Abstract 2799: BAY 1082439, a highly selective and balanced PI3Kα/β inhibitor demonstrated potent activity in tumors with activated PI3Kα and loss-of-function of PTEN

The PI3K pathway plays critical roles in cancer cell growth and survival, as well as in intrinsic and acquired resistance to both chemotherapy and targeted agents. These essential roles have led to the clinical development of PI3K pathway inhibitors. Due to the complexity derived from the existence of various PI3K isoforms (≥,α,α,α), and their differential roles in signal transduction, as well as cancer pathology, development of PI3K inhibitors with differential pharmacological profiles would allow exploration in different indications, combinations and dosing regimens. Having identified BAY 80-6946, an intravenously dosed, highly potent and selective PI3K inhibitor which is particularly effective in PIK3CAmut and/or Her2+ tumors, we sought to develop a novel oral PI3K inhibitor particularly effective in PTEN-loss tumors with coexisting mutation or amplification of PIK3CA and/or activation of PI3Kα (e.g., through RTKs). Herein we report the pharmacological profile of a highly selective PI3Kα/α-balanced inhibitor, BAY 1082439. BAY 1082439 has an IC50 ratio of 1:3 in biochemical assays of PI3Kα (4.9 nM) vs. PI3Kα (15.0 nM), and >1000-fold selectivity against mTOR kinase. The balanced PI3Kα and PI3Kα activity of BAY 1082439 is also reflected in cellular mechanistic (p-AKT473) and proliferation assays in PI3Kα- (KPL4, BT474) vs. PI3Kα-driven (PC3, LNCaP) tumor cells. In vivo, BAY 1082439 showed clear advantages over the strong PI3Kα inhibitor BAY 80-6946 in PTEN/PI3Kα-driven tumor models (e.g., PC3 and HEC-1B), when the two compounds were compared at their MTDs. Furthermore, BAY 1082439 has unique pharmacokinetic (PK) properties with very high plasma free fractions across all species tested (33-50%), large Vss, high clearance and intermediate T1/2. The relationship of PK vs. PD and the efficacy vs. dosing regimens were investigated. BAY 1082439 showed strong p-AKT inhibition at 2 and 5 hours post-treatment while p-AKT returned to levels comparable to the vehicle group at 24 hours in all 4 tumor models tested. Interestingly, with once daily dosing, BAY 1082439 could induce tumor regression in KPL4 (PIK3CAmut and HER2+), and tumor stasis in HEC-1B (PTENdel) and in HEC-1A (PIK3CAmut) tumor models, suggesting that continuous inhibition of p-AKT may not be required for anti-tumor efficacy. In addition, comparison of different dosing regimens (QD, Q2D, D1-2/W, D1-3/W, D1-4/W to QW) at MTD indicated that QD and QW dosing produced optimal anti-tumor efficacy. These res