Abstract
PET imaging using radiolabeled avid substrates of the ATP-binding cassette (ABC) transporter P-glycoprotein (ABCB1) has convincingly revealed the role of this major efflux transporter in limiting the influx of its substrates from blood into the brain across the blood–brain barrier (BBB). Many drugs, such as metoclopramide, are weak ABCB1 substrates and distribute into the brain even when ABCB1 is fully functional. In this study, we used kinetic modeling and validated simplified methods to highlight and quantify the impact of ABCB1 on the BBB influx and efflux of 11C-metoclopramide, as a model of a weak ABCB1 substrate, in nonhuman primates. Methods: The regional brain kinetics of a tracer dose of 11C-metoclopramide (298 ± 44 MBq) were assessed in baboons using PET without (n = 4) or with (n = 4) intravenous coinfusion of the ABCB1 inhibitor tariquidar (4 mg/kg/h). Metabolite-corrected arterial input functions were generated to estimate the regional volume of distribution (VT), as well as the influx (K1) and efflux (k2) rate constants, using a 1-tissue-compartment model. Modeling outcome parameters were correlated with image-derived parameters, that is, areas under the regional time–activity curves (AUCs) from 0 to 30 min and from 30 to 60 min (SUV⋅min) and the elimination slope (kE; min−1) from 30 to 60 min. Results: Tariquidar significantly increased the brain distribution of 11C-metoclopramide (VT = 4.3 ± 0.5 mL/cm3 and 8.7 ± 0.5 mL/cm3 for baseline and ABCB1 inhibition conditions, respectively, P < 0.001), with a 1.28-fold increase in K1 (P < 0.05) and a 1.64-fold decrease in k2 (P < 0.001). The effect of tariquidar was homogeneous across different brain regions. The parameters most sensitive to ABCB1 inhibition were VT (2.02-fold increase) and AUC from 30 to 60 min (2.02-fold increase). VT correlated significantly (P < 0.0001) with AUC from 30 to 60 min (r2 = 0.95), with AUC from 0 to 30 min (r2 = 0.87), and with kE (r2 = 0.62). Conclusion: 11C-metoclopramide PET imaging revealed the relative importance of both the influx hindrance and the efflux enhancement components of ABCB1 in a relevant model of the human BBB. The overall impact of ABCB1 on drug delivery to the brain can be noninvasively estimated from image-derived outcome parameters without the need for an arterial input function.
- 11C-metoclopramide
- P-glycoprotein
- ABC transporters
- blood-brain barrier
- positron emission tomography
- influx hindrance
- efflux enhancement
P-glycoprotein (ABCB1) is the most studied ATP-binding cassette (ABC) transporter expressed at the luminal (blood-facing) side of endothelial cells forming the blood–brain barrier (BBB) (1,2). At the cellular level, ABCB1 mediates the efflux of a large variety of compounds from the intracellular space. At the BBB, ABCB1 was shown to limit exposure of the brain to many xenobiotics and to protect the brain against potentially neurotoxic substances by restricting their influx from the blood into the brain parenchyma (3). The downside of this protective function is that ABCB1 is a bottleneck in drug development because it reduces the number of effective drug candidates to treat central nervous system (CNS) diseases (4). ABCB1-mediated transport is now systematically considered to be a parameter explaining low brain distribution of new chemical entities (5).
However, many marketed CNS-active drugs were shown to be weak ABCB1 substrates, including some antiepileptic drugs (6), opioids (7), antidepressants (8), and neuroleptics (9). These compounds are considered weak ABCB1 substrates because their permeability is sufficient to cross the BBB and distribute into the brain parenchyma, even when ABCB1 is fully functional (10–12). It can be hypothesized that for such drugs, the impact of ABCB1 on brain distribution may differ from that of avid ABCB1 substrates, with ABCB1 effectively inhibiting uptake of drugs from the blood into the brain, resulting in very low baseline brain distribution (influx hindrance) (13). Once inside the brain, weak ABCB1 substrates may undergo ABCB1-mediated efflux transport across the BBB. ABCB1 may thus control the clearance of its substrates from the brain into the blood (efflux enhancement). Under such conditions, ABCB1 would act as a detoxifying system, thus adding a new role for ABCB1 in controlling drug exposure to the brain after initial uptake (13).
