Pharmacogenetics of Serious Antipsychotic Side Effects


Gene

Polymorphism

Main findings

CYP1A2

(*1F, −163C>А, rs762551)

CC genotype and/or C allele are possible risk factors for tardive dyskinesia

SLC18A2

rs2015586

C allele can be a risk variant for tardive dyskinesia

PIP5K2A

rs10828317

Association of CC genotype with tardive dyskinesia

CNR1

rs806374

Possible risk for tardive dyskinesia in C-allele carriers

DPP6

rs6977820

A allele was associated with tardive dyskinesia

HSPG2

rs2445142

G allele was associated with tardive dyskinesia







2.2.2 Antipsychotic-Induced Weight Gain (AIWG)


AIWG is a highly heritable (h 2 = 0.6–0.8) and polygenic side effect of atypical antipsychotics. AIWG risk differs amongst SGAPs with “high-risk” AIWG drugs including clozapine and olanzapine, “moderate risk” including risperidone and quetiapine, and haloperidol and aripiprazole being “low risk.” Substantial weight gain leads to diabetes type II, metabolic syndrome, and cardiovascular diseases. AIWG results in social stigmatisation of weight gain, leading to patient incompliance, which further increases the need to determine the aetiology to alleviate AIWG in AP-treated patients [50]. While there have been no significant AIWG findings with FGAP treatment due to lack of adequate samples thus far [18], several interesting findings were reported with SGAPs.


2.2.2.1 Neurotransmitter Genes


Some neurotransmitters (e.g. serotonin and dopamine) are highly involved in appetite regulation thus they have been predominantly chosen as good candidates for AIWG studies [76].


Serotonin System Genes

The serotonin system has been extensively analyzed in relation to AIWG, as serotonin receptors are one of the main targets of SGAPs [50]. Serotonin transporters also affect central pathways that influence satiety and hunger [50].

There are several consistent results that described an association between the functional HTR2C gene −759C/T (rs3813929) polymorphism and AIWG. Two meta-analyses have confirmed the role of −759C/T variant in AIWG [15, 54, 76]. The C allele has been suggested to be a risk variant as it has been found to be more frequent in AIWG cases than in control (non-AIWG) populations. However, the role of −759C/T in metabolic syndrome is less clear, as some studies have found associations with metabolic syndrome [28, 59], whereas other studies have failed to replicate the same findings [35, 70]. While a recent meta-analysis have not confirmed an association of −759C/T and metabolic syndrome, a strong trend has been shown for the C allele to be associated with high olanzapine-induced weight gain [54].

Several genetic association studies have been conducted with the intronic HTR2C variant rs1414334 and AIWG, albeit with mixed results. A positive link for rs1414334 has been established in a meta-analysis [54]; however, a recent study has not able to replicate the association between rs1414334 and AIWG [41].

Lack of replication may be due to the differences in the ethnicities of populations analyzed, medications prescribed, previous exposure to AP, and importantly length of treatment observation [94]. Thus, more studies with stringent sample selection are required to avoid confounding effects and increase likelihood of replicated findings.


Dopamine System Genes

The dopamine system plays an important role in mediating AP response; additionally, disruptions in DRD2 have been linked with obesity [50].

The A allele of DRD2 rs2440390 variant has shown association with more severe AIWG [30] in patients without SCZ; however, this finding awaits replication. A recent study on a Polish population showed no effect of the DRD2 -141C Ins/Del (rs1799732), Taq1A, and DRD2 exon 8 variants on AIWG during ziprasidone, olanzapine, and perazine treatment [91]. A previous study on first-episode patients treated with risperidone and olanzapine has shown the del allele to be associated with AIWG [48]. A later study has shown an association between three DRD2 SNPs: rs6277 (or C957T), rs1079598 and rs1800497 (TaqIA), and AIWG [62]. Thus, some DRD2 gene variants appear to be suggestive risk factors for AIWG. Gene-gene interaction or pathway analyses may help to validate these risk variants further and detect contribution of each variant.


