Clinical Outcome and Non-Synonymous Mutation Clearance in Chinese Children with Acute Myeloid Leukemia Treated with a Low-Intensity Induction Chemotherapy Regimen

Administration of near-myeloablative courses of chemotherapy followed by hematopoietic stem cell transplantation (HSCT) is considered the standard of care for children with acute myeloid leukemia (AML). However, this approach is associated with severe impairment of hematopoiesis and immune response, high rate of toxic death, and a high economic burden for families. Recent genomic studies show that the failure associated with treatment is due to the selection of chemotherapy-resistant clones present at diagnosis. In this study, we (1) report the outcomes of children with AML who received low-intensity induction chemotherapy, and (2) evaluate the quality of remission by using next-generation sequencing.

From July 2012 to May 2016, 90 Chinese children with AML who had white blood cell counts less than 50.0×10⁹/L were treated at a single center. Of them, 31 (Group I) received dose-reduced induction chemotherapy because they had an active infection or had socioeconomic constraints. The remaining 59 children (Group II) received standard-dose chemotherapy. The induction regimen in Group I consisted of homoharringtonine 1 mg/m², i.v., once daily, 7 doses; cytarabine 10 mg/m², i.v., q12 h, 20 doses; and G-CSF 5 µg/kg, i.v., once daily, 10 doses (HAG). On day 28 of induction, Group I patients with <10% blasts received another course of HAG and the remaining patients received mitoxantrone 5 mg/m², once daily, 3 doses; cytarabine 10 mg/m², i.v., q12 h, 20 doses; and G-CSF 5 µg/kg, i.v., once daily, 10 doses (MAG). Group II patients were treated with homoharringtonine 3 mg/m², i.v., once daily, 7 doses; cytarabine 100 mg/m², i.v., q12 h, 20 doses; and etoposide 100 mg/m², i.v., once daily, 5 doses (HAE). Group II patients received the second induction consisting of daunorubicin 50 mg/m² i.v., once daily, 3 doses; cytarabine 100 mg/m², i.v., q12 h, 20 doses; and etoposide 100 mg/m², i.v., once daily, 5 doses (DAE). Post-remission therapy was the same for Group I and II patients (cytarabine/mitoxantrone; etoposide/cytarabine; high-dose cytarabine/L-asparaginase). Patients at high risk of relapse underwent HSCT. Median age, white blood cell count, FAB subtype, and favorable genotypes did not significantly differ between Group I and II patients. Complete remission, based on morphology, was 71.0% and 74.6% (P=0.71) in Group I and II patients, respectively. The 2-year event-free survival was also similar in both groups (66.7%±11.5% vs. 70.5%±6.6%, respectively; P=0.565). Moreover, there was a significant reduction in the number of Grade III or higher treatment-related adverse events in Group I compared with Group II patients (8.2±2.0 vs. 18.0±1.0, P<0.0001).

Whole-exome deep sequencing was performed for 20 patients who achieved morphologic remission. The variant allele frequency (VAF) of non-synonymous mutations (median, 5 mutations; range 2-11 mutations) was analyzed at day 28 of each induction course. After induction I, the average reduction in VAF for all mutations was 76.3% and 99.0% in Group I and II patients, respectively. After induction II, of 16 patients analyzed, all mutations cleared under the VAF threshold of 2.5% in 7 of 8 patients in Group I and 6 of 8 patients in Group II. 77.4% and 69.4% of mutations were undetectable in Group I and II patients, respectively. At the molecular level, both regimens achieved similar clearance efficiencies, but patients given the low-intensity regimen had a slower clearance than those given the high-intensity regimen. Interestingly, exome sequencing of the primary and relapsed tumors from 1 patient in each cohort revealed that the dominant clone remained the same as that at diagnosis in the patient treated with the low-intensity regimen but changed in the patient treated with the high-intensity regimen.

Our preliminary results suggest that children treated with HAG/MAG for induction, followed by conventional post-remission strategies, have similar outcomes as those treated with the high-intensity induction regimen. Although patients with active infections and a high financial burden were recruited to Group I, our clinical and genomic analyses suggest that HAG/MAG represents a valuable alternative to conventional chemotherapy and can be used in the general pediatric population with AML. The molecular signature of patients responsive to HAG/MAG and its prognostic implications remain to be determined.