Toxicokinetics and Hydrolysis of Artelinate and Artesunate in Malaria-Infected Rats

Abstract

Comparative toxicokinetic (TK) and hydrolysis studies of intravenously administered two new antimalarial agents, artelinate (AL) and artesunate (AS), were performed in malaria-infected rats using three daily equimolar doses (96 μmoles/kg). The TK evaluation was related to select one drug for severe malaria treatment in U.S. Army. Drug concentration of AS with daily dose of 36.7 mg/kg was one-third less on day 3 than on day 1, which resembled its active metabolite, dihydroartemisinin (DHA), suggesting an autoinduction of hepatic drug-metabolizing enzymes for AS. The results were similar to other artemisinin drugs, but not for AL. TK parameters of AL were very comparable from day 1 to day 3 at same AS molecular dose at 40.6 mg/kg. AS is the prodrug of DHA with the DHA/AS ratio of 5.26 compared to the ratio of 0.01 for DHA/AL. Other TK parameters revealed that the total AUC_{1–3 days} (84.4 μg · h ml⁻¹) of AL was fivefold higher than that of AS (15.7 mu;g h ml⁻¹ of AS plus DHA). The elimination half-life of AL (7.1 h) was much longer than that of AS (0.36 h) or DHA (0.72 h). The remarkable alteration of the TK shape of AL may be caused by poor conversion rates to DHA and an enterohepatic circulation, which is confirmed by the present TK and tissue distribution studies. Compared to AS, higher drug exposure levels and longer exposure time of AL in the rat blood may be the cause of its increased toxicity.

Artemisinin class compounds act rapidly against drug-resistant Plasmodium falciparum strains, and are widely used for the treatment of severe malaria in humans. Dihydroartemisinin (DHA) is obtained by sodium borohydride reduction of artemisinin an endoperoxide containing sesquiterpene lactone, which was isolated by Chinese researchers and characterized as the antimalarial principle of the plant Artemisia annua. In vitro bioassay tests have shown DHA to be more potent (3.8 to 5.2 times) than artemisinin. However, due to its poor solubility in water or oils, DHA has only been formulated as an oral preparation and has been used primarily as a semisynthetic compound for derivatization to the oil-soluble drugs, artemether and arteether, and the water-soluble drugs, artesunate (AS) and artelinate (AL) (de Vries and Dien 1996). The effectiveness of AS has been attributed to its rapid and extensive hydrolysis to DHA (Batty et al. 1998a; Davis et al. 2001; Li et al. 1998a; Navaratnam et al. 2000). DHA is three- to fivefold more active and more toxic than other artemisinin derivatives (Li et al. 2002; Mclean and Ward 1998). It can completely inhibit parasite growth within 2 to 4 h, and is the only artemisinin derivative with activity against all asexual blood stage parasites (Skinner et al. 1996). Thus, AS represents the gold standard against which any new candidate artemisinin derivative, for example AL, should be compared.

AL is generally considered to be a less potent antimalarial compound in vitro than either AS or DHA (Bustos, Gay, and Diquet 1994; Milhous et al. 1996; Shmuklarsky et al. 1993). However, AL has a superior antimalarial potency than AS in malaria-infected rats (Li et al. 2003). This is despite the fact that the metabolic conversion of AL to DHA is minimal (Li et al. 1998a; Grace, Skanchy, and Aguilar 1999) and may relate to a greater transfer of AL into erythrocytes, higher plasma concentrations, and longer residence time of AL (Li et al. 1998b). AL is also superior to AS in terms of its greater stability in vitro metabolic system and biological fluid (Li et al. 2003; Lin, Klayman, and Milhous 1987). Recently, we found out that AL also caused a mild to moderate renal failure and urinary excretion inhibition at the clearance dose (40.6 mg/kg) in malaria-infected rats (Xie et al. 2003), which may alter the pharmacokinetics (PK), efficacy, and toxicity parameters of AL.

