Introduction: In contemporary treatment protocols for childhood acute lymphoblastic leukemia (ALL), most chemotherapeutic agents are prescribed as a fixed dose per m2. 6-mercaptopurine (6-MP) is frequently used during consolidation therapy and mediates its effect through incorporation into DNA as thioguanine nucleotides (DNA-TGN), higher levels of which are associated with a decreased risk of relapse. In this randomized controlled trial on children with non-high risk ALL, we tested the hypothesis that individualized, toxicity-titrated 6-mercaptopurine (6-MP) increments during consolidation therapy with high-dose methotrexate were superior to the standard fixed-dose 6-MP regimen. The primary end point was minimal residual disease (MRD) positivity at end of consolidation, and the secondary end point was event-free survival. The trial was registered at clinicaltrials.gov (NCT00816049).

Methods: Between January 2009 and March 2016 we randomized 392 children aged 1-17 years to experimental and 396 to standard therapy. Patients allocated to standard therapy received oral 6-MP (25 mg/m2/day) from day 30 to 85, while the experimental arm received stepwise increments with 25 mg/m2/day beginning on days 50 and/or 71 unless they had experienced dose-limiting myelosuppression (neutrophil count ≤ 0.5 x 109/L or thrombocyte count < 50 x 109/L) following the preceding dose of high-dose methotrexate (days 36 and 57). MRD was measured using PCR for T-cell ALL and flow cytometry for B-cell precursor ALL (or PCR, if flow was not technically possible). MRD positivity was defined as any detectable marker with PCR and more than 10 leukemic events with flow cytometry. As a measure of treatment intensity and physician and patient compliance, we analyzed concentrations of DNA-TGN and methylated 6-MP metabolites (MeMP) in blood samples collected throughout consolidation therapy. Analyses were carried out on an intention-to-treat basis.

Results: In the experimental arm, 166 patients (42%) received one dose increment, and 62 (16%) received two. Of 387 patients in the experimental arm, 57 (15%) were MRD positive at end of consolidation versus 77 of 389 (20%) in the control arm (P=0.08). Five-year probability of event-free survival was 0.89 (95% CI: 0.85-0.93) in the experimental arm versus 0.93 (0.90-0.96) in the control arm (P=0.13). The median cumulative length of 6-mercaptopurine treatment interruptions was 7 days (IQR 2-12) in the experimental arm versus 4 (IQR 0-10) in the control arm (P=0.002), but duration of treatment interruption was not associated with risk of relapse (P=0.99). There were seven treatment-related deaths in the experimental arm versus three in the control arm (P=0.22), and none of the deaths in the experimental arm were related to 6-MP increments. At the final risk stratification on day 79, one patient in the experimental arm and seven in the control arm (P=0.07) had MRD values ≥0.1%, which is the cutoff for stratification to high risk treatment with stem cell transplantation. The mean DNA-TGN levels at end of consolidation in patients given zero, one, or two dose increments were 162 fmol/µg DNA (95% CI 144-179), 237 (194-280), and 293 (156-430) respectively, P<0.0001. The corresponding MeMP values were 1 879 nmol/mmol Hb (1 397-2 360), 5 930 (4 368-7 492), and 12 173 (7 625-16 721), P<0.0001.

Conclusion: Despite 6-MP dose intensifications in 58% of patients in the experimental arm and a strong association between dose increments and DNA-TGN levels, as well as a non-significant improvement in MRD at end of consolidation, we found no survival benefit from individualized 6-MP increments compared to standard therapy.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.