Pharmacogenomics for Clinical Utility

 

Professor Katherine Aitchison

BA(Hons), MA(Oxon), BM BCh, PhD, FRCPsych

Professor, Departments of Psychiatry and Medical Genetics, University of Alberta, and Lead Psychiatrist, Edmonton Early, Psychosis Intervention Clinic

 

 

Background

Although adverse drug reactions (ADRs) are known to be under-reported,1,2,3 they account for up to 30% of hospital admissions in Canada and the USA.4 In 2006-07 to 2010-11, 2.7% of seniors’ hospitalizations due to ADRs were caused by antipsychotics.5 The more medications a patient is on (particularly likely in the case of seniors), the higher the risk of ADRs.6 Furthermore, if ADRs occur, medications may be discontinued by the prescriber and/or the patient, or patients may be non-compliant, and if the ADRs are particularly bad, patients may lose trust in their prescriber. Of 4 billion outpatient prescriptions in the US, approximately 18% (72 million) could have the efficacy or tolerability of the medication improved if pharmacogenomic (PGx) information were made available to the patient and prescriber.7

 

Method

Review of publicly available literature and genotyping methodologies, plus analysis of our comparative data from various different technologies for drug metabolizing enzyme and transporter variants.

 

Results 

In the field of mental health, there are 20 drugs with strong evidence of gene-drug associations, of which 18 are antidepressants or antipsychotics.8 Sixteen of these have an at least actionable level of genetic association (meaning that the genetic variants are associated with changes in efficacy, dosage or toxicity; the medication may be contraindicated in a subset of patients; or genetic testing is required in order to prescribe the medication), mainly with CYP2D6, 14 out of 16. The most commonly used medications metabolized by CYP2D6 constitute 189 million prescriptions (75% of which are for mental health or heart disease9), and cost $12.8 billion USD annually in the US (5-10% of total cost of all outpatient prescriptions). Comparative data analyses reveal platform strengths and weaknesses.

 

Conclusion

These analyses point the way forward for methods for clinical utility. By using comprehensive and accurate genotyping together with clinical reporting, we will predict ADRs from psychoactive substances and thus enable improved prescribing practice.

 

Reference

1. Pirmohamed, M., James, S. et al. BMJ 329, 15-19 (2004)

2. Bartlett, C., Doyal, L. et al. Health Technol Assess 9, iii-iv, ix-x, 1-152 (2005)

3. Legge, S. E., Hamshere, M. et al. Schizophr Res 174, 113-119 (2016)

4. Howard, R. L., Avery, A. J. et al. Br J Clin Pharmacol 63, 136-147 (2007)

5. CIHI, (2017)

6. Chue, P. & Chue, J. in Side-effects of drugs Vol. 38 (ed S. D. Ray) 35-54 (Elsevier, 2016)

7. Relling, M. V. & Evans, W. E. Nature 526, 343-350 (2015)

8. CPIC, (2017)

9. Phillips, K. A. & Van Bebber, S. L. Nat Rev Drug Discov 4, 500-509 (2005)