Norepinephrine Transporter Gene Variants and Remission From Depression With Venlafaxine Treatment in Older Adults
Publication: American Journal of Psychiatry
Abstract
Objective:
The primary objective of this study was to investigate five putatively functional variants of the norepinephrine transporter (SLC6A2, NET) and serotonin transporter (SLC6A4, SERT) genes and remission in depressed older adults treated with venlafaxine. A secondary objective was to analyze 17 other variants in serotonergic system genes (HTR1A, HTR2A, HTR1B, HTR2C, TPH1, TPH2) potentially involved in the mechanism of action of venlafaxine.
Method:
The sample included 350 adults age 60 or older with DSM-IV-defined major depressive disorder and a score of at least 15 on the Montgomery-Åsberg Depression Rating Scale (MADRS). Participants received protocolized treatment with open-label venlafaxine, up to 300 mg/day for approximately 12 weeks, as part of a three-site clinical trial. Each individual was genotyped for 22 polymorphisms in eight genes, which were tested for association with venlafaxine remission (a MADRS score ≤10) and changes in MADRS score during treatment.
Results:
After adjusting for multiple comparisons, NET variant rs2242446 (T-182C) was significantly associated with remission (odds ratio=1.66, 95% CI=1.13, 2.42). Individuals with the rs2242446 C/C genotype were more likely to remit (73.1%) than those with either the C/T (51.8%) or the T/T genotype (47.3%). Individuals with the C/C genotype also had a shorter time to remission than those with the C/T or T/T genotypes and had a greater percentage change in MADRS score from baseline to end of treatment (up to week 12).
Conclusions:
NET rs2242446/T-182C may serve as a biomarker to predict the likelihood of remission with venlafaxine in older adults with major depression. These findings may help to optimize antidepressant outcomes in older adults.
Late-life depression is a disabling mental disorder affecting more than 2 million older adults in the United States (1). More than 50% of patients with late-life depression do not achieve remission with an antidepressant trial (1). In older adults, prolonged depressive symptoms are associated with greater caregiver burden, medical complications, functional impairment, and cognitive decline (2). Therefore, finding predictive biomarkers for antidepressant treatment success in older adults is imperative to achieve timely remission (3). Only a few genetic polymorphisms have been consistently investigated for associations with antidepressant response, resulting in varying levels of evidence for genetic biomarkers (4); therefore, more research is needed in this area.
Venlafaxine is a serotonin-norepinephrine reuptake inhibitor (SNRI) with evidence supporting its use for the treatment of late-life depression (5). Venlafaxine is typically prescribed at dosages between 75 and 300 mg/day, which have been correlated with differential binding affinity and inhibitory properties at norepinephrine and serotonin transporters in vivo. Previous investigations have shown that venlafaxine exerts dose-dependent norepinephrine reuptake inhibition; a dosage of 150 mg/day or higher is sufficient to produce noradrenergic activity (6). Genetic variants across serotonergic and noradrenergic neurotransmitter systems have been investigated for association with venlafaxine treatment response in younger adults, including the serotonin transporter gene (SLC6A4; 5-HTTLPR [7–9], VNTR [9–11], and rs25531 [10, 12]), the serotonin receptor 2A gene (HTR2A; rs7997012 [12] and G-1438A [9, 11]), the tryptophan hydroxylase gene (TPH; A218C [9, 13]), and the norepinephrine transporter gene (NET, SLC6A2; rs5569). Despite these observations in younger adults, to our knowledge, this is the first study to investigate the associations between late-life depression treatment remission variability, venlafaxine dosage, and variants in NET/SERT genes, as well as other potentially involved serotonergic system genes.
Method
Sample
Our sample included participants from the National Institutes of Health–funded multisite Incomplete Response in Late-Life Depression: Getting to Remission trial (14). Participants were age 60 or older and had a DSM-IV diagnosis of major depressive disorder and a score ≥15 on the Montgomery-Åsberg Depression Rating Scale (MADRS) (15). Participants received protocolized treatment with open-label venlafaxine, starting with 37.5 mg/day and titrated up to a maximum of 300 mg/day over approximately 12 weeks. To avoid confounding effects of potential comorbid neurological abnormalities (e.g., early dementia symptoms), participants were screened and excluded if they had a DSM-IV dementia diagnosis or a score <24 on the Mini-Mental State Examination (16). Full eligibility criteria have been described previously (5, 14). The study was approved by the institutional ethical review boards at each of the three recruitment sites, and all subjects provided written informed consent after receiving a complete description of the study procedures.
For this analysis, our primary outcome of interest was remission, defined as a MADRS score ≤10 at the end of treatment. For validation, secondary outcomes of interest included time to remission, defined as time to the first MADRS score ≤10; percentage change in MADRS score from baseline to end of treatment (up to week 12); and repeated-measures MADRS scores across eight time points (baseline, weeks 1, 2, 4, 6, 8, and 10, and treatment end).
