Serum autoantibodies, ulcerative colitis and primary sclerosing cholangitis.
Chapman RW., Cottone M., Selby WS., Shepherd HA., Sherlock S., Jewell DP.
The aetiology of primary sclerosing cholangitis is unknown, but it is closely associated with ulcerative colitis. Serum anticolon antibodies, crossreacting with portal tracts, have been reported in patients with ulcerative colitis but no studies have been carried out in primary sclerosing cholangitis. The frequency of serum anticolon antibodies and portal tract antibodies have been measured in 24 patients with primary sclerosing cholangitis and ulcerative colitis; 15 patients with primary sclerosing cholangitis without ulcerative colitis; 77 patients without primary sclerosing cholangitis: 25 patients with Crohn's colitis; 10 patients with primary biliary cirrhosis; 22 patients with extrahepatic biliary obstruction and 20 normal controls. Serum anticolon and portal tract antibodies were detected using immunoperoxidase techniques on normal colon and obstructed human liver. Tissue typing was undertaken using a standard microcytotoxicity technique. The frequency of anticolon antibodies was markedly increased in primary sclerosing cholangitis patients with ulcerative colitis (62.5%) compared with patients with ulcerative colitis (17%) and Crohn's colitis (16%) (chi 2 = 17.9; p less than 0.001). The antibodies were almost entirely of IgG and IgA classes in all groups. Anticolon antibodies were not found in sera from any other group. Sera from eight of 15 patients with primary sclerosing cholangitis, ulcerative colitis and anticolon antibody reacted with portal tracts of human obstructed liver. This reaction was also seen in four of nine patients with ulcerative colitis and primary sclerosing cholangitis and in three of 15 patients with primary sclerosing cholangitis alone. Portal tract antibody was of IgG class and was not present in sera from any other groups. Unlike anticolon antibody, there was a close relationship between HLA-B8 phenotype and the portal tract antibody (p<0.02; chi 2 = 6.04). Absorption studies confirmed that the anticolon antibody is distinct from portal tract antibody.