Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Aug 1.
Published in final edited form as: Psychiatry Res. 2017 Apr 14;254:8–11. doi: 10.1016/j.psychres.2017.04.025

Genome-Wide Association Study of Treatment Response to Venlafaxine XR in Generalized Anxiety Disorder

Jeesun Jung a, Elisabeth A Tawa b, Christine Muench b, Allison D Rosen b, Karl Rickels c, Falk W Lohoff b,c,*
PMCID: PMC5798606  NIHMSID: NIHMS904232  PMID: 28437668

Abstract

We conducted the first genome-wide association study (GWAS) in Generalized Anxiety Disorder (GAD) to identify potential predictors of venlafaxine XR treatment outcome. Ninety-eight European Americans (EA) patients participated in a venlafaxine XR clinical trial for GAD, with Hamilton Anxiety Scale (HAM-A) response/remission at 24 weeks as the primary outcome measure. All participants were genotyped with the Illumina PsychChip, and 266,820 common single nucleotide polymorphisms (SNPs) were analyzed. Although no SNPs reached genome-wide significance, 8 SNPs were marginally associated with treatment response/remission and HAM-A reduction at week 12 and 24 (p < 0.00001). Several identified genes may indicate markers crossing neuropsychiatric diagnostic categories.

Keywords: pharmacogenetics, generalized anxiety disorder, venlafaxine XR

1. Introduction

Generalized anxiety disorder (GAD), a chronic psychiatric disorder characterized by a state of excessive worry, afflicts roughly 4.1% of American adults (Grant et al., 2005). GAD is marked by significant morbidity and mortality. Recent studies found that compared to those without anxiety, individuals with GAD had a significantly higher likelihood of suicidal ideation (odds ratios (ORs) = 1.78 – 4.81) and attempted suicides (ORs = 2.70 – 5.59) (Cougle et al., 2009; Kanwar et al., 2013). GAD is most often treated with selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs). However, antidepressant treatment response is highly variable in individuals with GAD; while antidepressant treatments are effective for some GAD patients, up to 34% of patients fail to respond favorably (Baldwin and Nair, 2005; Gelenberg et al., 2000; Rickels et al., 1993). Pharmacogenetic studies investigating the effect of antidepressant drugs on mood disorders have primarily focused on major depressive disorder (MDD), while research on pharmacogenetic treatments for anxiety disorders is limited (Kato and Serretti, 2010; Multani et al., 2012; Tiwari et al., 2009). Because the research on pharmacogenetic treatment response in anxiety disorders is sparse, the genetic component underlying pharmacotherapy response remains unknown.

To identify genetic variants contributing to the etiology of primary anxiety disorders, a meta-analysis of genome-wide association studies (GWASs) with large, independent samples found multiple novel genetic variants to be significantly associated with anxiety disorder phenotypes (Otowa et al., 2016). GWASs could be useful for examining the pharmacogenetics of antidepressant treatment response in GAD to detect small effect sizes of associated genetic variants.

In this study, we tested the hypothesis that single nucleotide polymorphisms (SNPs) are associated with treatment response outcome in GAD. Treatment responses were quantified based on Hamilton Anxiety Scale (HAM-A) and Clinical Global Impressions-Severity (CGI) scale reductions following 24-week treatment with venlafaxine XR. Identifying genetic variants which potentially contribute to GAD could better inform pharmacological treatments based on individual genetic profiles.

2. Methods

2.1 Subjects

Participants with a diagnosis of GAD were enrolled in an 18-month relapse prevention study which included three treatment phases (Rickels et al., 2010): The first phase 24-week open-label venlafaxine XR flexible-dose treatment phase (75–225 mg day−1) was used to conduct primary pharmacogenetic analyses (supplementary materials). Overall, 156 patients (European-Americans [EA] n =112; African-Americans n =41; others n= 3) were evaluated for treatment response to venlafaxine XR. However, due to ethnic differences in allele frequencies and consequent population stratification, only the EA population (n= 112) was used in the pharmacogenetic analysis. The HAM-A score was used as a primary outcome measure, and the CGI of Improvement (CGI-I) score at 24 weeks was used as a secondary outcome measure.

