Pharmacokinetics of Saquinavir After Intravenous and Oral Dosing of

Dec 12, 2007 - Male Wistar−Hannover rats (weight 300–350 g) were obtained from RCC Ltd. (Laboratory Animal Services, CH-4414 Füllinsdorf, Switzerland)...
0 downloads 0 Views 605KB Size
Download PDF
Biomacromolecules 2008, 9, 305–313

305

Pharmacokinetics of Saquinavir After Intravenous and Oral Dosing of Saquinavir: Hydroxybutenyl-β-Cyclodextrin Formulations Charles M. Buchanan,*,† Norma L. Buchanan,† Kevin J. Edgar,†,‡ James L. Little,† Michael G. Ramsey,† Karen M. Ruble,† Vincent J. Wacher,†,§ and Michael F. Wempe†,| Eastman Chemical Company, Research Laboratories, P.O. Box 1972, Kingsport, Tennessee 37662, and Department of Pharmacology, East Tennessee State University, Johnson City, Tennessee 37614 Received July 27, 2007; Revised Manuscript Received October 5, 2007

The current research evaluated the ability of hydroxybutenyl-β-cyclodextrin (HBenBCD) to enhance saquinavir in Vitro solubility and in ViVo oral bioavailability; both the base and mesylate salt forms of saquinavir were investigated. HBenBCD was effective and significantly improved saquinavir solubility in aqueous media. In the presence of 10 wt % HBenBCD, saquinavir base solubility in water was increased to ca. 5.5 ( 0.4 mg/mL and represents a 27-fold increase from that observed in water (207 ( 5 µg/mL) in the absence of HBenBCD. Saquinavir-HBenBCD formulations were found to have rapid dissolution over a wide pH range (1.2–6.8), and saquinavir solubility in these media was maintained throughout the experiments. When saquinavir-HBenBCD formulations were administered to Wistar-Hannover rats, saquinavir was rapidly absorbed and rapidly eliminated. Rapid saquinavir elimination was particularly pronounced when saquinavir-HBenBCD formulations were given as an oral aqueous gavage. Saquinavir oral bioavailability in rats obtained from saquinavir mesylate capsules (2.0% ( 0.7%) was increased (9 ( 4)-fold (18.6% ( 7.3%) when dosed with saquinavir base-HBenBCD capsules. Clearly, HBenBCD can significantly improve the solubility and oral bioavailability of saquinavir; however, further formulation studies are required to optimize saquinavir oral delivery using this technology.

Introduction Human immunodeficiency virus (HIV) infection, which leads to acquired immunodeficiency syndrome (AIDS), remains a serious worldwide health problem. An estimated 40 million people are currently living with HIV/AIDS and approximately 25 million people have died from the disease since 1981.1 The discovery of HIV protease inhibitors introduced new and effective first-line therapies for HIV/AIDS. Helping to combat HIV-related diseases and prolong survival, protease inhibitors are commonly administered with reverse transcriptase inhibitors. However, poor patient compliance, noxious side effects, and viral resistance have led to a recommendation to treat with two different kinds of protease inhibitors. The most important HIV protease inhibitors in clinical use are saquinavir, nelfinavir, indinavir, lopinavir, ritonavir, atazanavir, and amprenavir. These protease inhibitors are metabolized by cytochrome P450 3A (CYP3A) enzymes, are efflux transporter substrates (i.e., P-glycoprotein, P-gp), or both.2,3 These metabolism and transport mechanisms often result in widely variable drug absorption. In addition to the metabolism and transport issues, many protease inhibitors have poor aqueous solubility, which produces very low and variable bioavailability. As a result, HIV/AIDS patients require frequent and large medication dosing and commonly are unable to adhere to their treatment regimes. * Corresponding author. Phone 423-229-8562; fax 423-229-4558; e-mail [email protected]. † Eastman Chemical Company, Research Laboratories. ‡ Present address: Virginia Tech, 230 Cheatham Hall, Blacksburg, VA 24061. § Present address: 1042 N. El Camino Real, Suite B-174, Encinitas, CA 92024-1322. | East Tennessee State University.

An archetypal protease inhibitor, saquinavir has poor water solubility and is reported to be an excellent P-gp and CYP3A substrate.4,5 As a result, the oral bioavailability has been reported to be very low (0.7-4.0%) and dependent upon the dosage form used. Saquinavir has been offered in two dosage forms, Invirase (Roche) and Fortovase (Roche).6 Available as hard gelatin capsules, Invirase contains saquinavir mesylate (200-mg strength as saquinavir free base). Fortovase has been supplied as soft gelatin capsules containing 200 mg of saquinavir free base. Invirase and Fortovase are not bioequivalent and should not be used interchangeably. In fact, Fortovase was discontinued in the United States in February 2006. Typically, Invirase is dosed 2-times daily as five 200-mg capsules in combination with ritonavir (100 mg bid.). It is recommended that Invirase be taken with meals. Ritonavir, a CYP3A and P-gp inhibitor, helps to increase saquinavir oral bioavailability. However, because HIV/ AIDS patients must take other drugs known to be metabolized by CYP3A or they are P-gp substrates, ritonavir has been shown to cause additional toxicity and safety issues. Therefore, novel pharmaceutical formulations that may safely enhance the bioavailability of protease inhibitors are needed. Cyclodextrins (CDs) are glucose cyclic oligomers and typically contain six, seven, or eight glucose monomers joined by R-1,4 linkages; these oligomers are commonly called R-CD, β-CD, and γ-CD, respectively. Topologically, CDs form a torus with a hydrophobic interior and a hydrophilic exterior. This allows the CD to be dissolved in water, where it acts as a host molecule and forms inclusion complexes with hydrophobic guest molecules. This feature has led to CD usage in pharmaceutical formulations.7,8 Unmodified CDs, particularly β-CD, are relatively crystalline and have limited aqueous solubility. In parenteral formulations, limited solubility can be a very serious issue because renal concentration of the unmodified CD can

