[Autoimmunity induced by drugs. Immunological characteristics and etiopathogenic hypotheses]

Drug-induced autoimmune diseases have two immunological peculiarities. Firstly, some autoantibodies are present, which are virtually never seen in spontaneous human diseases and may be regarded as specific. This applies to antimitochondria antibody type 3 (anti M3) in the lupus-like syndrome caused by Venocuran, to antimitochondria antibody type 6 (anti M6) in iproniazide-induced hepatitis, to anti-insulin antibody found after treatment with methimazole, and to anti liver/kidney microsome antibody type 2 (anti LKM2) associated with hepatitis induced by tielinic acid. Secondly, a search for other autoantibodies shows that the immune disorder is much more limited than in spontaneous autoimmune diseases. Thus, contrary to myasthenia and idiopathic autoimmune haemolytic anaemia, we never found autoantibodies specifically directed against the thyroid, the stomach or the adrenal gland during treatment with D-penicillamine and alpha-methyldopa. Only some hypotheses may account for these peculiarities. Cross-reaction between drug and autoantigen may occur, but the fact that the antigen-antibody reaction is not inhibited by the drug or its metabolites does not support this explanation. Much more attractive is the "T-cell bypass" theory, according to which autoreacting suppressor T-cells are circumvented by helper T-cells stimulated by the drug-modified autoantigen. In this case, the autoimmune reaction would indicate to which body substance the drug is bound, thus making it immunostimulant, and not a structural similarity between the drug and the autoantigen.