European Journal of Pharmaceutics and Biopharmaceutics (v.66, #1)

APV Diary (S1-S3).

The selection of a suitable vehicle for preclinical compound profiling is a very important task during the early developmental phases to ensure the quality of candidates and the speed of compound progression. Apart from biopharmaceutical and pharmaceutical technical considerations, i.e. solubility/dissolution improvement or route of application, other aspects have to be taken into account, as well: (i) availability and quality of the compound, (ii) tolerability of the vehicle in the selected animal model, (iii) developmental possibilities, i.e. whether the formulation can be transformed into a clinical one.The approach described in this paper is based on results of team collaboration between functions involved in the conduct of animal experiments (Pharmacology, Pharmacokinetics, Toxicology, and Pharmaceutical Sciences). Very early in vivo studies should be performed with dissolved API as available information on solid-state characteristics is usually limited at this time. Later studies should be performed with developable formulations, taking into consideration pharmacological, toxicological, and pharmaceutical requirements. At this stage, delivery strategies (i.e. advanced formulations and/or alternative routes of administration) should be considered, as well. In addition, a minimum level analytical characterization of compounds and formulations used in animal studies is required to explain unexpected results.
Keywords: Early formulation strategy; Poor solubility compounds; Physical-chemical properties; Standard formulations; Enabling formulations; Harmonization; Customized vehicles;

Cellulose acetate butyrate–pH/thermosensitive polymer microcapsules containing aminated poly(vinyl alcohol) microspheres for oral administration of DNA by Gheorghe Fundueanu; Marieta Constantin; Fabrizio Bortolotti; Rita Cortesi; Paolo Ascenzi; Enea Menegatti (11-20).
The aim of this work is to safely transport bioadhesive microspheres loaded with DNA to intestine and to test their bioadhesive properties. Poly(vinyl alcohol) (PVA) microspheres were prepared by dispersion reticulation with glutaraldehyde and further aminated. These microspheres were firstly loaded with plasmid DNA by electrostatic interactions and then entrapped in cellulose acetate butyrate (CAB) microcapsules for gastric protection. The entrapped PVA microspheres do not have enough force by swelling to produce the rupture of CAB shell, therefore the resistance of microcapsules was weakened by incorporating different amount of the pH/thermosensitive polymer (SP) based on poly(N-isopropylacrylamide-co-methyl methacrylate-co-methacrylic acid) (NIPAAm-co-MM-co-MA). This polymer is insoluble in gastric juice at pH 1.2 and 37 °C, but quickly solubilized in intestinal fluids (pH 6.8 and pH 7.4). Therefore, DNA loaded PVA microspheres were not expelled in acidic media but were almost entirely discharged in small intestine or colon. The integrity of DNA after entrapment was tested by agarose gel electrophoresis indicating that no DNA degradation occurs during encapsulation. The percentage of adhered microspheres on the mucus surface of everted intestinal tissue was 65 ± 18% for aminated PVA microspheres without DNA and almost 50 ± 15% for those loaded with DNA. Non-aminated PVA microspheres display the lowest adhesive properties (33 ± 12%). In conclusion DNA loaded microspheres were progressively discharged in intestine. The integrity of DNA was not modified after entrapment and release, as proved by agarose gel electrophoresis. Both loaded and un-loaded aminated microspheres display good bioadhesive properties.
Keywords: DNA; pH/thermosensitive polymers; Bioadhesive microspheres; Drug delivery systems;