The PK parameters of AS in humans are very well known but the same is not true for AL. Detailed pharmacokinetic data for AS and its active metabolite DHA have been reported in adults (Batty et al. 1998a, 1998b; Davis et al. 2001; Newton et al. 2000) and children (Bethell et al. 1997) with malaria, and in healthy volunteers (Benakis et al. 1997; Na-Bangchang et al. 1998). These results indicate that AS has an elimination half-life (t _1/2) of 2 to 5 min and DHA has a t _1/2 in the order of 40 to 60 min. In animals, both drugs have been clearly defined (Li et al. 1998b; Titulaer, Eling, and Zuidema 1993). In rats, AL and AS have been compared under the same conditions following intravenous (i.v.) administration. The results show that compared to AS, AL has much higher plasma concentrations (fivefold) with poor conversion to DHA, although AS is converted almost immediately into DHA with a very low Cmax and area under the curve (AUC). Both AS and AL present extremely fast PK with short half-lives from 0.35 to 1.35 h, respectively.

The conversion of AS to DHA in human seems to be much more efficient than in animal species. The ratio of AUC_DHA to AUC_AS in malaria-infected humans had a ratio in the range of 4.3 to 9.7 compared to uninfected rats and dogs with a ratio of only 0.5 to 0.6 (Batty 1998a, 1998b; Li et al. 1998a, 1999), indicating that the ratio in infected humans was 4.5 to 13.5 times higher. In animal species, DHA to AL ratio was found to be less than 0.001 to 0.017 (Li et al. 1998a, 1998b), suggesting that AL is very stable for the hydrolysis.

AL is currently under preclinical development for the treatment of multidrug-resistant and severe malaria. AS, another water-soluble artemisinin derivative, has been in clinical use as a treatment for severe malaria for more than 10 years. However, neither drug has been formally registered for use in the United States. Recently, the efficacy and toxicity of AL and AS have been systematically compared in rats and rhesus monkeys for the candidate selection in our institution.

Although both drugs (AL and AS) were found to be similarly effective, AL proved even better in the rat malaria model (Li et al. 2003). The data showed that the minimum dose required for parasitemia clearance in 100% of animals was 96 μmoles/kg (40.6 mg/kg) for AL, whereas AS at 36.7 mg/kg cleared parasitemia in 39% to 67% of animals, both following daily dosing for 3 days. At this dose level, a mild (in male animals) and moderate (in female rats) nephrotoxicity was seen with AL in malaria-infected rats (Xie et al. 2003). The purpose of the study was to determine which candidate drug (AL or AS) would yield a better toxicokinetics profile at the minimum clearance dose (96 μmoles/kg) in Plasmodium berghei rat malaria model.

MATERIALS AND METHODS

Chemicals

Artelinic acid (4-(10′ dihydroartemisinin-oxymethyl) benzoic acid hemihydrate) was manufactured as an l-lysine salt and obtained from the Walter Reed Chemical Inventory System. [13-¹⁴C]artelinic acid (specific activity 12.4 mCi/mmol) was synthesized by the Research Triangle Institute (Research Triangle Park, NC, USA). l-Lysine monohydrochloride was obtained from Sigma Chemical. Guilin Pharmaceutical II Factory (Guangxi, China) manufactured AS/NaHCO₃, which was imported by Atlantic Pharmaceutical, Bangkok, Thailand. Solutions of 0.9% saline was purchased from Abbott Labs. (Chicago, IL). Pentobarbital, sodium citrate, heparin, d-glucose, glycerol, and methanol were purchased from Sigma Chemical. Hema 3 stain was obtained from Fisher Scientific.

Animals and Parasites

ANKA strain of P. berghei to be used in this study had been rat-adapted from a mouse strain by three successive 4-week passages through 7-week-old male rats. Parasitized blood from these animals was cryopreserved in a large batch and was used to inoculate donor animals for the efficacy and toxicity experiments. Seven-week-old (young adults;180 to 200 g) Sprague-Dawley (S.D.) rats (Charles River Labs, Raleigh, NC) uninfected and infected with P. berghei were randomly assigned into study groups of 6 or 10 animals. Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals, National Research Council (NRC) Publication, 1996 edition.