2.2.2.2 Leptin and Leptin Receptor


Leptin is an adipocyte hormone that works as a long-term regulator of energy balance and is part of a negative feedback loop regulating body weight. Leptin released in adipose tissue signals the brain to decrease food intake and increase energy metabolism. Furthermore, mutations in the LEP gene have been reported in obesity and insulin resistance [7]. Taken together, leptin and its receptor are natural candidates for AIWG genetic studies.

The −2548A/G LEP gene variant has been reported to be associated with AIWG; however, studies produced inconsistent results. In a longitudinal (>1 year) study of an Asian population, the A allele and the AA genotype have been reported as risk factors for AIWG [34, 95]; however, other studies have suggested the G/A genotype to be the risk factor [34, 101]. In contrast, the G allele has been found to be associated with AIWG in a Caucasian sample [9, 19]. Further, replication studies have resulted in mostly negative results [7, 64]. A recent investigation on autistic children and adolescents treated with risperidone have shown increased AIWG risk for G allele carriers [63]. One recent meta-analysis study has yielded negative findings; however, studies were heterogeneous and phenotype definitions remained unclear [74]. Another recent retrospective association study in Finnish patients also yielded negative results [41]. Most plausible explanations for inconsistent findings are likely due to various ethnicities, different study durations, and heterogenous AP treatment. Significant associations have been mostly reported in Asian patients observed for longer time periods (>1 year), mostly on monotherapy. Although clinically significant weight gain may appear in the first weeks of treatment, study findings suggest at least 9 weeks of observation may be necessary to observe statistically significant genetic associations [81].

One study explored the gene-gene interaction between LEP (rs7799039) and leptin receptor, LEPR (Q223R) however, no significant interaction has been detected [7]. The Q223R LEPR variant was previously investigated in AIWG pharmacogenetic studies, albeit with mixed reuslts. The R allele and RR genotype have been associated with higher risk for obesity following AP treatment [19, 27]; however, other studies have failed to confirm the role of Q223R with AIWG [7, 67].


2.2.2.3 The Melanocortin 4 Receptor (MC4R)


The melanocortin 4 receptor (MC4R) gene locus has been shown to be one of the best and consistently replicated findings associated with AIWG [37, 73]. MC4R is a membrane-bound receptor that plays an essential role in the regulation of energy homeostasis [29]. Defects of MC4R are one of the causes for monogenic forms of obesity [29]. GWAS study of SGAP-treated sample has shown strong association of MC4R gene locus marker rs489693 in AIWG, which has been replicated in three independent samples [55]. In addition, the same variant has been associated with AIWG in another German study [13]. Notably, in a sample with children and adolescents treated with risperidone for autism symptoms, this marker has been again associated with AIWG, however, showing an opposite allele effect [63].

Other variants of the MC4R gene have been investigated in AIWG, including MC4R variant rs8087522. The A-allele carriers gained more weight than noncarriers. Results became marginal after correction for multiple testing; however, in vitro studies have suggested that the A allele might create a binding site for transcription factors [12].


2.2.2.4 The Cannabinoid Receptor 1 (CNR1) and the Neuropeptide Y (NPY) Genes


The cannabinoid receptor 1 (CNR1) is associated with appetite and satiety. One study has showed an association between CNR1 rs806378 variant and AIWG, in which T-allele carriers have been found to gain more weight than CC carriers [85]. Notably, another study in autism children and adolescents has shown the same trend for the T allele [63].

Based on the association with CNR1, Tiwari et al. conducted a study investigating the neuropeptide Y (NPY) gene since it has shown to interact with CNR1 in animal studies [82]. NPY is an orexigenic peptide that stimulates food intake [82].This study has found an interaction with NPY rs16147 and CNR1 rs806378 to be associated with AIWG [82].


2.2.2.5 Mitochondrial Genes


The mitochondria plays a key role in energy homeostasis, and it has been shown to be related with neuronal activity [36] and obesity [42]. Recently the rs6971 variant of the translocator protein 18kDA (TSPO) has been found to be associated with AIWG in two independent samples [68]. In another study, associations have been found between rs6435326 variant of NADH dehydrogenase (ubiquinone) Fe-S protein 1 (NDUFS1) and AIWG in two independent samples [25]. Furthermore, this study also discovered a significant gene-gene interaction between the TT genotype of rs6435326 and the AG genotype of rs3762883 of the cytochrome C oxidase assembly factor (COX18) [25]. These are the first studies conducted with mitochondrial genes to suggest that they may have a significant role in AIWG.


