Baseline Mutations Lack Impact on Clinical Outcomes and Molecular Response in Core Binding Factor Leukemia Treated with Highly Effective Regimen

Background: Several mutation classes e.g. epigenetic (e.g. ASXL2), kinase (e.g. KIT, JAK2 or 3), cohesion etc. have been implicated with poorer outcome of patients with core binding factor acute myelogenous leukemia (CBF-AML). Most of these patients have been treated with a cytarabine and anthracycline ‘3+7’ based induction and high dose cytarabine based consolidations. Fludarabine, high dose cytarabine and G-CSF (FLAG) based regimens have shown better relapse free survivals indicating deeper remissions. Presence of unique transcripts render CBF-AML amenable to be monitored by quantitative reduction of transcripts. Limited information is available about correlation of optimal transcript reduction with baseline mutation classes among CBF-AML patients treated with a highly effective regimen.

We reported RTPCR values of ≤0.1 at end of induction and ≤0.01 at 2-3 months of remission as optimal transcript reduction in bone marrow for our patients treated with FLAG based regimen.

Method: We analyzed 88 consecutive patients with CBF-AML treated with a FLAG based regimen who also had baseline mutation analysis of a minimum of 24 gene myeloid leukemia relevant next generation sequencing based panel and serial CBF-AML specific fusion transcript monitoring by reverse transcriptase polymerase chain reaction (RTPCR). Mutation classes analyzed were epigenetic (TET2, IDH1 and 2, DNMT3A, ASXL1 and 2, EZH2), transcription factor (RUNX1, CEBPA, NPM1, MLL, NOTCH1), signal transduction (FLT3, NRAS, KRAS, KIT, JAK2, PTPN11). Cohesin mutation data was available from 30 most recently treated patients.

Results: Median age was 51 years (range, 22-78 years), 80 patients had idarubicin added to the FLAG regimen (FLAG-Ida) and 8 patients had gemtuzumab ozogamicin added to FLAG (FLAG-GO) after reapproval of GO in the US. With a median follow up of 4+ years, median overall survival (OS) is not reached (NR), with a 3 year OS of 75%. Median remission duration (68% at 3 years) and relapse free survival (RFS) (60% at 3 years) has not been reached.

Mutations in signal transduction pathways were seen in 61%, epigenetic regulators in 41% and transcription factors in 22%. Presence of signal transduction (p=0.4), epigenetic (p= 0.8) and transcription factor (p=1) did not predict for optimal CBF-AML specific transcript reduction either at end of cycle 1 of treatment or at 2-3 months of remission. While signal transduction mutations did not predict for RFS, presence of RAS mutations trended for better RFS (p=0.06, median NR versus 43 mos) and KIT mutations (p=0.4) did not impact RFS.

Conclusions: Impact of adverse mutations on outcomes in CBF-AML can potentially be overcome with effective treatment regimen.