Prospective assessment of measurable residual disease kinetics in intensively treated patients with Acute Myeloid Leukemia (AML) - results of the PALG-AML1/2016 study

Abstract

Abstract BACKGROUND: Measurable residual disease (MRD) is an important biomarker in AML. Detectable MRD (MRD+) after 2 cycles of intensive chemotherapy, at the end of consolidation and/or before allogeneic stem cell transplantation (alloHSCT) predicts relapse and shorter overall survival (OS) (Short NJ, JAMA Oncol. 2020). However, whether eradicating MRD through subsequent chemotherapy improves outcomes remains an open question. AIM: The aims of this analysis were to evaluate factors predictive of MRD- remission, the kinetics of multiparametric flow cytometry (MFC) MRD during post-remission treatment, and their influence on outcomes. PATIENTS AND METHODS: In the PALG-AML1/2016 phase 3 open-label randomized trial (NCT03257241), patients (pts) with newly-diagnosed, untreated AML, ECOG performance status 0–2 and HCT-CI≤ 3 were randomized to DA-90 (n=220) or DAC (n=219) induction chemotherapy (IC). Pts with >10% blasts in non-aplastic bone marrow at day 14 received early second IC. Pts with CR/CRi/CRp (cCR) were offered IDAC consolidation with or without alloSCT according to predefined risk groups. Serial samples for MRD assessment were collected at cCR after 1 or 2 IC (MRD1), and after each consolidation cycle (MRD2-4). MFC MRD evaluation using LAIP-based analysis was performed by local laboratories (n=16) and subsequently reviewed by an independent central reviewer. MRD was assessed according to ELN recommendations using the threshold of ≥0.1% for MRD positivity. RESULTS: Between 2017 and 2023, 439 pts, median (med) age 47 (18-60) years were enrolled at multiple centers across Poland and at Weill Cornell Medicine in New York City; 428 pts received initial IC and 51 (11.6%) received early second IC per protocol. The overall cCR rate after 1 or 2 IC was 72.4% with no difference between treatment arms (p=0.338). After med follow-up of 55.9m, 5-year OS was 50.6% (48.6% in DA-90 vs. 51.8% in DAC; HR 1.01 [0.77-1.33], p=0.921). Centrally reviewed MRD data were available for 173 (54.4%) pts in cCR after 1 or 2 induction cycles. Reasons for missing data included lack of LAIP (n=22), missing samples (n=9), inaccessible FCS files (n=60), and sample processing issues (n=54). The overall rate of cCR MRD negativity (MRD-) was 89.1% (89.4% vs. 88.9%; p=0.92 for DA-90 and DAC, respectively). 11/174 pts underwent alloSCT after IC and did not have MRD2 evaluation. Centrally reviewed results of both MRD1 and MRD2 were available for 131 pts. After 1st consolidation, 116/131 (89%) pts did not change MRD status; 109 pts remained MRD- and 7 (5.3%) remained MRD+. Conversion from MRD+ to MRD-was noted in 9 pts (6.9%) and from MRD- to MRD+ in 6 (4.6%) pts. FLT3-ITD mutation and complex karyotype were associated with lower probability to achieve MRD1- and MRD2- respectively in uni- and in multivariable analysis. OS was significantly impacted by MRD kinetics, with med OS not reached (NR) in MRD1-/MRD2-; 9.9 mos in MRD1-/MRD2+; NR in MRD1+/MRD2- and 27.8 mos in MRD1+/MRD2+ (p<0.01). MRD1-/MRD2+ conversion was associated with significantly shorter OS compared to other groups. Similarly, med RFS was NR in MRD1-/MRD2-, 3 mos in MRD1-/MRD2+, 14.5 mos in MRD1+/MRD2- and 9 mos in MRD1+/MRD2+ (p<0.001). As expected, significantly shorter RFS was observed in MRD1-/MRD2+ group, compared with MRD1-/MRD2- (p<0.001) and MRD1+/MRD2- (p<0.004) pts. In a multivariable models with time-varying alloSCT status, de novo AML (HR=0.28; p<0.001), higher baseline WBC (HR=1,01; p=0.046), ELN 2022 intermediate-risk group (HR=2.72; p<0.001), and MRD2+ (HR=2.32; p=0.034) were significant predictors of OS, while de novo AML (HR=0.32; p<0.001), baseline WBC (HR=1.01; p=0.005), ELN 2022 high (HR=2.62; p<0.001) or intermediate-risk group (HR=3.42; p<0.001), alloSCT in CR1 (HR=0.42; p<0.001) as well as MRD2+ (HR=2.05; p=0.012) were independent predictors for RFS. CONCLUSIONS: Data from PALG-AML1/2016 confirm the prognostic value of MRD after 1 consolidation for both OS and RFS. Loss of MRD negativity after consolidation 1 (early MRD relapse) is observed in a proportion of patients and is associated with significantly shorter OS and RFS.