The Effect of 5-Hydroxytryptamine 3A and 3B Receptor Genes on Nausea Induced by Paroxetine

The effect of 5-hydroxytryptamine 3A and 3B receptor genes on nausea induced by paroxetine

T Sugai1, Y Suzuki1, We investigated the effect of 5-hydroxytryptamine 3A and 3B receptor 1 1 (HTR3A and HTR3B) gene polymorphisms on nausea induced by paroxetine K Sawamura , N Fukui , in Japanese psychiatric patients. Blood samples were collected from 78 2 1 Y Inoue and T Someya individuals after at least 2 weeks treatment with the same daily dose of paroxetine. The patients visited every 2 weeks and the paroxetine dose was 1Department of Psychiatry, Niigata University Graduate School of Medical and Dental Sciences, changed in response to their clinical symptoms. Nausea was assessed at each Niigata, Japan and 2MP-Technopharma visit. The Tyr129Ser polymorphism of the HTR3B gene had a significant effect Corporation Technology Department, Fukuoka, on the incidence of nausea (P ¼ 0.038). Logistic regression analysis also Japan showed that patients with the Tyr/Tyr genotype had a 3.95-fold (P ¼ 0.048) higher risk of developing nausea than patients with the Ser allele. HTR3A Correspondence: Professor T Someya, Department of Psychiatry, gene polymorphisms and the CYP2D6 gene polymorphisms had no Niigata University Graduate School of Medical significant effect on the incidence of nausea. The mean score of nausea and Dental Sciences, 757 Asahimachidori- severity was corrected by the Bonferroni test. HTR3B gene polymorphisms are ichibancho, Niigata 951-8510, Japan. significant predictors of paroxetine-induced nausea. E-mail: [email protected] The Pharmacogenomics Journal (2006) 6, 351–356. doi:10.1038/sj.tpj.6500382; published online 14 March 2006

Keywords: HTR3A; HTR3B; paroxetine; nausea; gene polymorphism

Introduction

Nausea is a severe side effect induced by selective serotonin reuptake inhibitors (SSRIs). In general, SSRIs are better tolerated than tricyclic antidepressants, although gastrointestinal side effects can have an incidence of up to 40%, which can be severe enough to lead to early treatment discontinuation.1,2 Recently, 5- hydroxytryptamine 3 receptors (HTR3) have been considered to have an important role in SSRI-induced gastrointestinal side effects, since HTR3 antagonists cisapride and ondansetron were reported to reduce SSRI-induced gastrointestinal side effects.3 HTR3 is a ligand-gated ion channel that mediates fast synaptic neurotransmis- sion.4 Central and peripheral HTR3 have different structures and different properties.5 HTR3 exist in the area called the chemoreceptor trigger zone of the medulla oblongata, and are thought to be involved with the vomiting reflex.6 HTR3 are also distributed in the autonomic, enteric and sensory nervous systems.7 They also regulate the control pain sensation, movement of the digestive tract and vomiting by prompting nerve depolarization.8 In particular, modification of HTR3 on the small intestinal mucosa is one of the mechanisms for regulation of antineoplastic-induced nausea and vomiting.9 We do not have any data about the binding affinity of paroxetine for the 5-HT3 receptors. However, HTR3 antagonists such as cisapride and ondansetron were used for the Received 24 October 2005; revised 20 December 2005; accepted 6 February 2006; therapy of nausea induced by SSRIs, and SSRIs such as fluvoxamine display a published online 14 March 2006 relatively high affinity for HTR2A. Thus, we considered that there might be some The relation between the HTR3 gene and nausea T Sugai et al 352

