Clonal Basis of Response and Resistance to Combination Anti-Apoptotic Therapy in Patients With Acute Myeloid Leukaemia

Commercial partner: AstraZeneca, Cambridge - https://www.astrazeneca.co.uk/

Background
Acute Myeloid Leukaemia (AML) is the most common aggressive adult leukaemia (~52 000 new cases/year in US & Europe). Most AML patients die within 6 months of diagnosis. Thus, there is a significant unmet need.

AML biology: Every human makes ~10 billion new mature, blood cells/daily. 10 billion blood cells die daily by exiting cell cycle and undergoing apoptosis or phagocytosis. In AML, this “conveyer belt” production is grossly perturbed. We[1-3], and others, have shown that AML is propagated by leukaemic stem cells (LSCs), marked by expression of specific cell surface molecules (e.g. ASCT2[4]). LSCs have deregulated transcriptional programmes dependent on cyclin dependent kinase 9 (CDK9) (reviewed in[5]), the catalytic subunit of the transcription elongation factor P-TEFb. AML LSCs avoid apoptosis through elevated expression of the anti-apoptotic protein myeloid cell leukaemia-1 (MCL-1)[6, 7], closely related to BCL-2. Notably, MCL-1 expression is dependent on CDK9. Small molecules targeting CDK9 and MCL-1 are therapeutically active in pre-clinical AML patient derived xenograft models[8].

Clinical trials: AZ have just started three first-in-man Phase I trials of inhibitors of MCL1 and CDK9 and an antibody-drug conjugate against ASCT2. Sequential, viably frozen patient bone marrow (BM) and peripheral blood (PB) cells, before and after treatment, are available with clinical, pharmacokinetic (PK) and pharmacodynamic (PD) metadata.

Vyas laboratory: internationally leading in studying AML patient samples, including state-of-the-art single cell assays[1-3,9-15]. Now also implemented single cell transcriptomic assays combined with: (i) mutational analyses[16,17], (ii) cell surface marker analysis (CITE-seq) and epigenetic analyses (scATAC-seq).

Project Aims & Experimental Approach & Feasibility
In all three trials, studies below will be correlated with clinical, PK and PD metadata.

(i) Are there genetic AML mutational biomarkers of response?
Approach: Test BM DNA from trial entry samples for recurrent AML mutations using our established methods[13].

(ii) Measure depth of anti-AML cell eradication.
Approach: The most sensitive way to measure amount BM leukaemic cells left after treatment is by flow cytometry[9] and sensitive next generation sequencing[18].
ASCT2 antibody-drug conjugate study

(iii) Determine if therapy removes ASCT2-expressing cells.
Approach: Flow cytometric measurement of BM ASCT2-expressing cells pre- and post-treatment using Vyas Lab’s existing protocols[14] and a commercial anti-human ASCT2 antibody not binding same epitope as the therapeutic antibody.

(iv) Determine if therapy removes ASCT2 LSCs. Correlate with PK and clinical response.
Approach: Up to106 ASCT2-expressing and non-expressing cells, pre- and post-therapy, will be purified by FACS-sorting, and tested for leukaemic stem cell function in immunodeficient NSG mice using established Vyas lab protocols[2].
MCL-1 inhibitor and CDK9 inhibitor trials

(v) Test if CDK9 or MCL-therapy inhibits CDK9 transcriptional programmes and promotes apoptosis. Correlate with PK and clinical response.
Approach: RNA-Seq on BM cells pre- and post-treatment using Vyas Lab protocol (only 200 cells required)[19] and single cell RNA-Seq (10X Chromium). Test for apoptosis by FACS.