PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of aacPermissionsJournals.ASM.orgJournalAAC ArticleJournal InfoAuthorsReviewers
 
Antimicrob Agents Chemother. 2009 November; 53(11): 4753–4761.
Published online 2009 August 31. doi:  10.1128/AAC.01541-08
PMCID: PMC2772352

Nonclinical Pharmacokinetics of Oseltamivir and Oseltamivir Carboxylate in the Central Nervous System[down-pointing small open triangle]

Abstract

Oseltamivir, a potent and selective inhibitor of influenza A and B virus neuraminidases, is a prodrug that is systemically converted into the active metabolite oseltamivir carboxylate. In light of reported neuropsychiatric events in influenza patients, including some taking oseltamivir, and as part of a full assessment to determine whether oseltamivir could contribute to, or exacerbate, such events, we undertook a series of nonclinical studies. In particular, we investigated (i) the distribution of oseltamivir and oseltamivir carboxylate in the central nervous system of rats after single intravenous doses of oseltamivir and oseltamivir carboxylate and oral doses of oseltamivir, (ii) the active transport of oseltamivir and oseltamivir carboxylate in vitro by transporters located in the blood-brain barrier, and (iii) the extent of local conversion of oseltamivir to oseltamivir carboxylate in brain fractions. In all experiments, results showed that the extent of partitioning of oseltamivir and especially oseltamivir carboxylate to the central nervous system was low. Brain-to-plasma exposure ratios were approximately 0.2 for oseltamivir and 0.01 for oseltamivir carboxylate. Apart from oseltamivir being a good substrate for the P-glycoprotein transporter, no other active transport processes were observed. The conversion of the prodrug to the active metabolite was slow and limited in human and rat brain S9 fractions. Overall, these studies indicate that the potential for oseltamivir and oseltamivir carboxylate to reach the central nervous system in high quantities is low and, together with other analyses and studies, that their involvement in neuropsychiatric events in influenza patients is unlikely.

Oseltamivir (Tamiflu; F. Hoffmann-La Roche Ltd.) is an orally administered antiviral drug for the treatment and prophylaxis of influenza A and B viruses. Following the oral administration of oseltamivir phosphate, the prodrug is converted by esterases to the active metabolite oseltamivir carboxylate (OC) (7). In influenza patients, OC selectively binds and potently inhibits the neuraminidase (NA) enzymes that are present on all influenza A and B viruses and are essential for the release of progeny viruses from infected host cells (20). In clinical studies, this action reduces the severity and duration of the symptoms of influenza virus in individuals with clinical illness or prevents the onset of symptoms in those with asymptomatic disease (24).

Recently, an increase in reports of neuropsychiatric adverse events (NPAEs) in influenza patients who were and were not exposed to oseltamivir was identified (29). The reporting frequency indicates that these events were rare. The majority of reports originated from Japan and usually involved children or young adolescents. No causal link with oseltamivir therapy could be identified, but the reports generated renewed interest in the central nervous system (CNS) tolerability profile of oseltamivir (5, 21). In response to these events, Roche, the manufacturer of oseltamivir, initiated a comprehensive safety review to firmly establish the nature of the reported events and investigate potential links with oseltamivir therapy (29). The review involved the assessment of all spontaneously reported NPAEs in the Roche Global Safety Database, review of the existing preclinical and clinical data sets, and assessment of epidemiological databases for information on the incidence of NPAEs in patients with and without influenza virus and/or drug exposure. From this comprehensive analysis, it was concluded that the incidence of NPAEs in patients with influenza virus receiving oseltamivir was no higher than that in those who had not received oseltamivir (1, 31) and that there was no plausible mechanism by which oseltamivir or OC could induce or exacerbate NPAEs (29).

In addition to the above-mentioned clinical safety review, a series of nonclinical in vitro and in vivo studies was performed with the objective of characterizing the distribution of oseltamivir and OC to the CNS, the active export of oseltamivir and OC from the CNS, and the conversion from oseltamivir to OC, including the potential for intracerebral conversion. The present article addresses the conduct and outcomes of these pharmacokinetic investigations.

MATERIALS AND METHODS

Distribution of oseltamivir and OC to the CNS and brain of rats.

Several studies were performed to characterize the pharmacokinetics of oseltamivir and OC in the plasma, cerebrospinal fluid (CSF), and brain of Sprague-Dawley rats following single-dose bolus administration of oseltamivir (intravenous [i.v.] and oral) and OC (i.v.). In the i.v. studies, nonfasted adult rats (two groups of 35 animals for each test substance) received a dose of 30 mg/kg body weight of either oseltamivir or OC in aqueous solution with sodium chloride (0.9%; pH 4.0) via slow injection into the tail vein over 20 to 30 s. In both i.v. studies, pharmacokinetic sampling took place at 5 min and at 0.25, 0.5, 1, 2, 4, and 8 h postdose (four or five rats/time point). In the oral study, rats received oseltamivir phosphate by oral gavage at a dose of 1,000 mg/kg free base, and sampling was performed at 1, 2, 4, 6, and 8 h postdose (four rats/time point).

Rats were terminally anesthetized using isoflurane (5% in oxygen) at each scheduled time point (at 5 min and at 0.25, 0.5, 1, 2, 4, and 8 h postdose for i.v. studies and at 1, 2, 4, 6, and 8 h postdose for oral studies), and approximately 0.5 ml of blood and as much CSF as possible were collected via puncture of the heart and cysterna magna, respectively. To investigate the effect of residual blood in brain tissue on the observed oseltamivir and OC concentrations in the i.v. studies, brain samples were obtained by whole-brain removal and homogenization in one group per test substance, while in the other group, brain tissue was perfused transcardially with physiological sodium chloride solution (0.9%; ca. 30 ml) before tissue collection and homogenization. Brain tissue perfusion was also performed in the oral study before collection and homogenization. All samples were stored at −20°C. All animal protocols were performed under the approval of an animal welfare officer and in accordance with Institutional Animal Care and Use Committee policies.

Bioanalytical assays were conducted using a specific and validated liquid chromatography-tandem mass spectrometry method, with quantification limits of 1 ng/ml and 10 ng/ml for oseltamivir and OC in plasma, respectively; 1 to 5 ng/ml and 10 to 50 ng/ml for oseltamivir and OC in CSF, respectively; and 60 ng/g for both analytes in brain homogenate. Generally, the interassay precisions (values for the coefficient of variation) were below 10% for plasma, CSF, and brain samples.