A powerful method to study ABCB1 function in vivo at the human BBB is PET imaging in combination with radiolabeled ABCB1 substrate radiotracers (2). The most commonly used ABCB1 probes for PET are racemic 11C-verapamil, (R)-11C-verapamil, and 11C-N-desmethyl-loperamide. These radiotracers are avid ABCB1 substrates, with a high transport rate, and show very low brain uptake when ABCB1 is functional. In humans, their brain uptake increased when ABCB1 was pharmacologically inhibited, thus providing convincing clinical data on the impact of this functional component of the BBB in vivo (2,14,15). These radiotracers may, however, not be representative of most clinically used CNS-active drugs. Better insight into the role of ABCB1 in controlling the brain distribution of CNS-active drugs is much needed to improve our understanding of the impact of ABCB1 on the variability in response to CNS-active drugs, in terms of both efficacy and adverse effects (16).
Metoclopramide is commonly prescribed as an antiemetic drug. Frequent CNS effects have been reported, suggesting substantial brain distribution (17,18). Metoclopramide is a moderate substrate of both human and rodent ABCB1 (12). We have shown the feasibility of PET imaging with 11C-metoclopramide to measure in vivo ABCB1 function at the rat BBB (19).
In the present study, 11C-metoclopramide PET imaging was performed on nonhuman primates, an animal model that more closely resembles humans in terms of ABCB1 expression levels at the BBB than rodents (20). Compartmental pharmacokinetic modeling was performed to highlight and quantify the specific contribution of ABCB1 to the BBB crossing and subsequent exposure to brain regions of this model CNS-active, moderate ABCB1 substrate. Simplified methods for quantification of ABCB1 function at the BBB, which do not require arterial blood sampling, were developed and validated.
MATERIALS AND METHODS
Animals
All animal-use procedures were in accordance with the recommendations of the European Community (86/809/CEE) and the French National Committees (law 87/848) for the care and use of laboratory animals. The experimental protocol was validated by a local ethics committee for animal use (CETA/APAFIS 892). Animal experiments were performed on 4 adult male Papio anubis baboons (27.58 ± 1.33 kg in weight during the study) obtained from Celphedia, Station of Primatology.
Chemicals and Radiochemicals
Tariquidar used for ABCB1 inhibition was purchased from ERAS Labo. Tariquidar solutions for intravenous injection (3 mg/mL) were freshly prepared on the day of the experiment by dissolving tariquidar dimesylate 2.35 H2O (∼300 mg) in a 5% (w/v) glucose solution (50 mL) followed by dilution with sterile water (50 mL). Ready-to-inject 11C-metoclopramide (4-amino-5-chloro-N-(2-(diethylamino)ethyl)-2-11C-methoxybenzamide) more than 99% radiochemically pure was prepared and controlled as previously described (19,21).
Experimental Conditions
Four baboons underwent 2 11C-metoclopramide PET scans. 11C-metoclopramide brain kinetics were compared in the absence (baseline scan) and presence of pharmacologic ABCB1 inhibition, achieved with a concurrent intravenous infusion of tariquidar (inhibition scan). The tariquidar infusion protocol was adapted from a clinically validated protocol (14). Tariquidar was infused at a dose of 4 mg/kg/h (37.5 mL/h). The infusion started 1 h before radiotracer injection and was continued during the entire PET acquisition. One baboon underwent a third PET experiment with 11C-metoclopramide coinjected with the maximal clinical dose of unlabeled metoclopramide (0.5 mg/kg of a 10 mg/2 mL solution; Laboratoire Renaudin) (Supplemental Fig. 1; supplemental materials are available at http://jnm.snmjournals.org).
Imaging Experiments
First, each baboon underwent a T1-weighted brain MR scan using an Achieva 1.5-T scanner (Philips Healthcare) under ketamine anesthesia (10 mg/kg, intramuscularly; Virbac).
PET experiments were performed on an HR+ Tomograph (Siemens Healthcare), with the animal anesthetized and monitored as previously described (22). The animal received ketamine (10 mg/kg, intramuscularly) to induce anesthesia. Once the animal was intubated, venous catheters were inserted for radiotracer injection (sural vein), propofol infusion (sural vein), and tariquidar infusion (brachial vein). Another catheter was inserted into the femoral artery for arterial blood sampling. The animal was positioned under the camera before administration of an intravenous bolus of propofol (2 mL) followed by a 16–22 mL/h intravenous infusion under oxygen ventilation.
