Antipsychotic Treatment of Behavioral and Psychological Symptoms of Dementia (BPSD): Management of Extrapyramidal Side Effects

Antipsychotic drugs are often used for the treatment of behavioral and psychological symptoms of dementia (BPSD), especially psychosis and behavioral disturbances (e.g., aggression and agitation). They are prescribed alone or in conjunction with anti-dementia (e.g., anti-Alzheimer’s disease drugs) and other psychotropic drugs (e.g., antidepressants). However, antipsychotic drugs frequently produce serious extrapyramidal side effects (EPS) including Parkinsonian symptoms (e.g., bradykinesia, akinesia, tremor, and muscle rigidity). Therefore, appropriate drug choice and combination strategy are important in the treatment of BPSD. Among anti-Alzheimer’s disease drugs, cholinesterase inhibitors (ChEIs, e.g., donepezil and galantamine) have a propensity to potentiate EPS associated with antipsychotic treatment in a synergistic manner. In contrast, the NMDA receptor antagonist memantine reduces antipsychotic-induced EPS. Antidepressant drugs, which inhibit 5-HT reuptake into the nerve terminals, also synergistically augment antipsychotic-induced EPS, while mirtazapine (α2, 5-HT2 and 5-HT3 antagonist) reduces the EPS induction. Importantly, previous studies showed that multiple 5-HT receptors play crucial roles in modulating EPS associated with antipsychotic treatment. Specifically, activation of 5-HT1A receptors or blockade of 5-HT2, 5-HT3 and 5-HT6 receptors can alleviate EPS induction both by antipsychotics alone and by combined antipsychotic treatments with ChEIs or 5-HT reuptake inhibitors. In this article, we review antipsychotic use in treating BPSD and discuss the favorable drug selection in terms of the management of antipsychotic-induced EPS.


INTRODUCTION
Dementia is a neurodegenerative brain disorder with diverse clinical symptoms including cognitive impairment (e.g., memory loss and learning deficits) and non-cognitive disorders (e.g., behavioral and psychological deficits). Nearly 50 million patients worldwide develop dementia and this population is expected to exceed 130 million in 2050 (Prince et al., 2015;Jin and Liu, 2019). The global cost associated with dementia was about 1,000 billion dollars in 2015, and this continues to increase rapidly. There are numerous causes of dementia including Alzheimer's disease, cerebrovascular diseases, September 2019 | Volume 10 | Article 1045 Frontiers in Pharmacology | www.frontiersin.org Parkinson's disease, Lewy body disease, and mixed types, among which Alzheimer's disease is the most frequent (Lee et al., 2004;Prince et al., 2015;Sturm et al., 2018).
It is well known that antipsychotic drugs commonly cause serious extrapyramidal side effects (EPS) (e.g., bradykinesia, muscle rigidity, tremor, and akathisia) by blocking dopamine D 2 receptors in the striatum (Remington and Kapur, 1999;Kapur and Remington, 2001;Ohno et al., 2013;Ohno et al., 2015;Ohno, 2019). Antipsychotic-induced EPS often leads to suboptimal treatment of BPSD or treatment discontinuation. In addition, recent studies showed that cholinesterase inhibitors (ChEIs), licensed drugs for cognitive impairment due to Alzheimer's disease, potentiate EPS induction with antipsychotic treatments . It is therefore important to understand the mechanism underlying antipsychotics-induced EPS and antipsychotic drug interactions with other medications in the treatment of BPSD.
In this article, we review the pharmacological features of antipsychotic drugs, especially those related to EPS, and discuss the proper usage and selection of antipsychotics in treating BPSD in terms of EPS management.

