Abstract A21: Compound PIK3CA mutations support a mutational dose-response model for oncogene activation and response to PI3K inhibitor targeted therapy in breast cancer

PIK3CA mutations represent the most frequent oncogenic driver lesion found across all human cancers. Given that single PIK3CA mutations are a predictive biomarker for response to PI3K inhibition, and that these clinical benefits are small, we hypothesized that additional genomic factors may cooperate with single PIK3CA mutations in oncogenesis and response to PI3K inhibitors. We analyzed several large pan-cancer datasets and have discovered a phenomenon of dual PIK3CA mutations in 15% of all PIK3CA-mutant cancers, including breast cancer. Dual PIK3CA mutations are clonal, are found in both primary untreated and metastatic tumors, and are enriched at conserved amino acid positions. We have proved that the majority of dual PIK3CA mutations—including the most frequent dual PIK3CA mutant combinations in breast cancer—are in cis compound mutations (i.e., on the same allele, resulting in a protein molecule with two mutations) in patient samples and cell lines, as shown through multiple sequencing techniques including long-range single-molecule real-time sequencing (SMRT-seq). Compound PIK3CA mutations increase PI3K pathway signaling and growth proliferation in nontransformed cells and cancer cells in vitro and in vivo. Biochemical experiments using recombinant PI3K complexes reveal a mechanism where compound mutants increase PI3K complex thermal instability, resulting in increased kinase activity and liposome binding as compared to single mutants. Compound PIK3CA mutations render cells more sensitive to the PI3Kα inhibitors BYL719 and GDC-0077 than single PIK3CA mutations. These data translate to the clinic, where patients with dual PIK3CA-mutant metastatic breast cancer exhibit more durable responses to PI3K inhibition versus patients with single PIK3CA-mutant breast tumors. Our data support a novel mutational dose-response model for oncogene activation, where compound mutations amplify PI3K signaling and growth to a greater degree than single mutants, resulting in increased sensitivity to PI3Kα inhibitor targeted therapy. These findings also have therapeutic relevance, where compound PIK3CA mutations may be a superior predictive biomarker for response to PI3Kα inhibitors than single PIK3CA mutations, across PIK3CA-mutant cancers. Citation Format: Neil Vasan, Jared Johnson, Hong Shao, Pedram Razavi, Alexander Gorelick, Erik Ladewig, Alesia Antoine, Hardik Shah, Eneda Toska, Guotai Xu, Abiha Kazmi, Barry Taylor, Komal Jhaveri, Maura Dickler, Elisa de Stanchina, Eduard Reznik, Raul Rabadan, Melissa Smith, Robert Sebra, Lewis Cantley, Maurizio Scaltriti, Jose Baselga. Compound PIK3CA mutations support a mutational dose-response model for oncogene activation and response to PI3K inhibitor targeted therapy in breast cancer [abstract]. In: Proceedings of the AACR Special Conference on Targeting PI3K/mTOR Signaling; 2018 Nov 30-Dec 8; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(10_Suppl):Abstract nr A21.