Liver is a frequent target for drug-induced hepatitis. They can be cla
ssified in two categories: the hepatitis in which the drug or a metabo
lite reach a vital target in the cell and the hepatitis in which the d
rug triggers an adverse immune response directed against the liver. We
mill discuss essentially this second kind of disease. They have key c
linical features such as the low frequency, the dose independence, the
delay between the beginning of drug intake and the triggering of the
disease, the shortening of the delay upon rechallenge and very often t
he presence of autoantibodies in the serum of the patients. Such signs
were found in hepatitis triggered by drugs such as halothane, tienili
c acid, dihydralazine, anticonvulsants. They will be taken as examples
to show the recent progress in the understanding of the mechanisms le
ading to the disease. It has been postulated that the drug is metaboli
sed into a reactive metabolite binding to the enzyme which generated i
t; therefore the neoantigen might trigger an immune response character
ised by the production of antibodies recognising the native and or the
modified protein. Most of these steps were proven in the cases of hal
othane, tienilic acid and dihydralazine. Several points seem important
in the development of the disease: the equilibrium between toxication
and detoxication pathways, the nature and amount of neoantigen, the i
ndividual immune response. However, many points remain unclear: for in
stance, the reason for the very low frequency of this kind of disease;
the precise mechanism of the adverse immune response; the risk factor
s for developing such adverse reactions. Efforts should be made to bet
ter understand the mechanisms of this kind of disease: for instance, a
n animal model, tests to identify drugs at risk for such reactions, th
e role of these drugs in the processing of P450s and the processing of
the neoantigens for their presentation to the immune system.