Previously, we prepared a biodegradable elastomeric device that can release different therapeutic proteins at a nearly constant rate in nanomolar concentrations with high bioactivity. The elastomer device was fabricated using a photo-initiated free radical cross-linking reaction of acrylated star(ε-caprolactone-co-d,l-lactide) in organic solvent in the presence of solid protein particles. The objective of this study was to examine the effect of various parameters used for fabricating the photo-cross-linked elastomeric device on the stability of a therapeutic protein, vascular endothelial growth factor (VEGF), to determine which factor plays the dominant role in protecting VEGF. VEGF was lyophilized with or without bovine serum albumin (BSA) and then suspended in solid state in a macromer (acrylated star-poly(ε-caprolactone-co-d,l-lactide)) solution containing different concentrations of a free radical initiator, 2,2-dimethoxy-2-phenylacetophenone (DMPA). The protein suspension was then UV-irradiated at different intensities. UV irradiation with the generation of free radicals was detrimental to VEGF stability. BSA preserved the VEGF bioactivity during UV irradiation but provided little protection in the presence of the photo-initiator DMPA. The acrylated macromer acted as a free radical scavenger and effectively preserved VEGF and BSA stability during UV-initiated photo-polymerization. The detrimental effect of UV radiation with free radical generation on VEGF stability during device manufacture can be eliminated by choosing the proper bulking agents coupled with an efficient photo-polymerization reaction.
Keywords: Photo-cross-linking; Vascular endothelial growth factor; Protein stabilization; Protein bioactivity;

A nuclear localization signal was non-covalently attached to DNA for the purpose of enhancing transfection efficiencies of non-viral gene carriers. Psoralen, a nucleic acid-intercalating agent, was chemically attached to a signal peptide. The conjugate spontaneously intercalated into DNA and then poly(ethyleneimine) [PEI] was added to prepare a DNA/PEI complex containing the signal peptide moieties. The existence of the conjugate did not alter the complexation process between DNA and PEI, which was confirmed by dynamic light scattering. The conjugate was slowly released from the DNA/PEI complex for 24 h, while a burst release was examined when the conjugated was added to DNA without PEI. The complex containing a signal peptide moiety increased transfection efficiencies on COS-1 cells, compared to a mutant signal peptide or a control. Cytotoxicity of the conjugate slowly increased as the amount of the conjugate increased, however, the cytotoxic effect of the conjugate was not significant at the effective concentration of the conjugate for transfections. Therefore, the psoralen-nuclear localization signal is expected to be a potent transfection enhancing agent without a covalent modification of transgenes.
Keywords: Nuclear localization signal; Transfection; PEI; Psoralen;

This study was aimed at developing a polymeric drug delivery system for a new and potent antitumor drug, 9-nitrocamptothecin (9-NC), intended for both intravenous administration and improving the therapeutic index of the drug. To achieve these goals, 9-NC loaded poly(dl-lactide-co-glycolide) (PLGA) nanoparticles were prepared by nanoprecipitation method and characterized. The full factorial experimental design was used to study the influence of four different independent variables on response of nanoparticle drug loading. Analysis of variance (ANOVA) was used to evaluate optimized conditions for the preparation of nanoparticles. The physical characteristics of PLGA nanospheres were evaluated using particle size analyzer, scanning electron microscopy, differential scanning calorimetry and X-ray diffractometry. The results of optimized formulations showed a narrow size distribution with a polydispersity index of 0.01%, an average diameter of 207 ± 26 nm, and a drug loading of more than 30%. The in vitro drug release profile showed a sustained 9-NC release up to 160 h indicating the suitability of PLGA nanoparticles in controlled 9-NC release. Thus prepared nanoparticles described here may be of clinical importance in both stabilizing and delivering camptothecins for cancer treatment.
Keywords: 9-Nitrocamptothecin; PLGA; Nanoparticles; Full factorial experimental design; Release kinetics;

