Cancer on-chip: Disease models for testing microbubble-mediated drug delivery

The effective delivery of drugs to tumours remains a major challenge in the successful treatment of cancer 1. Whilst many of the new therapeutic treatments developed show promising results when tested in vitro, such observations are rarely translated to clinical trials. This results in upwards of 97 % of anti-cancer therapeutics failing to receive approval for clinical use 2. One of the primary contributors to this issue is the inability of the simplistic 2D cancer cell in vitro models to effectively recreate the in vivo microenvironment of tumours. In recent years, it has become increasingly evident that the combined use of multicellular 3D cell cultures and microfluidics offers a means of producing organoid systems which closely mimic tumours – overcoming the limitations of 2D models. Indeed, several experiments conducted with pre-existing anti-cancer compounds using such systems have observed responses similar to that of in vivo observations, demonstrating their potential utility 3,4. Whilst many of the anti-cancer therapeutics successfully developed are capable of killing tumour cells, several factors limit their efficacy. Systemic side effects, insufficient drug accumulation and drug-resistant tumour cells all impact the ability of therapeutics to prevent tumour progression 5. The use of targeted, drug-loaded nanoparticles has previously been proposed as a means of delivering drugs exclusively to the tumour site, allowing for the localised accumulation of high drug concentrations whilst also minimising systemic side effects 6,7. Compounds such as Doxil, which consist of Doxorubicin (DOX) encapsulated in liposomes, have been successful in reducing systemic side effects and improving the overall clinical outcome after treatment 8,9. However, there still remains significant room for further improvements and it is evident that further methods of drug delivery must be developed. Microbubbles (MBs) offer a potential means of improving drug delivery to tumours through a mechanism known as sonoporation. Due to their size and compressibility, gas-filled MBs oscillate when exposed to ultrasound (US) which matches their resonant frequency. Under high acoustic pressures, MBs undergo inertial cavitation where they expand and contract rapidly, producing locally high pressures and temperatures. If this process occurs sufficiently close to a cell membrane, high shear stresses are exerted on the membrane which leads to the production of pores - this process is known as sonoporation. The opening of pores has been shown to allow for the increased passage of therapeutics into cells, presenting a method of enhancing drug delivery to tumours 10–13. Whilst several studies have investigated the use of MBs, many of these studies have been conducted on basic 2D cell models known to be inadequate in reproducing in vivo responses to treatments. This project sought to utilise the recent advances in the organ-on-chip field to further the pre-clinical assessment of MB-mediated drug delivery. Experiments conducted throughout this project present the first instance of MB-mediated therapeutic delivery being tested on novel organ-on-chip systems. Two microfluidic organ-on-chip systems were developed in this project. The first system consists of a series of microfluidic traps which capture solid 3D tumours known as spheroids. Both free and liposomal DOX was co-delivered with MBs under constant flow conditions. Dead cell imaging and analysis of spheroid viability revealed that MBs allowed for increased accumulation of DOX and a greater reduction in spheroid viability when co-delivered with either drug formulation. As MBs would be administered intravenously, the first barrier that must be overcome in the tumour-associated vasculature. The second system developed therefore consisted of a fully perfusable vasculature network which, through tumour media conditioning, mimicked tumour-associated vasculature. This system was used to evaluate the use of MBs in enhancing the accumulation of liposomes (LSs) in tumour tissues. LSs were targeted to a membrane integrin protein found to be upregulated in tumour vasculature. Results found that the use of targeting and MBs greatly improved LS accumulation in tumour vasculature compared to healthy vasculature. This provided evidence that the two-fold localised delivery mechanism of targeting and MBs is a promising method by which localised drug delivery to tumours can be enhanced