relationship between nausea induced by SSRIs such as 40 mg/day in response to clinical symptoms. We rated the paroxetine and the function of HTR3. side effects during the last 2 weeks and evaluated the Murphy et al.10 report that the T102C polymorphism of severity of nausea according to our original scale, which the HTR2A gene may predict the treatment discontinuation included five graded items: 0, no nausea; 1, mild nausea for caused by paroxetine-induced side effects in older patients less than during the last 2 weeks; 2, mild nausea for more with depression. HTR3A and 3B genes have been assigned to than 1 week during the last 2 weeks; 3, continuous, chromosome 11q23.1–q23.3,11 and several genetic varia- moderate nausea during the last 2 weeks; 4, continuous, tions have been reported. Tremblay et al.12 report that severe nausea and vomiting during the last 2 weeks. Subjects variations in the HTR3B gene predict the efficacy of with a score of 0 or 1 were defined as subjects without antiemetic treatment in cancer patients. However, no nausea, and those with a score or 2, 3 or 4 were defined as previous studies have investigated the effects of HTR3A having nausea. The side effect raters were blind to the and 3B gene polymorphisms on the occurrence of SSRI- patients’ genotypes. induced nausea. Cytochrome P450 (CYP) 2D6 has been shown to be Blood sampling involved in the metabolism of paroxetine, and CYP2D6 is Blood sampling was performed using a Venojects tube known to have genetic polymorphisms that affect enzyme containing EDTA-Na (Terumo Japan, Tokyo, Japan) at week 13 activity. These observations suggest that the CYP2D6 gene 1 for genotype detection and subsequently at the first polymorphism may be a predictor for paroxetine-induced appearance of nausea to measure the concentration of 14 side effects. On the other hand, Gerstenberg et al. showed paroxetine. Blood samples were taken at 12 h after the final that steady-state plasma concentrations of fluvoxamine are ingestion of paroxetine. Venous blood, 7 ml, were collected, not associated with incidence of nausea, and that CYP2D6 and genomic DNA was extracted from the peripheral genotype does not affect nausea development. Therefore, leukocytes by utilizing a QIAamp Blood Kit (QIAGEN, further studies are needed to clarify whether CYP2D6 gene Valencia, CA, USA) within 2 h of collection. polymorphisms affect SSRI-induced side effects. In this study, we investigated the effects of pharmacodynamic factors, such as HTR3A and 3B gene polymorphisms, Genotyping and determination of plasma concentration and the effects of pharmacokinetic factors, such as CYP2D6 Polymerase chain reaction (PCR) was used to determine the genotype, on the occurrence of paroxetine-induced nausea C195T and Pro16Ser genotypes of HTR3A gene according to 15 in Japanese psychiatric patients. the method of Niesler et al. and the Tyr129Ser genotype of HTR3B gene according to the method of Tremblay et al.12 CYP2D6*10 alleles causing decreased enzyme activity were Materials and methods identified by the C188T mutation using a two-step PCR as described by Johansson et al.16 A long-PCR analysis was used Subjects to detect the *5 allele causing a lack of enzyme activity as 17 This study was conducted at Niigata University Medical described by Steen et al. Hospital, Japan, and the study protocol was approved by the The plasma concentration of paroxetine was measured Hospital Ethics Committee. Each subject provided written using column-switching high-performance liquid chroma- informed consent before enrolment. The subjects comprised tography (HPLC) with ultraviolet detection. Paroxetine was 78 Japanese psychiatric outpatients (28 males, 50 females) extracted from plasma, to which cisapride had been added aged 38.4713.8 years (mean7s.d., range 18–70 years.). as an internal standard, with hexane–chloroform, and the Thirty-nine subjects had major depressive disorder, 25 had extract was subjected to automated column-switching HPLC anxiety disorders, six had adjustment disorder, seven had a using a TSK BSA-C8 precolumn (Tosoh, Tokyo, Japan) for depressive disorder not otherwise specified, and one had sample clean-up, and a TSK gel ODS-80TS column (Tosoh) other mood disorders. All patients were diagnosed according for separation. to DSM-IV-TR. The exclusion criteria were additional diagnoses of Axis I or II of DSM-IV-TR. Demographic data, Statistical analysis medical history and laboratory data, including hematology, Statistical analysis was performed using SPSSII for Windows. serology, electrolytes and urine analysis, were collected for Genotype and allele distributions were analyzed by each patient. Patients with obvious physical illnesses were the w2 test. The clinical and demographic characteristics, excluded from the study. All patients were orally treated sex, age, daily dose and paroxetine concentration with paroxetine for their psychiatric illness. No patients were compared among groups by the unpaired t-test. The were being treated with antiemetic medication during our mean score of nausea severity was compared among study. each genotype group by one-way analysis of variance, and post hoc analysis of the mean score of nausea severity Study design was carried out by using Bonferroni’s test. Logistic regression The patients visited the hospital every 2 weeks and side analysis was used to compare the probability of the effects, including nausea, were assessed at each visit. The incidence of nausea. The level of significance was set at paroxetine dose was increased from 10 or 20 to 30 and Po0.05.