Where possible, the following pharmacokinetic parameters were estimated for plasma, CSF, and brain using pooled samples from each time point by noncompartmental analysis of composite concentration data: maximum concentration (Cmax) and time to maximum concentration, area under the curve (AUC) from 0 to 8 h postdose (AUC0-8 h) and from 0 h to infinity (AUC0-∞), apparent terminal half-life (t1/2), plasma clearance (CL), and volume of distribution at steady state (Vss). CSF/plasma, plasma/brain and CSF/brain concentration ratios were determined for each analyte, and oseltamivir/OC concentration ratios were determined for oseltamivir administration. Apart from calculating mean values, medians, and coefficients of variation, no formal statistical analysis was performed.

Active transport processes involved in the export of oseltamivir and OC from the CNS in humans.

The active transport of oseltamivir and OC by vesicles stably expressing breast cancer resistance protein 1 (BCRP1) and multidrug resistance protein (MRP1, MRP2, or MRP3) and the directional transport of oseltamivir by P-glycoprotein (P-gp) (P-gp/MDR1) were investigated in vitro. For the experiments with BCRP1 and MRP, vesicles were obtained from Solvo Biotechnology, Budapest, Hungary, and were prepared from insect cells expressing recombinant human BCRP1, MRP1, MRP2, or MRP3. Vesicle transport assays were run using three determinations as one set. Vesicles (20 to 50 μg depending on the transporter) in a volume of 20 μl SMS (50 nM sucrose, 100 nM KNO3, 10 mM HEPES/Tris, pH 7.4) buffer were prewarmed for 1 min at 37°C. Incubation solution [80 μl containing 50 mM sucrose, 100 mM KNO3, 10 mM Mg(NO3)2, and 10 mM HEPES-Tris (pH 7.4)] with or without ATP (5 mM), containing a control substrate {[3H]leucotriene C4 (LTC4) and [3H]estradiol-17β-glucuronide for MRP1, MRP2, and MRP3 and [3H]estrone sulfate for BCRP1} or 14C-labeled test substrate (3.8 μM oseltamivir or 3.1 μM OC), was added to the vesicles and incubated for 60 s at 37°C. These conditions were chosen to produce linear uptake with respect to time with optimal assay sensitivity. Incubation was stopped with 3 ml ice-cold “cold stop” solution (50 mM sucrose, 100 mM KCl, 10 mM Tris-HCl [pH 7.4]) containing 1 mM unlabeled taurocholate. Nitrocellulose filters (0.45 μm) were prewashed with 3 ml “cold stop” solution containing 1 mM taurocholate. Vesicles were immediately filtered through a prewashed nitrocellulose filter and washed twice with 3 ml “cold stop” solution using a rapid-filtration system. Filters were dissolved in 10 ml filter count scintillation solution, and radioactivity was counted by use of a scintillation counter.

The method used for the P-gp directional transport experiments was previously reported (26). Transcellular transport measurements were initiated by adding the labeled substrate [14C]oseltamivir or [3H]digoxin together with an extracellular marker (Lucifer yellow) to either the apical or the basolateral (donor) side of monolayers of LLC-PK1 cells transfected with P-gp. Multiple substrate concentrations (oseltamivir, 0.1 to 10 mM) in the absence or in the presence of the P-gp inhibitor elacridar (1 μM) were used. For the inhibition studies, unlabeled oseltamivir was added to the donor and receiver sides (0.1 to 10 mM), and labeled digoxin (10 nM) was used as a substrate. The inserts were incubated at 37°C under an atmosphere of 5% CO2. Samples were taken from the opposite (acceptor) side after 30 min and 1, 2, and 3 h of incubation. Concentrations of substrate in the acceptor side were determined by scintillation counting. The extracellular marker was quantified using a Spectrafluor Plus reader. In each experiment, three different inserts were taken for each condition, and a mean was calculated.

The following equation was used for data evaluation: Pe = 1/(A·C0dQ/dt, where Pe, A, C0, and dQ/dt represent the apparent permeability coefficient, filter surface area, initial concentration, and amount transported per time period, respectively. Pe values were calculated on the basis of single time points (3 h).

Transport ratios were calculated as follows: Pe values in the basolateral-to-apical direction were divided by Pe values in the apical-to-basolateral direction. Pe values were not corrected for the flux of the extracellular marker Lucifer yellow, which was used to assess the quality of the cell monolayers. For kinetic determinations of oseltamivir, active A-to-B transport was calculated as the difference between passive diffusion (A-to-B transport in the presence of the P-gp inhibitor elacridar) and apparent transport (without inhibitor) as described elsewhere previously (19).

The following formula was used for data evaluation: v = Papp·S·CaVmax·Ca/(Km + Ca), where S is the filter surface area (33 mm2), Ca is the substrate concentration in the apical compartment, Papp is the apparent permeability, and Vmax and Km are the Michaelis-Menten parameters for P-gp (transport in the basolateral-to-apical direction). Evaluations of the kinetic parameters, including the curve fittings, were performed using Origin 7.0 data analysis and graphing software (OriginLab Corporation, Northampton, MA).

Conversion of oseltamivir to OC by the recombinant human carboxylesterases HCE1 and HCE2.

The hydrolysis of oseltamivir to OC was investigated using recombinant baculoviruses for HCE1 and HCE2. A wild-type baculovirus was used as a control. Sf9 insect cells (40 ml) were cultured with shaking at 100 rpm in non-CO2 incubators at 27°C and infected with baculoviruses using viable cell numbers of 1 to 2 million cells/ml. Culture supernatants from 5 to 7 days' culture time containing the same level of cell disruption by the baculovirus were used for activity testing. The protein concentration of each of the selected cell culture supernatants was then determined using the bicinchoninic acid method.

