Investigation of the transport of lipophilic drugs in structurally diverse lipid formulations through caco-2 cell monolayer using mathematical modeling

Introduction: To be absorbed from the gastrointestinal tract, a drug has to be sufficiently soluble,
because, with some exceptions, passive diffusion of dissolved drug molecules from high to low drug
concentration is the driving force of drug absorption. Different physicochemical and physiological
properties determine the reasons for poor drug absorption, which are poor water solubility, low
membrane permeability, carrier mediated drug efflux, drug metabolism, and pharmacological
interactions. A successful strategy to improve the oral bioavailability of poorly water soluble drugs in
vivo is the use of lipid containing dosage forms. Lipid formulation can reduce the inherent limitation of
slow and incomplete dissolution of poorly water soluble drugs by facilitating the formation of
solubilized phases containing the drug, from which absorption may occur. Only few commercially
available products on this basis have been approved so far. Reasons for this small number of
approved products may be the limited knowledge about formulation parameters that are responsible
for good in vivo performance because of limited understanding of the underlying mechanisms.
Compared to an aqueous suspension of lipophilic drug, it is generally agreed so far that improved drug
absorption takes place because the drug is solubilized already in a lipid containing dosage form. There
is little information in literature dealing with the effect of lipid containing dosage forms on the passive
permeation. The objective of this thesis was to elucidate mechanisms by which a lipophilic drug that is
contained in a lipid formulation is absorbed by the intestine. For this purpose, a theoretical model and
experimental procedures were developed, using Caco-2 cell monolayer.
Methods: Different formulations were tested as model formulations. Since it is known that several
formulation components may work as permeation enhancers by tight junction modulation,
trans-epithelial electrical resistance (TEER) was used as criteria to test monolayer integrity. As model
formulations phosphatidylcholine liposomes, an emulsion with a lipid phase consisting of 67% (m/m)
triglyceride (Captex 8000), 5% (m/m) mixture of mono- and diglycerides (Capmul MCM), 18% (m/m)
surfactant (Cremophor EL), and 10% (m/m) ethanol, and a microemulsion with a lipid phase consisting
of 35.05% (m/m) Captex 8000, 17.58% (m/m) Capmul MCM, 36.84% (m/m) Cremophor EL, and 10% (m/m)
ethanol were chosen. To determine the influence of these model formulations on the permeation of
lipophilic drugs, different drugs were evaluated as suitable model compounds. Propranolol,
progesterone, saquinavir, and triclabendazole were finally selected. An equilibrium dialysis method to
determine the free fraction of the drugs in the different formulations was developed. The influence of
liposomes, microemulsion, and emulsion on transport processes of the model drugs through Caco-2
monolayer was determined with a bi-directional Caco-2 assay, using purely aqueous drug solutions as
reference. At least three different lipid concentrations for each formulation in the range of 0.1-50 mg/ml
were tested. Within each lipid concentration at least three different drug concentrations were tested
per drug. Apparent passive permeability coefficient of the apical (Pa) and of the basal membrane (Pb),
formulation-to-cell partition coefficient, and carrier mediated apical efflux rate were deduced by fitting a
mathematical model to the experimental concentration data of the bi-directional assay using Easy Fit®
fitting software. Further, a biophysical model was developed to delineate the contribution of drug
transport in the diffusion boundary layer and drug permeation through cell membrane to the
determined apparent permeability coefficient. Additionally, a differentiation was introduced between
permeation of free drug through the cell membrane and permeation following direct drug transfer from
the lipid particles to the membrane upon collision. Drug uptake and passive drug efflux for selected
drugs and formulations were further studied in the Caco-2 cell monolayer.
Results and Discussion: Both, the model for the determination of absorption parameters in Caco-2
cells and the biophysical model for delineating the components of apparent permeability coefficient
explained the experimental data satisfactorily. Generally Pa, Pb, and free fraction decreased with
increasing lipid concentration. Within the same lipid concentration, no influence of drug concentration
on Pa, Pb, and free fraction was determined. Triclabendazole showed lower Pb than Pa whereas
permeability coefficients of all other drugs were equal for both membranes. Carrier mediated apical
drug efflux was found for saquinavir only and its rate, when expressed as zero order, decreased with
increasing lipid concentration and increased with increasing drug concentration. Formulation-to-cell
partition coefficient increased with increasing lipid concentration for all drugs and formulations.
Deduced permeability coefficients of diffusion boundary layer, reflecting drug transport in the apical
and basal solution, was smaller than overall permeability coefficient of cell membrane for all drugs
except saquinavir for which values were comparable. This indicates that the compounds are good
permeable for cell membrane. Permeability coefficient of the drug corresponding to direct mass
transfer from lipid particle to cell membrane (Pm,L) was for progesterone greater than the permeability
coefficient corresponding to permeation of free drug through cell membrane (Pm,d). For triclabendazole
Pm,L was smaller than Pm,d. For saquinavir Pm,L was comparable or smaller than Pm,d. Finally for
propranolol Pm,L was smaller than Pm,d for liposome formulation. For propranolol emulsion and
microemulsion, no interaction of formulation and drug was observed, therefore no meaningful values
were obtained for Pm,L. The rate limiting step of transport and the dominating mechanism of membrane
permeation depend on the corresponding permeability coefficients and the free and lipid bound drug
concentration. These observations apply to all three structurally different lipid formulations used in this
study. Permeability coefficients of drug uptake of progesterone formulations and passive drug efflux of
progesterone and triclabendazole formulations in the Caco-2 monolayer decreased with increasing
lipid concentration, which was consistent with the permeation experiments.
Drug fluxes increased with increasing drug concentration within the same lipid concentration and
decreased with increasing lipid concentration within the same drug concentration. Fluxes of
progesterone in experiments with equal free drug concentration increased with simultaneously
increasing drug and lipid concentration. This was demonstrated with progesterone liposomes and can
be related to the larger Pm,L compared to Pm,d found for this drug.
Conclusions: Lipid formulations containing the drug in a molecular form provide the possibility to
increase the concentration of poorly water soluble drugs in a macroscopically aqueous system.
Apparent drug permeability coefficient for the cell membrane is decreased by these formulations.
Apparent drug permeability coefficient depends on free fraction, whereas drug flux depends on
absolute amount of free drug in water phase. Therefore simultaneous increase of drug and lipid
concentration can provide an undiminished drug flux, which may improve bioavailability by prolonged
intestinal absorption at a sustained rate. These findings are independent of the composition and the
structure of the lipid formulation lending support to the universal nature of this conclusion. In addition
flux can be further increased by direct drug transfer from lipid particle to cellular membrane. This was
observed for only one drug in the present work. The necessary drug properties for this direct transfer
to take place should be investigated in the future. The results of this work shed light into the
mechanism of drug absorption from lipid formulations and demonstrate potential beneficial effects of
these formulations on absorption of lipophilic drugs in vivo. They may be used for the development of
efficient oral dosage forms to improve bioavailability for these drugs.