Long-term extra-corporeal assistance by continuous haemofiltration during fulminant hepatic failure
Denis, J.; Opolon, P.; Delorme, M.L.; Granger, A.; Darnis, F.
Gastroenterologie Clinique et Biologique 3(4): 337-347
1979
ISSN/ISBN: 0399-8320 PMID: 499696 Document Number: 140470
Continuous hemofiltration was carried out in 10 patients with fulminant hepatic failure and coma until persistent recovery of consciousness and normal EEG, death or decerebration. Etiology of fulminant hepatic failure was toxic in 2, due to cytomegalovirus in 1 and viral in 7. Hemofiltration started 30 .+-. 16 h (6-60 h) after the clinical onset of encephalopathy. When starting hemofiltration, 2 patients were in encephalopathy grade II, another one in grade III and the remaining 7 in complete coma grade IV. Hemofiltration duration averaged 92 .+-. 102 h (65-352 h), total ultrafiltration volume averaged 248 .+-. 286 l (7-968 l). Four patients died without awakening; in all these cases total liver cell destruction was found at post mortem histological examination. Another patient died after partial awakening (from encephalopathy grade IV to grade II-III) by septic shock on the 5th hemofiltration day. Five out of the 10 patients survived. Regain of consciousness occurred 2 h to 4 days after starting hemofiltration. In all cases normal EEG was obtained during hemofiltration only several days after clinical improvement. A women in 7th month of pregnancy delivered during hemofiltration a 1400 gm baby, well and presently alive; hepatic regeneration started on the 13th day, hemofiltration was interrupted on the 18th day. No biocompatibility problem occurred, except for a significant fall in platelet counts during the first 8 h of hemofiltration. Later, platelet count remained rather stable when dipyridamole was added to heparin. Transient bacteriemia was frequently observed, but septicemia occurred only in 2 patients. During hemofiltration, large amounts of amino acids and polypeptidic hormones (gastrin, insulin, renin) were removed. For these compounds, variations of plasma levels and amounts removed in ultrafiltrate seemed to better correlate with the extent of liver necrosis than with degree of encephalopathy.