To test the cell culture supernatants for activity, incubations of methyl anthranilate (MA) (HCE1 selective), procaine (HCE2 selective), and oseltamivir were prepared. For MA and procaine, triplicate 200-μl incubation mixtures containing 100 μM substrate in 100 mM Tris-HCl buffer (pH 7.4) (0.5% [vol/vol] dimethyl sulfoxide) were warmed to 37°C for 5 min before the addition of 0.125 mg/ml (MA) or 1.25 mg/ml (procaine) cell culture supernatant protein. For oseltamivir, triplicate 500-μl incubation mixtures containing 10 μM 14C-labeled oseltamivir in 100 mM Tris-HCl buffer were warmed to 37°C for 5 min before the addition of 1.0 mg/ml cell culture supernatant protein. Reactions were stopped after 10 min (MA), 20 min (procaine), and 30 min (oseltamivir) by the addition of cold quench reagent (3:2 mixture of 10% trichloroacetic acid-acetonitrile [67 μl for MA and P and 167 μl for oseltamivir]). Incubation mixtures were then centrifuged at 20,000 × g for 10 min, and 33 μl was analyzed by high-pressure liquid chromatography (HPLC) with UV detection and collection of chromatogram fractions for offline scintillation counting of 14C material. Positive control (human liver S9) and negative control (no protein) incubations were performed in parallel for each substrate and analyzed in the same way.

After incubation, the extent of MA, procaine, and oseltamivir hydrolysis was analyzed using a Shimadzu LC-20 high-pressure liquid chromatograph equipped with a Superspher RPselect 250-mm by 3-mm, 5-μm-particle-size column. Mobile phases were 20 mM ammonium acetate (pH 4.5) (A) and acetonitrile (B); the flow rate was 0.6 ml/min.

For methyl anthranilate analysis a linear gradient profile was used, with the acetonitrile concentration increasing from 10% initially to 90% after 20 min, followed by a wash-and-reequilibration cycle. UV (254 nm) analysis was used. MA and o-aminobenzoic acid (o-ABA) eluted after 14.3 and 9.2 min, respectively. The use of authentic reference standards established that MA was detected with 1.25-fold-greater sensitivity than o-ABA under these conditions. The fractional conversion of MA to o-ABA was calculated as follows: 1.25·o-ABA peak area/[MA peak area + (1.25·o-ABA peak area)].

For procaine analysis a linear gradient profile was used, with the acetonitrile concentration increasing from 10% initially to 90% after 10 min, followed by a wash-and-reequilibration cycle. UV (320 nm) analysis was used. Procaine and p-aminobenzoic acid (p-ABA) eluted after 6.8 and 6.3 min, respectively. The use of authentic reference standards established that procaine was detected with 2.7-fold-greater sensitivity than p-ABA under these conditions. The fractional conversion of procaine to p-ABA was calculated as follows: 2.7·p-ABA peak area/[procaine peak area + (2.7·p-ABA peak area)]. Correction for p-ABA matrix background was made by subtracting the apparent fraction conversion value calculated for samples at 0 min of incubation.

For oseltamivir analysis a linear gradient profile was used, with the acetonitrile concentration increasing from 10% initially to 50% after 25 min, followed by a wash-and-reequilibration cycle. The chromatograms were fractionated into 0.16-min fractions, which were collected in yttrium silicate-coated 96-well plates (deep-well Luma plates; Perkin-Elmer Life Sciences), and the radioactivity in each fraction was counted using a Perkin-Elmer TopCount NXT instrument. Oseltamivir and OC were collected in the 18.7- and 8.8-min fractions, respectively. The fractional conversion of oseltamivir to OC was calculated as follows: radioactivity in OC peak/total radioactivity in chromatogram. Correction for nonenzymatic hydrolysis was made by subtracting the apparent fraction conversion value calculated for samples incubated in the absence of enzymes.

The rates of hydrolysis were then determined for each compound as follows: (pmol/ml initial substrate in incubation × fraction hydrolyzed)/(mg protein/ml × incubation time).

Extent of local conversion of oseltamivir to OC in human and rat liver and brain fractions.

The brains and livers of 10 male and 10 female 7-day-old rats, two male and two female 42-day-old rats, and human brain samples (inferior frontal gyrus tissue) from two male donors aged 77 and 86 years were obtained. S9 fractions were prepared by adding ~4 ml Tris-sucrose buffer per gram of tissue and homogenizing. After centrifugation at 9,000 × g at 10°C, the supernatant material (S9 fraction) was removed, and the protein concentration of each sample was determined using the bicinchoninic acid method. Commercially available pooled human brain and liver S9 fractions were also obtained (Celsis InVitro Technologies, Brussels, Belgium). This pool came from four individual donors comprising two male Caucasians aged 68 and 80 years and two female Caucasians aged 85 and 86 years.

Control esterase activity incubation mixtures containing (finally) 100 μM paranitrophenyl acetate (PNP-Ac) and 0.05 mg brain or liver S9 protein in 1.5 ml Tris-HCl buffer (pH 7.4) were prepared. Samples were incubated at 37°C for 1 min and quenched with acetonitrile. Incubation mixtures were then centrifuged at 20,000 × g for 10 min, the supernatant was removed, and 50 μl was analyzed by HPLC with UV detection. Blank incubations (no protein) were also performed to establish the stability of PNP-Ac under these experimental and analysis conditions. Triplicate 1-ml oseltamivir incubation mixtures containing (finally) 10 μM 14C-labeled oseltamivir and 1.0 mg brain or liver S9 protein in 100 mM Tris-HCl buffer (pH 7.4) at 37°C were made up. All samples were incubated for 30 min at 37°C before being quenched with 0.33 volumes of a 3:2 mixture of 10% trichloroacetic acid-acetonitrile. Incubation mixtures were then centrifuged at 20,000 × g for 10 min, the supernatant was removed, and 333 μl was analyzed by radioflow HPLC with chromatogram fractionation and solid scintillation counting. Negative control (no-protein) incubations were also performed to establish the stability of oseltamivir under the experimental and analysis conditions.

For both PNP-Ac and oseltamivir analyses, HPLC with UV detection (calibrated using external standards) was used to determine the amount of hydrolysis product generated. For oseltamivir analysis, the fractional conversion of oseltamivir to OC was calculated from the amount of radioactivity in the OC fractions compared to the total chromatogram radioactivity. The rate of hydrolysis was determined as follows: (pmoles/ml initial substrate in incubation × fraction hydrolyzed)/(mg protein/ml × incubation time). The apparent background degradation rate determined from the negative control samples was then subtracted to give the enzymatic hydrolysis activity.

RESULTS

Distribution of oseltamivir and OC to the CNS and brain of rats.

