Stephanie van Hoppe

58 Chapter 3 Our results show that osimertinib is transported by hABCB1 and efficiently by mAbcg2, but not detectably by hABCG2 in vitro, and that this transport can be inhibited with specific inhibitors. A previous report demonstrated that osimertinib at low concentration (00.4 µM) can also significantly reverse hABCB1 and hABCG2 mediated multidrug resistance (MDR) via inhibiting their efflux activity in vitro [48], illustrating the various interactions of osimertinib with these transporters. In spite of the clear transport in vitro , in vivo we did not observe an obvious limiting effect of Abcb1a/1b or Abcg2 on the oral availability of osimertinib in mice. However, the accumulation of osimertinib in the brain was markedly restricted by Abcb1a/1b and Abcg2. The brain distribution of osimertinib was clearly increased by absence of the combination of Abcb1a/1b and Abcg2 in the BBB (6.4-fold compared to WT), but not substantially by absence of Abcg2 alone. This is in contrast to Abcb1a/1b, which by itself showed a clear limiting effect (by 4.1-fold) on osimertinib accumulation in the brain (Figure 3, Table 2). These data suggest that the brain penetration of osimertinib could be further enhanced by effectively inhibiting ABCB1 and ABCG activity in the BBB, for instance by coadministration of the dual ABCB1 and ABCG2 inhibitor elacridar with osimertinib. In this context it is worth noting that we did not observe any indication that the increased brain penetration of osimertinib in Abcb1a/1b;Abcg2 -/- mice resulted in noticeable toxicity. This is in contrast to the TKI brigatinib, which caused lethal toxicity in Abcb1a/1b;Abcg2 -/- mice, whereas WT mice were completely unaffected by the same oral dose of brigatinib [45]. This suggests that it may potentially be safe enough to boost osimertinib brain accumulation using ABCB1 and ABCG2 inhibitors, although obviously this will always first need to be carefully tested in appropriately designed clinical trials. The brain penetration of the most active metabolite of osimertinib, AZ5104, was also strongly limited by Abcb1a/1b activity in the BBB, and more notably so when Abcg2 was additionally deficient (Figure 5). This suggests that AZ5104 is similarly affected by the ABC transporters at the BBB as its parental compound. In contrast, the distribution of osimertinib and AZ5104 to the liver was not markedly affected by these efflux transporters (Figures 3 and 5). The difference we observed between a high impact on brain accumulation versus no significant impact on oral availability of osimertinib is a common observation for various other shared Abcb1a/1b and Abcg2 substrates such as sunitinib, sorafenib, imatinib, and gefitinib (27, 48-50). We have observed that when a drug is only modestly transported by ABCB1 and/or ABCG2 in vitro , we generally see a much more outspoken effect of these transporters in limiting the brain accumulation of this drug, than in reducing its oral availability (51, 52). Only drugs that are very efficiently transported in vitro , like afatinib, tend to show a clear role of the transporters in restricting their oral