[Diagnostic value of portal duplex ultrasound in liver cirrhosis]

In a prospective study we investigated if the endoscopical evaluation of the bleeding risk of esophageal varices in cirrhotics could be improved by additional duplex sonography of the portal vein.

The trial involved 41 patients with endoscopically diagnosed esophageal varices (27 male, 14 female) who were followed up over a period of 30 months (mean of 13 +/- 3). According to the Child-Pugh-Turcotte -classification, 14 patients were classified as grade A, another 14 as grade B, and 13 as grade C. The cause of the cirrhosis included virus infection (n = 14), alcohol (n = 17) and miscellaneous disorders (n = 9, primary biliary = 3, autoimmune = 2, idiopathic = 4).

During the observation time, 17 patients who developed an acute esophageal variceal hemorrhage were put in the bleeding group (mean portal flow velocity Vm = 9.29 +/- 3.31 cm/s, mean flow volume Fm = 371 +/- 173 ml/min), while the remaining 24 patients formed the non-bleeding group (Vm = 13.29 +/- 5.12 cm/s, FV = 500 +/- 200 ml/min). The bleeding group had significantly lower mean portal flow velocities (p < 0.017) and mean flow volumes (p < 0.05) than the non-bleeding group. By adopting cut-off values of 12 cm/s for Vm and 420 ml/min for FV we obtained a diagnostic sensitivity for predicting esophageal variceal hemorrhage of 0.88 and 0.65 respectively.

Portal duplex sonography may improve the evaluation of bleeding risk in patients wit endoscopically diagnosed esophageal varices and could influence the decision for prophylactic treatment.