Evaluation of dopamine D3 receptor occupancy by blonanserin using [11C]-(+)-PHNO in schizophrenia patients

Rationale Unlike other antipsychotics, our previous positron emission tomography (PET) study demonstrated that a single dose of blonanserin occupied dopamine D3 as well as dopamine D2 receptors in healthy subjects. However, there has been no study concerning the continued use of blonanserin. Objectives We examined D2 and D3 receptor occupancies in patients with schizophrenia who had been treated with blonanserin. Methods Thirteen patients with schizophrenia participated. PET examinations were performed on patients treated with clinical dosage of blonanserin or olanzapine alone. A crossover design was used in which seven patients switched drugs after the first scan, and PET examinations were conducted again. D2 and D3 receptor occupancies were evaluated by [11C]-(+)-PHNO. We used nondisplaceable binding potential (BPND) of 6 healthy subjects which we previously reported as baseline. To consider the effect of upregulation of D3 receptor by continued use of antipsychotics, D3 receptor occupancy by blonanserin in seven subjects who completed 2 PET scans were re-analyzed by using BPND of olanzapine condition as baseline. Results Average occupancy by olanzapine (10.8 ± 6.0 mg/day) was as follows: caudate 32.8 ± 18.3%, putamen 26.3 ± 18.2%, globus pallidus − 33.7 ± 34.9%, substantia nigra − 112.8 ± 90.7%. Average occupancy by blonanserin (12.8 ± 5.6 mg/day) was as follows: caudate 61.0 ± 8.3%, putamen 55.5 ± 9.5%, globus pallidus 48.9 ± 12.4%, substantia nigra 34.0 ± 20.6%. EC50 was 0.30 ng/mL for D2 receptor for caudate and putamen (df = 19, p < 0.0001) and 0.70 ng/mL for D3 receptor for globus pallidus and substantia nigra (df = 19, p < 0.0001). EC50 for D3 receptor of blonanserin changed to 0.22 ng/mL (df = 13, p = 0.0041) when we used BPND of olanzapine condition as baseline. Conclusions Our study confirmed that blonanserin occupied both D2 and D3 receptors in patients with schizophrenia.


Introduction
The dopamine D 2 receptor family contains 3 subtypes (D 2 , D 3 , and D 4 ), and they are known as the D 2 -like receptor family. In terms of the treatment of schizophrenia, D 2 receptor has been thought to be strongly associated with the pathology of schizophrenia and be a major target for its treatment. Dopamine D 3 receptor has similarities to the other members of the D 2 -like receptor family, but D 3 receptor has very high affinity for dopamine and modulates dopamine release as an autoreceptor (Gross and Drescher 2012). Dopamine D 3 receptors are predominantly located in the ventral striatum, thalamus, and hippocampus, which are important for psychotic symptoms and are thought to modulate normal dopaminergic function and cognition (Maramai et al. 2016). The results from PET studies with [ 11 C]-(+)-PHNO indicated that 100% of the signal in the substantia nigra (SN), 67% in the globus pallidus (GP), and 26% in the ventral striatum represent D 3 receptor sites (Searle et al. 2010). The distribution of D 3 receptor in the limbic areas indicated that D 3 receptor might regulate motivation and reward-related behavior (Leggio et al. 2013).
Selective D 3 receptor antagonists affect the firing of dopaminergic neurons in the ventral striatum in a manner similar to atypical antipsychotics, and they enhance dopamine and acetylcholine release in the prefrontal cortex (Millan et al. 2008). It has also been indicated that D 3 receptor antagonists can inhibit extrapyramidal symptoms and produce neither anhedonia nor metabolic adverse effects, mainly based on evidence from rodent studies (Richtand 2006;Young et al. 2012). D 3 receptor antagonists can improve a series of social and cognitive behaviors in rodents, including executive functions, which are particularly impaired in patients with schizophrenia, while D 2 antagonists do not have this effect (Gross et al. 2013). Since dopaminergic hypofunction in the prefrontal cortex has been implicated in the pathogenesis of negative symptoms (Davis et al. 1991) and cognitive dysfunctions of schizophrenia (Sawaguchi 2000), these findings led to the theoretical treatment model of D 3 receptor antagonism being a valuable approach for the treatment of schizophrenia (Maramai et al. 2016). Thus, it seems worth verifying whether D 3 receptor antagonism can improve the negative symptoms and cognitive deficits of schizophrenia.
Many antipsychotics have been reported with Ki values for D 2 receptors not differing much from those for D 3 receptors (McCormick et al. 2013), but occupancy of D 3 receptors is moderately less than that of D 2 receptors. It has been reported by a positron emission tomography (PET) study with [ 11 C]-(+ )-PHNO that several antipsychotics (i.e., clozapine, risperidone, olanzapine) did not decrease, or even increased the in vivo nondisplaceable binding potential (BP ND ) of D 3 receptors in human brain (Graff-Guerrero et al. 2009). These findings suggested that these antipsychotics hardly occupied D 3 receptors in a clinical setting. [ 11 C]-(+)-PHNO gives a mixed D 2/3 signal composed of differing D 2 and D 3 proportions, and therefore previous studies measured BP ND of D 3 receptors in D 3 -receptor rich-regions. Another study also reported that chronically administered antipsychotics (i.e., clozapine, olanzapine, haloperidol) showed lower selectivity for D 3 compared with D 2 receptors ex vivo than in vitro in rat brain (McCormick et al. 2010).
Blonanserin is a second-generation antipsychotic drug developed in Japan, and it is currently being used as a therapeutic agent for schizophrenia in Japan, South Korea, and China. Comparative studies with other antipsychotic drugs have also been carried out, suggesting the possibility of this drug contributing to the improvement of cognitive impairments and negative symptoms of mental disorders (Murasaki 2016;Kishi et al. 2019). Blonanserin reportedly occupied a D 3 -rich region (i.e., cerebellum lobes 9-10) similarly to a D 2 -rich region (i.e., striatum) in rat brain, while risperidone, olanzapine, and aripiprazole did not (Baba et al. 2015). We recently examined the occupancy of D 2 and D 3 receptors by blonanserin in healthy subjects (Tateno et al. 2018). Using [ 11 C]-(+)-PHNO and PET, we demonstrated that a single dose of 12 mg of blonanserin occupied D 3 receptors to the same degree as D 2 receptors (i.e., EC 50 for the D 2 -rich region was 0.39 ng/mL and for the D 3 -rich region was 0.40 ng/mL) (Tateno et al. 2018). This finding led us to suggest the possibility that some of the pharmacological effect of blonanserin in schizophrenia patients might be mediated via D 3 receptor antagonism. However, the result from the single-dose administration of blonanserin in healthy subjects may not reflect actual clinical practices as patients obtained antipsychotic effects by its continuous administration. Therefore, it is important to confirm how the continued use of blonanserin occupied D 3 receptor of patients with schizophrenia in a clinical setting.
We hypothesized that blonanserin would occupy D 3 receptor to the same degree as D 2 receptor in patients with schizophrenia, in a manner similar to healthy subjects. In the present study, we evaluated both D 2 and D 3 receptor occupancy by blonanserin in patients with schizophrenia and compared the results with those by olanzapine, which has been demonstrated as not occupying D 3 receptor (Mizrahi et al. 2011).

Subjects and study design
We selected a group of patients, aged 20 to 70 years, who met the criteria of Diagnostic and Statistical Manual of Mental Disorders (DSM)-IV for schizophrenia. Inclusion criteria were as follows: (1) schizophrenia patients treated with blonanserin or olanzapine alone for 4 weeks or more, and those who did not change their dose for at least 2 weeks; (2) subjects who agreed to change from one drug to the other, (3) subjects who scored less than 120 on the Positive and Negative Syndrome Scale (PANSS15) at screening. Exclusion criteria were as follows: (1) subjects with past or current serious medical illness and/or organic brain diseases, (2) subjects with contraindication for the use of magnetic resonance imaging (MRI), (3) subjects with contraindication for blonanserin and olanzapine, (4) subjects treated with electroconvulsive therapy within 3 months before the screening, (5) subjects taking tandospirone at the time of screening, as there was a report that buspirone, which is in the same drug family, occupies D 3 receptor (Le Foll et al. 2016), and (6) subjects who were judged to be unsuitable for participation in this study. We allowed concomitant drugs such as b e n z o d i a z e p i n e s , a n t i h y p e r t e n s i v e d r u g s , a n d antiparkinsonian drugs that do not act on dopamine.
This study was designed as an open-label protocol. PET examination was performed on subjects who had been treated with blonanserin or olanzapine. A crossover design was then used in which patients switched drugs after the first scan, and PET examination was performed again after 2 weeks or longer. The doses of both drugs were within their clinical dose range. The mean dose of olanzapine was 10.8 ± 6.0 (range: 2.5-20) mg/day, and that of blonanserin was 12.8 ± 5.6 (8-24) mg/day. After complete explanation of the study, written informed consent was obtained from all participants. This study was approved by the institutional review board of Nippon Medical School Hospital, Japan.
Thirteen subjects participated in this study. The patients' characteristics are listed in Table 1. None of them worsened greatly after the medication change. Seven patients completed 2 PET scans, five of whom took olanzapine first and two were given blonanserin first. Six subjects were examined only by the 1st PET scan, 3 of whom participated in the PET scan with only blonanserin, and 3 with only olanzapine. Three of them felt uneasy about the drug change after the 1st PET scan and decided to continue with the initial drug, and the other 3 discontinued because of their clinical condition before the 2nd PET scan; one was diagnosed with diabetes , one complained of insomnia, and one was stopped due to extrapyramidal symptoms.

MRI procedures
MRI of the brain was acquired with 1.5 T MR imaging, Intera 1.5 T Achieva Nova (Philips Medical Systems, Best, Netherlands) as proton density image (echo time = 17 ms; repetition time = 6000 ms; field of view = 22 cm, 2-dimensional, 256 × 256; slice thickness = 2 mm; number of excitations = 2). These images were used for analysis of the PET scans.

Measurement of plasma concentrations of blonanserin and olanzapine
Venous blood samples were taken just before the PET scans, collected in tubes containing EDTA-2Na, and centrifuged at 3000 rpm for 10 min at 4°C. Separated plasma samples were stored at − 80°C until analysis. The plasma concentration of blonanserin was measured by validated method using highperformance liquid chromatography-tandem mass spectrometry with a target lower quantification limit of 0.001 ng/mL (Sekisui Medical Co., Ltd., Tokyo, Japan). The plasma concentration of olanzapine was measured by validated method using high-performance liquid chromatography-tandem mass spectrometry with a target lower quantification limit of 0.0001 ng/mL (Sumika Chemical Analysis Service Co., Ltd., Osaka, Japan).

PET data analysis
MR images were co-registered to summated PET images with the mutual information algorithm using PMOD (version 3.4; PMOD Technologies Ltd., Zurich, Switzerland). Regions of interest (ROIs) were defined for the caudate, putamen, globus pallidus, substantia nigra, and cerebellum in accordance with Tziortzi's study (Tziortzi et al. 2011). We defined the caudate and putamen as D 2 -rich regions and the substantia nigra and globus pallidus as D 3 -rich regions, based on Searle's study with [ 11 C]-(+)-PHNO (Searle et al. 2010). ROIs were drawn manually on overlaid summated PET and co-registered MR images of each subject. By matching the targeted frame to the average of the first 10 frames (i.e., 0-10 min), motion corrections were conducted in all subjects.
Quantitative estimate of binding of [ 11 C]-(+)-PHNO was performed using a simplified reference tissue model (Lammertsma and Hume 1996), with the cerebellar cortex as reference region. We avoided cerebellum midline-structures because of measurable specific [ 11 C]-(+)-PHNO binding. This model has been validated to reliably estimate BP ND , which compares the concentration of radioligand in the receptor-rich region with the receptor-free region (Innis et al. 2007) for [ 11 C]-(+)-PHNO (Ginovart et al. 2007).
We used a 1-site binding model, the same as in a previous study . The relationship between plasma concentration and receptor occupancy was shown by the following equation: where C is the plasma concentration of drug, E max is the maximum occupancy, and EC 50 is the plasma concentration required to achieve 50% occupancy (Tateno et al. 2018;Graff-Guerrero et al. 2010). E max was fixed at 1 and EC 50 > 0, the same as in the previous occupancy studies (Tateno et al. 2018;Graff-Guerrero et al. 2010). Mizrahi et al. reported that continuous intake of atypical antipsychotic drugs upregulated D 3 receptors (Mizrahi et al. 2011). Upregulation of D 3 receptors in treated schizophrenia patients might increase BP ND , which induces the underestimation of occupancy of antipsychotics when using HV as baseline. To accurately compare D 3 receptor occupancy with D 2 receptor occupancy by blonanserin in consideration of the effect of the upregulation of D 3 receptors, we also calculated the D 3 receptor occupancy of blonanserin using individual BP ND of olanzapine as a baseline among 7 patients who were taking both blonanserin and olanzapine. The paired t test was used to statistically analyze the comparison between D 3 receptor occupancy of blonanserin by using BP ND of olanzapine as baseline and that of healthy control as baseline.

Results
The BP ND values of each of the ROIs by olanzapine and blonanserin are summarized in Table 1.

D 3 receptor occupancy by blonanserin using individual BP ND of olanzapine as baseline
We also calculated the D 3 receptor occupancy by blonanserin using individual BP ND of olanzapine condition as baseline. The results are shown in Table 3. The occupancy of D 3 was higher than when using the baseline BP ND in healthy volunteers (67.9 ± 11.8% versus 44.7 ± 16.6%) (df = 26, p = 0.0002). EC 50 of D 3 receptor occupancy was 0.22 ng/mL (df = 13, p = 0.0041, 95% CI [0.095-0.341]), which was close to that of D 2 receptor occupancy (Fig. 2).