Preparation of Infected Rats

All animals were quarantined (stabilized) for at least 7 days prior to infection. Rats were individually housed with food and water, supplied ad libitum. Rats were inoculated intraperitoneally (i.p.) with cryopreserved P. berghei–infected rat blood (2 × 10⁷/rat in 0.5 ml glucose citrate solution), obtained from donor rats infected 1 week earlier with cryopreserved parasites. Two pretreatment smears were taken from all animals to determine parasitemia. Animals with >4% of parasitemia were selected for the efficacy and tolerant dose studies. Eighteen post-treatment smears were obtained from each rat at 0, 3, 5, 8, and 12 h on day 6, and at 0, 3, and 6 h on days 7 and 8. From day 9 until day 21, blood smears were obtained from each animal once daily.

Formulations of AL and AS

The preformulated AL/lysine salt was prepared with 1:1 molecular weight of AL with lysine. The salt weight was equivalent to the real weight of AL for study use. The salt was dissolved in the solution containing 0.45% NaCl/0.1% l-lysine (w/v) and was prepared by mixing 0.9% NaCl with equal volume of 0.2% l-lysine in sterile water yielding 40.6 mg/ml of AL. AS/NaHCO₃ injection, a Chinese formulation, was selected to compare with the AL/lysine salt in the study. AS was purchased from Thailand as drug sets, containing 60 mg AS powder (per vial) and 1 ml of 5% sodium bicarbonate. A stock i.v. solution was prepared daily by dissolving 60 mg of AS powder in 0.6 ml of 5% sodium bicarbonate and 1.04 ml 5% glucose injection or normal saline injection so that each ml contained 36.7 mg/ml of AS. The dosing solution was prepared fresh every day before the experiments. The dosing solutions were analyzed by using a high-performance liquid chromatography–electrochemical detection (HPLC-ECD) system.

Multiple PK of AL and AS in Malaria-Infected Rats

Two groups of infected rats with 5% to 7% parasitemia received AL at 40.6 mg/kg and AS at 36.7 mg/kg intravenously once daily through jugular veins for 3 days. The first dose was given 6 days after parasite inoculation. A series of samples were collected during the absorption, distribution, and elimination phases for a full PK analysis on the first and last dosing days. Additionally, trough levels (predose) samples were collected in between the first and last doses to follow possible drug accumulation and a series PK approach (23 samples per animal). Thus, each animal was dosed and its plasma samples were obtained for up to 3 days. A total of 23 samples were collected (at 0, 5, 20, 40, and 60 min, and 1.5, 2, 3, 5, 8, 24, 25, 48, 48.08, 48.33, 48.67, 49, 49.5, 50, 51, 53, 56, and 72 h) in cooled vials. From predosing to 1 h post dosing, 50 μl of blood was collected for each sample point and 100 μl was obtained during each sample point between 1.5 and 24 h after dosing. Thus, total 1.9 ml blood was obtained from single rat during the 3 days treatment period (0.9 ml of blood on day 1, 0.1 ml on day 2, and 0.9 ml on day 3), which is 9.8% of total blood volume per rat. Blood samples were mixed with heparin/saline (1 unit ml⁻¹) to total volume of 250 μl. All samples were analyzed by using liquid chromatography–mass spectrometry–mass spectrometry (LC/MS/MS) or HPLC-ECD.

Single PK and Tissue Distribution of [¹⁴C]AL in Uninfected Rats

The PK of [¹⁴C]AL was evaluated in uninfected rats following i.v. administration of approximately 10 mg/20 μCi/kg. Radioactivity levels in blood and plasma after dosing were determined using liquid scintillation spectrometry. Population blood samples (approximately 0.5 ml) were collected from the jugular vein of each rat prior to dosing (0 h), and at 0.08, 0.25, 0.5, 1, 1.5, 2, 4, 8, 12, and 48 h after i.v. dosing. Additionally, blood was collected from the abdominal aorta of each rat at the time of euthanasia. Reverse-phase HPLC with ultraviolet (UV) detection was used to quantify plasma concentrations of unchanged AL.