ANTIPSYCHOTIC USE IN BPSD TREATMENT
Antipsychotic drugs are used to treat BPSD with a prescription rate of about 20-50% (Lee et al., 2004;Brimelow et al., 2019;Sturm et al., 2018). The target symptoms of antipsychotic drugs include agitation, aggression, psychosis, and inappropriate behaviors (Figure 1). None of the antipsychotics, except for haloperidol and risperidone in several countries, are approved to treat BPSD; therefore, these drugs are generally prescribed as off-label. Nonetheless, antipsychotic FIGURE 1 | Behavioral and psychological symptoms of dementia (BPSD) in Alzheimer's disease. Patients with dementia show not only core symptoms of dementia (e.g., memory loss and learning deficits), but also various associated symptoms including BPSD (e.g., psychosis, behavioral excitation, and mood disorders). Cognitive impairment in Alzheimer's disease is usually treated with cognitive enhancers such as the cholinesterase inhibitors (ChEIs, e.g., donepezil, galantamine, and rivastigmine) and the NMDA antagonist (e.g., memantine). In the treatment of BPSD, antipsychotic drugs are used for psychosis and behavioral disturbances, and antidepressants for depressive mood.
Antipsychotic drugs commonly possess dopamine D 2 blocking actions. It is known that D 2 receptor blockade by antipsychotics in the cortico-limbic regions (e.g., nucleus accumbens) contributes to antipsychotic activities, which alleviates psychosis (e.g., hallucinations and delusions) and behavioral excitation (e.g., agitation, aggression and hyperactivity) (Figure 2). However, it should be noted that all antipsychotic drugs frequently cause extrapyramidal motor disorders due to the striatal D 2 receptor blockade, which disrupts the effective treatment of BPSD.
Antipsychotic drugs are generally classified into two groups, typical and atypical (Ohno et al., 1997;Ohno et al., 2012). Typical antipsychotics are the classic standard drugs and frequently cause severe EPS. Based on their chemical structures, they are grouped into several classes, phenothiazines (e.g., chlorpromazine and fluphenazine), butyrophenones (e.g., haloperidol and spiperone), benzamides (e.g., sulpiride and tiapride), and others. On the other hand, atypical antipsychotics were developed as second generation, and are generally less potent than typical ones in inducing EPS (Figures 2 and 3). These include the serotonin and dopamine antagonists (SDAs) with potent blocking action for 5-HT 2 receptors, the multiple-acting receptor targeted antipsychotics (MARTAs) and the dopamine D 2 partial agonists (Ohno et al., 2012). Besides reduced EPS, these drugs were originally expected be superior to typical antipsychotics in terms of their efficacy to treat negative symptoms (e.g., apathy and emotional withdrawal) (Figure 2). However, comprehensive clinical studies including the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) and European First-Episode Schizophrenia Trial (EUFEST), revealed no clear advantages of atypical over typical drugs in terms of efficacy (Lieberman et al., 2005;Keefe et al., 2007;Davidson et al., 2009). Nonetheless, due to the reduced side effect profile, atypical antipsychotics are widely used as a first line drug in BPSD treatment as well as schizophrenia treatment.

Clinical Symptoms
Major EPS symptoms associated with antipsychotic treatment of BPSD include Parkinsonian symptoms, akathisia, and dystonia. Tardive dyskinesia (repeated abnormal involuntary movements) is another antipsychotic-induced EPS, but is rare during the relatively short-term BPSD treatment as it is a chronic side effect associated with long-term antipsychotic treatment and usually appears upon the cessation of treatment.

Parkinsonian Symptoms
Antipsychotic-induced Parkinsonian symptoms are involuntary movement disorders including bradykinesia, tremor and muscle rigidity (Samii et al., 2004;Haddad and Dursun, 2008;Ohno et al., 2015). Parkinsonian symptoms usually occur in a few weeks after starting the antipsychotic treatment. Bradykinesia refers to reduced motor activity and slowing movements, which leads to akinesia in more severe cases. Tremor is an involuntary, rhythmic muscle contraction and relaxation (oscillation or twitching movements), affecting the hands, feet and head especially during resting state. In addition, affected patients often exhibit a stooped posture with increased muscle tone (rigidity) and a slow gait without arm swing.

Akathisia
Patients with akathisia suffer from restlessness and repetitive movements of the legs and feet (Keefe et al., 2007;Haddad and Dursun, 2008). As a result, they cannot keep sitting and FIGURE 2 | Pharmacological actions of atypical antipsychotics. Likely typical antipsychotic drugs, D 2 blocking actions of serotonin and dopamine antagonists (SDA)type and multiple-acting receptor targeted antipsychotics (MARTA)-type antipsychotics ameliorate positive symptoms (e.g., hallucination, delusions, and excitation) in schizophrenia, but induce extrapyramidal side effects (EPS). On the other hand, SDA-type and MARTA-type antipsychotics show higher 5-HT 2 than D 2 binding affinities and possess potent 5-HT 2 antagonistic actions. The 5-HT 2 blocking activities of SDA-type antipsychotics ameliorate negative symptoms (e.g., apathy and social withdrawal) in schizophrenia and can reduce EPS. Thereby, overall EPS liability of SDA-type antipsychotics is lower than typical antipsychotics (D 2 antagonists).
frequently shift their body position. Akathisia usually appears soon after starting antipsychotics or after increasing the dose.

