P575: IMPACT OF MOLECULAR RESPONSE AND CHEMOTHERAPY REGIMEN ON OUTCOMES IN CORE BINDING FACTOR AML

Abstract

Background: Fludarabine based induction/consolidation regimens result in excellent outcomes in core binding factor acute myeloid leukemia (CBF-AML) Attainment of early polymerase chain reaction (PCR) based optimal minimal residual disease (MRD) responses positively impact long term outcomes and provide early decision timepoints for alternative therapies including stem cell transplantation (SCT). Aims: We studied factors that affected early and late optimal PCR responses (OPR) and survival in patients (pts) of CBF AML who were treated with fludarabine based regiemns (fludarabine, cytarabine and GCSF) with idarubicin (FLAG-IDA) or gemtuzumab (FLAG-GO). Methods: This is a prospective study of pts with N/D CBF-AML treated at our center on FLAG based regimens. Pts received up to 7 cycles of therapy with appropriate dose reductions based on age or toxicity. PCR for the disease defining transcripts (CBFB::MYH11 or RUNX1::RUNX1T1) were monitored after C1, C3 and end of therapy (EOT). Based on our earlier report (Boddu et. al., Leukemia 2018) OPR was defined as PCR transcript <0.1% after C1 and <0.01% for later time points. Results: Between 4/2007 to 12/2019, 174 pts, with a median age of 51 years (range, 19-78) were included. A total of 109 pts (63%) received FLAG-IDA, while the rest got FLAG-GO therapy. This was mostly due to withdrawal of approval of GO in the US for a period of time. Median number of cycles received was 6 (range, 1-7) Baseline parameters are described in Table 1. PCR data is available for 94% pts after C1, 90% after C3 and 92% at EOT. EOT was considered the timepoint at which pt stopped further FLAG based therapy beyond C1. The percent of patients achieving OPR at each time point by regimen are: 68% vs. 41% (p=0.002) after C1, 75% vs. 39% (P< 0.0001) after C3 and 92% vs. 55% (p<0.0001) at EOT for FLAG-GO and FLAG-IDA respectively. On univariate analysis, FLAG-GO regimen favored attainment of post C1 OPR over FLAG-IDA (OR, odds ratio=3.1, 95%CI 1.6-6.2) while age, baseline BM blast, cytogenetic subtype of CBF, ACA and kinase mutations did not. The benefit of FLAG-GO in attaining OPR after C1 was maintained even on a multivariate (MV) model. For EOT OPR, FLAG-GO (OR= 8.5, 95% CI 3.4-26) fared better over FLAG-IDA. Absence of KIT mutation (OR=2.4) and attainment of post C1 OPR (OR= 11.3) were the other relevant factors. On MV analysis, only chemo arm and post C1 OPR remained significant. At a median follow up of 92 months (mos), the RFS was 124 mos and OS not reached (NR) for the whole cohort (RFS 68 mos and OS 102 mos for FLAG-IDA treated pts vs. both NR for FLAG-GO treated pts, p=0.004 and 0.04 respectively). Across the 2 groups there was no difference in RFS or OS between pts who attained early or late OPR. However, on stratifying serial PCR responses with therapy arms, even in pts with early OPR (after C1), RFS was better (NR) in FLAG-GO arm vs 101 mos for FLAG-IDA arm (p= 0.03), though OS was similar, likely reflecting efficacy of salvage therapies in this subset of AML. In a Cox proportional hazards model that included age, therapy arm, kinase mutations, cumulative therapy cycles, post C1 and EOT PCR responses; only younger age and EOT OPR favorably affected RFS; however FLAG-GO treated pts had lower hazards for RFS at any level of PCR response compared to FLAG-IDA treated pts (Fig. 1). Image:Summary/Conclusion: In CBF AML, FLAG-GO regimen leads to higher rate of earlier and end of therapy OPR than FLAG-Ida and substantially better RFS. At any combination of PCR response, FLAG-GO treated patients had lower hazards for relapse than FLAG-IDA treated pts.