Er+ metastatic breast cancer targeted therapybiomarkers of response and mechanisms of resistance to pi3k and fgfr inhibitors

Resumen

Endocrine therapy has been one of the major advances in the treatment of breast cancer (BC) for the past 30 years. However, patients eventually relapse due to mechanisms of resistance. These involve several pathways related to proliferation and growth, including the PI3K/mTOR and FGFR pathways. The PI3K pathway is frequently altered in BC, especially in the luminal setting where 30-40% of patients harbor PIK3CA mutations. Specific inhibitors targeting key nodes in the PI3K pathway have been developed, such as mTOR and PI3K inhibitors (PI3Kinh), and some are FDA approved for metastatic ER+ BC. However, alterations downstream PI3K or in parallel pathways, such as 11q13 amplification, are recurrent in the metastatic disease, and it is important to understand if they confer resistance to PI3Kinh. FGFR1 amplification is found in 10% of metastatic luminal BC. Specific inhibitors against FGFR1/2/3 (FGFRinh) have shown limited efficacy in the FGFR1amp BC. This is in contrast with the clinical efficacy observed previously with Multi-targeted Tyrosine Kinase Inhibitors (MTKI, targeting FGFR, VEGFR and PDGFR families), albeit these agents showed severe adverse effects. In this study, we investigated mechanisms of resistance to PI3Kinh and biomarkers of response to FGFRinh. In the first part, we determined the antitumoral activity of alpelisib, an α-specific PI3Kinh, in a cohort of 24 genetically annotated Patient-Derived Xenografts (PDX). We observed an aggregation of genetic alterations downstream PI3K and 11q13 amplifications among the resistant models and validated these with in vitro models as mechanisms of resistance to alpelisib, including overexpression of cyclin D1 and high FGF-signaling. In addition, in 2 out of 3 PDX models with putative mTORC1 activating alterations, we observed higher efficacy of an mTORC1/2 inhibitor compared with a PI3Kinh. In the second part of this thesis we analyzed the antitumor activity of an FGFRinh, rogaratinib, in a cohort of 17 PDX harboring amplifications in FGFR1/4 and/or FGF3/4/19 (11q13 amplicon). Amongst the potential biomarkers of response tested (FGFR1-4 gene copy number, FGFR1-4 gene expression, and FGF ligand co-amplification and expression), we identified that high mRNA of FGFR1-4 could predict for FGFRinh response. We compared rogaratinib efficacy with a MTKI, lucitanib, and we observed higher efficacy of lucitanib across 7 models. Furthermore, we observed an increased blockade of proliferation and vascularization under lucitanib treatment. However, when testing an antiangiogenic therapy in two rogaratinib-resistant models, similar levels of efficacy as with lucitanib monotherapy were achieved, suggesting that lucitanib efficacy was due to vascular blockade in both models. In conclusion, we established that PDX models harboring genetic alterations consistent with mTORC1 activation (e.g. TSC1del) are resistant to PI3Kinh but sensitive to an mTORC1/2inh and that cyclin D1 overexpression, high FGF and both alterations together result in resistance to PI3Kinh. We further determined that high mRNA of FGFR1-4 predicts for FGFRinh response and that a MTKI exhibits higher efficacy than specific FGFRinh due to a higher blockade of proliferation and vascularization. These results will help for a better patient selection in future clinical trials for ER+ BC targeted therapy.