The animal was intravenously injected with 11C-metoclopramide (298 ± 44 MBq; 5.6 ± 3.1 μg). A dynamic PET acquisition (60 min) was performed over the brain with the animal supine.
Arterial Input Function and Metabolism
During the PET acquisition, arterial blood samples (0.5 mL) were withdrawn at selected times after radiotracer injection. Samples were centrifuged (5 min, 2,054g, 4°C) and the supernatant (200 μL) was counted for total radioactivity.
Additional plasma samples were withdrawn to measure the percentage of parent (unmetabolized) using high-performance liquid chromatography (HPLC). Time–activity curves of parent 11C-metoclopramide in plasma were expressed as SUV ([radioactivity/mL of plasma/injected radioactivity] × body weight). Plasma exposure to parent 11C-metoclopramide was estimated under all tested conditions by calculating the area under the time–activity curve (AUC) in plasma.
Dedicated arterial plasma samples were withdrawn immediately before 11C-metoclopramide injection to assess 11C-metoclopramide binding to plasma proteins. The fraction of 11C-metoclopramide that was not bound to baboon plasma proteins (fP) was measured as previously described (23).
Imaging Data Analysis
PET images were reconstructed as previously described (24). The PET data were analyzed using PMOD software (version 3.8; PMOD Technologies Ltd.). PET images were coregistered onto corresponding MR images for each baboon. A baboon T1-weighted MR template (25) was normalized onto individual MR images. Transformation matrices were then applied to the segmentation obtained from the template, which included 12 brain structures as volumes of interest: cingulate cortex, orbital cortex, occipital cortex, temporal cortex, frontal cortex, caudate nucleus, putamen, precentral gyrus, postcentral gyrus, thalamus, superior parietal lobe, and cerebellum.
Individual volumes of interest were applied to coregistered dynamic PET images to generate corresponding time–activity curves. Kinetic modeling was performed using a 1-tissue-compartment model (1-TCM) with the corresponding arterial plasma input function of parent 11C-metoclopramide to estimate the influx (K1; mL/min/cm3) and efflux (k2; min−1) rate constants and the total volume of distribution (VT = K1/k2 in the 1-TCM; mL/cm3) for each region under each tested condition. Additional PET kinetic parameters were calculated, including the AUC of regional time–activity curves from 0 to 30 min (AUC0–30 min) and from 30 to 60 min (AUC30–60 min). The elimination slope (kE; min−1) of 11C-metoclopramide washout from the brain was estimated from the log-transformed time–activity curves from 30 to 60 min after radiotracer injection.
HPLC Determination of Tariquidar in Plasma
Arterial blood samples (3 mL) withdrawn immediately before and during the PET acquisition were used to determine tariquidar concentrations using a newly developed HPLC-ultraviolet method.
Solvents and Chemicals
Elacridar hydrochloride (C34H33N3O5⋅HCl; molecular weight, 600.1 g⋅mol−1), used as an internal standard, was purchased from Syncom. Ultrapure water was obtained using an ultraviolet purification system (Purelab; ELGA LabWater).
HPLC Method
An HPLC-ultraviolet method was developed to determine tariquidar in plasma. HPLC was performed on an Alliance 2996 system (Waters), equipped with an autosampler, a binary pump, and a photodiode array detector. Separation was achieved using an AtlantisT3 C18 column (4.6 × 150 mm, 5 μm; Waters). The mobile phase consisted of water containing 0.1% (v/v) trifluoroacetic acid (solvent A) and acetonitrile containing 0.1% (v/v) trifluoroacetic acid (solvent B) delivered in a gradient elution mode at a flow rate of 1.2 mL⋅min−1: solvent B increased linearly from 20% to 65% from 0 to 15 min. Tariquidar and the internal standard were detected at a ultraviolet wavelength of 250 nm.
Sample Preparation
Stock solutions of tariquidar and the internal standard were prepared in water/acetonitrile (80/20 and 27/73, v/v, respectively) at a concentration of 1 mg⋅mL−1. The stock solutions were stored at −20°C.
