Human pharmacokinetics of XBD173 and etifoxine distinguish their potential for pharmacodynamic effects mediated by translocator protein

XBD173 and etifoxine are translocator protein (TSPO) ligands that modulate inflammatory responses in preclinical models. Limited human pharmacokinetic data is available for either molecule, and the binding affinity of etifoxine for human TSPO is unknown. To allow for design of human challenge experiments, we derived pharmacokinetic data for orally administered etifoxine (50 mg 3 times daily) and XBD173 (90 mg once daily) and determined the binding affinity of etifoxine for TSPO. For XBD173, maximum plasma concentration and free fraction measurements predicted a maximal free concentration of 1.0 nM, which is similar to XBD173 binding affinity. For etifoxine, maximum plasma concentration and free fraction measurements predicted a maximal free concentration of 0.31 nM, substantially lower than the K i for etifoxine in human brain derived here (7.8 μM, 95% CI 4.5–14.6 μM). We conclude that oral XBD173 dosing at 90 mg once daily will achieve pharmacologically relevant TSPO occupancy. However, the occupancy is too low for TSPO mediated effects after oral dosing of etifoxine at 50 mg 3 times daily.

XBD173 and etifoxine are translocator protein (TSPO) ligands that modulate inflammatory responses in preclinical models. Limited human pharmacokinetic data is available for either molecule, and the binding affinity of etifoxine for human TSPO is unknown. To allow for design of human challenge experiments, we derived pharmacokinetic data for orally administered etifoxine (50 mg 3 times daily) and XBD173 (90 mg once daily) and determined the binding affinity of etifoxine for TSPO. For XBD173, maximum plasma concentration and free fraction measurements predicted a maximal free concentration of 1.0 nM, which is similar to XBD173 binding affinity.
For etifoxine, maximum plasma concentration and free fraction measurements predicted a maximal free concentration of 0.31 nM, substantially lower than the K i for etifoxine in human brain derived here (7.8 μM, 95% CI 4.5-14.6 μM). We conclude that oral XBD173 dosing at 90 mg once daily will achieve pharmacologically relevant TSPO occupancy. However, the occupancy is too low for TSPO mediated effects after oral dosing of etifoxine at 50 mg 3 times daily.

K E Y W O R D S
anti-inflammatory drugs, brain, etifoxine, translocator protein, XBD173

| INTRODUCTION
Pharmacological and genetic modulation of the translocator protein (TSPO) is immunomodulatory and biases microglia towards expression of immunosuppressive phenotypes in vitro. [1][2][3][4][5][6] Preclinical in vivo evidence for an anti-inflammatory effect of TSPO ligands also is compelling; in a range of neurodegenerative and inflammatory mouse models, TSPO ligands inhibit proinflammatory activation and improve clinical scores. 1,[6][7][8][9][10][11][12][13][14][15] These experiments suggest that TSPO may be a novel target for central nervous systems disorders which are partly driven by neuroinflammatory mechanisms. TSPO ligands have also been investigated as tools to enhance neurosteroid synthesis and hence as potential treatments for disorders driven partly by reduced neurosteroid concentrations. 16 XBD173 and etifoxine are TSPO ligands that have been explored extensively in preclinical models 14,[17][18][19][20][21][22][23][24][25][26][27][28] and clinical studies. 16 Etifoxine is licensed for the treatment of anxiety in France. XBD173 has previously been administered to humans in an experimental medicine study. 16 However, while both have been believed to be acting through modulation of TSPO, differences in pharmacodynamics have been reported. For David R Owen and Alexandra Phillips contributed equally to the manuscript The authors confirm that the Principal Investigator for this paper is Professor Paul Matthews and that he had direct clinical responsibility for participants. example, etifoxine improved clinical scores in an experimental autoimmune encephalomyelitis mouse model, whereas XBD173 did not. 20 Here we sought to gather appropriate pharmacokinetic and binding affinity data to enable use of these molecules in challenge studies in humans for exploration of effects of TSPO modulation. Although etifoxine is used clinically, there are no publicly available pharmacokinetic data. The binding affinity of etifoxine for TSPO in the rodent brain is approximately 12 μM, 29 but the affinity in the human brain has not been reported publicly. The influence of differences in TSPO structure with the common rs6971 polymorphism, which changes the affinity of TSPO for many ligands, 30  Pharmacokinetic studies with XBD173 suggested that a 90-mg dose produces peak plasma concentrations in the micromolar range. 16 However, positron emission tomography (PET) data subsequently showed only 80% TSPO peak occupancy in people with the rs6971 common variant (high affinity binders, HAB) 30 following a 90-mg dose. 31 This occupancy figure is substantially lower than would be predicted, as the affinity of XBD173 for TSPO in the human brain is 2-3 nM and hence micromolar concentrations would be expected to saturate it. 32 To better understand this dose occupancy relationship, we repeated pharmacokinetic analyses of XBD173 in healthy volunteers and conducted measurements of the plasma-free fraction.

| Pharmacokinetics
Plasma samples for pharmacokinetic analysis were obtained on day 1 at 0.5, 1, 2, 3, 4 hours following the first dose. As there are no data on whether etifoxine accumulates following repeated dosing, a further sample was taken on the final day of etifoxine dosing, approximately 2-3 hours following the final dose of the drug. This analysis was not performed for XBD173 as this molecule does not accumulate. 16 Plasma samples were stored at À20 C or lower until analysis. Plasma • There are limited data available on plasma concentrations and plasma free fraction following administration to humans, and on the TSPO binding affinity of etifoxine.
These are required to best design these experiments.

