Examining Mechanisms of Sensitivity and Resistance to Phosphatidylinositol 3-kinase inhibitors in Head and Neck Squamous Cell Carcinoma

Abstract

Head and neck squamous cell carcinoma (HNSCC) is a common and debilitating form of cancer with few effective treatment options. HNSCC tumors display a complex array of molecular changes, and sequencing studies have identified the phosphatidylinositol 3-kinase pathway (PI3K) as the most frequently mutated oncogenic and targetable pathway in this cancer type. PI3K signaling contributes to cell growth and survival and is most commonly dysregulated by alterations in the gene PIK3CA, which encodes the catalytic subunit and alpha isoform of PI3K. In spite of this, PI3K inhibition has shown underwhelming efficacy in HNSCC clinical trials to date. Thus, my thesis seeks to evaluate the hypothesis that resistance to PI3K targeting therapies is the result of compensatory signals, which are activated in the presence of PI3K inhibitors. To test this, I examined how aberrant PI3K signaling was influenced by co-expression of EGFR, co-alteration of NOTCH1, and co-dependence of multiple RTKs, including ALK and IGF-1R. EGFR is overexpressed in most HNSCCs and its signaling is a widely studied means by which HNSCC cells evade death in the presence of PI3K inhibition. Consistent with previous studies, I demonstrated activation of the Ras-MEK-ERK pathway, downstream of EGFR, following treatment with PI3K inhibitor monotherapy in multiple PIK3CA amplified UM-SCC cell lines. I also showed that co-inhibition of PI3K with MEK or EGFR was synergistic in a further subset of these cell lines. I then tested several PI3K and EGFR inhibitor combinations in additional in vitro models. My pharmacologic analysis revealed that combinations including irreversible EGFR inhibitors were more effective than those utilizing reversible EGFR inhibitors. In HNSCC, NOTCH1 acts as a tumor suppressor, and inactivating alteration in this gene is observed in nearly 20% of tumors. Emerging data suggests interplay between PI3K and NOTCH signaling in this cancer type. Our CRISPR/Cas9 partial knockout model of PIK3CA in UM-SCC-47, reveal the cooperativity between the PI3K and NOTCH pathways. We confirmed this relationship and its potential importance using a transgenic mouse model: following treatment with 4-nitroquinoline N-oxide, mice with overexpression of mutant Pik3ca and knockout of Notch1 reach endpoint faster than animals with alterations in just one of these genes. Finally, in order to characterize additional signaling pathways driving compensatory PI3K inhibitor resistance, we developed and optimized an unbiased, high-throughput screening approach. We used this assay to test ~1400 inhibitors as monotherapies and in combination with PI3K inhibitors HS-173 and BKM120 in ten HNSCC cell lines. Our initial screening data suggested that combinations of PI3K inhibitors and ALK/IGF-1R inhibitors were among the most effective drug pairs. Using viability, apoptosis and cell cycle assays to test single-agent and combined treatments, we validated the combinatory effects of FDA-approved agents PI3K inhibitor pictilisib and ALK inhibitor brigatinib in a subset of cell lines. These inhibitors were similarly effective in a xenograft model. Furthermore, we identified additional synergistic dual-therapies; many of these inhibited PI3K in combination with upstream receptor tyrosine kinases, while combining PI3K inhibition with inhibition of downstream pathway members did not display synergy. Collectively, these data deepen our understanding of the combined effects of PI3K activation and aberration of an additional signaling pathway in HNSCC. In doing so, they inform the use of targeted PI3K inhibitors, motivate further analyses of PI3K combination treatments and suggest dual-therapies that may result in improved prognoses for HNSCC patients.