SNP Selection and Genotyping
For our primary analysis, we conducted targeted genotyping of three SLC6A4 (5-HTTLPR, VNTR, rs25531) (7–12) and two SLC6A2 (rs2242446, rs5569) (11, 17) variants. These two transporters are primary targets of venlafaxine; therefore, the five SLC6A4/SERT and SLC6A2/NET variants were strongly hypothesized to mediate venlafaxine response. Subsequently, we also selected 17 putatively functional variants in six genes that we hypothesized to moderate the serotonergic action of venlafaxine: HTR1A (rs6295) (18, 19), HTR1B (rs6296, rs130058, rs11568817) (20, 21), HTR2A (rs9567746, rs2274639, rs6311) (9, 11, 12), HTR2C (rs6318, rs10521432, rs1801412, rs3813929, rs17260600, rs518147) (22), TPH1 (rs1800532) (19), and TPH2 (rs11178998, rs11178997, rs4570625) (23, 24). The single-nucleotide polymorphisms (SNPs) were genotyped on an OpenArray platform and underwent quality control procedures in PLINK, version 1.07 (25), including control for minor allele frequency >5%, Hardy-Weinberg equilibrium (p>0.0001), genotype call rates >98% and individual missingness <10% (to ensure good DNA quality) (26). HTR2C variants rs1801412 and rs17260600 were excluded from the analysis because they had a minor allele frequency less than 5%. See Table S1 in the data supplement that accompanies the online edition of this article for all genotyped variants and their functional annotations.
Population Stratification Analysis
We conducted additional genotyping of 64 ancestry-informative markers, selected from a list of 128 validated markers (27), to obtain genetic information to assess continental ancestry (TaqMan Ancestry Informative Marker Assay, catalog no. 4404607, Life Technologies, Inc., Burlington, Ontario). It has been shown that as few as 24 ancestry-informative markers is sufficient to stratify admixed population samples (28). After quality control, 46 ancestry informative markers were used to conduct multidimensional scaling analysis in PLINK, version 1.9, with HapMap3 (phase 3, release 3, NCBI build 36) populations as reference (see Table S2 in the online data supplement ). Ancestry informative markers were used to verify self-reported ancestry and were not used as covariates to correct for fine population structure. Participants with mismatched ancestry were reclassified as “admixed” if their first and second principal components (obtained from multidimensional scaling analysis) fell six or more standard deviations in either direction from the ancestral cluster (29). To avoid spurious association due to population stratification and to be inclusive of all ethnic groups, we conducted a three-step analysis: 1) a mixed-ancestry sample controlling for the most predictive dimensions, 2) a European-ancestry subsample only, and 3) an African-ancestry subsample only.
Statistical Analysis
Before conducting genetic association analyses, we explored several potential covariates for independent associations across remission, percentage change in MADRS score, and time to remission. Only nonparametric statistical tests were used, as all investigated variables failed the Shapiro-Wilk test of normality and deviated from normal Q-Q distribution. For the remission phenotype, the Mann-Whitney U test was used to analyze potential continuous covariates, and Pearson’s chi-square was used for categorical variables. We used bivariate correlational analyses (Spearman’s rank correlation) to explore correlations between continuous covariates and percentage change in MADRS score. For binary covariates, a difference in percentage change in MADRS score was examined using Mann-Whitney U tests. Lastly, we checked for significant correlations between covariates to avoid overcorrection in our association models.
We examined whether selected SNPs were significantly associated with our primary treatment outcome after controlling for significant covariates. For dichotomous remission, variants were analyzed using multivariate binary logistic regressions (including significant covariates). The percentage change in MADRS score was not normally distributed and was normalized using a rank-aligned transformation (R package: GenABEL) (30). The normalized variable was included in analyses of covariance (ANCOVAs) including variants and covariates. Tukey’s test was performed post hoc to determine significant genotypes. Association of polymorphisms with time to remission was assessed using Kaplan-Meier survival analyses and log rank (Mantel-Cox) calculations. For HTR2C, located on the X chromosome, we conducted analyses separately for male and female subjects.
We investigated genotype association with response across eight treatment time points (baseline, weeks 1, 2, 4, 6, 8, and 10, and treatment end) using a linear mixed-effects model (package: nlme) (31) for repeated measurements in R, version 3.0.2 (R Foundation, r-project.org). For missing data, we conducted multiple imputations using the MICE package (multivariate imputation by chained equations) (32) in R. We created 10 imputed data sets based on the predictive mean matching imputation method. Aside from seven time points, we included significantly associated covariates in our models.
All association analyses were conducted in SPSS, version 22.0 (IBM, Armonk, N.Y.). We adjusted for multiple testing across primary and secondary analyses separately using Bonferroni correction, given the lack of correlation between genotyped SNPs. This resulted in adjusted p thresholds, for an alpha of 0.05, of 0.0125 and 0.0031 for the primary and secondary analyses, respectively. Achieved power was calculated using Quanto, version 1.2.4 (USC Biostats) (33). We used HaploReg (34) to investigate the functional consequences of the SNPs and SNAP (SNP Annotation and Proxy Search), version 2.2 (Broad Institute) to calculate linkage disequilibrium.
Lastly, neural-specific expression profiles based on expression quantitative trait loci analyses for SNPs were retrieved from BRAINEAC (UK Brain Expression Consortium) (35). In the BRAINEAC sample, 10 postmortem CNS tissues were obtained from healthy control individuals of European ancestry (N=134). RNA expression was conducted using the Affymetrix Exon 1.0 ST Array, and genotyping was conducted using the Illumina Infinium Omni1-Quad BeadChip (35).
