Orthograde whole gut irrigation with mannite versus paromomycine plus lactulose as prophylaxis of hepatic encephalopathy in patients with cirrhosis and upper gastrointestinal bleeding: Results of a controlled randomized trial

Background/Aims: In patients with liver cirrhosis and upper gastrointestina l bleeding development of hepatic encephalopathy is a major problem. The ai m of the present study was to evaluate the efficacy of the mannite lavage i n a controlled randomized trial with respect to the Child-Pugh classificati on. Methodology: After initial gastroscopy (+/- sclerotherapy) 39 patients with cirrhosis (18 F, 21 M; age: 57.5+/-11.9yr; Child A: 6, Child B: 16, Child C: 17) and upper gastrointestinal-bleeding were classified according to the Child-Pugh-criteria (A,B,C) and randomized in 2 groups (A,B) for each Chil d-Pugh level. Patients in group A (n=18) were initially treated with 2000mL mannite solution (10%) during the first 2 hours using a naso-gastric tube. Treatment was continued using 2000mL mannite solution (10%) per day until no rectal blood could be observed. Patients in group B (n=21) were treated with paromomycine ter in die (1g tid) and lactulose (10mL tid). There were no statistical differences between both groups concerning age, sex, Child-P ugh-scores, severity or source of bleeding, initial hemoglobin-levels, numb er of given blood-transfusions or number of patients with sclerotherapy. Results: Patients in group A were treated with a total of 3325+/-1897mL man nite solution. The application was well tolerated. In addition, kinetics of serum creatinine, potassium and sodium levels did not show any significant changes. No significant differences between both groups could be shown wit h respect to clinical criteria of encephalopathy according to O'Grady and t he length of intensive care unit treatment. Moreover, kinetic of ammonia-le vels showed a pronounced decrease (P=0.05) on day 2 versus day 1 in group A (110.0+/-24.2 vs. 156.4+/-98mg/dL) as compared to group B (210.0+/-52.7 vs . 162.0+/-45mg/dL). In group A, 6 patients (33.3%) died during the study as compared to 3 patients (14.3%) in group B (P>0.05). The lethality rate was strongly associated with the larger proportion of Child-C-patients in grou p A. Conclusions: The data indicate that whole gut irrigation with mannite is eq ually efficacious as compared to standard treatment for prophylaxis of hepa tic encephalopathy after upper gastrointestinal bleeding in liver cirrhosis . In contrast to previously published controlled studies, no impact of the lavage on the mortality rate or duration of intensive care unit treatment c ould be shown: With respect to the lower costs for the mannite solution as compared to paromomycine and lactulose (ROTE LISTE, Germany), the mannite l avage should be recommended for the prophylaxis of hepatic encephalopathy a fter upper gastrointestinal bleeding in patients with liver cirrhosis.