Tumor involvement in hepatic veins: comparison of MR imaging and US for preoperative assessment

To compare use of magnetic resonance (MR) imaging and ultrasonography (US) for diagnosis of vascular involvement by tumor at the hepatic vein confluence.

Thirty-seven consecutive patients with tumors at the hepatic vein confluence were prospectively evaluated with spin-echo and gradient-echo MR imaging and gray-scale and Doppler US. Encasement, thrombosis, occlusion, and nonvisualization were considered to be evidence of vascular involvement. Imaging results were compared with surgical and pathologic examination findings in 27 patients who underwent resection.

Sixteen hepatic veins (nine right, four middle, three left) were seen to be involved at surgery. Twelve of 16 involved veins were identified at MR imaging (75% sensitivity, 98% specificity, 92% positive predictive value, 94% negative predictive value). Thirteen of 16 involved veins were detected at US (81% sensitivity, 97% specificity, and 87% positive and 95% negative predictive values). There was one false-positive diagnosis of inferior vena cava involvement at both MR imaging and US. Ten patients had unresectable disease. One patient had motion artifact on MR images; in the remaining nine patients, MR imaging and US yielded identical findings at 26 of 27 hepatic vein sites.

MR imaging and US provide comparable results for assessment of hepatic vein involvement by tumor.