A high-performance liquid chromatography–tandem mass spectrometric method for the determination of pharmacokinetics of ganaxolone in rat, monkey, dog and human plasma

Abstract

A method for determining concentration levels of ganaxolone in rat, monkey, dog and human plasma was validated in the range of 5–1500 ng/ml using a 200-μl plasma sample volume. This validation report describes the linearity, specificity, sensitivity, reproducibility, accuracy, recovery and stability of the analytical method. The inter-day C.V. ranged from 0.5 to 9.2%, intra-day C.V. from 0.7 to 8.8% and intra-day accuracy (mean absolute percentage difference) ranged from 0.0 to 14.0% for rat, monkey, dog and human plasma. The method was used for the routine analysis of ganaxolone in rat, monkey, dog and human plasma and summary of the pharmacokinetic data are presented.

Introduction

Therapeutically useful anticonvulsants, anxiolytics and sedative-hypnotics such as benzodiazepines (BZ) and barbiturates mediate their action by binding to distinct allosteric modulatory sites on the GABA_A receptor-Cl⁻ channel complex. There is now a large body of evidence for an additional modulatory site on GABA_A receptors that binds neuroactive steroids [1]. The prototypical ligand for this binding site is epianopregnanolone (3a-hydroxy-5a-pregnan-20-one) an endogenous metabolite of progesterone that demonstrates potent modulatory effects at the GABA_A receptor.

Ganaxolone (CCD 1042) (1) is a 3β-methyl-substituted analog of the endogenous neuroactive steroid 3a-hydroxy-5a-pregnan-20-one (2). It is a high-affinity, stereoselective, positive allosteric modulator of the GABA_A receptor complex in the brain through a unique recognition site and exhibits potent anticonvulsant activity in a broad range of animal seizure models [2].

In radioligand-binding studies ganaxolone allosterically displaced [³⁵S]-t-butylbicyclophosphorothionate (TBPS) binding to the chloride channel and enhanced [³H]-flunitrazepam binding to the benzodiazepine receptor in the presence of GABA. Electrophysiological studies demonstrate the ability of ganaxolone to potentiate the GABA-evoked currents in Xenopus oocytes expressing the human GABA_A-receptor subunits a₁, β₁ and γ_2L [2].

In order to correlate activity with plasma levels and to understand the pharmacokinetics of ganaxolone, it was necessary to develop a sensitive and specific method for the determination of ganaxolone in animal plasma with the limit of quantification (LOQ) of at least 5 ng/ml. Due to lack of a chromophore and extremely poor UV absorbance of ganaxolone, the development of an assay based on high-performance liquid chromatography (HPLC) with UV absorption detection was not feasible. An HPLC–UV method was developed using derivatization of the 3-hydroxy group to increase the sensitivity and lower the limit of detection (LOD) but this method had poor reproducibility and was limiting in many ways. A GC–MS method was successfully validated and used in assaying for ganaxolone in rat and dog plasma, but the run time was 39 min. and LOQ was only 15 ng/ml.

An assay procedure with liquid phase extraction and gas chromatography (GC) with electron capture detection (ECD) was developed for selectively assaying ganaxolone in plasma samples. Electron capture detectability and specificity for ganaxolone was achieved by derivatization with pentafluorobenzyl hydroxylamine (PFBHA) in pyridine at 65°C for 1 h to form stable oximes. The procedure was used to quantitate plasma levels of unchanged ganaxolone following oral and parental administration to mice, rats, dogs and humans. Plasma samples (1 ml) containing ganaxolone and an internal standard (I.S.) were extracted with hexane and derivatization was conducted. The extraction recovery over the standard curve range averaged around 98% for both ganaxolone and I.S. The GC analysis was carried out on 10 μl of the reconstituted extracts injected on to a DB17 capillary column with hydrogen as carrier gas. GC column and oven temperature programs were used over a run time of ∼35 min. The retention times of ganaxolone and I.S. were 23.5 min. and 26.0 min. respectively. The assay was linear over the concentration range of 4–500 ng/ml with a minimum quantifiable limit was 4 ng/ml using 1 ml of plasma. Intra-day and inter-day coefficients of variation averaged around 4.2%. The absolute percentage differences found in the accuracy determination were less than 10.1%. The derivatization procedure [3] increased the sensitivity and lowered the limit of detection (LOD).

The above methods are limited in terms of long run times, having to deal with issues of stability of the derivatized product and the inability to detect metabolites. An alternative and more sensitive assay for ganaxolone and its metabolites in plasma was evaluated using HPLC with MS–MS detection method [4] with a run time of <5 min. ESI–LC–MS–MS is the method of choice for many classes of steroid compounds and is more sensitive a method [5]. Ganaxolone was found to be very sensitive in negative ion mode ESI [6], [7] but only at extremely high pH conditions (pH>12) and was poorly ionized in positive ion mode ESI. Ganaxolone predominantly formed sodium ion adducts in positive ion mode ESI and attempts to fragment this sodium adduct ion were unsuccessful. This difficulty in fragmenting ganaxolone–sodium adduct ions is consistent with the behavior of many other sodium adduct ions. However, the high pH condition was not ideal for chromatographic separation.

Steroids have been commonly analyzed using liquid–liquid extraction and APCI–LC–MS–MS [8], [9]. Ganaxolone was found to be quite fragile in APCI interface due to the fact that APCI probes were typically operated at high temperatures (above 400°C). Subsequently, ganaxolone loses one molecule of water easily regardless of the conditions (such as lower probe and source temperature and lower cone voltage) and thus the conditions were carefully optimized to gain maximum sensitivity for the protonated molecular ion. This could be due to the fact that dehydration of ganaxolone would involve the formation of a stable tertiary carbocation, which makes it easier for ganaxolone to lose a molecular of water as compared to any secondary alcohol. We were therefore left with a choice of developing a method based on the dehydrated molecular ion. In one respect, this increased the specificity of the method, since it enhanced selectivity in the MS–MS detection.

Influence of eluent composition and ionization efficiency has been extensively studied [10]. In the process of optimizing conditions for dehydrated ganaxolone, it was found that small amount of formic acid (0.05%, other acids such as acetic acid showed the same effect) helped the sensitivity. Additional amounts of acid (adding up to 0.5% formic acid in the mobile phase was tried) did not increase sensitivity. This effect could be explained easily, as the presence of small amount of acid could facilitate the dehydration process.

The objective of the studies was to develop a rugged, specific method with sufficient sensitivity that can be applied to analyze large amount of pre-clinical and clinical samples. A fully validated [11] bioanalytical method including inter and intra-day precision and accuracy was used to determine the pharmacokinetics of ganaxolone in rats, monkeys, dogs and human volunteers after oral administration.