Pancreas Allografts: Comparison of Three-dimensional Rotational Angiography with Standard Digital Subtraction Angiography

Abstract

To define the role of three-dimensional (3D) rotational angiography (RA) for the evaluation of pancreas allografts and compare 3D RA to standard digital subtraction angiography (DSA). DSA and 3D RA were performed in patients with vascular abnormalities diagnosed on contrast medium-enhanced magnetic resonance (MR) angiography. Patency of the allograft vasculature, confidence in the ability to make a therapeutic decision, and value of the study for definition of the optimal projection for an intervention was assessed on a graded scale. Seventeen standard DSA projections (mean, 3.4; range, 2-6) and 10 3D RA images (mean, 2; range, 1-3) were obtained in five patients. An average iodinated contrast agent dose of 14.4 mL (range, 8-22 mL) was administered for DSA. An average CO2 dose of 54 mL (range, 0-120 mL) was administered for 3D RA. Five 3D RA procedures were timed for the arterial phase and five were timed for the arterial and venous phases. Average contrast agent doses were 17.6 mL (range, 11-22 mL) for arterial 3D RA and 24.4 mL (range, 16-34 mL) for arterial- and venous-phase 3D RA. Of 68 vascular segments available for direct comparison of patency, complete concordance was present in 96%. There was no difference in the reviewers' diagnostic confidence (10 +/- 0 for both techniques). Three-dimensional RA was considered significantly superior for planning the optimal projection for intervention (10 +/- 0 for 3D RA vs 7.2 +/- 1.6 for DSA; P = .0052). Three-dimensional RA of pancreatic allografts is feasible and does not differ in accuracy from conventional DSA. It provides similar reviewer confidence in the ability to make an accurate treatment decision, but its key advantage is its superior ability to define the optimal projection for planned endovascular interventions.