oil beads as a new carrier for improving the oral bioavailability of lipophilic drugs

Journal of Controlled Release 122 (2007) 47 – 53 www.elsevier.com/locate/jconrel

α-Cyclodextrin/oil beads as a new carrier for improving the oral bioavailability of lipophilic drugs Laury Trichard, Elias Fattal, Madeleine Besnard, Amélie Bochot ⁎ Univ Paris-Sud, CNRS UMR 8612, Physico-chimie–Pharmacotechnie–Biopharmacie, Faculté de Pharmacie, 5 rue JB Clément, Châtenay-Malabry, F-92296 Received 20 March 2007; accepted 6 June 2007 Available online 14 June 2007

Abstract The purpose of the present work was to investigate the potential of novel lipid-carrier “beads” consisting of minispheres made of alphacyclodextrin and soybean oil for the encapsulation and the oral delivery of drugs. Isotretinoin was chosen as a model of poorly-stable and lipophilic molecule. Isotretinoin-loaded beads were prepared, characterised and administrated orally in rats. Isotretinoin previously dissolved in soybean oil had no significant effect upon bead preparation and characteristics. Drug encapsulation efficiency was found to be particularly high (93 ± 7%) and no isotretinoin degradation occurred during the preparation process. Freeze-drying advantageously concentrated isotretinoin in beads (3.4 ± 0.2 mg/g) and facilitating ease of handling and use for oral administration. Isotretinoin exhibited good stability for at least 4 months when beads were stored protected from light. Finally, pharmacokinetics of isotretinoin in rats demonstrated that the drug was successfully released from beads in the digestive tract and that isotretinoin absolute bioavailability was doubled compared to isotretinoin lipid solution (32% and 15% respectively). In conclusion, these beads constitute a novel and efficient system for encapsulation and oral delivery of lipophilic and fragile drugs. © 2007 Elsevier B.V. All rights reserved. Keywords: Bead; Cyclodextrin; Lipid carrier; Vegetable oil; Isotretinoin; Oral bioavailability

1. Introduction Poor water-solubility appears to be an intrinsic property of many drugs. To transport and deliver these drugs, a wide range of particulate systems composed either wholly or partially of lipids have been developed, such as: microemulsions (including selfemulsifying drug delivery systems), liposomes, micro- or nanocapsules and solid lipid nanoparticles (SLN). A number of drawbacks related to the composition and preparation of these lipid-based delivery systems undermine their suitability for use in oral administration. Quite apart from the fact that liposomes display poor stability in the gastrointestinal tract [1], the use of organic solvents in their manufacture (and in that of micro/ nanocapsules) gives rise to the possibility that they could contain toxic residues [2]. Toxicity may also result from the large quantities of surface-active agents required to obtain appropriate carrier characteristics such as particle or droplet size and stability

⁎ Corresponding author. Tel.: +33 146835579; fax: +33 146835308. E-mail address: [email protected] (A. Bochot). 0168-3659/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2007.06.004

[2–5]. Furthermore, in the fabrication of SLN the lipid phase is heated, possibly resulting in damage to labile drugs [4,5]. To overcome these problems, we propose an innovative particulate system, namely “beads” composed of α-cyclodextrin and vegetable oil [6], materials which are considered safe for oral administration [7]. Morphologically, these beads appear as minispheres consisting of a partial crystalline matrix of cyclodextrins surrounding micro-domains of oil [6]. Bead manufacture is very simple, since the continuous external orbital shaking of a mixture of an α-cyclodextrin aqueous solution and soybean oil at room temperature is all that is required to form the particles, thus avoiding the need for organic solvents, cross-linking or surface-active agents and heating [6]. Freeze-drying advantageously transforms beads into dry powder in which the oil content reaches 80% wt and also facilitates ease of handling and use for oral administration [6]. Retinoids are natural or synthetic vitamin A (retinol) derivatives used in the treatment of severe cystic acne, psoriasis and other keratinisation disorders. Numerous reviews have also highlighted their potential in cancer prevention and therapy [8,9]. Among the large number of retinoid compounds, 13-cis-retinoic acid

