Rapid solid-phase extraction method to quantify [(11)C]-verapamil, and its [(11)C]-metabolites, in human and macaque plasma

Disrupting P-glycoprotein function in clinical settings: what can we learn from the fundamental aspects of this transporter?

P-glycoprotein is one of the most well-studied drug transporters, significant for its role in cancer multiple drug resistance. However, using P-gp inhibitors with the aim of enhancing the therapeutic efficacy of anti-cancer drugs has led to disappointing outcomes. Furthermore, several lead compounds suggested by in vitro and pre-clinical studies have shown variable pharmacokinetics and therapeutic efficacies when applied in the clinical setting. This review will highlight the need to revisit a sound approach to better design and apply P-gp inhibitors in light of safety and efficacy. Challenges confronting the issue hinge upon myriad studies that do not necessarily represent the heterogeneous target population of this therapeutic approach. The application of P-gp modulators has also been complicated by the promiscuous substrate-binding behaviour of P-gp, as well as toxicities related to its intrinsic presence in healthy tissue. This review capitalizes on information spanning genetics, energetics, and pharmacology, bringing to light some fundamental aspects that ought to be reconsidered in order to improve upon and design the next generation of P-gp inhibitors.

Modulation of P-glycoprotein at the Human Blood-Brain Barrier by Quinidine or Rifampin Treatment: A Positron Emission Tomography Imaging Study

Permeability-glycoprotein (P-glycoprotein, P-gp), an efflux transporter at the human blood-brain barrier (BBB), is a significant obstacle to central nervous system (CNS) delivery of P-gp substrate drugs. Using positron emission tomography imaging, we investigated P-gp modulation at the human BBB by an approved P-gp inhibitor, quinidine, or the P-gp inducer, rifampin. Cerebral blood flow (CBF) and BBB P-gp activity were respectively measured by administration of (15)O-water followed by (11)C-verapamil. In a crossover design, healthy volunteers received quinidine and 11-29 days of rifampin treatment during different study periods. CBF and P-gp activity was measured in the absence (control; prior to quinidine treatment) and presence of P-gp modulation. At clinically relevant quinidine plasma concentrations, P-gp inhibition resulted in a 60% increase in (11)C-radioactivity distribution across the human BBB as measured by the brain extraction ratio (ER) of (11)C-radioactivity. Furthermore, the magnitude of BBB P-gp inhibition by quinidine was successfully predicted by a combination of in vitro and macaque data, but not by rat data. Although our findings demonstrated that quinidine did not completely inhibit P-gp at the human BBB, it has the potential to produce clinically significant CNS drug interactions with P-gp substrate drugs that exhibit a narrow therapeutic window and are significantly excluded from the brain by P-gp. Rifampin treatment induced systemic CYP3A metabolism of (11)C-verapamil; however, it reduced the ER by 6%. Therefore, we conclude that rifampin, at its usual clinical dose, cannot be used to induce P-gp at the human BBB to a clinically meaningful extent and is unlikely to cause inadvertent BBB-inductive drug interactions.

[(18)F]Fallypride: metabolism studies and quantification of the radiotracer and its radiometabolites in plasma using a simple and rapid solid-phase extraction method

INTRODUCTION

[(18)F]Fallypride, a fluorinated and substituted benzamide with high affinity for D2/D3 receptors, is a useful PET radioligand for the study of striatal/extrastriatal areas. Since [(18)F]fallypride is extensively metabolized in vivo and since PET examinations are long lasting in humans, the rapid measurement of the unchanged radiotracer in plasma is essential for the quantification of images. The present study aims: i) to evaluate if the radiometabolites of [(18)F]fallypride cross the blood-brain barrier in rodents, ii) to identify these radiometabolites in baboon plasma and iii) to develop a rapid solid phase extraction method (SPE) suitable for human applications to quantify both [(18)F]fallypride and its radiometabolites in plasma.

METHODS

The metabolites P450-dependant in rat and human liver microsomes were characterized by LC-MS-MS and compared to those detected in vivo. Sequential solvent elution on Oasis®-MCX-SPE cartridges was used to quantify [(18)F]fallypride and its radiometabolites.

RESULT

In rat microsomal incubations, five metabolites generated upon N/O-dealkylation or hydroxylation at the pyrrolidine and/or at the benzamide moiety were identified. No radiometabolite was detected in the rat brain. N-dealkylated and hydroxylated derivatives were detected in human microsomal incubations as well as in baboon plasma. The use of SPE (total recovery 100.2%± 2.8%, extraction yield 95.5%± 0.3%) allowed a complete separation of [(18)F]fallypride from its radiometabolites in plasma and evaluate [(18)F]fallypride at 150 min pi to be 22%± 5% of plasma radioactivity.

CONCLUSIONS

The major in vivo radiometabolites of [(18)F]fallypride were produced by N-dealkylation and hydroxylation. Allowing the rapid analysis of multiple plasma samples, SPE is a method of choice for the determination of [(18)F]fallypride until late images required for quantitative PET imaging in humans.