2.2.2.6 Other Genes


Few studies have investigated associations between other genes and AIWG. One such gene is the BDNF gene in relation to AIWG. The Val/Val (rs6265) genotype of BDNF has been found to be associated with increased AIWG in Han Chinese, risperidone-treated sample [45]. However, a study on a larger Asian sample has failed to replicate the original findings [87] but has reported an association of AIWG with another BDNF variant rs11030101. Subsequent analysis has shown association of BDNF haplotype rs6265-rs1519480 with strong, nominal association between Val/Val genotype and AIWG [100]. In summary, BDNF gene variants might be associated with AIWG; however, the genetic architecture of the BDNF gene requires further investigation.

The methylenetetrahydrofolate reductase (MTHFR) is involved in the homocysteine metabolism and thus potentially associated with AIWG [78]. The CC genotype of 677C/T (rs1801133) variant has been linked to significant AP-induced increase in BMI in both Spanish and Chinese samples [78]. An independent study has confirmed a possible risk of the C allele in both chronic and first-episode patients [38].

In 2011, a GWAS study of 738 patients, investigating 492 000 SNPs, has found an association between meis homeobox 2 (MEIS2), cyclic adenosine monophosphate-dependent, regulatory, type II, beta (PRKAR2B), forming homology 2 domain containing 3 (FHOD3), ring finger protein 144A (RNF144A, ASTN2), sex-determining region Y-box 5 (SOX5), and activating transcription factor 7 interacting protein 2 (ATF7IP2) and AIWG [1]. However, replication studies for the mentioned genes are still need to be conducted for validation.

Our group recently submitted a GWAS study conducted on a well-characterised, genetically homogenous subsample of European ancestry carefully selected from CATIE study. None of the variants from this study have reached genome-wide significance; however, strong, nominal associations were found for rs9346455 located upstream of opioid growth factor receptor-like 1 (OGFRL1) and rs1059778 located in iron-sulfur cluster assembly (IBA57). In addition, we investigated our top hit findings in a smaller replication cohort, in which the top SNP rs9346455 has shown significant association with AIWG. The combined meta-analysis p-value for rs9346455 was close to genome-wide significance (p < 10−7) [8a] (Table 2.2).


Table 2.2
Summary of most promising polymorphisms associated with antipsychotic-induced weight gain
































Gene
Polymorphism
Main findings

HTR2C

−759C/T (rs3813929)
The C allele was suggested to be a risk variant

LEP

−2548A/G
The A allele is a risk factor for antipsychotic-induced weight gain in Asians, whereas G allele is linked to antipsychotic-induced weight gain in Caucasians. Results were significant only in longitudinal studies of extended duration

MC4R

rs489693
AA genotype associated with increased weight gain

CNR1

rs806378
T allele associated with antipsychotic-induced weight gain

NDUFS1

rs1801318
TT genotype associated with increased weight gain


2.2.3 Clozapine-Induced Agranulocytosis (CIA)


Clozapine-induced agranulocytosis (CIA) is a severe adverse effect in treatment-resistant SCZ patient population requiring clozapine. CIA is present in up to 2 % of clozapine-treated SCZ patients [8]. It is usually linked with the immune-mediated response against neutrophils and toxic effect against bone marrow stromal cells [8].

