRV (i.e., positive for TS stenosis by at least two of three readers). Eight patients were suspected of having unilateral or bilateral TS stenosis on both SE T1-weighted imaging and 3D GRE T1-weighted imaging, but none of the readers suspected TS stenosis on contrast-enhanced MRV. The sensitivity, specificity, PPV, NPV, FP rate, and FN rate for TS stenosis per patient are listed in Table 3. The PPV and NPV of contrast-enhanced SE T1-weighted imaging for TS stenosis were 75% and 80%, respectively, and 91% and 92% for 3D GRE T1-weighted imaging in comparison with contrast-enhanced MRV. Kappa statistics for TS stenosis on SE T1-weighted imaging, 3D GRE T1-weighted imaging, and contrast-enhanced MRV per patient are listed in Table 4.

TABLE 3: Performance of Contrast-Enhanced Spin-Echo (SE) T1-Weighted Imaging and Contrast-Enhanced 3D Gradient-Recalled Echo (GRE) T1-Weighted Imaging for the Detection of Transverse Sinus Stenosis per Patient (n = 73) in Comparison With Contrast-Enhanced MR Venography (MRV)
TABLE 4: Interreader Concordance Rates per Patient (n = 73) for Transverse Sinus Stenosis on Contrast-Enhanced Spin-Echo (SE) T1-Weighted Imaging, Contrast-Enhanced 3D Gradient-Recalled Echo (GRE) T1-Weighted Imaging, and Contrast-Enhanced MR Venography (MRV)

Figure 4 shows a case with discordance between techniques in a patient with bilateral TS stenosis on contrast-enhanced MRV later given a clinical diagnosis of IIH.

figure
View larger version (121K)

Fig. 4A —38-year-old woman who presented with headaches and papilledema.

A, Focal narrowing is apparent on axial subtracted contrast-enhanced MR venography (MRV) image through right transverse sinus (TS)–sigmoid sinus junction (arrow). Image was read as positive for stenosis by all three readers.

figure
View larger version (147K)

Fig. 4B —38-year-old woman who presented with headaches and papilledema.

B, Coronal maximum-intensity-projection image from contrast-enhanced MRV dataset shows high-grade stenosis at right TS–sigmoid sinus junction (short arrow) and long-segment stenosis of left TS (long arrow).

figure
View larger version (146K)

Fig. 4C —38-year-old woman who presented with headaches and papilledema.

C, Sagittal contrast-enhanced 3D gradient-recalled echo (GRE) T1-weighted image through right TS fails to show convincing evidence of stenosis. No readers considered stenosis to be present on contrast-enhanced 3D GRE T1-weighted imaging or spin-echo T1-weighted imaging datasets.

Discussion
Previous sectionNext section

Various specific MRI techniques are available for evaluation of suspected abnormalities of the dural venous sinuses. However, in a patient presenting for MRI of the brain with headache or nonspecific symptoms, the referring service may not have considered the possibility of a dural venous sinus abnormality as a cause for the patient's symptoms and dedicated techniques such as contrast-enhanced MRV may not have been performed. Therefore it is important to understand the performance, pitfalls, and reliability of routine brain MRI sequences for the diagnosis of dural venous sinus thrombosis and TS stenosis.

The results of this study show that the finding of hyperintense thrombus on unenhanced SE T1-weighted imaging is a very insensitive finding for dural venous sinus thrombosis. It was present in only a third of the patients with dural venous sinus thrombosis in this series. The results of this study also show that SE T1-weighted imaging in general is an insensitive technique for the evaluation of dural venous sinus thrombosis. The relatively low percentage of patients in this series with dural venous sinus thrombosis, as in practice, explains why the NPV of SE T1-weighted imaging is very high. The overall specificity of the finding of hyperintense thrombus or a filling defect on SE T1-weighted imaging can be problematic because of flow-related artifacts and the relatively thick sections (4–5 mm) that are generally used for the unenhanced and contrast-enhanced techniques, respectively. This pitfall is particularly important when considering the possibility of TS stenosis because flow-related enhancement, flow-related signal void, and slice thickness effects make SE T1-weighted imaging extremely inaccurate. The interreader concordance rate for SE T1-weighted imaging was much lower than for the other techniques, particularly for TS stenosis, indicating the variability in interpretation of this technique for these indications.

