An in vitro model of the human alveolus to predict the efficacy of systemic antifungal therapy

Invasive pulmonary aspergillosis is a leading cause of mortality and morbidity in immunocompromised patients for which expanded therapeutic options are urgently required. The development and optimisation of antifungal agents requires the use of tens-of-thousands of laboratory animals.

The objective of the current project is to determine the extent to which a novel in vitro model of the human alveolus, in which murine pharmacokinetics are replicated via the application of micro-pumps, is predictive of the observed antifungal effect in mice. Such a model would represent an extremely powerful translational tool for the optimisation of antifungal therapy in humans.

The in vitro model consists of a cellular bilayer consisting of an upper layer of human alveolar epithelial cells and a lower layer of human pulmonary endothelial cells grown on a semipermeable membrane. The cellular bilayer delineates an upper and lower compartment which are representative of the air-space and pulmonary capillary, respectively. Aspergillus conidia ("spores") are introduced into the alveolar compartment, where they germinate to form hyphae; these constitute the invasive forms, which invade the cellular bilayer. Antifungal drugs can be administered within the endothelial compartment, to mimic systemic drug administration.

In the current proposal, we intend to use micro-pumps to enable the replication of the concentration-time curves of the antifungal agent, amphotericin B. In the first instance, murine pharmacokinetics will be simulated and predictions made as to the likely outcome in mice. The in vitro antifungal effect will be measured using a number of (clinically relevant) biomarkers, including the Aspergillus cell-wall antigen, galactomannan, fungal culture, total RNA and total DNA. The predicted in vitro results will be verified in mice with invasive pulmonary aspergillosis.

If the model is demonstrated to have predictive power, the clinical implications will be sought, by replicating human pharmacokinetics and measuring the antifungal effect in order to predict human dosages which are likely to result in a sub-maximal and near-maximal antifungal effect.

Possible applications of this research include:

1. A reduction in laboratory animal numbers and experiments at all points of the drug development process.
2. A further study of complicated questions, such as combination chemotherapy, which currently require many hundreds of animals.
3. Provision of a valid substitute for laboratory animals which could be used by researchers without access to animal facilities.
4. Demonstration of a novel means by which similar questions could be addressed for a wide range of pulmonary pathogens.