Bidirectional Flow in the Vertebral Artery Is Not Always Indicative of the Subclavian Steal Phenomenon

Objective-To evaluate the causes of bidirectional flow in the vertebral artery detected by Doppler sonography and its differential diagnosis. Methods-Twenty-nine patients with bidirectional flow in the vertebral artery were retrospectively studied. The vertebral artery parameters, including peak antegrade velocity (PAV), peak reversed velocity (PRV), maximum peak velocity (MPV), peak systolic velocity, resistive index (RI), and diameter, were measured. The MPV was defined as the MPV of bidirectional flow regardless of the velocity of antegrade or retrograde flow. To better predict the cause of bidirectional flow, receiver operating characteristic curves were constructed for these parameters, and the best cutoff values were obtained. The cause of bidirectional flow was determined by angiography. Results-The causes of bidirectional flow were classified as the subclavian steal phenomenon (n = 21) and factors unrelated to the steal phenomenon (n = 8, including a hypoplastic vertebral artery [n = 4] and proximal vertebral artery stenosis and occlusion [n = 4]). Significant differences were observed between the steal phenomenon and non steal phenomenon groups (P < .05) for MPV, PRV, PAV, target vertebral artery diameter, and contralateral RI. To determine the cause of bidirectional flow, areas under the receiver operating characteristic curves for the different parameters were obtained: 0.929 for MPV, 0.881 for PRV, 0.824 for PAV, 0.753 for target vertebral artery diameter, and 0.845 for contralateral RI. The cutoff value for MPV was 26.1 cm/s, and the accuracy was 93% (27 of 29). Conclusions-Bidirectional flow in the vertebral artery is not always indicative of the subclavian steal phenomenon. Measurement of hemodynamic parameters in the vertebral artery, such as MPV, can facilitate determination of the cause of bidirectional flow.