10.1021/bm700827h CCC: $40.75  2008 American Chemical Society Published on Web 12/12/2007

306 Biomacromolecules, Vol. 9, No. 1, 2008

Buchanan et al.

Experimental Procedures

Figure 1. Structures of (a) HBenBCD and (b) saquinavir, which were used in this study.

lead to crystallization, leading to necrotic damage.9 Fortunately, the CD solubility in water can be significantly increased by the addition of a small number of substituents to the hydroxyl groups of the anhydroglucose monomers.10 Different CDs have previously been evaluated for their ability to improve the solubility and bioavailability of protease inhibitors. Boudad et al. have described saquinavir-hydroxypropylβ-cyclodextrin (HPBCD, molar substitution ) 3) preparations and their incorporation into poly(alkylcyanoacrylate) nanoparticles.11 The saquinavir utilized in this study was isolated from Invirase capsules by chemical extraction. At 10 wt % HPBCD, Baudad et al. reports 15.8 and 9.3 mg/mL of saquinavir were solubilized in water and pH 2.0 buffers, respectively. Later work by Boudad et al. experimented with Caco-2 cells and suggested that the cyclodextrin contained in the nanoparticles could reduce cell cytotoxicity by complexing polymer degradation products resulting from nanoparticle degradation.12 Anderson et al. have examined the solubilization of Kynostatin, a tripeptide HIV protease inhibitor, using HPBCD and sulfobutyl-β-cyclodextrin (SBEBCD).13 They found that both CDs were able to increase Kynostatin solubility in aqueous media. Optimal results were obtained with high concentrations of SBEBCD (50 wt %) at low pH (pH ) 3.0). Recently we have described the preparation and characterization of hydroxybutenyl-β-cyclodextrin (HBenBCD, Figure 1).14 HBenBCD, an amorphous white solid (Tg ca. 210 °C), has high water and organic solvent solubility (water, >500 g/L; poly(ethylene glycol) 400, > 400 g/L). In this paper, we describe our saquinavir base and mesylate solubility investigations using HBenBCD as a complexing agent. We also describe the preparation of solid inclusion complexes (saquinavir base and mesylate with HBenBCD), as well as liquid formulations (saquinavir base and mesylate) dissolved in PEG400 with and without HBenBCD. We compare the dissolution behavior of these formulations to that of the parent drug. Additionally, we report our preclinical results after intravenous (iv) and oral (po) administration of saquinavir-HBenBCD formulations to male Wistar-Hannover rats. Our goals were to establish in Vitro solubility and dissolution profiles and to probe formulation effects on absorption, metabolism, elimination, and oral bioavailability of saquinavir in this animal model.

