PULMONARY DRUG DELIVERY: IN VITRO PREDICTION OF LUNG ABSORPTION

Within the past few decades, the lung has received increasing attention as a target for local and systemic drug delivery (1, 2). Drug administration via the respiratory tract appears promising because of the peculiar anatomical and physiological characteristics of the lung: large surface area, thin alveolar epithelium, high vascularization and relatively low level of metabolic enzymatic activity (3, 4). The optimal absorption characteristics of a pulmonary drug depend on the site of action. For systemically acting drugs, absorption from the lung determines the therapeutic effect profile (onset, intensity and duration of action). On the contrary, for locally acting drugs, the absorption process may determine the removal and consequently the termination of drug action in the lung, as well as the onset of any systemically mediated adverse effects. On the other hand, a very high lung retention is not targeted due to accumulation/toxicity issue (5, 6). Therefore, when designing drugs for pulmonary delivery, it is important to consider both lung-tissue retention and permeability in order to find the right balance between these two parameters (6). In vitro cell culture models of absorptive epithelia could be helpful and useful in early phase of drug discovery for the prediction of in vivodrug permeability and drug absorption.Advantages of cell culture models based on continuous cell lines are well demonstrated by the large use of the Caco-2 model of the gastrointestinal epithelium so that Caco-2 cells are now definitely recognized as the gold standard for permeability and transport studies of orally administered drug (7, 8). However, at present there is no lung equivalent of the Caco-2 cell line to serve as well-established and reliable in vitro model of the respiratory epithelium (9, 10). For this reason, in respiratory delivery, there is a need to have a functionally relevant, continuous and robust in vitro epithelial cell modelto be used as a permeability screening tool that is predictive of in vivo lung absorption and to establish reliable In vitro-In vivo (IVIV) correlation to guide drug discovery programs (7). Recently, the continuously growing bronchial epithelial cells, Calu-3, are quite often chosen as a model for the pulmonary epithelial barrier, so, in the PhD project, foremost cell culture model was set up and validated to demonstrate that in our hand, air- interfaced cell layers exhibit morphological and bioelectrical characteristics proper of the native epithelium. Once validated, Calu-3 cell line was used to assess the permeability (Papp) of a set of 8 locally acting drugs of varying molecular size, charge, lipophilicity and polar surface area. The permeability of these compounds was determined also in the well-recognized Caco-2 cells absorption model. Thus, the in vitro permeability values both from Caco-2 and Calu-3, were correlated with several in vivo pharmacokinetic (PK) parameters obtained after an intratracheal administration in rats to find out if an IVIV correlation could be established. Overall, the data demonstrate that Calu-3 cells exhibit many of the features of primary cells, forming a uniform layer of polarized, well-differentiated cells with mucus secretion, functional tight-junction and important transporter proteins, confirming its potential use as a model to assess lung permeability. Moreover, when Papp values determined both in Calu-3 and Caco-2 cell lines, were compared to in vivo absorption data in rats, a strong and reliable IVIV correlation was established with three in vivo PK parameters: lung drug half-life (T ½ ), lung mean residence time (MRT), and the percentage of drug retained in the lung at 24h after dose. Commonly, permeability data from cell culture, offer a convenient reproducible and quantifiable way to evaluate the absorption potential of a compound, leading to a rank order that highlights qualitatively those compounds that show favourable permeability characteristics. In the present study, it was demonstrated how Calu-3 and Caco-2 permeability data can be applied to quantitatively predict the extent of absorption in vivo. In future, in vitro Papp couldbe used to guide drug discovery programs allowing prediction and estimation of lung kinetic of inhaled drug. To deep analyse the predictive power of Caco-2 and Calu-3 cell lines for permeability studies in pulmonary drug delivery, the two cell models were evaluated and compared to each other through a novel method of in vivo prediction named physiologically-based pharmacokinetic (PBPK) modelling. To this aim, routine ADME ( Absorption, Distribution, Metabolism and Excretion) in vitro assays as well as non-standard in vitro assays were performed in order to provide all the in vitro experimental data and input information to fill in the PBPK model built in rat preclinical specie. As in vitro permeability data, both Caco-2 and Calu-3 Papp values were used to find out if a better simulation was achieved with one of the two cell lines. For all analysed compounds, the concentration-time curves in lungand plasma, generated starting from Calu-3 Papp values, were more close the actual PK profile. In the same way, also the estimation of key lung PK parameters was better achieved with Calu-3, since the ratio between the observed versus simulated parameter is often more proximal to unity in comparison to Caco-2 simulation. Overall these results showed that Calu-3 cells line used in a PBPK modelling gave the right and proper input parameters to better describe the absorption process of pulmonary drug, leading for some compounds, to an improvement in prediction in comparison to Caco-2 and even better, for other compounds, making predictionpossible. In future, application of PBPK modeling used in conjunction with IVIV correlation approach can provide a useful tool to study pulmonary drug disposition and enhance understanding of drug transport in the lung