, et al
Color-Doppler US of normal cerebral venous sinuses in neonates: a comparison with MR venography
Pediatr Radiol.
; doi:

Investigators from the Hospital for Sick Children, Toronto, studied the specificity, inter-modality and inter-reader agreement of color Doppler ultrasound (CDUS) for the evaluation of cerebral venous sinuses in neonates and compared the results to magnetic resonance venography (MRV). The authors prospectively evaluated consecutive hospitalized infants referred to radiology by a neonatologist with a MRI-MRV request. Patients were included if there was a clinically indicated need for a first-time MRV and the US could be performed within 10 hours.

CDUS was performed by a pediatric radiology fellow or experienced sonographer. The CDUS was performed through the anterior fontanelle or other intersutural acoustic windows. Each vein was imaged in 2 planes. Veins were interpreted as normal/patent versus abnormal/not patent versus indeterminate or not visualized by 2 experienced pediatric radiologists blinded to clinical information and any other imaging examination results. MRVs were interpreted by 2 experienced neuroradiologists. Intermodality and inter-reader agreement was determined for CDUS and MRV. Specificity was calculated for CDUS with MRV as the gold standard.

Fifty patients with mean gestational age of 37 weeks (range 25-41) and averaging 19 days old (range 0-119 days) were evaluated. Overall inter-reader agreement for CDUS findings was 97%, 99% for MRV, and 99% and 100% for intermodality (95% CI, 96%-100%). CDUS readers had 100% agreement regarding the superior sagittal sinus, vein of Galen, and sigmoid sinuses. Straight sinus, cerebral veins, left and right transverse sinuses, and torcular Herophili agreement was considered excellent (≥90% agreement). Agreement on the inferior sagittal sinus was 82% for CDUS and 90% for MRV. With CDUS, inferior sagittal sinus flow could be seen in 24 of 31 cases where it could not be seen with MRV. MRV had technical limitations (“flow gaps”) in 20% of cases; CDUS helped prove flow in these gaps. CDUS examinations took 30 to 40 minutes. Disagreements between CDUS and MRV readers (3 cases) were due to hypoplasia of a transverse sinus, bony artifact, and limited medial visualization of a transverse sinus. Overall specificity of CDUS for disease or thrombus in the examined vessel was 80% (95% CI, 66%-90%).

The authors conclude that CDUS and MRV show excellent interreader and inter-modal agreement, and that CDUS has high specificity in ruling out neonatal cerebral venous thrombosis (CVT).

The risk of CVT in children is greatest among neonates and has become recognized as a key cause of neonatal brain injury associated with high mortality and poor outcome. Venous sinus thrombosis has been associated with venous brain infarction in 60% of cases and treatment with antithrombotic agents can limit clot propagation, reduce complications, and improve prognosis. However, the symptoms are nonspecific and guidance relies on neuroimaging. The current gold standard for the analysis of the intracranial sinovenous system (SVS) is MRI, specifically MRV, which...

You do not currently have access to this content.