Results
Sample Characteristics
Our final sample included 350 participants of European (N=311, 88.9%), African (N=33, 9.4%), Asian (N=5, 1.4%), and admixed (N=1, 0.3%) ancestries (Table 1) (see also Figure S1 in the online data supplement ). The sample consisted predominantly of women (N=223, 63.7%), and the mean age was 68.6 years (SD=7.0). At baseline, participants were typically moderately depressed (mean MADRS score, 26.6 [SD=5.6]), and most had recurrent depression (N=255, 72.9%). On average, participants spent 94.6 days (SD=18.4) in the treatment protocol. The mean venlafaxine dosage was 241.6 mg/day (SD=70.3). Overall, 179 participants (51.1%) remitted, and the mean percentage change in MADRS score from baseline to end of treatment was −50.3 (SD=37.4%).
| Characteristic | Total (N=350) | Achieved Remission (N=179) | Did Not Achieve Remission (N=171) | |||
|---|---|---|---|---|---|---|
| % | N | % | N | % | N | |
| Female | 63.7 | 223 | 71.5 | 128 | 55.6 | 95 |
| Ethnic ancestry | ||||||
| European | 88.9 | 311 | 88.3 | 158 | 89.5 | 153 |
| African | 9.4 | 33 | 10.1 | 18 | 8.8 | 15 |
| Asian-Pacific | 1.4 | 5 | 1.1 | 2 | 1.8 | 3 |
| Admixed | 0.3 | 1 | 0.6 | 1 | — | — |
| Recruitment site | ||||||
| Pittsburgh | 41.7 | 146 | 48.6 | 87 | 34.5 | 59 |
| Toronto | 27.4 | 96 | 26.0 | 47 | 28.7 | 49 |
| St. Louis | 30.9 | 108 | 25.1 | 45 | 36.8 | 63 |
| Recurrent major depressive episode | 72.9 | 255 | 75.4 | 135 | 70.2 | 120 |
| Mean | SD | Mean | SD | Mean | SD | |
| Age (years) | 68.6 | 7.0 | 69.9 | 7.6 | 67.3 | 6.1 |
| Baseline MADRS score | 26.6 | 5.6 | 24.8 | 5.3 | 28.4 | 5.4 |
| Change in MADRS score (%) | –50.3 | 37.4 | –82.0 | 12.6 | –17.0 | 22.5 |
| Treatment duration (days) | 94.6 | 18.4 | 97.0 | 18.6 | 92.4 | 18.3 |
| Age at onset (years) | 41.1 | 21.8 | 42.3 | 21.9 | 39.0 | 21.3 |
| Duration of current episode (weeks) | 319.6 | 667.6 | 227.7 | 540.5 | 430.4 | 782.1 |
| Venlafaxine dosage (mg/day) | 241.6 | 70.3 | 206.8 | 72.8 | 278.3 | 44.1 |
a
MADRS=Montgomery-Åsberg Depression Rating Scale.
Exploration of Potential Clinical Covariates
Several clinical variables were found to be significantly and independently associated with remission status in the total (mixed ancestry) sample and the European sample. Compared with patients who did not remit, those who remitted were more likely to be older (mean=69.9 years [SD=7.6] compared with mean=67.4 years [SD=6.1]; p=0.001), to be female (χ2=9.63, df=1, p=0.002), to have a shorter duration of illness (mean=227.7 weeks [SD=540.5] compared with mean=430.4 weeks [SD=782.1]; t=2.47, df=335, p=0.014), and to have lower baseline MADRS scores (mean=24.8 [SD=5.3] compared with mean=28.4 [SD=5.4]; t=6.32, df=348, p<0.001). Patients who remitted spent a longer time in the treatment protocol (mean=97.0 days [SD=18.8] compared with mean=92.4 days [SD=18.4]; t=−2.58, df=348, p=0.010). Remission rates also varied across recruitment sites (χ2=8.23, df=2, p=0.016). All variables that were significantly associated with remission were also significantly associated with percentage change in MADRS score and time to remission. Duration of illness (for the current major depressive episode) was dropped as a potential covariate because of a significant correlation with baseline MADRS score (rs=0.13, p=0.017) and a small effect on remission status (β=−0.13, p=0.13). There was no association of ethnicity with remission status, percentage change in MADRS score, or time to remission. Based on the above univariate analysis and removal of significantly correlated covariates, we included in the ANCOVA model age, sex, baseline MADRS score, treatment duration, and site. See Table S3 in the online data supplement for a full description of these covariates across phenotypes.
Norepinephrine and Serotonin Transporter Variants and Remission Status
NET (SLC6A2) variant rs2242446 was significantly associated with remission/nonremission in the total (mixed ancestry) sample (odds ratio=1.66, 95% CI=1.13, 2.42, p=0.009) (Figure 1) and nominally associated in the European-ancestry subsample (odds ratio=1.67, 95% CI=1.11, 2.46, p=0.013; see Figure S2 in the data supplement ) after adjusting for covariates and multiple comparisons (see Table S4 in the data supplement for a full breakdown of the logistic regression). In the total sample, individuals with the rs2242446 C/C genotype were more likely to remit (73.1%) than those with either the C/T (51.8%) or the T/T genotype (47.3%). There were no significant differences among rs2242446 genotypes in terms of age, age at onset, duration of illness, duration of treatment, and baseline MADRS score. There was also no significant effect of ancestry (see Table S5 in the data supplement ). The same pattern of association of rs2242446 with remission but lack of genotype differences across other baseline variables were also seen in the European-ancestry subsample but not in the smaller African-ancestry subsample (N=31) (see Table S6 in the data supplement ).
