Diagnostic value of gadolinium-enhanced 3D magnetic resonance angiography in patients with suspected hepatic arterial complications after liver transplantation.

Purpose: To evaluate the accuracy and clinical role of gadolinium-enhanced 3D magnetic resonance angiography (MRA) in patients with suspected hepatic arterial complications after liver transplantation. Materials and Methods: Thirty-six consecutive MRA studies were performed in 33 liver transplant recipients after transplantation. MRA image quality was assessed subjectively. Thirty-two MRA studies were retrospectively reviewed and correlated with surgery (n = 2), conventional angiography (n = 18), or clinical follow-up (n = 12). MRA findings were also correlated with those of Doppler sonography in 30 of the cases. In 20 cases, concordance between MRA and surgery or conventional angiography was evaluated for each grade of hepatic artery stenosis (normal, mild [< 50%], moderate [50-75%], severe [> 75%], or occluded). Results: MRA image quality was degraded 13 of 36 cases (36.1%) studies. The sensitivity, specificity, and accuracy of MRA by consensus reading for more than 50% of hepatic artery stenosis or occlusion were 67%, 90%, and 81.3%, respectively. Of the 19 cases in which Doppler sonography was abnormal, MRA correctly characterized hepatic artery stenosis in 16 (84.2%). MRA also correctly identified all 5 occurrences of celiac artery stenosis. However, MRA overestimated the severity of hepatic arterial stenosis in 3 (15%) of 20 cases and underestimated 5 (25%) of 20 cases. Conclusion: MRA complements Doppler ultrasound to exclude significant hepatic artery stenosis. However, a substantial number of MRA studies were technically inadequate, and MRA demonstrated limited efficacy for correctly grading the severity of hepatic artery stenosis.