Stephanie van Hoppe

169 Conclusions and future perspectives In this thesis we investigated the pharmacological functions of ABC efflux transporters in vitro , by using transwell transport assays and in vivo , by using the single and combination knockout mouse models of ABCB1 and ABCG2. We demonstrated the usefulness of these mouse models to study the impact of ABCB1 and ABCG2 on oral availability of several transported targeted anticancer drugs and their organ exposure – especially for the brain. We have shown that dual deficiency of ABCB1 and ABCG2 most effectively increases the brain penetration of the studied tyrosine kinase inhibitors (TKIs) and, when applicable, their active metabolite(s). We further used CYP3A knockout and humanized transgenic mouse models to assess the in vivo impact of this multispecific drug-metabolizing complex on the study drugs. The findings in this thesis suggest that tumors expressing ABCB1 and/or ABCG2 may demonstrate resistance to afatinib-, osimertinib-, ibrutinib- and ponatinib-based cancer therapy. Inhibiting both these ABC transporters during therapy might therefore be beneficial for the response of these tumors. It may further be possible to increase the levels of the afore-mentioned TKIs in the brain of patients with CNS involvement for better treatment efficacy by co-administration of an efficacious dual inhibitor of ABCB1 and ABCG2 such as elacridar. From our studies it appears that this sort of co- administration might perhaps also increase the oral availability and consequently overall tissue levels of afatinib, which should be taken into account for possible dose adjustment, to avoid increased toxicity. The oral availability of osimertinib, ibrutinib or osimertinib, however, does not seem to be noticeably affected by the activity of ABCB1 or ABCG2. While afatinib is barely metabolized, osimertinib, ibrutinib and ponatinib are all metabolized into a subset of active and inactive metabolites. Their metabolism by CYP3A and possibly other CYPs in theintestinesandliverismorelikelytheprimaryfactorinrestrictingtheoralbioavailability of these TKIs. Therefore, the oral availability of these TKIs could potentially be enhanced by co-administration of a CYP3A4 inhibitor, such as ketoconazole or ritonavir. However, caution should be exercised in such approaches to prevent unexpected toxicity. Our results suggest that co-administration of ABCG2 and ABCB1 inhibitors may be considered to increase exposure of afatinib, osimertinib, ibrutinib and ponatinib in patients, especially in the brain, thus providing an option to better treat, amongst others, NSCLC and its metastases positioned in part or in whole behind a functionally intact blood-brain barrier. However, we feel that the best option for patients with brain (micro-) metastaseswouldbe if it were possible to treat themwithdrugs that are not substantially transported by these transporters. As few of such drugs are as yet available, it can still be considered for the current generation of TKIs to try and carefully inhibit ABCB1 and ABCG2 with a coadministered inhibitor, as this may give these patients a better chance of survival. However, the considerable risk of completely inhibiting ABCB1 and ABCG2 in