Following administration of equivalent oral doses (30mg) of either prednisone or prednisolone to 5 patients with chronic active liver disease who had failed to respond to therapy, 5 patients with chronic active liver disease in remission induced by prednisone, and 7 healthy volunteers, corticosteroid concentrations were measured in both serum and urine by radioimmunoassay. Prednisone and prednisolone concentrations in the urine were very similar in all groups, regardless of the drug given. After either treatment, the peak serum concentration and area under the prednisolone serum concentration-time curve were 4 to 5 times those of prednisone. Slight differences among the 3 groups studied were seen in prednisone and prednisolone pharmacokinetics, but none was significant. It is concluded that the use of prednisone instead of prednisolone to treat chronic active liver disease can not be implicated as a cause of treatment failure. Indeed, this study suggests that either medication is effective, and this is supported by serum concentrations which suggest a rapid interconversion equilibrium between the 2 corticosteroids.

To determine the effect of impaired liver function on conversion of prednisone to prednisolone, and to investigate the relationship of this to responses to treatment with prednisone, we measured serum prednisone and prednisolone by radioimmunoassay after 10 mg of prednisone was given by vein to 10 healthy volunteers, 6 untreated patients with severe chronic active liver disease (CALD), 10 patients with prednisone-induced remission of CALD, and 3 patients with CALD deteriorating despite treatment with prednisone. Prednisone disappearance was comparable in all groups and substantial values for serum prednisolone appeared in all groups within 0.3 hr. Minor differences between the groups included lower than normal serum prednisolone in severe CALD and treatment failure; a higher percentage of nonprotein bound prednisolone in such patients; and an association between earlier treatment with prednisone and an increased disappearance rate of prednisolone. We conclude that no major defect of prednisone metabolism occurs in CALD and that failure of this condition to respond to therapy with prednisone is caused by other factors.

Cholecystokinin-pancreozymin (CCK-PZ) was measured by a bioassay method based on pancreatic enzyme output in response to essential amino acid (EAA) perfusion of the small intestine. EAA perfusions of the duodenum, proximal jejunum, and distal jejunum evoked comparable trypsin outputs which exceeded basal secretion. In contrast, ileal EAA perfusions yielded no increase over basal secretion. Therefore, CCK-PZ is secreted throughout the human duodenum and jejunum, but not from the ileum. Simultaneous perfusions of EAA at the duodenal and both jejunal sites evoked less trypsin output than at any one of these sites and was accompanied by increases of plasma glucagon and alpha-amino nitrogen. Similar inhibition of the effect of endogenous CCK-PZ on the pancreas was also caused by intravenous protein hydrolysate and also accompanied by increases of plasma glucagon. Glucagon itself, given by vein, also inhibited the stimulatory effect of endogenous CCK-PZ on pancreatic enzyme output. It is proposed that products of protein digestion in the small intestine stimulate pancreatic enzyme secretion by liberating CCK-PZ and, after absorption, decrease pancreatic secretion probably by causing release of glucagon.

Impaired secretion of cholecystokinin-pancreozymin was identified in patients with npntrgp_ ieal.spr1Je, and was attributed to mucosal damage of the small intestine.Outputs of pancreatic enzymes and gallbladder contents in response to essential amino acids in the duodenum were reduced in sprue, although gallbladder and pancreatic function were normal after intravenous cholecystokinin-pancreozymin.The effects of impaired cholecystokinin-pancreozymin secretion in sprue on the components of fat digestion were tested after administering two test meals.In early postprandial periods, pancreatic lipase and bile acid concentrations were diminished, and these abnormalities were later intensified by excessive dilution of the meal.Impairment of fat digestion was greater after the feeding of the second meal, correlated best with decreases in intraluminal bile acid concentrations and micellar lipid, rather than lipase deficiency, and may contribute to steatorrhea in some patients with sprue.The authors 1 hypothesize that mucosal damage in nontropical sprue reduces secretion of the gastrointestinal hormone cholecystokinin-pancreozymin ( CCK-PZ), and thus prejudices intraluminal fat digestion because secretory and motor functions of the target organs, the pancreas, and the gallbladder are impaired.In sprue, subnormal pancreatic function in response to test meals, which depend on endogenous secretion of CCK-PZ and secretin for their effects, is consistent with the hypothesis. 2 Furthermore, failure of normal bicarbonate secretion from the pancreas in response to acidification of the duodenum has been interpreted as reflecting impaired secretin release in sprue. 3Alternatively, it has been speculated that subnormal pancreatic secretion in sprue may be attributable to pan-

Histologic study of serial biopsy specimens in a prospective, controlled, double blind, randomized trial of treatment involving 63 patients with predefined clinical and biochemical criteria of severe chronic active liver disease revealed five different histologic patterns of hepatic injury on initial biopsy: chronic persistent hepatitis, chronic aggressive hepatitis, subacute hepatitis with bridging, subacute hepatitis with multilobular necrosis, and cirrhosis. The initial biopsies in the fatal cases revealed cirrhosis, subacute hepatitis with multilobular necrosis, or subacute hepatitis with bridging. Patients with subacute hepatitis with bridging and with multilobular necrosis more frequently revealed stigmata of viral hepatitis and more frequently developed cirrhosis than chronic aggressive hepatitis, regardless of treatment. The severe forms of hepatitis more frequently remitted to the histologic features of chronic persistent hepatitis in patients treated with prednisone and a combination of prednisone and azathio-prine than with azathioprine or placebo alone. Serial biopsy specimens frequently revealed changes in the histologic pattern before remission to chronic persistent hepatitis. Histologic progression to cirrhosis may occur by bouts of intralobular and multilobular necrosis as well as by piecemeal necrosis. Sampling error in needle biopsy of cirrhosis is so great that the condition either may be missed or may be impossible to accurately classify into portal and postnecrotic cirrhosis. Observations at autopsy always revealed postnecrotic cirrhosis characterized by extensive parenchymal loss and incomplete regeneration. We conclude that the evolution of any lesion is influenced by at least two factors: the original histologic pattern and the therapy applied.

We have examined in vitro some requirements for an artificial support system for use in hepatic failure. These experiments compared methods of removing from plasma (1) certain unbound substances implicated in the pathogenesis of hepatic coma, and (2) some model protein-bound substances of variable toxic potential known to be extracted by the liver and excreted in the bile. A Craig thin film countercurrent dialyzer was used, and dialysis rates were related to protein binding by determining concentrations of unbound species by ultrafiltration. Compounds not bound to protein (methionine, glutamine, butyrate, ammonia, and urea) were readily dialyzed (>90%). Dialysis of protein-bound anions (unconjugated bilirubin, chenodeoxycholate, sulfobromophthalein, and methyl orange) was negligible and was enhanced little by addition of a competitive anion (salicylate) or an acceptor (plasma or a water-soluble polycation) to the dialysate. The major rate-limiting factor was the low concentration of unbound species, rather than the membrane. When plasma was passed over columns containing anion exchange or uncharged resins, protein-bound material was removed efficiently; anion exchange resins were superior to uncharged resins. Both hemoperfusion through anion exchange columns and hemodialysis may therefore be essential excretory components of an artificial liver.