Supplementary Figure 4. Impact of FGF ligand blockade in FGFRi treatment response. A) Analysis of mRNA expression levels of FGF ligands by HTG. FGFR1-4 and FGF3/4/19 CN and mRNA status are depicted in the box underneath. Gene expression levels are expressed from the lowest -grey- to the 90th percentile of expression -red-; samples above the 90th percentile are highlighted with stripes. B) Ex vivo analysis of Rogaratinib response (IC50) in 5 PDC models (PDX-derived cell cultures), with or without co-treatment with an FGF-trap (NSC12). The relative mRNA expression levels of relevant FGF ligands and FGFR2 in each model, and the corresponding FGFRi response is summarized in the panel underneath. Statistical analysis: Sidak's multiple comparisons test.
ARTICLE ABSTRACT
FGFR1 amplification (FGFR1amp) is recurrent in metastatic breast cancer (MBC) and is associated with resistance to endocrine therapy and CDK4/6 inhibitors (CDK4/6is). Multi-tyrosine kinase inhibitors (MTKIs) and selective pan-FGFR inhibitors (FGFRis) are being developed for FGFR1amp breast cancer. High-level FGFR amplification and protein expression by IHC have identified breast cancer responders to FGFRis or MTKIs, respectively.
Here, we used preclinical models and patient samples to identify predictive biomarkers to these drugs. We evaluated the antitumor activity of an FGFRi and an MTKI in a collection of 17 breast cancer patient–derived xenografts (PDXs) harboring amplification in FGFR1/2/3/4 and in 10 patients receiving either an FGFRi/MTKI. mRNA levels were measured on FFPE tumor samples using two commercial strategies. Proliferation and angiogenesis were evaluated by detecting Ki-67 and CD31 in viable areas by immunofluorescence.
High FGFR1–4 mRNA levels but not copy-number alteration (CNA) is associated with FGFRi response. Treatment with MTKIs showed higher response rates than with FGFRis (86% vs. 53%), regardless of the FGFR1–4 mRNA levels. FGFR-addicted PDXs exhibited an antiproliferative response to either FGFRis or MTKIs, and PDXs exclusively sensitive to MTKI exhibited an additional antiangiogenic response. Consistently, the clinical benefit of MTKIs was not associated with high FGFR1–4 mRNA levels and was observed in patients previously treated with antiangiogenic drugs.
Tailored therapy with FGFRis in molecularly selected MBC based on high FGFR1–4 mRNA levels warrants prospective validation in patients with CDK4/6i-resistant luminal breast cancer and in patients with TNBC without targeted therapeutic options.
