Stephanie van Hoppe

115 Brain accumulation of ponatinib and its active metabolite is limited by ABCB1 and ABCG2 membrane of hepatocytes. The disproportionate increase in brain accumulation of ponatinib (and of its metabolite DMP, Supplementary Figures 3 and 4) observed in the Abcb1a/1b;Abcg2 -/- mice compared to single knockout strains was similar to that found for a range of other TKIs (e.g. axitinib, vemurafenib, lapatinib, gefitinib, erlotinib, sunitinib, dasatinib, regorafenib and imatinib [29, 35, 38-43]), and can be explained by relatively straightforward pharmacokinetic models [40, 44]. These models indicate that if two transporters each have a high contribution of efflux transport relative to the background efflux at the BBB in the absence of both transporters, the effect of single transporter ablation on brain accumulation will be far less than the effect of combined ablation. An interesting implication of these models and our findings on ponatinib (and DMP) brain accumulation is that next to Abcg2 and Abcb1a/1b, there can be no other efflux transporters in the BBB that are remotely as efficient in keeping ponatinib and DMP out of the brain. The increase in treatment options and efficacy for patients with leukemia due to the advent of modern targeted drugs has increased the chance for isolated metastases behind the BBB to emerge, as effective control of peripheral malignancies greatly increases patient survival times. Malignancies in the brain are often hard to reach for drugs because of the presence of ABCB1 and ABCG2 in the continuous physical barrier between blood and brain tissue formed by the BBB. Pharmacological inhibition of these transporters during pharmacotherapy could therefore also improve treatment of metastases positioned in part or in whole behind the BBB. However, it should be noted that studies in mice can only provide qualitative information on the possible impact of these transporters in the blood-brain barrier of patients. In order to better assess the impact of such processes in patients, one can consider appropriate physiologically based pharmacokinetic modeling, or clinical positron-emission tomography (PET) studies with the drug of interest, in combination with highly efficacious ABCB1 and ABCG2 inhibitors, such as for instance elacridar [45-47]. Next to treatment of (metastatic) CML and ALL, a study by Whittle et al. has shown that ponatinib could potentially also be used for neuroblastoma therapy in vivo [48]. Also here pharmacological inhibition of the ABC efflux transporters in the BBB might possibly increase the therapeutic efficacy. Our finding that Cyp3a deficiency in mice increases ponatinib oral availability (Figure 2A, Table 1) is in line with a significant role of CYP3A in metabolizing ponatinib. The fact that we did not observe obvious changes in the pharmacokinetics of DMP, one of its CYP3A-generated metabolites, could be explained by the possibility that other mouse enzymes, for instance one or more of the many mouse Cyp2c isoforms, might form this metabolite as well [49]. Additionally, it may be possible that CYP3A further metabolizes DMP. In this context it is worth noting that patients receiving ponatinib and the CYP3A inhibitor ketoconazole showed both an increase in plasma ponatinib