The main pharmacokinetic parameters for oseltamivir and OC in plasma, CSF, and brain homogenate (from perfused and nonperfused brains) following single i.v. and oral dosing of oseltamivir and i.v. dosing of OC in rats are shown in Table Table1.1. For the i.v. studies, the concentration-time profiles of oseltamivir and OC are shown in Fig. Fig.11 and and22.

FIG. 1.
Composite concentration-time profiles for oseltamivir (OP) and OC after single i.v. (bolus) administration of 30 mg/kg oseltamivir to male rats for plasma, CSF, and brain (A) and brain only (B).
FIG. 2.
Mean composite plasma, CSF, and brain concentration-time profiles after single i.v. (bolus) administration of 30 mg/kg OC to male rats.
TABLE 1.
Pharmacokinetic parameters of oseltamivir and OC in plasma, CSF, and brain (composite data) after single i.v. bolus or oral administration of oseltamivir or OC to adult male ratsa

Following oseltamivir administration, Cmax concentrations of oseltamivir in plasma, CSF, and brain were observed at 5 min postdose. In plasma, oseltamivir concentrations declined in a multiexponential manner, and rats were exposed to oseltamivir throughout the sampling period. For oseltamivir, the systemic CL was 93.9 ml/min·kg, and the Vss was ~3.0 liters/kg. After oseltamivir administration, rapid cleavage to OC occurred in plasma. OC concentrations after oseltamivir administration decreased, with a t1/2 of 1.7 h (versus 1.5 h for oseltamivir), and AUCs for OC were ~2.5 times higher than those for oseltamivir. CSF concentrations of oseltamivir declined roughly in parallel to plasma concentrations but were markedly lower (Cmax of ~900 ng/ml versus ~12,000 ng/ml). After oseltamivir administration, OC concentrations were below the limit of quantification in CSF (<50 ng/ml). For oseltamivir, the CSF/plasma AUC ratio was ~0.09.

For samples from nonperfused brains, Cmax values for oseltamivir were ~50% higher than those for perfused ones (712 and 491 ng/g) (Fig. (Fig.1).1). However, the brain/plasma exposure ratio (AUC0-∞) of oseltamivir was only slightly higher for nonperfused samples (0.24) than for perfused samples (0.21). After i.v. oseltamivir administration, the Cmax for OC in brain occurred at 0.25 h postdose, with values of 164 and 75.5 ng/g in nonperfused and perfused brain samples, respectively. Brain/plasma concentration ratios of OC were substantially lower than those of oseltamivir, with Cmax values of 0.02/0.01 (nonperfused/perfused) and respective AUC ratios of ~0.01/<0.004 (nonperfused/perfused).

As with oseltamivir, the i.v. administration of OC produced Cmaxs for OC in plasma, CSF, and brain at 5 min postdose. OC concentrations declined in a multiexponential manner, and rats were exposed to OC throughout the whole sampling period, with an exposure almost sixfold higher following OC administration (AUC0-∞ values of 28,600 and 33,800 for perfused and nonperfused brains, respectively) than OC concentrations after oseltamivir administration. A systemic CL value of ~16 ml/min·kg and a Vss of ~0.8 liters/kg were observed. In CSF, the Cmax for OC after direct administration was 250 ng/ml, markedly lower than the plasma Cmax, and OC concentrations were below the limit of quantification (<50 ng/ml) at 2 h postdose and beyond. CSF/plasma ratios for Cmax were in the range of ~0.003 to 0.016 and increased with time until 1 h postdose.

In brain, the Cmax of OC was markedly higher for samples from nonperfused animals than for perfused samples (1,710 and 605 ng/g) (Fig. (Fig.2).2). Brain/plasma concentration ratios for nonperfused samples ranged from 0.02 to 0.05, while brain/plasma ratios for AUC were ~0.024. For perfused samples, brain/plasma ratios for AUC were ~0.009, 2.7-fold lower than those for nonperfused samples.

In the pharmacokinetic study, after the oral administration of 1,000 mg/kg oseltamivir, peak plasma concentrations were reached at 2 h postdose for oseltamivir and 8 h for OC (Table (Table1).1). Rats were exposed to oseltamivir over the whole sampling interval and had a ~2.7-fold-higher rate of exposure to OC than oseltamivir. In CSF, peak concentrations were reached at 2 h postdose for oseltamivir and 6 h for OC. CSF/plasma exposure ratios (AUC0-8 h) were ~0.07 for oseltamivir and 0.007 for OC. In perfused brain samples, peak concentrations were reached at 8 h postdose for oseltamivir and 6 h for OC. Brain/plasma exposure ratios (AUC0-8 h) of ~0.12 for oseltamivir and 0.01 for OC were recorded. Corresponding CSF/brain exposure ratios ranged between ~0.55 and 0.64 for both analytes. A further group of animals that received a single oral administration of oseltamivir at a lower dose produced similar results (data not shown).

Active transport processes involved in export of oseltamivir and OC from the CNS in humans.

In membrane vesicles expressing either BCRP1, MRP1, MRP2, or MRP3 and control vesicles from Sf9 cells, control substrates (estrone-3-sulfalte, LTC4, LTC4, and estradiol-17β-glucuronide for BCRP1, MRP1, MRP2, and MRP3, respectively) were transported in the presence of ATP (150, 30, 10, and 74 pmol/mg protein/min for BCRP1, MRP1, MRP2, and MRP3, respectively) but not in control vesicles. No active transport of oseltamivir or OC was observed. Transport was not ATP dependent for oseltamivir with any transporter and for OC with BCRP1, MRP1, and MRP3 and was comparable for wild-type vesicles, indicating mainly unspecific binding. OC showed a small but likely irrelevant transport by MRP2 in three independent experiments (3.3, 0.7, and 1.9 pmol/mg protein/min), and wild-type vesicles indicated mainly unspecific binding. No inhibition of LTC4 transport by MRP2 transporters was observed in the presence of up to 100 μM OC.

The results of kinetic studies performed for P-gp (MDR1) in LLC-MDR1 cells are shown in Tables Tables22 and and33 and Fig. Fig.33 and and4.4. The transport of oseltamivir was saturable (the apparent Km values in two different experiments were 4.2 and 10.5 mM, with Vmax values of 8.9 and 10.9 pmol/s, respectively). The passive diffusion values of oseltamivir (transport from A to B in the presence of the P-gp inhibitor elacridar) were calculated to be 59.8 and 38.9 nm/s in the two experiments. The effect of oseltamivir on the bidirectional transport of the MDR1 model substrate drug digoxin was also tested. Digoxin showed the expected high apically directed transport (export ratios of 9.9 and 8.8), indicating good MDR1 functionality. This transport was completely inhibited in the presence of the MDR1 inhibitor verapamil (80 μM) as a positive control. Oseltamivir inhibited the digoxin transport mediated by MDR1 with 50% inhibitory concentrations of 1.3 and 2.3 mM in the two experiments.