Dystonia
Dystonia causes sustained muscle contraction, often leading to postural distortion (Haddad and Dursun, 2008). Dystonia often attacks the neck muscles, tongue, trunk, and limbs. Acute dystonia usually appears in the first week after starting or increasing the dose of antipsychotics.

Neural Mechanism of EPS Induction
It is well known that antipsychotic-induced EPS are caused by the blockade of dopamine D 2 receptors in the striatum (caudateputamen) (Ohno et al., 1997;Ohno et al., 2013;Ohno et al., 2015;Ohno, 2019) (Figure 4). The GABAergic medium spiny neurons in the striatum receive excitatory glutamatergic inputs from the cerebral cortex and acetylcholinergic inputs from striatal interneurons. The medium spiny neurons also receive inhibitory dopaminergic inputs from the substantia nigra pars compacta (SNc) and express a high density of D 2 receptors . In addition, the dopaminergic neurons from the SNc also negatively regulate activities of the acetylcholinergic interneuron via D 2 receptors. Most antipsychotic drugs commonly act as dopamine D 2 receptor antagonists and activate the medium spiny neurons and acetylcholinergic interneurons in the striatum, eliciting various EPS symptoms (Ohno et al., 2013) (Figure 4).
To reduce EPS, a series of atypical antipsychotics, that show potent 5-HT 2 blocking activities have been developed in the last three decades (Ohno et al., 1997;Ohno et al., 2012) (Figures 2 and 3). These agents include risperidone, perospirone, olanzapine, quetiapine, lurasidone, and paliperidone, and they commonly exhibit higher 5-HT 2 than D 2 affinities. Since olanzapine and quetiapine also show high affinities for other multi-receptors (e.g., histamine H 1 , adrenergic α 1, and muscarinic acetylcholine (mACh) receptors), these drugs are sometimes called as MARTAs and distinguished from SDAs.
It is well documented that blockade of 5-HT 2 receptors attenuates antipsychotic-induced EPS associated with the striatal D 2 receptor blockade (Figure 2). 5-HT 2 receptors are located on nerve terminals and cell bodies of dopaminergic neurons in the striatum and the SNc, respectively, and inhibit dopaminergic neuron activities (Ohno et al., 1997;Ohno et al., 2012;Ohno et al., 2013). It is therefore proposed that blockade of 5-HT 2 receptors relieves 5-HT 2 receptor-mediated inhibition of dopamine release in the striatum and of dopamine neuron firing in the SNc, which leads to alleviation of EPS (Figure 5) (Remington and Kapur, 1999;Kapur and Remington, 2001). In fact, blockade of 5-HT 2 receptors can reverse various responses of striatal neurons to FIGURE 3 | Classification and characteristics of atypical antipsychotic drugs. Figure shows chemical structures of atypical antipsychotics with their receptor binding profiles (affinities). Most atypical antipsychotics possess potent 5-HT 2 blocking actions and act as serotonin and dopamine antagonists (SDAs). Among SDAs, clozapine derivatives (e.g., olanzapine and quetiapine) show various actions on other receptors than D 2 and 5-HT 2 receptors, including histamine H 1 , adrenergic α 1, and muscarinic blocking actions. Thereby, they are sometimes called as multi-acting receptor-targeted antipsychotics (MARTAs) and differentiated from SDAs (e.g., risperidone, perospirone, and lurasidone). Although aripiprazole possess moderate 5-HT 2 blocking activities, it primarily acts as a dopamine D 2 partial agonist. Furthermore, several atypical antipsychotics have own characteristics such as 5-HT 1A partial agonistic actions for perospirone, lurasidone and aripiprazole, 5-HT 6 blocking actions for olanzapine and quetiapine, and 5-HT 3 blocking actions for olanzapine.
antipsychotics (D 2 receptor blockade), such as the enhancement of acetylcholine (ACh) release, the increase in metabolic turnover rate of dopamine and the induction of Fos protein expression, in the striatum Ohno et al., 1997;Ohno et al., 2013).
EPS induction was also reduced by microinjection of SB-258585 into the striatum, implying that blockade of the striatal 5-HT 6 receptors is at least partly involved in alleviating EPS. Since 5-HT 6 receptors positively regulate the neural activities of the striatal ACh interneurons (Bonsi et al., 2007), it is conceivable that 5-HT 6 antagonists reduce antipsychotic-induced EPS by inhibiting them (Figure 5).

