The distribution of six HLA antigens in a population of 88 Portuguese chronic alcohol abusers with biopsyproven liver disease was compared to that in 66 Portuguese normal controls. Among the group of 88 alcohol abusers, the presence of HLA antigens Al, A9, A28 and Bw35 marked a significant 2.5-to %fold increased estimated risk for the development of alcohol-induced cirrhosis, while the presence of HLA B5 was found to be associated with a significantly decreased risk of alcohol-induced cirrhosis. Further, compared to controls, the estimated risk of any stage of alcohol-induced liver disease was significantly increased in alcoholic individuals with HLA A28 and Bw35, and the protective effect of HLA B5 was again observed. The findings of this study suggest that at least in a relatively homogeneous population group such as the Portuguese, the presence of HLA B5 may confer protection against alcohol-induced liver disease, including cirrhosis. The presence of HLA Bw35 and A28 appear to mark susceptibility to all histologic manifestations of alcohol-induced liver disease, while in addition to Bw35 and A28, A1 and A9 may mark increased risk for cirrhosis in particular.
Although some epidemiological studies have shown that the risk of developing alcoholic cirrhosis is related to the amount and duration of alcohol consumed (1, 2), data from Sorensen et al. (3) reinforce the hypothesis that there is not necessarily a simple relationship between duration of abuse and susceptibility to alcoholinduced liver disease. This prospective 13-year study provided evidence that a high daily alcohol ingestion rate alone did not necessarily lead to cirrhosis, and that alcohol may act only as a trigger mechanism in an individual predisposed to develop alcoholic liver disease. Consistent with such a hypothesis are data obtained in alcoholic families which show that a genetic influence appears to be a principal factor in the susceptibility of an individual to develop alcoholism (4). Thus, the puta-