Prognostic impact of minimal residual disease (MRD) in children is different in B or T lineage acute lymphoblastic leukemia: Results of trial AIEOP-BFM ALL 2000

Slow early response indicates poor prognosis in childhood ALL. We aimed to evaluate if post-induction MRD levels had different prognostic impact in precursor B-cell (pB) or T-cell ALL. From 07/2000 to 06/2006, 4730 pts with ALL were enrolled in trial AIEOP-BFM ALL 2000. MRD levels were centrally measured by real-time quantitative polymerase chain reaction using the identification of clone-specific T-cell receptor and immunoglobulin gene rearrangements. MRD study time-points (TP) were treatment day 33 (TP1, end of induction) and day 78 (TP2, after consolidation). To define MRD negativity, two markers with a sensitivity of at least 10–4 were required. Patients were treated with BFM induction (protocol I-A), consolidation (I-B), extra-compartment/intensified consolidation (HD-MTX in non-high-risk patients, pulses in high-risk patients), reinduction, and maintenance. MRD analysis at one or two time points suceeded in 3707 pts; the immunophenotype was available from 3636 pts. MRD levels and corresponding estimated 5-year event-free survival (5y-pEFS) comparing pB- and T-ALL are shown in Table 1 (3yrs median follow-up). MRD response in T-ALL was slower than in pB-ALL resulting in a higher percentage of pts with high MRD load in T-ALL. In pB-ALL as well as T-ALL, high MRD levels at TP2 were well predictive to identify pts with poor prognosis. For prediction of good prognostic subgroups, TP1 was more appropriate identifying a subgroup with excellent 5y-pEFS of >90% in case of MRD negativity. Specificity of TP1 was poor in T-ALL if the pB-ALL criteria of MRD negativity were applied. If MRD low positive and MRD negative T-ALL pts were combined, the discrimination was as good as in pB-ALL. The optimal choice of MRD evaluation time points depends on biological factors and treatment, and is most relevant for MRD-based risk stratification.