General Materials and Methods. HBenBCD (molar substitution ) 4.7) was prepared according to previously described methods.14 The HBenBCD was dried at 10–15 mmHg at room temperature for 14-60 h prior to use. Saquinavir base and mesylate were obtained from Apin Chemicals (Oxon, OX14 4RU, UK) and characterized prior to use. Ethylenediaminetetraacetic acid trisodium salt hydrate (EDTA), nonessential amino acids, streptomycin and penicillin, N-2-hydroxyethylpiperazine-N′-2-ethane sulfonic acid (HEPES), Dulbecco’s modified Eagle medium (DMEM), HPLC grade water, HPLC grade methanol, HPLC grade acetonitrile, ethanol, isopropyl alcohol, formic acid, ammonium acetate, and nicotinamide adenine dinucleotide phosphate (NADPH) were purchased from Sigma-Aldrich (St. Louis, MO). Heat-inactivated fetal calf serum (FBS) and Hank’s balanced salt solution (HBSS) were purchased from GIBCO (Invitrogen Corp.; Carlsbad, CA). Microsomes (rat, male Wistar; Lot TCF) were purchased from In Vitro Technologies (Baltimore, MD). Water used to conduct solubility testing and prepare formulations was filtered through a Milli-Q water system equipped with a 0.22 µm sterile biofilter (www.millipore.com/catalogue.nsf/docs/C7658) and had very low total organic and pyrogen content and low ionic strength. The water was not degassed prior to use and typically had a pH of 4.7. Buffers were prepared using Millipore water. All glassware and tools used in complex preparation were washed extensively with water and then with absolute EtOH and dried at 115 °C for 1–24 h. Solubility Studies. Solubility measurements were made using a modified version of the traditional shaking flask method.15,16 To each well of a 2 mL 96-well polypropylene mixing plate was added excess drug (ca. 5–10 mg). To each well of the preloaded 96-well mixing plate was added water, buffer, or the appropriate HBenBCD solution (300–500 µL). Each determination was made in triplicate. The blanks were used to determine the intrinsic drug solubility (So) in that solution and the wells containing the HBenBCD solutions were used to determine drug solubility due to HBenBCD (St). After addition of the stock solutions, the plate was sealed using aluminum foil with a nonvolatile adhesive on one surface. The plate was placed on a rotary shaking plate (Helidolph Titramax 1000) and was shaken at 800–1200 rpm at 23 ( 2 or 37 ( 2 °C for 48–72 h. Preliminary experiments showed that less than 24 h was required to reach equilibrium (i.e., drug concentration plateaued and did not change with time). Longer mixing times were used, both for convenience and to ensure that we were well above the time required to reach equilibrium. During the mixing period, the plate was inspected to ensure that each well contained undissolved, excess drug. Drug was added as necessary to maintain an excess. Following the mixing period, the solutions in each well were transferred to the corresponding wells of a 96-well 2 mL multiscreen filter plate using a multichannel pipet. The bottom of each well was a hydrophilic membrane. The filter plate was placed on top of a vacuum manifold, and the solutions were filtered at ca. 20 mmHg into the corresponding wells of a 2 mL storage plate. The duration of the filtration period was typically no longer than 60 s. The storage plate was then sealed with a silicon mat, and samples were removed for analysis as appropriate. The drug content in each well was determined by UV spectroscopy using a SpectraMax Plus 384 Molecular Devices multiwell plate reader. Typically, drug solution (10–20 µL) was transferred to the corresponding well of a 96-well measurement plate (UV-STAR plates from Greiner with a spectral range of 190–400 nm) and, if necessary, diluted with 1/1 water/ethanol so that the absorbance was in the linear response range. The UV measurements for saquinavir were made at 300 nm, well separated from the absorbance of HBenBCD (350 Ω · cm2 with background subtracted were used for transport studies. Saquinavir transport assays were performed in absorptive (apical to basolateral, Ap-to-Bl) and secretory (Bl-to-Ap) directions. To initiate the experiments, a formulation solution (10 or 50 µM, saquinavir base equivalent) in buffer solution, HBSS containing 25 mM HEPES (pH 6.5 for Ap-to-Bl; pH 7.4 for Bl-to-Ap; 37 °C), was added to the donor compartment, and buffer solution (pH 6.5 for apical side; pH 7.4 for basolateral side) was added to the receiver compartment. The Ap-toBl samples were collected from the receiver compartment by complete replacement of the receiver volume with fresh buffer solution (37 °C) at 30, 60, 90, 120, 150, 180, and 240 min. The Bl-to-Ap samples were collected as mentioned but only at 60 and 120 min. Compounds were quantified by LC/MS/MS. Flux was determined using receiver compartment compound steady-state appearance rates (∆Q/∆t). Papp across Caco-2 monolayers was calculated via Papp ) ∆Q/(∆tAC0) where Papp ) apparent permeability coefficient [cm/s], ∆Q/∆t ) permeability rate [µg/s; pmol/s], C0 ) initial concentration in donor chamber [µg/cm3; pmol/cm3], and A ) membrane surface area [cm2]. Efflux ratio (ER) was computed as ER ) PBlfAp /PApfBl . app app For the Caco-2 samples, an Applied Biosystems Sciex 4000-QTrap (Applied Biosystems; Foster City, CA) equipped with a Shimadzu HPLC (Shimadzu Scientific Instruments, Inc.; Columbia, MD), a PEAK Scientific API Systems gas generator (PEAK Scientific; Bedford, MA), and Leap autosampler (LEAP Technologies; Carrboro, NC) were used for liquid chromatography–mass spectrometry. An Agilent Technologies, Zorbax extended-C18 50 × 4.6 mm2, 5 µm column was used at 40 °C with a flow-rate of 0.4 mL/min. The mobile phase consisted of buffer A, 10 mM ammonium acetate, 0.1% formic acid in water, and buffer B, 50:50 acetonitrile/methanol. HPLC method had a 6.5 min

Pharmacokinetics after Saquinavir-HBenBCD Dosing

Figure 2. Saquinavir base equilibrium solubility in water and buffers with and without 10 wt % HBenBCD. N ) 3; error bars are one standard deviation.

run time, started at 5% B, ramped up to 95% B at 3 min and held until 4.5 min, ramped back down to 5% B at 5 min and held until end of run. The MS/MS conditions were electrospray ionization positive ion mode (ESI+) using mass transition 671.3-570.3 m/z with the following settings: (i) an ion-spray voltage of 5500 V; (ii) temperature 450 °C; (iii) nitrogen was used for the curtain gas (CUR) and for the collisionally activated dissociation (CAD) gas; (iv) the CAD gas was set at medium; (v) ion source gas one (GS1) and two (GS2) were air and set on 10; (vi) the entrance potential was set at 10.0 V; (vii) quadruple one (Q1) and three (Q3) were both set on unit resolution; (viii) dwell time was set at 200 ms; (ix) declustering potential (DP; 61.0), collision energy (CE; 47.0), and collision cell exit potential (CXP; 14.0) are voltages (V). A 1/x2 weighted (x ) analyte concentration) linear regression standard curve (N ) 3 ( SD) was used; LOD ) limit of detection was 0.07 ng/mL, and R ) correlation coefficient of 0.9942 was computed using the Pearson correlation with a two tailed p value test (p < 0.0001) at a 95% confidence interval.