The Pharmacogenomics Journal The relation between the HTR3 gene and nausea T Sugai et al 353

Results Effects of HTR3A and HTR3B gene polymorphisms The genotype distribution of HTR3A and HTR3B gene The genotype frequencies of the HTR3A, HTR3B and polymorphisms are shown in Tables 1 and 2. No significant CYP2D6 genes are shown in Tables 1–3. All of these genetic differences were demonstrated for sex, age, and paroxetine variations were in Hardy–Weinberg equilibrium. HTR3A daily dose and plasma concentration among each genotype C195T polymorphism genotype was not detected in four group. Association analysis revealed that genotype frequen- patients. There were no differences in nausea between each cies of HTR3A Pro16Ser and HTR3B C195T polymorphisms diagnostic groups (e.g. major depressive disorders and did not significantly differ between subjects with and anxiety disorders (P ¼ 0.12)). Therefore, we consider that without nausea (Pro16Ser genotype: w2 ¼ 0.912, df ¼ 2, the nausea observed in the present study was induced only P ¼ 0.634; C195T genotype: w2 ¼ 2.128, df ¼ 2, P ¼ 0.546). by paroxetine. There were also no differences in HTR3A, There was a significant difference in genotypic distribution HTR3B and CYP2D6 polymorphisms between major depres- associated with HTR3B Tyr129Ser polymorphism between sive disorders and anxiety disorders (P ¼ 0.341). patients with and without nausea (w2 ¼ 6.547, df ¼ 2,

Table 1 Genotypic distribution and demographic data of the HTR3A Pro16Ser and C195T

HTR3A Pro16Ser HTR3A C195T

Pro/Pro Pro/Ser Ser/Ser C/C C/T T/T

Sex (M/F) 23/34 5/13 0/3 15/22 11/16 1/9 P 0.261 0.291 Age (7s.d.) 37.8 (13.8) 38.2 (14.2) 52.0 (4.4) 37.8 (13.8) 38.2 (14.2) 52.0 (4.4) P 0.411 0.370 Daily dose of PRX (7s.d.) 23.0 (10.5) 20.0 (9.7) 20.0 (17.3) 20.5 (9.7) 25.9 (10.5) 19.0 (12.9) P 0.408 0.205 Concentration of PRX (7s.d.) 66.4 (89.3) 28.9 (31.3) 22.3 (29.8) 42.8 (61.1) 65.8 (90.6) 52.9 (80.6) P 0.167 0.707 Nausea (+) (n ¼ 15) 12 3 0 9 5 1 % (21.1%) (16.7%) (0.0%) (24.3%) (18.5%) (10.0%) Nausea (À)(n ¼ 63) 45 15 3 28 22 9 % (78.9%) (83.3%) (100.0%) (75.7%) (81.5%) (90.0%) P 0.634 0.546 Mean score of severity of nausea 0.21 (0.411) 0.17 (0.383) 0 (0) 0.1 (0.316) 0.19(0.396) 0.24(0.435) P 0.643 0.593

Table 2 Genotypic distribution and demographic data of the HTR3B Tyr129Ser

HTR3B Tyr129Ser

Tyr/Tyr Tyr/Ser Ser/Ser Tyr/Ser+Ser/Ser

Sex (M/F) 16/19 10/26 2/5 12/31 P 0.264 0.103 Age (7s.d.) 36.3 (11.5) 38.9 (15.1) 46.1 (16.6) 40.1 (15.4) P 0.099 0.267 Daily dose of PRX (7s.d.) 21.4 (10.6) 21.9 (10.6) 27.1 (9.5) 22.8 (10.5) P 0.4 0.939 Concentration of PRX (7s.d.) 37.6 (46.9) 66.4 (100.2) 95.0 (77.2) 71.1 (96.6) P 0.124 0.064 Nausea (+) (n ¼ 15) 11 4 0 4 (%) (31.4%) (11.1%) (0.0%) (9.3%) Nausea (À)(n ¼ 63) 24 32 7 39 (%) (68.6%) (88.9%) (100.0%) (90.7%) P 0.038* 0.014* Mean score of severity of nausea 0.54 (0.919) 0.14 (0.487) 0 (0) 0.12 (0.448) P 0.030** 0.009**

*Po0.05, significant difference from patients with nausea (w2 text). **Po0.05, significant difference among the three genotypes (one-way analysis of variance).

