A prospective randomized clinical trial of peripheral amino acid-glucose supplementation in acute alcoholic hepatitis.

Twenty-eight patients with biopsy or clinical acute alcoholic hepatitis were prospectively randomized to 21 days of conventional therapy (14 control patients) or, in addition, to 2 L/day of a peripheral iv infusion of a 900 mosmol amino acid-glucose solution (14 patients). Cirrhosis was present in 64% of controls and in 54% of infused patients. There were three deaths in controls and one in the infused group. There were no significant intergroup differences in mortality, clinical findings, liver tests, or functional mass by the galactose elimination capacity either at entry or after completion of the study. In controls, there was intragroup improvement in serum bilirubin (p = 0.033) and AST (p = 0.008) but not in other "liver tests" or in functional hepatic mass by galactose elimination capacity. In infused patients there was improvement in bilirubin (p = 0.001), AST (p = 0.008), and serum albumin (p = 0.016). Improvement in functional mass by galactose elimination capacity was significant at the p = 0.052 level. Hyaline was initially present in six of eight pairs of biopsies in both groups. After treatment, five of six pairs in controls but only one of six pairs in the infused group still had hyaline (p = 0.03). This latter finding, if confirmed in larger groups, may be of considerable clinical importance. It is suggested that 3- to 4-wk trials of such protocols may require a longer duration of follow-up in greater numbers of patients to detect important advantages of amino acid-glucose infusions which are only implicit in these findings.