Should pharmacogenetics be incorporated in major depression treatment? Economic evaluation in high- and middle-income European countries.

The serotonin transporter 5-HTTLPR polymorphism moderates response to SSRIs and side-effect burden. The aim of this study is to quantify the cost-utility of incorporating 5-HTTLPR genotyping in drug treatment of major depressive disorder (MDD). We previously reported a theoretical model to simulate antidepressant treatment with citalopram or bupropion for 12 weeks. The drugs were alternatively selected according to an 'as usual' algorithm or based on response and tolerability predicted by 5-HTTLPR profile. Here we apply this model to conduct a cost-utility analysis in three European regions with high GDP (Euro A), middle GDP (Euro B) and middle-high GDP (Euro C). In addition we test a verification scenario in which citalopram+bupropion augmentation is administered to individuals with the least favorable 5-HTTLPR genotype. Treatment outcomes are remission and Quality Adjusted-Life Weeks (QALW). Cost data (international $, year 2009) are retrieved from the World Health Organization (WHO) and national official sources. In base-case scenario incremental cost-effectiveness ratio (ICER) values are $1147 (Euro A), $1185 (Euro B) and $1178 (Euro C). From cost-effectiveness acceptability curve (CEAC), the probability of having an ICER value below WHO recommended cost-utility threshold (3 GDP per capita=$1926) is >90% in high-income countries (Euro A). In middle- income regions, these probabilities are <30% (Euro B) and <55% (Euro C) respectively. All estimates are robust against variations in treatment parameters, but if genetic test cost decreases to $100, pharmacogenetic approach becomes cost-effective in middle-income countries (Euro B). This simulation using data from 27 European states suggests that choosing antidepressant treatment from the results of 5-HTTLPR might be a cost-effective solution in high income countries. Its feasibility in middle income countries needs further research.