The aim of this study was to assess the sensitivity and specificity of pseudo-continuous arterial spin labeling (PCASL) magnetic resonance angiography (MRA) with 3-dimensional (3D) radial acquisition for the detection of intracranial arteriovenous (AV) shunts.Materials and Methods
A total of 32 patients who underwent PCASL-MRA, clinical magnetic resonance imaging (MRI)/MRA exam, and digital subtraction angiography (DSA) were included in this retrospective analysis. Twelve patients presented with AV shunts. Among these were 8 patients with AV malformations (AVM) and 4 patients with AV fistulas (AVF). The clinical MRI/MRA included 3D time-of-flight MRA in all cases and time-resolved, contrast-enhanced MRA in 9 cases (6 cases with AV shunting). Research MRI and clinical MRI were independently evaluated by 2 neuroradiologists blinded to patient history. A third radiologist evaluated DSA imaging. A diagnostic confidence score was used for the presence of abnormalities associated with AV shunting (1–5). The AVMs were characterized using the Spetzler-Martin scale, whereas AVFs were characterized using the Borden classification. κ Statistics were applied to assess intermodality agreement.Results
Compared with clinical MRA, noncontrast PCASL-MRA with 3D radial acquisition yielded excellent sensitivity and specificity for the detection of intracranial AV shunts (reader 1: 100%/100%, clinical MRA: 91.7%, 94.4%; reader 2: 91.7%/100%, clinical MRA: 91.7%/100%). Diagnostic confidence was 4.8/4.66 with PCASL-MRA and 4.25/4.66 with clinical MRA. For AVM characterization with PCASL-MRA, intermodality agreement with DSA showed κ values of 0.43 and 0.6 for readers 1 and 2, respectively. For AVF characterization, intermodality agreement showed κ values of 0.56 for both readers.Conclusion
Noncontrast PCASL-MRA with 3D radial acquisition is a potential tool for the detection and characterization of intracranial AV shunts with a sensitivity and specificity equivalent or higher than routine clinical MRA.