Prediction of drug transport through the blood-brain barrier in vivo: a comparison between two in vitro cell models

Abstract : Purpose. Studies were conducted to evaluate whether the use of an in vitro model of the blood-brain barrier (BBB) resulted in more accurate predictions of the in vivo transport of compounds compared to the use of a human intestinal cell line (Caco-2). Methods. The in vitro BBB model employs bovine brain capillary endothelial cells co-cultured with primary rat astrocytes. The Caco-2 cells originate from a human colorectal carcinoma. The rat was used as experimental animal for the in vivo studies. Results. Strong correlations (r = 0.93–0.95) were found between the results generated by the in vitro model of the BBB and two different methodologies to measure the permeability across the BBB in vivo. In contrast, a poor correlation (r = 0.68) was obtained between Caco-2 cell data and in vivo BBB transport. A relatively poor correlation (r = 0.74) was also found between the two in vitro models. Conclusion. The present study illustrates the limitations of the Caco-2 model to predict BBB permeability of compounds in vivo. The results emphasize the fact that the BBB and the intestinal mucosa are two fundamentally different biologic barriers, and to be able to make accurate predictions about the in vivo CNS penetration of potential drug candidates, it is important that the in vitro model possesses the main characteristics of the in vivo BBB.