Duplex ultrasound in the subclavian steal syndrome

The effect of subclavian steal on the contralateral vertebral flow and its possible effect on carotid flow were studied and the US results were compared to the angiographic findings.

The study consisted of the records of 74 patients with a duplex Doppler finding of subclavian steal syndrome. Of these, 48 patients had had both angiography and US of the neck arteries and were selected for comparison. For a comparison of the US flow values, a control series of 48 was selected from our carotid archive and consisted of patients without subclavian steal who had also been examined with both US and angiography.

Of the 48 patients, 44 had a subclavian steal syndrome at angiography, 31 on the left side and 13 on the right. Of the 44, 21 patients had subclavian occlusion, and 23 stenosis. In 84% of the subclavian occlusions, US showed a complete systodiastolic steal and in 16% a partial systolic steal, while the corresponding findings for subclavian stenoses were 17% and 83%. Of the 48 cases, 4 were not real subclavian steals: 2 had vertebral occlusion (1 with a 90% subclavian diameter stenosis) on the side of retrograde flow at US and a steal through the vertebral and collateral arteries to the spinal arteries; and 2 had a 70% diameter stenosis of the subclavian artery and to-and-fro flow in the vertebral artery at angiography. In the 21 cases of complete steals, the subclavian diameter stenosis was 97 +/- 8% at angiography, and in the 23 partial steals, it was 85 +/- 10%. There was a significant increase in contralateral vertebral and common carotid flow in the cases with retrograde vertebral flow compared to the vertebral and common carotid flow of the control subjects. The retrograde flow values, the ipsilateral vertebral lumen diameter, and the flow values in the common carotid arteries were higher in complete steals and subclavian occlusions than in partial steals.

A complete vertebral steal at US correlated well with subclavian occlusion and a partial steal suggested stenosis of the subclavian artery. There were also flow changes in the contralateral vertebral artery and the common carotid arteries that compensated for the steal. Retrograde vertebral flow at US was sometimes associated with vertebral occlusion in subclavian stenosis without a true subclavian steal.