PCM-10: PLASMA AND CEREBROSPINAL FLUID PHARMACOKINETICS OF 5-AZACYTIDINE FOLLOWING INTRAVENOUS, INTRANASAL, AND INTRATHECAL ADMINISTRATION IN A NON-HUMAN PRIMATE MODEL

BACKGROUND: Promoter-associated hypermethylation resulting in transcription silencing has been identified in epigenetic cancer pathways for pediatric high-grade gliomas. The pyrimidine analog, 5-azacytidine (AZA), is a DNA demethylating agent that covalently binds DNA methyltransferases. The resulting hypomethylation may reactivate tumor suppressor genes and generate an antitumor effect. AZA has demonstrated activity in a pre-clinical murine astrocytoma model following systemic administration ^{(Borodovsky, et al)}. To evaluate the CNS exposure of AZA, plasma and CSF pharmacokinetics (PK) were examined in a nonhuman primate model following systemic (intravenous [IV] and intranasal [IN]) and intrathecal (IT) administration. METHODS: AZA was administered to rhesus macaques; IV (n = 4), IN (n = 1), and IT (n = 3). Serial paired plasma and CSF samples were collected. AZA was quantified via an LC-MS/MS assay and PK parameters calculated using noncompartmental methods. RESULTS: IV-One animal was unevaluable. Mean peak plasma concentration (C_max) was 1.54 ug/ml, half-life 0.21 min, AUC_0-∞ 0.729 hr*ug/ml. No AZA concentrations were quantifiable in the CSF. IN-One plasma sample (0.102 ug/ml) at 5 min and no CSF samples were quantifiable. IT- Mean plasma C_max 0.0268 ug/ml, half-life 1.08 hr, AUC_0-∞ 0.0622 hr*ug/ml. Mean CSF values were: C_max 125.43ug/ml, half-life 1.16hr, AUC_0-∞ 296.58 ug/ml*hr. A moderate transient pleocytosis in the CSF was noted at 24hr, resolved by 48 hr, and no clinical or neurological toxicity was noted. CONCLUSIONS: Systemic delivery of AZA, following IV or IN administration, does not provide adequate CSF exposure. However, IT delivery was tolerable and resulted in substantial and sustained CSF drug levels without evidence of neurologic toxicity.