The Pharmacogenomics Journal The relation between the HTR3 gene and nausea T Sugai et al 354

Table 3 Genotypic distribution and demographic data of CYP2D6 genotype groups

CYP2D6

*1/*1 *1/*5, *1/10 *5/*10, *10/*10

Sex (M/F) 17/34 5/7 6/9 P 0.807 Age (7s.d.) 38.6 (14.9) 39.7 (10.7) 36.7 (12.9) P 0.126 Daily dose of PRX (7s.d.) 21.4 (10.2) 24.2 (11.7) 23.3 (11.1) P 0.925 Concentration of PRX (7s.d.) 50.7 (76.3) 89.5 (114.4) 47.5 (51.2) P 0.287 Nausea (+) (n ¼ 15) 9 2 4 (%) (17.6) (16.7) (26.7) Nausea (À)(n ¼ 63) 42 10 11 (%) (82.4) (83.3) (73.3) P 0.716 Mean score of severity of nausea 0.18 (0.385) 0.17 (0.389) 0.27 (0.458) P 0.725

Table 4 Logistic regression analysis of independent valiables to Nausea

Independent valiable Partial regression coefficients P Odd ratio (95% confidence interval)

Sex À0.040 0.111 1.115 (0.518–3.225) Age À0.793 0.041 1.200 (0.810–1.110) Daily dose of Paroxetine À0.538 0.320 0.447 (0.112–3.154) HTR3B Tyr129Ser genotype À0.148 0.048* 3.950 (1.009–15.455)

*Po0.05, statistical significance.

P ¼ 0.038). The proportion of Ser allele carriers (i.e., patients and paroxetine daily dose and plasma concentration with either Tyr/Ser or Ser/Ser) was significantly higher in the between those three genotype groups (Table 3). group without nausea (w2 ¼ 6.082, df ¼ 1, P ¼ 0.014). There There were no significant differences in the incidence of were significant differences in the severity score of nausea nausea between the three genotype groups (w2 ¼ 1.029, among the three genotypes (score: 0.5470.91, 0.1470.49 df ¼ 2, P ¼ 0.716). We also divided patients into two and 070, df ¼ 2, P ¼ 0.03). genotype groups: 61 patients with the *1/*1 or *1/*10 The results of logistic regression analysis are shown in genotype were termed normal metabolizers, and 17 patients Table 4. The incidence with or without nausea was used in with the *10/*10,*1/*5 or *5/*10 genotype were termed low the analysis as an independent variable, and sex, age, daily metabolizers. However, there were also no significant dose of paroxetine and the genotypes of HTR3B Tyr129Ser differences in the incidence of nausea between the two were added as potential confounders. This analysis also genotype groups. showed that there was a significant association between nausea and HTR3B Tyr129Ser genotype. (P ¼ 0.048; OR ¼ 3.95; 95% CI ¼ 1.009–15.455). Discussion

We screened for two polymorphisms in the HTR3A gene and Effect of CYP2D6 gene polymorphism one variant in the HTR3B gene as a pharmacodynamic Five CYP2D6 genotypes were identified: *1/*1 (n ¼ 51), *1/*5 factor, and CYP2D6 gene polymorphisms (*1,*5,*10 alleles) (n ¼ 1), *1/*10 (n ¼ 11), *5/*10 (n ¼ 3) and *10/*10 (n ¼ 12). as a pharmacokinetic factor. To our knowledge, the present The allele frequencies of the *5 and *10 alleles were 2.6 and study is the first demonstration that the HTR3B gene may 24.4%, respectively. Patients were divided into three geno- predict the incidence of paroxetine-induced nausea in type groups according to the number of mutated alleles: 51 Japanese psychiatric patients. patients with the *1/*1 genotype, 12 with the *1/*10 and *1/ Kaiser et al.18 reported that polymorphisms of the HTR3A *5 genotypes and 15 with the *5/*10 and *10/*10 genotypes. gene may not serve as pharmacogenetic predictors of No significant differences were demonstrated for sex, age, antiemetic treatment with HTR3 antagonists in cancer

The Pharmacogenomics Journal The relation between the HTR3 gene and nausea T Sugai et al 355