Results and Discussion Solubility Studies. Saquinavir base solubility in different media in the presence and absence of HBenBCD has been summarized in Figure 2. Saquinavir base has poor water solubility (So ) 207 ( 5 µg/mL) which was significantly increased by 10 wt % HBenBCD (St/So ) 26). Saquinavir base solubility at pH 1.4 (So ) 87 ( 6 µg/mL) was also poor but dramatically improved by complexation with HBenBCD (St/So ) 37). Saquinavir base solubility in buffers (pH 3.0), particularly the tartrate buffers, increased significantly in the absence (So ) 6.5–7.0 mg/mL) and in the presence of HBenBCD (St ) 29.7–31.3 mg/mL). Saquinavir base solubility at pH 4.8 or higher was diminished relative to pH 3.0 buffer. However, the HBenBCD did help to maintain significant drug solubility in these higher pH media (St ) 6–12 mg/mL) relative to water or pH 1.4 buffer. As expected, the intrinsic solubility of saquinavir mesylate (2.1 ( 0.3 mg/mL) was higher than that of saquinavir base (Figure 3). However, unlike saquinavir base (Figure 2), 10 wt % HBenBCD increased saquinavir mesylate solubility in all media over a broad pH range. In the presence of HBenBCD, saquinavir mesylate solubility was essentially the same (11–12 mg/mL) in all media except for pH 1.4 buffer where the drug solubility dropped to 3.8 ( 0.3 mg/mL (St/So ) 14) and with

Biomacromolecules, Vol. 9, No. 1, 2008 309

Figure 3. Saquinavir mesylate equilibrium solubility in water and buffers with and without 10 wt % HBenBCD. N ) 3; error bars are one standard deviation.

Figure 4. Saquinavir base equilibrium solubility in water and pH 3.0 buffers in the presence of 0-40 wt % HBenBCD: (0) ) water; (4) ) phosphate, pH 3.0; (O) ) citrate, pH 3.0; (]) ) L-tartrate, pH 3.0. N ) 3; error bars are one standard deviation.

pH 7.3 phosphate buffer where drug solubility dropped to 9.3 ( 0.4 mg/mL (St/So ) 42). Figures 4 and 5 present saquinavir base and saquinavir mesylate equilibrium solubilities in water and pH 3.0 buffers in the presence of variable amounts (0–40 wt %) of HBenBCD. Saquinavir base solubility in each medium increased with increasing HBenBCD concentration. At 40 wt % HBenBCD, saquinavir solubilization in the different media decreased in the order tartrate > citrate > phosphate > water and was consistent with the observations presented in Figure 2. In the tartrate and citrate buffers, some negative curvature in the solubility curves were evident at higher HBenBCD concentrations, a behavior known to be indicative of a soluble inclusion complex (AN solubility).7 In the case of saquinavir mesylate, increasing the concentration of HBenBCD resulted in an increase in drug solubility in all media; only a slight negative curvature in the solubility curves was observed. When the data presented in Figures 4 and 5 are converted to molar concentrations (Supporting Information), it is evident that multiple HBenBCD molecules are interacting with saquinavir. Moreover, the HBenBCD/saquinavir molar ratio is dependent upon the form of saquinavir (base versus salt) and, in the case

310 Biomacromolecules, Vol. 9, No. 1, 2008

Figure 5. Saquinavir mesylate equilibrium solubility in water and pH 3.0 buffers in the presence of 0-40 wt % HBenBCD: (0) ) water; (4) ) phosphate, pH 3.0; (O) ) citrate, pH 3.0; (]) ) L-tartrate, pH 3.0. N ) 3; error bars are one standard deviation.

Figure 6. Dissolution of saquinavir base at 37 °C from gelatin capsules filled with solid saquinavir base-MCC or saquinavir base-HBenBCD powder: (O) ) saquinavir base, pH 1.2; (b) ) saquinavir base, pH 6.8; (0) ) saquinavir base-HBenBCD, pH 1.2; (4) ) saquinavir base-HBenBCD, pH 4.5; (]) ) saquinavir base-HBenBCD, pH 6.8. N ) 3; error bars are one standard deviation.

of saquinavir base, upon the specific formulation. In the case of saquinavir mesylate, if only the initial linear portion of the solubility curve is considered, the HBenBCD/saquinavir molar ratio varies from ca. 2.5–4.2. Furthermore, the association or binding constant calculated from the equation KA/B ) slope/ So(1 - slope) is ca. 98 M-1 indicating weak binding.7 In the case of saquinavir base, the HBenBCD/saquinavir molar ratio varies in an analogous manner to that of saquinavir mesylate, and the binding constants are also low (12–86 M-1). Collectively, these observations suggest weak interactions between HBenBCD and saquinavir leading to formulations containing a mixture of complex with excess HBenBCD. Dissolution Studies. The dissolution profiles for saquinavir base-MCC and saquinavir base-HBenBCD powder filled capsules are summarized in Figure 6. In the case of saquinavir base-HBenBCD powder filled capsules, drug dissolution was rapid at each pH examined with approximately 100% of the drug being released into the medium within 30 min at pH 1.2 and pH 4.5; 95% was released within 30 min at pH 6.8. Once dissolved, the drug concentration in the presence of HBenBCD remained stable over the time course of the experiment; the HBenBCD effectively prevented crystallization of the drug

Buchanan et al.

