• cirrhosis;
  • chronic aggressive hepatitis;
  • prednisone;
  • survival

ABSTRACT— A liver biopsy material comprising 477 biopsies from 477 patients included in a prospective, unblinded, randomized trial of treatment of cirrhosis with prednisone has been re-evaluated using new and more restrictive histological criteria for the diagnosis of cirrhosis and chronic aggressive hepatitis (CAH). The material was divided according to the likelihood of cirrhosis being present: (A) cirrhosis (287 patients), (B) probably cirrhosis (101 patients) and (C) compatible with but not diagnostic for cirrhosis (89 patients). Each group was further divided according to the presence (I) or absence (II) of CAH. A total of 98 patients fulfilled the histological criteria for CAH. The effect of prednisone concerning survival was evaluated in each group. In the total group of patients with CAH a significant beneficial effect of prednisone was found (p=0.04). Among these patients the subgroup with cirrhosis in addition (group I A) also showed a significant effect of prednisone, while groups I B and I C only had a trend towards a beneficial effect (p=0.44 and p=0.36, respectively). In patients without CAH in the biopsy (Group II, A+B+C), no effect of prednisone was seen although a trend towards a harmful effect was found in patients with cirrhosis (Group II A). Control patients with CAH in all three subgroups had an insignificantly shorter survival than patients without CAH. All the CAH groups significantly more often included female patients with no history of alcoholism and a lower frequency of spider naevi. In addition, the CAH groups were more active as judged by biochemical and histological variables. It was further disclosed that the presence of large piecemeal necroses indicated a favourable effect of prednisone treatment, while alcoholism, ascites and male sex acted as indicators for an unfavourable treatment effect.