Pharmacogenomics refers to the study of genetic variation associated with drug response, toxicity and disposition. Cancer pharmacogenomics requires the study of both acquired (somatic) mutations in tumours, which occur during tumour development and evolution, and inherited (germline) variations. The study of germline variation is often termed pharmacogenetics.
The strategies used to identify pharmacogenomic biomarkers have evolved alongside supporting technological advances. Early work focused on family studies comparing monozygotic twins with dizygotic twins. These were followed by candidate gene and pathway studies1,2 that largely targeted pharmacokinetic or pharmacodynamic pathways related to the drugs under investigation. The major disadvantage of these approaches lies in the limited and prespecified gene coverage, thus restricting the potential for discovery outside these areas.
In the last two decades the availability and affordability of array and sequencing technology have led to more empirical approaches such as genome-wide association studies (GWAS).3 GWAS have allowed the discovery of common associations, with smaller effect sizes, that lie outside the known genes and pathways. The major disadvantages of this methodology, however, are the stringent statistical significance level of association and the independent validation that are required to ensure that false-positive associations are excluded. This can require large samples sizes which, with many pharmacogenomic phenotypes, can be challenging to achieve. More recently, the use of whole genome, exome or targeted sequencing has provided the ability to increase the depth at which we can examine these regions and has led to the discovery of rarer variants with larger effect sizes.4
Challenges in pharmacogenomics
The challenges facing pharmacogenomic studies include lack of homogeneity between comparable studies in terms of sample types used for analysis, the precise clinical phenotype, scoring systems, specific drugs, and patient populations under investigation. These issues have often been responsible for conflicting results in studies and delays in the implementation of potentially actionable variants/mutations.
Clinical implementation of pharmacogenomic biomarkers
Owing to the sheer volume of pharmacogenomic data being published, it is essential that there are systematic assessments of the quality of evidence available prior to implementation of an actionable variant/mutation. Various publications have provided different ways to prioritize the types of evidence available for both somatic5 and germline biomarkers. The Clinical Pharmacogenetics Implementation Consortium6 has developed guidelines, focusing on germline variants, which provide guidance regarding the clinical implementation of pharmacogenomic testing.
Clinical implementation requires that the biomarker meets the necessary standards of analytical validity, clinical validity and clinical utility before fulfilling all the regulatory requirements, ethical and educational issues involved in allowing its routine use. These are encompassed in the analytical validity, clinical validity, clinical utility and ethical, legal and social implications of genetic testing (ACCE) framework.7 Analytical validity refers to the accuracy of the clinical test in determining the genotype/mutation. Clinical validity refers to the accuracy of the test in determining ...