The hepatic toxicity of antituberculosis regimens containing isoniazid, rifampicin and pyrazinamide

Abstract

This paper reviews hepatic toxicity during chemoprophylactic treatment with isoniazid alone, and during the treatment or retreatment of active pulmonary tuberculosis with regimens containing one or more of the drugs isoniazid, rifampicin and pyrazinamide. Chemoprophylaxis with isoniazid carries a risk of drug-induced hepatitis, and this risk needs to be weighed against the advantages of preventing tuberculosis morbidity. The risks of hepatitis during standard treatment based on isoniazid are very small, and most patients who develop hepatitis recover. Moreover, it is often doubtful whether hepatitis is in fact drug-induced, and a proportion of patients who develop it already have liver disease at the time treatment is started. The risks are acceptable in the treatment of bacteriologically active disease. There is no consistent evidence that giving rifampicin with isoniazid in the initial treatment of tuberculosis increases the risk of hepatitis; in particular, transient abnormalities in the results of tests of liver function during the early weeks of treatment do not imply serious toxicity; patients who are rapid acetylators of isoniazid are not, as has been suggested, exposed to any special risk, and patients with known liver disease can also be treated without undue risk. Retreatment regimens based on rifampicin plus ethambutol carry a low risk of hepatitis, even though patients who need retreating have often experience toxicity during their initial treatment. Frist-line or second-line regimens containing pyrazinamide in currently accepted dosages, given daily or intermittently, carry a low and acceptable risk of hepatic toxicity. Finally, current studies of daily and intermittent short-course regimens based on isoniazid, rifampicin and pyrazinamide will extend our knowledge of hepatic toxicity. Because such regimens involve small total quantitites of drugs given over short periods they are likely to give rise to less hepatic toxicity than regimens of standard duration.