EPCT-09. CNS LEVELS OF PANOBINOSTAT IN A NON-HUMAN PRIMATE MODEL: COMPARISON OF BLOOD AND CEREBROSPINAL FLUID PHARMACOKINETIC METHODS AND MALDI MSI

Adequate exposure (effective concentration over time) of a therapeutic agent at its site of action is essential for antitumor efficacy. Given constraints of repeat tissue sampling, non-human primate models predictive of pharmacokinetics in pediatric patients have been utilized to assess central nervous system (CNS) exposure. Assessment of cerebrospinal fluid (CSF) drug levels have been used to extrapolate CNS penetration but the relationship of CSF drug levels with tissue distribution is unclear. Utilizing microdialysis, we previously demonstrated geographic variability of drug permeability across the blood:brain barrier (BBB), but this technique is complex and has a high standard deviation. We, therefore, explored a novel technique, matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI), to compare plasma, CSF, and tissue drug levels in a terminal non-human primate model. Panobinostat, an HDAC inhibitor in clinical trials for DIPG/DMG, was selected for study as it has previously demonstrated poor CNS tissue penetration but suggested modest clinical activity. METHODS: Panobinostat (p.o., dose 1.6 mg/kg) was administered to non-tumor bearing primates (n=2). One hour following administration (Tmax), blood and CSF were collected, the animal euthanized, brain and spinal cord extracted, and immediately frozen at -80. Panobinostat distribution was mapped on ex vivo sagittal tissue sections using MALDI MSI. To provide specificity and degree of permeability, anatomical structures were segmented for analysis to determine drug concentrations. Blood, CSF and tissue levels of panobinostat were measured via LC-MS/MS. RESULTS: Segmentation analysis revealed quantifiable panobinostat, particularly in the lateral ventricles and choroid plexus, and also in the subventricular zone and brainstem, although the overall panobinostat concentration was below the limit of quantitation in these areas. CONCLUSIONS: Although not reflected in CSF PK, panobinostat is widely distributed in brain tissue. MALDI MSI allows regional assessment of panobinostat penetration and complements CSF pharmacokinetics.