BACKGROUND: Adjustment of mercaptopurine and methotrexate maintenance therapy of acute lymphoblastic leukaemia by leucocyte count is confounded by natural variations. Cytotoxicity is primarily mediated by DNA-incorporated thioguanine nucleotides (DNA-TGN). The aim of this study was to establish whether DNA-TGN concentrations in blood leucocytes during maintenance therapy are associated with relapse-free survival.

METHODS: In this substudy of the NOPHO ALL2008 phase 3 trial done in 23 hospitals in seven European countries (Denmark, Estonia, Finland, Iceland, Lithuania, Norway, and Sweden), we analysed data from centralised and blinded analyses of 6-mercaptopurine and methotrexate metabolites in blood samples from patients with non-high-risk childhood acute lymphoblastic leukaemia. Eligible patients were aged 1·0-17·9 years; had been diagnosed with non-high-risk precursor B-cell or T-cell leukaemia; had been treated according to the Nordic Society of Pediatric Hematology and Oncology ALL2008 protocol; and had reached maintenance therapy in first remission. Maintenance therapy was (mercaptopurine 75 mg/m(2) once per day and methotrexate 20 mg/m(2) once per week, targeted to a leucocyte count of 1·5-3·0 × 10(9) cells per L). We measured DNA-TGN and erythrocyte concentrations of TGN nucleotides, methylated mercaptopurine metabolites, and methotrexate polyglutamates. The primary objective was the association of DNA-TGN concentrations and 6-mercaptopurine and methotrexate metabolites with relapse-free survival. The secondary endpoint was the assessment of DNA-TGN concentration and 6-mercaptopurine and methotrexate metabolites during maintenance therapy phase 2.

FINDINGS: Between Nov 26, 2008 and June 14, 2016, 1509 patients from the NOPHO ALL2008 study were assessed for eligibility in the DNA-TGN substudy, of which 918 (89%) of 1026 eligible patients had at least one DNA-TGN measurement and were included in the analyses. Median follow-up was 4·6 years (IQR 3·1-6·1). Relapse-free survival was significantly associated with DNA-TGN concentration (adjusted hazard ratio 0·81 per 100 fmol/μg DNA increase, 95% CI 0·67-0·98; p=0·029). In patients with at least five blood samples, erythrocyte concentrations of TGN, methylated mercaptopurine metabolites, and methotrexate polyglutamates were associated with DNA-TGN concentration (all p<0·0001).

INTERPRETATION: Our results suggest the need for intervention trials to identify clinically applicable strategies for individualised drug dosing to increase DNA-TGN concentration, and randomised studies to investigate whether such strategies improve cure rates compared with current dose adjustments based on white blood cell counts.

FUNDING: Danish Cancer Society, Childhood Cancer Foundation (Denmark), Childhood Cancer Foundation (Sweden), Nordic Cancer Union, Otto Christensen Foundation, University Hospital Rigshospitalet, and Novo Nordic Foundation.