Introduction

Antipsychotics are used to treat mental illnesses such as schizophrenia and other psychosis, as well as bipolar disorder and depression. These medications are divided into three groups: (1) first-generation antipsychotics (FGAs); (2) second-generation antipsychotics (SGAs); and (3) third-generation antipsychotics (TGAs) [1, 2]. The main difference between FGAs, SGAs, and TGAs is their pharmacological target. FGAs, such as haloperidol, chlorpromazine, and thioridazine, act mainly on the dopaminergic system as antagonists for the dopamine type 2 (D2) receptors [3]. They alleviate positive symptoms of schizophrenia; however, D2 blockade often induces numerous side effects, the most prominent are extrapyramidal symptoms (EPS) [3]. In addition, the blockade of D2 receptors causes an increase in prolactin levels that correlates with the dose [4]. SGAs, including risperidone and clozapine, have a higher affinity for serotonin receptors (5-HT) than D2 receptors [5]. Due to 5-HT2A/D2 antagonist properties, they are also called dopamine-serotonin antagonists. Moreover, SGAs exhibit an action on muscarinic cholinergic receptors (M3), histamine receptors (H1), and adrenergic receptors (α1 and α2) [6]. Although SGAs are associated with a lower risk of EPS they can cause metabolic effects, such as weight gain, diabetes mellitus, hyperlipidemia, QT modifications, and hyperprolactinemia [7]. TGAs, including aripiprazole (ARI), brexpiprazole, and cariprazine are partial dopamine receptor agonists and also act as antagonists or weak partial agonists on the serotonin receptors [1, 8,9,10].

ARI, apart from being described as an antipsychotic drug, is also a mood stabilizer. Antipsychotic medications are mainly used for treating schizophrenia; however, they are effective for other psychotic disorders and also other psychiatric disease entities, such as mania, bipolar affective disorder, depression, anxiety disorders, delusional disorders, irritability associated with autism, Tourette's syndrome, disorders associated with problems with impulse control, and behavioral disturbances in dementias or in children and adolescents. Short-acting ARI is also used intramuscularly for rapid tranquillization in acute agitation associated with schizophrenia or bipolar disorder [11, 12]. Detailed information about antipsychotic drug metabolism, mechanism of action, and adverse side effects is shown in Table 1.

Table 1 Characteristics of selected antipsychotic drugs
Full size table

Schizophrenia is a chronic, serious mental illness that affects about 1% of the world’s population. The pathomechanism of this disease is complex and still not sufficiently understood. It is plausible that genetic and environmental factors as well as other causes, such as brain chemistry, substance use, and autoimmune diseases or inflammation play a role in the risk of developing schizophrenia. The disease begins at a young age (usually before the age of 18) and causes different symptoms: (1) positive (e.g., hallucinations, delusions, disorganized behavior); (2) negative (e.g., social withdrawal, apathy, lack of energy, anhedonia, flattened affect); and (3) cognitive (e.g., memory and learning impairments or attention deficiencies) [9, 11]. Several SGAs are currently available for the treatment of schizophrenia including clozapine, olanzapine, and risperidone. However, these drugs can cause metabolic adverse effects, in turn, a TGA drug, ARI, appeared to have general advantages regarding side effects [13]. ARI is generally well-tolerated, it has a low propensity for EPS and causes lower incidences of excessive weight gain, glucose dysregulation, hypercholesterolemia, and hyperprolactinemia [14, 15]. The latter seems to be very important for women because high prolactin levels increase the risk of developing breast cancer [16, 17]. A study carried out on female patients with schizophrenia showed that long-term exposure to ARI (prolactin-sparing antipsychotic) was not associated with an increased risk of breast cancer [18]. Moreover, a clinically significant property is that ARI is not associated with impaired glucose tolerance, which is particularly important in pregnant women [19]. It is well known that the disturbance in glucose during pregnancy can increase the risk of gestational mellitus diabetes [20]. Several studies have shown that ARI, both oral form, and long-acting injection, was not associated with an increased risk of major congenital defects and neurological malformations [21,22,23]. Surprisingly, it was established that ARI (10–20 mg/day) has the potency to normalize significantly elevated levels of serum prolactin caused by antipsychotic-induced prolactinemia [24, 25]. Although the severity of ARI side effects, including EPS and metabolic syndromes, is less frequent than with other antipsychotics, some patients have experienced adverse drug effects.

