Globally, hepatocellular carcinoma (HCC) is the most common form of liver cancer and a leading cause of cancer death. One important risk factor is liver cirrhosis and the disease progression is characterized by deposition of extracellular matrix proteins that form a fibrous network, which increases liver stiffness and may limit the effectiveness of different treatment strategies. 

The overall aim of this thesis was to investigate the anticancer drug doxorubicin (DOX) and its clinically relevant drug delivery systems from an in vitro perspective. The focus was on developing and using qualitative and quantitative methods to better understand formulation performance, intracellular uptake and molecular diffusion. The experimental in vitro findings were then translated to clinical scenarios using physiologically based pharmacokinetic (PBPK) modelling.

The performance of clinically employed emulsion formulations containing DOX and the tumor accumulating oil Lipiodol® were evaluated in terms of their stability (Paper I). The most stable emulsion (> 72 h) was achieved when using an aqueous phase containing the contrast agent iohexol and with an aqueous to lipid phase ratio of 1:4 to assure formation of a water-in-oil emulsion.  This was followed by a cell-based study (Paper II) where nanoformulated DOX was compared to DOX in solution in terms of tumor cell toxicity, intracellular DOX uptake and intracellular formation of the main active metabolite doxorubicinol (DOXol). DOX in solution was more potent in all investigated cell lines, where the most sensitive cells (HepG2) displayed IC50 values that were approximately 100 times lower than the most resistant cell line (SNU449). This was explained by the rapid intracellular uptake in HepG2 cells which was also confirmed with a complimentary miniaturized chip technique in Paper IV.  In papers III and IV the focus was on molecular diffusion across biomimetic hydrogels mimicking tissue properties of cirrhotic liver and early stage HCC. The diffusion of DOX was significantly reduced in biomimetic gels as compared with more commonly used agarose gels, however the presence of human liver tumor cells did not significantly influence diffusion. Simulations using a developed PBPK and spatio-temporal tissue concentration model suggested that a liver tumor resembling SNU449 cells would not reach therapeutic exposure levels in a clinical scenario while the diffusion of DOX required further reduction by the tumor extracellular matrix in order to generate tumor concentration-time curves consistent with in vivo observations.

This thesis contributes to an increased understanding of using DOX and its drug delivery systems as a treatment option for HCC. The approach of translating in vitro experimental data to clinical scenarios using modelling will grow in relevance as methods become more complex and data more bio-relevant.