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

Through PET imaging, our laboratory has studied the dynamic biodistribution of (11)C-verapamil, a P-gp substrate, in the nonhuman primate Macaca nemestrina. To gain detailed insight into the kinetics of verapamil transport across the blood-brain barrier (BBB) and the blood-placental barrier (BPB), we analyzed these dynamic biodistribution data by compartmental modeling.

METHODS

Thirteen pregnant macaques (gestational age, 71-159 d; term, ∼172 d) underwent PET imaging with (11)C-verapamil before and during infusion (6, 12, or 24 mg/kg/h) of cyclosporine A (CsA, a P-glycoprotein [P-gp] inhibitor). Dynamic (11)C-verapamil brain or fetal liver (reporter of placental P-gp function) activity was assessed by a 1- or 2-tissue-compartment model.

RESULTS

The 1-tissue-compartment model best explained the observed brain and fetal liver distribution of (11)C-radioactivity. When P-gp was completely inhibited, the brain and fetal liver distribution clearance (K1) approximated tissue blood flow (Q); that is, extraction ratio (K1/Q) was approximately 1, indicating that in the absence of P-gp function, the distribution of (11)C-verapamil radioactivity into these compartments is limited by blood flow. The potency of CsA to inhibit P-gp was tissue-independent (maternal BBB half-maximal inhibitory concentration [IC50], 5.67 ± 1.07 μM, vs. BPB IC50, 7.63 ± 3.16 μM).

CONCLUSION

We propose that on deliberate or inadvertent P-gp inhibition, the upper boundary of increase in human brain (or fetal) distribution of lipophilic drugs such as verapamil will be limited by tissue blood flow. This finding provides a means to predict the magnitude of P-gp-based drug interactions at the BBB and BPB when only the baseline distribution of the drug (i.e., in the absence of P-gp inhibition) across these barriers is available through PET. Our data suggest that P-gp-based drug interactions at the human BBB and BPB can be clinically significant, particularly for those P-gp substrate drugs for which P-gp plays a significant role in excluding the drug from these privileged compartments.

Enantiomeric separation of verapamil and its active metabolite, norverapamil, and simultaneous quantification in human plasma by LC-ESI-MS-MS

A simple, selective and high-throughput liquid chromatography-tandem mass spectrometry method has been developed and validated for the chromatographic separation and quantification of (R)- and (S)-enantiomers of verapamil and its active metabolite, norverapamil, in human plasma. All four analytes along with deuterated internal standards (D(6)-verapamil and D(6)-norverapamil) were extracted from 50 µL human plasma by liquid-liquid extraction. Separation was achieved on a Chiralcel OD-RH (150 × 4.6 mm, 5 µm) analytical column with resolution factors of 1.4 and 1.9 for (R)- and (S)-enantiomers of verapamil and norverapamil, respectively. A mobile phase consisting of 0.05% trifluoroacetic acid in water-acetonitrile (70:30, v/v) afforded capacity factors of 2.45, 3.05, 2.27 and 3.13 for (R)- and (S)-enantiomers of verapamil and norverapamil, respectively. Detection was carried out on a triple quadrupole mass spectrometer, operating in the multiple reaction monitoring and positive ion modes. The method was validated over the concentration range of 1.0-250.0 ng/mL for all four analytes. Absolute recovery for the analytes ranged from 91.1 to 108.1%. Matrix factors calculated at three quality control levels varied from 0.96-1.07. The method was successfully applied to a pharmacokinetic study in 18 healthy Indian males after oral administration of a 240-mg verapamil tablet formulation under fasting conditions.

Removal of trace organics by MBR treatment: the role of molecular properties

This study examined the relationship between specific molecular features of trace organic contaminants and their removal efficiencies by a laboratory scale membrane bioreactor (MBR). Removal efficiencies of 40 trace organic compounds were assessed under stable operating conditions. The reported results demonstrate an apparent correlation between chemical structures and the removal of trace organic contaminants by the laboratory scale MBR system. The removal of all 14 very hydrophobic (Log D > 3.2) trace organic compounds selected in this study was consistently high and was above 85%. The occurrence and types of electron withdrawing or donating functional groups appear to be important factors governing their removal by MBR treatment. In this study, all hydrophilic and moderately hydrophobic (Log D<3.2) compounds possessing strong electron withdrawing functional groups showed removal efficiency of less than 20%. In contrast, high removal efficiencies were observed with most compounds bearing electron donating functional groups such as hydroxyl and primary amine groups. A qualitative framework for the assessment of trace organic removal by MBR treatment was proposed to provide further insights into the removal mechanisms.