What does this study add
• The pharmacokinetic and brain-binding affinity data provided here are consistent with potential anti-inflammatory activity of orally administered XBD173 that is mediated by TSPO.
• The free plasma concentration of etifoxine is too low for pharmacologically relevant TSPO occupancy by etifoxine at the standard clinical dose control plasma over the ranges 2-10 000 ng/mL (XDB173) or 1-5000 ng/mL (etifoxine), then preparing and analysing as for the study samples. Lower limit of detection was 2 ng/mL for XDB173 and 1 ng/mL for etifoxine.  Sigma, UK. Nomenclature related to drugs and molecular targets conforms to the IUPHAR/BPS Guide. 33

| Participant characteristics
Four participants (1 female, 3 male) with mean age 51.5 years were separately dosed with XBD173 and etifoxine. Three participants were HABs and 1 was a mixed affinity binder. The washout period in between the dosing periods ranged from 28 to 42 days. All participants completed all visits. No adverse events were reported.

| Pharmacokinetics
Following oral administration of 90 mg XBD173, the median plasma maximum plasma concentration (Cmax) and time to Cmax were 114 ng/mL and 2.5 hours, respectively (Table 1). Following oral administration of 50 mg etifoxine, the median plasma Cmax and time to Cmax were 32 ng/mL and 2.0 hours, respectively. We did not find evidence for significant etifoxine accumulation after multiple doses: plasma etifoxine concentrations following the final dose on the morning of day 7 were not higher than on day 1 (Supplementary Table 2).

| Plasma free fraction
We assessed plasma free fractions using equilibrium dialysis and found

| Estimation of etifoxine K i in brain tissue
Competition assays with unlabelled etifoxine were performed with brain tissue from 8 donors (4 HABs, 4 LABs). The mean K i value for the HABs (7.6 ± 2.2 μM, n = 4) was similar to that of the LABs (7.6 ± 1.7 μM, n = 4; P = .99). The estimated K i value for the whole population when fitted to a single site binding curve was 7.8 μM (95% CI 4.5-14.6 μM; Figure 1). Due to the limitation on etifoxine solubility, the competition curve did not fully plateau.

| Estimates of TSPO occupancy
The median Cmax of 114 ng/mL and free fraction of 0.34%, suggests a predicted free plasma XBD173 concentration of approximately 1.0 nM and a TSPO occupancy of $30% in HABs.
For etifoxine, with an average Cmax was 32 ng/mL and free fraction of 0.29%, the predicted free etifoxine concentration was estimated to be approximately 0.31 nM. Given the affinity of etifoxine for TSPO estimated here (7.6 μM), this implies that TSPO occupancy of etifoxine at peak concentration will be <0.01%.

| DISCUSSION
XBD173 and etifoxine have been explored as possible immunomodulatory TSPO ligands in preclinical and in vitro models. Here we sought to gather appropriate pharmacokinetic and binding affinity data to enable design of TSPO challenge studies in humans for future pharmacodynamic investigations.
We found that the affinity of etifoxine for TSPO in the human brain is approximately 7.8 μM, irrespective of rs6971 genotype. This estimate is similar to measures previously reported for the rodent brain (12.5 μM), heart (22.5 μM) and kidney (14 and 9 μM). 29,35,36 However, following a 50-mg oral dose, we also estimated a subnanomolar free plasma concentration from our pharmacokinetic studies. Assuming equilibrium between free plasma concentration and brain tissue concentrations, this would equate to TSPO occupancy of approximately 0.01%. We also did not find evidence for drug accumulation after dosing at 50 mg TDS for 7 days. Indeed, plasma etifoxine concentration on day 7 was lower than on day 1, which may reflect variability or that a participant took the day 7 morning dose later than T A B L E 1 Pharmacokinetic parameters following oral administration of a single dose of 90 mg XBD173 or 50 mg etifoxine directed. We therefore conclude that any pharmacodynamic effects of etifoxine with a dose regimen of 50 mg TDS are not mediated by its interaction with TSPO. These low free plasma concentrations of etifoxine also appear inconsistent with pharmacological activity at the γ-aminobutyric acid-a receptor, the presumed target for anxiety. 37,38 Although the affinity has not been estimated in human tissue, in the rodent brain, etifoxine binds the γ-aminobutyric acid-a receptor with an IC50 of approximately 6.7 μM, and pharmacodynamic actions at the receptor are elicited only above 1 μM. 35 If plasma free concentration reflects concentration in the brain, any pharmacodynamic effect due to etifoxine at the standard dose is likely to be independent of this target.
PET data with XBD173 showed that a 90-mg dose is associated with only approximately 80% occupancy of TSPO in the brain, 31 despite plasma concentrations reaching 1 μM. 16  with TSPO occupancy. The disparity between our estimates of etifoxine binding affinity and free plasma concentration is substantial. It is therefore unlikely that more precise estimates would materially alter the conclusions. When estimating TSPO occupancy of etifoxine in the brain, we made the assumption that free plasma and free brain concentrations of etifoxine are in equilibrium. This may not be the case. To definitively determine whether etifoxine binds TSPO in the brain at the administered doses, a PET occupancy study would be required. Finally, we did not formally assess compliance, beyond measuring etifoxine plasma concentration on day 7. It is therefore possible that day 7 etifoxine concentrations were no higher than day 1 concentrations because the participants did not take the drug as directed.

| CONCLUSION
The pharmacokinetic and brain binding affinity data are consistent with potential anti-inflammatory activity of orally administered XBD173 that is mediated by TSPO. However, the free plasma concentration of etifoxine is too low for pharmacologically relevant TSPO occupancy by etifoxine at the approved clinical dose of 50 mg TDS.