Genotype and prognosis in childhood leukaemia

5 July 2005

Thiopurine drugs such as mercaptopurine and thioguanine are used to treat malignancies such as acute lymphoblastic leukemias (ALLs), as well as rheumatic disease, inflammatory bowel disease, dermatologic and other conditions. These drugs are metabolised by a key enzyme, thiopurine methyltransferase (TPMT), the activity of which is controlled by a common genetic polymorphism that affects drug efficacy and toxicity. Several alleles are associated with reduced or absent TPMT activity (when present in heterozygous or homozygous form, respectively); around 10% of the Caucasian population are heterozygous and 0.3% homozygous for mutant (inactive) TPMT alleles. Such individuals cannot completely metabolise thiopurine drugs and will have higher residual levels of thiopurines and intermediate metabolic products, which can lead to toxic side-effects, if treated with standard doses.

A new study published in the Journal of the American Medical Association (JAMA) reports an investigation into the association between TPMT genotype with prognosis among children receiving mercaptopurine for the treatment of childhood ALL [Stanulla M et al. (2005) JAMA 293, 1485-1489]. Children with deficient TPMT activity treated for ALL have an increased risk of severe haematopoietic toxicity if treated with normal dosages of thiopurines, but high dose mercaptopurine has also been associated with a better prognosis. Chemotherapy for ALL is often adjusted according to the risk of treatment failure, as deduced from prognostic factors. One such prognostic measure is that of minimal residual disease analysis, polymerase chain reaction (PCR) based detection of leukaemic-specific sequences, which is reportedly highly predictive of disease recurrence.

A total of 814 ALL patients aged 1-18 participated in the study; all were genotyped with respect to the TPMT gene, and minimal residual disease for analysis was performed on days 33 and 78 of their treatments, before and after administration of a 4-week course of mercaptopurine. 92.8% of the patients were homozygous for wild-type TPMT, 6.8% heterozygous, and 0.5% homozygous for mutant TPMT alleles. This latter group was treated with a 10-fold lower dose of mercaptopurine to prevent toxicity, and were excluded from further study; the heterozgotes and wild-type homozygotes received standard doses. Minimal residual disease levels on treatment day 33 were equivalent in both groups, but on treatment day 78 (following mercaptopurine treatment), significant differences were observed between the wild-type and heterozygous patient groups. The heterozygous patients were found to have a 2.9-fold reduction in risk of having measurable minimal residual disease (and hence good prognosis) at this point, but no increased risk of toxic side-effects.

The authors conclude that TPMT genotype has a substantial impact on disease prognosis after administration of mercaptopurine for the treatment of childhood ALL, proposing that this observation is due to genetically determined modulation of mercaptopurine dose intensity. The heterozygous patients treated with equivalent doses to wild-type patients in this study received effectively greater levels of mercaptopurine, since they were less able to metabolise it, and subsequently were more likely to show good prognostic signs. The researchers note that long-term outcome data will be required on these patients, but suggest that they may call for increasing doses of mercaptopurine according to TPMT genotype (an increase above standard doses for wild-type patients) in the early treatment of ALL, since this may improve overall outcomes. The report also calls for future assessments of treatment response to specific drugs to include analysis of genetic variation in drug-metabolizing enzymes and minimal residual disease.

Comment: This study adds an interesting angle to the study of pharmacogenetics (genetic factors that influence individuals

More from us