LAUSR

Intramedullary spinal cord tumors (IMSCTs) are a rare, heterogeneous group of neoplasms with limited treatment options and high rates of morbidity and mortality. Next-generation sequencing has revealed opportunities for targeted therapies of the intracranial counterparts of IMSCT, but little is known about the molecular features of IMSCT. To better understand the genetic basis of these tumors we performed whole exome sequencing on fifty-one IMSCT and matched germline DNA, including 29 ependymomas, 16 astrocytomas, 4 gangliogliomas,1hemangioblastoma, and 1 oligodendroglioma. Whole-genome sequencing was further performed on 12 IMSCT to discover possible structural variants. Though recurrent somatic mutations in IMSCTs were rare, we identified NF2 mutations in 15.7% of tumors (ependymoma, N = 7; astrocytoma, N = 1), RP1 mutations in 5.9% of tumors (ependymoma, N = 3), and ESX1 mutations in 5.9% of tumors (ependymoma, N = 3). We further identified copy number amplifications in CTU1 in 25% of myxopapillary ependymomas. Given the paucity of somatic driver mutations, we further performed whole-genome sequencing of 12 tumors (ependymoma, N = 9; astrocytoma, N = 3). Overall, we observed that IMSCTs with intracranial histologic counterparts did not harbor the canonical mutations associated with their intracranial counterparts (eg glioblastoma). Our findings suggest that the origin of IMSCTs may be distinct from tumors arising within other compartments of the central nervous system and provides a framework to begin more biologically based therapeutic strategies.