patients. We also found no relationship between HTR3A reported that the effect of CYP2D6 on metabolism was less gene polymorphism and paroxetine-induced nausea. Trem- prominent at steady-state than after a single dose of blay et al.12 reported that the Tyr129Ser polymorphism of paroxetine, since CYP2D6 enzymatic activity seems to be the HTR3B gene did not alter the incidence of nausea and easily saturated upon increasing paroxetine dose. This may vomiting. However, in the present study, there was a be one of the reasons that there was no effect on paroxetine- significant relationship between the Tyr129Ser polymorph- induced nausea. Furthermore, the groups with two mutated ism of the HTR3B gene and paroxetine-induced nausea. On alleles were expected to have the highest concentration the other hand, Tremblay et al.12 also reported that an among these three genotype groups, but the concentration insertion/deletion polymorphism in the promoter region of in groups with two mutated alleles was lower than in groups the HTR3B gene had a significant effect on the incidence of with one mutated allele, and was the same as in groups with nausea and vomiting induced by cancer chemotherapy, no mutated allele (Table 3). This result suggests that there although we did not examine this insertion/deletion are patients who had noncompliance with paroxetine in the polymorphism. This discrepancy may occur because of the group with *5/*10 and *10/*10 alleles, since they discon- difference in medication, that is, cancer chemotherapy tinued because of various adverse events other than nausea, versus paroxetine. We cannot account for this discrepancy as a result of a great increase in plasma paroxetine between the previous and the present study because the concentration. function of the HTR3B gene polymorphism still remains We demonstrated that a pharmacodynamic factor such as unclear. To date, there are no in vitro data about the the Tyr129Ser polymorphism of the HTR3B gene may be a functional effects of the Tyr129Ser polymorphism in HTR3B predictor of paroxetine-induced nausea in Japanese psychia- gene. With regard to the functional effects of the other tric patients. Taking account of these findings, in clinical polymorphism in HTR3B gene, Cazzola et al.19 report that a situations, it may be possible to tailor paroxetine pharma- 6-bp deletion in the 50UTR of L-ferritin mRNA is a cause of cotherapy based on genetic factors. However, since the hereditary hyperferritinanemia-cataract syndrome, and results of some previous studies are not consistent with our Frank et al.20 report that the deletion *À100_À102delAAG current results, further study is needed to clarify these polymorphism may change the structure of mRNA com- discrepancies. pared to the wild type. However, the Tyr129Ser polymorphism of HTR3B gene had an amino-acid substitution in the Acknowledgments coding region, and it was possible that the Tyr129Ser polymorphism of HTR3B gene affected the expression level This research was supported by Health and Labor Sciences Reseach of the B subunit either by itself, or because of linkage Grants (Research on Psychiatric and Neurological Diseases and disequilibrium with other yet unknown functional variants, Mental Health) (H17-kokoro-002), a Grant in Aid for Scientific and this polymorphism of HTR3B gene may affect the Research (KAKENHI) from the Japan Society for the Promotion of Research (JSPS) (#17591199) and the 38th Pharmacopsychiatry occurrence of nausea by itself, or indirectly. Meanwhile, Research Grant of Mitsubishi Pharma Research Foundation. Murphy et al. 10 reported that the T102C polymorphism of the HTR2A gene could predict the treatment discontinua- Duality of Interest tion caused by paroxetine-induced side effects in older patients with depression. In future studies, we should also None declared examine the relationship between HTR2A gene polymorphism and paroxetine-induced nausea. References In our study, the CYP2D6 gene polymorphism had no significant effect on the incidence of paroxetine-induced 1 Kasper S, Fuger J, Moller HJ. Comparative efficacy of antidepressants. nausea. Murphy et al.10 reported that the CYP2D6 genotype Drugs 1992; 43: 11–22; discussion 22–3. did not influence side effects of paroxetine. Gerstenberg et 2 Trindade E, Menon D, Topfer LA, Coloma C. Adverse effects associated al.14 also reported that the number of CYP2D6 mutated with selective serotonin reuptake inhibitors and tricyclic antidepressants: a meta-analysis. CMAJ 1998; 159: 1245–1252. alleles was not related to the development of fluvoxamine- 3 Bergeron R, Blier P. Cisapride for the treatment of nausea produced by induced nausea. Our present results support these previous selective serotonin reuptake inhibitors. Am J Psychiatry 1994; 151: findings. 1084–1086. Kasper et al.1 reported that there was a relationship 4 Barnes NM, Sharp T. A review of central 5-HT receptors and their function. Neuropharmacology 1999; 38: 1083–1152. between plasma concentration of fluvoxamine and inci- 5 Morales M, Wang SD. Differential composition of 5-hydroxytrypta- dence of nausea. Sawamura et al.21 showed that plasma mine3 receptors synthesized in the rat CNS and peripheral nervous paroxetine concentration in patients with CYP2D6*10 system. JNeurosci 2002; 22: 6732–6741. alleles was significantly higher than those without *10 6 Tyers MB, Freeman AJ. Mechanism of the anti-emetic activity of 5-HT3 receptor antagonists. Oncology 1992; 49: 263–268. alleles, and plasma paroxetine concentration in patients 7 Tyers MB, Bunce KT, Humphrey PP. Pharmacological and antiemetic with *5 alleles also showed a tendency to be higher than properties of ondansetron. Eur J Cancer Clin Oncol 1989; 25: S15–S19. those without *5 alleles. Therefore, we hypothesized that the 8 Stewart DJ. Cancer therapy, vomiting, and antiemetics. Can J Physiol CYP2D6 gene polymorphism had a significant influence on Pharmacol 1990; 68: 304–313. 9 Ogata E, Yamada Y, Sugiura M. Analysis of 5-HT3 receptor antagonist, paroxetine-induced nausea, but it did not affect paroxetine- ramosetron hydrochlorid, based on receptor occupancy considering its induced nausea in the present study. Sindrup et al.22 active metabolite. Yakugaku Zasshi 2001; 121: 793–798.

