Peg-Asparaginase Administration on Day 3 of Remission Induction Accelerates MRD Response in Children with B-Cell Precursor Acute Lymphoblastic Leukemia

To improve the outcome of children with standard and intermediate risk acute lymphoblastic leukemia (ALL) by the addition of PEG-asparaginase (PEG) on day 3 of remission induction was one of the key aims of trial ALL MB-2008 (J. Roumjantseva et al, Pediatriya, 2009). Therapy was designed based on previous MB-group studies (A. Karachunsky et al, Leukemia, 2008). Aim of the present study was to evaluate the influence of day 3 PEG on the end-induction MRD-status which is one of the significant predictors of outcome in children with B-cell precursor ALL (BCP-ALL).

Patients and methods. In total, 474 patients (242 boys and 232 girls aged 1 to 17 years (median 5 years 8 months) were studied. Inclusion criteria were: BCP-ALL, treatment according to protocol ALL MB-2008, compliance with standard risk group (SRG, n=266) and intermediate risk group (ImRG, n=208) criteria according to the protocol stratification system, and availability of end-induction (day 36) bone marrow (BM) samples. Remission induction of ALL MB-2008 consisted of oral dexamethasone (6mg/m² daily), 5 injections of vincristine (1,5 mg/m²), 0-2 injections of daunorubicin (DNR, 45 mg/m²) and 6 triple intrathecal injections. Patients were randomized in respect to PEG administration on day 3. SRG patients were divided into three groups (PEG+DNR+, PEG+DNR- and PEG-DNR+); ImRG patients were split into two (PEG+DNR+ and PEG-DNR+). MRD detection was performed by multicolor flow cytometry (MFC) in three reference laboratories. Samples with a tumor cell level above 0.01% were considered MRD-positive. Outcome parameters were event-free survival (EFS) and cumulative incidence of relapse (CIR). Median of follow up was 5 years.

Results. Outcome in 474 study group patients was not significantly different from outcome of 2310 children with BCP-ALL enrolled in SRG and ImRG of ALL MB-2008 trial in whom MRD was not studied: EFS was 0.86±0.02 and 0.84±0.01, respectively, p=0.14, and CIR was 0.12±0.02 and 0.09±0.01, p=0.38.

MFC MRD data obtained at day 36 had a significant impact on prognosis in SRG patients. EFS as well as CIR were worse for patients with MRD >0.1% (n=26; EFS 0.47±0.15, CIR 0.49±0.17) compared with 248 patients with low/negative MRD (EFS 0.94±0.2, CIR 0.05±0.02, p<0.001). Nevertheless, the distribution of MRD levels was completely different dependent on randomization arms: in arm PEG-DNR+ (n=92) 18.5% of patients belonged to the group with unfavorable MRD compared with only 6.2% in arm PEG+DNR- (n=80, p=0.017) and 4.3% in arm PEG+DNR+ (n=94, p=0.002). Interestingly, in two PEG+ arms the number of patients with high MRD was not significantly different (p=0.558). Despite the lower proportion of high MRD levels in PEG+ arms MRD maintained its prognostic significance also in these patients. EFS was 0.65±0.16 and CIR 0.35±0.18 in 9 children with MRD ≥0.1 compared with EFS 0.94±0.02 and CIR 0.05±0.02 in 165 patients of the low MRD group (p<0.001 for both EFS and CIR).

In ImRG, MFC MRD data disclosed a prognostic value, as well. EFS was 0.92±0.03 and CIR 0.06±0.02 for 143 patients with MRD-negativity on day 36; in contrast, outcome of 78 MRD-positive patients was significantly inferior (EFS 0.73±0.06, CIR 0.25±0.06, p<0.001 for both comparisons). Like in SRG, in ImRG the distribution of MRD levels was different between two randomization arms. In PEG-DNR+ group (n=93) only 54.8% of patients were MRD-negative at the end of remission induction. In contrast, in PEG+DNR+ group (n=115) the proportion of patients with favorable MRD results was significantly higher (76.5%, p=0.001). Again even with low number of MRD-positive cases in PEG+ arm, day 36 MRD disclosed a significant influence on outcome: In 88 MRD-negative patients EFS was 0.93±0.03 and CIR 0.05+0.03, significantly better than in 27 MRD-positive patients (EFS 0.80+0.08 and CIR 0.17+0.08, p=0.025 and p=0.019, respectively).

Conclusions. Administration of PEG at day 3 of remission induction leads to a more rapid MRD response in children both with standard-risk and intermediate-risk BCP-ALL. In SRG, rapidity of MRD clearance in PEG+ arms was independent of the additional administration of daunorubicin. We conclude that end-induction MRD data are independent predictors of outcome despite the small proportion of MRD-positive patients after the implementation PEG during induction.