Prednisone and chronic liver disease. II. Clinical versus morphological criteria for selection of patients for prednisone treatment.

A prospective, unblinded, randomized trial of treatment of cirrhosis and chronic aggressive hepatitis (CAH) with prednisone included 484 patients who were divided using both the clinical criteria: non-alcoholic females without ascites, and the morphological criterion: presence of a liver biopsy showing CAH. A total of 164 patients (34%) satisfied the three clinical criteria and biopsies from 95 patients (20%) showed CAH. By a cross-tabulation method between the two sets of criteria, four groups were formed. Evaluation of the treatment effect by means of relative death rate showed that the clinical criteria identified both a group of patients with a beneficial effect of prednisone (p less than 0.05) and a group with a harmful effect of the treatment (p less than 0.05). Patients with CAH also had a significant beneficial treatment effect, but patients without CAH showed only an insignificantly harmful effect of treatment. Patients fulfilling both sets of criteria had a beneficial effect of prednisone (p less than 0.05) and patients lacking both sets had a significantly harmful treatment effect. A significantly better survival was found in patients satisfying the clinical criteria without CAH in the biopsy than in patients with CAH and not fulfilling the clinical criteria (p = 0.03). In both the groups of patients with disagreement between the two sets of criteria, a trend towards a beneficial effect of prednisone was found. It is concluded that the simple clinical criteria in comparison with the histological criteria identify a larger group of patients with a beneficial effect of treatment.