Background and objective: The activity of thiopurine S-methyltransferase (TPMT), a key enzyme in the metabolism of purine analogues, displays wide inter-subject variability partly due to a genetic polymorphism. Previous studies have suggested adjusting purine analogues dosing according to TPMT activity but measurements are costly and time-consuming. It is still unclear, especially under treatment, whether the simpler TPMT genotyping reliably predicts enzyme activity. Our aim was to study the possible correlation of TPMT genotype with phenotype.
Methods: We determined the genotypic status and TMPT activity, at diagnosis and after 6 months of maintenance therapy, of 118 children with acute lymphoblastic leukaemia (ALL).
Results and discussion: Eighty-nine per cent of the children had a homozygous wild-type genotype (group 1), 11% had one or two mutant allele(s) (group 2). At both time points, TPMT activity (U/mL peripheral red blood cell) was significantly higher in group 1 than in group 2 (P < 0.001) but inter-group levels overlapped considerably. There was considerable heterogeneity in the percentage increase in TPMT activity after therapy, and little correlation between metabolites ratio [6-methylmercaptopurine derivative/6-thioguanine nucleotides (6-TGN)] and TPMT activity at the end of 6 months' maintenance treatment. These results show that TPMT activity cannot be used as an accurate tool for 6-mercaptopurine monitoring.
Conclusion: Genotyping at diagnosis identifies patients with a homozygous mutant TPMT and may prevent severe and life-threatening toxicity. ALL treatment monitoring should preferentially be based on repeated determinations of intracellular active metabolites (6-TGN) and methylated metabolites.