FIG. 3.
Kinetics of directional transport (A to B) of oseltamivir in human LLC-MDR1 cells expressing P-gp in the absence or presence of the P-gp inhibitor elacridar (one of two experiments).
FIG. 4.
Inhibitory effect of oseltamivir on the directional transport (A to B and B to A) of digoxin in human LLC-MDR1 cells (two independent experiments; mean values for three wells and standard deviations are shown). IC50, 50% inhibitory concentration. ...
TABLE 2.
Kinetics of directional transport (A to B) of oseltamivir in human LLC-MDR1 cells in the absence or presence of the P-gp inhibitor elacridar
TABLE 3.
Apparent permeability and export ratio of directional transport of oseltamivir in human LLC-MDR1 cells in the absence or presence of the P-gp inhibitor elacridar

Conversion of oseltamivir to OC by the recombinant human carboxylesterases HCE1 and HCE2.

The principal MA, procaine, and oseltamivir esterase activity findings are presented in Fig. Fig.5.5. Consistent with previously reported findings (27), cell culture supernatants containing HCE1, but not those containing HCE2, catalyzed the ester hydrolysis of MA and oseltamivir, while the opposite was true for procaine.

FIG. 5.
MA, procaine, and oseltamivir ester hydrolysis activities of HCE1-containing, HCE2-containing, and wild-type virus control cell culture supernatants.

Extent of local conversion of oseltamivir to OC in human and rat liver and brain fractions.

The hydrolysis activities of rat and human S9 fractions toward PNP-Ac and oseltamivir are shown in Table Table4.4. All rat and human S9 fractions were active in the hydrolysis of PNP-Ac but exhibited very little oseltamivir esterase activity. Quantifiable data could not be generated for the 7-day-old rat brain samples. For both brain and liver, fractions from 42-day-old rats had higher oseltamivir esterase activity than did those from 7-day-old rats, although the OC generation rate was still two- to fourfold less than that of pooled liver S9 fractions from the same animals. A four- to sixfold increase in rat liver S9 activities was seen between 7-day-old and 42-day-old samples, indicating an age-dependent increase in liver oseltamivir esterase activity. Human brain S9 showed very little oseltamivir esterase activity (~0.1% conversion, equivalent to ~0.3 pmol/min/mg); human brain S9 fractions had ≥300-times-lower oseltamivir esterase activity per milligram of protein than human liver S9 fractions.

TABLE 4.
Esterase activities of rat and human liver and brain S9 fractions toward PNP-Ac and oseltamivira

DISCUSSION

The studies described here further clarify the pharmacokinetic profile of oseltamivir (prodrug) and OC (active metabolite) in the CNS. The main pathways of oseltamivir distribution, metabolism, and excretion are qualitatively similar in animals and humans (7). The main differences are the rate and location of carboxylester cleavage (15). While in the rat, oseltamivir is cleaved predominantly in the plasma (16), in humans (and monkeys), the cleavage takes place in the liver (7). The major pathways of oseltamivir and OC distribution in the body are illustrated in Fig. Fig.66.

FIG. 6.
Major pathways for CNS penetration and metabolic fate of oseltamivir and OC.

In the rat, the partitioning of oseltamivir and OC to the brain and CSF was very low following single i.v. doses of oseltamivir and OC and a single high oral dose of oseltamivir. This is consistent with data from previously described quantitative whole-body autoradiography experiments using ferrets and rats after oral doses of radiolabeled oseltamivir, which showed a limited distribution of drug to the brain (concentrations were generally <20% of those in plasma) (4, 7). Importantly, perfusion of the brain tissue to remove residual blood prior to homogenization resulted in lower observed brain concentrations of oseltamivir and, especially, OC than for nonperfused samples. Therefore, the present data for brain exposure should be considered more accurate than previously reported data from studies in which perfusion was not performed.

These data are also consistent with the low level of CNS exposure to oseltamivir and OC described for humans. A recent exploratory study of healthy volunteers (four each of Caucasian and Japanese origins) measured 24-h CSF and plasma pharmacokinetic profiles of oseltamivir and OC after a single 150-mg dose of oseltamivir. CSF penetration was very limited: the mean CSF/plasma Cmax ratios for all patients were 2.11% (standard deviation, 0.52%) for oseltamivir and 3.47% (2.94%) for OC; the ratios based on AUC0-last were 2.35% (1.09%) and 2.93% (4.06%), respectively (10). A low level of CNS exposure to oseltamivir and OC may be expected given the physicochemical and pharmacokinetic properties of each. Oseltamivir is more lipophilic than OC and therefore more likely to cross the blood-brain barrier (BBB) by passive transport, as shown previously for rodents (22), but is actively transported out of the brain. As for OC, penetration into the CNS appears very limited in rats, with exposure values >10-fold lower than those for oseltamivir (see also reference 4). This can be understood as a result of a concomitant low passive transport into and, possibly, active transport out of the brain.

Several systems found in the endothelial cells of the BBB are known to actively transport drugs out of the CNS, including P-gp (MDR1), BCRP, and MRP (17). In the active-transport experiments performed, oseltamivir was a good substrate for the P-gp system but not for BCRP1, MRP1, MRP2, or MRP3. The impairment or inhibition of the P-gp transport system is likely to increase oseltamivir brain concentrations, as demonstrated in P-gp knockout mice (the brain/plasma concentration ratio was five to six times that for wild-type mice) (19). Under normal conditions, therefore, P-gp is thought to actively transport oseltamivir out of the brain.