Arterial and venous blood samples from the animal experiments were collected in tubes containing lithium iodoacetate as an anticoagulant (BD Vacutainer) and were immediately centrifuged for 10 min (2,054g, 4°C). The plasma was stored at −80°C until analysis (<1 mo). For analysis, samples were thawed at room temperature. The internal standard solution (diluted with water to a concentration of 60 μg⋅mL−1 [100 μL]) and water containing 4% hydrochloric acid (v/v, 500 μL) were added to the plasma sample (600 μL). After being stirred in a vortex mixer, 1 mL of this mixture was deposited on a cation-exchange solid-phase extraction cartridge (MCX, 30 mg, Oasis; Waters) that had been preconditioned with 1 mL of methanol and 1 mL of water. The cartridge was washed with water containing 0.4% (v/v) hydrochloric acid (1 mL) followed by acetonitrile containing 0.4% (v/v) hydrochloric acid (1 mL). Tariquidar and the internal standard were finally eluted with acetonitrile containing 4% (v/v) ammonia (3 × 1 mL). The combined alkaline acetonitrile fractions were evaporated during 120 min (SPD1010 SpeedVac system; Thermo Scientific). The residue was dissolved in water/acetonitrile/trifluoroacetic acid (80/20/0.5, v/v/v, 100 μL), stirred in a vortex mixer for 30 s, and sonicated for 5 min. This solution (20 μL) was then injected into the HPLC system.
HPLC Determination of Parent 11C-Metoclopramide in Plasma
During the PET experiments, additional plasma samples withdrawn at 0, 5, 10, 15, 30, and 60 min were used to measure the percentage of parent (unmetabolized) 11C-metoclopramide, assessed using radio-HPLC analysis. After centrifugation, the total radioactivity in the plasma samples (500 μL) was counted. Samples were then deproteinated with acetonitrile (700 μL) and centrifuged (5 min, 2,054g, 4°C). The pellet of proteins was counted. Recovery of radioactivity in the supernatant was more than 90%. Then, 1,000 μL of the supernatant were injected into the radio-HPLC system.
The radio-HPLC system consisted of a gradient pump, an ASI100T autosampler, and a UVD170U ultraviolet–visible-light detector (Thermo Scientific) in line with a Flo-One scintillation analyzer (Packard). Separation was achieved using an Atlantis T3 5-μm, 10 × 250 mm column (Waters). The mobile phase consisted of 0.1% trifluoroacetic acid in water (solvent A) and acetonitrile (solvent B). A linear gradient from 20% to 40% of solvent B over 9 min was applied to the column at a flow rate of 5 mL/min.
For each experiment, a monoexponential decay function was fitted to the percentage of unmetabolized 11C-metoclopramide versus time and then applied to the corresponding total radioactivity in plasma.
Statistical Analysis
Statistical analysis was performed using GraphPad Prism software (version 7.0). Outcome parameters obtained from pharmacokinetic modeling were compared using 2-way ANOVA. A result was deemed significant when a 2-tailed P value was less than 0.05.
RESULTS
Plasma Pharmacokinetics
Radio-HPLC analysis showed the presence of 11C-metoclopramide and radiometabolites in plasma. In the first 30 min, the parent fraction of 11C-metoclopramide could be determined. After 30 min, the peak corresponding to parent 11C-metoclopramide was below the limit of quantitation and plasma radioactivity consisted of only radiometabolites (Fig. 1). Metabolite-corrected arterial input functions and corresponding AUCs in plasma are thus reported from 0 to 30 min, which corresponds to the time frame allowing for accurate quantitation of parent 11C-metoclopramide in plasma. 11C-metoclopramide plasma kinetics were not affected by tariquidar infusion, with AUCs in plasma of 8.9 ± 4.8 and 9.6 ± 2.9 SUV⋅min in the baseline and inhibition scans, respectively (P = 0.797). Plasma concentrations peaked rapidly, followed by a fast washout of radioactivity to reach equilibrium (∼0.1 SUV) at 10 min after 11C-metoclopramide injection (Fig. 1).
