Duplex Ultrasonography-Detected Positional Vertebral Artery Occlusion in Upper Cervical Rheumatoid Arthritis.

STUDY DESIGN Prospective imaging study. OBJECTIVE To clarify the frequency of positional vertebral artery (VA) occlusion using duplex ultrasonography in patients with rheumatoid arthritis (RA). SUMMARY OF BACKGROUND DATA Some patients with upper cervical RA develop thromboembolic stroke related to positional and transient VA occlusions; however, whether RA patients have positional VA occlusion without neurological symptoms is unclear. METHODS Outpatients with RA were enrolled. Clinical data were collected, and radiograph examinations were performed to measure the anterior atlantodental interval (AADI), the posterior atlantodental interval (PADI), and the Ranawat method. Patients underwent duplex ultrasonography during rotation to the contralateral side of the examination side, flexion, and extension of their neck. If positional VA occlusion was detected, CT angiography was conducted in the neutral position and in the same position that showed VA occlusion on duplex ultrasonography. Clinical and radiological data were compared between the VA occlusion (VAO) group and the non-VAO group. Sensitivity-specificity curve analyses were performed to clarify optimal threshold values of AADI, PADI, and the Ranawat method for predicting positional VA occlusion. RESULTS Of the 132 RA patients, dynamic duplex ultrasonography showed positional VA occlusion in eight (6%) patients. Patients in the VAO group had a greater AADI (median, 7.4 vs. 2.3 mm; P < 0.001), a shorter PADI (median, 13.7 vs. 19.6 mm; P = 0.002), and a lower Ranawat value (median, 13.7 vs. 16.8 mm; P = 0.006) than those in the non-VAO group. Cut-off values of AADI, PADI, and the Ranawat method for predicting positional VA occlusion were 6.5, 14.0, and 15.5 mm, respectively. CONCLUSION A subset of RA patients developed positional VA occlusion associated with cervical spine involvement.