Background: The hypothesis evaluated in the study was the existence of correlation between blood flow direction and velocity in portal collateral circulation and the portal vein,the size of esophageal varices and variceal bleeding. The aim of the study was to confirm the afore- mentioned correlation and to determine whether the results would allow to optimize indications for preventive obliteration of esophageal varices in the group of patients characterized by high risk of hemorrhages. Material/Methods: From 1999 to 2002,118 patients (group I)with the history of variceal hemorrhages ,but not treated yet, and 80 patients (group II)with no history of hemorrhagic events despite the signs of portal hypertension and the presence of esophageal varices were included in the analysis.The characteristics evaluated in both groups included size of the esophageal varices ,direction and volume (mean velocity)of blood flow in the portal vein –if it was patent –or in the formed collateral circulation. The appearance and size of the esophageal varices were assessed by endoscopy. Portal hemodynamics disorders were evaluated by ultrasonography using a Doppler probe for blood flow mea- surements. Patients with the history of hemorrhages (group I) were examined once. The obtained results were used to determine the hemodynamic disorders characteristic of variceal bleeding and the corresponding size of varices. Group II was subjected to prospective evaluations. The baseline results constituted the reference point for further periodic check-ups monitoring the changes taking place in the portal system. The follow-up of the patients involved endoscopy and sonography performed at 2 –3 month intervals to evaluate the progression of the changes (according to the above criteria), with special attention paid to the onset of bleeding. The mean observation time in the prospectively evaluated group II was 24 months and ranged from 3 to 36 months.Results:In all the patients from group I, large and medium-sized varices were found. Patients with hepatopetal blood flow with mean flow velocity of 12 cm/s predominated. n most patients, collateral circulation with hepatofugal flow direction had been formed, which in 25 cases was effected through the left gastric vein. Small varices were predominant on base- line endoscopic assessment of group II. Patients with large and medium-sized varices constituted 28%of the group. During the analyzed follow-up period, the number of patients large and medium-sized varices increased. On final assessment they accounted for 52% of the whole group. Conversely, the number of patients with small varices decreased. Baseline ultrasonographic assessment (group II) demonstrated the predominance of patients with hepatopetal blood flow of 16 cm/s mean velocity. On final assessment, the number of such patients decreased along with the reduction of mean flow velocity in the portal vein. On the other hand, the number of patients with portal venous stasis or reversed flow direction increased during the analyzed period On baseline assessment, patients with hepatopetal or bidirectional portal flow in collateral circulation were found to predominate, whereas final assessment demonstrated the increase of the number of patients with hepatofugal collateral circulation formed during the observation period. The number of patients with hepatopetal blood flow in collateral portal circulation decreased. Variceal bleeding was noted in 19 (25%) patients from group II.Statistical analysis confirmed the existence of statistically significant differences at p

Keywords: portal hypertension, portal hemodynamics disorders, Doppler monitoring of portal flow and collateral circulation