A functional polymorphism in the transcriptional control region upstream, the serotonin transporter (SLC6A) coding sequence (5-HTTLPR) has been consistently associated with antidepressant response. More recently, another functional single nucleotide polymorphism within l variant has been reported, and the l variant with an adenosine at single nucleotide polymorphism rs25531 (lA) has been reported to have higher activity compared with the long variant with a guanine (lG)1 and lG expression is nearly equal to the s allele. Therefore, in the present work, we aimed to compare the effects of paroxetine and fluvoxamine in our previously reported sample expanding the analysis to rs25531 variants in a larger sample. The subjects of this study were in part previously reported investigating 5-HTTLPR.2 The method and design of this clinical trial were described elsewhere.2 Briefly, a total of 100 Japanese patients affected by major depression were randomly assigned to either paroxetine or fluvoxamine in a 6-week study. Response was evaluated by the percentage of the Hamilton Rating Scale for Depression (HAM-D) score change and by the rate of “response” defined as an at least 50% decrease after medication. The study was approved by the ethical committee of the Kansai Medical University and Osaka University. Written informed consent was obtained from each subject. Plasma levels of selective serotonin reuptake inhibitors (SSRIs) were determined after at least 2 weeks of stable dosage to evaluate compliance. Genomic DNA was isolated, and we detected lA and lG alleles by real-time polymerase chain reaction method following a previous study.1 Repeated-measures analysis of variance or analysis of covariance (ANCOVA) with baseline HAM-D scores were included in the model as covariant was performed. The low-expression s and lG alleles were grouped together as S' compared with the high-expression lA allele as L' following evidence from previous studies.1 For response rate, a logistic regression analysis was performed using response rate as the dependent variable and age, sex, dose, with or without sleep-induced drugs, HAM-D baseline, and the type of SSRIs as independent variables. Results were considered significant with an [alpha] level lower than 0.05. Eighty-one patients completed the 6-week study, whereas 19 patients withdrew during the course of the study. Genotype frequencies were s/s = 60.4%, s/lA = 22.2%, s/lG = 9.9%, lA/lA = 3.7%, lA/lG = 3.7%, and lG/lG = 0%. All were in Hardy-Weinberg equilibrium. Sociodemographic and clinical characteristics of the samples are shown in Supplemental Table 1, Supplemental Digital Content 1, http://links.lww.com/JCP/A159. No significant differences were found for all sociodemographic and clinical characteristic factors between 2 SSRIs in each genotype group. The triallelic polymorphism was marginally associated with HAM-D score change at week 6 in the total sample (P = 0.04). Subsequent comparison of paroxetine and fluvoxamine in each genotype group showed that fluvoxamine-treated subjects were significantly less improved compared with paroxetine in S'/S' carriers on the HAM-D change over time (P = 0.0002), at week 2 (P = 0.016), at week 4 (P = 0.0004), and at week 6 (P = 0.001); mean HAM-D change of paroxetine and fluvoxamine at week 6 was 76.6% and 55.5%, respectively, whereas in L' allele carriers, no significant difference was found, and mean HAM-D change of both SSRIs was approximately 75% (Fig. 1). Response rate of subjects prescribed with paroxetine (week 4, 80.0%; week 6, 88.0%) was significantly higher compared with fluvoxamine (week 4, 43.8%; week 6, 62.5%) in S' allele homozygotes (week 4, P = 0.008; week 6, P = 0.038), whereas in L' allele carriers, response rate of paroxetine and fluvoxamine were, respectively, 78.6% and 100% at week 4 and 92.9% and 100% at week 6 without significant difference between the 2 agents (Supplemental Table 1, Supplemental Digital Content 1, http://links.lww.com/JCP/A159). When the sample was stratified by l and s variants only, therefore repeating our previously published analysis in this expanded sample not considering rs25531, a similar but less pronounced difference between 2 SSRIs was observed in s homozygotes only (repeated-measures ANCOVA, P = 0.013). However, at week 6, a