Clinical Translation of Cell-Based Pharmacogenomic Discovery

Autor: Nancy J. Cox, M E Dolan, Eric R. Gamazon, Heather E. Wheeler
Rok vydání: 2012
Předmět:
Zdroj: Clinical Pharmacology & Therapeutics. 92:425-427
ISSN: 1532-6535
0009-9236
Popis: The use of cell-based models has emerged as a promising means to discover and validate pharmacologic phenotype-genotype relationships. The availability of large-scale genome studies in both human and model systems is now allowing us unprecedented opportunity to understand how well cell-based models identify clinically relevant genetic variants associated with drug response and toxicity. Here, we review these studies and the emerging translational information. The use of human Epstein-Barr virus transformed lymphoblastoid cell lines (LCLs) has emerged as a promising model system in the study of the genetics of drug response because the cells provide a cost-effective testing system in which environmental factors such as concomitant medications can be controlled. The advantages of HapMap LCLs include a controlled, unlimited resource to evaluate toxic drugs without the in vivo confounders such as concomitant medications. Most importantly, publicly available genotype and sequencing data is available, to allow for genome-wide association studies between HapMap/1000 Genomes variants and pharmacologic phenotypes measured in the LCLs after drug or radiation treatment such as cell growth inhibition, changes in gene expression, intracellular concentration of drug/metabolite and/or apoptosis. Limitations of the model are that they represent one cell type (B-lymphoblasts), they are subject to in vitro confounders and that most CYP450 genes are not expressed and therefore the model is most useful for identifying genes and genetic variants within drug transporter and pharmacodynamic genes as opposed to pharmacokinetic genes. A recent review has described studies from a number of groups using preclinical cell-based models have taken top genome-wide association study (GWAS) hits from these models and validated them in prospective clinical trials (1). For example, top SNPs from a GWAS of cisplatin-induced cytotoxicity in LCLs were found to be associated with overall survival in patients treated with platinum for non-small and small cell lung cancer and a top SNP from a GWAS of carboplatin-induced cytotoxicity in LCLs was shown to be associated with outcome in ovarian cancer patients treated with carboplatin-based induction chemotherapy. In addition, variation in the expression of FKBP5 was found to associate with cytarabine-induced cyototoxicity in a genome-wide screen in LCLs and follow-up clinical studies showed SNPs within FKBP5 associated with both event-free and overall survival in pediatric AML patients treated with cytarabine. Moreover, the cell-based model has been utilized to generate information on associations between germline variation and a wide variety of functional phenotypes, including gene expression and miRNA levels. Comprehensive eQTL maps have been generated in LCLs to enable the identification of cis- and trans- effects of (pharmacologic) trait-associated SNPs on gene expression (2). Such maps may facilitate the functional interpretation of findings from GWAS studies of drug response phenotypes in both cell-based and clinical studies. These studies led to the finding that SNPs associated with chemotherapeutic-induced cytotoxicity in LCLs are enriched in expression quantitative trait loci (eQTLs) (3). Since most pharmacogenetic studies previous to GWAS were focused on variation in coding regions of known candidate genes, this was an important finding since it opened up the possibility that SNPs in introns or intergenic regions associated with gene expression contributed significantly to variation in pharmacologic phenotypes. Furthermore, connections between pharmacologically important variants and eQTLs may lay the basis for understanding the mechanisms underlying the observed associations with potentially important implications for individualized therapy. One consideration in the use of eQTLs identified in LCLs for pharmacogenetic studies is that a proportion of eQTLs may be tissue-specific and therefore LCLs may provide a limited subset of functional SNPs of relevance to toxicity (eg. bladder or renal toxicity, peripheral neuropathy). However, a recent study (4) estimated that 65–70% of the cis eQTLs identified in LCLs are also present in fibroblasts and T cells after adjusting for statistical power. This report is consistent with the status of current eQTL mapping studies as typically underpowered, which results in an underestimation of shared eQTLs across tissues. Although it is indeed gratifying to observe that the very top signals from cell-based models can be validated in prospective clinical trials, it is critical to understand how well the overall genetic architecture of the response to chemotherapy can be captured by the cell-based models. Thus, it is important to compare the results of genetic studies in the cell-based models to studies conducted in human clinical trials beyond just the top few signals. A first approach to such a comparison might be to ask whether SNPs with nominally significant associations to cytotoxicity in studies in cell-based models are enriched among the top SNPs (e.g. those at least nominally significant) showing association to outcomes or adverse events in human clinical trials. The genome-wide criteria for significance in GWAS is widely accepted to be on the order of 5 × 10−8 and there are generally a large number of individual SNPs that are nominally associated (p
Databáze: OpenAIRE
Externí odkaz: