Brain histamine H1 receptor occupancy of orally administered antihistamines, bepotastine and diphenhydramine, measured by PET with 11C-doxepin


Dr Manabu Tashiro, Division of Cyclotron Nuclear Medicine, Cyclotron and Radioisotope Centre, Tohoku University, 6-3 Aoba, Aramaki, Aoba-ku, Sendai, Miyagi 980-8578, Japan.
Tel: + 81 22 795 7797
Fax: + 81 22 795 7797



• ‘Bepotastine besilate’ is a novel second-generation antihistamine developed in Japan and its antiallergic effects have already been demonstrated by various studies.

• However, only a few clinical studies regarding its sedative property are available.

• In addition, histamine H1 receptor occupancy (H1RO) of this new antihistamine has never been measured by positron emission tomography (PET).


• This paper provides the first measurement result of cerebral H1RO of bepotastine besilate (approximately 15%) as determined by PET.

• This result is in accordance with the clinical classification of bepotastine as a second-generation antihistamine.

• In addition, the relationship between subjective sleepiness and cerebral H1RO of this second-generation antihistamine is demonstrated for the first time using a placebo-controlled crossover study design.


Antihistamines are frequently used for treating various allergic diseases, but often induce sedation. The degree of sedation can be evaluated by measuring histamine H1 receptor occupancy (H1RO) in the brain using positron emission tomography (PET). The aim was to measure H1RO of bepotastine, a new second-generation antihistamine, and to compare it with that of diphenhydramine.


Eight healthy male volunteers (mean age ± SD 24.4 ± 3.3 years) were studied after single oral administration of bepotastine (10 mg), diphenhydramine (30 mg) or placebo, by PET imaging with 11C-doxepin in a crossover study design. Binding potential ratio and H1ROs were calculated using placebo data and were compared between bepotastine and diphenhydramine in the anterior and posterior cingulate gyri (ACG and PCG, respectively), superior and inferior frontal cortices (SFC and IFC, respectively), orbitofrontal cortex (OFC), insular cortex (IC), lateral and medial temporal cortices (LTC and MTC, respectively), parietal cortex (PC), occipital cortex (OC) and sensorimotor cortex (SMC). Plasma concentration of each antihistamine was measured, and its correlation to H1RO was examined.


H1RO after bepotastine treatment was significantly lower than that after diphenhydramine treatment in all cortical regions (P < 0.001). Mean H1ROs of bepotastine and diphenhydramine were 14.7% and 56.4%, respectively. H1ROs of both bepotastine and diphenhydramine correlated to their respective drug plasma concentration (P < 0.001).


Oral bepotastine (10 mg), with its relatively low H1RO and thus minimal sedation, has the potential for use as a mildly or slightly sedative antihistamine in the treatment of various allergic disorders.


Histamine H1 receptor (H1R) antagonists, or antihistamines, are often used for treating allergic disorders such as seasonal rhinitis. Antihistamines mainly act on peripheral tissues, but can induce sedation as a central side-effect. This undesirable side-effect is caused by blockade of nerve transmission in the histaminergic neuron system which projects from the tuberomammillary nucleus in the posterior hypothalamus to almost all cortical areas [1–5]. First-generation antihistamines that can easily penetrate the blood–brain barrier (BBB), such as diphenhydramine and d-chlorpheniramine, tend to occupy a large proportion of postsynaptic H1Rs as demonstrated by positron emission tomography (PET) [1, 6–8]. PET also reveals that second-generation antihistamines (mildly or slightly sedative antihistamines), such as cetirizine and olopatadine, can slightly penetrate the BBB and occupy some H1Rs [1, 6, 9, 10]. Users who take these second-generation antihistamines at doubled or tripled doses to achieve desired effects may suffer from dose-related cognitive impairment. Third-generation antihistamines (truly nonsedative antihistamines), such as fexofenadine and ebastine, hardly penetrate the BBB and do not occupy H1Rs even at high doses, as demonstrated by 11C-doxepin PET [9]. Thus, the sedative property of antihistamines can be evaluated in terms of H1R occupancy (H1RO). Such variations in BBB permeability are caused by various factors, including differences in lipophilicity, molecular size and actions of drug transporters.

Bepotastine besilate ({d-(S)-4-[4-[(4-chlorophenyl) (2-pyridyl)methoxy]piperidino} butyric acid monobenzenesulphonate, betotastine besilate, CAS 125602-71-3, TAU-284 or Talion), a new second-generation antihistamine developed in Japan, is now used as an oral tablet for allergic rhinitis and chronic urticaria (Figure 1) [11–13]. Previous studies have demonstrated its excellent antiallergic effects compared with other antihistamines such as ketotifen, cetirizine, epinastine and terfenadine [14–18], whereas only a few studies have shown its central effects [18, 19]. Only one available animal behavioural study by Kato and colleagues has demonstrated that bepotastine is a highly specific drug to H1R, having no significant binding affinity for histamine H3, adrenergic α1, α2, β, dopaminergic D2, serotonergic 5HT2, muscarinic or benzodiazepine receptors, and that it poorly penetrates the BBB [19]. Takahashi and colleagues first conducted a double-blind, placebo-controlled, crossover study to measure subjective sedation and psychomotor activities following administration of bepotastine, cetirizine, fexofenadine and olopatadine [18], where bepotastine had the least sedative effect [18].

Figure 1.

Chemical structure of bepotastine besilate

To date, we have measured H1ROs of various second-generation and third-generation antihistamines, but not that of bepotastine. It is of great interest to examine H1RO of bepotastine in humans. Thus, the primary aim of the present study was to measure subjective sedation and cerebral H1RO of bepotastine and to compare the results with those of diphenhydramine, a typical sedative antihistamine [20], using a placebo-controlled crossover study design that would make the interpretation of results clearer and easier by minimizing potential errors due to intersubject variability [10]. Another aim was to determine whether bepotastine should be classified as a truly nonsedative or mildly sedative antihistamine.


The present study was approved by the Committee on Clinical Investigation at Tohoku University Graduate School of Medicine, Japan, and was performed in accordance with the principles of the Declaration of Helsinki. All experiments were performed at the Cyclotron and Radioisotope Centre, Tohoku University.

Subjects and study design

Eight male Japanese volunteers (mean age ± SD 24.4 ± 3.3 years), who provided written informed consent after receiving a detailed description of the study, were recruited. All subjects were in good health with no clinical history of major physical or mental illness, showed no abnormality in brain magnetic resonance imaging (MRI), and were not receiving any concomitant medication likely to interfere with the study results. Nicotine, caffeine, grapefruit and grapefruit juice were not allowed on the test day, and alcohol was not allowed on the test day or the day before PET measurement.

All subjects underwent PET measurement after single oral administration of bepotastine (10 mg), diphenhydramine (30 mg) or a lactobacteria preparation (6 mg) used as placebo in a three-way crossover study, with minimum wash-out intervals of 7 days between treatments. The lactobacteria preparation has been widely used as placebo in Japan, and its administration has produced no statistical difference between pre- and post-administration in previous cognitive studies [7, 9, 10, 21]. The present PET study was conducted in a single-blinded manner, and after drug administration each subject was asked to remain seated comfortably on a sofa. To determine bepotastine and diphenhydramine plasma concentrations, blood samples were collected from each subject before drug administration and at 0, 60, 120 and 180 min post administration. Subjective sleepiness of each subject was also measured at 0, 60, 120 and 180 min post administration using the Line Analogue Rating Scale (LARS) [22] and the Stanford Sleepiness Scale (SSS) as used in previous studies [9, 23].

Measurement of plasma concentrations of bepotastine and diphenhydramine

Plasma concentrations of bepotastine and diphenhydramine were measured using liquid chromatography/tandem mass spectrometry (LC/MS/MS) together with an electrospray ionization method [24]. LC was performed on an Agilent 1100 Series LC instrument (Agilent Technologies Inc., Santa Clara, CA, USA) equipped with an analytical column. The MS/MS system was the API 4000 (Applied Biosystems/MDS Sciex, ON, Canada). The Solid Phase Extraction (SPE) cartridge (OASIS HLB 3 ml/60 mg; Waters Corp., Milford, MA, USA) was pretreated with 2 ml of methanol, 2 ml of water and 2 ml of 0.2 m Na2CO3/HCl buffer (pH 11).

For measurement, an internal standard solution (10 μl) and methanol (10 μl) were added to each plasma sample (50 μl). To the resulting solution, 930 ml of 0.2 m Na2CO3/HCl buffer was added for bepotastine measurement, and 1000 μl of 0.1% formic acid containing acetonitrile/methanol (50 : 50, v/v) was added for diphenhydramine measurement. The mixture was applied onto the SPE cartridge after pretreatment as mentioned above. Separations were carried out on a high-performance liquid chromatography (HPLC) column [CAPCELL PAK C18 MG II (3 μm) 2.0 mmφ × 100 mm; Shiseido Co., Ltd, Tokyo, Japan) at a flow rate of 0.2 ml min−1 and at a column temperature of 40 °C. The reconstituted extract (5 μl) was injected onto an HPLC system with mobile phases for bepotastine measurement including 10 mmol l−1 ammonium acetate and acetonitrile of varied concentrations, namely, 32% (0–9 min), 70% (9.5–12.5 min) and 32% (12.6–24 min), and with mobile phases for diphenhydramine measurement including 0.1% heptafluorobutyric acid and acetonitrile of varied concentrations, namely, 40% (0–7 min), 70% (7.5–10.5 min) and 40% (11–21 min).

Detection of bepotastine was based on fragmentation of the precursor ion (m/z = 389 to product ion m/z = 202 with collision energy of 29 eV for bepotastine, and m/z = 256 to product ion m/z = 167 with collision energy of 19 eV for diphenhydramine), and that of the internal standard was based on fragmentation of the precursor ion (m/z = 389 to product ion m/z = 201 with collision energy of 29 eV for bepotastine, and m/z = 270 to product ion m/z = 181 with collision energy of 17 eV for diphenhydramine) in positive multiple reaction monitoring (MRM) mode. Positive ions were detected using an API 4000 system at 550 °C nebulizer gas temperature, with 5000 V ion spray voltage, 68.9 kPa (nitrogen) curtain gas and Level 4 collision gas for bepotastine, and 206.8 kPa curtain gas and Level 4 collision gas for diphenhydramine. Chromatographic data for positive MRM were collected using Analyst software (ver. 1.2, Applied Biosystems/MDS Sciex) with cycle times of 1.010 s per cycle for bepotastine and 0.5200 s per cycle for diphenhydramine. The lowest detectable concentration was around 1 ng ml−1 for both antihistamines, and some values slightly under the threshold (only for diphenhydramine) were extrapolated. As for validation, the following items were checked for bepotastine and diphenhydramine, respectively: accuracy (100.7% and 102.2%), correlation coefficients to standard solutions (r > 0.99 for both), and coefficients of variation (CVs) of three different concentrations (n = 5) (1.7–2.3%, 1.8–3.8%).

For examination of the relationship between estimated binding potential ratio (BPR) of 11C-doxepin and plasma concentration of each antihistamine, the areas under the curves (AUCs) of bepotastine and diphenhydramine were calculated for 0–180 min (AUC0−3 h) post administration (Table 1).

Table 1. 
Plasma concentrations of bepotastine and diphenhydramine (n = 8)
Bepotastine time (min)Mean (ng ml−1)SEMCV, %Diphenhydramine time (min)Mean (ng ml−1)SEMCV, %
  1. CV, coefficient of variation; SEM, standard error of mean.
AUC0−3 h196.124.335.0AUC0−3 h32.34.035.3

PET tracer and image acquisition

Doxepin is one of the tricyclic antidepressants that has binding affinity to other receptors such as muscarinic receptors to some extent. However, its affinity to histamine H1Rs is much higher than to other receptors and is very high compared with other antidepressants [25]. Thus, doxepin's affinity to other receptor systems is negligible in this imaging study, as also confirmed by experiments using histamine H1R knock-out mice, where doxepin binding in the brain was nearly zero [26]. Thereafter, 11C-doxepin has been used to evaluate the distribution of histamine H1Rs. 11C-doxepin kinetics in plasma and the brain are not affected by the sedative antihistamine d-chlorpheniramine using arterial sampling data combined with metabolite analysis [6].

In the present study, 11C-doxepin was prepared by 11C-methylation of desmethyl doxepin with 11C-methyl triflate as described previously [10, 27]. 11C-doxepin radiochemical purity was >99%, and its specific radioactivity at the time of injection was 120.9 ± 80.55 GBq μmol−1 (3268 ± 2177 mCi μmol−1). 11C-doxepin-containing saline solution was intravenously injected into each subject at 90 min post administration, a time close to Tmax of both antihistamines (1.2 h for bepotastine and 2–3 h for diphenhydramine). The injected dose and cold mass of 11C-doxepin were 135.4 ± 19.83 MBq (3.660 ± 0.536 mCi) and 1.587 ± 0.895 nmol, respectively, and the radiological dose was calculated based on a previous study on radiological exposure [28].

Approximately 60 min after 11C-doxepin injection, the subjects were positioned on the coach of the PET scanner (SET2400W; Shimadzu Co., Kyoto, Japan) for transmission scan (6 min) and emission scan in the three-dimensional (3D) mode lasting for 15 min (70–85 min post injection of 11C-doxepin) in a similar fashion to our previous work [10, 29]. PET brain images from a 15-min-long emission scan were corrected for scattering based on a previous study [30] and for tissue attenuation using post-injection transmission scan data according to previous work [31]. Brain images were reconstructed with a filtered back-projection algorithm, with the aid of a supercomputer SX-7 at the Information Synergy Centre, Tohoku University, Sendai, Japan. The brain images were then normalized by plasma radioactivity at 10 min post injection to yield images reflecting distribution volume (DV) based on our static scan protocol reported previously [10, 32]. Validation using venous sampling instead of arterial sampling was carried out by another group, giving no difference between venous and arterial sampling at 10 min post injection (M. Senda, personal communication, 23 August 2007).

Three brain images of each subject, following oral administration of bepotastine, diphenhydramine and placebo, were coregistered to an identical stereotaxic brain coordinate system using an MRI-T1 image of each subject, with the aid of Statistical Parametric Mapping (SPM2, Wellcome Department, UK) software package [33]. MRI images were obtained with a 1.5-T magnetic resonance (MR) scanner (HiSpeed, Ver. 9.1; General Electric Inc., WI, USA) at Sendai Seiryo Clinic (miyagi, Japan). T1-weighted images (Vascular TOF SPGR: TR/TE 50/2.4 ms, FA 45°, number of excitations 1, matrix size 256 × 256, spatial resolution: x, y, z = 0.86, 0.86, 20.0 mm, respectively) were collected from all subjects.

Regions of interest (ROIs) were first placed on the following brain regions on the T1 images that had precise anatomical information, i.e. anterior and posterior cingulate gyri (ACG and PCG, respectively), prefrontal cortices (PFC), orbitofrontal cortex (OFC), insular cortex (IC), temporal cortex (TC), parietal cortex (PC), occipital cortex (OC), primary sensorimotor cortex (SMC), thalamus, striatum, midbrain, pons, and cerebellum. ROI was defined for each cortical region by two to five circles with a diameter of 7.6 mm for each hemisphere in four to five consecutive brain transaxial slices, as indicated in Figure 2A. For the thalamus, striatum, pons and midbrain, the margin of each region was traced in MRI T1 images. An averaged value from all ROIs was used as a representative value of each region. Information on ROI location was automatically transferred to the coregistered three PET images reflecting DV, and the binding potential ratio (BPR) was calculated for each region using the following equation: BPR = [(DV of each region − DV of cerebellum)/DV of cerebellum][8, 9]. Finally, H1ROs of bepotastine and diphenhydramine were calculated for each cortical region using the following equation: H1RO = [(BPR of placebo − BPR of antihistamine)/BPR of placebo] × 100. BPR brain images were also created by applying the same equation to each DV brain image [8–10, 34, 35]. BPR brain images were analysed statistically on a voxel-by-voxel basis using SPM2 [33], following spatial normalization and smoothing using the same method as in our previous work. Differences in parameter values between bepotastine, diphenhydramine and placebo (control) were statistically examined, and regional maxima of statistical significance (P < 0.001) were projected onto surface-rendered MRI-T1 standard brain images. Precise locations of statistically significant regions were identified with the Co-Planar Stereotaxic Atlas [36].

Figure 2.

Binding potential ratio (BPR) images of 11C-doxepin in the human brain (A) and results of voxel-by-voxel comparison (B). BPR of 11C-doxepin was calculated in healthy male subjects (n = 8) by positron emission tomography following oral administrations of placebo (left), bepotastine (10 mg, middle) or diphenhydramine (30 mg, right), and their magnetic resonance imaging-T1 images (far right), demonstrated in the transaxial (top), coronal (middle) and sagittal (bottom) sections for each treatment condition, were compared (A). White circles in the transaxial images indicate the regions of interest (ROIs). The brain image of each subject was transformed to fit stereotaxic brain space (spatial normalization) and was averaged across each drug condition to generate the mean images displayed (A). The images demonstrate that diphenhydramine treatment results in BPR significantly lower than those of other drug conditions (B). Height threshold of voxel values was set at P < 0.001 and extent threshold was set at 10 voxel minimum. Results were not corrected for multiple comparisons. There are no areas with significantly lower BPR after bepotastine treatment compared with those after placebo treatment (‘bepotastine 10 mg < placebo’ in the left columns). In contrast, red colour shows areas with significantly lower BPR after diphenhydramine treatment compared with those after placebo treatment (‘diphenhydramine 30 mg < placebo’ in the right columns). In both columns, significant areas are demonstrated in four aspects, namely, left and right medial (L. MED and R. MED) and right and left lateral (R. LAT and L. LAT) aspects (P < 0.001, uncorrected, using SPM2) (B)

Statistical analysis

Differences in subjective sleepiness and BPR between bepotastine, diphenhydramine and placebo were examined using anova followed by multiple comparison with Bonferroni correction. The difference in H1RO between bepotastine and diphenhydramine was examined using paired Student's t-test. The relationship between plasma drug concentration (AUC) and H1RO was examined using Pearson's correlation test. A probability of P < 0.05 was considered statistically significant. All statistical examinations were performed using SPSS for Windows 13.0.1 (Japanese version).


Plasma concentrations of bepotastine and diphenhydramine

Mean plasma concentrations and AUCs of bepotastine and diphenhydramine are shown in Table 1. The peak mean plasma concentration of bepotastine ranged from 60 to 180 min post administration because, in five of the eight subjects, it peaked at 60 min post administration, whereas in the other three subjects it peaked at 120 or 180 min post administration. Mean plasma concentration of diphenhydramine was maximal from 120 to 180 min post administration (Table 1).

Subjective sleepiness

Results of mean subjective sleepiness are shown in Figure 3. Mean subjective sleepiness of diphenhydramine peaked at 120–180 min post administration and that of bepotastine at 120 min post administration. Both LARS and SSS manifested similar patterns. Multiple comparison following repeated anovademonstrated that subjective sleepiness following diphenhydramine administration was significantly stronger (P < 0.001) than that of both bepotastine and placebo at 120 and 180 min post administration, and that of bepotastine was not significantly different from that of placebo (Figure 3).

Figure 3.

Subjective sleepiness evaluated using the Line Analogue Rating Scale (LARS). Eight healthy subjects were studied following oral administration of bepotastine (BEP, 10 mg), diphenhydramine (DIP, 30 mg) or placebo (PLA). *P < 0.001 by anova followed by multiple comparison with Bonferroni correction. Error bars represent interindividual variability (SEM). BEP, (inline image); DIP, (▪); PLA, (▴)

Brain distribution of 11C-doxepin

Radioactivity distribution patterns of 11C-doxepin are shown in Figure 2A. The mean BPR image, averaged from eight subjects, following bepotastine treatment was similar to that following placebo treatment in an individual subject. Namely, high radioactivity was observed in ACG, PCG, PFC, OFC, IC, TC, PC, OC and SMC following both treatments, whereas the radioactivity distribution pattern following diphenhydramine treatment was much lower than that following the other two treatments (Figure 2A).

Comparison of parametric BPR images (bepotastine vs. diphenhydramine)

Using SPM2 on a voxel-by-voxel basis, parametric brain BPR images following bepotastine or diphenhydramine treatment were statistically compared with those following placebo treatment. In Figure 2B, red areas show brain regions where BPRs were significantly lower (P < 0.001) following diphenhydramine treatment than those following placebo treatment (Figure 2B, right; Table 3). ACG, PFC and TC demonstrated significantly lower BPR after diphenhydramine treatment than after placebo treatment (Table 2). On the other hand, SPM analysis showed no brain areas where BPR was significantly lower after bepotastine treatment than after placebo treatment (Figure 2B, left).

Table 3. 
Regions with significantly lower specific binding following diphenhydramine treatment compared with those following placebo treatment
RegionsBrodmann's areaHemispherex, y, z (mm)Cluster sizeT-valueZ-valueP-value
  1. Cluster size is represented by the number of voxels (voxel size: 2.0 × 2.0 × 2.0 mm3). S.C., the same cluster as above. Results are not corrected for multiple comparisons.

Superior temporal gyrus22R54–56 1629 54311.556.41<0.001
Medial temporal gyrus21L−58-12-12S.C.11.416.37<0.001
Precuneus7R26–58 48S.C.11.086.29<0.001
Medial temporal gyrus21L−48 4–32328.145.41<0.001
Superior frontal gyrus10L−24 60 121047.535.18<0.001
Superior frontal gyrus8L−28 24 50387.075<0.001
Medial frontal gyrus10R28 54 1413428.535.55<0.001
Medial frontal gyrus9R42 6 40S.C.8.465.52<0.001
Medial frontal gyrus6R24 4 58S.C.8.985.7<0.001
Inferior frontal gyrus46L−40 44 12388.965.69<0.001
Table 2. 
Binding potential ratios and histamine H1 receptor occupancies in placebo, bepotastine and diphenhydramine conditions
  1. Results of statistical evaluation (P-values) are shown in Figure 3. BEP, bepotastine; BPR, binding potential ratio; C.E., correlation efficient of H1RO to plasma concentration of each antihistamine; 95% CI, 95% confidence interval; DIP, diphenhydramine; H1RO, histamine H1 receptor occupancy; Pla, placebo; SD, standard deviation.

Anterior cingulate gyrus (ACG)0.780.220.660.110.330.090.28, 0.640.24–0.4214.311.957.78.933.1, 53.80.54−0.30
Posterior cingulate gyrus (PCG)0.850.180.710.090.400.090.28, 0.610.18–0.4314.812.352.09.724.1, 50.20.51−0.57
Prefrontal cortex (PFC)0.650.160.550., 0.510.20–0.3714.212.259.312.231.7, 58.50.66−0.28
Orbitofrontal cortex (OFC)0.600.200.500., 0.510.18–0.3515.411.860.56.934.1, 56.2−0.01−0.28
Insular cortex (IC)0.800.180.660.090.360.090.29, 0.570.21–0.3715.99.954.08.028.7, 47.40.52−0.29
Temporal cortex (TC)0.610.180.510., 0.480.21–0.2814.412.657.211.030.6, 55.00.04−0.20
Parietal cortex (PC)0.540.170.440., 0.430.17–0.3016.016.962.816.530.6, 62.90.340.16
Occipital cortex (OC)0.490.100.430., 0.310.11–0.2211.012.846.313.523.5, 47.00.310.06
Somatosensory cortex (SMC)0.440.140.360., 0.380.06–0.2916.515.957.621.618.9, 63.30.78−0.18
Mean0.640.140.540.120.280.07  14.71.656.45.0 0.48−0.19

ROI-based comparison of BPR and H1RO

BPRs in H1R-rich regions such as ACG, PFC, IC, TC, PC and OC were evaluated based on ROI analysis (Figure 4A). BPRs following bepotastine treatment were only slightly different from those following placebo treatment. However, BPRs following diphenhydramine treatment were significantly lower than those following placebo or bepotastine treatment (P < 0.001) for all cortical regions. In the thalamus, striatum, pons and midbrain, there was no significant difference in BPRs between diphenhydramine, bepotastine and placebo treatments.

Figure 4.

Region of interest (ROI)-based analyses of binding potential ratios (BPR) (A) and histamine H1 receptor occupancy (H1RO) (B) in the cortex. ROI measurements were performed in the anterior and posterior cingulate gyri (ACG and PCG, respectively), prefrontal cortex (PFC), orbitofrontal cortex (OFC), insular cortex (IC), temporal cortex (TC), sensorimotor cortex (SMC), parietal cortex (PC) and occipital cortex (OC) after treatments with placebo (PLA), bepotastine (BEP) and diphenhydramine (DIP). Comparison of BPRs shows differences in the sedative properties of the three drugs (A). H1ROs due to BEP and DIP are shown, taking H1RO due to placebo as 0% (B). *P < 0.001, anova followed by multiple comparison with Bonferroni correction. Error bars represent interindividual variability (SEM). PLA, (□); BEP, (inline image); DIP, (▪)

