One of the first disorders where genetic factors were found to affect drug metabolism was glucose-6-phosphate dehydrogenase (G6PD) deficiency. In the 1950s G6PD deficiency was identified in people with an illness related to consumption of fava beans (a variety of broad bean) and an X-linked inherited association was determined. 

G6PD is an enzyme involved in carbohydrate metabolism and protects red blood cells from potentially harmful oxidative toxins. Fava beans contain toxins readily broken down in people with the enzyme but can cause severe haemolysis in those with relative deficiency of G6PD.

Certain drugs also become ‘toxic’ if red blood cells are not protected from oxidation by G6PD: primaquine, nitrofurantoin, quinolones and some sulfonamides (such as co-trimoxazole) should be avoided in those with known G6PD deficiency as their use may precipitate a haemolytic crisis, requiring blood transfusion. 

Aspirin, quinine, chloroquine, hydroxy-chloroquine and sulfonylureas are less toxic but should still be used with caution. These drugs are flagged up in the BNF. In several countries G6PD deficiency is screened for by blood spot tests in newborn babies, but this is not done routinely in the UK. 

Some patients tolerate codeine very poorly and become nauseous on small doses, while others may fail to get pain relief even at high doses. Codeine is an opioid prodrug that is converted to its active form, morphine, by a liver enzyme that is part of the CYP2D6 pathway of cytochrome P450 enzymes. 

There are over 100 genetic variants of CYP2D6, of which several result in ultra-rapid, high activity, and some in reduced or no enzyme activity at all. 

Those with ultra-rapid activity (1-2 per cent of the population) tolerate the drug poorly as they get a morphine ‘overdose’. Non-opioid analgesia or, even paradoxically, very low dose morphine may be a better, more predictable option for codeine ultra-metabolisers.  

At the other extreme, in patients who are poor or intermediate metabolisers (as many as 5-10
per cent of the population), the analgesic effects of codeine may be inadequate due to lower levels of active morphine. In these patients dihydrocodeine may prove to be effective or low-dose morphine may be required. 

This is useful in our understanding of the variation in response to codeine but, at present, the UK has no consensus guidelines on how to screen patients for CYP2D6 variants. Pharmacists may wish to make suggestions to prescribers for alternative analgesics where a conversation with a patient indicates lack of efficacy or heightened side-effects.

A fairly recent development in the UK is the use of blood tests to determine the optimum dose of azathioprine – or whether to avoid it altogether. This also applies to the closely related drugs mercaptopurine and tioguanine. 

Azathioprine is often prescribed in ulcerative colitis, Crohn’s disease, severe eczema, rheumatoid arthritis and some other autoimmune conditions. It is a prodrug converted in the body to an active immunosuppressant. An enzyme, thiopurine methyltransferase (TPMT), acts in a separate pathway in the breakdown of azathioprine, so less is available to be converted to the active form. The enzyme activity in individuals varies considerably according to a range of inherited genetic factors. 

This variation is an example of genetic polymorphism and explains why even small doses are highly toxic in some people with low levels of TPMT (the active immunosuppressant accumulates), while others tolerate relatively high doses (azathioprine is readily broken down). 

The main toxic effects of excess azathioprine activity are severe anaemia and myelosuppression (white blood cell counts can be dangerously low). Polymorphism with TPMT that results in relatively decreased enzyme activity affects 5-10 per cent of the population, who thus require lower doses. In up to one in 300 patients TPMT is absent, so azathioprine should be avoided. 

Measuring TPMT activity using a blood test is now recommended to identify these patients prior to commencing therapy. This is in order to achieve optimal therapeutic levels while minimising the risk of toxicity. The BNF gives recommended starting doses of azathioprine based on TPMT test results.

In certain ethnic groups allopurinol is associated with a high prevalence of severe allergic reactions. These are described as severe cutaneous adverse reactions (SCAR), including drug reactions with eosinophilia and systemic symptoms (DRESS), toxic epidermal necrolysis (TEN), Stevens-Johnson syndrome (SJS) and allopurinol hypersensitivity syndrome (AHS). By measuring certain human leukocyte antigens (HLA), which vary in genetic expression from one person to another, the immune cell response can be predicted. 

