Stephanie van Hoppe

51 Brain accumulation of osimertinib and its active metabolite is restricted by ABCB1 & ABCG2 No s ub s t an t i a l e f f e c t o f Ab c b1 and Ab c g2 , o r Cy p3a on p l a sma pha rma c o k i ne t i c s o f o r a l o s ime r t i n i b In view of the in vitro transport results, we studied the impact of Abcb1 and Abcg2 on the plasma and tissue pharmacokinetics of osimertinib in a pilot experiment in male wild-type (WT) and Abcb1a/1b;Abcg2 -/- mice. In addition, although metabolism of osimertinib in mice appears to be primarily mediated by mouse Cyp2d proteins and not by Cyp3a as in humans [44], we included Cyp3a -/- mice in this pilot. We orally administered osimertinib to the mice at a dosage of 10 mg/kg, physiologically roughly equivalent to the recommended human dose (80 mg oral, once daily). We analyzed the plasma concentrations of osimertinib over 24 h. As shown in Supplemental Figure 2 and Table 1, we found a 1.4-fold higher plasma AUC 0-24h in Abcb1a/1b;Abcg2 -/- mice than in WT mice, but this was not statistically significant. Also the absence of Cyp3a enzymes did not cause a statistically significant shift compared to the WT mice in the plasma exposure between 0 and 24 h (Supplemental Figure 2, Table 1). We next assessed the tissue distribution of osimertinib at 24 h. As the plasma level of osimertinib at this time point was below the lower limit of detection in all strains, we could not directly calculate tissue-to-plasma ratios, but we could calculate the relative tissue accumulation (P) by correcting for the plasma AUC 0-24h . Interestingly, the tissue concentrations in brain but also liver of Abcb1a/1b;Abcg2 -/- mice were markedly higher than those in WT or Cyp3a -/- mice, and this also applied when assessing the tissue accumulations (Table 1). By far the strongest effect was seen in brain. Although brain concentrations in WT mice were below the detection limit, and experimental variation was high, for instance the average brain concentration of osimertinib compared to that in liver was nearly 10-fold higher in Abcb1a/1b;Abcg2 -/- mice (0.82) than in Cyp3a -/- mice (0.083). This parameter for other tissues such as spleen and kidney was much less affected (Supplemental Figure 3). These data suggest that, at this late distribution phase, Abcb1a/1b and/or Abcg2 play a major role in limiting osimertinib concentrations in the brain, and a smaller role in limiting the concentrations in liver and spleen, in the latter cases perhaps by mediating late elimination from these organs (Table 1 and Supplemental Figure 3). As for Cyp3a -/- mice, there were no tissue parameters significantly different from WT values at 24 h, but the C brain and P brain were markedly lower than in the Abcb1a/1b;Abcg2 -/- mice (Table 1, Supplemental Figure 3).