Autoradiography study of the radioactivity AL ([¹⁴C]AL) was carried out after a single i.v. administration of 8.8 mg/kg (54 μCi/kg) into a tail vein. The rats were sacrificed at 1, 6, 24, 48, 96, and 192 h following dosing. Terminal blood samples were obtained for radioanalysis. The rats were embedded, and sagittal cryostatic sections were prepared from three regions of each rat for the whole body autoradiography. All films were developed in a Kodak X-OMAT Clinical Processor (Eastman Kodak Company, Rochester, NY). The autoradiographs of trial sections developed at intervals indicated that 3- and 6-week exposure intervals would be appropriate for evaluation of autoradiographs from rats sacrificed at early and late time points following treatment. The comparison described in this report was based on films exposed for the same intervals.

Automated Blood Sampler (ABS)

The Culex ABS is a blood sampling and metabolism monitoring machine, which is controlled by its own internal computer. We were successful in evaluating TK for long-term periods by using this system. After jugular vein cannulation, study animals were individually placed in the system, which allowed for a micropump to infuse heparinized saline (10 units heparin ml⁻¹ saline) at a rate of 1 to 2 μl/min to prevent coagulation within the catheter. Culex obtained precise blood volumes from the study animals and maintained total collected volume at less than 6% per day and 14% during the entire study period. Blood samples were maintained at 3°C using Culex’s temperature-controlled fractional collector. Plasma was separated from the blood samples and processed for drug extraction. The detailed sampling process was described previously (Reisen, Lothrop, and Meyer 1997; Tian et al. 2002).

LC/MS/MS Assay and Sample Preparation

Rat plasma samples were analyzed for AL with an LC/MS/MS procedure in a PE SCIEX API III triple-quadrupole mass spectrometer (Perkin-Elmer/ABI, Foster City, CA) with a C₈ column, 35% CH₃CN, 35% MeOH, 0.1% TFA mobile phase and mass spectrometric detection with sample inlet by heated nebulizer (Thornhili, Ontario, Canada), positive ionization by APCI (atmospheric pressure chemical ionization), and mass scanning by MRM (multiple reaction monitoring) analysis. Collisional activations were accomplished for MS/MS using 90% argon/10% nitrogen as the collision gas at energy of 25 eV.

Sample were prepared by adding 2 times volume of acetonitrile containing indomethacin internal standard (IS), mixing of the mixture for 1 min, centrifugation for 5 min, and transferring the supernatant to an autosampler injection vial prior to separation by LC/MS/MS. Standard curve and quality control (QC) samples were generated by spiking interference free rat plasma samples with known amounts of AL and IS. Standard curve, QC, and assay samples were prepared. Then, ∼40-μl aliquots were injected into the LC/MS/MS system for chromatographic separation and subsequent mass spectrometric detection. The peak area ratios of AL (high range daughter ion at 220 m/z and low range daughter at 162 m/z from parent ion at 373 m/z) to IS (daughter ion at 139 m/z from parent ion at 359 m/z) were calculated for each sample from the measured peak areas obtained by LC/MS/MS. Spiked concentrations and peak area ratios of the standard curve samples were fit by 1/y weighted least squares linear regression to the equations for the best straight lines (y = mx + b, where y = peak area ratio and x = AL concentrations), and drug concentrations in assay samples were calculated by these equations from the AL to IS peak area ratios obtained by LC/MS/MS.

HPLC-ECD Assay

HPLC with reductive ECD was performed utilizing a model BAS 200B liquid chromatography system (Bioanalytical Systems, West Lafayette, IN). This system has three mobile-phase reservoirs, solenoid proportioning valves, a dual-piston pump, a pulse dampener, a column and detector oven, dual thin-layer electrodes with Ag/AgCl reference electrode, and a Rheodyne injector for manual injection that was modified for reductive work (Li et al. 1998a). The system is also equipped for mobile-phase heating and sparging. Stainless steel connectors and tubing were used throughout the system. For simultaneous determination of AS and its hydrolytic metabolite DHA (artemisinin as an internal standard), the Waters μBondapack, CN column (4.6 mm × 30 cm), and a mobile phase consisting of 30% acetonitrile:70% 0.1 M acetic acid/NaOH buffer (pH 4) was used. Compounds were detected via reductive electrochemical detection, as described previously (Li et al. 1998a). Data were acquired and analyzed using a Waters model 820 chromatography data system, Maxima program (Waters Associates, Milford, MA). There was good reproducibility and a lower quantitation limit of 2.5 ng ml⁻¹ for AS and 4 ng ml⁻¹ for DHA. The inter-and intraday coefficient of variation for accuracy and precision was within ±10%.