Superoxide dismutase (SOD) is the most potent antioxidant enzyme. In this study, SOD was encapsulated in chitosan microspheres to obtain suitable sustained protein delivery. Protein-loaded chitosan microspheres with various formulations were prepared based on complex coacervation process. Due to the inherent characteristic of SOD, high encapsulation efficiency could not be obtained with simple preparation method. The pH of chitosan solution is 3.0; when the chitosan microspheres were prepared with this solution, encapsulation was low. Therefore, several strategies have been tested to increase the encapsulation efficiency and good results have been obtained. 70–80% protein encapsulation efficiency was obtained. The addition of PEG to the protein solution enhanced the encapsulation efficiency also. Mean sizes of microspheres were between 1.38 and 1.94 μm. Factors affecting the release behaviour of SOD from microspheres have been studied. They included pH values of chitosan solution (the pH of chitosan solution is 3.0), addition of PEG to the protein solution and the use of adsorption technique. In general, biphasic release profiles were obtained with these formulations. The protein activity changed between 70 and 100% during the release. In general, the protein activity remained in acceptable limits. The SOD encapsulated chitosan microspheres can be prepared by changing the pH or addition of PEG, allowing the safe incorporation of protein for controlled release.
Keywords: Chitosan; Superoxide dismutase; Microspheres; PEG; Controlled release; Protein;

Encapsulation of ketoprofen for controlled drug release by Adi I. Arida; Moawia M. Al-Tabakha (48-54).
Ketoprofen particles were encapsulated with polyions and gelatin to control the release of the drug in aqueous solutions. Charged linear polyions and gelatin were alternatively deposited on 6 μm drug microcrystals through layer-by-layer (LbL) assembly. Sequential layers of poly(dimethyldiallyl ammonium chloride) (PDDA) and poly(styrenesulfonate) (PSS) were followed by adsorption of two to six gelatin/PSS bilayers with corresponding capsule wall thicknesses ranging from 41 to 111 nm. The release of Ketoprofen from the coated microparticles was measured in aqueous solutions of pH 1.4, 4.1, and 7.4. The release rate has changed at these different pH values. At pH 7.4 the release rate of Ketoprofen from the encapsulated particles was less by 107 times compared to uncoated Ketoprofen. The results provide a method of achieving prolonged drug release through self-assembly of polymeric shells on drug crystals.
Keywords: Ketoprofen; Encapsulation; Controlled release;

Estradiol sustained release from high affinity cyclodextrin hydrogels by C. Rodriguez-Tenreiro; C. Alvarez-Lorenzo; A. Rodriguez-Perez; A. Concheiro; J.J. Torres-Labandeira (55-62).
Hydrogels for loading estradiol and controlling its release were prepared cross-linking various cyclodextrins with ethyleneglycol diglycidylether. To select the more adequate cyclodextrins, estradiol solubility diagrams in water with β-cyclodextrin (βCD), methyl-β-cyclodextrin (MβCD), hydroxypropyl-β-cyclodextrin (HPβCD), and sulfobutyl-β-cyclodextrin (SBβCD) were made in absence and presence of hydroxypropyl methylcellulose (HPMC) applying or not autoclaving. Although all cyclodextrins showed enough complexation capability, the low solubility of βCD and the high anionic character of SBβCD hindered the cross-linking process, and these cyclodextrins were discarded for preparing hydrogels. Hydrogels prepared with MβCD (20%, 25%) or HPβCD (20%, 25%, and 30%), with or without HPMC 0.25%, absorbed 4–10 times their weight in water and loaded up to 24 mg estradiol per gram, which is 500 times greater than the amount of drug that can be dissolved in their aqueous phase. Positive linear correlation was found between the stability constant and the network/water partition coefficients of drug. The hydrogels sustained the release up to one week; the affinity of estradiol for the cyclodextrin units controlling the process, as shown by the negative correlation with the release rate constants. These results highlight the potential of cyclodextrin complexation for the development of hydrogels useful in loading hydrophobic drugs and controlling their release.
Keywords: Hydroxypropyl-β-cyclodextrin; Methyl-β-cyclodextrin; Sulfobutyl-β-cyclodextrin; Inclusion complexes; Hydrophobic drugs; Chemically cross-linked hydrogels; Controlled release;