H1RO following diphenhydramine or bepotastine treatment was also calculated using H1RO after placebo treatment as baseline (0%) (Figure 4B). Cortical mean H1RO following bepotastine treatment was approximately 14.7% and that following diphenhydramine treatment was approximately 56.4%. H1ROs following bepotastine treatment were significantly lower than those following diphenhydramine treatment (P < 0.001) in all cortical regions. These data demonstrate that BPR following bepotastine treatment is substantially higher than that following diphenhydramine treatment in all cortical regions studied.

Relationships between subjective sleepiness, plasma drug concentration and H1RO

Subjective sleepiness (AUC of LARS curve) did not significantly correlate to plasma concentration of bepotastine (r = 0.04), but did correlate well to plasma concentration of diphenhydramine (r = 0.72). H1ROs following bepotastine administration significantly correlated to plasma concentration of bepotastine in ACG, PCG, PFC, IC and SMC, whereas those following diphenhydramine administration were all negatively correlated (Table 2). A similar trend was observed between cortical mean H1RO and plasma concentration of both antihistamines (Figure 5A,B). However, when the baseline data are plotted together, mean H1RO due to diphenhydramine tended to increase rapidly with diphenhydramine concentration, whereas that due to bepotastine gradually increased with bepotastine concentration (Figure 5A,B). H1RO following bepotastine administration did not significantly correlate to subjective sleepiness (r = 0.01) (Figure 5D), whereas that following diphenhydramine administration negatively correlated to subjective sleepiness when the baseline data were plotted together (Figure 5C).

Figure 5.

Relationship between mean H1 receptor occupancy (H1RO), plasma concentration and subjective sleepiness following administrations of bepotastine and diphenhydramine. Relationship between mean H1RO (across brain regions) and plasma concentrations was examined for diphenhydramine (A) and bepotastine (B). Plasma concentrations of the two antihistamines are presented as area under the curve (AUC). H1RO of diphenhydramine rapidly increases with plasma concentration, whereas H1RO of bepotastine gradually increases with plasma concentration. Error bars represent intraindividual variability (SD). Dotted curves reflect the estimated curves of relationship between plasma concentration and the H1RO analysed by the Michaelis–Menten equation (A and B). Subjective sleepiness (presented as AUC of line-analogue rating scale curve: AUC_LARS) demonstrates mild correlation to mean H1RO due to diphenhydramine (C), whereas subjective sleepiness due to bepotastine demonstrates no correlation to mean H1RO due to bepotastine (D)


Recently, molecular imaging techniques have been actively applied in clinical pharmacology research [1, 8–10, 37]. One of the primary aims of such work is to evaluate the relationship between doses of therapeutic drugs and their adverse reactions in terms of changes in ‘endophenotypes’ in the brain [38]. The primary aim of the present study was to clarify whether bepotastine belongs to second-generation antihistamines in terms of H1RO. Another aim was to determine whether bepotastine should be classified as a truly nonsedative or mildly sedative antihistamine, as previously discussed in the Consensus Group on New Generation Antihistamines (CONGA) [3].

In the present study, H1ROs following a single oral administration of bepotastine (10 mg) or diphenhydramine (30 mg) were 14.7 and 56.4%, respectively. The relatively high H1RO due to diphenhydramine agrees with those of other reported first-generation antihistamines [1, 6, 38]. Single oral administration of d-chlorpheniramine (2 mg) achieved 50–77% H1RO, and this high H1RO was associated with high prevalence of sleepiness and cognitive decline [6, 8, 39]. On the other hand, bepotastine's low H1RO value suggests that this antihistamine can be classified into the category of second-generation antihistamines [6, 8, 9, 39–42, 38].

