CPX-351 Induces High Rate of Complete Remission in the New ICC/ELN 2022 Adverse-Risk Entity ‘AML with MDS-Related Gene Mutations’: Evidence from a Large Real-World Italian Study

Introduction

CPX-351 is a liposomal formulation of daunorubicin and cytosine arabinoside approved for the treatment of therapy-related AML (t-AML) or AML with Myelodysplastic-related changes (AML-MRC). Despite the pivotal study showed significant improvements in OS compared to ‘7+3’ conventional chemotherapy, a recent study (UK NCRI AML19 trial) suggests that in patients <60 years of age CPX-351 does not improve results as compared to FLAG-IDA chemotherapy. Specifically, in this study, overall response rates (CR/CRi) were 64% and 76% with CPX-351 and FLAG-IDA, respectively, whilst there were no differences in terms of OS (13.3 vs 11.4 months). However, in an exploratory subgroup analysis, a significant improvement in OS was observed with CPX-351 in AML with MDS-related gene mutations (median, 38.4 vs 16.3 months). This is a new high-risk entity introduced in ICC/ELN 2022 characterized by the presence of at least one mutation among SRSF2, ASXL1, BCOR, EZH2, STAG2, SF3B1, U2AF1, ZRSR2 and RUNX1, in absence of TP53 mutations and with or without cytogenetics abnormalities. Actually, CPX-351 is not specifically indicated in this AML.

Methods

We conduct a multicenter retrospective study in 109 patients from a large cohort of AML patients who received commercially available CPX-351 treatment since the approval of the drug at 16 Italian Hematology centers focusing on the new ‘AML with MDS-related gene mutations’ entity as defined by ICC 2022. Primary endpoint was to evaluate the ORR (CR/CRi) to CPX-351 induction (1 or 2 cycles) and the rate of allo-HSCT. Secondary endpoint was to evaluate the impact of specific MDS-related gene mutations on ORR. Other secondary endpoints included OS between AML with MDS-related gene mutations and AML with MDS-related cytogenetic abnormalities. Moreover, to improve statistical power for specific subgroups analysis we generated synthetic data from our dataset using Synthpop package, expanding our cohort of 300% to obtain 327 pts.

Results

109 patients were selected from the database for available data on MDS-related gene mutations. Patients were treated between September 2019 and June 2024 (median age 65 yrs, range 19-80). Seventy-five patients (52%) satisfy the criteria to be classified as AML with MDS-related gene mutations, 30/70 (43%) with available karyotype showed MDS-related cytogenetic abnormalities. A group of 35 patients classified as AML with MDS-related cytogenetic abnormalities (i.e. without Tp53 mutations and without MDS-related gene mutations, according to ICC 2022) was used as comparison group.

In AML with MDS-related gene mutations we observed higher ORR as compared to AML with MDS-related cytogenetic abnormalities: 52/70 (74%) vs 18/32 (56%) respectively (p 0.0685). Strikingly, these results were also observed in our newly generated synthetic dataset (79% vs 54%)(p <0.0001). Median OS was 1031 dd and 349 dd between the two groups (p 0.0003). In AML with MDS-related gene mutations, subgroup analysis based on cytogenetics abnormalities showed no impact of the karyotype on the ORR (73% vs 75%)(p 0.83) neither on OS (median OS: 1014 vs 874 dd)(p 0.23). We then evaluated the rate of bridge to transplant in this adverse-risk categories; among patients who achieved CR/CRi, thirty-one out of 48 evaluable patients (65%) undergone to allo-HSCT as compared to 10/17 (59%) with AML with MDS-related cytogenetic abnormalities (p 0.51).

Interestingly, focusing on specific MDS-related gene mutations, we found ASXL1 mutation predictive of worst ORR (56% vs 84%, p 0,0091) and worst OS (p 0.0005).

Conclusions

Our results show that CPX-351 is highly effective in the new ICC/ELN 2022 adverse-risk entity ‘AML with MDS-related gene mutations’ in terms of ORR giving the opportunity to go to transplant in CR to the majority of patients. The presence of specific mutations such as ASXL1 have a strong prognostic impact opening new possibilities to further stratify this high-risk AML sub group. This study, based on one of the largest database available, also highlights the potential of synthetic augmentation analysis in facilitating advanced research.