The Pharmacogenomics Journal The relation between the HTR3 gene and nausea T Sugai et al 356

10 Murphy Jr GM, Kremer C, Rodrigues HE, Schatzberg AF. Pharmacoge- diminished capacity for deberisoquine hydroxylation. Mol Pharmacol netics of antidepressant medication intolerance. Am J Psychiatry 2003; 1994; 46: 452–459. 160: 1830–1835. 17 Steen VM, Andreassen OA, Daly AK, Tefre T, Borresen A-L, Idle JR et al. 11 Gutierrez B, Arranz MJ, Huezo-Diaz P et al. Novel mutations in 5-HT3A Detection of the poor metabolizer-associated CYP2D6(D) gene and 5-HT3B receptor genes not associated with clozapine response. deletion allele by long-PCR technology. Pharmacogenetics 1995; 5: Schizophr Res 2002; 58: 93–97. 215–223. 12 Tremblay PB, Kaiser R, Sezer O, Rosler N, Schelenz C, Possinger K et al. 18 Kaiser R, Tremblay PB, Sezer O, Possinger K, Roots I, Brockmoller J. Variations in the 5-hydroxytryptamine type 3B receptor gene as Investigation of the association between 5-HT3A receptor gene predictors of the efficacy of antiemetic treatment in cancer patients. J polymorphisms and efficiency of antiemetic treatment with 5-HT3 Clin Oncol 2003; 21: 2147–2155. receptor antagonists. Pharmacogenetics 2004; 14: 271–278. 13 Kaye CM, Haddock RE, Langley PF, Mellows G, Tasker TC, 19 Cazzola M, Foglieni B, Bergamaschi G, Levi S, Lazzarino M, Arosio P. A Zussman BD et al. A review of the metabolism and pharmacokinetics novel deletion of the L-ferritin iron-responsive element responsible for of paroxetine in man. Acta Psychiatr Scand 1989; 350: 60–75. severe hereditary hyperferritinaemia-cataract syndrome. Br J Haematol 14 Gerstenberg G, Aoshima T, Fukasawa T, Yoshida K, Takahashi H, 2002; 116: 667–670. Higuchi H et al. Relationship between clinical effects of fluvoxamine and 20 Frank B, Niesler B, Nothen MM, Neidt H, Propping P, Bondy B et al. the steady-state plasma concentrations of fluvoxamine and its major Investigation of the human serotonin receptor gene HTR3B in bipolar metabolite fluvoxamino acid in Japanese depressed patients. Psycho- affective and schizophrenic patients. Am J Med Genet B Neuropsychiatr pharmacology (Berlin) 2003; 167: 443–448. Genet 2004; 131: 1–5. 15 Niesler B, Flohr T, Nothen MM, Fischer C, Rietschel M, Franzek E et al. 21 Sawamura K, Suzuki Y, Someya T. Effects of dosage and CYP2D6- Association between the 50 UTR variant C178T of the serotonin receptor gene mutated allele on plasma concentration of paroxetine. Eur J Clin HTR3A and bipolar affective disorder. Pharmacogenetics 2001; 11: 471–475. Pharmacol 2004; 60: 553–557. 16 Johansson I, Oscarson M, Yue Q, Bertilsson L, Sjo¨qvist F, Ingelman- 22 Sindrup SH, Brosen K, Gram LF, Hallas J, Skjelbo E, Allen A et al. The Sundberg M. Genetic analysis of the Chinese cytochrome P4502D6 relationship between paroxetine and the aparteine oxidation poly- locus: characterization of variant CYP2D6 genes present in subjects with morphism. Clin Pharmacol Ther 1992; 51: 278–287.

The Pharmacogenomics Journal