different HAM-D change of paroxetine and fluvoxamine was observed in both variants (s: 73.7% and 58.9%; ANCOVA, P = 0.026; l: 80.4% and 62.4%; ANCOVA, P = 0.028, respectively) Supplemetal Figure 1, Supplemental Digital Content 2, http://links.lww.com/JCP/A160. Figure 1 Back to Top DISCUSSION The association of 5-HTTLPR with SSRI response is not always consistent, especially in Asian populations. Among other explanations, this heterogeneity could be explained by different types of SSRIs. In fact, we previously reported a different effect on paroxetine and fluvoxamine. In the present article, we observed that, taking into consideration rs25531 variants, the difference was even larger. The rs25531 is putatively located in the sixth repeat of the 5-HTTLPR, and the lG allele reduces SLC6A4 messenger RNA expression to levels nearly equivalent to those of the short allele,1 whereas the lA allele confers higher SLC6A4 expression. We observed that S' homozygotes (s and lG) benefit less from SSRI treatments compared with L' (lA) carriers. Interestingly, even for the same class of antidepressants, response was differently influenced by these variants in each compound. L' carriers had a better response than S' homozygotes only for fluvoxamine, whereas most of the patients with paroxetine had a good response regardless of genotype. When these 2 SSRIs were compared, stratified by only long and short variants of 5-HTTLPR without consideration of rs25531, the result was similar to a certain degree but not the same. That is, s/s carriers influenced less fluvoxamine outcome compared with when rs25531 was also considered, whereas both agents had a beneficial effect over time on l allele carriers, although with some difference at week 6. Therefore, a potential confounding effect of rs25531 was demonstrated. A possible reason lies in the pharmacodynamic profile. Paroxetine [Ki = 0.83 (0.06), mean (SE)] is a more potent serotonin reuptake inhibitor than fluvoxamine [Ki = 14 (1), mean (SE)] in the human brain.3 In addition, paroxetine has also a higher degree of affinity for norepinephrine transporter [Ki = 85 (5), mean (SE)] compared with fluvoxamine [Ki = 2950 (103), mean (SE)]. Thus, paroxetine could preserve the effectiveness regardless of genetic impact of SLC6A4, even in S' homozygotes. This is, to our best knowledge, the first open-label randomized controlled trial to compare SSRIs stratified by genetic influence of rs25531A > G on 5-HTTLPR. Two previous randomized controlled trials have demonstrated no significant differences in clinical response to treatment of depression between paroxetine and fluvoxamine in white populations,4,5 and there are no direct comparisons of paroxetine versus fluvoxamine in Asians. A much higher frequency of lA allele in whites compared with Japanese population could explain the different findings. Some previous studies investigated the association of SSRIs response with lA, lG, and s variants in white depressed patients. Our results were consistent with a previous study that tested its impact on paroxetine treatment showing no association.6 However, our results are in part opposite to a study using fluvoxamine that showed favorable response of lG compared with lA.7 This inconsistency could be due to low lG allele frequency and to the inclusion of bipolar disorder. There are possible limitations of the present study linked to the sample. The sample was already used in our previous articles, and results must therefore be interpreted with caution. In S'/S' carrier, we had a power of 0.80 to detect a large effect size of d = 0.76, which corresponded to a difference of approximately 16.8% HAM-D change between 2 SSRIs. Although in L' allele carriers, small sample size with insufficient power could lead type II error; however, the significant difference observed between 2 SSRIs in L allele carriers stratified by simple 5-HTTLPR not considering rs25531 could suggest considerable effect of rs25531 polymorphism. In conclusion, we expanded our previous study in terms of sample size and adding a further variant. This allowed us to evidence that rs25531 has an effect, particularly in fluvoxamine-treated subjects.

5-HTTLPR rs25531A > G differentially influence paroxetine and fluvoxamine antidepressant efficacy: a randomized, controlled trial.

