Venoocclusive liver disease after bone marrow transplantation: findings at duplex sonography

Ascites, thickening of the gallbladder wall, and reversal of portal flow are documented sonographic findings in venoocclusive disease of the liver. The frequency and specificity of these findings and their relationship to the severity of this disease have not been studied. In an attempt to clarify these issues, 65 patients who had bone marrow transplantations were prospectively studied with serial B-scans and duplex color Doppler sonography. For all patients, assessment included liver size and texture, thickening of the gallbladder wall (greater than 10 mm), and presence of ascites. Doppler flow velocity profiles were obtained from the portal vein, hepatic veins, and inferior vena cava. The hepatic artery resistive index (RI) was calculated. Twenty volunteers were also studied to establish normal flow values. Nineteen patients had documented venoocclusive disease, nine had hepatic graft-vs-host disease (GVHD) (five after proved venoocclusive disease), two had hepatitis, and 40 had no clinical or biochemical evidence of liver injury after bone marrow transplantation. Ascites (n = 16), thickening of the gallbladder wall (n = 8), hepatomegaly (n = 8), and altered liver texture (n = 3) were not distinguishing features of venoocclusive disease. Mean hepatic artery RI was as follows (ranges are in parentheses): control group, 0.69 (0.58-0.76); venoocclusive disease patients, 0.81 (0.75-0.87); GVHD patients, 0.69 (0.63-0.71); all other patients after bone marrow transplantation, 0.66 (0.61-0.71). The RI values in venoocclusive disease were significantly elevated, but an incremental rise in RI with increasing severity of the disease was not seen. Abnormalities in portal vein flow were seen in only two patients: in one with fatal venoocclusive disease, reversed portal flow developed, and in one with GVHD, portal vein thrombosis developed. Contrary to previous reports, no correlation between abnormalities in portal flow and venoocclusive disease was seen. Flow velocities in the hepatic veins and the inferior vena cava were not significantly different from values in the volunteer group. These results suggest that a significant elevation of the hepatic artery RI may be a sensitive index of liver damage related to venoocclusive disease after bone marrow transplantation and an important distinguishing sonographic feature.