48

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53

(isotretinoin) is a lipophilic molecule which is unstable by virtue of its sensitivity to heat, oxygen and light [10,11]. This drug is currently formulated as soft capsules mainly composed of soybean oil for oral administration. Isotretinoin was chosen as a model of fragile and poorly-water soluble drug to evaluate the potential of αCD/oil beads for drug encapsulation and oral drug delivery. The present work focuses on the preparation and the characterisation of isotretinoin-loaded beads. The ability of the beads to enhance isotretinoin bioavailability was then investigated in rats and compared to isotretinoin lipid solution. 2. Materials and methods 2.1. Materials Alpha-cyclodextrin (α-CD) (Pharma grade) was purchased from Wacker-Chemie (München, Germany) and Cropure® soybean oil from Croda (Trappes, France). Methylcellulose (MW = 3500–5000 g mol− 1) and N-methyl-2-pyrrolydone were provided by Prolabo (Paris, France), polysorbate 80 was produced by Seppic (Paris, France) and aqueous glucose solution 5% was delivered by Cooper (Melun, France). 13cis-retinoic acid (isotretinoin, MW = 300.45 g mol− 1), 13-trans retinoic acid (tretinoin, MW = 300.45 g mol− 1) and 9-cisretinoic acid (alitretinoin, MW = 300.45 g mol− 1) were provided by Sigma Chemical Co. (St Louis, USA). 1,4-Dioxane, propionic acid and heptane were obtained HPLC grade from Carlo Erba Reagenti (Val de Reuil, France).

2.3.3. Encapsulation efficiency and drug loading Separation and quantification of isotretinoin were achieved by HPLC using the following equipment (Waters, Milford, USA): a mobile phase delivery pump (model 501), an autosampler (model 712 WISP), a tunable absorbance UV–Visible detector (model 486) and a data module (model 746). Samples were introduced into glass insert and the injection volume was set at 20 μL. A Lichrosorb Si column (Modulo-Cart QS Lichrosorb 5 silice 250 × 4 mm; Interchim, Montluçon, France) was employed with dioxan-propionic acid-heptan mixture (20:2:978 v/v/v) as mobile phase and 1.5 mL/min flow rate. These conditions allowed separation of isotretinoin from its main cis or trans isomers, i.e. tretinoin (all-trans-retinoic acid), alitretinoin (9-cis-retinoic acid), 11-cis-retinoic acid, 9,11-dicisretinoic acid and 9,13-dicis-retinoic acid (Fig. 1). Absorbance measurement was performed at λ = 360 nm because this wavelength was suitable to detect most of retinoic acid isomers [12]. Isotretinoin, tretinoin and alitretinoin peaks were identified on the chromatograms at the retention time of 9.6, 12.1 and 11.0 min respectively (Fig. 1). Calibration curves were performed for isotretinoin from 2 to 100 μg/mL and for tretinoin from 5 to 50 μg/mL. The analytical method was validated in terms of accuracy, sensitivity (limit of quantification of 2 μg/mL), repeatability (relative standard deviation ≤5%) and specificity (Rs N 1.5).

2.2. Preparation of isotretinoin-loaded beads Isotretinoin (4.0 mg/mL) was dissolved in soybean oil. 5.8 mL of this oily solution was then added to 20 mL of a α-CD aqueous solution (8.1% w/v). This preparation was continuously shaken at 200 rpm in a gyratory shaker (Salvis, Bioblock Scientific, Illkirch, France) at 28 °C until a monodisperse population of beads was obtained. Beads were then washed and freeze-dried for 48 h to eliminate their water content (Christ LDC-1 alpha1-4 freeze-dryer, Bioblock Scientific). Samples were protected from light exposure during the entire process. Unloaded beads were prepared using the same protocol. 2.3. Characterisation of beads 2.3.1. Bead size Bead diameter was determined before freeze-drying (n = 50 beads) using an optical microscope (Leitz Diaplan microscope, Leica Microsystèmes, France) equipped with a Coolsnap ES camera (Roper Scientific). 2.3.2. Bead fabrication yield Bead fabrication yield was calculated after freeze-drying using the following calculation: Bead fabrication yield ð%Þ weight of freeze  dried beads ¼  100 weightða  CD þ oily solutionÞ