Detector response linearity studies were performed by preparing six duplicate calibrations covering the range of 2.5 to 1000 ng ml⁻¹for AS and DHA. Linear regression lines were obtained by plotting the peak area ratios versus target peak areas divided by internal standard peak areas. In order to evaluate repeatability (within-day precision) and reproducibility (between-day precision) of the method, replicate analyses (n = 6) of plasma spiked with AS were carried out. The absolute recoveries of the AS were determined by comparing the peak areas of spiked plasma samples and reference samples.

The reference samples were injected directly into the HPLC for both detections. The limit of detection for AS and DHA was determined as the lowest concentration giving a response with 1.0 to 2.5 ng ml⁻¹ plasma in spiked serum sample for AS and DHA (signal-to-noise ratio of 3). The limit of quantitative (lowest concentration of the calibration curve) was 1.8 and 4.2 ng ml⁻¹ in spiked plasma for AS and AL, respectively.

Data Analysis

For TK, the concentration-time data of AL, AS, and DHA, collected during first day and last day, were fitted to a two-compartment open model using a nonlinear, extended least-square fitting procedure (WinNonlin 4.0; Scientific Consulting, Apex, NC). The AUC was determined by the linear trapezoidal rule with extrapolation to infinity based on the concentration of the last time point divided by the terminal rate constant. Extrapolations to time 0 were done using zero concentration for intragastric dosing and using C ₀ values determined from the two-compartment model equation at time 0 by i.v. route. Mean clearance rate (CL) was determined by dividing the dose by the AUC_infinity for i.v. injection. Mean residence time (MRT) was determined by dividing the area under the first moment curve (AUMC) by AUC. The volume of distribution at steady state (Vss) was calculated as the product of CL and MRT. The ratio of DHA to AL or AS was calculated by AUC_DHA/AUC_ALorAS.

RESULTS

Severe malaria-rat model (parasitemia of 5% to 63%) was established by infection of 143 rats with an achieved infection rate of 97.9% and a 4.9% delayed infection rate, which was similar to our previous efficacy study. (Li et al. 2003). In P. berghei–infected animals (166 to 224 g), the mean parasitemia slowly increased during the first 3 to 4 days from 0% to 6%, and then plateaued at 5% to 9% for the next 3 days. After day 6, parasitemia increased rapidly to reach a mean peak of 43% ± 12% (CV 36.5%) on day 11. The model was given a long therapeutic time window within 5 to 7 days (from day 5 to day 11 post inoculation). Then, parasitemia declined (self cleared) to undetectable levels over the next 7 to 11 days. The treatments were carried out on days 6, 7, and 8. Animals were under clinical observation until the final pharmacokinetic blood sample was taken, after which they were sacrificed.

TK and Biotransformation of AL in Infected Rats

AL/lysine was administered intravenously as multiple doses in the malaria-infected rats. The individual and computer fitted plasma concentration-time curves (dashed line) following multiple i.v. injections of AL are shown in Figure 1 (top). AL results show that a biphasic pattern of disposition and two-compartment open TK model fit the data on day 1, and the same biphasic pattern of disposition and two-compartment open TK model fit the result obtained on the last dosing day. The TK parameter estimates on day 1 and day 3 after multiple i.v. doses of 40.6 mg/kg AL are summarized in Table 1. The results revealed that the 2-day (day 1 and day 3) TK parameters were very similar, and no significant accumulation and decline were found for the plasma concentration during the 3-day dosing. The mean Cmax of AL on day 1 was 48453 ng ml⁻¹ and 44101 ng ml⁻¹ on day 3. The mean AUC_{0–24 h} on the last dosing day (24650 ng·h ml⁻¹) was same as the AUC on day 1 (25857 ng · h ml⁻¹). The elimination half-lives of AL were 6.1 to 7.1 h and the total clearance rates were 25.9–29.4 ml/min/kg.

