Diverse Drug Combinations with an Eed Inhibitor Confer Context-Specific Benefit across Multiple Tumor Types

The polycomb repressive complex 2(PRC2), consisting of the core subunits EED, EZH2, and SUZ12, is frequently overexpressed and deregulated in hematological and solid malignancies. Clinical validation for targeting the PRC2 complex has been demonstrated with several small molecule inhibitors of EED or EZH2 that are either approved or being evaluated in clinical trials in different cancer settings. Like other treatment modalities, single-agent activity is limited. Given that the PRC2 complex is involved in multiple diverse oncogenic/immune pathways, we thus explored the combination potential of inhibiting PRC2 in a wide array of biological contexts. To this end, we have developed a potent, selective, and orally bioavailable EED ligand. The compound binds to EED (K D =3.2 nM) and inhibits PRC2 enzyme activity (IC 50= 7nM). The compound inhibits proliferation of an EZH2 dependent Karpas422 GCB-DLBCL cell line (GI 50= 27nM) in vitro. In vivo, the inhibitor is well tolerated and drives tumor regression at 10 mpk and 50 mpk doses in a DLBCL xenograft model. Given the compound has favorable selectivity, potency and pharmacokinetic profile, we next evaluated the combinability of this EEDi with select compounds targeting diverse oncogenic pathways. We find that EED and PI3K/AKT inhibitors show marked combination activity in DLBCL, in part through downregulation of AKT signaling. Combination activity is also observed with the EED inhibitor in combination with the BTK inhibitor, acalabrutinib, in DLBCL, although this appears to be context dependent. In addition to DLBCL models, we have also assessed the therapeutic potential of EED inhibitor combinations in solid tumors. In SCLC, the EED inhibitor increased expression of SLFN11, a biomarker that has been linked to clinical response to a PARP inhibitor / temozolomide combination (Pietanza et al., 2018). Therefore, we evaluated the EED inhibitor in combination with the PARP inhibitor, olaparib, in an aggressive SCLC xenograft model and observed reduced tumor growth in the combination arm but not in monotherapy arms. In ovarian cancer, ARID1A mutations are proposed to sensitize to EZH2 and ATR inhibitors (Bitler et al., 2015, Williamson et al., 2016) which prompted evaluation of the EED inhibitor with the ATR inhibitor, AZD6738. Indeed, the combination resulted in tumor stasis whereas the monotherapies produced modest tumor growth inhibition. Last, we have evaluated the EED inhibitor alone, and in combination with a CTLA4 antibody, in a tumor immunity setting using an immunologically cold syngeneic melanoma model. Interestingly, we find that EEDi has significant anti-tumor activity alone, as compared with either control or CTLA4 antibody treatment alone. Together, these data show that while our EEDi has strong single-agent activity in vitro and in vivo, combining the EEDi together with inhibitors of select oncogenic pathways may bring deeper therapeutic response in a context dependent manner.