The purpose of this review is to highlight the role of various genetic polymorphisms in the pharmacokinetics and pharmacodynamics of ARI. This review also offers a brief discussion of the relationship between genetic variants and metabolic side effects. Understanding the role of polymorphisms in the efficacy and safety of aripiprazole therapy may in the future result in the translation of pharmacogenomic knowledge into clinical practice.

Mechanism of action of ARI

ARI was approved in the USA in 2002, in Europe in 2004, and in Japan in 2006 for indication of schizophrenia [71]. It acts as a partial D2 dopamine and serotonin 5-HT1A receptors agonist as well as a serotonin 5-HT2A receptor antagonist. Furthermore, the affinity of ARI for other crucial nervous system receptors has been demonstrated (Table 2).

Table 2 Aripiprazole affinity for human receptors
Full size table

ARI is defined as a dopamine-system stabilizer (DSS) because of its higher affinity for the D2 receptor (Ki = 0.34 nM) than for 5-HT1A and 5-HT2A receptors (Ki = 1.7 nM and Ki = 3.4 nM, respectively) and its stabilizing effect on dopamine (DA) neurotransmission [72]. DSS partially activates DA receptors stabilizing the balance between stimulation and blockade of DA receptors [73]. DSS blocks D2 receptors in brain regions where DA activity needs to be reduced, at the same time, it does not reduce dopamine activity in brain regions where normal DA levels are needed [74]. Indeed, a positron emission tomography (PET) performed in healthy men who received single doses of ARI (3–9 mg) has shown that ARI decreases or increases DA synthesis in individuals with high or low baseline DA levels, respectively [75]. Thus, these findings suggest that the therapeutic effects of ARI may be related to a stabilizing effect on DA synthesis capacity and dopaminergic neurotransmission. Although the ARI occupancy rate on D2 receptors needs to be greater than 90% to have to achieve a therapeutic effect, it does not produce EPS [76]. Because ARI has lower intrinsic activity than a full agonist (i.e., endogenous dopamine), therefore signal transmission is lower than that of dopamine, but not completely blocked as with an antagonist (i.e., conventional antipsychotics) (Fig. 1A).

Fig. 1
figure 1

Mechanism of action of ARI. A ARI, as a partial agonist, reduces dopamine-mediated transmission but does not block it like an antagonist. B Hyperactive dopamine transmission in mesolimbic brain regions mediates positive symptoms. ARI works as a function antagonist in areas of too high dopamine levels. C Hypoactive dopamine transmission in mesocortical brain regions mediates negative symptoms. ARI works as a function agonist in areas of too-low dopamine levels. DA dopamine, ARI aripiprazole, ANT antagonist, D2R dopamine D2 receptor

Full size image

It is postulated that the positive symptoms of schizophrenia are related to hyperactive dopamine transmission in the mesolimbic brain regions in turn, hypoactive dopamine transmission in the mesocortical system underlies the negative symptoms [77]. Due to the unique dopamine-dependent action of ARI, it helps to control both positive and negative symptoms of schizophrenia (Fig. 1B, C).

However, the unique mechanism of action is more complex possibly due to the functional selectivity of ARI. Indeed, studies indicate that ARI is a functional selective D2 ligand that exerts an effect on intracellular signaling pathways [76, 78]. An in vitro study showed that ARI caused the activation of mitogen-activated protein kinases (MAPK) and arachidonic acid pathways [78]. In addition, an in vivo functional selectivity study revealed different effects on protein kinase A (PKA), protein kinase B (Akt) and glycogen synthase kinase 3 beta (GSK3β) depending on brain regions [79]. It is worth emphasizing that elevated PKA levels in the nucleus accumbens correlated with increased expression of the GABAA (β-1) receptor as well as GSK3β signaling probably modulating NMDA and GABAA expression [80,81,82]. A study in patients with schizophrenia suggested that ARI increased GABA transmission in the prefrontal regions and this may have clinical benefits in terms of improved social competence [83]. Another study on the effects of ARI exposure on NMDA and GABAA receptor binding levels revealed that ARI modulates the neurotransmission of both receptors in juvenile rats [84].

To summarize, different properties such as partial agonism and functional selectivity as well as actions at other receptor systems may be responsible for the action of ARI and the effective management of positive and negative symptoms in schizophrenia.

