Hierarchical lineage tracing reveals diverse pathways of cytarabine resistance

Cancer cells adapt to treatment, leading to the emergence of clones that are more aggressive and resistant to anti-cancer therapies. We have a limited understanding of resistance mechanisms as we lack technologies to map cancer evolution under the selective pressure of treatment. To address this, we present a hierarchical, dynamic lineage-tracing method, FLARE (Following Lineage Adaptation and Resistance Evolution). We use FLARE to track the progression of acute myeloid leukemia (AML) cell lines treated with Cytarabine (AraC), a front-line treatment in AML, both in vitro and in vivo. We map distinct cellular lineages in both murine and human AML cell lines that are predisposed to AraC resistance. Using FLARE, we identify treatment-naïve populations responsible for seeding resistance that are characterized by upregulation of stemness markers and a cell adhesion-associated AraC-resistant lineage signature. We find that expression of this signature in pediatric AML is associated with the expansion of HSC-like malignant cells at relapse and significantly shorter overall survival. These findings underscore the role of pre-existing lineage states in driving relapse and establish FLARE as a platform for uncovering the evolving, heritable transcriptional programs that underlie tumor evolution. We still have a limited understanding of how cancer evolves under therapy and develops resistance. Here, the authors develop FLARE, a dynamic lineage-tracing method to identify clonal patterns of drug response and expansion in vitro and in vivo, and apply this method to study the response of acute myeloid leukaemia cells to chemotherapy.