Pharmacogenomics

The proteins that interact with drugs, that is, transporters, receptors and enzymes, are all encoded by genes and account for the majority of variability in both pharmacokinetics and pharmacodynamics. These genetic variations in individual patients are predictive of the clinical effect of some medications. Between two patients, the same dose of a drug may result in an adverse reaction in one and be ineffective in the other. Genetic variation in just four genes accounts for more than 95% of drug–gene interactions,6 and it has been estimated that almost the entire UK population has at least one genetic variant in their drug metabolising competency.1

The majority of medicines are metabolised by cytochrome P450 (CYP) enzymes, primarily, but not exclusively, in the liver.7 Two enzymes, CYP2D6 and CYP3A4, are responsible for the biotransformation of more than half of prescribed medicines,8 while each accounting for less than one-third of the CYPs in the liver (Figure 1). These CYPs are under genetic control (pharmacogenes) and generate four phenotypes of metabolic competency:

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The relative abundance of cytochrome P450 (CYP) metabolising enzymes in the liver.

• Ultra-rapid metabolisers have multiple copies of dominant allele coding an active enzyme.

• Competent metabolisers have two copies of the dominant allele coding an active enzyme.

• Intermediate metabolisers have one copy of the dominant allele coding an active enzyme and one variant allele coding an inactive or reduced function enzyme.

• Poor metabolisers have two copies of a variant allele coding an inactive or reduced function enzyme.

• See visual abstract (Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Visual abstract, infographic with representations of alleles encoding for cytochrome P450 metabolising enzymes that express one of four metabolic phenotypes, which cause differences in patient response to pharmaceutical medications. Created using https://mindthegraph.com. Images under Creative Commons licences CC BY-SA 4.0 DEED.

The incidence of the metabolic phenotypes can be correlated with ancestry. The ultra-rapid metabolising version of the CYP2D6 enzyme has a low (<2%) prevalence in people with northern and middle European ancestry, whereas it is higher (12%) in those with Mediterranean backgrounds, and highest in Saudi-Arabian (21%) and Ethiopian (29%) ancestry (Figure 3).9 This information can be used to support prescribing if a particular phenotype is known to be associated with adverse effects or to be ineffective in certain populations. Where this information is missing, it provides a focus for future pharmacogenomic research. Adverse reactions to prescribed medication account for 6.5% of UK hospital admissions with the resulting median hospitalisation of 8 days costing an estimated £530 million annually.1 Pharmacogene testing for specific CYP alleles is already available, meaning that informed prescribing is possible and could reduce the risk of patient harm and its associated costs.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Median incidence of CYP2D6 phenotypes with increased metabolic activity (ultra-rapid metabolisers), stratified by genetic ancestry region.9 Ethiopian, Northern European, Mediterranean European and Saudi Arabian phenotypic incidence demonstrates the difficulty in stratifying by such large groupings. The North America ethnic group is Afro-American black. North American white ethnic group is similar to the European white ethnic group. The Pacific Ocean result displayed is for the Oceania ethnic group. The World map was designed by Freepik.com. The Country maps highlighted were designed by freevectormaps.com (https://freevectormaps.com/saudi-arabia/SA-EPS-01-0001?ref=atr and https://freevectormaps.com/ethiopia/ET-EPS-01-0001?ref=atr). Images under Creative Commons licences CC BY-SA 4.0 DEED.

The Netherlands,2 Spain3 and the Nordic countries4 have initiated pharmacogene testing for the introduction of precision medicine for their populations. While these tests are not yet routinely available to the NHS, the UK Genomic Medicine Service is part of the ‘NHS Long Term Plan’10 and aims to be ‘the first national health care system to offer whole genome sequencing as part of routine care’.5 Working towards this aim, a Genome England project completed sequencing of 100 000 patient genomes in December 2018.11 The follow-on research programme, ‘Our Future Health’, aims to sequence the genomes from 5 million participants by the end of 2025.11 This offers routes to improve health outcomes, particularly for rare disease and cancers, but also to work towards personalised testing for pharmaceutical treatment throughout the NHS, as called for in a joint report published by the British Pharmaceutical Society and the Royal College of Physicians.1

Clinical applications

The treatment of moderate pain often uses opioid analgesics such as codeine, tramadol, hydrocodone or oxycodone. The individual response to these analgesics involves the metabolic conversion of their parent molecule into pharmacologically active metabolites by CYP2D6 (Figure 4). Poor metabolisers are at risk of having inadequate analgesia and ultra-rapid metabolisers may suffer side effects due to quick, extensive conversion of the medication to the active metabolites.12 This, in turn, may result in higher rates of addiction for ultra-rapid metabolisers and increased anxiety for poor metabolisers.13 Anxiety stemming from inadequate pain relief is a ‘yellow flag’ for clinicians as it may have an impact on the recovery from musculoskeletal injury14; a particularly important consideration in a military population. Knowing the metabolic phenotype of a patient allows a clinician to amend their prescribing at an individual level. One mild opioid can be substituted with a different ‘pain killer’ or opioid that does not need metabolic conversion to a pharmacologically active form to achieve the desired clinical effect.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Codeine metabolism, reproduced from Codeine and Morphine Pathway, Pharmacokinetics (pharmgkb.org). Data Usage Policy (pharmgkb.org). Ovals represent enzyme genes and transporter genes. Stars indicate significant enzymes. Arrows indicate metabolic pathways. T-bars indicate repression of activity.