Discussion
In this study, we confirmed that blonanserin indeed occupied D 3 receptors in the globus pallidus and substantia nigra, although to a lesser degree than D 2 receptors in the caudate nucleus and putamen, in patients with schizophrenia. On the other hand, olanzapine occupied 30% of the evaluation sites of D 2 receptor, but hardly those of D 3 receptor. These findings were consistent with a previous animal study and an in vivo human study (Baba et al. 2015;Graff-Guerrero et al. 2009).
The occupancy of D 3 receptor by blonanserin was a little lower than in a previous study of healthy volunteers (Tateno et al. 2018), but it was similar when recalculated using individual BP ND under olanzapine condition. First, upregulation of D 3 receptors in treated schizophrenia patients with antipsychotics has been thought to influence BP ND and the occupancy of antipsychotics (Graff-Guerrero et al. 2009;Mizrahi et al. 2011), and it might decrease the apparent D 3 receptor occupancy. Regarding the upregulation by antipsychotics, it was earlier reported that occupancy of the globus pallidus by clozapine, olanzapine, and risperidon was − 70.7 ± 86.5% (Graff-Guerrero et al. 2009). Another study reported that occupancy of the globus pallidus by olanzapine and risperidone was − 50.28 ± 29.37% (Mizrahi et al. 2011). In the current study, we also used BP ND of D 3 receptors of patients under olanzapine treatment as a baseline for the calculation of D 3 receptor occupancy to reduce the influence of upregulation. Although it was expected to show a similar value to those of previous studies, our result was that the D 3 receptor occupancy by blonanserin (34.0 to 48.8%) was slightly lower than the D 2 receptor occupancy (55.5 to 61.0%) using HV as baseline. This result seemed to be due to the influence of upregulation. Second, schizophrenia patients showed increased [ 11 C]-(+)-PHNO binding compared to healthy subjects even if they were untreated (Weidenauer et al. 2020). For these reasons, individual baseline values would be more desirable. EC 50 of D 3 receptor by blonanserin changed from 0.70 to 0.22 ng/mL when switching baseline BP ND from the average of healthy volunteers to the individual patient's value with olanzapine in the 7 patients who had completed the 2 PET scans. This study assumes that the degree of upregulation was similar with olanzapine and blonanserin. This value was lower than EC 50 of D 2 receptor (0.40 ng/mL) for blonanserin in the same 7 subjects by using HV as baseline. Thus, our results confirmed that blonanserin occupied D 3 receptor as well as D 2 receptor in patients with schizophrenia. Blonanserin might be an important target for further studies regarding the therapeutic efficacy of D 3 receptor blockade by antipsychotic drugs.
In this study, the degree of D 3 receptor occupancy in the substantia nigra by olanzapine in 8 patients with  (Mizrahi et al. 2011). To our knowledge, this is the first report regarding the upregulation of dopamine D 3 receptor by olanzapine; however, the evaluation of a large number of subjects and/or using same subjects both before and after its administration will be needed for clarification.
We should acknowledge several limitations to this study. First, our sample size was small. Furthermore, 6 of the 13 subjects underwent only one PET scan. Therefore, additional studies including larger numbers of subjects and longitudinal designs are essential for the generalization of our findings. Second, we used a younger-aged control group compared to the patients. BP ND of D 2 receptor was negatively correlated with age in the caudate, while that of D 3 receptor was not correlated with age in the globus pallidus and substantia nigra (Nakajima et al. 2015). Our results of D 2 receptor using younger-aged controls for baseline might be influenced by an age effect if controls would be older. Third, we used all-male HVs, whereas 46% of the patients were female. In this regard, a previous study indicated that D 3 receptor differed between male and female rhesus monkeys (Martelle et al. 2014). Fourth, it is uncertain whether the degree of upregulation was similar or not between olanzapine and blonanserin. This study assumes that the degrees were comparable, although we could not estimate D 3 upregulation exactly as there was no drug-free baseline condition. Fifth, 7 of the 13 participants in this study were smokers while all HVs were non-smokers. We could not rule out the effects of smoking, as it has been shown to have an effect on the dopamine system (Le Foll et al. 2014), In conclusion, our study confirmed that continuous usage of blonanserin occupied dopamine D 3 receptors to the same degree as D 2 receptors in the brains of schizophrenia patients. By more discussions on the therapeutic effects of blonanserin, which is now known to clearly possess in vivo D 3 receptor antagonism, we may be able to consider the relevance of anti-dopamine D 3 receptor acivities as well as the therapeutic effects on cognitive impairments and negative symptoms of mental disorders.