2.2 Genotype and Quality Controls

All participants were genotyped using the Illumina PsychChip capturing 571K SNPs. Among those SNPs, only common variants SNPs whose minor allele frequency is greater than 5% were selected due to our sample size. Common variants are defined by over 1% of a minor allele frequency with decent sample size such as 1,000 (Schork et al., 2009), but we selected common SNPs having 5% or greater because at least 5 out of 100 samples were needed for a suitable statistical analysis due to our sample size of 96.

Quality control (QC) procedures were applied to individual samples and SNPs. Using common SNPs (n = 146,257) available in both our samples, and in samples in the 11 populations of HapMap phase 3 data (n = 1,184), we performed principal components analysis to determine non-European ancestry by Eigenstrat (Price et al., 2006). Non-European ancestry samples were removed. Samples were excluded if the missing rate exceeded 2%. Related or duplicated samples were identified through identity-by-state sharing analysis and removed. SNPs were excluded if the missing rate exceeded 5%, their minor allele frequency was < 5%, or they showed departure from the Hardy Weinberg equilibrium test (p-value < 1.0E-5). After the series of QC, 98 European samples with 266,820 common SNPs remained for further statistical analysis (supplementary materials).

2.3 Statistical Analysis

The HAM-A reduction scores from baseline to 24 weeks and from baseline to 12 weeks respectively were tested for association with each SNP using a linear regression model based on an additive model. Response and remission to venlafaxine XR at 12 and 24 weeks, including CGI-I response and remission, were tested using an allelic based chi-square test that compared allele frequency between two groups. Instead of using Fisher Exact test that is more accurate than the chi-square test when the sample size is small and a genetic variant is rare, we utilized chi-square test because our sample size of 98 and common variants provide a reasonable approximation of test statistics[reference].

3. Results

None of the SNPs were identified with genome-wide significance (1.9×10−7 = 0.05/266,820 by Bonferroni correction based on the number of SNPs, 266,820 SNPs) for either the main outcome measures or secondary outcome measures at the main study end point at week 24. All p-values except HAM-A reduction score were driven by an allelic comparison test without any assumption of genetic model such as a co-dominant model. Additional results are available in the supplementary materials. Table 1 illustrates the 8 SNPs associated with all treatment response/remission measures and HAM-A score at week 24 and 12 by a threshold of p-value < 0.01: rs10483832 (MED6), rs13216187 (SGK1), rs17154827, rs1993919 (STAB2), rs2136474 (SPATA3), rs7060140 (OPHN1), rs7342064 and rs7897283 (PARD3).

Table 1.

Results (p-values) of significant SNPs across all outcomes at both week 24 and week 12 at significance level of 1.9×10−5.

CHRa SNP Physical locationb A1 A2 Week RESPONSE CGI RESPONSE REMISSION CGI REMISSION HAM-A GENE symbol SNP function
14 rs10483832 71060837 G A 24 0.0008 1.22E-05 0.0020 0.0086 0.0030 MED6 Intronic
12 0.0010 0.001 0.0083 0.0010 0.003
6 rs13216187 134627600 G A 24 0.0001 0.0033 0.0001 2.99E-05 4.42E-05 SGK1 Intronic
12 0.0016 0.0033 0.0053 0.0012 0.0018
10 rs17154827 44490532 C A 24 1.36E-05 2.80E-05 0.0017 0.0002 4.36E-05 LINC00841c Intergenic
12 0.0007 2.80E-05 0.0024 0.0029 4.07E-05
12 rs1993919 104027817 G A 24 0.0024 0.0038 0.0063 0.0085 0.0071 STAB2 Intronic
12 0.0082 0.0005 0.0008 0.0017 0.0022
2 rs2136474 231864289 C A 24 0.0028 0.0015 0.0005 0.0007 0.0047 SPATA3 Intronic
12 6.36E-05 0.0004 0.0095 0.0012 0.0065
X rs7060140 67407442 C A 24 0.0015 0.0058 0.0008 0.0043 0.0004 OPHN1 Intronic
12 0.0004 0.0004 0.0086 0.0016 6.36E-05
10 rs7342064 35099799 A G 24 6.01E-05 0.0002 0.0023 0.0003 0.0009 PARD3 Intronic
12 0.0005 0.0047 3.27E-05 0.0019 0.0003
10 rs7897283 35075817 C A 24 7.40E-05 0.0002 0.0014 0.0003 0.0010 PARD3 Intronic
12 0.0006 0.0053 2.01E-05 0.0023 0.0003
Open in a new tab
a