Contrast-enhanced 3D GRE T1-weighted imaging sequences have the potential advantage of a higher-resolution (1-mm isotropic) volumetric acquisition with a decreased tendency for flow-related venous artifacts and diminished partial-volume effects. The results of this study indicate that the 3D GRE T1-weighted imaging sequence was highly specific for dural venous sinus thrombosis (100%) per segment evaluated and per patient. There were, however, FN cases that slightly diminished the NPV of 3D GRE T1-weighted imaging for dural venous sinus thrombosis. In these cases, readers consistently (3/3 readers) determined there was thrombosis on contrast-enhanced MRV but did not suggest dural venous sinus thrombosis on 3D GRE T1-weighted imaging. A potential explanation for this finding is that, although both techniques are GRE-based, the acquisition timing is different in the protocol. The contrast-enhanced MRV sequence is performed 60 seconds after IV gadolinium administration during a time when there is peak venous enhancement. The 3D GRE T1-weighted imaging sequence in this study was performed approximately 5 minutes after contrast injections and takes approximately 4–5 minutes to acquire. In theory, gadolinium contrast material could diffuse into intravascular thrombus over this time and result in enhancement, limiting detection over time on 3D GRE T1-weighted imaging and representing a potential pitfall resulting in a FN study. Interobserver concordance for 3D GRE T1-weighted imaging was higher than for SE T1-weighted imaging but did not approach the 95–98% that contrast-enhanced MRV showed for both dural venous sinus thrombosis and TS stenosis.

In a study of 28 patients without venous sinus thrombosis, Mermuys et al. [23] reported that the contrast-enhanced 3D GRE T1-weighted imaging sequence achieved superior contrast-to-noise and signal-to-noise ratios compared with contrast-enhanced MRV but that the two techniques offered no significant difference in visualization of intracranial venous structures on qualitative review of the images. Although the study did not assess venous sinus thrombosis or stenosis, one might extrapolate that the two techniques would be equivalent or perhaps that the 3D GRE T1-weighted imaging technique would be superior in the evaluation of these conditions. Our finding that the MRV is superior for detection of dural venous sinus thrombosis supports the idea that the imaging time after contrast injection affects the likelihood of detecting dural venous sinus thrombosis.

Three-dimensional GRE T1-weighted imaging performed well for the detection of TS stenosis but did not perform nearly as well as contrast-enhanced MRV for this purpose. Because this difference in performance is unlikely a contrast-timing issue, it may be related to the generation of subtracted datasets and MIPs for contrast-enhanced MRV. The information provided in rotating MIP images adds 3D information about stenoses that often occur in an oblique orientation and that may be unclear on axial and even on multi-planar reconstructed imaging used in evaluation of these patients.

This study used contrast-enhanced MRV as the reference standard by which the other techniques were assessed. Although it is possible that the contrast-enhanced MRV technique could be systematically inaccurate for the detection of dural venous sinus thrombosis and TS stenosis, previous studies have documented its accuracy and superiority in comparison with phase contrast and TOF MRV techniques. Liang et al. [18] used digital subtraction angiography as the diagnostic gold standard to compare a contrast-enhanced 3D GRE T1-weighted imaging sequence with 2D TOF MRV and conventional SE imaging in the diagnosis of venous sinus thrombosis. The study showed superior performance of the 3D GRE T1-weighted imaging sequence (sensitivity, 83.3%; specificity, 99.6%) over 2D TOF MRV (sensitivity, 51.0%; specificity, 92.5%) and SE T1-weighted imaging (sensitivity and specificity: unenhanced, 33.3% and 84.3%, respectively; contrast-enhanced, 14.7% and 80.0%). Interobserver agreement was also superior on 3D GRE T1-weighted imaging (kappa coefficient = 0.76) compared with the other techniques. Farb et al. [32] performed contrast-enhanced MRV of 29 patients with a clinical diagnosis of IIH and 59 control patients with normal neurologic findings. Their results revealed a higher degree of correlation between venous sinus stenosis and IIH (sensitivity and specificity of 93% for the presence of stenosis to predict the clinical diagnosis of IIH) with improved visualization of the sigmoid sinus–TS junctions on contrast-enhanced MRV compared with prior studies using TOF MRV.

