No abstract

To date, the Caco-2 model is considered as the gold standard to predict intestinal drug absorption. Often, aqueous phosphate buffers are used as apical medium. The purpose of this study was to use fasted state human intestinal fluid (FaHIF) as apical solvent system to generate biorelevant permeability values for a series of 16 model drugs that can be used as reference data to critically evaluate fasted state simulated intestinal fluid (FaSSIF) as possible substitute medium. Caco-2 compatibility with FaHIF was achieved when 50mg/ml mucus was applied on top of the cells before adding the apical medium. The use of FaHIF as solvent system generated a broad range of apparent permeability values (Papp) for the series of model compounds. When Papp values obtained with FaHIF were compared to those obtained with FaSSIF, a strong correlation was observed (R=0.951). The use of FaSSIF in the absence of mucus did not significantly alter this correlation. For FaHIF, FaSSIF and reference phosphate buffer blank FaSSIF, a strong sigmoidal relationship was found between Papp and fahuman, illustrated by correlation coefficients of 0.961, 0.893 and 0.868, respectively. In terms of inter-subject variability, the use of FaHIF from different volunteers originating from two distinct age groups (18-25 years; 65-72 years) exhibited an average coefficient of variance (CV) of 30%. However, no age dependency in permeability could be observed. In conclusion, the data generated in this article justify the use of FaSSIF as biorelevant apical medium in the Caco-2 assay to accurately predict in vivo drug absorption. Also, the optimized mucus-containing Caco-2 model can be used in combination with intestinal fluid samples aspirated after drug administration to further investigate intraluminal drug and formulation behavior.

No abstract

The Ussing chambers model is almost exclusively used in the presence of plain aqueous phosphate buffers as solvent system. In an attempt to further elucidate the effect of luminal ingredients and postprandial conditions on intestinal permeability, pooled fasted and fed state human intestinal fluids (FaHIFpool, FeHIFpool) were used. In addition, simulated intestinal fluids of both nutritional states (FaSSIF, FeSSIF) were evaluated as possible surrogate media for HIF. The use of FaHIFpool generated a broad range of Papp values for a series of 16 model drugs, ranging from 0.03×10(-6)cm/s (carvedilol) to 33.8×10(-6)cm/s (naproxen). A linear correlation was observed between Papp values using FaSSIF and FaHIFpool as solvent system (R=0.990), justifying the use of FaSSIF as surrogate medium for FaHIF in the Ussing chambers. In exclusion of the outlier carvedilol, a strong sigmoidal relationship was found between Papp and fahuman of 15 model drugs, illustrated by correlation coefficients of 0.961 and 0.936 for FaHIFpool and FaSSIF, respectively. When addressing food effects on intestinal permeability, the use of FeHIFpool resulted in a significantly lower Papp value for nine out of sixteen compounds compared to fasting conditions. FeSSIF as solvent system significantly overestimated Papp values in FeHIFpool. To conclude, the optimized Ussing chambers model using biorelevant media as apical solvent system holds great potential to investigate food effects in a more integrative approach, taking into account drug solubilisation, supersaturation and formulation effects.

No abstract