FIGURE 1. Association of SCL6A2 Variant rs2242446 With Remission, Percentage Change From Baseline to End of Treatment in MADRS Score, and Time to Remission in the Total Sample (N=350)a
a In panel A, there was a significant association between remission status and SLC6A2 rs2242446 genotype (odds ratio=1.67, 95% CI=1.13, 2.42, p=0.009). Panel B is a box-and-whisker plot showing the first quartile (25%, lower end), the median, and the third quartile (75%, upper end) for the variable percentage change in MADRS score. There was a significant association between percentage change in MADRS score and rs2242446 (partial eta squared, η2=0.03, p=0.006). Panel C illustrates time to remission by rs2242446 status. There was a significant association between time to remission and rs2242446 (Mantel-Cox χ2=9.47, df=2, p=0.009). MADRS=Montgomery-Åsberg Depression Rating Scale.
NET variant rs2242446 was also associated with percentage change in MADRS score (partial eta squared [η2]=0.03, p=0.006) in the total (mixed ancestry) sample and nominally associated in the European-ancestry subsample (η2=0.028, p=0.013). Furthermore, rs2242446 genotype was associated with time to remission in the total sample (Mantel-Cox χ2=9.47, p=0.009) and nominally associated in the European-ancestry sample (χ2=7.84, p=0.02). In the total sample, individuals with the C/C genotype had a shorter time to remission (mean=8.13 weeks [SD=4.64]) than those with the C/T (mean=10.20 weeks [SD=4.19], p=0.089) or T/T (mean=10.65 weeks [SD=4.84], p=0.023) genotypes. In linear mixed-model analyses, NET variant rs2242446 was also nominally associated with change in MADRS score across treatment time points (F=8.08, df=2, p=0.018) before correction for multiple testing. After only 2 weeks of treatment, individuals with the rs2242446 C/C genotype already showed a nominally significant greater reduction in MADRS score (mean change, −38.84% [SD=22.42) than those with the C/T (mean change, −24.51% [SD=26.66]; p=0.034) or T/T genotypes (mean change, −22.46% [SD=27.20]; p=0.010) (Figure 2). Given that individuals with the C/C genotype showed earlier and better response, they also required a significantly lower dosage of venlafaxine on average (mean=209.1 mg [SD=78.8]) at the end of treatment than those with the T/T genotype (mean=248.1 mg [SD=67.4]; p=0.022) but not those with the C/T genotype (mean=239.4 mg [SD=71.2], p=0.11).
FIGURE 2. MADRS Scores Across Treatment Time Points for the SLC6A2 rs2242446 Genotype in the Total Sample (N=350)a
a Each point represents the mean MADRS score in the total (mixed ancestry) sample at a particular time point. There was a significant association between score across time points and rs2242446 genotype (F=8.08, df=2, 347, p=0.018). Error bars indicate standard deviation. MADRS=Montgomery-Åsberg Depression Rating Scale.
For the serotonin transporter, we observed no significant associations of rs25531/5-HTTLPR and 5-HTTVNTR genotypes with our outcomes of interest in the total (mixed ancestry) sample and the ethnic (European and African) subsamples (p>0.05). Similarly, there were no significant associations with time to remission or across treatment time points.
Other Serotonergic System Variants
Other investigated serotonergic system genes showed inconsistent association across several outcomes before correction for multiple testing. HTR2A variant rs6311 was nominally associated with percentage change in MADRS score in the total sample (η2=0.018, p=0.047) and the European-ancestry (η2=0.020, p=0.044) sample. However, rs6311 was not associated with response across time points, remission, or time to remission in any sample. Similarly, none of the other serotonergic system genes (HTR1A, HTR1B, HTR2C, TPH1, and TPH2) showed any consistent significant associations across multiple outcomes (Table 2) (see also Table S6 in the online data supplement ).