OC is a weak substrate for organic anion transporter 1 (OAT1) (8) and, as recently reported by Ose et al. (23), a substrate of OAT3 and MRP4. All three transporters are expressed in the BBB (11, 12, 25) and could theoretically contribute to the active transport of OC out of the brain. A functional role for MRP4 was suggested by the finding that in Mrp4 knockout mice, the exposure of OC in brain was fourfold higher than that in wild-type animals (23). In Oat3 knockout mice, however, no overall differences in exposure were observed. Nevertheless, those authors speculated that the brain distribution of OC is limited by active efflux by both Mrp4 and Oat3 across the BBB (23). However, very recent data suggest that the clinical relevance of these observations may be questionable because of a low or even undetectable level of expression of MRP4 and OAT3 in the human BBB compared with rodent tissue (T. Terasaki, Y. Uchida, K. Ito, Y. Katsukura, C. Ikeda, H. Kawakami, J. Kamiie, and S. Ohtsuki, presented at the 8th Cerebral Vascular Biology International Conference, Sendai, Japan, 28 June to 2 July 2009). Another possibility for the distribution of OC into the brain is the potential local conversion of oseltamivir to OC by brain esterases. As demonstrated here, HCE1, but not HCE2, was active in the conversion of oseltamivir to OC. This is consistent with data reported previously by Shi et al. (27).

Importantly, the current data indicate only a limited and slow conversion of oseltamivir to OC in rat and human brain S9 fractions compared with equivalent liver fractions, in agreement with previously reported findings showing that in rat and human brains, only low levels of mRNA expression of the carboxylesterase HCE1 (and HCE2) were detected (33). Together, these data suggest that the conversion of oseltamivir to OC in the brain is unlikely to play a significant role. Additionally, a third human carboxylesterase, HCE3, was previously reported to be particularly expressed in brain tissue (9). It is currently not known whether HCE3 could contribute to the hydrolysis of oseltamivir in humans. Mori et al. previously speculated that HCE3 may contribute to protecting the CNS from ester compounds (18). The low overall rate of conversion of oseltamivir to OC in rat and human brain S9 fractions, however, also points to a low activity of this HCE isoform.

Taking all these factors together (enzymes and transporters), predictions of the limits of brain exposure to oseltamivir can be obtained and compared to extrapolated human exposure data to obtain safety margins. Oseltamivir brain exposure is driven by passive influx and active efflux by P-gp, whereas conversion in the brain is not thought to play an important role. As brain exposure will be driven by systemic exposure, potential effects of carboxylesterase variability need to be considered as well. In other words, an increase of systemic concentrations can indirectly lead to high oseltamivir concentrations in the CNS. In a previous report by Shi et al. (27), several naturally occurring HCE1 polymorphic variants with different in vitro hydrolysis rates from the wild-type enzyme were described, and genetic variation in the HCE1 gene was also reported previously by other groups (3, 28). The clinical relevance of these polymorphisms is unclear, and there are currently only sparse clinical data on the effect of genetic esterase polymorphisms on oseltamivir (and OC) pharmacokinetics in humans (30). Nevertheless, a complete blockade of the conversion of oseltamivir to OC could theoretically occur and lead to enhanced oseltamivir concentrations. Estimates of the maximum achievable oseltamivir exposure were investigated with a population pharmacokinetic model constructed with data from five healthy-volunteer studies using nonlinear mixed-effect modeling (29). Simulations assuming a complete inhibition of oseltamivir-to-OC conversion following standard 75-mg twice-daily dosing produced a mean steady-state Cmax plasma value for oseltamivir of 555 ng/ml, 14 times higher than those simulated with the normal metabolism scenario (29). It could be speculated that similar increases would be observed in the brains of such patients. However, the predicted theoretical maximum increase of oseltamivir in the highly unlikely situation of a complete inhibition of oseltamivir-to-OC conversion in combination with a complete absence of P-gp (factors of 14- and 5- to 6-fold, respectively) would still not be expected to be clinically relevant due to the very large safety margins of the compound in rodents (plasma, ≥300-fold; brain, >500-fold) and monkeys (plasma, ≥250-fold), compared to plasma and to extrapolated brain concentrations for humans with standard treatment (Roche).

For OC, such predictions of the limits of brain exposure are more difficult to establish. OC cannot easily cross the BBB by passive influx due to its polarity, and in addition, active efflux by MRP4 and possibly OAT3 as well as only low local conversion of oseltamivir by carboxylesterases in the brain further add to the low observed exposure levels in the brain. The combined effects of these different processes on human brain OC levels are difficult to quantify based on in vitro and in vivo rodent data; that is, they are mostly interdependent and can work in different directions (e.g., whereas a reduction in CES1 activity potentially leads to a decrease in OC brain levels, reduced levels of MRP4 activity would have an opposite effect). Nevertheless, even if significant increases in brain OC levels occurred, these would not be clinically meaningful because of the very large safety margins of the compound in rodents (plasma, >240-fold; brain, >114-fold) and monkeys (plasma, no observed adverse effect level, 74-fold, 251-fold; no CNS-related effects) compared to plasma and to extrapolated brain concentrations in humans with standard treatment (Roche).

To determine the likely effect, if any, of the presence of small concentrations of oseltamivir and OC in the CNS, several pharmacological investigations were conducted. Animal studies showed no specific CNS and/or behavioral effects after oral dosing equivalent to ≥100 times the clinical dose (29), and neither oseltamivir nor OC showed any relevant pharmacological off-target activities on human neuraminidases or 155 known molecular targets in radioligand binding and functional assays up to high concentrations (29).

Human infection with the highly pathogenic avian influenza A/H5N1 virus has been associated with systemic infection, and virus has been recovered from the CNS (32). While OC is systemically distributed to all sites relevant for viral replication outside the CNS (13), oseltamivir treatment of H5N1 infection in mice and ferrets is also associated with decreased viral titers in the brain (2, 6, 14). Although OC is not present in the CNS in relevant quantities, the inhibition of viral penetration to the CNS may be the consequence of an inhibition of viral replication in the lung, resulting in limited systemic infection.