Metabolism of ARI

The bioavailability of the tablet formulation of ARI is 87%, maximum plasma concentrations (Cmax) occur 2.8–6.8 h after drug intake (depending on the dose), and its pharmacokinetics is linear [9]. ARI is metabolized by the hepatic cytochrome P450 (CYP450) enzyme system via three biotransformation pathways: dehydrogenation, hydroxylation, and N-dealkylation [85]. The two isoenzymes, CYP2D6 and CYP3A4, are mainly involved in the metabolism and elimination of ARI. However, the CYP3A4 shows a less significant influence on the metabolism of ARI [9]. The active metabolite dehydro-aripiprazole (D-ARI) arises as a result of a dehydrogenation pathway mediated by both isoenzymes. D-ARI accounts for approximately 40% of the drug concentration in plasma [10]. Although ARI and D-ARI exhibit similar pharmacological properties, their half-lives differ significantly (ARI—75 h versus D-ARI—94 h). Several studies have shown the impact of genetic polymorphisms on the pharmacokinetic and pharmacodynamic parameters of ARI. The Food Drug Administration (FDA) and the Dutch Pharmacogenetics Working group (DPWG) recommend adjusting the dose of ARI based on the CYP2D6 genotype. Applying a pharmacogenetic approach to ARI management can help determine a specific dosage for a patient, to ensure maximum efficacy with minimal side effects.

Pharmacogenetics

The cytochrome P450 monooxygenases metabolize approximately 70–80% of all used drugs, including antipsychotic drugs. Their expression depends on both genetic and non-genetic factors, such as age, sex, comorbidities, and other medications [86]. The CYP450 system-mediated drug conversion can lead to detoxification, creating new, reactive molecules accelerating the process of toxic compounds elimination, and hence, general response to the therapy may differ according to the individual metabolic capacity presented by patients [87]. Thus, the overall response to therapy may vary depending on the patient's individual metabolic rate.

A recent study, performed in a population of healthy volunteers receiving a single oral dose of ARI, confirmed that the pharmacokinetic parameters are influenced by the polymorphisms of genes encoding metabolizing enzymes (CYP2D6, CYP3A4, and CYP3A5) and in the drug transporter (ABCB1) [66]. It is postulated that the pharmacodynamics of ARI can be affected by polymorphisms in dopamine D2- and serotonin-5-HT2A receptors [88].

Gene polymorphisms and drug response

CYP2D6

Although CYP2D6 constitutes only 2% of the hepatic CYPs, it is an essential isoform involved in the metabolism of approximately 20–25% of drugs, including antidepressants, antipsychotics, β-blockers, analgesics, and tamoxifen [89]. The CYP2D6 gene is highly polymorphic and more than 130 allelic variants have been identified so far. These variants include single nucleotide polymorphisms (SNPs), small insertions/deletions (Ins/Del) of nucleotides, deletion of the entire CYP2D6 gene, gene duplication or multiplications as well as hybrid alleles [90, 91]. The activity of the enzyme encoded by each allele, as defined by the clinical pharmacogenomics implementation consortium (CPIC), can be either normal, reduced, or absent. The CYP2D6*1 allele is considered as a wild-type (so-called normal) allele that encodes enzyme with normal activity. An individual with two or one 2D6*1 alleles has a normal metabolic rate and is classified as a normal metabolizer (NM) or extensive metabolizer (EM). It is possible to predict metabolizer status based on the specific combination of alleles: ≥ 3 normal function gene copies—ultrarapid metabolizer (UM); 1 or 2 normal function alleles—normal metabolizer (NM); ≥ 2 decreased function alleles or 1 decreased function and 1 no function allele—intermediate metabolizer (IM); ≥ 2 no function alleles—poor metabolizer (PM) [92, 93]. It is well known that the frequency of CYP2D6 alleles varies among racial and ethnic groups [94]. A study by Gaedigk and colleagues predicted phenotypes in major populations from allele frequency data [95]. The frequencies of the CYP2D6 alleles and genetically predicted phenotypes are presented in Fig. 2.

