Liver transplantation is currently the only effective therapy for patients with fulminant hepatic failure. The availability of an artificial liver could bridge these patients through the relatively brief crisis period and allow their own livers to regenerate, providing a more favorable outcome and sparing the trauma and expense of transplant. We have developed a device consisting of a highly differentiated human liver cell line cultured in a hollow fiber cartridge. This device is capable of supporting dogs with acetaminophen-induced fulminant hepatic failure for a period long enough for their own livers to resume function. Even though liver function tests such as albumin and prothrombin time became extremely abnormal during the course of the experiment, the dogs did not become encephalopathic. Two of the three treated animals recovered sufficient liver function after 42 to 48 hr of treatment that they could be disconnected from the device, and they survived the experiment. Histological results and serum ALT levels suggest that the device affected the course of the disease in two animals, allowing recovery of hepatocytes that would otherwise have lysed. In the third animal, regenerative nodules demonstrated that, even in the presence of severe liver injury, the device was capable of supporting total liver function.

We have established an improved model of fulminant hepatic failure in dogs. Buthionine sulfoximine is used to inactivate glutathione synthesis, and small increments of acetaminophen are given intravenously to maintain the plasma level at approximately 200 micrograms/ml for 20 hr. This regimen produces severe liver injury along with many of the features seen in humans with acetaminophen poisoning. The first sign of impending liver failure is hypoglycemia. This occurs about 15 hr into the experiment and requires treatment with a continuous infusion of glucose. Between 15 and 20 hr, serum ALT activity begins to rise, indicating the onset of liver necrosis. Over the following 15 to 20 hr ALT activity continues to rise and is accompanied by an increase in bilirubin, a prolongation of the prothrombin time and the development of fetor hepaticus. Thirty to 48 hr after the initial acetaminophen dose, the animals begin to exhibit symptoms of encephalopathy and progress from lethargy to the inability to maintain posture and then coma, seizures and death. Liver biopsy specimens obtained at several stages throughout the study showed progressive necrosis, ultimately resulting in the complete destruction of zones 2 and 3.

We have used an anhepatic dog model to demonstrate the efficacy of a bioartificial liver assist device. Six dogs underwent total hepatectomy. Three received only medical care (controls) while the remainder were connected to an extracorporeal liver assist device (ELAD). The control dogs failed to regain consciousness after anesthesia although all lived 4–5 h postoperativlely. Plasma ammonia concentration increased by an average of 2–50 μmol/L between the end of surgery and the demise of the animals. The treated dogs lived 3–12.5 h, and 2 of them required repeated doses of thiamylal sodium to maintain sedation. Plasma ammonia concentration was unchanged after connection to the ELAD except in the longest survivor, whose ammonia began to rise after 8 h on the ELAD. The short survival in the other 2 treated dogs was the result of uncontrolled intraabdominal bleeding. This device is capable of replacing the metabolic function of the liver, and might provide hepatic support in patients awaiting transplantation or in fulrninant hepatic failure.

Patients with chronic epigastric to right upper quadrant pain are often considered to have gallbladder of sphincter of Oddi dysfunction, but standard tests are nondiagnostic. In 62 consecutive patients with this compliant undergoing antroduodenal manometry, we correlated a change on duodenal motility with spasm of the ampulla of Vater/duodenal wall. This distinctive motility pattern occurred and was analyzed in 35% of patients. It is characterized by increased duodenal wall tone with phasic contractions of 19-22 or 41-44 contractions/min or by phasic activity alone. The subjects with spasm also underwent cholecystokinin cholescintigraphy, and 50% showed either significantly delayed gallbladder emptying of hilum to small intestine emptying, or both. The disorder appears to be secondary to a loss of neural inhibitory control and a dysfunctional small-bowel pacemaker. Antroduodenal manometry is an essential diagnostic procedure that complements sphincter of Oddi manometry in evaluation of unexplained right upper quadrant pain.

No abstract