The role of efflux transporters in refractory epilepsy: evaluation with PET tracers

Epilepsy is a neurological disorder characterized by seizures. Epilepsy is usually controlled, but not cured, with antiepileptic medication. However, 30% of epileptic patients do not have seizure control even with the best available antiepileptic drugs. A possible explanation of this form of epilepsy, also called refractory epilepsy, is postulated as the transporter hypothesis. This theory proposes that refractory epilepsy may be the consequence of an overexpression of efflux transport proteins that prevents antiepileptic drugs from penetrating the blood–brain barrier in sufficient concentration. Why patients are resistant to multiple AEDs with distinct mechanisms of action can thereby be explained by this hypothesis. Although several studies already investigated the role of the efflux transporters, P-gp and MRP, the aim of this thesis was visualizing the P-gp and their interaction with antiepileptic drugs with the use of a new PET tracer, 11C-desmethylloperamide. In chapter 1, an overview of the current findings in epilepsy is mentioned, with some greater attention to refractory epilepsy and the recovery strategies. In this chapter, the role of animal models in epilepsy research is also described. The structure of the brain, as well as the transport across the BBB is discussed in chapter 2. Moreover, the structure, localization in tissues, and the known substrates and inhibitors of the efflux transporters are cited, whereby the characteristics of P-gp are highlighted. The role of PET and SPECT in molecular imaging and the practical facts about kinetic modeling is discussed in chapter 3. Chapter 5 gives information about general experimental procedures, while chapter 6 deals with the radiosynthesis and in vivo evaluation of a novel PET tracer, 11C-desmethylloperamide, to visualize P-gp. The tracer was synthesized with a good radiochemical yield and specific activity. Although low brain uptake in wild-type mice was observed, P-gp inhibition with cyclosporine A resulted in significantly increased tracer uptake. Moreover, in P-gp knock-out mice an eight fold higher uptake was observed, suggesting that 11C-desmethylloperamide is an avid tracer to investigated P-gp. Moreover, no metabolisation in the brain occurs, while only one polar metabolite is observed in plasma. Nevertheless, the depletion of P-gp has no effect on the tracer metabolisation pattern. In chapter 7, the first acquirement, which has to be fulfilled in the transporter hypothesis, is investigated. By an indirect method the interaction of phenytoin, levetiracetam, topiramate, and sodium valproate on the P-gp at the BBB are observed. Therapeutic doses of phenytoin, levetiracetam, and topiramate in combination with 11C-desmethylloperamide resulted in brain uptake, which is statistically increased compared to baseline tracer uptake. Higher doses of antiepileptic drugs reduced the brain uptake again to baseline levels. These results confirm the fact that these AED shows interaction with P-gp, while P-gp probably plays no role in the transport of sodium valproate across the BBB, since no different tracer brain uptake is observed with and without antiepileptic drug administrations. Kinetic modeling with 11C-desmethylloperamide to visualize the P-gp mediated transport in the brain is the key topic of chapter 8. First, a kinetic model was accompliced in wild-type, with or without cyclosporine A pretreatment and in P-gp knock-out mice to identify a typical rate constant for P-gp transport. We proposed and evaluated an alternative method to determine the input function, essential for kinetic modeling, by using an 18F-FDG scan to delineate the left heart ventricle. Subsequently, a model, which best fitted the brain data, was determined as well as the rate constants. K1 in wild-type mice is statistical smaller than in P-gp knock-out mice and in cyclosporine A pretreated mice, while the k2 is in the same range in all mice. Since it was expected that K1, representing influx in the brain, should not change between the different groups tested, while k2, indicating efflux out of the brain, was supposed to be lower in P-gp knock-out mice and in the wild-type mice pretreated with cyclosporine A, we postulated a new theory to explain the results. . So, we propose K1 as a pseudo value in mice, representing a combination of passive influx of 11C-dLop through the BBB and a rapid energy dependent output by P-gp, while k2 corresponds to slow passive efflux out of the brain. Since 11C-desmethylloperamide kinetic modeling is useful to predict the presence of absence of functional P-gp, it was postulated that overexpression of P-gp in epileptic rats could be predicted by this non invasive method of kinetic modeling. The kinetic parameters (K1 and k2) obtained from a two-tissue compartment model were statistical different between non epileptic Sprague-Dawley and epileptic rats, when P-gp was partially blocked by cyclosporine A. Higher expression of P-gp, as indicated by immunohistochemical staining, can thus be indicated by the increase of k2 observed in the epileptic rat group. Thereby, kinetic modeling with 11C-desmethylloperamide is a useful non invasive method to evaluate the P-gp functionality in the brain. Chapter 9 focussed on the influence of specific activity of PET tracers on the brain tracer uptake. Our results clearly demonstrate statistical higher brain uptake of 11C-laniquidar when administered as low specific activity solutions, while high specific activity 11C-laniquidar solutions showed lower brain uptake. In comparison, no different 11C-dLop brain uptake was observed after high or low S.A. solutions. These results confirm the important role of specific activity, and subsequently the mass amount of PET tracers when used to investigate P-gp mediated efflux at the BBB.