Abstract 1784: Dynamic control of cerebrospinal fluid to achieve personalized drug delivery in the central nervous system

Background: Systemically delivered therapeutics have limited access to the brain due to the tightly regulated blood brain barrier. There is a fine balance between production (20 mL/h), volume (150 mL), and removal/turnover (every 5h) of cerebrospinal fluid (CSF). The high turnover of CSF and the active efflux makes CNS therapeutics particularly challenging. Direct delivery of therapeutics to the CSF intrathecally is increasingly utilized in clinical practice but penetration deep into the brain parenchyma is poor. This poses a significant challenge for the efficacy of intrathecally delivered small molecules, antibodies and gene therapies. Ideally, a CSF delivery system would have the capability to deliver therapeutics broadly within the CSF compartment, while simultaneously also improving parenchymal delivery. We show with our device how we can move dye between the two ventricles in a controlled fashion resulting in improved distribution of methotrexate. Methods: Sheep (n=8) were used as a model, which have a brain size and geometry closer to humans when compared to rodents. EnClear’s drug delivery system utilizes two CSF access points; Group 1 accessed the right and left lateral ventricle (V2V) and Group 2 accessed the right or left lateral ventricle and lumbar thecal sac (V2L). An extracorporeal pump was used to control the speed and direction of CSF flow at 50 μL/mL. Sensors continuously monitored pressure, respiration, and heart rate. Imaging studies with Omnipaque dye determined the best parameters for controlling CSF flow. Administration of methotrexate was followed by a protocol to move CSF between the two access points. Methotrexate levels were measured with LC/MS in CSF, blood, brain regions, spinal cord, kidney, and liver. Results: Omnipaque dye in V2V demonstrated the dye was observed to move between the two ventricles, whereas, in V2L, the dye disappeared within minutes from the ventricles. In Group 1 (V2V), safe pressures (< 25 mmHg or 34 cm water) were sustained in all subjects. In both groups, effective levels of >1 ug/mL methotrexate were measured in all brain tissues, including deeper brain tissues. There was a three-fold higher level of methotrexate in striatum with the V2V compared to V2L subjects. Consistently, CSF levels were two-fold higher at 2h. Importantly, methotrexate levels were significantly lower in lumbar spinal cord, kidney, and liver-potential sources of toxicity. Conclusions: In this large animal study, EnClear’s system delivered high(er) levels of methotrexate in the parenchyma of deep brain structures than traditional IT injection. To the authors knowledge, this first intraCSF drug delivery device to show improved distribution throughout the CNS including deep brain structures, while keeping lumbar spinal cord, liver, and kidney levels very low. Clinical trials with the device are currently being planned. Citation Format: Priya Kumthekar, Marcie Glicksman, Margaret Schwartz, Seema Nagpal, Isabella Glitza, Peter Forsyth, Ernesto Ruiz, Adam Sonabend. Dynamic control of cerebrospinal fluid to achieve personalized drug delivery in the central nervous system [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1784.