Abstract PO2-18-04: STX-478 is a potentially best-in-class mutant-selective PI3Kα inhibitor that demonstrates robust efficacy in ER+ breast cancer models as monotherapy and in combination with standard of care agents

Abstract Activating mutations in PI3Kα are among the most prevalent genetic aberrations in breast cancer. Constitutive signaling of PI3Kα is an established oncogenic driver in these cancers, as these mutations are predictors of response to alpelisib, an approved PI3Kα inhibitor, which inhibits both wild type and mutant PI3Kα. While inhibiting PI3Kα has proven to be efficacious, the concomitant inhibition of wild type enzyme in normal tissue results in metabolic dysfunction and ultimately dose-limiting toxicities. Selectively targeting the mutant PI3Kα enzyme is expected to inhibit tumor growth while sparing normal tissues, thus increasing the therapeutic index compared to non-mutant selective PI3Kα inhibitors. Therefore, we have developed STX-478, an allosteric, CNS penetrant, mutant-selective PI3Kα inhibitor. In a high-throughput viability screen, STX-478 selectively inhibited growth of cell lines with kinase domain and helical domain mutations in PI3Kα. In a panel of PI3Kα mutant CDX and PDX models, STX-478 provided efficacy that was similar or superior to higher than clinically achievable doses of alpelisib. Robust suppression of pharmacodynamic (PD) markers, pAKT, was observed in tumors, but not in skeletal muscle. Importantly, the efficacy achieved with STX-478 occurred at doses that did not cause metabolic dysfunction, in contrast to alpelisib which caused insulin release and suppressed glucose uptake in skeletal muscle in a 13C glucose tolerance test. Collectively, these data indicate that STX-478 has an expanded therapeutic window to selectively target mutant PI3Kα in vivo. In efficacy studies using human tumor xenograft models, STX-478 caused tumor regressions as a single agent. In multiple advanced ER+ breast cancer PDX models including models that are insensitive to the approved CDK4/6 inhibitor palbociclib, STX-478 combined with standards of care was highly efficacious and well tolerated. In a representative PDX model, STX-478 combined with the ER degrader fulvestrant and palbociclib was well tolerated for 90 days of dosing in mice and resulted in durable tumor regressions that were superior to STX-478 alone. In conclusion, these data demonstrate STX-478 is a highly efficacious, well tolerated, mutant-specific PI3Kα inhibitor that can be combined safely with standards of care in breast cancer and provide the therapeutic benefits of PI3Kα inhibition without the side effects associated with inhibiting WT PI3Kα. These properties combined with the potential for CNS penetration and excellent pharmacokinetic profile in higher species make STX-478 a potentially best-in-class mutant-selective PI3Kα inhibitor. STX-478 is currently being evaluated in a Phase I clinical trial (NCT05768139). Citation Format: Trang Tieu, Leonard Buckbinder, David St. Jean, Samantha Manimala, Gregory Dowdell, Michael Huff, Jacob Alltucker, Erica Jackson, Angel Guzman-Perez, Darrin Stuart. STX-478 is a potentially best-in-class mutant-selective PI3Kα inhibitor that demonstrates robust efficacy in ER+ breast cancer models as monotherapy and in combination with standard of care agents abstract. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-18-04.