Object: The purpose of this study was to evaluate the accuracy of multidetector-row CT angiography (MDCTA) in demonstrating spinal dural arteriovenous fistulas (SDAVFs).
Methods: The authors studied 10 patients with SDAVFs, including 2 with spinal epidural AVFs, who underwent preoperative MR imaging, MDCTA, and digital subtraction angiography (DSA). In the evaluation of coronal sections of multiplanar reformation MDCTA images, inspection was focused on the presence of the following findings: 1) dilated perimedullary veins in the spinal canal; 2) focal enhancement of the nerve root, suggesting the location of the AVF, around the dural sleeve; and 3) a radicular vein that drains the AVF into perimedullary veins. The utility of MDCTA was assessed by comparing its findings with those of DSA in each case.
Results: Digital subtraction angiography confirmed that the AVFs were located in the thoracic spine in 4 patients and in the lumbar spine in 6 patients, and MDCTA detected dilated perimedullary veins in all 10 patients. In 8 patients, there was focal enhancement of the nerve root. The radicular vein that drains the AVF into the perimedullary veins was found in 8 cases. In 8 cases, the MDCTA-derived level and side of the AVF and its feeder corresponded with those shown by DSA. In 2 patients, however, the MDCTA-derived side of the feeder was on the side contralateral to the feeding artery confirmed by DSA. These lesions were interpreted as spinal epidural AVFs with perimedullary drainage. In 2 cases, MDCTA could not detect the multiplicity of their feeders.
Conclusions: The use of MDCTA preceding DSA can be helpful to focus the selective catheter angiography on certain spinal levels. However, one should keep in mind that epidural AVFs with perimedullary drainage may resemble SDAVFs and also that MDCTA cannot exclude the possibility of multiple feeders. Further research should elucidate how broadly selective angiography should explore around the MDCTA-suggested target.