Fig. 2
figure 2

The frequencies of CYP2D6 alleles and predicted phenotypes in the five populations. The sum of allelic frequencies is not 100% as they are average values in the given population. I increased functioning allele, N-F non-functioning allele, R reduced functioning allele, F functioning (normal) allele, UM ultrarapid metabolizer, PM poor metabolizer, IM intermediate metabolizer, NM normal metabolizer. (Diagrams have been prepared based on data provided by Gaedigk et al. 2017 [95])

Full size image

Interestingly, the frequency of UM phenotypes is much higher in South-East compared to North-West Europe (6% in Greece and Turkey to 1% in Sweden and Denmark, except Finland—3.4%). Inversely, the frequency of loss-of-function alleles (2D6*4 and 2D6*5) was lower in Mediterranean countries and highest in Northern Europe [96].

CYP2D6 metabolizer phenotype influences the half-life of ARI, patients with PM phenotype have almost double extended mean elimination half-life (146 versus 75 h) [10], as they cannot metabolize ARI. It has been observed that when the number of active CYP2D6 alleles decreased, AUC0-t and T1/2 were higher for ARI, and AUC0-t and Cmax were decreased for D-ARI [66]. As recommended by the FDA and the DPWG, the standard dose should be reduced by 50% or 67% (respectively), regardless of the administration route (oral and long-acting injectable). Moreover, a quarter of the usual dose should be used in poor metabolizers (PM) taking strong inhibitors of the CYP3A4 enzyme. In addition, no action is recommended for IM or UM; however, recent studies have suggested that IM patients may require a lower dose of ARI [97, 98]. Surprisingly, a recent study in Chinese subjects has shown that CYP2D6 rs1058164 and rs28371699 also affected the pharmacokinetics of ARI, T1/2, and AUC0-∞ but differed significantly between CYP2D6 NM and IM [99]. Due to the relatively high frequency of these SNPs in the Chinese population dose adjustment should probably be considered for IM.

CYP3A4

There are two main allelic variants of CYP3A4, *20 and *22, involved in the metabolism of ARI [66]. The CYP3A4*20 loss-of-function allele resulted in a higher AUC0-t of ARI, and a lower AUC0-t of D-ARI, thereby increasing the patient's plasma levels of ARI [66]. It seems that the CYP3A4*22 reduced functioning allele can also affect the metabolism of antipsychotics [100], but this allele did not affect the pharmacokinetics of either ARI or D-ARI [66].

CYP3A5

The role of CYP3A5 in ARI metabolism is much less significant in comparison to CYP2D6 [66]. However, a study has shown that the CYP3A5*3 allele may influence D-ARI/ARI ratio—lower values of this parameter were observed in individuals with genotype *3/*3 (no CYP3A5 enzyme production) compared to *1/*1 and *1/*3 genotypes [66].

CYP1A2 and UGT1A1

Although ARI is not a substrate for CYP1A2 and UGT enzymes, a recent study suggested that polymorphism in CYP1A2 and UGT1A1 genes may be involved in ARI and D-ARI pharmacokinetics [101]. However, this study was performed on a small number of healthy volunteers, and thus, more studies are needed including studies in patients with schizophrenia, in order to confirm the involvement of these polymorphisms in ARI metabolism.

ABCB1

ABCB1 gene encodes the membrane-associated protein (P-glycoprotein), a member of the superfamily of ATP-binding cassette (ABC) transporters, responsible for ATP-dependent active transport of drugs. ABCB1 protein is involved in processes, such as drug absorption, distribution, and elimination. It is postulated that the synonymous C1236T polymorphism influenced the expression level of the ABCB1 gene [102, 103]; however, the results are contradictory and further studies are needed to evaluate the association between the C1236T polymorphism and gene expression. Interestingly, the pharmacokinetic parameters of ARI and D-ARI were influenced by the synonymous C1236T polymorphism in the ABCB1 gene. The clearance of ARI, AUC0-t, and Cmax for D-ARI as well as the D-ARI/ARI ratio had higher values in C/C subjects compared to T/T subjects [66].

