Pharmacogenomics of treatment response in first episode schizophrenia

点击翻译按钮获取中文摘要

For the last decade, pharmacogenetics has held out the promise of improved clinical prediction in psychiatry, with individualized treatment serving as the ultimate goal. To date, however, that promise has remained unfulfilled due to a variety of technological and methodological limitations, such as: 1) limited and unsystematic genotyping, usually involving a very small number of single nucleotide polymorphisms (SNPs); 2) limited and arbitrarily-defined clinical response phenotypes; 3) examination of an acute, single time-point response only; 4) small sample sizes; and 5) heterogeneity of prior treatment history. The recent introduction of novel genetic techniques to rapidly and efficiently screen hundreds of thousands of SNPs may provide renewed impetus for pharmacogenetic/pharmacogenomic investigations in psychiatry; however, these technologies, combined with small sample sizes, raise the possibility of very large numbers of unreplicated results (false positives). The proposed study is designed to overcome these limitations by applying state-of-the-art whole-genome association (WGA) methods in a three-stage design to several large samples of patients with schizophrenia ascertained at the Zucker Hillside Hospital (ZHH) and its affiliates. The first two stages involve genomewide screening of >500,000 SNPs in 1600 patients with schizophrenia for four key phenotypes: positive symptom treatment response, negative symptoms, weight, and cognitive function. Half of this sample (400 Caucasian patients and 400 African-American patients) has already been collected at our site, and the Caucasian patients have already been genotyped with WGA (Affymetrix 500K microarray) methods. Our WGA work to date has demonstrated feasibility of high-quality data yields (>97% call rates and >99.5% reproducibility) and has resulted in a published finding of a novel schizophrenia susceptibility locus. Full WGA genotyping of all 1600 patients will be utilized to generate an empirically-based set of candidate genes for these four treatment response phenotypes. These candidate genes will then be comprehensively examined with highly dense mapping in 240 well-characterized patients in the first episode of schizophrenia studied prospectively for one year under controlled treatment (Stage 3). The data analytic strategy is designed to "drill down" from an initial broad screen in a more heterogeneous but very large sample, towards the dense genotyping of detailed outcome variables (both short- and longer-term) in a predominantly treatment-naTve population of patients in the first episode of schizophrenia.