CHR stands for Chromosome

b

the chromosome position was obtained from hg19 (GRCh37).

c

rs17154827 is intergenic SNP between LINC00841(distance = 25177) and C10orf142(distance = 297666)

4. Discussion

In this study, we conducted the first GWAS analysis of antidepressant treatment response in GAD. We found that none of the SNPs tested reached a genome-wide significance threshold, either in categorical outcomes or HAM-A response/remission. We did not explore demographic and clinical characteristics, such as sex, age, and time spent in treatment, as covariates in the association analysis of our sample due to the overall negative result. The following 8 SNPs were marginally associated with treatment response/remission and HAM-A scores at both week 12 and 24 (p < 0.00001): rs10483832-MED6, rs13216187-SGK1, rs17154827, rs1993919-STAB2, rs2136474-SPATA3, rs7060140-OPHN1, rs7342064-PARD3, and rs7897283-PARD3 (Table 1). The finding that these 8 SNPs were consistently associated with treatment outcome across HAM-A measures and time points indicates encouraging trends to pursue for further study.

SGK1 is the most clinically notable of the identified genes. A growing body of literature indicates that SGK1 may be involved in the pathophysiology of mood, anxiety, and trauma-related disorders. Chronic stress exposure in mice has been found to increase SGK1 in corpus callosum oligodendrocytes via hypothalamus-pituitary-adrenal axis activation, inducing morphological changes which may contribute to the pathogenesis of major depressive disorder (Miyata et al., 2015). Furthermore, downregulation of SGK1 in the hippocampus resulted in a reversal of corticosterone-induced depressive symptoms in a rodent model of depression (Li et al., 2015).

Recent studies have also shown that the effect of glucocorticoid signaling on glucocorticoid receptor function may be mediated by SGK1 upregulation, which has been demonstrated in both rodents and humans (Anacker et al., 2011; Sato et al., 2008; Yuen et al., 2011). Traumatic stress has been shown to induce learned helplessness and anhedonic-like behaviors in rats through decreased expression of SGK1 and synaptic significant decrease in spine density in medial prefrontal cortex (PFC) neurons (Licznerski et al., 2015). The same study found the SGK1 gene to be down-regulated by more than 80% in postmortem PFC samples of PTSD individuals compared to healthy controls (Licznerski et al., 2015). Thus these studies support an association between glucocorticoid-induced increases in SGK1 and the development of anxiety and mood disorders. Although a larger sample is needed to replicate the present study’s findings, SGK1 genetic variants may contribute to treatment response and GAD susceptibility.

STAB2 and OPHN1, two additional genes associated with antidepressant treatment outcome in GAD in the present study, have been identified as potential genetic determinants of schizophrenia. An Identify candidate Causal SNPs and Pathways (ICSNPathway) analysis on a schizophrenia GWAS data set implicated STAB2 as a candidate gene in schizophrenia susceptibility. Though OPHN1 has largely been associated with mental retardation (Nakano-Kobayashi et al., 2014), intellectual disability (Powell et al., 2014), and autism spectrum disorders (Piton et al., 2011; Won et al., 2013), reductions in gamma oscillatory activity observed in OPHN1 knockout mice have also been associated with Alzheimer’s disease, aging, and schizophrenia phenotypes, suggesting that reductions in OPHN1 expression increase the likelihood of cognitive and psychiatric pathology (Powell et al., 2014). The common correlations between STAB2 and OPHN1 genetic variants and neuropsychiatric illnesses suggest an underlying genetic basis for the symptomatology shared by these disorders.