Plasma pharmacokinetics of tariquidar and 11C-metoclopramide during PET experiments. (A) Plasma pharmacokinetics of tariquidar (n = 3). (B) fP of parent 11C-metoclopramide. (C) Arterial input function of parent (unmetabolized) 11C-metoclopramide in animals without (baseline) or with coinfusion of tariquidar (n = 4 for both conditions). Data are mean ± SD.
The fP of 11C-metoclopramide was high in baseline scans (72.2% ± 2.5%), as was consistent with low metoclopramide binding to plasma proteins in humans (26). In tariquidar-treated animals, fP was significantly lower (61.6% ± 1.9%; P < 0.001, Student t test).
The pharmacokinetics of tariquidar could be measured in only 3 animals because of technical issues. The tariquidar concentration in arterial plasma increased rapidly to reach a range of 3.1–3.9 μg/mL at the start of the PET experiments. The concentration ranged from 4.1 to 5.0 μg/mL at the end of the PET experiments (Fig. 1).
PET Experiments
Baseline PET images showed substantial brain uptake of 11C-metoclopramide–associated radioactivity, which did not accumulate in specific brain regions (Fig. 2). Whole-brain radioactivity peaked rapidly (SUVmax, 1.3 ± 0.1; time of maximum uptake, ∼4.5 min) followed by a slow decrease, with an SUV of 0.4 ± 0.1 at 60 min after injection (Fig. 2). ABCB1 inhibition resulted in a pronounced and significant increase in SUVmax (1.8 ± 0.1, P < 0.001) and a delay in the time to maximum uptake, which occurred at about 7 min. The SUV was higher at 60 min (0.9 ± 0.1) than at baseline (P < 0.001) (Figs. 1B and 2B). The kinetics of 11C-metoclopramide were similar across different brain regions, in either the presence or the absence of tariquidar (Fig. 2).
11C-metoclopramide PET data in baboon brain. (A) Representative PET summation images (0–60 min) of brain of baboon without (baseline, left) or with (right) coinfusion of tariquidar. Radioactivity concentration is normalized to injected dose per body weight and expressed as SUV. (B) Mean time–activity curves obtained under both conditions using either linear (black) or logarithmic (red) scale. (C) Regional time–activity curves for both conditions.
The PET data were modeled using a 1-TCM under the 2 tested conditions. The baseline distribution of 11C-metoclopramide to the whole brain (VT = 4.3 ± 0.5 mL/cm3; VT/fP = 5.9 ± 0.6 mL/cm3) was significantly increased by ABCB1 inhibition (VT = 8.7 ± 0.5 mL/cm3; VT/fP = 14.2 ± 0.4 mL/cm3). Tariquidar induced a significant increase in brain AUC0–30 min and AUC30–60 min. Estimation of the transfer rate constants for 11C-metoclopramide across the BBB showed a significant increase in K1, associated with a significant decrease in k2. The effect of tariquidar on k2 was consistent with a significant decrease in the elimination slope of radioactivity from the brain (kE) from 30 to 60 min (Fig. 3; Supplemental Table 1).
Impact of ABCB1 inhibition on 11C-metoclopramide pharmacokinetic parameters. Percentage change in each parameter in whole brain in inhibition scan relative to baseline scan is reported. Data are individual values, as well as mean ± SD over 4 baboons.
For all outcome parameters (VT, K1, k2, kE, AUC0–30 min, and AUC30–60 min), the response to ABCB1 inhibition did not significantly differ among brain regions (P = 0.14), suggesting that the impact of tariquidar on ABCB1 function was homogeneous in the different brain regions. The impact of tariquidar was highest on VT and AUC30–60 min and lowest on K1 and kE (Fig. 3; Supplemental Table 1). The variability in response to tariquidar was highest for K1 (coefficient of variation, 51%) and lowest for AUC30–60 min (9%) and kE (11%) (Fig. 3).
The brain distribution of 11C-metoclopramide (VT), estimated in each region from 0 to 30 min, was used as a reference parameter to evaluate simplified methods to quantify ABCB1 function at the BBB. A significant correlation was found between VT and AUC0–30 min (P < 0.0001; r2 = 0.87) and between VT and AUC30–60 min (r2 = 0.95). The kE of the regional time–activity curves from 30 to 60 min correlated with VT and k2 (P < 0.0001; r2 = 0.62) (Fig. 4).
