< Terug naar vorige pagina

Publicatie

Drug-induced cholestasis: in vitro detection and mechanistic insights

Boek - Dissertatie

Drug-induced liver injury (DILI) is the primary cause for the discontinuation of clinical trials and post-marketing withdrawal of drugs, leading to considerable losses for the pharmaceutical industry. Despite the immense financial losses that are related to DILI, its impact on patient healthcare is of much greater importance. Indeed, numerous cases have been reported in literature that illustrate the potentially fatal consequences of DILI. As DILI presents itself in multiple clinical forms (e.g., acute hepatitis, cholestasis and steatosis), the plethora of possible underlying mechanisms is an immense hurdle for the development of next generation prediction tools. Therefore, a better understanding of these mechanisms is an absolute priority for the pharmaceutical industry. During the last decade, both academia and industry have put a great deal of effort into unravelling these underlying mechanisms. Based on this, it has become more and more apparent that the pathological disturbance of bile acid homeostasis plays a key role in the onset of liver injury, making cholestasis one of the major causes of DILI. However, the early detection of drug-induced cholestasis (DIC) remains challenging during the drug development process. Indeed, the majority of currently developed in vitro tools that evaluate DIC are based on the assumption that assessing a drug's potency to interfere with certain hepatobiliary transporters suffices to accurately predict the drug's in vivo cholestatic potential. However, as these tools do not adequately mimic the in vivo situation where various types of cellular responses (both adaptive and adverse) regulate bile acid homeostasis, they are unable to thoroughly assess human DIC risks. This shortcoming can be overcome by using more in vivo-relevant in vitro models such as sandwich-cultured human hepatocytes (SCHH). Indeed, our group recently developed a SCHH-based in vitro assay which allows for the detection of cholestatic compounds based on their ability to modulate bile acid homeostasis. Although the assay is able to distinguish cholestatic compounds from both non-hepatotoxic and non-cholestatic but hepatotoxic compounds, it was not designed to yield mechanistic insights into drug-induced changes in bile acid homeostasis. Such information is extremely valuable, especially when trying to make an informed decision regarding the cholestatic potential of a given drug. Consequently, the goal of the current doctoral thesis was to investigate human DIC by profiling in vitro bile acid disposition in SCHH and by evaluating these data using both non-compartmental analysis (NCA) techniques as well as a more advanced cellular mechanistic bile acid disposition model. In vitro data were obtained using an improved experimental setup that consisted of incubating 10 ╬╝M of the prototypical bile acid chenodeoxycholic acid (CDCA) in absence and presence of therapeutically relevant bosentan concentrations.As cryopreserved human hepatocytes were used throughout this thesis, the aim of the first study was to evaluate whether cryopreserved cells are a viable alternative to freshly-isolated cells. The results from this study clearly showed that cryopreservation does not affect the hepatocyte's biochemical integrity nor its application potential for drug disposition studies. Moreover, transporter studies, which were conducted using fluorescent probes, indicated that the organic anion transporting polypeptide (OATP) and the multidrug resistance-associated protein 2 (MRP2) activity levels remained unaltered following cryopreservation. In the same study, disposition of telmisartan and telmisartan-glucuronide was evaluated using an in-house developed mechanistic cellular disposition model which allowed us to distinguish between the susceptibilities of the individual disposition pathways to cryopreservation. The model predicted that the relative contribution of uptake, metabolism and efflux of telmisartan and telmisartan-glucuronide remained unchanged following cryopreservation, indicating that cryopreserved hepatocytes are a suitable alternative for freshly-isolated hepatocytes. This result also represents an inspiring example of how optimization of hepatocyte cryopreservation protocols can support implementation of the 3R concept for animal experimentation. In addition to assessing the effects of cryopreservation on hepatocyte longevity and functionality, the first study conducted in this doctoral thesis aimed to establish a generic modelling framework that could be used in future studies. Indeed, in a second study, the framework was utilized to develop a mechanistic model that was able to quantitatively evaluate the effects of bosentan on CDCA and GCDCA disposition in SCHH originating from five different human liver tissue donors. The model consisted of seven compartments to which CDCA and GCDCA could distribute using both linear and non-linear kinetics. Model predictions showed that the amidation of CDCA as well as biliary efflux clearance of GCDCA decreased in presence of bosentan, in line with the non-compartmental analysis results and reports from other research groups. Interestingly, not all tested donors were affected by in vitro bosentan treatment, underlining the importance of using an in vivo-relevant in vitro model when assessing human DIC. Indeed, human hepatocytes, that originate from various donors, allow for the evaluation of interindividual differences in susceptibility to cholestatic compounds, something which cannot be achieved with for instance hepatic cell lines as they do not reflect the interindividual variability. In a last study, we aimed to expand on our study regarding bosentan-mediated cholestasis, by pursuing new insights into the role of Ro 47-8634 (O-demethylation of the phenolic methyl ether of bosentan), Ro 48-5033 (hydroxylation at the t-butyl group of bosentan) and Ro 64-1056 (combination of hydroxylation and O-demethylation of bosentan), the three known phase I metabolites of bosentan. To do so, disposition of CDCA and GCDCA in presence of bosentan was evaluated using SCHH originating from three donors which showed distinct capacities in terms of enzymes that play a role in the metabolic pathways of bosentan. As expected, the donor with greatest metabolic capacity showed increased formation of all metabolites as compared to the other donors. However, this was also the only donor which showed a significant decrease in CDCA uptake and its subsequent conjugation to GCDCA following bosentan treatment, suggesting that formation of Ro 47-8634, Ro 48-5033 and/or Ro 64-1056 could drive the observed effects. Indeed, linear regression analysis indicated that inhibition of CDCA's uptake could (at least in part) be attributed to formation of bosentan's metabolites, while inhibition of CDCA conjugation most likely resulted from interaction with the parent.In summary, this doctoral thesis provided new insights into human DIC using an alternative experimental setup that consisted of incubating SCHH with CDCA in presence and absence of therapeutically relevant bosentan concentrations. By quantifying subtle changes in CDCA and GCDCA disposition using NCA techniques and an in-house developed cellular mechanistic disposition model, quantitative insights into the interplay of bosentan's mechanisms of toxicity were gained. More specifically, based on our results, it has become apparent that bosentan's metabolites play a role in the inhibition of bile acid uptake, leading to the prehepatic accumulation of bile acids, while, in addition, bosentan decreases the further conjugation of unconjugated bile acids to their conjugated forms.
Jaar van publicatie:2020