EFFECTS OF ANTI-ALZHEIMER'S DISEASE DRUGS ON ANTIPSYCHOTIC-INDUCED EPS
Alzheimer's disease is the major component of elderly dementia. Since Alzheimer's disease accompanies the loss of ACh neurons (Fibiger, 1991;Silva et al., 2014), several ChEIs such as donepezil, galantamine, and rivastigmine, which can increase the ACh level by inhibiting cholinesterase, are widely used to treat the cognitive impairment in Alzheimer's disease. In addition, an NMDA receptor antagonist, memantine, is also used to alleviate the cognitive impairment. These anti-Alzheimer's disease drugs are often prescribed in combination with antipsychotic drugs which can reduce BPSD (Salamone et al., 2001;Kozman et al., 2006), giving greater efficacy than monotherapy (Schmitt et al., 2004).
Although information on the drug interactions between antipsychotic and anti-Alzheimer's disease drugs is limited, our previous study revealed that they markedly potentiated antipsychotic-induced EPS induction .
Specifically, donepezil and galantamine rarely induce EPS signs when taken alone; however, they markedly potentiated bradykinesia induction by low dose of haloperidol in a dose-dependent and synergistic manner (Figure 6). In addition, the bradykinesia potentiation by galantamine was significantly reversed by a 5-HT 1A agonist (8-OH-DPAT), a 5-HT 2 antagonist (ritanserin) and a 5-HT 6 antagonist (SB-258585) . These findings indicate that caution is needed in the combined usage of antipsychotics and ChEIs in BPSD treatment. Furthermore, antipsychotics that can stimulate 5-HT 1A receptors or antagonize 5-HT 2 and 5-HT 6 receptors appear favorable as an adjunctive therapy for BPSD. Interestingly, in contrast to ChEIs, memantine, which antagonizes NMDA receptors, attenuated antipsychotic-induced EPS (Figure 6). Therefore, it seems likely that memantine is more favorable than ChEIs in the combined therapy of BPSD with antipsychotics.
Precise mechanisms underlying the synergistic potentiation of EPS by ChEIs is still unknown. However, the action of antipsychotics on cholinergic interneurons in the striatum seems to be involved since the firing of striatal cholinergic interneurons is negatively regulated by dopaminergic neurons and is reportedly facilitated by antipsychotics, increasing the ACh release (Damsma et al., 1990;DeBoer and Abercrombie, 1996). Therefore, ChEIs may augment the induction of EPS more potently in the presence of antipsychotics than their monotherapy.
In contrast to SSRIs and TCAs, tetracyclic antidepressants (mirtazapine and mianserin) did not augment, but rather attenuated antipsychotic-induced EPS   (Figure 7). Thus, it seems likely that tetracyclic antidepressants are superior to SSRIs or TCAs in modulating EPS in combined treatment of BPSD with antipsychotics. Since the blockade of α 2 receptors reportedly reduced antipsychotic-induced EPS , EPS reduction by tetracyclic antidepressants is probably due to the α2 blocking action in addition to their 5-HT 2 and 5-HT 3 blocking activities.
FIGURE 6 | Interactions between anti-Alzheimer's disease drugs and antipsychotics in induction of extrapyramidal side effects (EPS). Bradykinesia was estimated by the pole test, where mice were placed head-upward at the top of a pole (45 cm in height) and the time for mice to rotate downward (T turn ) and to descend to the floor (T descent ) was measured (Ohno et al., 2008a). Bradykinesia was evaluated as the prolongation of T turn or T descent Values. Although low dose (0.5 mg/kg) of haloperidol showed marginal effects in the pole test, combined treatment with cholinesterase inhibitors, donepezil, and galantamine, markedly potentiated haloperidolinduced bradykinesia in a synergistic manner. By contrast, the NMDA antagonist, memantine, significantly reduced bradykinesia induced by a high dose (1 mg/kg) of haloperidol. *P<0.05, **P,0.01; Significantly different from the control values. This figure is partly quoted and arranged from J. Pharmacol. Sci. 127, 439, 2015. September 2019 | Volume 10 | Article 1045 Frontiers in Pharmacology | www.frontiersin.org