Regional P-glycoprotein activity and inhibition at the human blood-brain barrier as imaged by positron emission tomography

We used positron emission tomography (PET) to evaluate the contribution of P-glycoprotein (P-gp), present at the human blood-brain barrier (BBB), to regional drug distribution in the brain. Eleven healthy volunteers underwent PET imaging with [(11)C]-verapamil before and during cyclosporine A infusion. Regional P-gp inhibition was expressed as cyclosporine A-induced percentage change in the distributional clearance of verapamil (K(1)) in the brain, normalized to the regional blood flow (rCBF). K(1) estimates were similar across gray-matter regions of the brain and lower in the white matter regions, but all these estimates were considerably lower than rCBF. Normalization of K(1) by rCBF diminished the differences in estimates related to gray matter and white matter. In contrast, the K(1) for the pituitary, which is situated outside the BBB, approximated the rCBF. The magnitude of P-gp inhibition was comparable across BBB-protected brain structures. Our results indicate that P-gp and its inhibition equally affect the distribution of drugs (and therefore their neuro-efficacy and toxicity) in the various brain regions protected by the BBB.

Positron emission tomography imaging of tissue P-glycoprotein activity during pregnancy in the non-human primate

BACKGROUND AND PURPOSE

Changes in tissue P-glycoprotein (P-gp) activity during pregnancy could affect the pharmacokinetics and thus the efficacy and toxicity of many drugs. Therefore, using positron emission tomography (PET) imaging, we tested whether gestational age affects tissue P-gp activity in the pregnant non-human primate, Macaca nemestrina.

EXPERIMENTAL APPROACH

Mid-gestational (day 75 +/- 13, n= 7) and late-gestational (day 150 +/- 10, n= 5) age macaques were imaged after administration of a prototypic P-gp substrate, (11)C-verapamil (13.7-75.4 MBq.kg(-1)), before and during intravenous infusion of a P-gp inhibitor, cyclosporin A (CsA) (12 or 24 mg.kg(-1).h(-1)). Accumulation of radioactivity in the fetal liver served as a reporter of placental P-gp activity. P-gp activity was expressed as CsA-induced percent change in the ratio of the area (0-9 min) under the (11)C-radioactivity concentration-time curve in the tissue (AUC(tissue)) to that in the maternal plasma (AUC(plasma)).

KEY RESULTS

The CsA-induced change in AUC(fetal liver)/AUC(maternal)(plasma) of (11)C-radioactivity significantly increased from mid- (35 +/- 25%) to late gestation (125 +/- 66%). Likewise, the CsA-induced change in AUC(maternal brain)/AUC(plasma) increased from mid- (172 +/- 80%) to late gestation (337 +/- 148%). The AUC ratio for the other maternal tissues was not significantly affected. Neither the CsA blood concentrations nor the level of circulating (11)C-verapamil metabolites were significantly affected by gestational age.

CONCLUSIONS AND IMPLICATIONS

P-gp activity at the blood-brain barrier and the placental barrier in the macaque increased with gestational age. If replicated in humans, the exposure of the fetus and maternal brain to P-gp substrate drugs, and therefore their efficacy and toxicity, will change during pregnancy.

Simultaneous PET imaging of P-glycoprotein inhibition in multiple tissues in the pregnant nonhuman primate

Studies in rodents indicate that the disruption of P-glycoprotein (P-gp) function increases drug distribution into the developing fetus and organs such as the brain. To simultaneously and serially evaluate the effect of P-gp activity and inhibition on the tissue distribution of drugs in a more representative animal model, we tested the feasibility of conducting whole-body PET of the pregnant nonhuman primate (Macaca nemestrina). We used (11)C-verapamil as the prototypic P-gp substrate and cyclosporine A (CsA) as the prototypic inhibitor.

METHODS

Four pregnant macaques (gestational age, 145-159 d; gestational term, 172 d) were imaged after the intravenous administration of (11)C-verapamil (30-72 MBq/kg) before and during intravenous infusion of CsA (12 or 24 mg/kg/h, n = 2 each). The content of verapamil and its metabolites in plasma samples was determined using a rapid solid-phase extraction method. The plasma and tissue time-radioactivity concentration curves of (11)C were integrated over 0-9 min after each verapamil injection. The tissue or arterial plasma area under the time-concentration curve (AUC(tissue)/AUC(plasma)) served as a measure of the tissue distribution of (11)C radioactivity. CsA effect on (11)C radioactivity distribution was interpreted as P-gp inhibition. The change in the fetal liver AUC ratio served as a reporter of placental P-gp inhibition.

RESULTS

CsA effect on tissue distribution of (11)C radioactivity (AUC ratios) did not increase with the mean blood concentration of CsA, indicating a near-maximal P-gp inhibition. CsA increased maternal brain and fetal liver distribution of (11)C radioactivity by 276% +/- 88% (P < 0.05) and 122% +/- 75% (P < 0.05), respectively. Changes in other measured tissues were not statistically significant.

CONCLUSION

These data demonstrate for the first time, to our knowledge, the feasibility of simultaneous, serial, noninvasive imaging of P-gp activity and inhibition in multiple maternal organs and the placenta in the nonhuman primate. Our findings, consistent with previous data in rodents, indicate that the activity of P-gp in the placenta and the blood-brain barrier is high and that the inhibition of P-gp facilitates drug distribution across these barriers.