In 2005, an association between HLA-B*5801 and an increased risk of allopurinol-related SCAR in Han Chinese was identified and subsequently this association was also observed in other ethnic populations. This genetic variant is most common in East Asian populations, including those of Han Chinese (13.3-20.4 per cent), Korean (12.2 per cent) and Thai (8.1 per cent) descent and is found much less frequently in those of Japanese (0.6 per cent) and European descent (1.5-5.2 per cent). 

As these genetic factors are not modifiable, rapid tests to identify the presence of HLA-B*5801 are becoming available, which enable screening before commencing therapy with allopurinol. This is increasingly recommended before starting allopurinol in people of East Asian descent although is not yet standard practice in the UK.

A similar situation is seen with carbamazepine (and closely related drugs oxcarbazepine and eslicarbazepine), where another HLA-B marker, HLA-B*1502, can indicate significantly increased risk of severe allergic reactions. These are potentially life-threatening, skin-related adverse drug reactions, including SJS and TEN. These reactions are estimated to occur in one to six per 10,000 new users in countries with mainly Caucasian populations, but the risk in some Asian countries is estimated to be about 10 times higher. 

Since 2008 the MHRA has advised screening for HLA-B*1502 in patients of Han Chinese, Hong Kong Chinese and Thai origin before starting carbamazepine treatment, and avoiding the drug if this is present. 

More recently another genetic marker, HLA-A*3101, has been identified in Japanese people and individuals of European descent for serious carbamazepine-induced cutaneous adverse drug reactions such as SJS, TEN and DRESS. However, as yet, the MHRA says there are insufficient data supporting a recommendation for HLA-A*3101 screening before starting carbamazepine or chemically-related drugs.

There has been increasing interest in using a genetic test kit to establish whether a newborn baby is vulnerable to deafness if treated with gentamicin. It is known that some babies (around 0.2 per cent) have a rare mitochondrial m.1555A>G gene variant. This variation allows gentamicin to bind more strongly to the hair cells in their ears, where it becomes toxic and can lead to permanent deafness. 

The deafness is a rare but well-known adverse effect of gentamicin use and the discovery of this gene goes some way to explain this. 

The test works by detecting the m.1555A>G variant from a swab of DNA taken from inside a newborn baby’s cheek, with results available in under an hour. If the m.1555A>G variant is found, the baby can be treated with alternative antibiotics. NICE has provisionally approved use of this test kit so that more data are collected to understand how effective it is.

Interest in the potential for pre-emptive screening has led to guidelines in some countries (such as the US and Netherlands) to support the prescribing of more than 50 medicines where PGx biomarkers are predicted to be of clinical relevance. 

However, considerable research is still required to understand the utility of using PGx tests and how pharmacists can best advise prescribers and patients in their choice of medicines based on these methods. 

The European Union-funded programme, Ubiquitous Pharmacogenomics, runs the Preventing Adverse Drug Reactions (PREPARE) study across seven countries. Some early promising results were published in February.  

This study used a 12-gene panel test covering 42 drugs in hospital and community settings, including pharmacies, to see if ADRs could be prevented pre-emptively prior to a drug first being prescribed. 

The study involved 6,944 adult patients and there was a reduced number of ADRs in those who received the testing versus controls (725 vs. 833). The main drugs associated with the adverse reactions were atorvastatin, clopidogrel, tacrolimus, simvastatin, various opioids, a number of cytotoxics and several antidepressants.  

The NHS in England is introducing a feasibility study for conducting PGx testing prior to commencing certain drugs and started a primary care pilot in several general practices for antidepressants, PPIs and statins in 2023, with plans for it to become a national programme if successful. Some of the genetic variations related to these drugs, including clopidogrel, and potential interventions, are illustrated in Table 1 (based on US FDA guidance). 

In the past few years there has been increasing interest in pre-emptive PGx profiling using commercial kits to determine the likely response to certain drugs, including predicting potential adverse drug reactions. 

A swab of cells from the mouth, inside the cheek, is the common tissue sample method. The intention is to help the patient, alongside prescribers, use the profile to help choose drugs most suited to the individual. 

A number of DNA tests sold in the UK now also give information on drug response and some pharmacy groups have said they intend to introduce a service to support their use.