Post Induction Minimal Residual Disease Levels Identifies a Group of High Risk Relapsed Childhood Acute Lymphoblastic Leukemia (rALL) with a Favorable Outcome Independent of Induction Therapy

BACKGROUND: The determinants of outcome in children with rALL are the duration of first remission (CR1), site of relapse and immunophenotype. High risk (HR) relapses are defined as those occurring with a CR1 of <18 months; B-cell precursor (BCP) with bone marrow (BM) relapse within 6 months of stopping therapy and T-cell BM or combined relapses at any time. All other relapses are defined as standard risk (SR). In the UKALLR3 clinical trial for rALL, HR patients had a lower CR2 rate, higher post induction MRD and inferior survival when compared to SR patients treated in identical fashion. We investigated the effect of further intensifying induction therapy with clofarabine in HR patients.

METHODS: Clofarabine was added to the UKALLR3 consolidation block of cyclophosphamide, etoposide (CCE) and used as induction therapy, with dexamethasone and PEG-Asparaginase for HR patients. The previous induction block with mitoxantrone (M) was given as consolidation and all patients were eligible for stem cell transplantation (SCT) with any donor after a third intensification block. The outcomes assessed were improvements in CR2, MRD and progression-free survival (PFS) when compared to historical controls of patients receiving idarubicin (I) or M induction in UKALLR3. A Fleming-style design, based on observed response and toxicity, was incorporated to allow an increase in the dose of cyclophosphamide from 300 mg/m² to 440 mg/m².

RESULTS: 61, 39 at lower and 22 at the higher dose of cyclophosphamide,  CCE patients were compared to 30 I and 69 M patients with HR rALL. Patients in the CCE group had a lower median age at presentation, but other prognostic variables were comparable. CR2 rates of 73%, 83%, 71% and low MRD (≤10^-4) was seen in 32%, 0%, 25% of CCE, I and M groups. The higher cyclophosphamide dose was associated with improved CR rates, lower MRD but also increased toxicity levels in CCE compared to M group patients.  The proportions of patients reaching transplantation were 43%, 60% and 55% of CCE, I and M patients respectively. 73/82 eligible patients received a SCT, 48 (66%) with matched and 25 (34%) with mismatched donors. The 2-year PFS with CCE, M and I regimens were 17% (11,23), 27% (19,34) and 30% (25,36) respectively (p=0.08). Outcomes of matched sibling, matched unrelated and mismatched SCT were comparable (p=0.9). Seventeen patients with a post induction MRD<10^-4, had a 2-year PFS of 63% (50,75), compared to 21% (15,27) for 53 patients with MRD≥10^-4 and 21% (17, 25) for the 90 patients with unknown MRD (p=0.005).  All 4 patients with MRD≥10^-3prior to SCT and 8/9 not transplanted suffered a second relapse. Overall outcomes of BCP (2-year PFS 21% (15,28)) and T-cell ALL (2-year PFS 26% (16,35)) were comparable (p=0.9). PFS in BCP-ALL was 31% (24,38) and 13% (6,20) (p=0.1) for those receiving M and CCE respectively.

CONCLUSIONS: We define two groups of HR rALL patients based on MRD levels attained post induction, independent of the induction regimen. Approximately a quarter of HR patients continue to have chemosensitive disease as evidenced by rapid MRD clearance (<10^-4 at week 5). This group includes high-risk cytogenetics and T-cell rALL with MRD as the single discriminatory factor for outcome. These patients have a favorable outcome after SCT with any donor. In the other group (MRD≥10^-4) over half of HR patients do not reach SCT primarily due to refractory disease (27%) or disease recurrence (14%).  One third of patients relapse post SCT. For this group novel agents and newer treatment strategies are urgently required.