Correlations between different parameters describing kinetics of 11C-metoclopramide in brain. Regional outcome parameter values obtained in 4 animals in presence and absence of ABCB1 inhibition are plotted. Linear regression analysis was performed. Statistical significance and goodness of fit (r2) are reported for each correlation.
DISCUSSION
PET imaging using radiolabeled ABCB1 substrates convincingly highlighted the importance of ABCB1 for drug distribution to the brain in animals and humans (2). The currently available PET probes revealed and quantified the influx hindrance (K1) component of ABCB1. In this study, 11C-metoclopramide PET imaging and kinetic modeling were performed on nonhuman primates to gain information about the kinetic role of ABCB1 on the brain distribution of a weak ABCB1 substrate.
11C-metoclopramide benefits from favorable pharmacokinetic properties that allow for quantitative determination of modeling outcome parameters describing its brain distribution. 11C-metoclopramide behaves like a weak substrate for ABCB1, with substantial brain uptake (baseline VT = 4.3 mL/cm3) despite fully functional ABCB1 activity. In vitro experiments have shown that 11C-metoclopramide is specifically transported by human ABCB1 but not ABCG2, the other major ABC transporter expressed at the human BBB (19). In rats, the BBB was poorly crossed by 11C-metoclopramide radiometabolites, compared with the parent compound, thus ensuring the radiochemical purity of the PET signal in the brain even when ABCB1 was inhibited (19).
In nonhuman primates, the brain kinetics of a tracer dose of 11C-metoclopramide could be described by a simple 1-TCM. This model allows for the estimation of rate constants to describe the exchange of 11C-metoclopramide–associated radioactivity between the plasma and the brain compartments (27). Metoclopramide is a dopamine D2 receptor antagonist and was expected to bind to dopamine D2 receptors in the brain (28). In our study, radioactivity was homogeneously distributed among brain regions and did not predominantly accumulate in dopamine D2 receptor–rich regions such as the striatum (Fig. 2). Our previous rat experiments have shown that ABCB1 inhibition resulted in a significant increase in 11C-metoclopramide nondisplaceable binding potential in the whole brain, estimated using a 2-tissue-compartment model (2-TCM). However, injection of a high dose of unlabeled metoclopramide during the PET scan did not displace the brain radioactivity, suggesting that receptor binding is not the main mechanism for 11C-metoclopramide retention in the brain (19). In our nonhuman primate study, the 2-TCM model did not allow for accurate estimation of 11C-metoclopramide nondisplaceable binding potential and was therefore not used. Coinjection of unlabeled metoclopramide had no impact on 11C-metoclopramide brain kinetics (Supplemental Fig. 1). The mechanism of 11C-metoclopramide retention in the brain remains to be elucidated to determine a possible interference with quantification of ABCB1 function at the BBB in different pathophysiologic conditions.
We chose a clinically feasible infusion protocol for tariquidar to highlight the impact of impaired ABCB1 function on the brain kinetics of 11C-metoclopramide (14). This protocol allowed for maintained and controlled plasma concentrations during the PET experiments to limit any variation in ABCB1 function. ABCB1 inhibition induced a significant, approximately 2-fold, increase in the VT of 11C-metoclopramide. Tariquidar did not cause any change in the arterial input function and did not increase the 11C-metoclopramide fP (23). As a consequence, VT significantly correlated with the concomitant brain exposure (AUC0–30 min; Fig. 4).
Pharmacokinetic modeling revealed that ABCB1 induced a significant 28.5% increase in K1. This effect corresponds to the influx hindrance component of ABCB1 function at the BBB. Compared with established avid ABCB1 PET probes, this effect was modest, as might be explained by the lower transport ability of metoclopramide by ABCB1 or a higher passive diffusion component of the baseline 11C-metoclopramide transport across the BBB (12). ABCB1 inhibition also induced a concomitant 36% decrease in k2. This result shows that once in the brain, ABCB1 can mediate clearance of its substrates back into the blood compartment. From a pharmacokinetic perspective, the efflux enhancement function of ABCB1 (described by k2), in addition to its effect on initial uptake (described by K1), accounts for the overall brain exposure (described by VT and AUC), which was increased by approximately 2-fold after ABCB1 inhibition (29).