Military applications

The US military are actively investigating precision medicine and have published early findings; however, their pharmacogenomic-tested cohort is reportedly fewer than 1000 service personnel.15 16 The Canadian Armed Forces are similarly exploring pharmacogenomic testing and report, ‘the tool has the expected potential to increase drug safety and improve clinical efficiency. A host of unintended positive outcomes emerged, including perceived increased patient engagement and a more robust team-based approach, bringing together prescribers, pharmacists, and patients’.17

The range of medications where pharmacogenomics has an impact on the clinical effect is broad. Of the medications contained in UK military primary care medical modules there are 13 that have pharmacogenetic warnings related to their use. The majority of these are not commonly prescribed in the military population, but codeine, ibuprofen, omeprazole and tramadol are all regularly used. Using codeine in the UK Armed Forces population as an example, the phenotype distribution suggests that between 7% and 10% of the predominantly white European personnel would not receive adequate analgesia, and 34 to 38%, who are intermediate metabolisers, would have an uncertain clinical effect from their prescription.

The medical modules contain a limited set of medications and military personnel may have been prescribed and be taking medications not contained within the modules. For example, statins, selective serotonin reuptake inhibitors, antihypertensives, proton-pump inhibitors, antiemetics and some antibiotics have pharmacogenomic associations.12 Prescribers are often taught that antibiotics are inhibitors of the CYP system, which causes drug interactions. Antibiotics can have genetic variations in their pharmacokinetics but they may also act competitively or non-competitively as substrates for enzymes that show pharmacogenetic variation. This may therefore result in an increase in the amount of opioid receptor agonist for ultra-rapid and normal metabolisers, but little or no effect for intermediate or poor metabolisers. From the military role 1 perspective this has little impact as the main antibiotic inhibitors are not usually available in this environment. Commonly used first-line antibiotics are less likely to inhibit the CYP enzymes. For any drugs commonly prescribed to military populations, knowing either an individual genetic profile, or even a population-based phenotype stratification, will only improve the clinical effectiveness of prescribing.

In the UK Armed Forces, there is good phenotypic information at a population level for those personnel with white European ancestry, whereas there is little or no information for minority ethnic groups who comprise almost 10% of the Armed Forces population.18 Specifically, personnel with Nepalese and Pacific Islander ancestry, who form a large proportion of the minority ethnic groups within the UK Armed Forces, have different CYP allelic variations to the white European population (Figure 5). The Genomics England Diverse Data Initiative11 seeks to enable equity in the provision of genomic medicine and, by proxy, improve the equity of precision prescribing in these groups. However, while these specific populations are present in the UK Armed Forces, they represent a very small minority within the UK population ethnic breakdown. There is therefore a risk that relying on Genomics England will not result in data that are directly relevant to the UK Armed Forces, especially those patients not matching the ethnicity profile of the UK as a whole.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Median incidence of CYP2D6 allelic variants causing loss of/reduced metabolism function (poor metabolisers), stratified by genetic ancestry region.9 The North America ethnic group is Afro-American black. North American white ethnic group is similar to the European white ethnic group. The Pacific Ocean results displayed are the Oceania ethnic group. The World map was designed by Freepik.com. Images under Creative Commons licences CC BY-SA 4.0 DEED.

Future direction

Precision medicine offers clinical, patient safety and economic advantage to a military population. Full genome sequencing of a military population raises ethical considerations, not least through the risks of identifying genetic predictors of disease that may limit employability and deployability that cannot be mitigated against or treated. It could be argued that such phenotypic selection already occurs with conditions such as colour-blindness. Specific pharmacogenomic testing offers the benefit of more effective prescribing based on individual genotype/phenotype. Analysis of an entire military force to allow personalised prescribing could be prohibitively expensive. The option of testing a military force and targeting populations where there are insufficient data available, such as the prevalent minority ethnic groups, should be explored. If the UK were to take on this task, it would be world leading, not only by being an early adopter of precision prescribing, but augmenting the international data on pharmacogenomics in otherwise under-represented populations, thereby reducing ancestry-based health inequalities.