In summary, oseltamivir was present in the CNS of rats at low concentrations following oral or i.v. dosing. The active export of oseltamivir by P-gp was demonstrated, explaining why oseltamivir did not reach high concentrations in the CNS. For OC, the active export from the CNS by the MRP4 transporter was reported recently. This active process, together with very limited passive transport due to its low lipophilicity, is thought to explain the minimal CNS penetration observed in rats. The conversion of oseltamivir to OC was shown to proceed via a single pathway (HCE1), confirming previously reported observations. There is no clinical evidence of a relevant impact on oseltamivir conversion with HCE1 polymorphisms. However, even a complete absence of ester cleavage would not be expected to lead to oseltamivir concentrations that would have an impact on safety. Finally, it was demonstrated that the intra-CNS conversion of oseltamivir to OC was both limited and slow, suggesting no substantial potential for the CNS build-up of OC over time via intracerebral conversion from oseltamivir. Taken within the wider context of a comprehensive drug safety review, the findings from the investigations presented here confirm the low potential for oseltamivir and OC to cause the NPAEs observed in a small proportion of subjects infected with influenza virus who received oseltamivir. The presented data are consistent with reviews of the Roche Global Safety Database, existing clinical and nonclinical data sets, and assessments of epidemiological databases, which concluded that the incidence of NPAEs in influenza patients is likely to be disease rather than drug related (29).

Acknowledgments

The manuscript was prepared with the assistance of Stephen Purver and colleagues at Gardiner-Caldwell Communications Ltd., Macclesfield, United Kingdom. We give special thank to all team members at F. Hoffmann-La Roche, especially to Alberto Guenzi, Herbert Birnboeck, Michael Pantze, Christian Freichel, and Muriel Bellot for their support of the studies; Philippe Coassolo, David Reddy, James Smith, Brian Davies, and Klaus Klumpp for their scientific advice; and Laurent Gand, Renée Portmann, Marie-Elise Brun, Paul Schmid, Georg Schmid, Marie Stella Gruyer, Christelle Rapp-Mary, Véronique Dall'Asen; and Caroline Kreuzer for their excellent technical assistance.

Financial support for the preparation of this paper was provided by F. Hoffmann-La Roche Ltd., the manufacturers of oseltamivir. All authors are current employees of Roche, except C. Rayner, who was a Roche employee at the time of study conduct and manuscript submission.

Footnotes

[down-pointing small open triangle]Published ahead of print on 31 August 2009.