Influence of moisture content on the mechanical properties of methyl methacrylate–starch copolymers by I. Bravo-Osuna; C. Ferrero; M.R. Jiménez-Castellanos (63-72).
The water vapour sorption–desorption behaviour of graft copolymers (hydroxypropylstarch–methyl methacrylate -HSMMA- and carboxymethylstarch–methyl methacrylate -CSMMA-) synthetised by free-radical polymerisation and alternatively dried by oven (OD) or freeze-drying (FD) techniques was investigated in a previous paper. The aim of the present study was to analyse the influence of the amount and distribution of water molecules on the flow and compaction characteristics of this family of methyl methacrylate–starch copolymers. Products were stored at constant temperature (25 °C) and different relative humidity conditions (RH). Flow properties of the powdered materials were evaluated using glass and stainless-steel funnels and the densification behaviour was studied in detail by means of Heckel treatment and compression parameters. Results revealed that the storage at 25–50% RH was the optimum condition relating flowability for HSMMA and OD-CSMMA copolymers. At higher RH values, the flow characteristics worsened, due to an increment in cohesive forces. Compaction experiments showed that the 25–50% RH range improved also the compression performance of the copolymers, due to increasing powder compressibility and reduced compact relaxation. Under these circumstances, absorbed water might act as plasticiser and adsorbed water as lubricant.
Keywords: Methyl methacrylate–starch copolymers; Flow properties; Heckel analysis; Compression behaviour; Moisture content;

The aim of the present research work was to systemically device a model of factors that would yield an optimized sustained release dosage form of an anti-hypertensive agent, losartan potassium, using response surface methodology by employing a 3-factor, 3-level Box-Behnken statistical design. Independent variables studied were the amount of the release retardant polymers – HPMC K15M (X 1), HPMC K100M (X 2) and sodium carboxymethyl cellulose (X 3). The dependent variables were the burst release in 15 min (Y 1), cumulative percentage release of drug after 60 min (Y 2) and hardness (Y 3) of the tablets with constraints on the Y 2  = 31–35%. Statistical validity of the polynomials was established. In vitro release and swelling studies were carried out for the optimized formulation and the data were fitted to kinetic equations. The polynomial mathematical relationship obtained Y 2 = 32.91 - 2.30 X 1 - 5.69 X 2 - 0.97 X 3 - 0.41 X 1 X 2 + 0.21 X 1 X 3 - 0.92 X 1 2 - 1.89 X 2 2 ( r 2 = 0.9944 ) explained the main and quadratic effects, and the interactions of factors influencing the drug release from matrix tablets. The adjusted (0.9842) and predicted values (0.9893) of r 2 for Y 2 were in close agreement. Validation of the optimization study indicated high degree of prognostic ability of response surface methodology. Tablets showed an initial burst release preceding a more gradual sustained release phase following a non-fickian diffusion process. The Box-Behnken experimental design facilitated the formulation and optimization of sustained release hydrophilic matrix systems of losartan potassium.
Keywords: Hydrophilic matrix tablets; Box-Behnken statistical design; Response surface methodology; Sustained release;

Development of starch-based pellets via extrusion/spheronisation by Aleksandra Dukić; Raoul Mens; Peter Adriaensens; Paul Foreman; Jan Gelan; Jean Paul Remon; Chris Vervaet (83-94).
A modified starch (high-amylose, crystalline and resistant starch) was evaluated as an alternative excipient to microcrystalline cellulose for pellets prepared via extrusion/spheronisation. Theophylline anhydrous (25%, w/w) was used as a model drug. A binder was necessary to obtain an acceptable yield and the addition of sorbitol improved the surface properties of the pellets. A surface response design with three formulation variables (binder, sorbitol and water level) and one process variable (spheronisation speed) was used to optimise the process and to evaluate pellet yield, sphericity (aspect ratio and two-dimensional shape factor, e R), size (mean Feret diameter), friability and disintegration properties. Mixer torque rheometry and solid-state NMR revealed a significant influence of sorbitol on wet mass consistency and pellet properties. A high pellet yield (>90%), acceptable sphericity (AR < 1.2), low friability (<0.01%), fast disintegration (<10 min) and complete drug release in less than 20 min for all formulations, demonstrated the potential of this modified starch in formulations intended for extrusion/spheronisation.
Keywords: Extrusion/spheronisation; Pellets; Starch; Sorbitol; NMR; Experimental design;