SERRETTI, ALESSANDRO;
2013

Abstract

A functional polymorphism in the transcriptional control region upstream, the serotonin transporter (SLC6A) coding sequence (5-HTTLPR) has been consistently associated with antidepressant response. More recently, another functional single nucleotide polymorphism within l variant has been reported, and the l variant with an adenosine at single nucleotide polymorphism rs25531 (lA) has been reported to have higher activity compared with the long variant with a guanine (lG)1 and lG expression is nearly equal to the s allele. Therefore, in the present work, we aimed to compare the effects of paroxetine and fluvoxamine in our previously reported sample expanding the analysis to rs25531 variants in a larger sample. The subjects of this study were in part previously reported investigating 5-HTTLPR.2 The method and design of this clinical trial were described elsewhere.2 Briefly, a total of 100 Japanese patients affected by major depression were randomly assigned to either paroxetine or fluvoxamine in a 6-week study. Response was evaluated by the percentage of the Hamilton Rating Scale for Depression (HAM-D) score change and by the rate of “response” defined as an at least 50% decrease after medication. The study was approved by the ethical committee of the Kansai Medical University and Osaka University. Written informed consent was obtained from each subject. Plasma levels of selective serotonin reuptake inhibitors (SSRIs) were determined after at least 2 weeks of stable dosage to evaluate compliance. Genomic DNA was isolated, and we detected lA and lG alleles by real-time polymerase chain reaction method following a previous study.1 Repeated-measures analysis of variance or analysis of covariance (ANCOVA) with baseline HAM-D scores were included in the model as covariant was performed. The low-expression s and lG alleles were grouped together as S' compared with the high-expression lA allele as L' following evidence from previous studies.1 For response rate, a logistic regression analysis was performed using response rate as the dependent variable and age, sex, dose, with or without sleep-induced drugs, HAM-D baseline, and the type of SSRIs as independent variables. Results were considered significant with an [alpha] level lower than 0.05. Eighty-one patients completed the 6-week study, whereas 19 patients withdrew during the course of the study. Genotype frequencies were s/s = 60.4%, s/lA = 22.2%, s/lG = 9.9%, lA/lA = 3.7%, lA/lG = 3.7%, and lG/lG = 0%. All were in Hardy-Weinberg equilibrium. Sociodemographic and clinical characteristics of the samples are shown in Supplemental Table 1, Supplemental Digital Content 1, http://links.lww.com/JCP/A159. No significant differences were found for all sociodemographic and clinical characteristic factors between 2 SSRIs in each genotype group. The triallelic polymorphism was marginally associated with HAM-D score change at week 6 in the total sample (P = 0.04). Subsequent comparison of paroxetine and fluvoxamine in each genotype group showed that fluvoxamine-treated subjects were significantly less improved compared with paroxetine in S'/S' carriers on the HAM-D change over time (P = 0.0002), at week 2 (P = 0.016), at week 4 (P = 0.0004), and at week 6 (P = 0.001); mean HAM-D change of paroxetine and fluvoxamine at week 6 was 76.6% and 55.5%, respectively, whereas in L' allele carriers, no significant difference was found, and mean HAM-D change of both SSRIs was approximately 75% (Fig. 1). Response rate of subjects prescribed with paroxetine (week 4, 80.0%; week 6, 88.0%) was significantly higher compared with fluvoxamine (week 4, 43.8%; week 6, 62.5%) in S' allele homozygotes (week 4, P = 0.008; week 6, P = 0.038), whereas in L' allele carriers, response rate of paroxetine and fluvoxamine were, respectively, 78.6% and 100% at week 4 and 92.9% and 100% at week 6 without significant difference between the 2 agents (Supplemental Table 1, Supplemental Digital Content 1, http://links.lww.com/JCP/A159). When the sample was stratified by l and s variants only, therefore repeating our previously published analysis in this expanded sample not considering rs25531, a similar but less pronounced difference between 2 SSRIs was observed in s homozygotes only (repeated-measures ANCOVA, P = 0.013). However, at week 6, a different HAM-D change of paroxetine and