Figure 7. Dissolution of saquinavir mesylate at 37 °C from gelatin capsules filled with solid saquinavir mesylate-MCC or saquinavir mesylate-HBenBCD powder: (O) ) saquinavir mesylate, pH 1.2; (b) ) saquinavir mesylate, pH 4.5; ([) ) saquinavir mesylate, pH 6.8; (0) ) saquinavir mesylate-HBenBCD, pH 1.2; (4) ) saquinavir mesylate-HBenBCD, pH 4.5; (]) ) saquinavir mesylate-HBenBCD, pH 6.8. N ) 3; error bars are one standard deviation.

even at concentrations that would have been supersaturated in the absence of HBenBCD. In contrast, saquinavir base (no HBenBCD) dissolution was slow and incomplete with 15–20% being dissolved after 30 min. After 6 h, the maximum percentages of saquinavir base (no HBenBCD) dissolved were 16.2% and 24.9% at pH 1.2 and 6.8, respectively. The saquinavir mesylate-MCC and saquinavir mesylateHBenBCD dissolution profiles for the powder filled capsules were similar to their saquinavir base counterparts (Figure 7). In the case of saquinavir mesylate-HBenBCD powder filled capsules, dissolution was rapid at each pH examined with ca. 100% of the drug being released into the medium within 30 min at pH 1.2 and pH 4.5; ca. 90% was released within 30 min at pH 6.8. Again, once dissolved, the drug remained dissolved in the presence of HBenBCD over the time course of the experiment. In contrast, saquinavir mesylate (no HBenBCD) dissolution was slow and incomplete with ca. 4–11% being dissolved after 30 min. After 6 h, the maximum percentages of saquinavir mesylate (no HBenBCD) dissolved were 9.4%, 19.2%, and 4.4% at pH 1.2, 4.5, and 6.8, respectively. The dissolution profiles for saquinavir base and saquinavir base-HBenBCD dissolved in PEG400 are summarized in Figure 8. In the instance of saquinavir base-HBenBCD-PEG400 liquid filled capsule, drug dissolution was rapid (30 min) and complete at pH 1.2 and pH 4.5. At pH 6.8, drug dissolution was slower; the drug concentration reached a plateau near 90% dissolution after ca. 2 h. Once dissolved, the drug solution concentration remained constant over the time course of the experiment. In the absence of HBenBCD, saquinavir base in PEG400 also dissolved rapidly in pH 1.2 or 4.5 buffer; ca. 100% dissolution was achieved within ca. 30 min and maintained over the experiment time course. As in the case of the saquinavir base-HBenBCD-PEG400 solutions, drug dissolution at pH 6.8 was slower; ca. 2 h was required to achieve 60% drug dissolution. Pharmacokinetic (PK) Study. Figure 9 presents the plasma concentration versus time profile for saquinavir after saquinavir base-HBenBCD intravenous dosing of rats. Saquinavir elimination was rapid (t1/2 ca. 3.2 h), and the plasma concentration of saquinavir after 24 h was quite low. It should be noted that metabolites were not observed in this or any other dosing groups. Plasma concentration versus time profiles for saquinavir after oral dosing with saquinavir mesylate-MCC and saquinavir

Pharmacokinetics after Saquinavir-HBenBCD Dosing

Figure 8. Dissolution of saquinavir base at 37 °C from gelatin capsules filled with saquinavir base or saquinavir base-HBenBCD dissolved in PEG400: (O) ) saquinavir base, pH 1.2; (b) ) saquinavir base, pH 4.5; ([) ) saquinavir base, pH 6.8; (0) ) saquinavir base-HBenBCD, pH 1.2; (4) ) saquinavir base-HBenBCD, pH 4.5; (]) ) saquinavir base-HBenBCD, pH 6.8. N ) 3; error bars are one standard deviation.

Figure 9. Saquinavir plasma concentration-time curve after intravenous dosing with saquinavir base-HBenBCD aqueous solution (group 1). N ) 4; error bars are standard error of the mean.

Figure 10. Saquinavir plasma concentration-time curves after oral dosing with capsules of saquinavir mesylate-MCC (0 ) group 2) and saquinavir base-HBenBCD (O ) group 3) powder. N ) 4; error bars are standard error of the mean.

base-HBenBCD powder filled capsules are illustrated in Figure 10. Following oral dosing with saquinavir base-HBenBCD powder filled capsules (group 3), initial drug absorption was

Biomacromolecules, Vol. 9, No. 1, 2008 311

Figure 11. Saquinavir plasma concentration-time curves after oral dosing with saquinavir mesylate-HBenBCD (0 ) group 4) and saquinavir base-HBenBCD (O ) group 7) aqueous solutions. N ) 4; error bars are standard error of the mean.

