AUTHORS
James Lynch, Liliana Greger2, Rebecca Harris I3, Martin Gerard O'Rourke Sr4, Ivan Del Barco Barrantes I5, Patrick Connerty I, PhD6, Vivien Nguyen7, Laura Casaeres Perez8, Aleksandra Azevedo I9, Colin Davies10, Stuart Nelis11, Giles A Brown12, Martin Pass13, Praful Gokhale I14, Richard Lock I, PhD15, Louise Kelly Modis I, PhD16
AFFILIATIONS
1 Amphista Therapeutics, Cambridge, United Kingdom
2 Amphista Therapeutics, Cambridge, United Kingdom
3 Amphista Therapeutics, Cambridge, United Kingdom
4 Amphista Therapeutics, Cambridge, United Kingdom
5 Amphista Therapeutics, Cambridge, United Kingdom
6 Children's Cancer Institute at Minderoo Children’s Comprehensive Cancer Centre,, Sydney,, New South Wales, Australia, School of Clinical Medicine, Sydney, New South Wales, Australia
7 Children's Cancer Institute at Minderoo Children’s Comprehensive Cancer Centre,, Sydney,, New South Wales, Australia, School of Clinical Medicine, Sydney, New South Wales, Australia
8 Amphista Therapeutics, Cambridge, United Kingdom
9 Amphista Therapeutics, Cambridge, United Kingdom
10 Amphista Therapeutics, Cambridge, United Kingdom
11 Amphista Therapeutics, Cambridge, United Kingdom
12 Amphista Therapeutics, Cambridge, United Kingdom
13 Amphista Therapeutics, Cambridge, United Kingdom
14 Dana-Farber Cancer Institute, Boston, Massachusetts, United States
15 Children's Cancer Institute at Minderoo Children’s Comprehensive Cancer Centre,, Sydney,, New South Wales, Australia, School of Clinical Medicine, Sydney, New South Wales, Australia
16 Amphista Therapeutics, Cambridge, United Kingdom
ABSTRACT
AMX-883 is an orally bioavailable BRD9 Targeted Glue™ degrader acting via the E3 ligase DCAF16, with sub-nanomolar potency and exquisite degradation selectivity over related bromodomain proteins. AMX-883 induces deep and potent BRD9 degradation and increases markers of maturation and differentiation, across a range of acute myeloid leukemia (AML) cell lines representing diverse cytogenetic backgrounds, including those harboring TP53 mutations. The aims of this study were to further elucidate the mechanism of action of AMX-883 and to assess its in vivo efficacy against patient-derived xenograft (PDX) models of AML both as a single agent and in combination with the BCL-2 inhibitor venetoclax.
To elucidate the mechanisms by which degradation of BRD9 releases the differentiation block in AML, a multi-omics approach was taken across five AML cell lines. Six hours of AMX-883 treatment led to significant gene expression changes, consistent with the rapid rate of BRD9 degradation induced by AMX-883. Gene set enrichment analysis revealed significantly enriched gene sets shared across all tested AML cell lines. This includes activation of myeloid differentiation signatures and suppression of MYC and leukemic stem cell (LSC) signatures. ATAC-seq following AMX-883 treatment revealed a rapid loss of chromatin accessibility in regions significantly enriched with transcription factor (TF) motifs for MYB, RUNX, CEBP and PU.1. At later time points greater modulation of chromatin accessibility was observed, significantly enriched with motifs for a larger set of hematopoietic TFs. Significant reductions in MYC and IRF8 mRNA levels were observed across 4/5 cell lines. Loss of MYC and IRF8 has been shown to be downstream of MYB, and is consistent with the ATAC-seq data, where AMX-883 reduces chromatin accessibility to regions enriched with MYB binding site motifs.
To demonstrate that BRD9 degradation leads to anti-leukemic activity, single-agent AMX-883 efficacy studies were performed against a number of AML PDX models. AMX-883 significantly increased mouse event-free survival (EFS) in all models and the survival benefit was accompanied by significant reductions in leukemia infiltration (% human CD45+ (%hCD45+) cells) in the bone marrow and peripheral blood.
We also tested in vivo efficacy of AMX-883 in combination with venetoclax. The combination of AMX-883 and venetoclax markedly increased mouse EFS compared with vehicle control mice and more than doubled the survival advantage compared with both single-agent treatment arms. Moreover, the AMX-833/venetoclax combination almost completely eradicated leukemic infiltration in the BM (median 0.1% hCD45+ cells compared with a median of 98.3% in vehicle control mice).
To determine whether BRD9 is involved in the maintenance of AML LSCs, we examined the impact of AMX-883 treatment on LSC frequency using an in vivo functional assay. Equal numbers of BM-derived hCD45+ cells from mice treated with AMX-883 or vehicle control were transplanted into secondary recipient mice in an extreme limit dilution assay. AMX-883 treatment reduced the LSC frequency by approximately 10-fold (1 in 35,473 cells from vehicle control mice versus 1 in 316,913 cells from AMX-883-treated mice).
Taken together, our data suggest that BRD9 plays a critical role in AML, maintaining the activity of a number of hematopoietic TFs. Degradation of BRD9 using AMX-883 restricts these TF programs, releases the differentiation block and reduces the frequency of AML LSCs. AMX-883 is scheduled for evaluation in a phase 1 study to assess the safety, pharmacokinetics, and preliminary efficacy in participants with AML and high-risk myelodysplastic syndrome, with recruitment initiating in Q4 2026 (NCT07723703).
DISCLOSURES
Louise Kelly Modis I, PhD: No relevant conflicts of interest to declare.