Fig. 1. Chromatograms of isotretinoin dissolved in soybean oil and its main degradation products (a) before illumination and (b) after illumination. Identification peak was as follows: (1) isotretinoin, (2) alitretinoin, (3) tretinoin, (4) other cis/trans retinoic acids isomers and (5) autoxidation products of retinoic acids. Peaks 4 and 5 were attributed according to Bempong et al. [19].

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53

Isotretinoin was extracted from beads before and after freezedrying using the following method. Beads (100 to 300 mg) were precisely weighted and gently destroyed. A known volume (5 to 8 mL) of a mixed solvent of chloroform–methanol 2:1 (v/v) was added to the preparation. The insoluble fraction was eliminated by centrifugation for 5 min at 2000 g in a minicentrifuge (Capsule HF 120, Tomy Tokyo, Japan) and the drug content in the supernatant was quantified by HPLC. Isotretinoin encapsulation efficiency and drug loading were calculated using the following equations according to isotretinoin concentration and bead weights before and after freeze-drying: Encapsulation efficiency ð%Þ Isotretinoin amount within beads ¼  100 Isotretinoin amount in oily solution Drug loading ðmg of Isotretinoin=g of beadsÞ Isotretinoin amount within beads ¼ Bead amount Encapsulation efficiency and drug loading were expressed as mean and standard deviation values (n ≥ 4 batches). Isotretinoin concentrations were compared using Student T test; values of p b 0.05 were considered statistically significant.

49

starts equal to (A + B = 100%) and decreases to a plateau (= B). The isotretinoin degradation half-time (t1/2 degradation) was equal to ln2/Kd and was expressed in hours. Two-way ANOVA test was used to compare values of t1/2 degradation. A p-value of b0.05 was considered statistically significant. Isotretinoin photoconversion into tretinoin was fitted to a first order kinetic model using the following equation:
Isotretinoin converted ð%Þ ¼ C  1  eKc t where t corresponds to time expressed in hours, Kc is the isotretinoin conversion into tretinoin constant (h− 1). Isotretinoin conversion starts equal to zero and increases to a plateau (= C). The conversion half-time (t1/2 conversion) was equal to ln2/ Kc and was expressed in hours. 2.4.2. Protected from light Isotretinoin stability was also examined at room temperature when freeze-dried beads and soybean oil solutions were protected from light. Samples were stored either under a nitrogen or air atmosphere. Aliquots from each formulation were collected after 4 months of storage. Isotretinoin and tretinoin were quantified by HPLC as described above. Two-way ANOVA test was used to compare the values of t1/2 degradation. A p-value b 0.05 was considered statistically significant.

2.4. Stability of isotretinoin beads 2.5. Pharmacokinetic study in rats 2.4.1. Under light exposure Isotretinoin stability was assessed using freeze-dried beads after light exposure according to ICH Guidelines for photostability testing (Guideline Q1B, 1996). The stability of isotretinoin in soybean oil solution (4.0 mg/mL) was carried out as reference. Aliquots of isotretinoin-loaded beads (200 mg) or isotretinoin soybean oil solution (40 μL) were introduced either in hermetic vials under nitrogen or in contact with air using pierced parafilm to close the vials. Samples were placed 30 cm from a full spectrum fluorescent lamp (Master PL E 20 W/865, 230–240 V, colour temperature of about 6500° Kelvin Philips, Paris, France) and exposed to artificial daylight (illumination sample around 100 000 lx) at room temperature for 0, 1, 3, 6, 9, 18, 24 or 48 h. At each time, isotretinoin extraction and quantification were performed as described above. Chromatograms of isotretinoin were analysed before and after sample illumination (Fig. 1). Isotretinoin stability was expressed as the percentage of isotretinoin remaining intact in the formulations (100% corresponding to the initial time point). From the same chromatograms, tretinoin was quantified (μg/mL) to measure the percentage of isotretinoin converted into tretinoin (Fig. 1). The results are presented as mean and standard deviation values (n = 3 for both isotretinoin-loaded beads and isotretinoin soybean oil solutions). Isotretinoin photodegradation was fitted to a first order kinetic model using the following equation: Isotretinoin recovered ð%Þ ¼ A  eKd t þ B where t corresponds to time expressed in hours, Kd is the isotretinoin degradation constant (h− 1). Isotretinoin recovered

