Schwerpunktthema

POST study on Pharmaco- genetics

Genetics is an important factor in the overall impact of a person's reaction to a specific drug. With increasing knowledge about the human genome, pharmacogenetic testing will become more important, as it opens up the prospect of personalised medicine, where each patient receives the treatment most appropriate to him. It poses however also challenges to regulators and raises questions about the nature of the information derived by phamacogenetic profiling.

What is pharmacogenetics?

Pharmacogenetics is an attempt to explain this variation in the way individuals respond to drugs in terms of genetics. This is not to suggest that genetics is the only factor involved. An individual's response to a drug will also be influenced by a range of other (environmental) factors such as what and when they last ate. Other factors such as compliance with a drug regime (the extent to which an individual takes the right drug at the right time) will also contribute to the overall impact of the drug. But it has become increasingly apparent in recent years that genetics is also an important factor - and the more that is known about the human genome, the more important pharmacogenetics may become.

Drug responses and genetic variation

While we all share the same genome, we each have our own unique version of it. For instance, coinciding with the February 2001 publication (Nature) of the first draft of the human genome, the International SNP Map Working Group issued a map documenting 1.42 million sites of known individual variations within the genome (SNPs are single nucleotide polymorphisms - sites at which single base differences occur).

Some of these genetic variations will occur in the genes that code for proteins that interact with drugs. Such variations may be important for two main reasons. First, proteins may themselves be drug targets. Because proteins perform many essential tasks within the body (messengers, carriers, receptors, etc.), drugs are often designed to bind a specific protein in order to exert their therapeutic effect. Variations in these target proteins may thus affect the way an individual responds to a drug. Researchers may increasingly be able to apply knowledge of variations in target proteins to drug design.

Second, proteins are also responsible for breaking down drugs in the body (Wolf et al. 2000).In particular, one family of proteins - the so-called cytochrome P450s - does this for most of the drugs used in modern medicine today. As discussed in more detail below, variations in these proteins can affect the rate at which a drug is broken down and the nature of the breakdown products. Such variations have been the focus of much recent attention because they can affect:

• Drug efficacy. For instance, some individuals have multiple copies of a gene (cyp2d6) that codes for a protein that (inter alia) breaks down tricyclic antidepressants such as nortriptyline. Such individuals break the drug down at such a high rate that it is virtually impossible for the drug to achieve a therapeutic dose.

• Dose required. The rate at which a drug is broken down affects the dose required to achieve a therapeutic effect. For instance, some individuals have inactivating mutations in a gene (cyp2c9) coding for another member of the cytochrome P450 family that is normally involved in breaking down drugs such as warfarin (an anticoagulant). Such individuals thus need only very low doses of the drug.

• Adverse reactions and toxicity. Variations in genes coding for the cytochrome P450 family of proteins also affect the metabolic route by which drugs are broken down. Some individuals may thus break down drugs into harmful metabolites that can cause adverse reactions, whereas others will not.

• Drug activation. Some drugs are given as inactive pro-drugs, relying on enzyme action to convert them to the active form. One such example is codeine, which is converted to its active form (morphine) by cyp2d6; people with inactivating mutations in this gene derive no therapeutic benefit from the drug.

Pharmacogenetic testing to predict drug responses

Implications for policy makers

Ethical issues concerning genetic tests have been the subject of much debate in recent years. To date this has been concerned with genetic tests for genes involved with disease. These have largely been tests for relatively rare single gene disorders such as cystic fibrosis, but may increasingly include tests for genes implicated in common complex disorders such as cancer, diabetes, etc. (see Fig. 1). Issues raised by "disease genetic testing" and considered by regulatory bodies throughout Europe include:

• informed consent (e.g. the need to provide counselling before and after a test);
• whether or not tests should be conducted if no appropriate treatment is available;
• test validation (particularly for predictive tests for common complex disorders);
• whether third parties such as employers or insurers should have access to genetic information;
• and, if so, under what circumstances.

