Treatment of hepatic encephalopathy.

The mortality of coma complicating acute hepatic necrosis remains high at about 80% and recently exchange transfusion has shown some promising results. This line of treatment is based on the assumption that improvement depends on the removal of non-dialysable toxic substances from the circulation although some authors have suggested that the benefit might be due to the addition of deficient factors such as antibody (Morris, Gocke & Sardi, 1967). In fact the first successful treatment of hepatic coma by exchange transfusion was more than 10 years ago (Lee & Tink, 1958) but there has been a resurgence in interest in this therapy following the complete recovery of six patients out of eleven treated by Trey, Burns & Saunders (1966). Other workers have failed to achieve such good results (Jones et al., 1967) but the relative simplicity of the method and the high mortality of coma in acute necrosis would seem to indicate its use in patients not responding to conventional measures of protein restriction and gut sterilization. The majority of reports indicate that exchange of at least one blood-volume must be carried out for appreciable clinical improvement to occur and this must be repeated on subsequent days. Whenever possible fresh citrated blood should be used but exchange transfusion may be complicated by coagulation defects due to both the hepatitis and the actual blood exchange itself (Samson et al., 1967). The same authors have discussed the importance of hypoglycaemia which may also extend the hazards of exchange transfusion in hepatic necrosis. We have had the opportunity of treating five patients with exchange transfusion (Table 1). Three of these patients had coma complicating acute hepatic necrosis and two had very severe portosystemic encephalopathy complicating cirrhosis. In retrospect the volumes of blood exchanged may have been too small and the procedure performed too late. The technical difficulties of obtaining fresh matched blood may be avoided by plasma-exchange which has been exploited by Sabin & Merritt (1968). They removed the patient's plasma and returned the patient's packed red cells and fresh frozen donor plasma. Encouraging results were obtained in these patients with cirrhosis and refractory coma. Further methods of treatment are at present being examined but, as yet, it is too early to assess their effectiveness. Burnell et al. (1967) have performed cross circulation exchanges with some success, but with this method there are ethical difficulties in the choice of suitable partners and practical problems with donor reactions. Extra-corporeal circulation of patients' blood through an isolated perfused cadaveric liver (Sen et al., 1966) and heterologous perfusion using pig's liver (Eisman, Liem & Raffucci, 1965) may produce some neurological improvement, but such measures require extensive facilities and have not been used in this country.