Stephanie van Hoppe

30 Chapter 2 Abcb1a/1b -/- mice (Supplemental Fig. 4). This indicates that especially Abcg2 but also Abcb1a/1b can contribute to the plasma clearance of afatinib in mice. Ab c g2 and Ab c b1 e a c h s t r ong l y l imi t a f a t i n i b b r a i n a c c umu l a t i on Tissue concentrations were measured at the last time point of the above-mentioned pharmacokinetic experiments. At 24 h, brain concentrations in Abcb1a/1b;Abcg2 -/- mice showed highly significant increases of 1208-fold (P < 0.001) compared toWT mice (Fig. 3A; Table I). The Abcg2 -/- and Abcb1a/1b -/- separate knockout strains also showed a significant increase compared towild typemice, although far lower, with increases of 4.6- and 3.2-fold, respectively (P < 0.001). However, as the afatinib plasma concentrations at 24 h were also very different between the strains, this required appropriate correction. Unfortunately, plasma concentrations at 24 h in the WT mice were undetectable, precluding a direct comparison with afatinib brain-to-plasma ratios in theWT mice (Fig. 3B). Correcting the afatinib brain concentrations for the corresponding plasma AUCs resulted in no significant differences in afatinib accumulation in the brains of Abcg2 -/- and Abcb1a/1b -/- mice compared to WT mice (Fig. 3C). However, for Abcb1a/1b;Abcg2 -/- mice the brain accumulation was 175-fold increased (P < 0.001) compared to that inWT mice, with similar differences observed when compared with Abcg2 -/- and Abcb1a/1b -/- mice (Fig. 3C; Table I). This indicates that Abcg2 and Abcb1a/1b each by themselves can drastically restrict the brain accumulation of afatinib, and that a combined deficiency of these transporters is needed for a pronounced increase of the brain accumulation of afatinib. The liver often equilibrates quickly with plasma levels of a drug, which makes it complicated to assess differences between strains when plasma levels are widely divergent (and undetectable in the WT strain), as is the case at 24 h. As shown in Table I and Fig. 3D, after oral administration, Abcg2 -/- mice showed a 29-fold increase (P < 0.001) in liver afatinib concentrations compared to WT mice, whereas Abcb1a/1b -/- mice did not show a significant difference compared to the WT mice. With both Abcb1a/1b and Abcg2 knocked out, we observed a more pronounced increase in liver concentration of afatinib compared to WT mice (286-fold, Table I, Fig. 3D). However, these pronounced differencesmaywell mainly reflect the large differences in plasma afatinib concentration at this time point, as is suggested by the much smaller differences in liver-to-plasma ratios between the strains as far as these could be assessed (Fig. 3E). Rapid equilibration of liver afatinib with plasma afatinib concentrations renders interpretation of a correction for the plasma AUCs (Fig. 3F) problematic, as the strain differences in plasma concentration at 24 h were far greater than the strain differences in plasma AUCs. Similar complications preclude a straightforward interpretation of kidney and spleen parameters at 24 h (Supplemental Figure 5). Such complications can be circumvented