Bleeding from esophageal varices in cirrhosis of the liver. Hemodynamic and radiological criteria for the selection of potential bleeders through hepatic and umbilicoportal catheterization studies.

Combined hepatic and umbilicoportal catheterization was performed in 38 compensated cirrhotics. Portohepatography with opacification of the coronary vein was obtained in all cases. The free portal venous pressure (FPVP) and the wedged (WHVP) and free (FHVP) hepatic venous pressures were recorded. The portohepatic gradient (FPVP-FHVP) was used as an index of portal hypertension. The coronary vein was separately re-evaluated for varices and graded as 1+ to 4+.Eighteen patients had varices graded as 3+ or 4+ (Group A) and all had a portohepatic gradient of 10 mm. Hg or more. The other 20 cirrhotics (Group B) had varices graded as 1+ or 2+ and 15 had a portohepatic gradient of less than 10 mm. Hg. The difference between gradients of Group A and Group B was highly significant.Of the 38 cirrhotics studied, eight had bled from varices and all are included in Group A. There is no significant difference between the gradients of both bleeders and non-bleeders of Group A.There is a significant correlation between the presence of large varices with a portohepatic gradient of 10 mm. Hg or more and a high risk of variceal bleeding. The radiological and hemodynamic data obtained by combined hepatic and umbilicoportal catheterization in cirrhosis of the liver can be of significant help in the selection of potential bleeders.