The treatment of TS stenosis is controversial, with some groups suggesting a role for stenting in cases unresponsive to maximal medical therapy [33]. In these cases, accurate determination of the presence and extent of TS stenosis using an optimal noninvasive imaging technique (contrast-enhanced MRV) may prove useful for patient selection before catheter-based venographic imaging.

There are limitations of the current study. First, only nine patients with venous sinus thrombosis were captured by the inclusion criteria, resulting in a small total number of positive cases and a small ratio of positive-to-negative cases. Although this study population simulated the true proportion of positive cases in our practice and helped minimize the readers' bias that might have existed with a larger percentage of positive cases, the small number of positive cases could decrease the power of this study to detect differences between the imaging techniques. Because of the rarity of venous sinus thrombosis and the constant refinement of MRI protocols, amassing a larger cohort with identical MRI sequences for comparison is a difficult task; indeed the largest series of patients with venous sinus thrombosis evaluated with contrast-enhanced MRV included only 20 positive cases [28]. Second, we made no distinction between acute, sub-acute, and chronic thrombosis when selecting or evaluating patients. Prior studies have revealed that chronic, organized thrombus can show enhancement on contrast-enhanced 3D GRE T1-weighted imaging, simulating normal flow [16, 17]. This could potentially account for the discrepancy between thrombus detection rates in the 3D GRE T1-weighted imaging and contrast-enhanced MRV sequences in our study. Newer techniques involving time-resolved contrast-enhanced MRV have shown promise in accurately characterizing chronic, organized thrombus and partially recanalized thrombus [26, 28].

Overall, this study indicates that SE T1-weighted imaging is less accurate and has lower interobserver concordance than the other techniques evaluated. Contrast-enhanced 3D GRE T1-weighted imaging had greater interobserver concordance, reasonable diagnostic accuracy for TS stenosis, and high specificity for dural venous sinus thrombosis but did not approach the sensitivity of contrast-enhanced MRV. It is therefore important to realize the pitfalls and limitations of the SE T1-weighted imaging and 3D GRE T1-weighted imaging techniques for the detection of dural venous sinus abnormalities and to have a low threshold for recommending and performing dedicated venous imaging. We suggest that for screening MR protocols for imaging of headache, the additional acquisition time for a contrast-enhanced 3D GRE T1-weighted imaging sequence is offset by the superior ability to detect dural venous sinus thrombosis and TS stenosis as potential causes than protocols using only SE-based T1-weighted imaging. For cases in which a dural venous sinus abnormality is specifically in question, contrast-enhanced MRV should be performed to minimize the likelihood of a FN result for both dural venous sinus thrombosis and TS stenosis.

Presented at the 2011 annual meeting of the American Society of Neuroradiology, Seattle, WA.

