Variation in the gene encoding the Serotonin 2A receptor is associated with outcome of antidepressant treatment
McMahon FJ, Buervenich S, Charney D, Lipsky R, Rush AJ, Wilson AF, et al.;
Commented by , 28 Apr 2006
Background
Only 60 to 70% of patients with depression responds to an adequate trial with a single antidepressants and 40 to 50% obtain complete remission of symptoms. Although we have no direct evidence it is plausible that some of the variation in treatment outcome has a genetic basis. Numerous studies have investigated gene variation in relation to outcome of antidepressant treatment but it has so far not been possible to reproduce the findings.
Method
The analyses of the STAR*D study (ref. 1) included data from 1953 outpatients with nonpsychotic major depression with a Hamilton Depression Rating Scale score 17 items of > 14. Patients were recruited from 14 regional centers across the United States. All patients received citalopram with regular assessment of outcome (using the Quick Inventory of Depressive Symptomatology scale, QIDS) and side effects.
Response was defined as a 50% reduction in symptoms on the QIDS over a 6-week trial and remission as only few symptoms on the QIDS. Sixty-eight genes were chosen among a list of plausible candidate genes according to a scoring system of the prior evidence. For each candidate gene, genotype data spanning the coding region was downloaded from the International HapMap Project to select an optimal set of SNPs to genotype.
Analyses were conducted according to a split sample design. The discovery sample consisted of two-thirds of the sample (n= 1380) and the replication sample consisted of the remaining one-third.
Results
In the discovery sample, only 12 SNPs met the significance level (p < 0.01) for either response or remission. Of these 12 SNPs only one SNP met or exceeded the significance level (p < 0.05) in the replication sample. This SNP recides in the gene HTR2A which encodes the serotonin 2A receptor. Patients who had the A allele had an 18% reduction in absolute risk of nonresponse to citalopram, compared to patients with homozygous for other allele.
None of the other 767 SNPs met these strict criteria for association in the discovery sample and replication in the replication sample.
Professor Kessing's comments
This is the first demonstration of a significant and reproducible association between genetic variation and outcome of antidepressant treatment. The study is remarkably by its large sample size and by its spilt sample design using a discovery sample and a replication sample.
Further, the STAR*D study (ref. 1) employed a naturalistic ascertainment and treatment protocol. Thus, the results have general relevance for a clinical population in contrast to results from many randomised trials. However, the effect size was modest. Additional alleles predictive of treatment outcome will need to be discovered before a clinically more relevant effect sizes are obtainable.
It should be noted that no association between polymorphisms in the serotonin tranporter and outcome to citalopram was found.
References
1. Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA, et al. Sequenced treatment alternatives to relieve depression (STAR*D): rationale and design. Controlled Clinical Trials 2004; 25 (1); 119-42