Characterization of hot-melt extruded drug delivery systems for onychomycosis by Praveen K. Mididoddi; Michael A. Repka (95-105).
The objectives of this investigation were to study the physico-chemical properties of hot-melt extruded (HME) films for onychomycosis and to determine the stability of the model antifungal drug incorporated within these films. The influence of etching and instrument variables on the bioadhesion of these drug delivery systems for the human nail was also studied. Six 250 g batches (F1–F6) of hydroxypropyl cellulose (HPC) and/or poly(ethylene oxide) films containing ketoconazole (20%) were extruded using a Killion extruder (Model KLB-100). The thermal properties of HME films were investigated using differential scanning calorimetry (DSC). Scanning electron microscopy (SEM) was used to examine the surface morphology of the films and X-ray diffraction (XRD) was used to investigate the crystalline properties of the drugs, physical mixtures as well as the HME films. Stability studies were performed on the films stored at 25 °C/60%RH. The bioadhesive properties of these films were investigated on the human nail (ex vivo) using a Texture Analyzer®. The nail samples tested were either non-treated (control) or treated with an etching gel. The parameters measured were peak adhesion force (PAF) and area under the curve (AUC). The Hansen solubility parameter was calculated using a combination of Hoy and Hoftyzer/Van Krevelen methods to estimate the likelihood of drug–polymer miscibility. SEM provided direct physical evidence of the physical state of the drug within the films. The theoretical post-extrusion content of ketoconazole remaining in the six film batches ranged from 90.3% (±2.2) to 102.4% (±9.0) for up to 6 months and from 83.9% (±3.6) to 91.6% (±3.0) for up to 12 months. Bioadhesion studies of HPC film tested on ‘etched’ nails recorded significantly higher PAF and AUC than that of the non-treated ‘control’ nails. Ketoconazole was found to be relatively stable during the extrusion process. Melting points corresponding to the crystalline drugs were not observed in the processed films. The Hansen solubility parameters predicted miscibility between the polymers and the drug. The predictions of the solubility parameters were in agreement with DSC, XRD and SEM results. Bioadhesion measurements of the film on the human nail substrate were generally higher for the etched nails than that of the control nails.
Keywords: Nail; Hot-melt extrusion; Stability; Bioadhesion; Texture analyzer;

Previous studies revealed that solid dispersions containing nimodipine and polyethylene glycol 2000 can be effectively prevented from recrystallization by adding povidone K17. These systems are characterized by a high dissolution rate and a remarkable supersaturation of the drug in the dissolution media. It is still unknown if these characteristics are achievable with all polyethylene glycol and povidone mixtures. The objective of the present study is to find out, whether povidone K17 has to be dissolved in melted polyethylene glycol during the preparation process of solid dispersions by the melting method in order to avoid recrystallization of the drug and to ensure storage stability. Solid dispersions consisting of 20% (m/m) nimodipine, 16% (m/m) povidone K17 and 64% (m/m) of six different mixtures of polyethylene glycol 2000 and 8000 were prepared by the melting method and investigated by dissolution testing, thermal analysis and X-ray diffraction. As the solubility of povidone K17 in polyethylene glycol 2000 is about 70% at 65 °C and decreases with increasing molecular weight of the polyethylene glycol, mixtures containing different amounts of dissolved povidone K17 are obtained by varying the mixing ratio of polyethylene glycol 2000 and 8000. Recrystallization is inhibited in the formulations, containing mainly polyethylene glycol 2000 whereas recrystallization occurs in systems consisting predominantly of polyethylene glycol 8000. These results show clearly that dissolution of povidone in melted polyethylene glycol is a prerequisite in order to prevent recrystallization.
Keywords: Dissolution; Nimodipine; Polyethylene glycol; Povidone; Recrystallization; Solid dispersions; Supersaturation; Stability;