Figure 12. Saquinavir plasma concentration-time curves after oral dosing with capsules filled with PEG400 solutions of saquinavir base-HBenBCD (O ) group 5) and saquinavir base (0 ) group 6). N ) 4; error bars are standard error of the mean.

rapid (Tmax ca. 0.9 h, Cmax ) 268.2 ( 125.4 ng/mL) and was followed by rapid drug elimination. At ca. 4 h, a second saquinavir absorption peak was observed followed again by rapid elimination. In the case of saquinavir mesylate oral dosing (group 2), the observed Cmax (14.4 ( 10.7 ng/mL) and Tmax (ca. 5.3 h) were not truly distinct because the drug plasma concentration appeared to be nearly constant from about 4 to 8 h. Given that the absorption and elimination of saquinavir was apparently rapid (cf. group 3), this observation was consistent with slower drug dissolution, overlapping absorptionelimination-reabsorption, or both. The net effect was that the AUC0–24 h for group 3 (675.8 ( 264.4 ng · h/mL) was more than 9-fold larger than that for group 2 (73.4 ( 25.1 ng · h/mL). Figure 11 shows the plasma concentration versus time profiles for saquinavir after oral aqueous gavage dosing with saquinavir mesylate-HBenBCD and saquinavir base-HBenBCD (groups 4 and 7). In both cases, Tmax was extremely short (ca. 0.3–0.6 h) and the initial elimination was rapid (t1/2 ca. 0.5 h). As was observed with group 3 (saquinavir base-HBenBCD oral capsules), a second absorption peak appeared at ca. 3 h. The AUC0–24h for group 4 (156.5 ( 102.6 ng · h/mL) was not statistically different from that for group 7 (141.7 ( 31.8 ng · h/ mL). Figure 12 compares the saquinavir plasma concentration versus time profiles after oral dosing using capsules filled with

312 Biomacromolecules, Vol. 9, No. 1, 2008

Buchanan et al.

Table 2. Pharmacokinetic Parameters for Saquinavir after Dosing Wistar Rats (n ) 4) with Saquinavir Mesylate or Saquinavir Base Formulationsa group saquinavir saquinavir saquinavir saquinavir saquinavir saquinavir saquinavir

base-HBenBCD (iv) (G1) mesylate capsules (G2) base-HBenBCD capsules (G3) mesylate-HBenBCD oral gavage (G4) base-HBenBCD-PEG400 capsules (G5) base-PEG400 capsules (G6) base-HBenBCD oral gavage (G7)

Cmax (ng/mL)

Tmax (h)

AUC0–24h (ng · h/mL)

%F(0–24h)

464.2 ( 90.4 14.4 ( 10.7 268.2 ( 125.4 103.0 ( 59.5 127.5 ( 144.1 128.2 ( 65.6 70.3 ( 34.2

0.3 5.3 (4.0–8.0) 0.9 (0.7–1.0) 0.3 1.2 (0.7–2.0) 1.9 (0.7–3.0) 0.6 (0.3–1.0)

910.3 ( 211.9 73.4 ( 25.1 675.8 ( 264.4 156.5 ( 102.6 276.1 ( 124.6 599.2 ( 509.9 141.7 ( 31.8

2.0 ( 0.7 18.6 ( 7.3b 4.3 ( 2.8 7.6 ( 3.4 16.5 ( 14.0 3.9 ( 0.9

a Abbreviations: HBenBCD, hydroxybutenyl-β-cyclodextrin; AUC0-24h, total area under the plasma concentration-time curve from 0 to 24 h. Values are means ( standard deviations. b The oral bioavailability for group 3 was significantly different from groups 2, 4, 7 (p < 0.01), and 5 (p < 0.05).

PEG400 solutions of saquinavir base-HBenBCD (group 5) and saquinavir base (group 6). The Tmax (1.2–1.9 h) and Cmax (group 5 ) 127.5 ( 144.1 ng/mL; group 6 ) 128.2 ( 65.6 ng/mL) were similar for both groups. However, a second absorption peak at ca. 3 h for group 6 was far more pronounced relative to group 5. Also, a third absorption peak was observed with group 6 but absent in group 5. The net effect was that the AUC0–24h for group 6 (599.2 ( 509.9 ng · h/mL) was larger than that for group 5 (276.1 ( 124.6 ng · h/mL). Due to the variability in the AUC0–24h for group 6, this difference was not statistically significant. The pharmacokinetic data and oral bioavailability for each oral dosage form has been summarized in Table 2. Saquinavir oral bioavailability (F) after dosing with saquinavir mesylateMCC powder filled capsules was 2.0% ( 0.7% (group 2). The three liquid dosage forms (groups 4, 5, and 7) afforded similar oral bioavailability, with no or only marginal improvements relative to group 2. Also relative to group 2, the Tmax values for groups 4, 5, and 7 were significantly reduced and Cmax was increased and saquinavir elimination rates were much higher. The result was that the AUC0–24h and oral bioavailabilities were quite similar and not significantly different. Oral bioavailabilities for group 3 (saquinavir base-HBenBCD powder filled capsules) and group 6 (capsules filled with saquinavir base PEG400 solution) were 8–9-fold higher than those for control group 2. Groups 3 and 6 were different from the other groups in the study; the Cmax was much larger, and both groups had a significant second absorption peak (120–200 ng/mL) sustained over a longer period of time. Consequently, the AUC0–24h and F for these two groups (3 and 6) were greater. However, due to the AUC0–24h and F variability for group 6, the AUC0–24h and F were not significantly different from group 2. In the case of group 3, F was significantly different relative to groups 2, 4, 7 (p < 0.01), and 5 (p < 0.05). We have previously shown that increasing the solubility of poorly soluble drugs with HBenBCD can significantly increase oral bioavailability in animal models.22–25 The solubility and dissolution studies discussed in this manuscript illustrate that saquinavir solubility and dissolution profiles were improved by formulating with HBenBCD or PEG400. The PK data illustrates that absorption and elimination of solubilized saquinavir was very rapid. Once in systemic circulation, rapid saquinavir elimination was consistent with a very high hepatic extraction followed by rapid metabolism or transport into bile (e.g., hepatic P-gp).2,3 Saquinavir transport into bile provides reabsorption potential and explains the observed second absorption peak (enterohepatic recirculation). As others have shown,26 Caco-2 transport data from our laboratories (Table 3) illustrate that saquinavir (10 µM; ApfBl ) (1.4 ( 0.2) × 10-6 cm/s; BlfAp ) (11.3 ( 0.7) × 10-6 cm/s; n ) 4 ( SD) had a significant efflux ratio (8.0 ( 0.7) consistent with in Vitro efflux via P-gp. Subsequently, we investigated saquinavir (50 µM) in Vitro