2.5.1. Formulation preparation To facilitate oral administration, beads (126 mg of beads/mL, i.e. 0.43 mg of isotretinoin /mL) were dispersed in a solution of methylcellulose (0.6% w/v) and polysorbate 80 (0.5% w/v). Capsules of isotretinoin (Roaccutane® 5 mg, Roche Pharma, France) were opened and their contents were diluted in soybean oil to produce 0.43 mg/mL isotretinoin solution. Intravenous solution was prepared by dissolving isotretinoin (0.48 mg/mL) in a mixture of N-methyl-2-pyrrolydone/ Polysorbate 80/aqueous glucose solution 5% (30/5/65 v/v/v). 2.5.2. Formulation administration Isotretinoin formulations were administered to male Sprague–Dawley OFA rats as single dose and as serial design, 6 catheterised animals per experimental treatment. Isotretinoinloaded bead suspension and isotretinoin soybean oil solution were given orally by gavage at a dose of 3 mg/kg for both formulations and isotretinoin solution by intravenous injection via the caudal vein at a dose of 0.5 mg/kg. Blood sampling was performed at 0.25, 0.5, 1, 2, 4, 6, 8 and 24 h following oral administration and at 0.083, 0.5, 1, 2, 4, 6, 8 and 24 h after intravenous injection. 2.5.3. Quantification of plasma concentration After protein precipitation, plasma samples were added with a 50 μM N-ethylmaleimide (thiol-blocking agent) + 150 μM vitamin C (antioxidant) solution to stabilise isotretinoin [13]. Isotretinoin was separated and eluted on a Symmetry C18 analytical column (5 μm, 150 × 2.1 mm) at 0.3 mL/min (isocratic

50

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53

flow) with a mixture of ammonium acetate (12.5 mM) and acetonitrile (10:90 v/v). Injection volume was set at 30 μL and temperature controlled at 35 °C. Isotretinoin quantification was performed using a tandem spectrometric detector, named API 4000 Sciex system including a Turbo V Source, a TurboIonSpray (TIS) probe and Analyst software, version 1.4.1 (Applied Biosystem, Foster City, CA USA). Mass spectral data were collected using electrospray ionisation in negative mode (ESI−) and parameters were set as follows: curtain gas pressure, 10 A. U.; TIS nebuliser gas pressure, 50 psig; TIS heater gas pressure, 45 psig; TIS voltage, − 4500 V, TIS temperature, 500 °C; collisionally activated dissociation (CAD) gas pressure, 4 A.U. and entrance potential, −10 V. Nilutamide was used as internal standard and daily-prepared calibration curve was done from 0.005 to 1.000 μg/mL. The limit of quantification was 0.005 μg/ mL for isotretinoin in rat plasma. Isotretinoin plasma concentrations were expressed in μg/mL as mean ± S.E. (n = 6). 2.5.4. Data analysis The pharmacokinetic parameters were calculated based on a mono-compartmental model with first order elimination kinetic using the following equation Civ = C0 × e− ClT × T / Vd (R2 = 0.9960). The area under the concentration–time curve (AUC) was calculated from 0.083 to 4 h and from 0.25 to 6 h for intravenous and per os data respectively. Plasma concentration

