Abstract

The accuracy of non-invasive radionuclide angiography in detecting portal vein occlusion was assessed in 61 patients--10 with portal vein occlusion confirmed by conventional portography, 25 with chronic liver disease and a patent portal vein (mild = 12, severe = 13), and 26 with normal liver function, who served as controls. The median percentage portal venous flow for the portal vein occlusion group was 8% (range 1-30) (consistent with negligible flow) compared with 78% (52-87) for control subjects (p < 0.005) and 68% (61-80) and 49% (23-59) respectively for patients with mild and severe liver disease (p < 0.001 and p < 0.005). At a portal venous inflow of < 20%, the procedure had a specificity of 100% and sensitivity of 90% in diagnosing portal vein occlusion. Non-invasive radionuclide angiography provides a safe and accurate screening method for evaluating portal vein patency or occlusion in the investigation of portal hypertension or before liver transplantation.