Because SGK1, STAB2, and OPHN1 have been associated with a) treatment response in GAD in our study, and b) MDD, PTSD, and schizophrenia across studies, these results suggest that some genetic markers might cross diagnostic categories for multiple psychiatric disorders.

From the standpoints of both genetics and clinical experimental design, the current study may have been limited by several factors, including the retroactive collection of DNA, the absence of a placebo arm in the first phase of the trial, and the use of pill counts as the only measure of medication adherence. It should also be noted that the Illumina PsychChip array is designed on the basis of literature findings, and it contains at least 50,000 markers which have been previously associated with common psychiatric disorders. Hence, it is possible that our method of genotyping increased the probability of identifying genes of prior association with psychiatric disorders.

Limiting factors in the present study were small sample sizes and lack of statistical power for a genome-wide association study. At the same time our phenotypic characteristics were a response/remission to the drug which often drives relatively larger effects by genetic variants than complex disease status does. Considering our study as an exploratory study, the sample size of 98 was suitable enough to provide nominal association signals of SNPs on treatment outcomes of venlafaxine XR due to relatively large effects of SNPs selected from psychiatric disease genetic studies. In addition, instead of using a mixed effective model incorporating different time points for HAM-A score, we took into account HAM-A reduction scores from a baseline of HAM-A to investigate how genetic variants contributed to an improvement of HAM-A scores.

The Illumina Psych Chip that we used for genotyping was customized SNPs selected from results of psychiatric disease GWAS and many Exome studies, therefore SNP markers were not well-covered by common variants. Accurate imputation requires well-covered common variants with strong linkage disequilibrium (LD) across whole genomes (Halperin and Stephan, 2009)(Marchini and Howie, 2010). Due to our customized common SNPs screened by 5% of minor allele frequency with our samples size, the accuracy of imputation was not guaranteed to improve statistical power to discover associations.

In summary, we conducted the first GWAS of antidepressant treatment response in GAD. Although we found no significant genome-wide association results, it is promising that 8 SNPs were marginally associated with treatment response/remission and HAM-A at both months 3 and 6 (p < 0.00001). Because three genes (SGK1, STAB2, and OPHN1) included in our SNPs of interest have been previously associated with mood and anxiety disorders, our results suggest that certain genetic markers may underlie the shared phenotypic characteristics of comorbid psychiatric pathologies. Avenues for future research include the replication of results with larger samples sizes to increase statistical power and further elucidate the treatment effects of antidepressant venlafaxine XR on GAD.

Supplementary Material

Supplemental

Acknowledgments

This research was supported by the US Public Health Research grant MH065963 (KR) and K08MH080372 (FWL) and the NIAAA Intramural Research Program. We thank the Wyeth Laboratories for providing all study medication. The study was registered under clinical trials.gov. Identifier: NCT00183274.

Footnotes

Conflict of Interest

The authors declare no conflict of interest.