Most studies have linked CIA to the human leucocyte antigen (HLA) system genes, which are part of the major histocompatibility complex (MHC). HLA genes have been linked with several immune and nonimmune diseases, and adverse drug reactions to xenobiotics [90]. The strongest CIA associations exist for HLA-DQB1 and HLA-B38 variants; however, the effects of particular genes and genetic variants in CIA are poorly replicated [8, 11]. The myeloperoxidase (MPO) and nicotinamide adenine dinucleotide phosphate oxidase (NOX) have also been suggested to be associated with CIA, albeit results are not as consistent as for HLA [8]. Meanwhile, another recent paper published on CIA in a Finnish sample using whole-exome sequencing identified protein tyrosine phosphatase, receptor type, f polypeptide, interacting protein, alpha 4 (PPFIA4), ubiquitin specific peptidase 43 (USP43), actinin, alpha 1 (ACTN1), podocan-like 1 (PODNL1), and spermatogenesis associated, serine-rich 1 (SPATS1) as the top five hits [84]. Although these genes have not reached whole-exome significance, they show a trend towards immunologically associated genes in CIA.

One GWAS has suggested a role of MyoD family inhibitor domain containing (MDFIC) and proteoglycan 4 (PRG4) loci in risk for CIA [16]. Further replications are required to validate these results.

Of note, in 2007, a first commercial test kit for CIA, the PGxPredict:CLOZAPINE test (Clinical Data, Inc, New Haven, CT) was launched. The test reached a high specificity of 98.4 % but sensitivity scores remained low at 21.5 %, thus failing to detect patients at high risk for CIA [8]. More pharmacogenetic research is needed to develop newer and more precise genetic screening tests for patients. Polygenic risk scores, derived from gene-gene interaction studies, may improve the initial algorithm.



2.3 Summary


Genetic associations of common and serious AP side effects have been extensively studied in the recent years. As for TD, gene variants in CYP1A2, SLC18A2 PIP5K2A, CNR1, DPP6, and HSPG2 have yielded promising results. However, this area of research would substantially benefit from further investigation in larger, well-characterised samples allowing for additional GWAS.

For AIWG, HTR2C, LEP, MC4R, NDUFS1, and CNR1 have recently yielded the strongest findings. The most clinically relevant finding was obtained in MC4R homozygote carriers for rs489693 who on average gained twice as much weight than noncarriers. In addition to other risk variants, such as HTR2C, polygenic risk tests for AIWG might become available for use in clinical practice.

As for CIA, a few studies have been conducted in the past years linking CIA further to immunological system genes. These findings suggest that gene variants associated with CIA might show relatively large effect sizes such as the HLA-B*15:02 allele which is associated with an increased risk of carbamazepine-induced Stevens-Johnson syndrome [46]. While an early genetic test for CIA failed to detect high-risk group due to low sensitivity, it is very likely that further research will aid in developing an improved genetic test to predict patients at risk for CIA. This would allow for the discontinuation of regular blood draws required to screen for the development of CIA.

In reviewing the literature, certain limitations in these genetic studies need to be addressed. The samples used were relatively small compared to disease genetic studies since high-standard pharmacogenetic studies require the collection of prospectively assessed samples. However, in many cases, genes involved in response and side effects to medication often show larger effect sizes than disease-risk genes. Many genetic association analyses faced issues with heterogeneity, either caused by ancestry and/or by current/previous medication exposure. Given these limitations, it is important to conduct more studies with standardised methodologies to obtain more consistent and comparable results, to develop new or refine existing genetic tests. First tests have been introduced mainly to identify nonnormal metabolizers for CYP enzymes (particularly for CYP2D6) to optimise AP drug treatment. The United States Food and Drug Administration (FDA) labelled over 100 medications for genetic testing, including 32 in psychiatry/neurology [17]. Expert groups such as the Clinical Pharmacogenetics Implementation Consortium (CPIC) are providing guidelines to help clinicians use genetic information to select type/dosage of various drugs to decrease trial-and-error switches in medications [10].

Future studies should analyse gene-gene interactions in order to explain higher degree of variants. More studies with stringent sample selection are required to avoid confounding effects (e.g. assessing medication exposure, considering ethnic differences) in order to increase the likelihood to replicate initial findings. Nonetheless, research in pharmacogenetics of APs have made some substantial progresses in the past years raising hope for an accelerated development of genetic testing in clinical practice.

Tags:
Mar 10, 2017 | Posted by in PSYCHOLOGY | Comments Off on Pharmacogenetics of Serious Antipsychotic Side Effects

Full access? Get Clinical Tree
Get Clinical Tree app for offline access