To what extent do similar considerations apply to pharmacogenetic tests? Some have argued that all genetic tests raise similar issues by virtue of the fact that they reveal genetic information about the person taking the test. Others, including many in the pharmaceutical industry, feel that it is the nature of the information that is important, not the nature of the test by which it was derived (Roses 2000).They see a clear distinction between information relating to genetic disease and information concerning likely response to a drug (see Fig. 1). The former may be used to the patient's medical advantage but may also be of interest to third parties such as insurers and employers, who might seek to use it in a way that discriminates against the patient. They argue that pharmacogenetic data can only be used to the patient's medical benefit; it is of no interest to third parties and is thus unlikely to be used in a discriminatory way. This may mean that some of the ethical considerations that apply to disease genetics may be avoided by pharmacogenetics.

Not all agree with this view. Some have suggested that a person's pharmacogenetic profile may itself be of interest to third parties, for example, if it showed that the individual was particularly difficult to "match up" to currently available treatments. Others have questioned the distinction between disease genetics and pharmacogenetics, pointing out that it depends on the assumption that the latter will yield information only on likely responses to drugs. Suppose it subsequently turned out that the same variations that affect drug metabolism (and thus predict drug responses) could also be used to deduce something about an individual's future susceptibility to genetic disease. Under these circumstances, the information might be of interest to third parties and consideration may have to be given to issues such as counselling, confidentiality, etc. It must be stressed that there is no evidence to suggest that this will be the case; it is more a matter of speculating that it is not entirely implausible that genetic variations in metabolic enzymes might have more far-reaching biological consequences. One issue for policy makers here is to consider mechanisms by which the "usefulness" of the information yielded by pharmacogenetic tests can be monitored.

Clinical trials and adverse drug reactions

The key feature of pharmacogenetics is that it allows drugs to be targeted at sub-groups of the population whose genetic profile predicts that they will respond to them best. Indeed many pharmaceutical companies have already started collecting genetic information in the early stages of clinical trials. By the time a new drug is ready to be tested in thousands of patients in Phase III trials, researchers may already have a pretty good idea of which patients are likely to benefit most from it. If the drug can be tested on patients with the appropriate genetic profile in the final phase clinical trial, then fewer patients may be needed to establish efficacy. This is good news for pharmaceutical companies since it raises the prospect of smaller, faster and cheaper clinical trials.

One concern about clinical trials in general is that they may miss very rare adverse drug reactions that become apparent only after the drug has been given to tens of thousands of patients. Current trials may already involve too few patients to detect some such adverse reactions; any reduction in the number of patients involved in clinical trials (e.g. because of pharmacogenetic profiling) might therefore further reduce the likelihood of detection. However, pharmacogenetics also offers the potential for improved surveillance for adverse drug reactions. Keeping blood samples from all patients receiving a new drug in the first few years it is on the market would allow pharmacogenetic profiling of any adverse reactions detected in this larger group of people. Information linking specific genetic variations with very rare adverse reactions to a drug could thus be "fed back" into the marketing approval. In this way, the pharmacogenetic profile of the sub-group of patients for which the drug was approved could be continually "fine tuned".

Overview

Healthcare providers such as the UK's National Health Service (NHS) will thus have to consider the potential financial and organisational issues raised by such trends. For instance, NHS genetic testing in the UK has evolved to meet the needs of patients suffering from relatively rare genetic disorders: it is currently centred on 50 or so regional laboratories. Basing drug prescriptions on pharmacogenetic profiles or involving other types of genetic testing in diagnosis/classification of more common disorders implies a more accessible, local, organisation of services.

It also implies that doctors, nurses, and medical technicians will require more knowledge of recent developments in clinical genetics than is currently the case. This can be addressed through changing curricula and by continued professional training, although such initiatives require time and money to set up and to take effect. Greater numbers of genetic counsellors may also be needed for some types of genetic tests (although as discussed above not necessarily for pharmacogenetic profiling), and consideration may need to be given to novel ways of delivering counselling.

Overall then, pharmacogenetics creates both opportunities and challenges for policy makers. It opens up the prospect of personalised medicine, with each patient receiving the treatment most appropriate to them. It also promises to speed up the clinical trials process and reduce the numbers of patients suffering adverse reactions to new drugs. However, this poses challenges for regulators and may also raise questions about the nature of the information derived by pharmacogenetic profiling.

References

Wolf C. R. et al., BMJ 320, 987-90, 2000

Roses, A. D., Lancet, 35, 1358-61, 2000

Contact: