Modeling Cyclosporine A Inhibition of the Distribution of a P-Glycoprotein PET Ligand, 11C-Verapamil, into the Maternal Brain and Fetal Liver of the Pregnant Nonhuman Primate: Impact of Tissue Blood Flow and Site of Inhibition

The multidrug-resistance protein P-glycoprotein (P-gp) is considered, on the basis of rodent studies, to be the most important efflux transporter at the blood–brain barrier (BBB) because of its high level of expression at the luminal membrane in brain capillary endothelial cells and its ability to exclude a wide variety of drugs and endogenous substances from the central nervous system (1). Drug removal protects the central nervous system from potential neurotoxic effects but also prevents effective pharmacotherapy of neurologic diseases. These rodent studies have also shown that P-gp is highly expressed at the blood–placental barrier (BPB), where it limits drug delivery to the fetus (e.g., HIV protease inhibitors) (2). Alternatively, the barrier may protect the fetus from toxicity from maternal cancer chemotherapy (3). Although these rodent studies have demonstrated significant P-gp activity at the BBB and BPB, it is not clear whether the same magnitude of activity is present at the human BBB and BPB. We and others have begun to address this question through measurement of P-gp activity at the human BBB, in the presence and absence of P-gp inhibitors, using various P-gp PET ligands (e.g., 11C-verapamil) and inhibitors (e.g., cyclosporine A [CsA]; tariquidar) (4–7). However, for ethical reasons, it is not possible to conduct such studies to determine the magnitude of P-gp activity at the human BPB. Moreover, complete inhibition of P-gp at the human BBB (and possibly BPB) with the prototypic P-gp inhibitor, CsA, is not possible because of the potential toxicity of this inhibitor when administered at doses necessary to produce such an effect (8). Therefore, we conducted 11C-verapamil dynamic biodistribution studies in a representative animal model, the nonhuman primate Macaca nemestrina, in which it is possible to administer doses of CsA that can completely inhibit P-gp. We published the nonparametric analysis of the results of these whole-body PET studies and showed an increased distribution (as measured by the ratio of the areas under the concentration–time curve [AUCs] for tissue and plasma) of 11C-radioactivity across the BBB and the BPB, with minimal or no changes in distribution of 11C-radioactivity into other organs such as the maternal liver, spleen, and kidneys (9,10). To gain detailed insight into the transport kinetics of 11C-verapamil radioactivity into macaque tissues with and without P-gp modulation, we analyzed these macaque data using compartmental modeling. Our human PET study examining inhibition of P-gp at the BBB using 11C-verapamil as a model P-gp substrate and CsA as a model P-gp inhibitor showed that inhibition of P-gp increased the plasma (or blood)-to-brain distribution clearance (K1) of 11C-verapamil radioactivity rather than the efflux rate constant k2 (6). Moreover, by examining regional P-gp activity at the BBB with 11C-verapamil and CsA, we showed that in the absence of functional P-gp activity (e.g., pituitary), the K1 of lipophilic P-gp ligands (such as verapamil) into the brain is limited by delivery (i.e., by regional cerebral blood flow) (Q) (7). Hence, we proposed that the extraction ratio (ER = K1/Q) is a better index of P-gp activity than is K1. In the presence of P-gp, the brain ER of lipophilic ligands such as 11C-verapamil can be low, and therefore the blood-flow dependence of K1 into the tissue may not be apparent. However, the ER of these tracers can increase substantially once P-gp activity is inhibited, and consequently, the delivery of P-gp substrates can become perfusion-dependent (11). Our ability to completely inhibit P-gp by CsA in the macaque provided us an opportunity to test the hypothesis that, as is the case in humans, inhibition of P-gp at both the BBB and BPB in the macaque is reported by 11C-verapamil K1 (and ER), not k2, and that the upper boundary of K1 of a lipophilic P-gp ligand (e.g., 11C-verapamil) is limited by tissue blood flow. Several groups have shown that inhibition of P-gp may be tissue-dependent (12). On the basis of these data, we also asked whether CsA inhibition of P-gp at the BBB and the BPB in the macaque is tissue-dependent. Therefore, the goals of our investigation were 3-fold: first, to confirm, through compartmental modeling, that inhibition of P-gp at both the BBB and BPB in the macaque is reported by 11C-verapamil K1 and ER and not k2; second, to address whether the magnitude of K1 of a lipophilic drug (e.g., verapamil), in the absence of P-gp function, is limited by tissue blood flow; and third, to determine whether the in vivo potency (half-maximal [50%] inhibitory concentration, or IC50) of P-gp inhibition by CsA at both the BBB and BPB is tissue-dependent.