References
Previous sectionNext section
1. Stam J. Thrombosis of the cerebral veins and sinuses. N Engl J Med 2005; 352:1791–1798 [Google Scholar]
2. Bousser MG, Ferro JM. Cerebral venous thrombosis: an update. Lancet Neurol 2007; 6:162–170 [Google Scholar]
3. Stam J, De Bruijn SF, DeVeber G. Anticoagulation for cerebral sinus thrombosis. Cochrane Database Syst Rev 2002; 4:CD002005 [Google Scholar]
4. Ferro JM, Canhao P, Stam J, Bousser MG, Barinagarrementeria F. Prognosis of cerebral vein and dural sinus thrombosis: results of the International Study on Cerebral Vein and Dural Sinus Thrombosis (ISCVT). Stroke 2004; 35:664–670 [Google Scholar]
5. Stienen A, Weinzierl M, Ludolph A, Tibussek D, Hausler M. Obstruction of cerebral venous sinus secondary to idiopathic intracranial hypertension. Eur J Neurol 2008; 15:1416–1418 [Google Scholar]
6. De Simone R, Ranieri A, Bonavita V. Advancement in idiopathic intracranial hypertension pathogenesis: focus on sinus venous stenosis. Neurol Sci 2010; 31(suppl 1):S33–S39 [Google Scholar]
7. Dhungana S, Sharrack B, Woodroofe N. Idiopathic intracranial hypertension. Acta Neurol Scand 2010; 121:71–82 [Google Scholar]
8. Friedman DI, Jacobson DM. Diagnostic criteria for idiopathic intracranial hypertension. Neurology 2002; 59:1492–1495 [Google Scholar]
9. Celebisoy N, Gokcay F, Sirin H, Akyurekli O. Treatment of idiopathic intracranial hypertension: topiramate vs acetazolamide, an open-label study. Acta Neurol Scand 2007; 116:322–327 [Google Scholar]
10. Bynke G, Zemack G, Bynke H, Romner B. Ventriculoperitoneal shunting for idiopathic intracranial hypertension. Neurology 2004; 63:1314–1316 [Google Scholar]
11. Lueck C, McIlwaine G. Interventions for idiopathic intracranial hypertension. Cochrane Database Syst Rev 2005; 3:CD003434 [Google Scholar]
12. Rohr A, Dorner L, Stingele R, Buhl R, Alfke K, Jansen O. Reversibility of venous sinus obstruction in idiopathic intracranial hypertension. AJNR 2007; 28:656–659 [Google Scholar]
13. Bussière M, Falero R, Nicolle D, Proulx A, Patel V, Pelz D. Unilateral transverse sinus stenting of patients with idiopathic intracranial hypertension. AJNR 2010; 31:645–650 [Google Scholar]
14. 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–S41; discussion, S42–S43 [Google Scholar]
15. Bianchi D, Maeder P, Bogousslavsky J, Schnyder P, Meuli RA. Diagnosis of cerebral venous thrombosis with routine magnetic resonance: an update. Eur Neurol 1998; 40:179–190 [Google Scholar]
16. Dormont D, Sag K, Biondi A, Wechsler B, Marsault C. Gadolinium-enhanced MR of chronic dural sinus thrombosis. AJNR 1995; 16:1347–1352 [Google Scholar]
17. Leach JL, Wolujewicz M, Strub WM. Partially recanalized chronic dural sinus thrombosis: findings on MR imaging, time-of-flight MR venography, and contrast-enhanced MR venography. AJNR 2007; 28:782–789 [Google Scholar]
18. Liang L, Korogi Y, Sugahara T, et al. Evaluation of the intracranial dural sinuses with a 3D contrast-enhanced MP-RAGE sequence: prospective comparison with 2D-TOF MR venography and digital subtraction angiography. AJNR 2001; 22:481–492 [Google Scholar]
19. Liang L, Korogi Y, Sugahara T, et al. Normal structures in the intracranial dural sinuses: delineation with 3D contrast-enhanced magnetization prepared rapid acquisition gradient-echo imaging sequence. AJNR 2002; 23:1739–1746 [Google Scholar]
20. Kirchhof K, Welzel T, Jansen O, Sartor K. More reliable noninvasive visualization of the cerebral veins and dural sinuses: comparison of three MR angio-graphic techniques. Radiology 2002; 224:804–810 [Google Scholar]
21. Farb RI, Scott JN, Willinsky RA, Montanera WJ, Wright GA, terBrugge KG. Intracranial venous system: gadolinium-enhanced three-dimensional MR venography with auto-triggered elliptic centric-ordered sequence—initial experience. Radiology 2003; 226:203–209 [Google Scholar]
22. Wetzel SG, Law M, Lee VS, Cha S, Johnson G, Nelson K. Imaging of the intracranial venous system with a contrast-enhanced volumetric interpolated examination. Eur Radiol 2003; 13:1010–1018 [Google Scholar]
23. Mermuys KP, Vanhoenacker PK, Chappel P, Van Hoe L. Three-dimensional venography of the brain with a volumetric interpolated sequence. Radiology 2005; 234:901–908 [Google Scholar]
24. Hu HH, Campeau NG, Huston J 3rd, Kruger DG, Haider CR, Riederer SJ. High-spatial-resolution contrast-enhanced MR angiography of the intracranial venous system with fourfold accelerated two-dimensional sensitivity encoding. Radiology 2007; 243:853–861 [Google Scholar]
25. Klingebiel R, Bauknecht HC, Bohner G, Kirsch R, Berger J, Masuhr F. Comparative evaluation of 2D time-of-flight and 3D elliptic centric contrast-enhanced MR venography in patients with presumptive cerebral venous and sinus thrombosis. Eur J Neurol 2007; 14:139–143 [Google Scholar]
26. Meckel S, Glucker TM, Kretzschmar M, Scheffler K, Radu EW, Wetzel SG. Display of dural sinuses with time-resolved, contrast-enhanced three-dimensional MR venography. Cerebrovasc Dis 2008; 25:217–224 [Google Scholar]
27. Fu JH, Lai PH, Hsiao CC, et al. Comparison of real-time three-dimensional gadolinium-enhanced elliptic centric-ordered MR venography and two-dimensional time-of-flight MR venography of the intracranial venous system. J Chin Med Assoc 2010; 73:131–138 [Google Scholar]
28. Meckel S, Reisinger C, Bremerich J, et al. Cerebral venous thrombosis: diagnostic accuracy of combined, dynamic and static, contrast-enhanced 4D MR venography. AJNR 2010; 31:527–535 [Google Scholar]
29. 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; 59:1021–1026 [Google Scholar]
30. Idbaih A, Boukobza M, Crassard I, Porcher R, Bousser MG, Chabriat H. MRI of clot in cerebral venous thrombosis: high diagnostic value of susceptibility-weighted images. Stroke 2006; 37:991–995 [Google Scholar]
31. Leach JL, Strub WM, Gaskill-Shipley MF. Cerebral venous thrombus signal intensity and susceptibility effects on gradient recalled-echo MR imaging. AJNR 2007; 28:940–945 [Google Scholar]
32. Farb RI, Vanek I, Scott JN, et al. Idiopathic intracranial hypertension: the prevalence and morphology of sinovenous stenosis. Neurology 2003; 60:1418–1424 [Google Scholar]
33. Ahmed RM, Wilkinson M, Parker GD, et al. Transverse sinus stenting for idiopathic intracranial hypertension: a review of 52 patients and of model predictions. AJNR 2011; 32:1408–1414 [Google Scholar]
Address correspondence to A. M. Saindane ().

