Between Scylla and Charybdis.

The ethical dilemmas of treating fulminant (or subfulminant) hepatic failure with liver replacement have been apparent since the earliest trials of this procedure. Acute hepatic failure was the indication for orthotopic liver transplantation in only I of our first 237 recipients I and accounted for only 8 of the next 300.2 Our reluctance to proceed with more cases was that unlike candidates whose livers are chronically diseased, those with acute or subacute hepatic failure may spontaneously recover. The vast majority of patients who pass through the acute crisis are able to regenerate their own livers and return to a normal life expectancy, absent the burden of chronic immunosuppression. The incidence of a favorable outcome is strongly influenced by the cause of the hepatic failure, highest with hepatitis A (67%) and acetaminophen overdose (53%). The lowest recovery is with nonA. non-B, and non-C hepatitis and after poisoning with halothane or other hepatotoxic drugs. No matter what the etiology, the prognosis for recovery. short ofliver transplantation, currently is better than the 5% to 10% frequently cited from the literature of the I 970s. which discussed the option of conservative management. The improvement requires specialized care teams armed with protocols designed to prevent brain injury. Hopital Paul Brousse. from which the report by Bismuth et al. 3 comes, has such services. However. survival is not as good as that after liver transplantation. at least within the context of the 1to 5-year follow-up. Consequently. liver surgeons currently are recapitulating the era preceding 1962 in kidney transplantation. when artificial kidney support was not available widely-if at all-for the treatment of acute renal failure. Then with the kidney. as currently with the liver. the first objective Was to differentiate those patients who would recover from those who were doomed without draconian interVention. In a further analogy to the history of acute renal tubular necrosis and kidney transplantation. the inherent reversibility of fulminant hepatic failure has driven research that eventuallv mav reduce the need for liver transplantation. Whil~ the' body's hepatic based metabolic machinery grinds to a halt, all of the organ systems are threatened. but the most dreaded insult is to the central nervous system. The minimal extra function provided by extraor intracorporeal hepatic allografts or xenografts, and more recently by hepatocytes injected intravenously or lining the capillary tubes of "artificial livers," has been credited with amelioration of brain deterioration and other complications of liver failure (including renal dysfunction). Buying limited time in this way with a borrowed animal or human liver (or hepatocytes) could be a hollow gesture, but since 1975, a family of hepatic growth factors has been discovered which may speed the regeneration of the devastated native liver.4 These molecules have not yet been successfully exploited clinically, partially because of the inability to maintain life support long enough for their effect to be evaluated. The key to success appears to be artificial liver support combined with iatrogenic promotion of hepatic regeneration. When these technologies are developed. emergency liver transplantation for fulminant hepatic failure will become largely obsolete. Until then and even afterward. the choice of aggressive medical versus transplant therapy will remain. With either decision. the best results always will be with the patients who are the least ill, the ones who historically have been the most likely to recover with no specific treatment at all. Unfortunately. in such cases. a highly visible and professionally damaging error will lie in wait for the physician or surgeon who delays operation until a patient who might have been saved by transplantation has lost that chance. In contrast. the invisible error of operating on patients who would have recovered spontaneously is seldom discussed and rarely is provable. The frequency of the latter mistake and its lifetime implications for the recipient is certain to increase the earlier the timing of transplantation is decided. The experience of Bismuth et al. 3 was acquired in this treacherous landscape. between Scylla and Charybdis. Of the 139 patients entered into their study. 23 were withdrawn-22 by death after a mean of 1.3 days and the other by spontaneous recovery. The 23 candidates who did not make it to the operating room cannot be con-