Clinical Pharmacodynamic Markers and Combinations with SY1425 (tamibarotene) in a Genomically-Defined Subset of Non-APL AML

SY-1425, a potent and selective agonist of the retinoic acid receptor RARα, is being investigated in a Ph2 trial in a novel genomically-defined subset of non-APL AML and MDS patients (clinicaltrials.gov NCT02807558). RARa is a nuclear hormone receptor and transcription factor that regulates genes involved in cell differentiation and proliferation. We identified a super-enhancer (SE) at the RARA locus, the gene encoding RARa, in a subset of primary non-APL AML blasts. Preclinical models demonstrated a correlation between the presence of a RARA SE and sensitivity to SY-1425, providing the rationale for clinical investigation. Further research has investigated pharmacodynamics (PD) markers and combinations of drugs to support clinical development of SY-1425. In this study we identified DHRS3mRNA induction as a measure of RARα target engagement with SY-1425. We also demonstrated synergy in preclinical models with SY-1425 and hypomethylating agents.

Since RARα is a transcription factor that regulates target genes when bound by a retinoid, we characterized the dynamic expression changes of a panel of RARA enhancer- high and - low non-APL AML cell lines (hereafter referred to as RARA-high and -low) in response to SY-1425 treatment. DHRS3 showed the largest expression increase following treatment in 3 RARA-high cell lines, with a range of 29 to 115 fold. In contrast, there was a much lower DHRS3 induction in 3 RARA-low cell lines (range of 1.6 to 6.1 fold). Induction was found to be both time- and dose-dependent with maximal induction at approximately 6 hours and half maximal induction near the EC50 for the anti-proliferative effect in RARA-high cell lines. DHRS3 encodes dehydrogenase/reductase (SDR family) member 3, a metabolic enzyme involved in maintaining cellular retinol homeostasis and had previously been shown to be induced by retinoids. Thus, DHRS3induction in tumor cells represents a potentially useful PD marker for clinical studies of SY-1425.

To better understand the mechanism of induction of DHRS3 by SY-1425 we examined the chromosomal localization of RARα as well as the epigenomic state of the DHRS3 locus by ChIP-seq for RARα and H3K27 acetylation, the latter being an indicator of active enhancers and promoters. In the untreated state, OCI-AML3 (a typical RARA-high AML cell line) was found to have multiple RARα binding sites both within and distal to the DHRS3 gene but minimal H3K27 acetylation. Following treatment with SY-1425, the level of H3K27 acetylation at DHRS3 increased, resulting in the formation of a SE. Moreover, the SE encompassed the RARα binding sites, consistent with the model in which SY-1425 converts RARα into an activator of DHRS3expression. Similar results were seen for the CD38 locus in which SY-1425 treatment increased expression, H3K27 acetylation, and RARα binding. CD38 is a cell surface antigen and marker of myeloid maturation readily analyzed by FACS analysis, suggesting it could be an additional PD marker to be used in clinical studies. Indeed, it was found that SY-1425 induced CD38 cell surface expression at similar levels in RARA-high AML cell lines and the NB-4 APL cell line, but not in RARA-low cell lines.

We also investigated combinations of SY-1425 with approved or investigational AML and MDS agents in in vitro and in vivo models to inform future clinical studies and to further explore potential PD markers unique to the combined action of the drugs. Several standard of care agents and drugs in current development were found to have synergistic interactions with SY-1425 in RARA-high but not RARA-low cell lines. In particular, azacitidine and decitabine each showed strong in vitro synergy with SY-1425. Evaluation of SY-1425 plus azacitidine in a RARA-high PDX model of non-APL AML demonstrated a better response compared to either agent alone. Additional genome-wide ChIP-seq and expression studies of RARA-high cells treated with various combinations are being investigated to identify optimal PD markers for these combinations. These studies support the use of DHRS3 mRNA induction in tumor cells as a PD marker in the recently initiated Ph2 study of SY-1425 in genomically-defined non-APL AML and MDS patients (clinicaltrials.gov NCT02807558) and further exploration as a PD marker for future combination studies.