fluvoxamine was observed in both variants (s: 73.7% and 58.9%; ANCOVA, P = 0.026; l: 80.4% and 62.4%; ANCOVA, P = 0.028, respectively) Supplemetal Figure 1, Supplemental Digital Content 2, http://links.lww.com/JCP/A160. Figure 1 Back to Top DISCUSSION The association of 5-HTTLPR with SSRI response is not always consistent, especially in Asian populations. Among other explanations, this heterogeneity could be explained by different types of SSRIs. In fact, we previously reported a different effect on paroxetine and fluvoxamine. In the present article, we observed that, taking into consideration rs25531 variants, the difference was even larger. The rs25531 is putatively located in the sixth repeat of the 5-HTTLPR, and the lG allele reduces SLC6A4 messenger RNA expression to levels nearly equivalent to those of the short allele,1 whereas the lA allele confers higher SLC6A4 expression. We observed that S' homozygotes (s and lG) benefit less from SSRI treatments compared with L' (lA) carriers. Interestingly, even for the same class of antidepressants, response was differently influenced by these variants in each compound. L' carriers had a better response than S' homozygotes only for fluvoxamine, whereas most of the patients with paroxetine had a good response regardless of genotype. When these 2 SSRIs were compared, stratified by only long and short variants of 5-HTTLPR without consideration of rs25531, the result was similar to a certain degree but not the same. That is, s/s carriers influenced less fluvoxamine outcome compared with when rs25531 was also considered, whereas both agents had a beneficial effect over time on l allele carriers, although with some difference at week 6. Therefore, a potential confounding effect of rs25531 was demonstrated. A possible reason lies in the pharmacodynamic profile. Paroxetine [Ki = 0.83 (0.06), mean (SE)] is a more potent serotonin reuptake inhibitor than fluvoxamine [Ki = 14 (1), mean (SE)] in the human brain.3 In addition, paroxetine has also a higher degree of affinity for norepinephrine transporter [Ki = 85 (5), mean (SE)] compared with fluvoxamine [Ki = 2950 (103), mean (SE)]. Thus, paroxetine could preserve the effectiveness regardless of genetic impact of SLC6A4, even in S' homozygotes. This is, to our best knowledge, the first open-label randomized controlled trial to compare SSRIs stratified by genetic influence of rs25531A > G on 5-HTTLPR. Two previous randomized controlled trials have demonstrated no significant differences in clinical response to treatment of depression between paroxetine and fluvoxamine in white populations,4,5 and there are no direct comparisons of paroxetine versus fluvoxamine in Asians. A much higher frequency of lA allele in whites compared with Japanese population could explain the different findings. Some previous studies investigated the association of SSRIs response with lA, lG, and s variants in white depressed patients. Our results were consistent with a previous study that tested its impact on paroxetine treatment showing no association.6 However, our results are in part opposite to a study using fluvoxamine that showed favorable response of lG compared with lA.7 This inconsistency could be due to low lG allele frequency and to the inclusion of bipolar disorder. There are possible limitations of the present study linked to the sample. The sample was already used in our previous articles, and results must therefore be interpreted with caution. In S'/S' carrier, we had a power of 0.80 to detect a large effect size of d = 0.76, which corresponded to a difference of approximately 16.8% HAM-D change between 2 SSRIs. Although in L' allele carriers, small sample size with insufficient power could lead type II error; however, the significant difference observed between 2 SSRIs in L allele carriers stratified by simple 5-HTTLPR not considering rs25531 could suggest considerable effect of rs25531 polymorphism. In conclusion, we expanded our previous study in terms of sample size and adding a further variant. This allowed us to evidence that rs25531 has an effect, particularly in fluvoxamine-treated subjects.
Kato M; Nonen S; Serretti A; Tetsuo S; Takekita Y; Azuma J; Kinoshita T.
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/145597
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? 1
  • Scopus 14
  • ???jsp.display-item.citation.isi??? 13
social impact