The thiopurine drugs mercaptopurine (6MP) and thioguanine (6TG) are a major part of long-term chemotherapy in childhood acute lymphoblastic leukaemia (ALL). One major factor influencing thiopurine metabolism is variable thiopurine methyltransferase (TPMT) activity. Approximately 89% of the population are homozygous wildtype (high activity), 11% are heterozygotes (intermediate activity) and 1 in 300 are homozygous variant (no activity). Thioguanine nucleotides (TGNs) are cytotoxic metabolites of 6MP and 6TG. Children with lower TPMT activities accumulate high TGN concentrations and experience more cytotoxicity. At disease diagnosis there is poor concordance between TPMT phenotype and genotype at lower TPMT activities. This is due to an excess of old red cells with decayed TPMT activity (Lennard et al., 2001). The aim of this study was to investigate the concordance between TPMT activity and genotype in children with ALL during thiopurine chemotherapy.

As part of the UK national MRC ALL97/99 trial, blood samples (5ml) were forwarded from children during thiopurine chemotherapy for the measurement of TGN metabolites, TPMT activity and genotype (Otterness et al., 1997). Samples were taken under standard conditions at the target protocol (75mg/m² 6MP, 40 mg/m² 6TG), or maximum tolerated, dose. Blood samples forwarded to the Sheffield Unit include an excess from problem patients, eg children who are sensitive to thiopurines and those with compliance problems.

A consecutive group of 214 children were studied, 113 children were randomised to 6MP and 101 to 6TG. TPMT activities ranged from 0 to 23 units ml/l RBCs (median 14.7). The break point between intermediate and high TPMT activity is 11.5 units (Lennard et al., 1990). The children were genotyped for the TPMT*3 family, which account for over 90% of variant alleles. Forty children had TPMT activities <11.5 units, 2 were TPMT deficient (*3A/*3A, *3C/*3C), 22 were *1/*3A, 5 were *1/*3C and 11 *1/*1. Of the 174 children with activities >11.5 units, 171 were *1/*1 and 3 were *1/*3A. Two of the latter had TPMT activities at the break point. TGN metabolite concentrations (pmol/8 x 10⁸ red cells) in the TPMT deficient children were 978pmol and 1784 pmol at 5 and 20 mg/m² 6MP respectively. For those taking 6MP, median TGNs were 724pmol for those with intermediate TPMT activity and 371pmol for those with high activity (median difference 300pmol, 95%CI 144 to 473, p<0.001). For those taking 6TG, median TGNs were 2682 and 1690pmol for the intermediate and high activity TPMT groups respectively (median difference 1009pmol, 95%CI 624 to 1474, p<0.001). During chemotherapy 91% of children with TPMT*3 variant alleles lacked, or had intermediate, TPMT activity as did 6% of children wild-type with respect to TPMT*3. This group of children accumulated higher TGN concentrations and experience more myelosuppression.

Lennard L et al. (1990) Lancet 336: 225.
Lennard L et al.(2001) Br J Clin Pharmacol 52: 539.
Otterness D et al (1997) Clin Pharmacol Ther 62: 60.

This work was supported by the Leukaemia Research Fund and the MRC Childhood Leukaemia Working Party.