Anticentromere antibodies in subjects with no apparent connective tissue disease.

OBJECTIVES--To study the association of anticentromere antibodies (ACA) in various diseases. METHODS--A total of 4800 consecutive serum samples were tested for ACA by indirect immunofluorescence using HEp-2 cells as substrates and by immunoblotting of Molt-4 cell mitotic chromosomal antigens and recombinant CENP-B protein. RESULTS--Anticentromere antibodies were identified in the serum samples of 24 subjects, including eight without apparent connective tissue diseases, six with primary biliary cirrhosis, two with diffuse scleroderma, one with pulmonary hypertension, one with primary Raynaud's phenomenon, one with CREST syndrome (calcinosis, Raynaud's phenomenon, oesophageal dysmotility, sclerodactyly, telangiectasia), and five with other connective tissue diseases. By immunoblotting using Molt-4 cells mitotic chromosomal antigens three centromere antigens were recognised by these serum samples. These were: CENP-A (17 kilodalton recognised by 22 of 24 ACA positive serum samples); CENP-B (80 kilodalton recognised by 22 of 24 ACA positive serum samples); and CENP-C (140 kilodalton recognised by 19 of 24 ACA positive serum samples). There was no specific pattern for serum samples from patients with different groups of diseases on immunoblotting. Recombinant CENP-B proteins were all recognised by these samples. Patients without apparent connective tissue disease often had a lower ACA titre than patients with primary biliary cirrhosis. CONCLUSIONS--These data suggest that a positive result for ACA does not always indicate the presence of a connective tissue disease.