DRUG SELECTION IN BPSD TREATMENT
We reviewed antipsychotic use in BPSD treatment focusing on EPS, the most frequent side effects associated with the striatal D 2 receptor blockade. Antipsychotic-induced EPS significantly disrupts activities of daily life and impairs the quality of life in the elderly patients with dementia. Therefore, information on the mechanisms and the drug interactions in modulating EPS induction are necessary to achieve proper pharmacotherapy of BPSD. In this regard, we should be very careful not only about EPS liability of antipsychotics by itself, but also about the interaction of antipsychotics with anti-Alzheimer's disease drugs and antidepressant drugs.
Atypical antipsychotics (e.g., SDAs, MARTAs, and D 2 partial agonists) are now the first line drug to treat psychosis and inappropriate behaviors (e.g., agitation and aggression) in patients with dementia. But, we should pay more attention to individual pharmacological characteristics of the atypical drug, especially their interactions with 5-HT receptor subtypes. Although most SDAs or MARTAs commonly possess high affinities to 5-HT 2 receptors, many atypical antipsychotics shows a differential binding profile each other, interacting with various monoamine receptors (Farah, 2005). In fact, olanzapine additionally show high affinities for 5-HT 3 and 5-HT 6 receptors and acts as antagonist (Bymaster et al., 2001). In addition to 5-HT 2 receptors, the SDA antagonist lurasidone also binds to 5-HT 1A receptors and acts as a partial agonist (Ishibashi et al., 2010). Furthermore, the dopamine D 2 partial agonist aripiprazole also binds to 5-HT 1A and 5-HT 2 receptors, and acts as a partial agonist and an antagonist, respectively (Stark et al., 2007). Since the actions of these agents with 5-HT receptor subtypes can reduce EPS caused by combined treatment of antipsychotics with anti-Alzheimer's disease drugs and antidepressants, they could be a favorable BPSD treatment in terms of EPS management.
Among anti-Alzheimer's disease drugs, the NMDA antagonist memantine appears superior to ChEIs in the combined BPSD therapy with antipsychotics as it attenuates antipsychotic-induced EPS. Likewise, the tetracyclic antidepressants (mirtazapine and mianserin) are recommended for combined use with antipsychotics to treat BPSD. Unlike 5-HT reuptake inhibitors (e.g., SSRIs, SNRI, and TCAs), these agents do not augment EPS induction, but alleviate antipsychotic-induced EPS, which is possibly by blocking α 2 , 5-HT 2 and 5-HT 3 receptors Ohno et al., 2011).

CLOSING REMARKS
This article provides information on the safe usage of antipsychotics in adjunctive therapy for BPSD in patients with dementia. The crucial roles of 5-HT receptors, especially 5-HT 1A , 5-HT 2 , 5-HT 3 , and 5-HT 6 receptors, in modulating antipsychoticinduced EPS were revealed. Although antipsychotic drugs are effective for psychosis, agitation, excitation, and abnormal behaviors, we should be very careful about drug selection FIGURE 7 | Interactions between antidepressants and antipsychotics in induction of extrapyramidal side effects (EPS). Bradykinesia was estimated by the pole test as described in Figure 6 legend. Although low dose (0.3 mg/kg) of haloperidol showed only weak effects in the pole test, combined treatment with selective serotonin reuptake inhibitors, fluoxetine and paroxetine, markedly potentiated haloperidol-induced bradykinesia in a synergistic manner. By contrast, the tetracyclic antidepressant mirtazapine, which possesses α 2 , 5-HT 2 and 5-HT 3 antagonistic actions, reduced bradykinesia induced by a moderate dose (0.5 mg/ kg) of haloperidol. *P<0.05, **P,0.01; Significantly different from the control values. This figure is quoted and arranged from Prog. Neuro-Psychopharmacol. Biol. Psychiatry 38, 252, 2012. in the combined use of antipsychotics with anti-Alzheimer's disease drugs or antidepressants. Specifically, ChEIs and 5-HT reuptake inhibitors (SSRIs, SNRI, and TCAs) markedly potentiate antipsychotic-induced EPS in a synergistic manner. In contrast, the NMDA antagonist (memantine) or the tetracyclic antidepressants (mirtazapine and mianserin) seem to be more suitable for adjunctive therapy of cognitive impairment and mood disorders of BPSD, respectively. Furthermore, antipsychotics which have 5-HT 1A agonistic actions or 5-HT 2 , 5-HT 3, and 5-HT 6 antagonistic actions appear to be useful for adjunctive BPSD treatment.

AUTHOR CONTRIBUTIONS
YO drafted the initial manuscript. All authors (YO, NK, SS) improved, contributed to and agreed on the final version of the manuscript.

FUNDING
This study was partly supported by a research grant from by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (YO:17K08324, SS:16K21501) and from the Smoking Research Foundation (YO).