The Diagnostic Value Of Susceptibility-Weighted Images For Cerebral Venous Thrombosis

Objective: Cerebral Venous Thrombosis (CVT) is a rare disorder, and itsdiagnosis is challenging. Comparing different imaging modalities, the aim ofour study was to investigate the value of susceptibility weighted imaging (SWI)in the detection of cerebral vein thrombosis.Method: A retrospective review of28consecutive inpatients with CVTand20control patients without any venous pathologies from May2006toDecember2011was undertaken in our hospital. SWI was available for all ofthem. Medical records including age, sex, clinical manifestations wereanalyzed retrospectively. The analysis of neuroimaging was performedindependently by two blinded experienced neuroradiologists, who evaluatedthe imaging modalities and sequences separately. The sensitivity, specificity,positive predictive values (PPV) and negative predictive values (NPV) andinter-observer agreement of the different modalities were calculated. When athrombosis was suspected by a reader, the involved sinuses and cerebral veinsand in addition, the presence of intracerebral venous infarction or venoushemorrhage, were noted. Consensus reading with interpretation of thedifferent imaging modalities served to establish the definite diagnosis. If thetwo readers in imaging diagnosis is inconsistent, refered to thethirdneuroradiologist opinion.Results: In the case group,18are females, the average age of thepatients is33.93+10.32years. Of the28CVT patients, nineteen have definiterisk factors while nine have not. The most common risk factor is bloodhypercoagulability such as gestation, puerperium, etc. Clinical manifestationsof CVT are divided into four kinds:(1)Isolated intracranial hypertension: themanifestations are headache, vomiting (especially injection) and papillaedema, blurred vision etc, totally13cases;(2)Focal deficits or seizures: Manifested as partial or generalized seizures, aphasia, hemiplegia, hemianopiaetc. totally11cases;(3)Subacute encephalopathy: Disturbance ofconsciousness as the main performance, totally3cases;(4)Cavernous sinusthrombosis: Mainly as eye symptoms, such as orbital pain, restricted eyemovements, just1case. According of the imaging diagnosis, Most of ourpatients present with venous sinus thrombosis, we identify only two patientswith isolated cortical venous thrombosis. In our patients, the frontal veins areinvolved in most of the cases, following by the parietal veins as the secondmost frequent localization. Dural sinus affected most is superior sagittal sinusand lateral sinus (including sigmoid sinus and transverse sinus). Usually morethan one dural sinus are affected. It is not common to see only one sinusinvolved. SWI shows the highest sensitivity for the detection of isolatedcortical venous thrombosis (95%), following by T₁WI (60%). FLAIR andMRV have a sensitivity of50%and35%, respectively, whereas the sensitivityof CT is below30%. SWI and MRV are the preferred methods for diagnosisof dural sinus thrombosis. FLAIR shows the highest sensitivity forintracerebral edema,while SWI was given precedence for hemorrhage. Thespecificity of all modalities is more than90%, with good to perfectinter-observer agreement. To determine the inter-observer agreementregarding the presence of a CVT or SVT, multi-rater κ values are calculated.The inter-observer agreement for CVT is perfect for FLAIR (κ=1), good forT₁WI and MRV (κ=0.73and0.75, respectively), and excellent (κ>0.80) forall other modalities.Conclusions: Susceptibility Weighted Image is the superior MR imagingsequence for diagnosing isolated cortical venous thrombosis and hemorrhage,especially at the acute stage. MRV and SWI are the preferred methods fordiagnosis of dural sinus thrombosis.