The in vitro transport of solutes across cell monolayers with polarized (apical) ABCB1 expression shows that the predominant impact of ABCB1 is usually along the concentration gradient (basolateral-to-apical) rather than against the concentration gradient (apical-to-basolateral) (10,12). This finding suggests that ABCB1 is more efficient in promoting efflux clearance than in competing with passive diffusion to limit the net influx across biologic barriers (10,12). In vivo, the efflux enhancement component of ABCB1 at the BBB has been suggested from invasive experiments. The brain efflux index can be assessed by injecting solutes into the brain and measuring the remaining brain concentrations during washout (30). Brain microdialysis allows for concomitant consideration of the brain and plasma kinetics of solutes and estimation of the corresponding transfer constant (13). Our PET data suggest an efflux-component role for 11C-metoclopramide distribution across the BBB, but the presence of brain radiometabolites prevented us from addressing this role using 11C-verapamil PET imaging. 11C-N-desmethyl-loperamide was shown to be trapped in brain cells, and the radiotracer may therefore not be available for ABCB1-mediated brain clearance in vivo (31).
The ABCB1-mediated efflux clearance could also be intuitively estimated from the Log-transformed brain time–activity curves for 11C-metoclopramide. Thirty minutes after injection, 11C-metoclopramide–associated radioactivity was about 7-fold higher in the brain than in plasma when ABCB1 was fully functional. Thereafter, parent 11C-metoclopramide could not be detected in plasma anymore. This situation is similar to the kinetic properties of 99mTc-sestamibi, for which the efflux rate was shown to be more accurate than the initial uptake in predicting ABCB1-mediated multidrug resistance in patients (32).
We sought to propose simplified parameters for quantitative estimation of the role of ABCB1 in the brain kinetics of 11C-metoclopramide. Using VT as a reference parameter, brain AUC was investigated as a surrogate parameter, as previously reported for 11C-N-desmethyl-loperamide (33). We found a better correlation between VT and AUC30–60 min than between VT and AUC0–30 min, the latter being more likely dependent on the variability in plasma and brain kinetics at early time points. AUC30–60 min should preferentially be used as a surrogate parameter to describe VT, as AUC30–60 min reflects the impact of ABCB1 on the overall brain exposure to 11C-metoclopramide without the need for an arterial input function. Interestingly, kE correlated with VT and k2 (Fig. 4). kE depends neither on the arterial input nor on the initial uptake of the radiotracer and corresponding SUVmax. Moreover, a low variability in the response of kE to ABCB1 inhibition was observed (Fig. 4). In some situations, kE may therefore be preferred as a pharmacokinetically relevant outcome parameter to noninvasively describe and quantify the efflux enhancement role of ABCB1 at the BBB.
Tracer-dose 11C-metoclopramide PET imaging can be safely and noninvasively performed on humans with or without ABCB1 inhibition. Compared with existing radiotracers, the lower transport rate of 11C-metoclopramide by ABCB1 may offer new opportunities for the study of ABCB1 function at the BBB. Further investigations are thus needed to assess the sensitivity of 11C-metoclopramide PET in detecting the functional impact of disease-induced reduction or induction of ABCB1 at the BBB (34).
CONCLUSION
We performed 11C-metoclopramide PET imaging and pharmacokinetic modeling on baboons to validate simplified methods for quantification of ABCB1 function at the BBB for practical clinical use. 11C-metoclopramide PET imaging provides novel insight into the kinetic impact of ABCB1 at the BBB on brain distribution of weak substrates.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
Acknowledgments
We thank Thierry Lekieffre, Maud Goislard, and Vincent Brulon for helpful technical assistance.
Footnotes
Published online May 10, 2018.
- © 2018 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication February 20, 2018.
- Accepted for publication April 23, 2018.
Jump to section
Related Articles
Cited By...
- Imaging P-Glycoprotein Induction at the Blood-Brain Barrier of a {beta}-Amyloidosis Mouse Model with 11C-Metoclopramide PET
- Impact of P-Glycoprotein Function on the Brain Kinetics of the Weak Substrate 11C-Metoclopramide Assessed with PET Imaging in Humans
- Overcoming P-Glycoprotein-Mediated Drug Resistance with Noscapine Derivatives