Diagnostic Accuracy of CT and MR Venography in Acute Cerebral Venous Thrombosis

Abstract

Background: Cerebral venous thrombosis (CVT) is a rare cause of stroke with variable clinical presentation and challenging diagnosis. Computed tomography venography (CT-V) or magnetic resonance venography (MR-V) are recommended for detecting CVT, but their diagnostic performance is unclear. This systematic review examines the accuracy of CT-V and MR-V in diagnosing CVT.

Methods: We performed a systematic literature search up to 02 July 2023 to retrieve original articles studying the diagnostic accuracy of CT-V or MR-V in patients with suspected CVT. Studies with a reference standard comprising at least one other imaging method alone or in addition to clinical data were included. The quality of studies was assessed with the QUADAS-2 tool.

Results: We found eight studies (478 subjects) published between 1995 and 2022 that compared one or more diagnostic methods to a reference: CT-V (four studies, n = 95) and/or MR-V (three studies, n = 83) versus clinical and radiological consensus, CT-V versus MR-V (two studies, n = 292), or MR-V versus digital subtraction angiography (DSA) (one study, n = 52). They reported excellent diagnostic accuracy for both CT-V (sensitivity 100%, specificity 94%–100%) and MR-V (sensitivity 90%–100%, specificity 71%–100%) compared to clinical and radiological consensus, for CT-V compared to MR-V (sensitivity 96%–100%, specificity 99%–100%), and for MR-V compared to DSA (sensitivity 100%, specificity 71%). The heterogeneity of the studies allowed only an exploratory meta-analysis. The risk of bias in the use of the reference standard was high.

Conclusions: Our systematic review found a high accuracy of CT-V and MR-V for detecting CVT, suggesting that CT-V is a feasible option for centers with limited access to MR-V. However, the small sample sizes and variable reference methods impair conclusions drawn from the current literature.