Fig. 3. Isotretinoin photoconversion into tretinoin kinetic: a) under air atmosphere; data were fitted against one order kinetic equation: R2 = 0.9978, t1/2 = 1.5 h in soybean oil solution (square) and R2 = 0.9307, t1/2 = 2.0 h in beads (circle); b) under nitrogen atmosphere. Data were fitted against one order kinetic equation: R2 = 0.9602, t1/2 = 1.4 h in soybean oil solution (square) and R2 = 0.9768, t1/2 = 2.4 h in beads (circle).

maximum (Cp max) and its corresponding time (Tmax) were obtained directly from the plasma concentration–time profiles. Volume of distribution at steady state (Vd = doseiv / C0), total plasma clearance (ClT = constant rate decrease ⁎ Vd) and elimination half-life (T1/2) were calculated after intravenous administration. Absolute post-oral bioavailability (F) of isotretinoin was established for soybean oil solution and beads as F = (AUCpo ⁎ doseiv) / (AUCiv ⁎ dosepo). Cmax and AUC values were compared using Student T test; values of p less than 0.05 were considered statistically different. 3. Results and discussion 3.1. Isotretinoin-loaded bead characteristics

Fig. 2. Isotretinoin photodegradation kinetic: a) under air atmosphere; data were fitted against one order kinetic equation: R2 = 0.9554, t1/2 = 2.4 h in soybean oil solution (square) and R2 = 0.9862, t1/2 = 4.2 h in beads (circle); b) under nitrogen atmosphere; data were fitted against one order kinetic equation: R2 = 0.9603, t1/2 = 2.6 h in soybean oil solution (square) and R2 = 0.9789, t1/2 = 6.0 h in beads (circle).

To evaluate the potential for α-cyclodextrin/oil beads to entrap and deliver lipophilic and fragile compounds, isotretinoin was chosen as a model drug. Isotretinoin is soluble in vegetable oil and more particularly in soybean oil until 4.0 mg/ mL. Therefore, a saturated solution of isotretinoin in soybean oil was processed in the conditions previously described and optimised [6]. Our study clearly demonstrated that beads containing a therapeutic molecule such as isotretinoin can be prepared by the method developed. Indeed, introduction of isotretinoin in soybean oil had no adverse effect upon bead formation or fabrication yield (= 88 ± 2 and 84 ± 4% with or

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53

without drug respectively). However, isotretinoin-loaded bead diameter was larger than the unloaded bead diameter (2.3 ± 0.2 mm and 1.6 ± 0.2 mm respectively) and their preparation required a longer time to achieve a monodisperse population of particles (7 days instead of 2.5 days). As observed for unloaded beads, isotretinoin-loaded bead diameter increased as a function of shaking time even if fabrication yield reached its maximum value after 5 days. It can be hypothesized that isotretinoin previously dissolved in soybean oil might modify the interactions occurring between cyclodextrin and oil components (triglycerides) at the oil/water interface. These interactions are essential to bead formation [6] and isotretinoin would therefore slightly modify bead characteristics. Quantification of isotretinoin entrapped in beads clearly shows that the process was highly efficient in encapsulating isotretinoin (93 ± 7% before freeze-drying). Bead structure (a matrix surrounding micro-domains of oil) and high oil content [6] facilitates the encapsulation of high amounts of isotretinoin (1.3 mg per grams of beads before freeze-drying). Interestingly, no degradation and no conversion of isotretinoin into isomers were detected (tretinoin concentration was below the detection limit) showing that the manufacturing process is mild and therefore suitable for such a poorly-stable drug. Beads were submitted to freeze-drying in order to facilitate ease of handling and storage. Freeze-drying also permitted the concentration of isotretinoin to reach 3.4 ± 0.2 mg per grams of beads. This stage of the process did not induce significant isotretinoin degradation (p = 0.13). Finally, compared to other lipid-based carriers, such as liposomes, beads appear to be 3.5 times more efficient with respect to isotretinoin entrapment [14] and the quantity of isotretinoin in freeze-dried beads is 5.7 times higher than in isotretinoin-loaded SLN [15]. 3.2. Isotretinoin stability study All isotretinoin photodegradation data followed a first order kinetics with regression coefficient (R2) values ranging from 0.9554 to 0.9862. In contact with air, degradation half-times were significantly longer when isotretinoin was formulated in beads compared to a soybean oil solution (t1/2 degradation = 4.2 and 2.4 h respectively). Similar results were observed for beads and soybean

