Investigating the effect of P-glycoprotein efflux on ondansetron pharmacokinetics in the central nervous system

Neuropathic pain results from diseases that damage the somatosensory nervous system, which can cause sensory disturbances and spontaneous pain in patients. Despite available treatments, less than half of the patients with neuropathic pain experience sufficient pain relief, and there is an urgent need to find effective treatments for this disease. It has been reported that the serotonin 3 receptors (5HT3R) at the spinal dorsal horn are activated by descending serotonergic neurons to facilitate pain transmission in neuropathic pain. And ondansetron is a selective 5HT3R antagonist that has been investigated for the treatment of neuropathic pain. However, preclinical and clinical studies in the literature reported inconsistent results, and we hypothesize that this variability in drug response could be due to insufficient concentrations of ondansetron in the central nervous system (CNS) because of the P-glycoprotein (Pgp) efflux transporter at the blood-brain-barrier (BBB). The aim of this thesis was to assess whether Pgp inhibition will increase ondansetron distribution in the CNS for the treatment of neuropathic pain. In the introductory Chapter 1, the pathophysiology of neuropathic pain, literature evidence for the use of ondansetron in neuropathic pain, and species differences in Pgp are discussed. An overview of the microdialysis method and examples of different CNS pharmacokinetic models are also summarized. In Chapters 2 and 3, bioanalytical methods were developed and validated based on the FDA and EMA criteria for the quantification of ondansetron concentrations in plasma and CSF matrices of rats and humans using high-performance liquid chromatography with ultraviolet detection (HPLC-UV) and in rat serum and brain microdialysates using liquid chromatography-tandem mass spectrometry (LC-MS/MS). In Chapter 4, a microdialysis study was performed to determine the effect of Pgp on ondansetron distribution in specific regions of the brain that are relevant to pain modulation. Ondansetron was administered to genetic Pgp knock-out (KO) and wild-type (WT) rats, and unbound ondansetron concentrations were measured in the prefrontal cortex and the hippocampus of each rat. In this study, Pgp inhibition increased ondansetron distribution in the two brain regions by approximately 2-fold. However, there was no difference in ondansetron distribution between the prefrontal cortex and the hippocampus. This study confirmed our previous findings that Pgp at the BBB reduces ondansetron uptake into the brain.In Chapter 5, a physiologically-based pharmacokinetic model (PBPK) was developed to describe ondansetron pharmacokinetics in multiple CNS compartments with P-glycoprotein (Pgp) transport. This PBPK model consists of physiological CNS compartments, such as the brain ECF, CSF, and spinal cord, physiological flows of CSF and brain ECF, and Pgp efflux that determine ondansetron disposition in the CNS. For model fitting, ondansetron concentration-time profiles in the brain, spinal cord, CSF, and brain ECF were used. Our CNS PBPK model adequately described ondansetron distribution in various CNS compartments in both Pgp KO and WT animals, and parameters were estimated with sufficient precision. This PBPK model can be used to predict ondansetron CNS disposition in humans. The purpose of Chapter 6 was to evaluate the effect of Pgp on ondansetron disposition in the CNS in patients with neuropathic pain. Patients received two infusions of ondansetron (16 mg) that were administered separately on two occasions with either tariquidar (4 mg/kg), a Pgp inhibitor, or placebo. Serial plasma samples and a single CSF sample were collected from each patient of each group, and the ratio of ondansetron concentrations in CSF to plasma (Kp) was calculated in each patient to assess for ondansetron CNS distribution. And population pharmacokinetic modeling was used to evaluate plasma and CSF disposition of ondansetron in patients with neuropathic pain. In this clinical study, Pgp inhibition with tariquidar did not increase ondansetron CSF distribution in patients with neuropathic pain: the population estimate of Kp was 0.12 in both treatment and control groups. This finding contrasts with the difference that was observed in ondansetron CNS distribution between Pgp KO and WT animals in Chapter 4. Chapter 7 provides the general discussion and the future work of the thesis. In this thesis, the effect of Pgp on ondansetron disposition in the CNS was assessed in preclinical and clinical studies that produced contrasting results: Pgp reduced the brain distribution of ondansetron in animals, but this was not observed in patients with neuropathic pain. Further research is needed to ascertain the effect of Pgp on ondansetron CNS pharmacokinetics in humans by using mathematical modeling to predict human brain distribution of ondansetron.