Table 3. Absorptive Transport (Ap-to-Bl), Secretory Transport (Bl-to-Ap), and Efflux Ratio (ER) across Caco-2 Monolayers for Different Saquinavir Formulationsa compound saquinavir baseb saquinavir base saquinavir mesylate saquinavir-HBenBCD saquinavir mesylate-HBenBCD

Ap-to-Bl Papp ( SDc

Bl-to-Ap Papp ( SDc

efflux ratio ( SD

1.4 ( 0.2 2.1 ( 0.1 2.0 ( 0.1 2.2 ( 0.1 2.1 ( 0.1

11.3 ( 0.7 29.3 ( 4.1 28.8 ( 9.4 27.0 ( 7.2 24.1 ( 3.9

8.5 ( 1.0 14.1 ( 2.0 14.4 ( 4.1 12.4 ( 3.3 11.4 ( 0.7

a 50 µM; mean ( SD, n ) 6. b Saquinavir at 10 µM and n ) 4. c Papp × 10-6 cm/s.

permeability using saquinavir base, saquinavir-HBenBCD, saquinavir mesylate, and saquinavir mesylate-HBenBCD (Table 3). At 50 µM, the saquinavir and saquinavir mesylate HBSS solutions (37 °C) were essentially saturated. The Caco-2 results did not afford evidence that HBenBCD can alter saquinavir permeability or alter the efflux ratio. The Papp values and efflux ratios were higher at 50 µM and may be attributed to the increased cellular toxicity at the higher dose.27 The HBenBCD formulation values were not statistically different than control; therefore, there was no in Vitro experimental data to suggest that HBenBCD inhibited P-gp or influenced influx pathways. The experimental results clearly indicate that the solubility of saquinavir was greatly enhanced by complexation with HBenBCD. The increase in oral bioavailability from these studies using HBenBCD formulations is likely due to enhanced solubility and faster mean dissolution times of saquinavir relative to the formulations without HBenBCD.28

Conclusions In this study, we have shown that saquinavir solubility in aqueous media can be significantly increased by formulating the base or salt form with HBenBCD. Complexation of saquinavir base with HBenBCD in the presence of pH 3.0 tartrate or citrate buffers provides the greatest drug solubility increase. Dissolution of saquinavir-HBenBCD formulations was found to be very rapid in the pH range of 1.2–6.8, and the solubility of the drug in these media was maintained over the time course of the experiments. When saquinavir-HBenBCD formulations were administered to Wistar-Hannover rats, saquinavir was rapidly absorbed but rapidly eliminated. Elimination of the drug was particularly rapid when the doses were given as oral aqueous gavage. Although we also observed rapid saquinavir absorption and elimination after oral administration of saquinavir base-HBenBCD powder filled capsules, the AUC and oral bioavailability were significantly greater than groups 2, 4, 5, and 7. Saquinavir transport assays in Caco-2 cells showed that there was no in Vitro experimental data to suggest that HBenBCD inhibited P-gp or influenced influx pathways.

