PROJECT SUMMARY High-grade Spinal Cord Glioma (SCG) is an orphan disease that results in significant morbidity and mortality, with no effective treatment options available. Despite significant advances in our knowledge of the disease process, there have unfortunately been limited changes to the clinical outcomes. In part, this represents the malignant nature of a disease that is refractory to the standard of care. On the other hand, this raises the question of the translational value of existing preclinical animal models, especially from a surgical standpoint – where widely scalable large animal models of SCG were previously unavailable. To this end, the Boulis and Canoll laboratories partnered to begin addressing this gap in the field by developing a minipig SCG model. Through lentiviral targeting of the well implicated RTK/RAS/PI3K and p53 pathways, our preliminary data demonstrates the induction of high-grade astrocytoma with histopathologic, radiologic, and transcriptomic characterization in 100% of minipigs. Consequently, we posit that the next steps to advancement of this model system are to modulate tumor phenotype and to demonstrate its utility in a directly translatable surgical application. In the enclosed proposal, we will begin by evaluating the induction of SCG by targeting common genetic lesions implicated in the human disease including PDGFB, P53, CDKN2A, EGFR, and PTEN (AIM 1). This represents the opportunity to produce highly characterized SCG lesions for therapeutic testing in an immunocompetent, more anatomically relevant, large animal model. In parallel, we will apply our existing minipig SCG model (AIM 2) to perform the first intra-tumoral convection enhanced delivery (CED) study for SCG in a large animal. Rodent studies of chemotherapeutic CED for SCG have reported suppression of tumor growth and amelioration of neurologic deficits. However, these data cannot be readily scaled for translation due to anatomic limitations of rodent systems. Despite an ongoing Phase I human trial for CED in SCG, drug distribution and CED parameters are poorly understood. Indeed, failures of CED in human trials for intracranial glioma can be attributed to both ineffective drug distribution and single treatments. As such, our study will employ implanted pumps for prolonged intratumoral CED. We will investigate parameters (flow rate, volume of infusion) to evaluate optimal readouts (volume of distribution, reflux, safety, radiologic vs chemotherapeutic distribution). These data will have immediate translational impact on present and future trials.