Stephanie van Hoppe

53 Brain accumulation of osimertinib and its active metabolite is restricted by ABCB1 & ABCG2 Ab c b1 and Ab c g2 l imi t t he b r a i n a c c umu l a t i on o f o s ime r t i n i b In all three testedmouse strains osimertinib reached its maximumplasma concentration approximately 1 h after oral administration (Supplemental Figure 2). To better assess the separate and combined impact of Abcb1a/1b and Abcg2 on tissue distribution of osimertinib at, or shortly after, peak plasma exposure, we performed a 1.5 hour pharmacokinetic experiment with oral administration of osimertinib (10 mg/kg) tomale WT, Abcb1a/1b -/- , Abcg2 -/- , and Abcb1a/1b;Abcg2 -/- mice. Although, not unexpectedly, the interindividual variation this shortly after oral administration of osimertinib was high, the plasma AUCs of the four strains were very similar, with a C max again occurring around 1 h (Figure 2, Table 2). Only at the 90 min time point the Abcg2 -/- plasma concentrations were significantly lower than those in the other strains, but this was probably related to the substantial experimental variation (Figure 2). These results were generally in line with the 24 h data and suggest that the absence of Abcb1 and Abcg2, alone or combined, has no substantial effect on the plasma exposure of osimertinib in mice during the first hours after administration. Similar to the results at 24 h after administration, the brain concentration of osimertinib in Abcb1a/1b;Abcg2 -/- mice showed a highly significant, 5.1-fold increase (P < 0.001) compared to WT mice (Figure 3A). Additionally, single Abcb1a/1b -/- mice displayed a statistically significant 3.5-fold increase (P < 0.01) compared to WT mice. In contrast, no significant difference was found between the single Abcg2 -/- and WT mice. Correcting the osimertinib brain concentrations for the corresponding plasma concentrations (Figure 3B) or plasma AUCs (Figure 3C) yielded similar results. The brain-to-plasma ratios showed a highly significant, 6.4-fold increase (P < 0.001) for Abcb1a/1b;Abcg2 -/- mice compared toWT mice, and a 4.1-fold increase for Abcb1a/1b -/- mice (P < 0.05), whereas values in the Abcg2 -/- mice were not significantly different from those in theWTmice (Figure 3B; Table 2). Analyzing the same parameters for the liver did not show significant differences between the strains (Figure 3 D-F) except for the Abcg2 - /- liver-to-plasma ratio, which was, however, determined only by the unexpectedly low plasma concentration in this strain at the single 1.5 h time point (Figure 2). Thus, also at 1.5 h, especially Abcb1a/1b could profoundly restrict the brain accumulation of osimertinib, and the combined deficiency for both transporters resulted in a further increased brain penetration of osimertinib. Strikingly, in the absence of Abcb1a/1b and/ or Abcg2, the brain-to-plasma ratios of osimertinib (60 to 90) were even higher than the liver-to-plasma ratios (30 to 70), suggesting a high intrinsic capacity of osimertinib to accumulate in the brain as well as the liver (Figure 3B and E). Indeed, even in WT mice the brain-to-plasma ratio of osimertinib was still relatively high (14.5, Figure 3B, Table 2), illustrating the propensity of this drug to accumulate into the brain.