Transradial cerebral angiography: predicting left ICA selective angiography success using pre-diagnostic aortic arch factors.

Abstract

Background: Transradial cerebral angiography (TRA) is a convenient but challenging procedure, particularly for selecting the left internal carotid artery (ICA) and vertebral artery.

Objective: To predict the selection of the left ICA using CT and MR images acquired before TRA.

Methods: Overall, 306 patients with TRA were enrolled and divided into either the group with success (264 patients) or the failure (42 patients) group. The following anatomical factors were measured: A1 (subclavian artery angle), A2 (right subclavian-innominate artery angle), A3 (innominate-left common carotid artery angle), D1 (aorta to right subclavian artery length), and D2 (innominate-to-left common carotid artery length).

Results: The median values for A1, A2, A3, D1, and D2 were 81.57° (IQR 69.26-94.14), 147.03° (125.73-161.09), 24.73 (15.85-37.72°), 34.73 mm (29.68-38.48), and 13.15 mm (11.33-15.64), respectively, with significant differences observed between the successful and failure groups in A3 (26.88° vs 15.50°; P<0.001), D1 (34.24 mm vs 37.62 mm; P<0.001), and D2 (12.78 mm vs 14.91 mm; P<0.001). The aortic arch type did not affect success (P=0.134), while patients in the failure group were significantly older (P<0.001). A predictive logistic regression model was developed, revealing differing factor impacts when controlling variables. The model (area under the curve 0.87) highlights data complexity and enables user-friendly prediction of left ICA-selective TRA success (https://je0000000342227505.shinyapps.io/icatra/).

Conclusion: This study demonstrated that the success of left ICA selective angiography can be predicted using aortic arch images, providing a basis for the extension of TRA.