(P 425) EXPLORING MECHANISMS OF EGFR INHIBITOR RESISTANCE IN CHORDOMA

Objective: Target discovery studies have identified EGFR inhibition as a promising therapeutic strategy in chordoma, motivating Phase 2 repurposing trials with afatinib and cetuximab. However, EGFR is not mutated in chordoma, leaving the mechanisms associated with EGFR inhibitor (EGFRi) sensitivity and resistance unclear. The objective of this study is to profile the genomes, transcriptomes, and EGFRi sensitivity of preclinical chordoma models to explore molecular features associated with drug sensitivity or resistance.

Methods: A panel of 14 chordoma cell lines and 12 patient-derived xenograft (PDX) models were profiled by whole-exome sequencing (100X) and whole-transcriptome sequencing (80M reads/sample, N=4 biological replicates). EGFR inhibitor sensitivity was assessed in CellTiter-Glo proliferation assays; for each cell line seeding densities and assay durations were used such that drug treatments were carried out between 15-90% confluence. In vivo xenograft studies were carried out in 6-8 week old female athymic nude mice at XenoSTART. Tumor fragments were implanted subcutaneously and mice were randomized into treatment groups once tumors reached 150-300 mm3 in size. RNA-Seq data files generated by the RSEM pipeline were used to perform differential gene expression analysis. RSEM gene-based count data were used to create a DESeq2 object and the full DESeq2 pipeline was run to generate an expression top table with FDR p-value adjustment. Cell imaging experiments were performed using a BioTek Cytation 5 multi-mode imaging reader, and western blots were performed using standard approaches and specified antibodies.

Results: We profiled a panel of 14 chordoma cell lines to categorize their sensitivity to afatinib, a second-generation covalent EGFRi with potency against the wild-type receptor, and looked for associations between response and gene expression to identify features associated with sensitivity or resistance. Proliferation assays revealed striking differential sensitivity of chordoma cell lines to afatinib: sensitive cell lines have Absolute EC50 values in the range of ~5-30 nM, whereas resistant cell lines have Absolute EC50 values greater than 1 µM. This differential sensitivity is mirrored in xenograft models, with tumor growth inhibition ranging from 0% in resistant models to ~100% in sensitive models. In matched cell line/xenograft models, afatinib pharmacotypes are conserved. Analysis of differentially-expressed genes in resistant versus sensitive cell lines revealed elevated type I interferon (IFN) signaling in afatinib-resistant cell lines. Modulation of IFN receptor signaling did not reverse sensitivity or resistance to afatinib, indicating that IFN does not drive the resistant phenotype per se, and implicating an upstream IFN-regulatory pathway in afatinib resistance. Studies are ongoing to identify the upstream resistance driver using CRISPR/Cas9 gene editing of immunostimulatory cytosolic nucleotide sensors, and to further understand the causes and consequences of high IFN signaling in chordoma.

Conclusion: Our data reveal a striking differential sensitivity of chordoma cell lines to afatinib, which was mirrored in vivo in xenograft models. Combining in vitro afatinib sensitivity data with gene expression analysis identified an enrichment of interferon (IFN) signaling in resistant cell lines. Despite its association with EGFRi resistance in lung cancer, IFN signaling per se does not appear to promote resistance in chordoma cell lines. Instead, it may serve as a biomarker to guide patient selection for EGFR-targeted therapy in chordoma, or reflect underlying genotoxic stress due to the accumulation of immunostimulatory cytosolic nucleotides. Studies are ongoing to understand the relationship between IFN signaling and EGFRi resistance.