Accuracy and Value of Time-resolved MRA for Spinal Vascular Malformations

Study Design. Retrospective. Objective. To explore the value of time-resolved CE-MRA in evaluating and locating the SVM before digital subtraction angiography (DSA). Summary of Background Data. Spinal vascular malformations (SVM) can be detected with time-resolved contrast-enhanced MRA (CE-MRA). Materials and Methods. One hundred seventy-eight patients with suspected SVM who underwent time-resolved CE-MRA examination and DSA were included in this study. DSA served as the reference standard. The type of SVM, feeding arteries, fistula/nidus, and proximal segment of draining veins were evaluated on time-resolved CE-MRA. The diagnostic performance and classification performance of time-resolved CE-MRA in the diagnosis of SVM is summarized in terms of overall accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The difference in catheterized vessels during spinal vascular DSA between the actual and estimated number was also analyzed. Results. One hundred forty-seven patients were diagnosed with SVM (20 cervical, 118 thoracolumbar, and 9 deep lumbosacral), and 31 patients were diagnosed with non-SVM, according to DSA findings. The diagnostic sensitivity, specificity, PPV, NPV, and accuracy of time-resolved CE-MRA for a subtype of SVM were 0.961, 0.961, 0.993, 0.806, and 0.961, respectively. The overall accuracy of time-resolved CE-MRA for the diagnosis of SVM was 0.821 and 0.783, 0.793, and 0.778 for cervical, thoracolumbar, and deep lumbosacral SVM, respectively. The actual number of catheterized vessels during spinal vascular DSA with time-resolved CE-MRA as the reference was lower than the estimated number of catheterized vessels in both SVM and non-SVM patients (P<0.001). Conclusions. Time-resolved CE-MRA could accurately evaluate SVM and reduce the number of catheterized vessels during spinal vascular DSA.