Azathioprine is a cytotoxic immunosuppressant drug used widely in clinical neurology as an adjunct to steroid treatment for autoimmune and inflammatory conditions. As a result of the relatively high cost and modest benefit of the newly licensed immunomodulatory therapies in the treatment of multiple sclerosis there has been a resurgence of interest in the possible benefits of azathioprine. A meta-analysis in 1997 suggested that it was as effective as newer treatments in increasing the proportion of patients who remain free of relapse at 2 years.1

The mode of action of azathioprine at the immune cell level remains unclear. It is converted rapidly in vivo to 6-mercaptopurine, which is extensively metabolised along three competitive routes (table1).2 Methylation catalysed by thiopurine transferase (TPMT) leads to the production of 6-methyl mercaptopurine. Wide variations in TPMT activity exist between patients and are determined by a common genetic polymorphism; 89% of the population have high TPMT concentrations, 11% intermediate concentrations, and 1 in 300 low or absent TPMT concentrations. A second catabolic route is oxidation with xanthine oxidase, which exhibits little interindividual variation in activity. The third route, catalysed by hypoxanthine guanine phosphoribosyl transferase, results in the formation of active thiopurine metabolites including 6-thioguanine nucleotides (6-TGN) which are thought to be responsible for the cytotoxicity of azathioprine.

In 1980 Weinshilboum and Sladek3 proposed that the inherited variation in TPMT activity might represent one factor in individual variations in sensitivity to thiopurine drugs. The importance of a low activity of TPMT with an associated increase in 6-TGN in red blood cells in patients with bone marrow failure treated with azathioprine was reported by Leonard et al 4 in 1989. Since 1963 61 fatalities suspected to be associated with azathioprine have been reported to the Committee on Safety of Medicines (personal communication); 25 of these were classed as haemopoietic disorders. The importance of an inherited deficiency in thiopurine methyltransferase is mentioned specifically in the data sheet for Imuran (Glaxo-Wellcome).

Full blood count is a poor method of detecting early bone marrow toxicity as by the time changes have occurred dangerously high concentrations of 6-TGN may have accumulated. An alternative which allows this complication to be anticipated is the measurement of TPMT activity in red blood cells. This is a relatively cheap test (∼ £26), performed on an EDTA blood sample, and genotypic testing may soon be clinically available Late onset myelosuppression has a more gradual onset and can be detected by changes in blood counts.

The overview of azathioprine treatment in multiple sclerosis published as a meta-analysis in the Lancet in October 19915 showed that the probability of freedom from relapse during the first, second, and third year of treatment was significantly greater in the azathioprine group, but the change in the expanded disability status scale (EDSS) was not significantly different. The authors concluded that it was debatable whether the slight clinical benefits outweigh side effects and that it is still not possible to predict which patients are likely to benefit from treatment with azathioprine. Although studies included in the Cochrane database mention morbidity in terms of decreased haematological indices and three mortalities said to be unrelated to multiple sclerosis or azathioprine, TPMT concentrations were not measured in any of the trials included in the meta-analysis.

Individual variation in TPMT may explain the variable toxicity and treatment response with azathioprine in multiple sclerosis. In addition, knowledge of TPMT status in patients with multiple sclerosis could identify those unsuitable for azathioprine treatment and those in whom the dose could be increased to the top of the therapeutic range secure in the knowledge of a very low probability of toxicity. It has been suggested that studies using azathioprine may fail to detect a therapeutic effect due to underdosage—if TPMT is measured this can be avoided.

Anticipation of azathioprine related toxicity and the tailoring of dose to the metaboliser status of individual patients might have considerable implications in routine clinical practice.