Abstract

Hepatitis C virus antibody titres (anti-HCV) were measured in serum from 122 patients with autoimmune liver disease (96 with primary biliary cirrhosis and 26 with autoimmune chronic active hepatitis using three generations of enzyme immunoassay (EIA): first generation--Ortho, EIA1; second generation--Abbott, EIA2; and third generation--Murex, EIA3. Anti-HCV was below the positive cut-off level in all 26 autoimmune chronic active hepatitis patients for all tests, while seropositivity values in primary biliary cirrhosis were 31% (EIA1), 14% (EIA2), and 0% (EIA3). In primary biliary cirrhosis, anti-HCV values as measured by all three tests correlated positively with serum IgG concentrations, serum storage time, and a number of other indices of hepatic dysfunction. Multiple regression analysis showed that anti-HCV values were independently affected by both serum IgG and the length of storage time, although the magnitude of the effects varied between tests. When all three multiple regression models were applied to an extreme clinical example, however, EIA3 was least likely to give a false-positive result. The difference in test performance was emphasised further by examination of anti-HCV values in nine primary biliary cirrhosis patients (confirmed negative by recombinant immunoblot assay 2) in whom serial samples were tested (seven to 14 per patient, stored for one to 138 months). Apparent anti-HCV values (EIA1 and EIA2) increased with increasing serum storage time, but were unchanged when measured by EIA3. A similar pattern was evident in a limited study of five autoimmune chronic active hepatitis patients. As the second generation EIA is in widespread use and confirmatory testing is not always available, the effect of serum storage in addition to hyperglobulinaemia should be considered in the interpretation of positive results in auto immune and in other types of chronic liver disease.