Menu

Section menu

Predicting the effects of drugs

No drug is effective for everyone, and in a few patients, usually safe drugs can produce serious adverse effects. Pharmacogenetic research is starting to provide the basis for predicting individual response to an increasing range of drugs.

 

 

Codeine acts as an effective painkiller after conversion to morphine, which is in turn detoxified and excreted. If too much morphine is produced or if its excretion is impaired, normal doses of codeine may become toxic. Individual genetic differences, as well as concomitantly prescribed drugs, may affect these metabolic pathways with clinical implications.

 

 

A PDF document

When usual doses lead to adverse effects: polymorphisms of metabolic enzymes (The genomic basis of therapeutics. Part 2)

Our bodies produce metabolic enzymes that detoxify and process most drugs. These enzymes may be impaired by genetic mutations or interactions with other drugs or foods, and lead to serious clinical effects.

 

 

A few HIV-AIDS patients given Abacavir, an antiretroviral drug, develop a potentially lethal hypersensitivity reaction. HLA-B*5701 testing prior to prescribing abacavir now allows us to identify many of them.

 

 

TPMT measurement is recommended prior to prescribing azathioprine for a range of autoimmune diseases

Thiopurine s-methyltransferase (TPMT) test for azathiaprine

Some Chinese patients develop a sever skin condition when prescribed carbamazepine for epilepsy. The results of an HLA-test can predict which individuals are likely to develop the skin reaction.

Human leukocyte antigen (HLA) test for Carbamazepine


The metabolism and pharmacodynamics (biological effects) of warfarin are complex and involve many genes. There are clear associations between the dose-requirement and effects of warfarin and variants of the metabolic enzyme (CYP 2C19) and of the pharmacodynamics enzyme (VKORC1). Whilst some experts believe that genetic testing should be undertaken when prescribing warfarin, others argue that the clinical utility of the available mathematical models for predictive purposes are not yet robust enough for general use. They point out that other factors need to be taken into account, such as diet, age and broader genetic background related to population ancestry, as they have major impacts on dose-requirements and effect. It is argued that the case for routine pre-prescription testing of CYP2C19 or VKORC1 is not yet strong enough for routine application.