| Gene | Variant | European-Ancestry Subsample (N=311) | Total (Mixed Ancestry) Sample (N=350) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Analysis for % Change in MADRS Score (η2) | Remission | Analysis for Time to Remission (χ2) | Analysis for % Change in MADRS Score (η2) | Remission | Analysis for Time to Remission (χ2) | ||||
| Odds Ratio | 95% CI | Odds Ratio | 95% CI | ||||||
| Primary analyses | |||||||||
| SLC6A2 | rs2242446 | 0.028* | 1.67*b | 1.11, 2.46 | 7.84*b | 0.030**b | 1.66**b | 1.13, 2.42 | 9.47**b |
| rs5569 | 0.005 | 0.80 | 0.55, 1.16 | 0.19 | 0.004 | 0.82 | 0.58, 1.17 | 0.21 | |
| SLC6A4 | rs25531-LPR | 0.003 | 0.98 | 0.84, 1.14 | 3.08 | 0.006 | 0.98 | 0.86, 1.11 | 3.23 |
| VNTR | 0.027 | 1.33 | 0.97, 1.83 | 21.6 | 0.026 | 1.23 | 0.91, 1.66 | 21.67***b | |
| Secondary analyses | |||||||||
| HTR1A | rs6295 | 0.001 | 1.02 | 0.72, 1.44 | 0.04 | <0.001 | 1.00 | 0.72, 1.40 | 0.14 |
| HTR1B | rs11568817 | 0.001 | 0.98 | 0.70, 1.39 | 1.61 | 0.002 | 0.99 | 0.71, 1.36 | 1.27 |
| rs130058 | 0.012 | 1.12 | 0.76, 1.65 | 6.00* | 0.011 | 1.12 | 0.77, 1.63 | 6.36* | |
| rs6296 | 0.002 | 1.08 | 0.73, 1.60 | 0.23 | 0.001 | 1.05 | 0.73, 1.52 | 0.16 | |
| HTR2A | rs2274639 | 0.002 | 0.86 | 0.49, 1.51 | 0.84 | 0.005 | 0.89 | 0.54, 1.46 | 0.66 |
| rs6311 | 0.020* | 0.73 | 0.52, 1.02 | 1.43 | 0.018* | 0.74 | 0.54, 1.03 | 1.50 | |
| rs9567746 | 0.004 | 1.12 | 0.73, 1.73 | 8.53* | 0.004 | 1.04 | 0.69, 1.57 | 5.61 | |
| HTR2Cc | rs3813929F | 0.016 | 1.02 | 0.58, 1.78 | 0.09 | 0.013 | 0.94 | 0.56, 1.59 | 0.10 |
| rs3813929M | 0.005 | 0.95 | 0.55, 1.65 | 0.002 | 0.004 | 0.91 | 0.53, 1.57 | 0.06 | |
| rs51814F | 0.008 | 1.03 | 0.65, 1.62 | 1.09 | 0.005 | 1.05 | 0.69, 1.58 | 0.53 | |
| rs51814M | 0.004 | 0.90 | 0.66, 1.68 | 0.29 | 0.004 | 0.97 | 0.63, 1.49 | 0.04 | |
| rs6318F | 0.021 | 1.10 | 0.57, 2.11 | 7.58* | 0.008 | 1.11 | 0.66, 1.87 | 1.43 | |
| rs6318M | 0.002 | 0.88 | 0.51, 1.51 | 0.001 | <0.001 | 0.80 | 0.47, 1.37 | 0.13 | |
| rs6644093F | 0.006 | 1.78 | 0.92, 3.45 | 4.67 | 0.003 | 1.58 | 0.84, 2.98 | 4.56 | |
| rs6644093M | 0.004 | 0.99 | 0.57, 1.75 | 0.02 | 0.006 | 0.98 | 0.55, 1.72 | 0.01 | |
| rs1801412F | 0.008 | 1.48 | 0.58, 3.80 | 0.54 | 0.004 | 1.33 | 0.52, 3.36 | 0.09 | |
| rs1801412M | 0.013 | 0.89 | 0.33, 2.39 | 0.08 | 0.014 | 0.87 | 0.31, 2.40 | 0.54 | |
| TPH1 | rs1800532 | 0.010 | 0.76 | 0.53, 1.09 | 3.44 | 0.005 | 0.80 | 0.57, 1.13 | 1.97 |
| TPH2 | rs11178997 | 0.002 | 1.20 | 0.64, 2.22 | 0.22 | 0.002 | 1.09 | 0.63, 1.88 | 0.57 |
| rs11178998 | 0.002 | 1.21 | 0.65, 2.23 | 0.20 | 0.001 | 1.16 | 0.64, 2.12 | 0.20 | |
| rs4570625 | 0.007 | 1.33 | 0.86, 2.04 | 3.91 | 0.009 | 1.31 | 0.88, 1.94 | 4.49 | |
a
Nominal significance levels (before correction for multiple testing) are represented as follows: ***p≤0.005, **p≤0.01, *p≤0.05.
b
Significant after correction for multiple testing. Bonferroni-adjusted thresholds are 0.0125 and 0.0031 for the primary and secondary analyses, respectively.
c
Analyses were conducted separately for males and females because of the location of HTR2C on the X chromosome. For the variants, F=female, M=male.
Functional Effects of rs2242446 Neural NET Expression
In order to understand the functional effects of rs2242446 on the expression of the NET gene across different brain regions, we conducted gene expression analyses (35) using data from the UK Brain Expression Consortium (BRAINEAC; http://braineac.org). Using BRAINEAC data, we found that rs2242446 genotypes differentially affect NET gene expression within the hippocampus (p=0.037), the thalamus (p=0.037), and the temporal cortex (p=0.048) (see Figure S3A in the data supplement ). In these regions, individuals with the C/C genotype show significantly lower expression of NET than do either those with the C/T genotype or those with the T/T genotype.
Discussion
This study is, to our knowledge, the first to examine the association between NET and serotonergic system gene polymorphisms in association with venlafaxine remission in late-life depression, using a large and well-characterized sample. We found that the NET gene variant rs2242446 (T-182C) was associated with multiple measures of treatment success in older depressed adults treated with venlafaxine: individuals with the C/C genotype had greater odds of remission, shorter time to remission, and a greater percentage change in depressive symptoms than those with either the T/T or the C/T genotype.