51

Fig. 5. Plasma concentration of isotretinoin (n = 6 rats, mean and SE) after intravenous injection of 0.5 mg/kg IT solution (diamond), per os administration of IT soybean oil solution 3 mg/kg (square) and IT-loaded bead 3 mg/kg (circle).

oil solution under nitrogen atmosphere (t1/2 degradation = 6.0 and 2.6 h respectively) (Fig. 2). During a short period of time, formulation in beads delayed isotretinoin photodegradation by a factor 2 compared to soybean solution. As isotretinoin does not fit within the α-CD cavity [16,17], the protection insured by α-CD/ oil beads does not result from the formation of an inclusion complex, which is a strategy widely used to stabilise fragile drugs [18]. The protective ability exhibited by beads can, however, be explained by their inner structure [6]. Since isotretinoin is likely to be localised within the micro-compartments of oil, the cyclodextrin-based matrix would therefore play the role of a short-term protective screen from light. Compared to other systems used to protect fragile drugs from light, beads can be considered at least as efficient as liposomes (t1/2 degradation doubled compared to free drug) [14] or as HPβCD (t1/2 degradation doubled for the highest cyclodextrin concentration) [17] in retarding isotretinoin photodegradation. After light exposure, isomerisation clearly appeared to be the main mechanism of isotretinoin degradation whatever the preparation (Fig. 3). Isotretinoin was mainly converted into tretinoin following first order kinetics (R2 values ranging from 0.9307 to 0.9978). For example, at 1 h, 11 ± 2% of isotretinoin in beads was degraded (Fig. 2a and b) and 9 ± 2% was converted into tretinoin (Fig. 3a and b). After longer exposure, other

Table 1 Pharmacokinetic parameters derived from isotretinoin plasma level data following oral administrations of isotretinoin soybean oil solution and isotretinoin-loaded beads in suspension

Fig. 4. Isotretinoin recovered after 4 months of storage (no light exposure) in soybean oil solution (open bars) and in beads (solid bars); under air (a) or nitrogen (b) atmosphere.

Dosage form

Cp max (ng/mL)

Tmax (h)

Soybean oil solution (3 mg/kg) Beads (3 mg/kg)

275 ± 51

1

703 ± 74

15

584 ± 116

1

1479 ± 326

32

AUC (h ng/mL)

Absolute bioavailability (%)

Data are expressed as the mean ± SE (n = 6). Cp max: maximum observed plasma concentration; Tmax: Time of Cp max; AUC: area under the curve over a period of 0.25 to 6 h.