Thus, whether bepotastine can be classified as truly nonsedative or not is an additional point of clinical importance. In the present study, subjective sleepiness following bepotastine administration (10 mg) was negligible (Figure 3). This finding seems to be in accordance with that of Takahashi and colleagues [18], where bepotastine (10 mg) induced slight subjective sedation among healthy volunteers. In their study, the best performance was achieved following bepotastine treatment in a word-processing task compared with those following cetirizine (10 mg), fexofenadine (60 mg) or olopatadine (5 mg) treatment [18]. Another study [11] has demonstrated that results of serial arithmetic tests following oral administration of bepotastine (2.5, 5, 10 and 20 mg) showed no significant differences from those following placebo treatment. These clinical studies tried to demonstrate the nonsedative property of bepotastine, but they failed to include an active placebo (sedative antihistamine) to prove sensitivity of the selected tasks as recommended in the CONGA guideline [3]. An additional animal study [19] has demonstrated that bepotastine manifested a sedative profile similar to that of cetirizine and terfenadine. As a whole, these results suggest the low liability of bepotastine to produce sedative side-effects at a therapeutic dose of 10 mg. Considering the significant correlation between H1RO and plasma drug concentration (AUC) found in the present study (Figure 5B), bepotastine may be classified as a mildly sedative antihistamine.

In general, H1RO is used as an index of BBB permeability, but it could also be affected by gut absorption that can raise plasma drug concentration. At the molecular level, variation in BBB permeability has been determined by factors such as lipophilicity, molecular size and different actions of various drug transporters including P-glycoprotein (P-gp), an efflux pump expressed in capillary endothelial cells in the BBB [10]. Many lipophilic first-generation antihistamines are absorbed in full amount in the gut and can freely enter the brain tissue, whereas gut absorption and brain penetration of second-generation antihistamines tend to be reduced. For fexofenadine, a substrate of P-gp, both gut absorption and BBB permeability are highly reduced because of its low membrane permeability and high action of P-gp. For bepotastine, also a substrate of P-gp, the same extent of reduction as that of fexofenadine is observed [43]. The gradual increase in H1RO with plasma bepotastine concentration, suggesting its relatively high membrane permeability (Figure 5B), may be associated with the action of P-gp in the BBB. On the other hand, gut absorption of bepotastine tends to be much higher than that of fexofenadine, presumably because of bepotastine's higher membrane permeability in the upper part of the small intestine where P-gp expression is low [43]. Such difference is one of the possible reasons for explaining the difference between bepotastine and fexofenadine. Another reason is that fexofenadine is transported not only by P-gp but also by the organic anion protein transporter family, further reducing its absorption in the gut [44, 45].

The limitations of the present method are as follows. The short scanning PET protocol would be useful especially in conducting a placebo-controlled crossover study because of the enormous mental and physical stress of volunteers observed in a previous study, where they were requested to complete four sets of 100-min-long PET examinations [34]. It is therefore important to develop a short scan protocol, although such protocols would include a larger amount of noise. Users of the PET system should consider these limitations. We note here that not exact values but, rather, approximations of DV and BP were measured in this study.

We were interested in the effects of an acute single dose of an antihistamine, and we planned to start PET scanning at the time point near Tmaxfor each drug, but the tracer injection for diphenhydramine condition seemed to be slightly early since the plasma concentration of diphenhydramine was still increasing (Table 1). Therefore, H1RO due to diphenhydramine might give slightly different values when measured in the equilibrium state. This might be the reason why H1RO following diphenhydramine treatment did not correlate positively to the plasma concentration of diphenhydramine, contrary to our expectation based on a previous study using d-chlorpheniramine [8].

In summary, we have examined H1RO of bepotastine (10 mg; 14.7%) and compared it with that of diphenhydramine (30 mg; 56.4%) given as a single oral administration, using PET in a placebo-controlled crossover study. It is suggested that oral administration of bepotastine (10 mg), with its low H1RO and minimal sedation effects, is useful for treating various allergic disorders. As for the dose dependency of its sedative effects, another cognitive study involving an active placebo is needed in order to draw a definitive conclusion. The dose dependency of H1RO should also be examined by PET measurements at higher doses.

This work was in part supported by Grants-in-Aid for Scientific Research (nos. 17390156 for K.Y. and 16790308 for M.T.) from the Japan Society of Promotion of Science (JSPS) and the Ministry of Education, Culture, Sports, Science and Technology in Japan, as well as by a grant from the Japan Society of Technology (JST) on research and education in ‘molecular imaging’. We thank the volunteers of the PET study and Mrs Kazuko Takeda for her care of them. We also thank TANABE R&D SERVICE Co., Ltd (Osaka, Japan) for technical support in measuring the plasma concentration of antihistamines.