Pharmacogenomics in psychiatry – Clinical innovation utilised by the Therapeutic Goods Administration and Food and Drug Administration

The recent debate article on pharmacogenomics by Foulds et al. (2016) presents one side of the argument and ignores other evidence. A single randomised controlled trial (Peters et al.’s (2008) analysis of STAR*D study) and one observational study is presented to refute the therapeutic application of testing CYP450s in depression. Their article ignored other important trials and observational studies in this area (Hall-Flavin et al., 2013) as well as the reanalysis of the STAR*D study by Mrazek et al. (2011).

Mrazek et al. reanalysed the STAR*D study and drew the opposite conclusion to Peters et al. Mrazek summarised some of the differences, including different outcome definitions in the remission analyses and different groupings of genotypes for analyses. Each group did their own genotype testing, but Peters et al. examined only single single nucleotide polymorphisms (SNPs) for relationship with clinical outcomes, while Mrazek examined genotype – as is appropriate for clinical practice.

In fact, Peters et al. did find a relationship with pharmacogenomics in their analysis. They did this by splitting the data into half – a discovery data set and a validation data set. The discovery data set found a relationship with CYP2C19 changes and intolerance to citalopram, but this was not confirmed in the validation data set. However, Mrazek, studying the same population, confirmed that citalopram intolerance was related to CYP2C19 changes and thus replicated Peters’ initial discovery data set finding.

This lack of tolerance with reduced function of CYP2C19 is recognised by the Therapeutic Goods Administration (TGA) and Food and Drug Administration (FDA), who both have specific warnings suggesting reducing the dose of citalopram by 50% in poor metabolisers of CYP2C19. Poor metabolisers of CYP2C19 have an increased risk of prolonged QT interval. The Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for selective serotonin re-uptake inhibitors (SSRIs) summarise seven studies in the literature that show this (Hicks et al., 2015).

Ethnicity is a significant factor in the pharmacogenomic testing process. This is especially relevant for CYP2C19-reduced function alleles in Australia and New Zealand because of the growing proportion of Asian people in the population. Poor and intermediate metabolisers are much more frequent in Southeast (SE) Asian and Japanese ethnic groups. Importantly, the recent Royal Australian and New Zealand College of Psychiatrists’ (RANZCP) Clinical Practice Guidelines (CPG) on mood disorders recommend human leukocyte antigen (HLA) testing in patients of Asian ancestry prior to trialling carbamazepine to gauge the risk of Stevens–Johnson Syndrome (Malhi et al., 2015). HLA testing for carbamazepine adverse effects was not developed from a randomised controlled trial but a series of observational studies showing a strong relationship between a rare HLA-B allele in SE Asians.

The CPIC guidelines and online supplement summarise the evidence for the use of SSRIs in psychiatry as the CPIC guidelines for tricyclic antidepressants do for tricyclic antidepressants (Hicks et al., 2015). CPIC guidelines are not based on prospective comparator trials, but as a tool the validity of pharmacogenomics testing is increased by gathering more data in specific conditions. Personalised prescribing in psychiatry based on measurable drug metabolism parameters is a step forward that our patients desperately need.