Single IGH/TCR MRD Marker Is More Informative and Cost-Effective in Risk-Stratified Therapy for Childhood ALL: Prospective Treatment Trial Results from the Malaysia-Singapore (Ma-Spore) ALL 2003 Study

Background: Several large cohort studies–like from BFM and St Jude groups-have shown that minimal residual disease (MRD) is the single most important predictor of treatment outcome in childhood acute lymphoblastic leukemia. However, lingering concerns about clonal evolution of IgH/TCR rearrangements may result in false-negative MRD. Hence, the AIEOP-BFM ALL 2000 study utilized at least 2 sensitive MRD markers with at least 10⁻⁴ sensitivity for risk-stratification. This stringent minimum of 2 sensitive MRD markers criteria limits the applicability of MRD stratification to <80% of patients and it is more costly and time consuming. The Malaysia-Singapore ALL 2003 study trial adopted a similar strategy of primarily early response assessment like Day-8 prednisolone response (PR) and a simplified methodology of at least one MRD marker (IgH/TCR) for risk stratification. We hypothesize that clonal evolution does not occur early during therapy and that a single marker PCR-based methodology is more informative and cost-effective, without hampering the accuracy of MRD risk stratification.

Methods: A total of 362 patients (B-lineage=333, T-lineage=29) were enrolled in Ma-Spore ALL 2003 from July 2002 to August 2008, with a median age of 5.61 years (range 0.14 to 15.29). The treatment protocol is based on the modified BFM ALL 2000 backbone without randomization. Risk-stratification is based primarily on MRD levels at end of induction (Day-33) and week 12, Day-8 PR, and presence of BCR-ABL or MLL. In patients whom we are unable to quantify MRD, are assigned IR risk group. Screening for IgH/TCR rearrangements is carried out using multiplex BIOMED-2 primers (JJM van Dongen et al 2003) with standardization with the European MRD Study Group, which we are an active member. Selection of MRD markers is based on frequency and stability of IgH/TCR rearrangements. MRD quantifications were carried out using Real-time PCR (LightCycler 1.0, Roche Diagnostics).

We define three distinct treatment subgroups: Standard-Risk (SR) who had MRD <10⁻⁴ at both time-points; High-Risk (HR) who had poor PR or no clinical remission or BCR-ABL or infant MLL or MRD at week 12 >1×10⁻³; remaining patients formed Intermediate-Risk (IR). The SR patients received decelerated therapy while HR group was treated under intensified therapy or bone marrow transplant.

Results: Of 362 patients enrolled, Ma-Spore risk stratification is possible in 360 patients (n=2 delayed treatment), out of which 239 have completed 2 years therapy, 15 had induction failures, and 22 had relapses (isolated BM=17, isolated CNS=4, isolated testis=1.) Patients stratified into SR=29% (n=106/360); IR=50% (n=180/360), and HR=21% (n=74/360). At the end of the study, 4-years overall survival (0S) was SR=94.2±2.5%; IR=95.3±1.6%; HR=66.2±6.5%; event-free survival (EFS) was SR=91.8±3%; IR=86.2±3.4%; HR=52.9±6.4%, and leukemia-free survival (LFS) was SR=94.8±2.6%; IR=90.5±3.2%; HR=62.7±6.6%. Within the HR subgroup, using multivariate analysis, high MRD levels at week 12 (>10⁻³) is significantly associated with adverse treatment outcome (OS=27.5±15.8%, EFS=17.9±14.4%; LFS=23.1±17.8%), whilst poor prednisolone responders alone in the absence of high risk MRD or high risk BCR-AL or MLL-AF4 did well (OS=84.1±7.4%, EFS=84.1±7.4%, LFS= no event).

Of the 362 patients, at least 91% (n=309/339) of patients had at least 1 sensitive MRD marker; the remaining patients have no diagnostic samples (n=9), died before time-point 1(n=10), and default before time-point 1(n=4). For risk-stratification using only MRD levels, 45.7% of the patients (n=128/280) were low-risk with MRD<10⁻⁴ at both time-points; 3.2% of the patients (n=9) were high-risk with MRD>10⁻³ at week 12; and the remaining 51.1% (n=143) were stratified as intermediate-risk MRD. Overall survival in MRD low-risk group was 93.2±2.5%, MRD intermediate-risk was 93.6±2.4% and MRD high-risk MRD was 45±18.8%.

Conclusion: We have demonstrated that a simplified MRD methodology using a single PCR-based marker (IgH/TCR) can be successfully implemented to tailor therapy for childhood ALL without adversely affecting outcome. It is cost-effective and can be applied to a larger group of patients (91%). This is particularly useful for countries with limited resources to pursue risk stratification incorporating MRD. Deceleration of therapy in the SR group would also help to reduce long-term side effects and better quality life in these children.