Accuracy and significance of computed tomographic scan assessment of hepatic volume in patients undergoing liver transplantation

Background. A small liver volume is considered to be a poor prognostic factor in cirrhosis, often indicative of advanced liver disease. Radiologic assessment of liver volume before liver transplant is routinely performed in many transplant centers. We sought to assess the accuracy and significance of computed tomographic (CT) scanning in hepatic volumetric analysis by correlating CT-derived estimation of liver volume with that of corresponding liver explants. Methods. A chart review of all patients aged 17 years or older undergoing liver transplant at Mount Sinai Medical Center between 1989 and 1995 was performed. Each patient underwent conventional CT scanning with measurement of liver volume (CTLV). Recipient liver volume (RLV) was defined as weight of liver explant after all attached ligaments, portal structures, and gallbladder were dissected free. Expected liver volume was calculated pretransplant based on age, gender, height, and weight. Patients were categorized into three groups based on etiology of liver disease: (1) hepatocellular (e.g., viral hepatitis, alcohol-related), (2) cholestatic (e.g., primary biliary cirrhosis), and (3) cryptogenic. The ratio of CTLV to RLV was used as a measure of the accuracy of CT volumetric analysis. Results. A total of 579 patients was studied (group 1=376, group 2=139, group 3=64). All three groups were statistically similar with regard to age, prothrombin time and total bilirubin. Median CT liver volume was 1308 ml (range: 338-3847), 1651 ml (range: 641-3861), and 1210 ml (range: 348-2575) in groups 1-3, respectively; median ratio of CTLV to RLV was 1.02 (range: 0.50-2.31), 1.05 (range: 0.52-2.22), and 1.05 (range: 0.50-1.56) for groups 1-3, respectively. When RLV was small, it tended to be overestimated by CTLV. In contrast, when RLV was large, it was often underestimated. Clinical features such as history of esophageal variceal bleed, encephalopathy or ascites, and laboratory data did not influence accuracy of CT volumetric analysis. Conclusions. CT-derived estimation of liver volume appears to correlate closely with actual weight of liver explant regardless of the etiology of chronic liver disease. With extremes in CT volumetric analysis, actual liver volume tends to be under- or overestimated. For patients with end-stage liver disease, both CT-derived and actual liver volume are greater in cholestatic than in hepatocellular disorders.