The second peaks of the plasma concentration (solid line) of AL were observed during the 3-day i.v. treatment (Figure 1, top ), which were leveled at 317 and 369 ng ml⁻¹ on day 1 and day 3, respectively, and at 5 h after the i.v. injections.

The conversion of AL to DHA is presented in Table 1 after multiple i.v. administrations in the rats. DHA, an active metabolite of AL, had peak plasma drug concentrations of 124 and 117 ng ml⁻¹ on days 1 and 3, respectively. The AUC on the last dosing day (204 ng·h ml⁻¹) was slightly greater than the AUC on the first dosing day (160 ng·h ml⁻¹). The ratio of total AUC_DHA (524 ng·h ml⁻¹) to AUC_AL (84406 ng·h ml⁻¹) was 0.006 during the 3-day treatment.

TK and Biotransformation of AS in Infected Rats

The individual and computer-fitted plasma concentration-time curves following multiple i.v. injections of AS/NaHCO₃ and its active metabolite DHA are shown in Figure 1 (bottom). AS results show that a biphasic pattern of disposition and two-compartment open TK model fit the data obtained on day 1, and the same biphasic pattern of disposition and two-compartment open TK model also fit the result of the last dosing day. However, the TK parameter estimates on day 1 were different that on day 3 after multiple i.v. doses of 36.7 mg/kg AS. The results revealed that a significant decline was found for the drug plasma concentration during the last i.v. dosing day. The mean AUC_{0–24 h} on the last dosing day (491 ng·h ml⁻¹) was one-third of the AUC on first day (1324 ng·h ml⁻¹). Similar decline was found for its active metabolite, DHA, with that AUC on the last dosing day being 3327 ng·h ml⁻¹ compared to the AUC of 4893 ng·h ml⁻¹ on the first day. The elimination half-lives were 0.36 to 0.46 h for AS and 0.65 to 0.72 h for DHA. The mean total clearance on the day 3 was 2172 ml/min/kg, which was fourfold higher than that on the day 1 with 474 ml/min/kg.

The conversion of AS to DHA is presented in Table 1. DHA had peak plasma drug concentrations of 10842 and 6485 ng ml⁻¹ on days 1 and 3, respectively. The ratio of total AUC_DHA (13051 ng·h ml⁻¹) to AUC_AS (2717 ng·h ml⁻¹) was 5.26 during the 3 days’ treatment.

Tissue Distribution of [¹⁴C]AL in Uninfected Rats

Autoradiographs indicated that at 1 h following single administration of [¹⁴C]AL, the relative autoradiographic density per unit area of tissue was mostly amassed in the heart and the gastrointestinal (GI) system (liver, intestines, and colon) and very light densities were distributed in other tissues (Figure 2, top, left ). Six hours after the i.v. injection, high density of radioactivity was still in the GI system with slow loss, but more drug distribution was seen in the brain, muscles, kidneys, and other tissues (Figure 2, top, right ). By 48 h, high density of the radioactivity was concentrated in GI system, whereas minimal drug distribution was still seen in the brain, heart, and other tissues (Figure 2, bottom, left ). At 96 h after injection, high density of the radioactivity was still concentrated in the GI system, and a greater rate of loss was indicated from heart and other tissues. By 192 h, the residual activity was still detected in the peripheries of the kidneys and the spleen, and trace density was found in other tissues.

The levels of radioactivity in the blood samples as a function of time after administration of the [¹⁴C]AL are shown in Figure 3. Total radioactivity in the blood lasted for a very long period (192 h) after the single i.v. injection. The concentration in whole blood was always higher (two- to fivefolds) than in plasma throughout the period of the treatment. Unchanged AL was eliminated very fast (within 1 to 2 h), indicating that the long lasting radioactivity should be the metabolites of AL.

