The detailed molecular and cellular mechanisms underpinning the clinical activity of most chemotherapies in cancers remain incompletely understood. Understanding how these drugs really act is a prerequisite for their rational therapeutic optimization. Recent observations suggest that early molecular and cellular changes in cancer cells upon chemotherapy exposure may dictate their long-term fate. We aim to address this question in previously untreated adult Acute Myeloid Leukemia (AML) patients treated with anthracycline/cytarabine association (either as free drugs, '7+3' regimen, or in liposomal formulation, CPX-351) by sequentially sampling peripheral blood during the first course of therapy, and by performing an early bone marrow reassessment. We will apply single cell RNA sequencing and multiparameter flow cytometry to correlate dynamic phenotypic landscapes with clinical outcomes (remission achievement and relapse-free survival). The study will be carried in two phases. First, a feasibility phase will be carried in the first 20 patients irrespective of the genetic make-up of their leukemic cells to identify the optimal pre-analytical conditions for single-cell transcriptional profiling. Second, an expansion phase will be carried focusing on two genetically subsets of patients chosen on the basis of their relative abundance and variability of clinical outcome, namely NPM1c-mutated AML (30% of patients, 60% cure rate) and NPM1-wildtype intermediate-risk AML (25% of patients, 40% cure rate), to correlate single-cell fates with remission and with long-term remission-free survival.