DRD2

There are many variants of the DRD2 gene, including − 141 Ins/Del, Ser311Cys, C957T, and Taq1A, that may affect antipsychotic response. The − 141 Ins/Del polymorphism is a deletion of one nucleotide (cytosine) at position − 141 of the 5' promoter region. Imaging studies in healthy volunteers showed that carriers of the − 141 Del allele have increased striatal D2 receptor density [104]. PGx testing indicated that carriers of the Del allele had reduced response to antipsychotic drugs [105]. The Ser311Cys DRD2 polymorphism results in a substitution of an amino acid at position 311 (serine to cysteine). Patients with schizophrenia and the Ser311 allele are more resistant to treatment with risperidone than patients carrying the Cys311 allele [106]. Although the C957T polymorphism is a synonymous variant and does not change the amino acid sequence of the resulting protein, it can alter mRNA stability [107]. Reduced translation and mRNA stability were associated with the T allele [108], moreover, the T allele showed a protective effect against schizophrenia [109]. The Taq1A polymorphism is a missense variant (cytosine is replaced with thymine) resulting in an amino acid substitution at position 713 (Glu713Lys, glutamic acid to lysine). The Taq1A*1A polymorphism seems to be especially important regarding the effectiveness of antipsychotic treatments. The risk allele, A1 (thymine) allele that reduce the expression of the DRD2 gene decreases D2 receptor density in the striatum [104]. A further study performed on healthy volunteers, found that the A1/A1 subjects showed increased metabolic activity in the frontal lobe compared to the A2/A2 (wild-type) subjects. Thus, patients with the A1/A1 genotype may respond better to ARI treatment [110]. Another study in patients with schizophrenia evaluated an association between the response to ARI treatment and four polymorphisms in the DRD2 gene mentioned above [108]. The carriers of the A1 allele with positive symptoms respond much better to ARI relative to individuals with A2/A2 genotype. Furthermore, regarding the C957T polymorphism, patients with T/T genotype had better ARI response for excitement symptoms compared to C/C genotype. This study also revealed no association with the ARI response and two polymorphisms (− 141 Ins/Del and Ser311Cys) [108].

5-HTR2A and 5-HTR1A

Among various polymorphisms of the 5-HTR2A gene, the T102C variant is the most studied. The C allele decreases receptor expression and receptor binding potentials [111, 112]. A study by Lane et al. showed that patients with the C/C genotype respond better to risperidone treatment (especially for negative symptoms of schizophrenia) [113]. Likewise, a significantly better response to olanzapine treatment was observed in patients with positive symptoms of schizophrenia and the C/C genotype [114]. In contrast, another study identified that the C allele is associated with less effective ARI treatment on negative symptoms of schizophrenia [115]. Another polymorphism, C1354T, is a missense variant of the 5-HTR2A gene resulting in an amino acid substitution at position 452 (Hys452Tyr, histidine to tyrosine). This polymorphism may alter the tertiary structure of the protein and thus may disrupt the function of the receptor. The homozygous (His/His) respond better to olanzapine treatment, this association was noticed in terms of positive symptoms [114].

The − 1019C/G polymorphism in the promoter region of the 5HTR1A gene increases its expression level both in animal models and humans. Individuals with G/G genotype have increased density of 5-HT1A receptor density in presynaptic raphe neurons [116]. In patients with the C/C genotype olanzapine or perospirone more effectively improved the cognitive deficit of schizophrenia (attention) than in patients having G/C and G/G genotypes [116]. Previous studies showed that the response to treatment with various antipsychotics was limited when the G allele was present, possibly due to the increased receptor density, which may result in the lower efficacy of antipsychotic drugs [117, 118].

Gene polymorphisms and metabolic side effects

Generally, ARI is well tolerated and not associated with significant EPS or raised prolactin concentrations. However, in some patients, it can cause side effects, such as increased blood glucose or cholesterol levels. Typically, ARI leads to low prolactin elevation, but in less than 5% of patients can sometimes cause hyperprolactinemia [119]. A recent study performed on healthy volunteers revealed that polymorphisms in specific genes can affect the levels of prolactin, C-peptide, insulin, and cholesterol [120]. Table 3 presents the relationship between metabolic parameters and gene polymorphisms identified after 5 days of ARI administration.

Table 3 Metabolic effects of aripiprazole in healthy volunteers
Full size table

Conclusions

In this review, we discuss the possible association between gene polymorphisms and ARI response. Although the FDA and DPWG recommended dosage adjustments for patients who are CYP2D6 poor metabolizers, it seems that other genetic variations are also related to pharmacokinetic, pharmacodynamics, and side effects of the drug. The specific genetic profile of a patient can determine the effectiveness and tolerability of ARI. We believe that more targeted pharmacogenetics testing prior to prescribing ARI will provide the opportunity for personalized medicine to treat schizophrenia, thereby improving clinical outcomes and patient satisfaction. However, extensive pharmacogenetic studies are needed to assess the relevance of specific gene polymorphisms in response to the drug, which will be included in future diagnostic panels.