Pharmacokinetics after Saquinavir-HBenBCD Dosing

The experimental results from these studies are very promising for the potential to improve protease inhibitor therapy. Saquinavir oral bioavailability was increased in this animal model from 2.0% ( 0.7% when dosing orally with saquinavir mesylate capsules to 18.6% ( 7.3% (a 9-fold increase) using capsules containing solid saquinavir base-HBenBCD. Additional studies utilizing a larger number of animals and additional formulations are warranted to further probe the observations described in this paper. Supporting Information Available. Saquinavir base and mesylate equilibrium solubility (mmol) in water and pH 3.0 buffers in the presence of 0-40 wt % HBenBCD. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) UNAIDS/WHO AIDS Epidemic Update; www.unaids.org/en/HIV_ data/epi2006/default.asp, December 2006. (2) Huisman, M. T.; Smit, J. W.; Wiltshire, H. R.; Hoetelmans, R. M. W.; Beijnen, J. H.; Schinkel, A. H. Mol. Pharm. 2001, 59, 806–813. (3) Gao, W.; Kishida, T.; Kageyama, M.; Kimura, K.; Yoshikawa. Y.; Shibata, N.; Takada, K. AntiViral Chem. Chemother. 2002, 13, 17– 26. (4) Tam-Zaman, N.; Tam, Y. K.; Tawfik, S.; Wiltshire, H. Pharm. Res. 2004, 21, 436–442. (5) Sinko, P. J.; Kunta, J. R.; Usansky, H. H.; Perry, B. A. J. Pharmacol. Exp. Ther. 2004, 310, 359–366. (6) Invirase; Food and Drug Administration, www.fda.gov/medwatch/ SAFETY/2003/03DEC_PI/Invirase_PI.pdf. Fortovase; Food and Drug Administration, www.fda.gov/cder/foi/label/2002/20828s10lbl.pdf. (7) Thompson, D. O. Crit. ReV. Ther. Drug Carrier Syst. 1997, 14, 1– 104. (8) Szejtli, J. Supramol. Chem. 1995, 6, 217–223. (9) Uekama, K.; Hirayama, F.; Irie, T. Chem. ReV. 1998, 98, 2045–2076. (10) Szejtli, J. Pharm. Technol. 1991, 15, 24–38. (11) Boudad, H.; Legrand, P.; Lebas, G.; Cheron, M.; Duchene, D.; Ponchel, G. Int. J. Pharm. 2001, 218, 113–124. (12) Boudad, H.; Legrand, P.; Appel, M.; Coconnier, M.-H.; Ponchel, G. S.T.P. Pharma Sci. 2001, 11, 369–375.

Biomacromolecules, Vol. 9, No. 1, 2008 313 (13) Johnson, M. D.; Hoesterey, B. L.; Anderson, B. D. J. Pharm. Sci. 1994, 83, 1142–1146. (14) Buchanan, C. M.; Alderson, S. R.; Cleven, C. D.; Dixon, D. W.; Ivanyi, R.; Lambert, J. L.; Lowman, D. W.; Offerman, R. J.; Szejtli, J.; Szente, L. Carbohydr. Res. 2002, 327, 493–507. (15) Chen, T. M.; Shen, H.; Zhu, C. Comb. Chem. High Throughput Screening 2002, 5, 575–81. (16) Pan, L.; Ho, Q.; Tsutsui, K.; Takahashi, L. J. Pharm. Sci. 2001, 90, 521–529. (17) USP 28-NF 23 711; The United States Pharmacopoeial Convention, Inc., Webcom Limited, Canada 2412-24142005. (18) Little, J.; Wempe, M.; Buchanan, C. J. Chromatogr. B 2006, 833, 219–230. (19) Yang, L.; Wu, N.; Clement, R. P.; Rudewicz, P. J. J. Chromatogr. B 2004, 799, 271–280. (20) Food and Drug Administration. Guidance for Industry Bioanalytical Method Validation; http://www.fda.gov/cder/guidance/4252fnl.pdf; U.S. Department of Health and Human Service, Food and Drug Administration, Center for Drug Evaluation and Research, Center for Veterinary Medicine, 2001. (21) Rowland, M.; Tozer, T. N. Clinical Pharmacokinetics; Lippincott, Williams, & Wilkins: Philadelphia, PA, 1995. (22) Buchanan, C. M.; Buchanan, N. L.; Edgar, K. J.; Little, J. L.; Malcolm, M.; Ruble, K. M.; Wacher, V. J.; Wempe, M. F. J. Pharm. Sci. 2007, 96, 644–660. (23) Buchanan, C. M.; Buchanan, N. L.; Edgar, K. J.; Klein, S.; Little, J. L.; Ruble, K. M.; Wacher, V. J.; Wempe, M. F. J. Pharm. Sci. 2007, in press, doi:10.1002/jps.20878. (24) Wempe, M. F.; Rice, P. J.; Buchanan, C. M.; Buchanan, N. L.; Edgar, K. J.; Hanley, G. A.; Ramsey, M. G.; Skotty, J. S. J. Pharmacy and Pharmacology 2007, in press, doi.10.1211/jpp.59.6.0006. (25) Wempe, M. F.; Wacher, V. J.; Ruble, K. M.; Ramsey, M. G.; Edgar, K. J.; Buchanan, N. L.; Buchanan, C. M. Int. J. Pharm. 2007, in press, doi:10.1016/j.ijpharm.2007.06.002. (26) Alsenz, J.; Haenel, E. Pharm. Res. 2003, 20, 1961–1969. (27) Storch, C. H.; Theile, D.; Lindenmaier, H.; Haefeli, E.; Weiss, J Biochem. Pharmacol. 2007, 73, 1573–1581. (28) Ritschel, W. A.; Kearns, G. L. Handbook of Basic Pharmacokinetics Including Clinical Applications, 6th ed.American Pharmacists Association: Washington, DC 2004; Chapter 23.

BM700827H