Impact of Flow Cytometry-Based Measurable Residual Disease Clearance Kinetics in Patients with Acute Myeloid Leukemia Undergoing Intensive Chemotherapy

Background: Measurable residual disease (MRD) in acute myeloid leukemia (AML) is highly prognostic. However, the optimal timepoint for achieving undetectable MRD is unclear. Whether early MRD dynamics can inform decisions regarding allogeneic stem cell transplant (SCT) in first remission (CR1) is also controversial.

Methods: We retrospectively evaluated patients (pts) with newly diagnosed AML treated with intensive chemotherapy at our institution. Pts must have achieved CR or CRi after 1 cycle of induction (time point 1; TP1) and had an evaluable multiparameter flow cytometry (MFC)-based MRD assessment at both TP1 and after 1-2 cycles of consolidation (TP2). Pts with core-binding factor AML, acute promyelocytic leukemia or treated secondary AML (i.e. those who received prior therapy for antecedent myeloid malignancy) were excluded. The sensitivity of the MFC MRD assay was 0.1-0.01%; MRD negativity was defined as absence of any detectable residual disease by MFC.

Results: Between 4/2010 to 8/2021, 277 pts with newly diagnosed AML achieved CR/CRi and had an evaluable MRD assessment at TP1 and TP2. The median age of the cohort was 53 years (range, 19-82 years). 42 (15%), 139 (50%) and 96 (35%) pts were classified as ELN 2022 favorable, intermediate, and adverse risk, respectively. Pts were classified by MRD response at TP1/TP2 as Neg/Neg (n = 187 [68%]), Pos/Neg (n = 43 [16%]) or Pos/Pos (n = 47 [17%]).

The median duration of follow-up was 46 months. Pts in the Neg/Neg, Pos/Neg and Pos/Pos groups had a 5-year RFS of 53%, 33%, and 13%, respectively (P< 0.01). Notably, RFS was significantly worse for Pos/Neg vs Neg/Neg pts (P=0.05), suggesting an RFS benefit to achievement of MRD negativity after induction, rather than later in consolidation. The 5-year OS rates were 54%, 48%, and 28%, respectively. However, the difference between Neg/Neg and Pos/Neg pts was not significant (P=0.19).

The proportion of pts undergoing allogeneic SCT in CR1 was similar among the MRD groups (63%, 61% and 47% in the Neg/Neg, Pos/Neg and Pos/Pos groups, respectively; P=0.27). MRD dynamics were predictive for RFS in non-transplanted pts, but not in transplanted pts. In non-transplanted pts, the median RFS was 17 and 8 months for Neg/Neg and Pos/Neg pts, respectively (P=0.02), while among transplanted pts, the difference between these groups was not significant (median RFS: 117 months vs not reached; P=0.41).

In a landmark analysis, SCT benefitted all analyzed subgroups, regardless of MRD dynamics. Pts with intermediate/adverse-risk AML and early MRD clearance (Neg/Neg) had superior RFS and OS with SCT (P<0.01 for both). Importantly, the benefit of SCT was also observed even when the analysis was restricted to only intermediate-risk pts, where the median RFS for SCT vs. no SCT was not reached and 15 months, respectively (P<0.01) and the median OS was not reached and 28 months, respectively (P<0.01).

By multivariate analysis, early attainment of MRD negativity was independently associated with superior RFS but not OS. The hazard ratio (HR) for relapse or death for Pos/Neg vs Neg/Neg pts was 1.73 (95% CI, 1.09-2.75; P=0.02). However, early MRD negativity was not independently prognostic for OS (HR for death for Pos/Neg vs. Neg/Neg: 1.45 (95% CI, 0.88-2.38; P=0.15). Factors that independently predicted for OS were age (P=0.02), TP53 mutation and/or complex cytogenetics (P<0.01) and SCT in CR1 (P<0.01).

Conclusions: Among pts with AML receiving frontline intensive chemotherapy, undetectable MFC MRD after induction is independently associated with better RFS compared with undetectable MRD achieved later during consolidation. However, early MRD kinetics did not significantly impact OS, which was driven primarily by age, cytomolecular risk and SCT consolidation. Among pts with intermediate-risk AML—a group were SCT decisions are often individualized—SCT improved outcomes irrespective of MRD kinetics, suggesting that MFC MRD may not be adequate to inform SCT decisions in these pts.