| Summary: | <p>Acute myeloid leukaemia (AML) is an aggressive malignant disorder of immature myeloid cells in the bone marrow. The outcome of older AML patients who are unfit for intensive chemotherapy remains dismal. Relapse, which is the main cause of mortality, occurs due to Darwinian clonal selection and persistence of leukaemia stem cells (LSCs) throughout treatment. Understanding the kinetics of clonal selection and the molecular mechanisms promoting LSC phenotypes is thus urgently required.</p>
<br>
<p>To address the first point, I focused on <em>IDH1</em>-mutated patients enrolled in a phase Ib/II clinical trial of ivosidenib and venetoclax ± azacitidine. Despite an initial response rate of 94%, relapse was still common. To define the clonal and cellular basis of acquired resistance and sustained response, studies were conducted integrating clonal and immunophenotypic identity of single cells on longitudinal samples from 9 patients. Whilst all patients attained an initial clinical response, 7 relapsed and 2 remained in sustained remission. In all relapsing patients, therapy-resistant clones were selected early, within 3 treatment cycles, preceding relapse by months or years. Treatment failure was associated with either newly detected genetically evolved clones or impaired differentiation of pre-existing clones. In both cases, selection occurred within small LSC populations, whilst patients were in clinical remission. In patients attaining sustained response, all leukaemic clones were eradicated and rapidly replaced by clonal and wild type haematopoiesis. This study serves as proof-of-principle that early clonal selection within LSCs correlates with long-term patient outcome. If validated in larger cohorts, clonal assessment within LSCs rapidly post-treatment initiation may predict patient outcome and
enable to alter treatment in those patients destined to relapse.</p>
<br>
<p>However, efficient therapeutic eradication of LSCs requires an understanding of how somatic mutations corrupt gene regulation to sustain LSC phenotypes. To address this gap, in the second part of this thesis, I developed a method enabling simultaneous genotyping of genomic loci with chromatin accessibility profiling at single-cell resolution, termed GTAC. In cell lines, GTAC enabled accurate and high-confidence genotyping, whilst yielding robust chromatin accessibility readouts. In human AML, GTAC efficiently mapped genetic clonal evolution to epigenomic variation, revealing chromatin changes underpinning the transformation of pre-leukaemic stem cells into <em>RUNX1</em>-mutated LSCs. Additionally, GTAC revealed that subclonal <em>BCOR</em> mutations cooperate with the <em>RUNX1</em> mutation within the leukaemic hierarchy, promoting an LSC-like epigenetic state whilst disrupting differentiation. Mechanistically, <em>BCOR</em> mutations associate with de-repression of proximal cis-regulatory elements related to stemness genes and alter the activity of transcription factors driving self-renewal and differentiation. These analyses identified a putative therapeutic target for this patient subgroup, that warrants further functional and molecular investigation. Overall, GTAC is a robust method applicable to a broader field of cancer evolution.</p>
|
|---|