Outcomes of Children with First Marrow Relapse: Results from Children’s Oncology Group (COG) Study AALL01P2.

Treatment of relapsed acute lymphoblastic leukemia (ALL) remains a significant challenge. Objectives of the COG AALL01P2 study were: a) to define a clinically acceptable re-induction platform to which novel agents might be added and b) to track response and minimal residual disease (MRD) through 4 months of treatment. The initial regimen consisted of 3, 35-day blocks of chemotherapy (block 1: vincristine, dexamethasone, PEG-asparaginase, idarubicin, and intrathecal cytarabine and methotrexate (CNS-negative), or triple intrathecal therapy (CNS-positive); block 2: etoposide, cyclophosphamide, 5 g/m² methotrexate and intrathecal methotrexate (CNS-negative), or triple intrathecal therapy (CNS-positive); block 3: high dose cytarabine and L-asparaginase). Patients with Philadelphia chromosome positive (Ph+) ALL received imatinib mesylate in combination with all 3 blocks of chemotherapy. AALL01P2 opened in January 2003. The study was suspended in May 2003, however, due to unacceptable toxicity among the first 21 patients. Block 1 therapy was subsequently modified substituting doxorubicin for idarubicin, and prednisone for dexamethasone, and the study reopened accruing a total of 127 eligible patients (120 B-precursor-ALL, 7 T-ALL) from August 2003 through December 2005. Seventy-two patients had early marrow relapse (< 36 months from initial diagnosis) (ER) and 55 patients had late marrow relapse (LR). Four B-lineage patients were Ph+. Subsequent analyses include only patients enrolled following the modification in block 1 therapy. Complete remission (CR2) rates at the end of block 1 were 67% ± 5.9% for ER (n=63) and 96% ± 2.8% for LR (n=48). Patients with ≥ 5% residual blasts (M2 or M3 marrows) at the end of block 1 continued on protocol therapy. However, only 2 of 13 B-lineage patients with M2/3 marrows at the end of block 1 responded to subsequent chemotherapy. Event-free survival (EFS) at 4 months was 64% ± 6.7% for ER (n=72) and 93% ± 4.3% for LR (n=55). Outcome for T-ALL patients (n=7, 6 ER) was poor with complete remission achieved in only 2 of 7 patients. Toxicity was acceptable during all 3 blocks. The most common toxicities were hematopoietic and infectious, with the highest incidence during blocks 1 and 3. Five toxic deaths occurred (4%), all from infections: 3 during block 1, and 2 during block 3. MRD was measured by flow cytometry at the end of each block. MRD was detected in 79% ± 6.2% of ER (n=43), vs. 50%± 8.3% of LR (n=36) at the end of block 1 in an analysis restricted to B-lineage, CNS-negative, Ph-negative patients, regardless of their morphological response. Four month EFS was 95% ± 3.5% vs. 78% ± 5.3% in patients who were MRD-negative vs. positive at the end of block 1 (p<0.01). The AALL01P2 regimen is a tolerable and active re-induction platform, acceptable for testing in combination with novel agents. The COG ADVL04P2 study, which uses this regimen in combination with epratuzumab (anti-CD22), is currently underway. MRD measurements are also feasible at serial time points during re-induction, and early response is prognostic. Alternative strategies are urgently needed for patients with relapsed T-ALL and for early marrow relapse.