DISCUSSION

The toxicokinetic profiles have been compared between AL/lysine and AS/NaHCO₃ injections in rats infected with P. berghei ANKA following daily i.v. administration for 3 days. At the same molecular dose level of 96 μmoles/kg, AL (40.6 mg/kg) yielded 100% parasitemia clearance in all malaria-infected animals, whereas AS (36.7 mg/kg) resulted in a clearance rate of only 39% to 67% (Li et al. 2003). In the present toxicokinetic study this dose rate (96 μmoles/kg), which induced significantly hematological changes with reversible anemia and reticulocytopenia (Q. Li et al., submitted) was selected for both drugs with the different i.v. formulations.

The plasma concentration of AL was very similar on day 1 when compared to day 3 following daily dose of 40.6 mg/kg for 3 days. However, AS exhibited a declining level in plasma during the i.v. treatments at dose of 36.7 mg/kg in malaria-infected rats. AS concentration was one-third lower on day 3 than that on day 1, resembling its active metabolite DHA, suggesting that an autoinduction of hepatic drug-metabolizing enzymes occurred for AS (Ashton et al. 1996; Khanh et al. 1999) but not for AL in the infected rats. Hyperparasitemia (>5%), hemolytic anemia, and renal failure are symptoms of P. berghei ANKA–infected rats in the several malaria model. Humans experience very similar symptoms when infected with P. falciparum (Li et al. 2003). Therefore, the P. berghei ANKA–infected rat model is sufficient to screen human drug used in efficacy and relevant PK analysis. The toxicity and tissue distribution study in uninfected rats requires that the toxicity should be related to drug treatment, not malaria.

Declining drug concentrations have also been reported in humans. Four artemisinin drugs (artemisinin, artemether, AS, and DHA) have revealed declining concentrations in plasma during multiple oral treatments in malaria patients and healthy subjects. The Cmax and AUC values for AS were markedly reduced to about one-third to one-seventh on last dose day as compared to the first day. The decrease in plasma concentration-time during multiple treatments is indicative of an increase in metabolic capacity due to autoinduction of hepatic drug-metabolizing enzymes (Ashton et al. 1996; Khanh et al. 1999) in patients and in healthy subjects (van Agtmael et al. 1999; Ashton, Hai, and Sy 1998; Park et al. 1998).

The present study also demonstrated that AS is the prodrug of DHA in malaria-infected rats. The ratio of AUC_DHA/AUC_AS was 5.26 during the entire 3-day treatment, which was very similar to the data in humans where the ratio of AUC_DHA to AUC_AS was 4.3 to 9.7 in malaria patients (Batty et al. 1998a, 1998b). Unlike AS, AL seems to be stable in malaria-infected rats, with an AUC_DHA/AUC_AL ratio of 0.01. Similar results have been published previously showing that a low conversion rate (1% to 2%) of AL was found in in vitro studies with human liver cytochrome P4503A and in uninfected rats (Grace, Skanchy, Aguilar 1999; Li et al. 1998b). The data indicated that the two water-soluble compounds of artemisinin have totally different metabolic profiles.

TK data showed a higher plasma concentration for AL along with longer elimination half-lives that were over five times higher than that of AS plus DHA (Table 1). Similar results with high plasma concentrations and longer half-lives of AL were also reported in uninfected rats with a single administration at the same dose level (Li et al. 1998a). One possible explanation for the increase in plasma concentration and half-life of AL during treatment could be due to the lower conversion (poor metabolism rate) to DHA in the malaria-infected rats as compared to AS. Another possible explanation may be an enterohepatic circulation of AL that was first observed in the present study.

The secondary drug peak in plasma indicated the enterohepatic cycling of AL. A blood AL concentration versus time curve displaying two peaks was present in all subjects. The drug plasma concentration profile revealed that AL formed a secondary peak at 5 h after each i.v. injection (Figure 1, top ), suggesting that a secondary increase in drug concentration due to enterohepatic circulation or a second absorption phase was present (Lindholm, Henricsson, and Dahlqvist 1990; Matsuzawa and Nakase 1984; Vermorken et al. 1984; Westerling, Frigren, and Hoglund 1993). Usually, the secondary peaks in oral concentration-time profiles following drug administration result from discontinuous absorption along the GI tract, postabsorptive storage and release, and/or enterohepatic circulation (Suttle and Brouwer 1994). Thus, double peaks after the i.v. injection of AL in the present study is strongly suggestive of the presence of an enterohepatic circulation (Ogiso et al. 1997, 2001).