References

  1. Anacker C, Zunszain PA, Cattaneo A, Carvalho LA, Garabedian MJ, Thuret S, Price J, Pariante CM. Antidepressants increase human hippocampal neurogenesis by activating the glucocorticoid receptor. Molecular psychiatry. 2011;16:738–750. doi: 10.1038/mp.2011.26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baldwin DS, Nair RV. Escitalopram in the treatment of generalized anxiety disorder. Expert review of neurotherapeutics. 2005;5:443–449. doi: 10.1586/14737175.5.4.443. [DOI] [PubMed] [Google Scholar]
  3. Cougle JR, Keough ME, Riccardi CJ, Sachs-Ericsson N. Anxiety disorders and suicidality in the National Comorbidity Survey-Replication. Journal of psychiatric research. 2009;43:825–829. doi: 10.1016/j.jpsychires.2008.12.004. [DOI] [PubMed] [Google Scholar]
  4. Gelenberg AJ, Lydiard RB, Rudolph RL, Aguiar L, Haskins JT, Salinas E. Efficacy of venlafaxine extended-release capsules in nondepressed outpatients with generalized anxiety disorder: A 6-month randomized controlled trial. Jama. 2000;283:3082–3088. doi: 10.1001/jama.283.23.3082. [DOI] [PubMed] [Google Scholar]
  5. Grant BF, Hasin DS, Stinson FS, Dawson DA, June Ruan W, Goldstein RB, Smith SM, Saha TD, Huang B. Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychological medicine. 2005;35:1747–1759. doi: 10.1017/S0033291705006069. [DOI] [PubMed] [Google Scholar]
  6. Halperin E, Stephan DA. SNP imputation in association studies. Nature biotechnology. 2009;27:349–351. doi: 10.1038/nbt0409-349. [DOI] [PubMed] [Google Scholar]
  7. Kanwar A, Malik S, Prokop LJ, Sim LA, Feldstein D, Wang Z, Murad MH. The association between anxiety disorders and suicidal behaviors: a systematic review and meta-analysis. Depression and anxiety. 2013;30:917–929. doi: 10.1002/da.22074. [DOI] [PubMed] [Google Scholar]
  8. Kato M, Serretti A. Review and meta-analysis of antidepressant pharmacogenetic findings in major depressive disorder. Mol Psychiatry. 2010;15:473–500. doi: 10.1038/mp.2008.116. [DOI] [PubMed] [Google Scholar]
  9. Li YC, Wang LL, Pei YY, Shen JD, Li HB, Wang BY, Bai M. Baicalin decreases SGK1 expression in the hippocampus and reverses depressive-like behaviors induced by corticosterone. Neuroscience. 2015;311:130–137. doi: 10.1016/j.neuroscience.2015.10.023. [DOI] [PubMed] [Google Scholar]
  10. Licznerski P, Duric V, Banasr M, Alavian KN, Ota KT, Kang HJ, Jonas EA, Ursano R, Krystal JH, Duman RS, Traumatic Stress Brain, Study G. Decreased SGK1 Expression and Function Contributes to Behavioral Deficits Induced by Traumatic Stress. PLoS biology. 2015;13:e1002282. doi: 10.1371/journal.pbio.1002282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Marchini J, Howie B. Genotype imputation for genome-wide association studies. Nature reviews Genetics. 2010;11:499–511. doi: 10.1038/nrg2796. [DOI] [PubMed] [Google Scholar]
  12. Miyata S, Yoshikawa K, Taniguchi M, Ishikawa T, Tanaka T, Shimizu S, Tohyama M. Sgk1 regulates desmoglein 1 expression levels in oligodendrocytes in the mouse corpus callosum after chronic stress exposure. Biochemical and biophysical research communications. 2015;464:76–82. doi: 10.1016/j.bbrc.2015.05.109. [DOI] [PubMed] [Google Scholar]
