I M Roitt

The sera of virtually all patients with primary biliary cirrhosis contained non-organ specific antibodies, probably directed against mitochondria, which give rise to cytoplasmic `M' immunofluorescence in unfixed sections from various organs in high titres. They were also found in 31% of patients with cryptogenic cirrhosis particularly those with obstructive features, and 28% of cases of active chronic (lupoid) hepatitis. In contrast only two patients out of twenty-eight with extrahepatic biliary obstruction and one patient out of twenty-five with infective hepatitis gave positive reactions and then in very low titre. Uniformly negative results for `M' immunofluorescence were obtained in alcoholic cirrhosis, cholestatic drug jaundice and cholestasis associated with ulcerative colitis. These findings suggest that `M' antibodies do not arise merely in response to liver damage and confirm the value of the test for the differential diagnosis between primary biliary cirrhosis and extrahepatic biliary obstruction although care must be taken in interpretation of the results in cases with associated connective tissue disorders where a significant incidence of positive reactions was observed. Seventy-seven per cent of juvenile cirrhosis, 46% of primary biliary cirrhosis and 38% of the cryptogenic cirrhosis patients had anti-nuclear factors. The incidence of ANF in the other liver groups was no higher than in mixed hospital controls. A high incidence of antibodies staining smooth muscle was observed in the juvenile and primary biliary cirrhosis groups. No liver specific antibodies could be detected by precipitation or tanned red cell agglutination although non-organ-specific antibodies were demonstrated in some instances with the latter technique. Thyroid specific autoantibodies were increased four-fold in females with active chronic hepatitis but the incidence of thyroid and gastric autoantibodies was not significantly different from the controls in all other liver conditions studied. Autoimmune reactions are prominent in primary biliary cirrhosis and in active chronic hepatitis but their role in the pathogenesis of these diseases is still undefined.

The presence of an inhibitor of intrinsic factor activity was demonstrated in the sera of some patients with pernicious anaemia who had never been treated with hog intrinsic factor (Taylor, 1959b; Schwartz, 1960). It was proposed that this inhibitor, which was precipitated with the serum globulin fraction, might be an autoantibody. This supported the earlier suggestions, based upon the histological appearances of the gastric mucosa, that pernicious anaemia might have an autoimmune component (Taylor, 1959a). Initial attempts to determine the nature of the serum factor by precipitation and tanned red-cell techniques were inconclusive (Taylor and Boyden, quoted by Schwartz, 1960; Abels et al., 1962).

A new, sensitive and quantitative technique for the routine estimation of immune complexes has been devised. The assay involved the binding of immune complexes to Clq linked to the surface of plastic tubes; the amount of immune complex bound is then determined by adding radiolabelled anti-human IgG. Immune complexes were detected in twelve out of thirty-one patients with systemic lupus erythematosus. The test should prove useful in the diagnosis of immune complex diseases and will be valuable in the evaluation of disease activity.