52

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53

isomers were observed on chromatograms (data not shown) and isotretinoin and tretinoin contents rapidly reached constant values (Figs. 2 and 3). This is in a good agreement with continuous and simultaneous retinoic acid cis or trans isomer interconversion, which results in a photostationary isomer mixture [12,19]. It appears that oxidation was not involved in isotretinoin degradation after 48 h of light exposure, since samples kept under both air and nitrogen atmospheres exhibited similar t1/2 degradation values (Fig. 2a and b). In another set of experiments, isotretinoin stability was evaluated when solutions and beads were protected from light during the period of 4 months at room temperature (Fig. 4). In these conditions, isotretinoin exhibited a good stability and no significant difference in isotretinoin content was observed either between soybean oil solution and beads or between air and nitrogen atmospheres. Tretinoin concentrations were below the limit of detection in these samples and a small peak, which could be attributed to an autoxidation product of isotretinoin, was eluted at 23 min [19]. 3.3. Isotretinoin pharmacokinetic study Absolute bioavailabilities for isotretinoin from drug-loaded beads and the reference formulation (prepared by diluting the commercially available medicine) were determined after first obtaining data from animals to which the drug was administered by intravenous injection. Pharmacokinetic parameters relative to intravenous administration were as follows: AUCiv = 783 h ng/mL, Vd = 0.47 L/kg, ClT = 0.64 L/h/kg and T1/2 = 0.5 h. Quantification of isotretinoin concentration in rat plasma revealed that the drug was successfully released from beads (Fig. 5) and that isotretinoin absolute bioavailability was enhanced twofold when employing this new delivery system compared with that of an oily solution (32% and 15% respectively; AUC values were statistically different with p = 0.0427). Cp max was doubled (p = 0.0384) employing beads whereas Tmax remained identical for both formulations (Table 1). Isotretinoin dose could thus be substantially decreased with beads to reach the plasmatic profile of the diluted marketed medicine. Beads and the reference solution have similar lipid composition. Indeed, freeze-dried beads are composed of soybean oil and α-CD (80 and 20% respectively) [6] whereas the oily solution was prepared by diluting isotretinoin soft capsule contents [soybean oil corresponding to around 66% wt., wax and vegetable oils partially hydrogenated (data furnished by Roche Pharma, France)] with soybean oil. However, some differences between the preparations could explain the higher bioavailability found with beads. Firstly, the oily phase is dispersed as micrometric globules within the beads [6]. In contact with biological fluids, the beads would probably disintegrate, releasing the oily components as globules. Isotretinoin peroral performance would be improved by the small size of oily globules, just as has been described for other drugs in dispersed formulations such as microemulsions [20,3]. Secondly, cyclodextrins have been reported to interact with biological membranes, especially with those containing both

cholesterol and phospholipids [21,22]. Thus membrane permeability could be modified and consequently improve isotretinoin absorption. Finally, beads contain partial inclusion complexes formed between α-CD and one fatty acid chain of triglycerides which are the main component of soybean oil [6]. This supermolecule constitutes a surface-active agent due to its amphiphilic property (a hydrophilic head and a hydrophobic tail) [23]. These complexes, once released from beads in biological fluids, could also interact with biological membranes and behave as a penetration enhancer. 4. Conclusions The present study clearly demonstrates that α-cyclodextrin/oil beads are able to efficiently encapsulate and facilitate oral delivery of lipophilic and fragile drugs. These beads are composed of safe materials and are prepared using a mild process. Their inner structure (micro-domains of oil) favours high drug loading. Beads are also able to dramatically increase oral bioavailability of a model drug in rats. By virtue of these properties, beads may open up new prospects for oral delivery of lipophilic drugs. References [1] J.F. Woodley, Liposomes for oral administration of drugs, Crit. Rev. Ther. Drug Carr. Syst. 2 (1) (1985) 1–18. [2] C. Mayer, Nanocapsules as drug delivery systems, Int. J. Artif. Organs 28 (11) (2005) 1163–1171. [3] R.N. Gursoy, S. Benita, Self-emulsifying drug delivery systems (SEDDS) for improved oral delivery of lipophilic drugs, Biomed. Pharmacother. 58 (3) (2004) 173–182. [4] P.M. Bummer, Physical chemical considerations of lipid-based oral drug delivery-solid lipid nanoparticles, Crit. Rev. Ther. Drug Carr. Syst. 21 (1) (2004) 1–20. [5] M. Uner, Preparation, characterization and physico-chemical properties of solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC): their benefits as colloidal drug carrier systems, Pharmazie 61 (5) (2006) 375–386. [6] A. Bochot, L. Trichard, G. Le Bas, H. Alphandary, J.-L. Grossiord, D. Duchêne, E. Fattal, α-cyclodextrin/oil beads: an innovative self-assembling system, Int. J. Pharm. 339 (1–2) (2007) 121–129. [7] T. Irie, K. Uekama, Pharmaceutical applications of cyclodextrins. III. Toxicological issues and safety evaluation, J. Pharm. Sci. 86 (2) (1997) 147–162. [8] M. Okuno, S. Kojima, R. Matsushima-Nishiwaki, H. Tsurumi, Y. Muto, S.L. Friedman, H. Moriwaki, Retinoids in cancer chemoprevention, Curr. Cancer Drug Targets 4 (3) (2004) 285–298. [9] Z. Buletic, K.J. Soprano, D.R. Soprano, Retinoid targets for the treatment of cancer, Crit. Rev. Eukaryot. Gene Expr. 16 (3) (2006) 193–210. [10] X. Tan, N. Meltzer, S. Lindenbaum, Solid state stability studies of 13-cisretinoic acid and all-trans-retinoic acid using microcalorimetry and HPLC analysis, Pharm. Res. 9 (9) (1992) 1203–1208. [11] X. Tan, N. Meltzer, S. Lindenbaum, Determination of the kinetics of degradation of 13-cis-retinoic acid and all-trans-retinoic acid in solution, J. Pharm. Biomed. Anal. 11 (9) (1993) 817–822. [12] M.G. Motto, K.L. Facchine, P.F. Hamburg, D.J. Burinsky, R. Dunphy, A.R. Oyler, M.L. Cotter, Separation and identification of retinoic acid photoisomers, J. Chromatogr. A 481 (1989) 255–262. [13] C.J. Wang, L.H. Pao, C.H. Hsiong, C.Y. Wu, J.J. Whang-Peng, O.Y. Hu, Novel inhibition of cis/trans retinoic acid interconversion in biological fluids — an accurate method for determination of trans and 13-cis retinoic acid in biological fluids, J. Chromatogr. B, Biomed. Sci. Appl. 796 (2) (2003) 283–291.