  13. Multani PK, Clarke TK, Narasimhan S, Ambrose-Lanci L, Kampman KM, Pettinati HM, Oslin DW, O'Brien CP, Berrettini WH, Lohoff FW. Neuronal calcium sensor-1 and cocaine addiction: a genetic association study in African-Americans and European Americans. Neuroscience letters. 2012;531:46–51. doi: 10.1016/j.neulet.2012.09.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Nakano-Kobayashi A, Tai Y, Nadif Kasri N, Van Aelst L. The X-linked mental retardation protein OPHN1 interacts with Homer1b/c to control spine endocytic zone positioning and expression of synaptic potentiation. The Journal of neuroscience : the official journal of the Society for Neuroscience. 2014;34:8665–8671. doi: 10.1523/JNEUROSCI.0894-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Otowa T, Hek K, Lee M, Byrne EM, Mirza SS, Nivard MG, Bigdeli T, Aggen SH, Adkins D, Wolen A, Fanous A, Keller MC, Castelao E, Kutalik Z, der Auwera SV, Homuth G, Nauck M, Teumer A, Milaneschi Y, Hottenga JJ, Direk N, Hofman A, Uitterlinden A, Mulder CL, Henders AK, Medland SE, Gordon S, Heath AC, Madden PA, Pergadia ML, van der Most PJ, Nolte IM, van Oort FV, Hartman CA, Oldehinkel AJ, Preisig M, Grabe HJ, Middeldorp CM, Penninx BW, Boomsma D, Martin NG, Montgomery G, Maher BS, van den Oord EJ, Wray NR, Tiemeier H, Hettema JM. Meta-analysis of genome-wide association studies of anxiety disorders. Mol Psychiatry. 2016 doi: 10.1038/mp.2016.11. [DOI] [PubMed] [Google Scholar]
  16. Piton A, Gauthier J, Hamdan FF, Lafreniere RG, Yang Y, Henrion E, Laurent S, Noreau A, Thibodeau P, Karemera L, Spiegelman D, Kuku F, Duguay J, Destroismaisons L, Jolivet P, Cote M, Lachapelle K, Diallo O, Raymond A, Marineau C, Champagne N, Xiong L, Gaspar C, Riviere JB, Tarabeux J, Cossette P, Krebs MO, Rapoport JL, Addington A, Delisi LE, Mottron L, Joober R, Fombonne E, Drapeau P, Rouleau GA. Systematic resequencing of X-chromosome synaptic genes in autism spectrum disorder and schizophrenia. Mol Psychiatry. 2011;16:867–880. doi: 10.1038/mp.2010.54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Powell AD, Saintot PP, Gill KK, Bharathan A, Buck SC, Morris G, Jiruska P, Jefferys JG. Reduced gamma oscillations in a mouse model of intellectual disability: a role for impaired repetitive neurotransmission? PLoS One. 2014;9:e95871. doi: 10.1371/journal.pone.0095871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nature genetics. 2006;38:904–909. doi: 10.1038/ng1847. [DOI] [PubMed] [Google Scholar]
  19. Rickels K, Downing R, Schweizer E, Hassman H. Antidepressants for the treatment of generalized anxiety disorder. A placebo-controlled comparison of imipramine, trazodone, and diazepam. Arch Gen Psychiatry. 1993;50:884–895. doi: 10.1001/archpsyc.1993.01820230054005. [DOI] [PubMed] [Google Scholar]
  20. Sato H, Horikawa Y, Iizuka K, Sakurai N, Tanaka T, Shihara N, Oshima A, Takeda J, Mikuni M. Large, scale analysis of glucocorticoid target genes in rat hypothalamus. Journal of neurochemistry. 2008;106:805–814. doi: 10.1111/j.1471-4159.2008.05489.x. [DOI] [PubMed] [Google Scholar]
  21. Schork NJ, Murray SS, Frazer KA, Topol EJ. Common vs. rare allele hypotheses for complex diseases. Curr Opin Genet Dev. 2009;19:212–219. doi: 10.1016/j.gde.2009.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Tiwari AK, Souza RP, Muller DJ. Pharmacogenetics of anxiolytic drugs. J Neural Transm (Vienna) 2009;116:667–677. doi: 10.1007/s00702-009-0229-6. [DOI] [PubMed] [Google Scholar]
  23. Won H, Mah W, Kim E. Autism spectrum disorder causes, mechanisms, and treatments: focus on neuronal synapses. Frontiers in molecular neuroscience. 2013;6:19. doi: 10.3389/fnmol.2013.00019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Yuen EY, Liu W, Karatsoreos IN, Ren Y, Feng J, McEwen BS, Yan Z. Mechanisms for acute stress-induced enhancement of glutamatergic transmission and working memory. Molecular psychiatry. 2011;16:156–170. doi: 10.1038/mp.2010.50. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental
NIHMS904232-supplement-Supplemental.pdf (1.9MB, pdf)

RESOURCES