Endoscopic sclerotherapy is widely employed for esophageal variceal hemorrhage. However it has side effects and can aggravate portal hypertension by suppression of portosystemic shunt. The purpose of the present investigation was to study the effect of variceal thrombosis on hepatic venous pressure gradient and azygos blood flow. Eight alcoholic cirrhotic patients with a first variceal hemorrhage were included. According to Child Pugh's classification, 4 patients were group A, 2 group B and 2 group C. At each session 40 to 60 ml of 1 p. 100 polidocanol were injected into the varices. A hemodynamic study was performed in each patient before and about one week after variceal obliteration (mean 3.3 procedures). Mean value of hepatic venous pressure gradient was 16.6 +/- 5.5 mm Hg and 17.0 +/- 3.8, respectively, before sclerotherapy and after eradication of varices; azygos blood flow 663 +/- 506 ml/mn before and 682 +/- 522 after; cardiac, output was 6.5 +/- 0.7 ml/min before and 6.5 +/- 0.8 after. None of these differences were significant. These results suggest that endoscopic sclerotherapy using polidocanol does not change hepatic venous pressure gradient and azygos blood flow, and does not lower blood flow through the gastroesophageal collaterals draining into the azygos vein. This is consistent with the hypothesis that thrombosis remains localized.

Concerned by mortality due to recurrent bleeding and associated risk factors in cirrhotic patients, we attempted to compare the efficiency of perivascular sclerotherapy using quinine urea with that of intravascular polidocanol and to determine the predictive factors of clinical outcome. Of 74 patients admitted for bleeding esophageal varices, 31 were treated with perivascular sclerotherapy (group I), and 43 with intravascular sclerotherapy (group II). Three months later, only 63 p. 100 of patients in group I had not rebled, compared to 90 p. 100 of patients in group II (p less than 0.001). After 6 months, no significative difference was found between the two groups. The one year survival rate was 44 p. 100 in group I and 79 p. 100 in group II (p less than 0.002). Encephalopathy and ascites were found to be predictive factors of mortality, whereas neither clinical or biologic factors were found to be predictive for recurrent bleeding. These results suggest that in esophageal sclerotherapy, the intravascular route with polidocanol provides earlier results and a better one year survival rate, compared with the perivascular route. However, survival rates depend on encephalopathy and ascites criteria.

Concerned by mortality due to recurrent bleeding and associated risk factors in cirrhotic patients, we attempted to compare the efficiency of perivascular sclerotherapy using quinine urea with that of intravascular polidocanol and to determine the predictive factors of clinical outcome. Of 74 patients admitted for bleeding esophageal varices, 31 were treated with perivascular sclerotherapy (group I), and 43 with intravascular sclerotherapy (group II). Three months later, only 63 p. 100 of patients in group I had not rebled, compared to 90 p. 100 of patients in group II (p less than 0.001). After 6 months, no significative difference was found between the two groups. The one year survival rate was 44 p. 100 in group I and 79 p. 100 in group II (p less than 0.002). Encephalopathy and ascites were found to be predictive factors of mortality, whereas neither clinical or biologic factors were found to be predictive for recurrent bleeding. These results suggest that in esophageal sclerotherapy, the intravascular route with polidocanol provides earlier results and a better one year survival rate, compared with the perivascular route. However, survival rates depend on encephalopathy and ascites criteria.

The purpose of this work was to study postoperative mortality and morbidity with respect to preoperative prognostic factors in 67 patients with alcoholic or posthepatitis cirrhosis. Surgical procedures involved the biliary tract (n = 20), stomach (n = 16), colon or rectum (n = 12), and hernia (n = 7). Thirteen preoperative clinical and biological variables were subjected to mono- and multivariate statistical analysis. The mortality rate was 23 p. 100. There was no statistical difference between the three main surgical procedures. No patients died after herniorrhaphy. The rate of morbidity was 37 p. 100. The most common complications were sepsis, organ failure, and ascites. Three preoperative variables were found to be different between survivors and non survivors: ascites, prothrombin time and the Child-Pugh score. Multidimensional analysis demonstrated that the only variable to have an independent unfavorable prognostic value was albuminemia. These results suggest that postoperative mortality following extrahepatic abdominal surgery in cirrhotic patients is: 1) especially high after digestive procedures, 2) increased by ascites, low prothrombin time and high Child-Pugh score. Only hypoalbuminemia had a significant independent explanatory value regarding prognosis.