Mechanistic analysis of drug release from tablets with membrane controlled drug delivery by Sandra Strübing; Hendrik Metz; Karsten Mäder (113-119).
The objective of this study was to receive detailed information on the mechanism of drug release from polyvinyl acetate (PVAc)/polyvinyl alcohol–polyethylene glycol graft copolymer (PVA–PEG) coated Propranolol HCl and Theophylline tablets. For this purpose the coating composition (PVAc/PVA–PEG: 90/10 and 80/20) and the amount of the coating layer have been varied. Due to its better solubility Propranolol HCl showed higher release rates than Theophylline. As expected, a higher percentage of the water soluble polymer accelerated drug release. Increased coating thickness led to amplified lag times of drug release. The water uptake of the tablets was quantified by gravimetric analysis. Furthermore, the microenvironment of the tablet core was monitored by EPR spectroscopy. For this purpose a hydrophilic EPR spin probe was incorporated into tablets. Surprisingly, despite a lag phase at the beginning and a controlled drug release over 24 h, the results of the EPR studies indicated an immediate water penetration through the coating layer into the tablet core. The water is able to solubilise the majority of water soluble compounds within minutes. The results obtained in this study demonstrate, that EPR is a powerful method to monitor the first steps of diffusion processes and the physicochemical state of coated dosage forms.
Keywords: EPR; ESR; Controlled release; Tablet; Film; Coated dosage forms;

The purpose of the present study was to investigate the influence of different drugs exhibiting different solubility on the viscoelastic properties and on the skin diffusion profile of a ringing gel. In a preliminary rheology study with the placebo gel predominating elastic properties were confirmed and a temperature influence was indicated. Fluconazole, fludrocortisone-acetate, flumethasone-pivalate, flutamide and flufenamic-acid each 1% (w/w) were incorporated into the preparation and oscillatory measurements were performed at temperatures of 25, 28, 32 and 37 °C. In all drug containing formulations a high elastic G′ value predominated the viscous G″ value. The highest G′ value could be obtained with the incorporated flumethasone-pivalate. Additionally in almost all cases the G′ values decreased with increasing temperature compared to the placebo gel. Additionally in vitro standard diffusion experiments using Franz-type cells and porcine skin were performed. Following rank order of the cumulative drug release after 48 h was obtained: fluconazole > flufenamic-acid > flumethasone-pivalate > flutamide > fludrocortisone-acetate. Furthermore an excellent chemical stability of all incorporated drugs was confirmed over 10 weeks.
Keywords: Fluconazole; Fludrocortisone-acetate; Flumethasone-pivalate; Flutamide; Flufenamic-acid; Skin permeation; Rheology; Chemical stability;

Permeability of the reconstructed human epidermis model Episkin® in comparison to various human skin preparations by Frank Netzlaff; Monika Kaca; Udo Bock; Eleonore Haltner-Ukomadu; Peter Meiers; Claus-Michael Lehr; Ulrich F. Schaefer (127-134).
The objective of this work was to compare the barrier function of the small diameter reconstructed human epidermis model Episkin® (d  = 12 mm) to human skin in vitro. For that purpose a modification for the Franz diffusion cell (d  = 15 mm) had to be developed so as to allow direct comparison with the following human skin preparations: Full thickness skin (FTS), split thickness skin (STS), heat-separated epidermis (HSE), and trypsin isolated stratum corneum (TISC). Among the tested preparations, HSE appeared to be the most preferable due to its clear morphological structure and ease of preparation. The lipid profile of HSE and Episkin® was analyzed and showed significant differences in terms of cholesterol, ceramides and triglycerides contents, whereas cholesterol esters and fatty acids were not different. Permeation data with HSE and Episkin® were then gathered using caffeine and testosterone. Both test compounds permeated much faster through Episkin® than through HSE. Moreover, opposed to Episkin®, HSE differentiated between the two test compounds. In spite of the remarkable progress in developing RHEs in the past years at this time Episkin® can obviously not yet fully replace human skin for in vitro permeability experiments.
Keywords: Episkin®; Testosterone; Caffeine; Stratum corneum; Heat-separated human epidermis; Split thickness skin; Full thickness skin; Skin lipids;

