Molecular response of Acute Myeloid Leukemia to induction with CPX-351 and its impact on prognosis

Abstract

Abstract Background: CPX-351 has emerged as a preferred induction therapy for patients (pts.) with acute myeloid leukemia (AML) with myelodysplasia-related changes and therapy-related AML, demonstrating superior outcomes compared to standard 7+3. Initial response assessment in clinical routine is done by bone marrow cytomorphology after 1st induction. However, genetic mutations and cytogenetic abnormalities significantly impact remission rates and may have predictive value. Molecular profiling post-induction is essential for refining prognosis, guide consolidation strategies, and inform the potential for allogeneic stem cell transplantation (allo-tx). Aim: This study aims to describe the molecular response of pts. with AML to induction with CPX-351 and its impact on prognosis. We assess the concise genetic background of the disease with a focus on molecular aberrations and pursue the response of these mutations to treatment. Furthermore, we correlate to survival data. Methods: We screened patient charts for induction with CPX-351 at our tertiary medical center. Pts. were included in the study if they had obtained initial NGS workup with a panel of up to 53 genes and at least one more NGS analysis after induction and excluded if either of those were missing. Treatment, response and survival data were retrieved by chart review. Molecular data was provided by our genetical laboratory INDIGHO. For molecular response, we defined complete molecular remission (CMR) as variant allele frequency (VAF) <1%, very good partial MR (VGPMR) as VAF <5%, partial MR as reduction of VAF >10% but not reaching residual VAF <5%, idem as VAF reduction <10% or increase <5%, and progression (PROGR) as VAF increase >5%. Fisher's exact test and the method of Kaplan and Meier were used to compare groups and perform survival analyses. Results: A total of 73 consecutive pts. with AML treated with CPX-351 induction were identified at our center between 03/2019 and 06/2025. Of those, 51 were evaluable with at least one NGS analysis before and after induction. The median age is 64 years (range 41-75), there are 30 males and 21 females. 45 pts. have AML-MR, 5 myeloid neoplasia post cytotoxic therapy and 1 AML with MECOM rearrangement per WHO 2022. 39 pts. received allo-tx. Conventional response assessment by cytomorphology resulted in 39 CR (78%), 4 PR (8%), 7 (14%) refractory and 1 pt. not evaluable due to dry tap. Molecular response assessment stratified by the worst responding gene mutation per pt. resulted in 8 CMR, 4 VGPMR, 17 PMR, 3 idem, 19 PROGR. Notably, most pts. with molecular PROGR were cytomorphologically defined as CR (15/18 evaluable pts., 83.3%). A total of 159 gene mutations in 33 genes were detected. The most frequently mutated genes were DNTM3A (n=16), RUNX1 (16), TET2 (13), and TP53 (12). Of the 159 mutations, 38% achieved a CMR, 12% a VGPMR, 28% a PMR after induction. 6 % stayed idem and 16% PROGR. For pts. achieving CMR or VGPMR after 1st induction, the alive:death ratio is 11:1, while the ratio is 22:17 for pts. with PMR, idem, or PROGR (p=0.0373). Comparing overall survival (OS), pts. achieving CMR or VGPMR demonstrated a trend towards longer OS than pts. in morphological CR (median OS not reached vs. 28 months, p=0.055), with an encouraging signal in the allo-tx subgroup as well (median OS not reached vs. 53 months, p=0.076). Conclusions: Our data demonstrates that molecular response assessment is feasible in a real-world setting after AML induction therapy with CPX-351. CPX-351 induces CMR or VGPMR in 24% and PMR in 33%, resulting in molecular responses in 57% of pts. While conventional cytomorphology demonstrates a CR in 78% of pts., molecular analysis improves detection of residual disease. Pts. achieving CMR or VGPMR after the 1st induction experienced a significantly lower mortality rate than those with worse molecular responses (p=0.0373). Furthermore, they show a trend to superior survival for the whole cohort (p=0.055) and for the allo-tx subgroup (p=0.076) than pts. in morphological CR. This analysis is limited due to the sample size and short follow-up. Taken together, CPX-351 induces a high rate of molecular remissions in pts. with AML. Molecular remission status refines prognostic accuracy and should be considered for disease assessment and treatment guidance. Additional analyses including cytogenetics and molecular response during subsequent course of treatment and expansion of the cohort are under investigation and will be reported.