Abstract A03: Transcriptome comparison of chordoma of the skull base and spine

Abstract

Abstract Differential gene expression profiles from formalin-fixed, paraffin-embedded tissue specimens of chordoma from skull base and spine versus normal specimens were generated and compared to each other. This genomic analysis closely supports our previous transcriptomics results for skull base chordoma, presents for the first time the RNA-seq transcriptome for spine chordoma, and provides clues for explaining severity and outcome of this disease. Using strict criteria, we identified 149 differentially expressed transcripts common in skull base and spine chordoma, 141 transcripts unique in skull base chordoma, and 243 transcripts unique in chordoma of the spine. Further analysis of these three groups of transcripts allowed selection of three subsets of highly differentially expressed genes as potential candidates for biomarkers, drivers, and therapy targets for both chordoma subtypes. Candidate genes were confirmed on protein level by immuno-histochemistry, 2 common for both chordoma subtypes, 2 unique for skull base chordoma, and 2 unique for spinal chordoma. In both chordoma subtypes the highest expressed genes were predominantly developmental-related genes, mostly transcription factors. We assume that these genes play important oncogenic roles in chordoma, mainly causing the high plasticity, adaptability, and resistance to therapy in both of these cancer subtypes but also determining the differentiation status/state and proliferation rate, pointing to features expected from heterogeneous stem cell-like tissues with similarities to its notochord origins. In contrast, some of the overexpressed genes rather imply tumor-suppressing functions like SALL3, and differentiation driving homeobox A genes, like HOXA5 in spinal chordoma, and SOX6 in skull base chordoma, which probably reduce the full potential of the oncogenic genes. Significance: Genomic profiling of the two chordoma subtypes of skull base and spine was performed, and compared to each other to reveal potential subtype specific candidates for biomarkers, disease drivers, and targets for therapy. For validation some of the highest differentially expressed candidates were analyzed by IHC. Conclusion: We assume that these chordoma subtypes are mainly defined by the high number of overexpressed developmental genes, mostly homeobox genes, which provide the high potential for transformation and plasticity, along with the differentiation and proliferation behavior of these cancer cells. The high plasticity probably is the main reason for the high variety in phenotypes of chordoma, from its high heterogeneity on cellular level to its subtype variations depending on tissue location, from its nearly dormant form to an aggressive cancer with extreme adaptability and resistance to any treatments or drugs. Additionally we assume that the impact of the oncogenes of these chordoma subtypes is reduced to different degrees by tumor-suppressing functions of SOX6 in skull base and SALL3, and developmental genes of the homeobox A family (HOXA5) in spine, which rather drive cells towards a stable differentiation, than preserving dedifferentiated multipotent stem cell-like features. We assume that inactivation of these suppressor genes, e.g., by mutations, would allow the overexpressed oncogenic developmental genes to fully exert their influence on these chordoma types. Citation Format: Achim H. Bell, Franco Demonte, Shaan M. Raza, Gregory N. Fuller, Diana Bell. Transcriptome comparison of chordoma of the skull base and spine [abstract]. In: Proceedings of the AACR Conference on Advances in Sarcomas: From Basic Science to Clinical Translation; May 16-19, 2017; Philadelphia, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(2_Suppl):Abstract nr A03.