Gemtuzumab Ozogamicin Added to Fludarabine, Cytarabine and G-CSF (FLAG-GO) Leads to Superior Molecular Response and Survival Outcomes Than Idarubicin (FLAG-IDA): 200 Patients Long-Term Follow up

Background:

CBF-AML are a favorable risk subtype of AML having 5-year overall survival (OS) of over 50% with 7+3 regimen. Fludarabine, Cytarabine and G-CSF (FLAG) dual nucleoside analogue based therapy have led to superior survival outcomes with 5-year OS >70%. Assessment of measurable residual disease (MRD) through quantitative estimation of specific CBF defining polymerase chain reaction (PCR) transcripts enables precise disease monitoring, can help determine need for allogeneic stem cell transplantation (SCT) and overall prognosis. Optimal PCR response (OPR) to therapy is defined by transcript <0.1% after cycle (C) 1 (induction) and ≤0.01% during and after consolidation. Gemtuzumab ozogamicin (GO) added to FLAG showed better cadence of PCR response than Idarubicin (IDA) and superior RFS (Senapati et al AJH, 2023). We report here the long-term follow up (FU) from this ongoing phase 2 trial.

Methods:

The treatment regimen has been published before (NCT00801489). Adult patients (pts) with CBF-AML were eligible for frontline therapy. Up to 7 cycles of FLAG based therapy was administered. For the period when GO was withdrawn from the US market, the regimen was modified to add IDA to FLAG (FLAG-IDA). Pts could have received up to one prior cycle of therapy before enrollment. These pts were excluded form post C1 OPR analysis. PCR was done from the peripheral blood or bone marrow after C1, C3 and after end of therapy (EOT). For pts who received <7 cycles of therapy, the last PCR on study was considered as EOT PCR.

Results:

From April 2007-Dec 2023 200 pts were treated on this study. The median (med) age of pts was 52 years (19-80 years), 59 (30%) pts were ≥60 years of age and 95 pts (48%) were female. 92 pts (46%) were treated with FLAG-GO and 108 pts (54%) with FLAG-IDA. 109 pts (55%) had Inv16/t(16;16) [FLAG-GO=50, FLAG-IDA=59] and 91 pts (45%) had t(8;21) [FLAG-GO=42, FLAG-IDA=49]. Amongst 186 evaluable pts, 84 (45%) had an additional cytogenetic abnormality (ACA) (38/86 [44%] in GO and 46/100 [46%] in IDA arm]. 192 pts had assessment for KIT mutation 38 (20%) of who were mutated (18/87 [21%] in GO and 20/105 [19%] in IDA arm); 70/180 (39%) evaluated pts were RAS mutated (37/76 [49%] in GO and 33/104 [32%] in IDA arm).

A best response of CRc (CR+CRi) was attained in 197 pts (99%) and 3 pts died before response evaluation (all in the FLAG-GO arm; on D1, D2 and D16 of therapy). Amongst evaluable pts (n=166), post C1 OPR was attained in 80 pts (48%), 49/77 (64%) vs. 31/89 (35%) in the GO and IDA arms resp., p<0.01. Amongst evaluated pts, post C3 OPR was attained in 70/134 pts (52%) (50/66 [76%] with GO vs. 20/68 [29%] with IDA, p<0.01) and EOT OPR was attained in 127/170 pts (75%) (70/76 [92%] vs. 57/94 with IDA [61%], p<0.01). On multivariate (MV) logistic regression analysis (adjusting for age<60/≥60, RAS/KIT mutations and type of CBF-AML), absence of KIT mutation (OR=5.1, 95% CI 1.8-14.2) and treatment with FLAG-GO (OR=15.9, 95%CI 4.6-55.1) favored odds of EOT OPR.

At a med estimated FU of 102 mos (95% CI 89-114 mos), 73 mos for the FLAG-GO arm and 104 mos for the FLAG-IDA arm, the med RFS and OS were both not reached (NR) (8-year RFS=62%; OS=67%) for the full cohort. The med RFS was NR vs 110 mos (5-year RFS 77% vs. 59%, p=0.007) and OS was NR vs 116 mos (5-year OS 79% vs. 69%, p=0.03) for FLAG-GO and FLAG-IDA arms resp. Stratified by age, the med OS was NR for both arms in pts <60 years of age (5-year OS= 85% vs. 75% in the FLAG-GO and FLAG-IDA arms resp., p=0.04); for pts ≥60 years of age it was 123 mos vs. 35 mos (5-year OS=69% vs. 49%, p=0.04) resp. Amongst pts who attained an EOT OPR (n=127), the med RFS were both NR (5-year RFS 85% vs. 71%, p=0.03) and med OS also both NR (5-year OS 88% vs. 80%, p=0.06) for FLAG-GO and FLAG-IDA treated pts respectively, but OS was still superior with FLAG-GO; On competing event analysis (cumulative incidence of relapse [CIR] and non-relapse mortality), the 5-year CIR in pts with EOT OPR was 14% vs. 28% resp (Gray’s p=0.04) for FLAG-GO and FLAG-IDA resp. On MV Cox analysis stratified by age and adjusting for the type of CBF-AML, KIT/RAS mutations and treatment arm, therapy on FLAG-GO independently reduced hazards of relapse (HR=0.46, 95%CI 0.27-0.80) and death (HR=0.47, 95%CI 0.25-0.84).

Conclusion

On long-term F/U FLAG-GO leads to superior OPR rates and survival outcomes than FLAG-IDA. Even in pts with EOT OPR, FLAG-GO leads to better RFS and OS than FLAG-IDA. KIT mutation had no independent prognostic impact on survival in our analysis.