To prevent relapse, high risk paediatric acute lymphoblastic leukaemia (ALL) is treated very intensively. However, most patients who eventually relapse have standard or medium risk ALL with low minimal residual disease (MRD) levels. We analysed recurrent microdeletions and other clinical prognostic factors in a cohort of 475 uniformly treated non-high risk precursor B-cell ALL patients with the aim of better predicting relapse and refining risk stratification. Lower relapse-free survival at 7 years (RFS) was associated with IKZF1 intragenic deletions (P < 0·0001); P2RY8-CRLF2 gene fusion (P < 0·0004); Day 33 MRD>5 × 10−5 (P < 0·0001) and High National Cancer Institute (NCI) risk (P < 0·0001). We created a predictive model based on a risk score (RS) for deletions, MRD and NCI risk, extending from an RS of 0 (RS0) for patients with no unfavourable factors to RS2 + for patients with 2 or 3 high risk factors. RS0, RS1, and RS2 + groups had RFS of 93%, 78% and 49%, respectively, and overall survival (OS) of 99%, 91% and 71%. The RS provided greater discrimination than MRD-based risk stratification into standard (89% RFS, 96% OS) and medium risk groups (79% RFS, 91% OS). We conclude that this RS may enable better early therapeutic stratification and thus improve cure rates for childhood ALL.

, , , ,,
British Journal of Haematology

Sutton, R., Venn, N., Law, T., Boer, J.M. (Judith M.), Trahair, T.N. (Toby N.), Ng, A., … Marshall, G. M. (2018). A risk score including microdeletions improves relapse prediction for standard and medium risk precursor B-cell acute lymphoblastic leukaemia in children. British Journal of Haematology, 180(4), 550–562. doi:10.1111/bjh.15056