BACKGROUND: Antimitochondrial antibodies (AMA) can occur in autoimmune hepatitis, but their durability and prognostic significance are uncertain. OBJECTIVES: Determine the frequency, behavior, and prognostic implications of AMA in type 1 autoimmune hepatitis. METHODS: One hundred thirty patients were tested for AMA by the same enzyme-linked immunosorbent assay at accession and during 123+/-9 months of observation (mean, 6+/-0.2 determinations/patient). Findings were correlated with clinical and histologic features and treatment outcomes. RESULTS: Twenty-four patients (18%) had AMA, including 17 who had the antibodies at accession (71%) and 7 who developed them during follow-up (29%). AMA persisted in 13 patients (54%) who were tested on 5+/-1 occasions during 141+/-33 months. Cholestatic histologic features occurred as commonly in patients with and without AMA at presentation (18% vs. 10%, P=0.4). AMA disappeared in 4 patients, and they developed in 7 patients after 34+/-7 months. Late expression of antibodies was not associated with a higher frequency of cholestatic histologic changes than continued seronegativity (14% vs. 9%, P=0.5). Remission (83% vs. 76%, P=0.4) and treatment failure (8% vs. 11%, P>0.9) occurred as commonly in both groups. CONCLUSIONS: AMA can appear and disappear. They do not identify a subgroup that requires different treatment or that evolves quickly into a cholestatic syndrome.
BACKGROUND: Antimitochondrial antibodies (AMA) can occur in autoimmune hepatitis, but their durability and prognostic significance are uncertain. OBJECTIVES: Determine the frequency, behavior, and prognostic implications of AMA in type 1 autoimmune hepatitis. METHODS: One hundred thirty patients were tested for AMA by the same enzyme-linked immunosorbent assay at accession and during 123+/-9 months of observation (mean, 6+/-0.2 determinations/patient). Findings were correlated with clinical and histologic features and treatment outcomes. RESULTS: Twenty-four patients (18%) had AMA, including 17 who had the antibodies at accession (71%) and 7 who developed them during follow-up (29%). AMA persisted in 13 patients (54%) who were tested on 5+/-1 occasions during 141+/-33 months. Cholestatic histologic features occurred as commonly in patients with and without AMA at presentation (18% vs. 10%, P=0.4). AMA disappeared in 4 patients, and they developed in 7 patients after 34+/-7 months. Late expression of antibodies was not associated with a higher frequency of cholestatic histologic changes than continued seronegativity (14% vs. 9%, P=0.5). Remission (83% vs. 76%, P=0.4) and treatment failure (8% vs. 11%, P>0.9) occurred as commonly in both groups. CONCLUSIONS: AMA can appear and disappear. They do not identify a subgroup that requires different treatment or that evolves quickly into a cholestatic syndrome.