Strikingly, the association of NET rs2242446 with remission was also accompanied by an earlier response. Given that the majority of individuals with the C/C genotype received at least 150 mg/day at the end of treatment (92.3%, N=24), there is reasonable evidence suggesting that the norepinephrine transporter was being sufficiently inhibited to produce clinically meaningful effects. Because of their favorable response, individuals with the C/C genotype required lower dosages of venlafaxine on average. Since response to venlafaxine is likely influenced by additional factors, our findings warrant further investigation of NET to assess the clinical utility of pharmacogenetic testing in older adults. In particular, this preliminary finding requires replication in comparable independent samples to determine its potential to be integrated into currently available genetic algorithms.
There is functional evidence to suggest that NET variant rs2242446 may play a significant role in the regulation of NET expression. Located 1.1 kb upstream of the NET transcription start site, rs2242446 has promoter histone markers and a binding site for the nuclear transcriptional repressor CTCF (CCCTC-binding factor) (34), which are indicative of the ability to enhance the expression of NET. Interestingly, rs2242446 is also in high linkage disequilibrium (linkage disequilibrium r2=0.97) with the variant rs28386840 (T-3081A) polymorphism, which is downstream of NET (2.7 kb). Given their high linkage disequilibrium, this suggests that rs2242446 and rs28386840 may serve as proxy indicators of the functional effect either variant would have on NET expression. In vitro functional evidence has shown that the wild-type rs28386840 T allele significantly reduces NET promoter function, resulting in decreased expression (36). Individuals with the rs28386840 T/T show significantly lower NET expression in the hippocampus (p=0.013), which is similar to the lower expression seen in those with the rs2242446 C/C genotypes (see Figure S3B and S3A in the online data supplement ) (35). Given the lower neural expression of NET and observed better response with C/C genotypes, further investigations are warranted to explore the functional consequences of rs2242446 in NET gene expression and venlafaxine action.
Our findings should be placed in the context of previous studies: There have been two other studies, in younger individuals, that examined the NET rs2242446 polymorphism in relation to antidepressant response (11, 37). In an Australian sample of European ancestry, Singh et al. (37) reported that individuals with a history of childhood abuse carrying the C allele of the rs2242446 marker were more likely to respond to venlafaxine or escitalopram treatment (N=38). Although these findings are consistent with our results, the sample used by Singh et al. was small and the authors did not conduct genotypic analysis—rather, allelic analysis was conducted comparing C allele carriers (N=29) and T/T genotype carriers (N=9). In a study of Japanese patients with major depression treated with the SNRI milnacipran (N=96), analyses indicated that the rs2242446 T allele was associated with better response, but no genotypic association was detected (11). This negative finding with respect to genotype is likely due to the smaller sample size, differences in age and in population investigated (i.e., European versus Asian ancestries), and differential affinities for the norepinephrine transporter between venlafaxine and milnacipran. In sum, findings from three studies support the association of rs2242446 with SNRI response. However, the effects may be modulated by additional factors, such as antidepressant type, patient ethnicity, early childhood adverse events, and age, which should be further dissected.
We also observed a nominal association between HTR2A variant rs6311 (A-1438G) and percentage change in MADRS score across in the total (mixed ancestry) sample and the ethnic subsamples. However, the association was seen only when we adjusted for covariates and not independently with genotype. Variant rs6311 is an upstream (308 base pairs) promoter variant of HTR2A that may initiate gene transcription. Although functional evidence warrants further investigation of rs6311, a meta-analysis showed no significant association with rs6311 and remission with various antidepressants across 10 studies (38). Our results add to previous findings (7, 8, 10, 11) suggesting a minor role for HTR2A rs6311 in contributing to remission during treatment with venlafaxine.
Overall, our multivariate logistic regression models were able to account for about 25% of the initial variance, suggesting that treatment response may be a complex and polygenic phenotype involving other genes that may also affect venlafaxine metabolism (e.g., CYP2D6) or other mechanisms, such as epigenetic regulation. Nonetheless, our post hoc power analyses for genetic studies revealed that we had a 74% power to detect a multivariate odds ratio of 1.67 or a partial eta squared of 0.03 for rs2242446 with a major allele frequency of 0.74.
Conclusions
In older adults, remission for depression is particularly difficult to achieve (39). We provided further evidence for the involvement of the NET gene in predicting remission to venlafaxine treatment in late-life depression. NET variant rs2242446 (T-182C) was significantly associated with greater and earlier treatment success in our sample of geriatric patients treated with venlafaxine. Our findings serve as a step toward predictive pharmacogenetic testing to deliver personalized medicine.
Footnotes
The samples and clinical data were provided by the principal investigators of Incomplete Response in Late Life Depression: Getting to Remission study. ClinicalTrials.gov identifier: NCT00892047.