L. Trichard et al. / Journal of Controlled Release 122 (2007) 47–53 [14] G. Ioele, E. Cione, A. Risoli, G. Genchi, G. Ragno, Accelerated photostability study of tretinoin and isotretinoin in liposome formulations, Int. J. Pharm. 293 (1–2) (2005) 251–260. [15] J. Liu, W. Hu, H. Chen, Q. Ni, H. Xu, X. Yang, Isotretinoin-loaded solid lipid nanoparticles with skin targeting for topical delivery, Int. J. Pharm. 382 (2) (2007) 191–195. [16] S. Munoz Botella, D.A. Lerner, B. del Castello, M.A. Martin, Analytical applications of retinoid-cyclodextrin inclusion complexes. Part 2. Luminescent properties at room temperature, Analyst 121 (11) (1996) 1557–1560. [17] K.L. Yap, X. Liu, J.C. Thenmozhiyal, P.C. Ho, Characterization of the 13cis-retinoic acid/cyclodextrin inclusion complexes by phase solubility, photostability, physicochemical and computational analysis, Eur. J. Pharm. Sci. 25 (1) (2005) 49–56. [18] T. Loftsson, P. Jarho, M. Masson, T. Jarvinen, Cyclodextrins in drug delivery, Expert Opin. Drug Deliv. 2 (2) (2005) 335–351.

53

[19] D.K. Bempong, I.L. Honigberg, N.M. Metlzer, Normal phase LC-MS determination of retinoic acid degradation products, J. Pharm. Biomed. Anal. 13 (3) (1995) 285–291. [20] S. Tenjarla, Microemulsions: an overview and pharmaceutical applications, Crit. Rev. Ther. Drug Carr. Syst. 16 (5) (1999) 461–521. [21] F. Leroy-Lechat, D. Wouessidjewe, J.P. Andreux, F. Puisieux, D. Duchêne, Evaluation of the cytotoxicity of cyclodextrins and hydroxypropylated derivates, Int. J. Pharm. 101 (1–2) (1994) 97–103. [22] R.A. Rajewski, V.J. Stella, Pharmaceutical applications of cyclodextrins.2. In vivo drug delivery, J. Pharm. Sci. 85 (11) (1996) 1142–1169. [23] K. Shimada, K. Kawano, J. Ihsii, T. Nakamura, Structure of inclusion complexes of cyclodextrins with triglyceride at vegetable oil/water interface, J. Food Sci. 57 (1992) 655–656.