Approximately 15–25% of patients with acute myeloid leukaemia (AML) harbour IDH2 mutations and are candidates for IDH2 inhibition with enasidenib. Now, a longitudinal analysis of the clonal landscape of AML has provided important insights into responsiveness and resistance to enasidenib.

To characterize both differentiation arrest in and the clonal architecture of AML, bone marrow cells from patients included in the pivotal trial of enasidenib were studied using flow cytometry and single-cell genotyping. “We tracked the behaviour of these clones through treatment and at relapse, showing that restored haematopoiesis during remission can be mediated by differentiation of IDH2-mutant ancestral clones or progeny subclones,” explains Lynn Quek. “We believe that this is the first time the clonal dynamics of AML have been studied at a single-cell level in response to therapy,” she adds.

IDH2 mutation-related differentiation arrest might be dependent on the particular co-mutational context, explaining why different populations of clones differentiate upon IDH2 inhibition and potentially also why ~60% of patients have intrinsic resistance to enasidenib. Moreover, “IDH2-mutant clones persisting during therapy acquire additional mutations at disease relapse, providing clues to potential mechanisms of acquired enasidenib-resistance,” states Quek. No second-site mutations in the mutant IDH2 allele were detected; however, IDH2-mutant subclones with neomorphic mutations in IDH1 or aberrations affecting components of other potential bypass pathways that restore differentiation arrest, such as cytokine receptors and haematological transcription factors, were clonally selected.

Together, these findings might enable improvements in patient selection in future trials or inform on novel drug combinations.