Dr. Blumberger has received research support from the Canadian Institutes of Health Research, NIH, Brain Canada, and the Temerty Family through the Centre for Addiction and Mental Health Foundation and the Campbell Research Institute; he receives research support and in-kind equipment support for an investigator-initiated study from Brainsway Ltd., and he is the site principal investigator for three sponsor-initiated studies for Brainsway Ltd.; he also receives in-kind equipment support from Magventure for an investigator-initiated study and medication supplies for an investigator-initiated trial from Indivior. Dr. Karp has received medication supplies for investigator-initiated trials from Indivior and Pfizer. Dr. Lenze has received research support from the McKnight Brain Research Foundation, the Sidney R. Baer Foundation, the Taylor Family Institute for Innovative Psychiatric Research, the Barnes Jewish Foundation, Takeda, Lundbeck, and Roche. Dr. Reynolds has received pharmaceutical support for NIH-sponsored research studies from Bristol-Myers Squibb, Eli Lilly, Forest, and Pfizer; grants from NIMH, the National Heart, Lung, and Blood Institute, the Center for Medicare and Medicaid Services, the Patient-Centered Outcomes Research Institute, the Commonwealth of Pennsylvania, the John A. Hartford Foundation, the National Palliative Care Research Center, the Clinical and Translational Science Institute, and the American Foundation for Suicide Prevention; he has also served on the American Association for Geriatric Psychiatry editorial review board. Dr. Reynolds receives an honorarium for serving as Editor-in-Chief for the American Journal of Geriatric Psychiatry, and he has received a speaking honorarium from MedScape/WebMD; is the coinventor of psychometric analysis of the Pittsburgh Sleep Quality Index (licensed intellectual property PRO10050447). Dr. Mulsant receives research support from Brain Canada, the Canadian Institutes of Health Research, NIH, the Centre for Addiction and Mental Health Foundation; he has received medications for NIH-funded clinical trials from Bristol-Myers Squibb, Eli Lilly, Pfizer, and Pfizer/Wyeth and software used in studies from Capital Solution Design and HAPPYneuron; he owns stock in General Electric. Dr. Müller has received research support from the Canadian Institutes of Health Research, NIH, the Centre for Addiction and Mental Health Foundation (Joanne Murphy Professorship), the Canadian Biomarker Integration Network in Depression, and the Ontario Brain Institute. The other authors report no financial relationships with commercial interests.
Supplementary Material
File (appi.ajp.2016.16050617.ds001.pdf)
- View/Download
- 768.86 KB
References
1.
Whyte EM, Basinski J, Farhi P, et al: Geriatric depression treatment in nonresponders to selective serotonin reuptake inhibitors. J Clin Psychiatry 2004; 65:1634–1641
2.
Lenze EJ, Schulz R, Martire LM, et al: The course of functional decline in older people with persistently elevated depressive symptoms: longitudinal findings from the Cardiovascular Health Study. J Am Geriatr Soc 2005; 53:569–575
3.
Reynolds CF 3rd, Dew MA, Pollock BG, et al: Maintenance treatment of major depression in old age. N Engl J Med 2006; 354:1130–1138
4.
Fabbri C, Porcelli S, Serretti A: From pharmacogenetics to pharmacogenomics: the way toward the personalization of antidepressant treatment. Can J Psychiatry 2014; 59:62–75
5.
Joel I, Begley AE, Mulsant BH, et al: Dynamic prediction of treatment response in late-life depression. Am J Geriatr Psychiatry 2014; 22:167–176
6.
Melichar JK, Haida A, Rhodes C, et al: Venlafaxine occupation at the noradrenaline reuptake site: in-vivo determination in healthy volunteers. J Psychopharmacol 2001; 15:9–12
7.
Lee SH, Choi TK, Lee E, et al: Serotonin transporter gene polymorphism associated with short-term treatment response to venlafaxine. Neuropsychobiology 2010; 62:198–206
8.
Ng C, Sarris J, Singh A, et al: Pharmacogenetic polymorphisms and response to escitalopram and venlafaxine over 8 weeks in major depression. Hum Psychopharmacol 2013; 28:516–522
9.
Higuchi H, Takahashi H, Kamata M, et al: Influence of serotonergic/noradrenergic gene polymorphisms on nausea and sweating induced by milnacipran in the treatment of depression. Neuropsychiatr Dis Treat 2009; 5:393–398
10.
Perlis RH, Fijal B, Dharia S, et al: Failure to replicate genetic associations with antidepressant treatment response in duloxetine-treated patients. Biol Psychiatry 2010; 67:1110–1113
11.
Yoshida K, Takahashi H, Higuchi H, et al: Prediction of antidepressant response to milnacipran by norepinephrine transporter gene polymorphisms. Am J Psychiatry 2004; 161:1575–1580
12.
Lohoff FW, Narasimhan S, Rickels K: Interaction between polymorphisms in serotonin transporter (SLC6A4) and serotonin receptor 2A (HTR2A) genes predict treatment response to venlafaxine XR in generalized anxiety disorder. Pharmacogenomics J 2013; 13:464–469
13.
Wang H-C, Yeh TL, Chang HH, et al: TPH1 is associated with major depressive disorder but not with SSRI/SNRI response in Taiwanese patients. Psychopharmacology (Berl) 2011; 213:773–779
14.
Lenze EJ, Mulsant BH, Blumberger DM, et al: Efficacy, safety, and tolerability of augmentation pharmacotherapy with aripiprazole for treatment-resistant depression in late life: a randomised, double-blind, placebo-controlled trial. Lancet 2015; 386:2404–2412
15.
Montgomery SA, Åsberg M: A new depression scale designed to be sensitive to change. Br J Psychiatry 1979; 134:382–389
16.
Folstein MF, Robins LN, Helzer JE: The Mini-Mental State Examination. Arch Gen Psychiatry 1983; 40:812
17.