Anthranoid laxatives influence the absorption of poorly permeable drugs in human intestinal cell culture model (Caco-2) by Leena Laitinen; Elina Takala; Heikki Vuorela; Pia Vuorela; Ann Marie Kaukonen; Martti Marvola (135-145).
Interactions between widely used anthranoid laxatives and other simultaneously administered drugs are not known. In this paper, the influence of rhein, danthron, sennidins A/B, sennosides A/B, and senna leaf infusion was investigated on the permeability of furosemide, ketoprofen, paracetamol, propranolol, verapamil, digoxin, and Rhodamine 123 across Caco-2 monolayers. The effects on monolayer integrity ([14C]mannitol permeability, TEER) were also determined.The in vitro absorption of highly permeable drugs was not strongly affected during co-administration of the laxatives. Furosemide permeability was enhanced by rhein and danthron (3.6 and 3.0-fold), which may partly be due to opening of the paracellular spaces and/or effects on active efflux. However, the secretory permeability of digoxin and Rho 123 was not strongly affected by rhein and danthron, suggesting that inhibition of MDR1 was not responsible for the increased permeation of furosemide. The absorptive permeability of digoxin was decreased by rhein and danthron, offering evidence for effects on apical membranes. The effects on monolayer integrity were detectable, but reversible. According to presented experiments, daily use of laxatives with well-absorbing drugs would seem unlikely to affect drug permeability, but the effects on the absorption of poorly permeable drugs cannot be excluded.
Keywords: Anthranoid laxatives; Sennosides; Caco-2 permeability; Absorption of drugs;

Absorption of poorly water soluble drugs subject to apical efflux using phospholipids as solubilizers in the Caco-2 cell model by Susanne B. Kapitza; Bettina R. Michel; Peter van Hoogevest; Mathew L.S. Leigh; Georgios Imanidis (146-158).
The purpose of this work was to determine the influence of liposomal solubilization of poorly water soluble drugs exhibiting apical efflux on permeation kinetics and cell toxicity in Caco-2 cells. The HIV-protease inhibitors indinavir and saquinavir were incorporated in phosphatidylcholine liposomes at maximal drug-to-lipid mass ratios and their absorption was determined in Caco-2 cell cultures grown on Transwell inserts using purely aqueous drug solutions as reference. A novel mathematical model was developed to quantitatively delineate the contribution of passive membrane permeation and carrier mediated efflux to transport across the cell monolayer and passive permeability coefficient and maximal efflux rate and affinity constant of the transporter system were determined. Cell toxicity of phospholipids was evaluated by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) and the lactate dehydrogenase (LDH) assay. Cell integrity was not significantly affected by phospholipid concentrations of up to 150 mg/ml with respect to the used standard tests. Maximum drug concentration was increased 10- and 750-fold for indinavir and saquinavir, respectively, by the use of liposomes. The passive membrane permeability coefficient differed between the two drugs in accordance with their lipophilicity and the affinity for apical efflux transporters was on average 4-fold greater for saquinavir than for indinavir. Liposomal solubilization diminished the passive permeability coefficient of both drugs but the passive apical-to-basal delivery rate was increased by the liposomes compared to the purely aqueous solutions at maximal donor concentrations for at least one of the two drugs. Efflux rate reached a maximum for the liposomal formulations reflecting transporter saturation. Hence, liposomal solubilization considerably increased drug concentration in the media and altered absorption behavior by affecting both the passive diffusion and the carrier mediated efflux components of cell monolayer permeation.
Keywords: Caco-2; Drug absorption; HIV-protease inhibitors; Efflux; Liposomes; Phospholipids; Kinetic modeling;