REFERENCES

1. Blumentals, W. A., and X. Song. 2007. The safety of oseltamivir in patients with influenza: analysis of healthcare claims data from six influenza seasons. Med. Gen. Med. 9:23. [PubMed]
2. Boltz, D. A., J. E. Rehg, J. McClaren, R. G. Webster, and E. A. Govorkova. 2008. Oseltamivir prophylactic regimens prevent H5N1 influenza morbidity and mortality in a ferret model. J. Infect. Dis. 197:1315-1323. [PubMed]
3. Charasson, V., R. Bellott, D. Meynard, M. Longy, P. Gorry, and J. Robert. 2004. Pharmacogenetics of human carboxylesterase 2, an enzyme involved in the activation of irinotecan into SN-38. Clin. Pharmacol. Ther. 76:528-535. [PubMed]
4. Freichel, C., E. Prinssen, G. Hoffmann, L. Gand, M. Beck, T. Weiser, and A. Breidenbach. Oseltamivir is devoid of specific behavioral and other central nervous system effects in juvenile rats at supratherapeutic oral doses. Int. J. Virol., in press.
5. Goshima, N., T. Nakano, M. Nagao, and T. Ihara. 2006. A clinical study of abnormal behaviors in patients with influenza. Infect. Immun. Child. 18:371-376.
6. Govorkova, E. A., I. A. Leneva, O. G. Goloubeva, K. Bush, and R. G. Webster. 2001. Comparison of efficacies of RWJ-270201, zanamivir, and oseltamivir against H5N1, H9N2, and other avian influenza viruses. Antimicrob. Agents Chemother. 45:2723-2732. [PMC free article] [PubMed]
7. He, G., J. Massarella, and P. Ward. 1999. Clinical pharmacokinetics of the prodrug oseltamivir and its active metabolite Ro 64-0802. Clin. Pharmacokinet. 37:471-484. [PubMed]
8. Hill, G., T. Cihlar, C. Oo, E. S. Ho, K. Prior, H. Wiltshire, J. Barrett, B. Liu, and P. Ward. 2002. The anti-influenza drug oseltamivir exhibits low potential to induce pharmacokinetic drug interactions via renal secretion—correlation of in vivo and in vitro studies. Drug. Metab. Dispos. 30:13-19. [PubMed]
9. Imai, T. 2006. Human carboxylesterase isozymes: catalytic properties and rational drug design. Drug Metab. Pharmacokinet. 21:173-185. [PubMed]
10. Jhee, S. S., M. Yen, L. Ereshefsky, M. Leibowitz, M. Schulte, B. Kaeser, L. Boak, A. Patel, G. Hoffmann, E. P. Prinssen, and C. R. Rayner. 2008. Low penetration of oseltamivir and its carboxylate into cerebrospinal fluid in healthy Japanese and Caucasian volunteers. Antimicrob. Agents Chemother. 52:3687-3693. [PMC free article] [PubMed]
11. Kikuchi, R., H. Kusuhara, T. Abe, H. Endou, and Y. Sugiyama. 2004. Involvement of multiple transporters in the efflux of 3-hydroxy-3-methylglutaryl-CoA reductase inhibitors across the blood-brain barrier. J. Pharmacol. Exp. Ther. 311:1147-1153. [PubMed]
12. Kikuchi, R., H. Kusuhara, D. Sugiyama, and Y. Sugiyama. 2003. Contribution of organic anion transporter 3 (Slc22a8) to the elimination of p-aminohippuric acid and benzylpenicillin across the blood-brain barrier. J. Pharmacol. Exp. Ther. 306:51-58. [PubMed]
13. Kurowski, M., C. Oo, H. Wiltshire, and J. Barrett. 2004. Oseltamivir distributes to influenza virus replication sites in the middle ear and sinuses. Clin. Drug Investig. 24:49-53. [PubMed]
14. Leneva, I. A., N. Roberts, E. A. Govorkova, O. G. Goloubeva, and R. G. Webster. 2000. The neuraminidase inhibitor GS 4104 (oseltamivir phosphate) is efficacious against A/Hong Kong/156/97 (H5N1) and A/Hong Kong/1074/99 (H9N2) influenza viruses. Antivir. Res. 48:101-115. [PubMed]
15. Li, B., M. Sedlacek, I. Manoharan, R. Boopathy, E. G. Duysen, P. Masson, and O. Lockridge. 2005. Butyrylcholinesterase, paraoxonase, and albumin esterase, but not carboxylesterase, are present in human plasma. Biochem. Pharmacol. 70:1673-1684. [PubMed]
16. Li, W., P. A. Escarpe, E. J. Eisenberg, K. C. Cundy, C. Sweet, K. J. Jakeman, J. Merson, W. Lew, M. Williams, L. Zhang, C. U. Kim, N. Bischofberger, M. S. Chen, and D. B. Mendel. 1998. Identification of GS 4104 as an orally bioavailable prodrug of the influenza virus neuraminidase inhibitor GS 4071. Antimicrob. Agents Chemother. 42:647-653. [PMC free article] [PubMed]
17. Loscher, W., and H. Potschka. 2005. Blood-brain barrier active efflux transporters: ATP-binding cassette gene family. NeuroRx 2:86-98. [PubMed]
18. Mori, M., M. Hosokawa, Y. Ogasawara, E. Tsukada, and K. Chiba. 1999. cDNA cloning, characterization and stable expression of novel human brain carboxylesterase. FEBS Lett. 458:17-22. [PubMed]
19. Morimoto, K., M. Nakakariya, Y. Shirasaka, C. Kakinuma, T. Fujita, I. Tamai, and T. Ogihara. 2008. Oseltamivir (Tamiflu) efflux transport at the blood-brain barrier via P-glycoprotein. Drug. Metab. Dispos. 36:6-9. [PubMed]
20. Moscona, A. 2005. Neuraminidase inhibitors for influenza. N. Engl. J. Med. 353:1363-1373. [PubMed]
21. Okumura, A., T. Kubota, T. Kato, and T. Morishima. 2006. Oseltamivir and delirious behavior in children with influenza. Pediatr. Infect. Dis. J. 25:572. [PubMed]
22. Ose, A., H. Kusuhara, K. Yamatsugu, M. Kanai, M. Shibasaki, T. Fujita, A. Yamamoto, and Y. Sugiyama. 2008. P-glycoprotein restricts the penetration of oseltamivir across the blood-brain barrier. Drug Metab. Dispos. 36:427-434. [PubMed]
23. Ose, A., M. Ito, H. Kusuhara, K. Yamatsugu, M. Kanai, M. Shibasaki, M. Hosokawa, J. D. Schuetz, and Y. Sugiyama. 2009. Limited brain distribution of [3R,4R,5S]-4-acetamido-5-amino-3-(1-ethylpropoxy)-1-cyclohexene-1-carboxy- late phosphate (Ro 64-0802), a pharmacologically active form of oseltamivir, by active efflux across the blood-brain barrier mediated by organic anion transporter 3 (Oat3/Slc22a8) and multidrug resistance-associated protein 4 (Mrp4/Abcc4). Drug Metab. Dispos. 37:315-321. [PubMed]
24. Oxford, J. 2005. Oseltamivir in the management of influenza. Expert Opin. Pharmacother. 6:2493-2500. [PubMed]
25. Pavlova, A., H. Sakurai, B. Leclercq, D. R. Beier, A. S. L. Yu, and S. K. Nigam. 2000. Developmentally regulated expression of organic ion transporters NKT (OAT1), OCT1, NLT (OAT2), and Roct. Am. J. Physiol. Renal Physiol. 278:635-643. [PubMed]
26. Schwab, D., H. Fischer, A. Tabatabaei, S. Poli, and J. Huwyler. 2003. Comparison of in vitro P-glycoprotein screening assays: recommendations for their use in drug discovery. J. Med. Chem. 46:1716-1725. [PubMed]
27. Shi, D., J. Yang, D. Yang, E. L. Lecluyse, C. Black, L. You, F. Akhlaghi, and B. Yan. 2006. Anti-influenza prodrug oseltamivir is activated by carboxylesterase human carboxylesterase 1, and the activation is inhibited by antiplatelet agent clopidogrel. J. Pharmacol. Exp. Ther. 319:1477-1484. [PubMed]
28. Tanimoto, K., M. Kaneyasu, T. Shimokuni, K. Hiyama, and M. Nishiyama. 2007. Human carboxylesterase 1A2 expressed from carboxylesterase 1A1 and 1A2 genes is a potent predictor of CPT-11 cytotoxicity in vitro. Pharmacogenet. Genomics 17:1-10. [PubMed]
29. Toovey, S., C. Rayner, E. Prinssen, T. Chu, B. Donner, B. Thakrar, R. Dutkowski, G. Hoffmann, A. Breidenbach, L. Lindemann, E. Carey, L. Boak, R. Gieschke, S. Sacks, J. Solsky, I. Small, and D. Reddy. 2008. Assessment of neuropsychiatric adverse events in influenza patients treated with oseltamivir: a comprehensive review. Drug Saf. 31:1097-1114. [PubMed]
30. Wattanagoon, Y., K. Stepniewska, N. Lindegårdh, S. Pukrittayakamee, U. Silachamroon, W. Piyaphanee, T. Singtoroj, W. Hanpithakpong, G. Davies, J. Tarning, W. Pongtavornpinyo, C. Fukuda, P. Singhasivanon, N. P. Day, and N. J. White. 2009. Pharmacokinetics of high-dose oseltamivir in healthy volunteers. Antimicrob. Agents Chemother. 53:945-952. [PMC free article] [PubMed]
31. Wilcox, M., and S. Zhu. 2007. Oseltamivir therapy appears not to affect the incidence of neuropsychiatric adverse events in influenza patients, poster 114. Abstr. 47th Intersci. Conf. Antimicrob. Agents Chemother. American Society for Microbiology, Washington, DC.
32. Writing Committee of the Second World Health Organization Consultation on Clinical Aspects of Human Infection with Avian Influenza A (H5N1) Virus, A. N. Abdel-Ghafar, T. Chotpitayasunondh, Z. Gao, F. G. Hayden, D. H. Nguyen, M. D. de Jong, A. Naghdaliyev, J. S. Peiris, N. Shindo, S. Soeroso, and T. M. Uyeki. 2008. Update on avian influenza A (H5N1) virus infection in humans. N. Engl. J. Med. 358:261-273. [PubMed]
33. Xie, M., D. Yang, L. Liu, B. Xue, and B. Yan. 2002. Human and rodent carboxylesterases: immunorelatedness, overlapping substrate specificity, differential sensitivity to serine enzyme inhibitors, and tumor-related expression. Drug Metab. Dispos. 30:541-547. [PubMed]

Articles from Antimicrobial Agents and Chemotherapy are provided here courtesy of American Society for Microbiology (ASM)