The enterohepatic circulation of AL was also demonstrated by our tissue distribution data. One hour after a single i.v. injection, total radioactivity of AL was immediately concentrated in the heart and the gastrointestinal system (liver and intestines), and very light density of the [¹⁴C]AL was noted in other tissues. At 6 h, high density of the radioactivity was still in the GI system with slow loss, but compared to 1 h radiophotographs, an increase in the density was found in the brain, muscles, kidneys, and other tissues. The results indicated that 1 hour after dosing, most of the [¹⁴C]AL was extracted into the liver from where it was excreted into the intestines via bile ductules. The increase of the radioactivity of AL in other tissues seen after 6 h should be attributed to the reabsorption of the drug from the intestines (Figure 2). The disposition of AL was mostly by biliary excretion (Li et al. 1998b) and intestinal reabsorption. [¹⁴C]AL’s (metabolites) lengthy presence (192 h) in the rat may be due to the continual enterohepatic cycling in spite of the short half-life of unchanged AL (Figure 3).

Enterohepatic circulation occurs by biliary excretion and intestinal reabsorption of a solute, sometime with hepatic conjugation and intestinal deconjugation (Roberts et al. 2002). Cycling is often associated with multiple peaks and a longer apparent half-life in a plasma concentration-time profile. Pharmacokinetics and bioavailability of drug are also affected by the extent of intestinal absorption, gut-wall P-glycoprotein efflux, and gut-wall metabolism. Depending on the efficiency of enterohepatic circulation, small variability in drug pharmacokinetic properties can cause high variance of drug bioavailability (Horkovics-Kovats 1999). The much higher (two- to sevenfold) plasma concentration and half-life of AL, as compared to other four artemisinin drugs that were reported previously (Li et al. 1998a) and in the present study, seems to be due to the enterohepatic circulation after i.v. administration.

Animals treated with 40.6 mg/kg of AL daily for 3 days exhibited acute renal failure with haemoglobinuria, decreased urinary excretion, and renal damages (Q. Li et al., submitted). Following second AL dosing, the urinary output decreased 73% from 19.9 ml to 6.6 ml in the infected rats. After the third dosing, urinary output remained 60% below normal. Similar inhibition (about 50%) was also observed in uninfected rats. Histopathological data confirmed renal lesions in both infected and uninfected rats receiving 40.6 mg/kg daily for 3 days. In male rats, 5% to 20% of tubules were affected with signs of nephrotoxicity, including single cell necrosis and degeneration, cytoplasmic vacuolation, and nuclear hypertrophy. In the female rats, a moderate renal failure was confirmed with >40% of tubules affected with multicellular necrosis and degeneration, intratubular cellular debris or casts, and tubular attenuation (Xie et al. 2003).

Did the renal failure affect toxicokinetic (TK) parameters in present study? When compared to the TK data at 10 mg/kg (previously reported; Li et al. 1998a), the AUC of AL in the present study (25857 ng·h ml⁻¹at 40.6 mg/kg) on the first day was still 18% higher than the earlier study (5481 ng·h ml⁻¹ at 10 mg/kg) after an adjustment of the dose level. The 18% higher plasma concentrations in the present study may be the result decreased urinary excretion or drug elimination at the higher dose.

In conclusion, an immediate and complete hydrolysis was occurred for AS, but not for AL, in P. berghei–infected rats. Similar to human clinical data, artesunate appears to be the prodrug of DHA in this model. The much higher plasma concentrations and longer half-lives of AL observed in the present study, as compared to AS, may be attributed to three factors: the poor conversion rate of AL to DHA, the drug’s enterohepatic circulation, and the nephrotoxic effects of AL at 40.6 mg/kg dose level. The high drug exposure level and the long exposure time of AL in the malaria-infected rats may bring about its higher antimalarial potential and more severe toxicity as compared to AS.