Therapy related core binding factor Acute Myeloid Leukemia (CBF-AML) responds favorably to fludarabine, cytarabine, GCSF (FLAG) based therapy: Exploratory Analysis from a Phase 2 trial

Abstract

Abstract Background: FLAG based therapy, especially with gemtuzumab ozogamicin (GO) has led to excellent relapse free (RFS) and overall survival (OS) among patients with CBF-AML. We previously reported inferior outcomes among patients with therapy related CBF-AML (T-CBF) (after exposure to prior chemo-radiotherapy for non-myeloid cancers) compared to de novo CBF-AML (DN-CBF)(Borthakur et al, Cancer 2009), but their outcomes after FLAG based therapy is not known. Methods: We did an exploratory analysis of the ongoing phase 2 clinical trial of FLAG therapy in newly diagnosed adult patients (≥18 years) with CBF-AML. GO is added to FLAG during induction and up to 2 cycles in consolidation, however for the period from 2010-2017, when GO was held in US, idarubicin (IDA) was added to FLAG. Pts were prospectively monitored for response with quantitative polymerase chain reaction (qPCR) of CBF-specific transcripts; a qPCR <0.1% after induction and <0.01% during/after consolidation have been identified as optimal PCR response (OPR) (Boddu et al, Leukemia 2018; Senapati et al AJH, 2022). Relapse free survival (RFS) was tabulated from the time of first response to relapse/death and overall survival (OS) from treatment initiation to death; both censored for last follow up. Results: From April 2007 to Dec 2024, 219 pts with CBF-AML were treated on the trial (NCT00801489), among whom 32 (15%) pts had T-CBF. The median age of these patients at AML diagnosis was 64 years (range 25-80) and 19 patients (59%) were ≥60 years of age. Among the patients with T-CBF, 17 (53%) had received both chemotherapy and radiotherapy (RT), 9 (28%) only chemotherapy and 6 (19%) only RT for prior non-myeloid cancers. Fourteen (44%) patients had RUNX1::RUNX1T1 and 18 (56%) had CBFB::MYH11 subtype of CBF-AML. Fifteen (47%) patients were treated with FLAG-GO and the rest with FLAG-IDA. Sixteen patients (50%) had additional cytogenetic anomaly (most common being trisomy 8). Based on available mutational (mut) profile, 6/26 (23%) patients had a KIT mut, 9/30 (30%) patients had a RAS mut, 1 patient each had a FLT3-ITD and FLT3-TKD mut, and no patients had a TP53 mutation. A best response of composite complete response (CRc= CR+CRi) was achieved by all patients, all after the first cycle. An OPR was achieved post induction in 13 (41%) patients which included 8/15 (53%) patients treated with FLAG-GO and 5/17 (29%) treated with FLAG-IDA, p=0.28. Among patients who received ≥3 cycles (C) of therapy and with serial qPCR data (n=19), end of C3 OPR was achieved in 12/19 (63%) patients; 8/9 (89%) treated with FLAG-GO and 4/10 (40%) treated with FLAG-IDA, p=0.05. A best response of OPR during/after consolidation was achieved by 19/25 (76%) patients, 11/12 (92%) treated with FLAG-GO and 8/13 treated with FLAG-IDA, p=0.16. At a median follow-up of 112.9 months (95% CI 64.2-143.1 months), the median RFS and OS was 113.0 months (95% CI 72.2-NR) and 115.8 months (95% CI 75.2-NR) while 5-year rates were 74.7% and 74.2% respectively. The 5-year RFS and OS rates was 81.5% vs. 70.6% (p=0.78) and 80.8% vs. 70.5% (p=0.84) among patients treated with FLAG-GO and FLAG-IDA respectively. At last follow-up, 6 patients have relapsed (and died); among the remainder 26 patients, 7 have died (non-relapse mortality), 19 are alive and in CR (4 post allo-HSCT, and one on ongoing study therapy). Among the 187 patients with de novo CBF-AML (DN-CBF) in this cohort, the median age was 47 (range 19-80) and significantly fewer patients (47 [25.1%]) were ≥60 years of age than T-CBF, p<0.01. At a median follow up of 69.4 months (95% CI 51-5-112.4), the median RFS and OS of the full DN-CBF group were both NR, 5-year rates were 65.9% and 74.2% respectively and similar to patients with T-CBF AML (p=0.70 and 0.27). Among patients with DN-CBF treated with FLAG-GO, the 5-year RFS and OS were 75.9% and 79.9% and similar to T-CBF patients treated with FLAG-GO (p=0.49 and 0.33 respectively). Five of 32 (15.6%) patients with T-CBF underwent allo-HSCT in ongoing remission after study therapy (with/without intervening therapy) compared to 11/187 (5.9%) patients with DN-CBF. Conclusion: Among adult patients with CBF-AML, treatment with FLAG based therapy leads to comparable survival between T-CBF and DN-CBF-AML. FLAG-GO trends towards higher rates of OPR in T-CBF AML while RFS and OS appear similar with FLAG-GO and FLAG-IDA. Even among FLAG-GO treated patients the RFS and OS of T-CBF and DN-CBF pts appear comparable.