Recommended Articles

Performance of Spin-Echo and Gradient-Echo T1-Weighted Sequences for Evaluation of Dural Venous Sinus Thrombosis and Stenosis

Full Access, ,
American Journal of Roentgenology. 2016;206:1298-1306. 10.2214/AJR.15.15719
Abstract | Full Text | PDF (979 KB) | PDF Plus (848 KB) | Supplemental Material 
Full Access, , , ,
American Journal of Roentgenology. 2007;189:S64-S75. 10.2214/AJR.07.7015
Abstract | Full Text | PDF (6409 KB) | PDF Plus (2747 KB) 
Full Access,
American Journal of Roentgenology. 2011;196:23-31. 10.2214/AJR.10.5323
Abstract | Full Text | PDF (1131 KB) | PDF Plus (1102 KB) | Erratum 1 | Erratum 2 
Full Access, , ,
American Journal of Roentgenology. 2013;200:W1-W16. 10.2214/AJR.11.8080
Citation | Full Text | PDF (1473 KB) | PDF Plus (1383 KB) 
Full Access,
American Journal of Roentgenology. 2011;197:1309-1321. 10.2214/AJR.11.7420
Abstract | Full Text | PDF (1195 KB) | PDF Plus (1215 KB) 
Full Access, ,
American Journal of Roentgenology. 2009;193:879-886. 10.2214/AJR.08.2257
Abstract | Full Text | PDF (930 KB) | PDF Plus (646 KB)