Hahn MK, Blackford JU, Haman K, et al: Multivariate permutation analysis associates multiple polymorphisms with subphenotypes of major depression. Genes Brain Behav 2008; 7:487–495
18.
Villafuerte SM, Vallabhaneni K, Sliwerska E, et al: SSRI response in depression may be influenced by SNPs in HTR1B and HTR1A. Psychiatr Genet 2009; 19:281–291
19.
Illi A, Setälä-Soikkeli E, Viikki M, et al: 5-HTR1A, 5-HTR2A, 5-HTR6, TPH1, and TPH2 polymorphisms and major depression. Neuroreport 2009; 20:1125–1128
20.
Xu Z, Zhang Z, Shi Y, et al: Influence and interaction of genetic polymorphisms in the serotonin system and life stress on antidepressant drug response. J Psychopharmacol 2012; 26:349–359
21.
Zhang K, Xu Q, Xu Y, et al: The combined effects of the 5-HTTLPR and 5-HTR1A genes modulates the relationship between negative life events and major depressive disorder in a Chinese population. J Affect Disord 2009; 114:224–231
22.
Arias B, Fabbri C, Gressier F, et al: TPH1, MAOA, serotonin receptor 2A and 2C genes in citalopram response: possible effect in melancholic and psychotic depression. Neuropsychobiology 2013; 67:41–47
23.
Scheuch K, Lautenschlager M, Grohmann M, et al: Characterization of a functional promoter polymorphism of the human tryptophan hydroxylase 2 gene in serotonergic raphe neurons. Biol Psychiatry 2007; 62:1288–1294
24.
Zhang X, Gainetdinov RR, Beaulieu JM, et al: Loss-of-function mutation in tryptophan hydroxylase-2 identified in unipolar major depression. Neuron 2005; 45:11–16
25.
Purcell S, Neale B, Todd-Brown K, et al: PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81:559–575
26.
Anderson CA, Pettersson FH, Clarke GM, et al: Data quality control in genetic case-control association studies. Nat Protoc 2010; 5:1564–1573
27.
Nassir R, Kosoy R, Tian C, et al: An ancestry informative marker set for determining continental origin: validation and extension using human genome diversity panels. BMC Genet 2009; 10:39
28.
Kosoy R, Nassir R, Tian C, et al: Ancestry informative marker sets for determining continental origin and admixture proportions in common populations in America. Hum Mutat 2009; 30:69–78
29.
Price AL, Patterson NJ, Plenge RM, et al: Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006; 38:904–909
30.
Aulchenko YS, Ripke S, Isaacs A, et al: GenABEL: an R library for genome-wide association analysis. Bioinformatics 2007; 23:1294–1296
31.
Pinheiro J, Bates D, DebRoy S, et al: nlme: Linear and Nonlinear Mixed Effects Models, 2012 (https://cran.r-project.org/web/packages/nlme/index.html)
32.
van Buuren S, Groothui-Oudshoorn K: MICE: multivariate imputation by chained equations in R. J Stat Softw 2011; 45:67
33.
Gauderman W, Morrison J: QUANTO 1.1: A computer program for power and sample size calculations for genetic-epidemiology studies. 2006 (http://biostats.usc.edu/software)
34.
Ward LD, Kellis M: HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res 2012; 40:D930–D934
35.
Trabzuni D, Ryten M, Walker R, et al: Quality control parameters on a large dataset of regionally dissected human control brains for whole genome expression studies. J Neurochem 2011; 119:275–282
36.
Kim YK, Hwang JA, Lee HJ, et al: Association between norepinephrine transporter gene (SLC6A2) polymorphisms and suicide in patients with major depressive disorder. J Affect Disord 2014; 158:127–132
37.
Singh AB, Bousman CA, Ng CH, et al: Effects of persisting emotional impact from child abuse and norepinephrine transporter genetic variation on antidepressant efficacy in major depression: a pilot study. Clin Psychopharmacol Neurosci 2015; 13:53–61
38.
Niitsu T, Fabbri C, Bentini F, et al: Pharmacogenetics in major depression: a comprehensive meta-analysis. Prog Neuropsychopharmacol Biol Psychiatry 2013; 45:183–194
39.
Lenze EJ, Sheffrin M, Driscoll HC, et al: Incomplete response in late-life depression: getting to remission. Dialogues Clin Neurosci 2008; 10:419–430
Information & Authors
Information
Published In
History
Received: 28 May 2016
Revision received: 23 September 2016
Accepted: 3 October 2016
Published online: 10 January 2017
Published in print: May 01, 2017
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.
For more information or tips please see 'Downloading to a citation manager' in the Help menu.
View Options
View options
PDF/EPUB
View PDF/EPUBLogin options
Already a subscriber? Access your subscription through your login credentials or your institution for full access to this article.
Personal login Institutional Login Open Athens loginNot a subscriber?
PsychiatryOnline subscription options offer access to the DSM-5-TR® library, books, journals, CME, and patient resources. This all-in-one virtual library provides psychiatrists and mental health professionals with key resources for diagnosis, treatment, research, and professional development.
Need more help? PsychiatryOnline Customer Service may be reached by emailing [email protected] or by calling 800-368-5777 (in the U.S.) or 703-907-7322 (outside the U.S.).