Inaugural Asian Drug-Drug Interactions (DDI) conference ADDI 2018 was held on December 3-6 in the Chinese University of Hong Kong. The event was especially interesting, as PXB-Mouse® and PXB-cells® can support fruitful research in many of discussed areas and can help in achieving research goals in predicting clinically relevant DDI.
For example, it was pointed that BSEP down-regulation may impact drug-induced liver injury (DILI). If reader recalls, in 2012 Dr. Foster (AstraZeneca team at that time) published the results of the study of troglitazone. BSEP and MRP2 were down-regulated as the result of treatment, which, most likely, was the reason for human liver toxicity. So, it is not surprising that the inhibition of efflux transporters can cause DILI, and the PXB-Mouse® studies may help in predicting human-specific response to xenobiotics.
Another important message brought by the speaker was the confirmation of tight connection between metabolism and toxicity in drug development. It is obvious that sometimes toxicity is caused by the reactive metabolite formation. On the other hand, metabolism studies (both in vitro and in vivo) may be impacted by the toxicity of the compound. It does make sense when pharmaceutical companies re-design their pre-clinical development groups to bring drug metabolism and toxicity divisions under the same umbrella.
The slide with human radio-labeled mass balance study results triggered the discussion that it is very challenging to predict human metabolites without actual human radio-labeled mass balance study. In the meantime, it has been shown by Igawa et al. in 2014 that the radio-labeled mass balance study with PXB-Mouse® can help in predicting human metabolites prior to first-in-human studies.
On the second day of the conference one of the presentations tried to investigate the factors involved into variability of IC50 values in PBPK modeling. The following results were revealed: experimental systems (cells/cell lines, recommended by FDA and EMA for DDI evaluation) end up with ~6-fold difference between systems used; experimental conditions result in ~3-fold difference and experimental analysis gives up to 10-fold variability in calculated IC50. Thus, uniformity is desirable, but barely achievable at this point.
It was stressed, that DDI needs human-relevant models. Current methodologies in DDI research include various assays (uptake assay, metabolism assay, transporter assay, etc.), which makes it difficult to see the whole picture and provide clinically relevant conclusions.
Great industry-originated presentation on the third day of the conference discussed an importance of reaction phenotyping and associated challenges. It was highlighted that the low clearance compounds are the most difficult for the reaction phenotyping, and most of the tools used in drug development process for these leads have their own limitations. Additional complication in phenotyping is associated with the non-CYP metabolized compounds (i.e., AO-metabolized compounds). It was concluded that low clearance and non-CYP metabolized compounds cannot be characterized using less physiologically relevant systems (like human liver microsomes) and require the use of more physiologically relevant approach (like human hepatocytes). And again, the combination if PXB-cells® and PXB-Mouse® prior to first-in-human stage may help in bridging the gap between less physiologically relevant systems (like in vivo studies with rodents) and human observations in clinic.
It was pleasant to hear presentation from Astellas (Japan) showing the possibility to improve prediction of clinical DDI using chimeric mouse with humanized liver (PXB-Mouse®). The combination of retrospective and perspective approaches helped to confirm biliary excretion of compound in study.
In addition to drug-drug interactions, vitamins and herbs were brought to the discussion. It was pointed that drug-induced nutrient-vitamin deficiency is underappreciated.
ADDI 2018 closing session was devoted to herb-drug interactions. Clinical evidence shows that it is important to consider potential interactions between herbs and drugs, as the change in systemic exposure may lead to the treatment failure or toxicity issues.
To summarize all the above, the integration of multiple mechanisms to predict DILI remains challenging; and the lack of relevant models, capable to bridge the gap between pre-clinical studies and clinical observation, makes it difficult to reliably predict DDI. Better familiarity of DDI researchers with the chimeric mice with humanized liver and its features may help to streamline DDI research.