| It is widely known that people respond differently to medicines. Why
does one patient suffer an adverse reaction and another not? And why
does a particular drug work well for one patient yet have little or no
effect on someone else? Take warfarin, for example. The dose of warfarin
required varies from patient to patient, and a therapeutic dose for one
person can cause severe bleeding in another. Although it will not provide
all the answers, pharmacogenetics should increase understanding of how
people respond to medicines and then help increase efficacy and improve
safety. In warfarin’s case, response is known to be related to
variations in metabolism via the enzyme cytochrome P450 (CYP) 2C9. These
variations are determined by genetics.
Few examples of the use of pharmacogenetics in clinical practice exist.
Even if the technology becomes available, the use of pharmacogenetics
raises a host of questions. This week, the Nuffield Council on Bioethics
attempts to answer a number of these questions through the publication
of its report “Pharmacogenetics: ethical issues”. It hopes
that its recommendations will ensure that future delivery of pharmacogenetics
tests is “as straightforward as possible”. The report is
available here.
Professor Peter Lipton, chairman of the working party that produced the
report, commented: “It is too early to predict whether ‘personalised
medicines’ will become a reality. Claims of ‘the right medicine
for the right patient at the right dose’ may be overstated. But
it is important to encourage discussion of ethical and policy issues
raised by the introduction of pharmacogenetics.”
Professor Tony Moffat, chief scientist at the Royal Pharmaceutical Society,
said: “The Society’s working party on pharmacogenetics is
examining the role of the pharmacist in pharmacogenetics and what the
Society can do to assist pharmacists in developing the new knowledge
that they will need for the wider roles in couselling and information
provision that they will perform.”
Next week, a national strategy for genetics education for health professionals
is expected to be recommended by a new report. This ties in with concerns
raised in the Nuffield Council report that initatives to provide independent
and impartial information about pharmacogenetic tests are needed. Next
week’s report — called “Addressing genetics, delivering
health” — is written by the Public
Health Genetics Unit in
Cambridge and was commissioned by the Department of Health and the Wellcome
Trust. It says that all health care professionals need to be better informed
about how genetics affects a patient’s likelihood of developing
a disease or response to medicines. It recommends that a genetics education
centre should be set up to support this.
How pharmacogenetics can help
“Some common treatments for conditions including diabetes, depression
and asthma are only effective in around 60 per cent of patients, and
for medicines used to treat cancer this figure may be as low as 25 per
cent,” the Nuffield Council says.
Pharmacogenetics could help improve understanding of these responses.
It can be used to identify an individual’s genetic variations that
are associated with adverse reactions or with differing metabolism so
that medicines could be tailored accordingly. It can also provide information
about the genetic characteristics of a disease, which could be used to
improve drug design.
Whether or not pharmacogenetics will be used to improve existing medicines
is unclear. There is a long list of currently prescribed medicines where
genetic variations are known to play a part in the drug’s metabolism
(see Panel below). How significant some of these variations are is
not known, but some do have an impact on response to therapy yet routine
testing for genetic variations does not occur.
Variations with existing drugs
How people metabolise drugs is determined by genes. The causes
of these differences can be grouped into three categories: enzyme
variations, transportation differences and receptor variations.
What is not known is how significant these variations in metabolism
are.
Professor Tony Moffat, chief scientist at the Royal Pharmaceutical
Society, explains that if one route of metabolism is affected then
the body can compensate by changing the rate of metabolism by another
route. So although a list of drugs with known genetic variations
in metabolism can be produced, it does not mean these differences
will have any significance. “The clinical outcome in most
cases is unknown,” says Professor Moffat.
There are only four drugs where evidence has demonstrated that
genetic variations do have a clinical effect on metabolism: warfarin,
omeprazole, triptans and metoprolol.
• Enzyme variations Differences in enzyme status can result in
a lack of efficacy or toxicity at standard doses as a result of
altered absorption, distribution or metabolism. The most important
affected enzymes are the cytochrome P450 (CYP) group since many
drugs are metabolised by them. For example, CYP2D6 metabolises
30 widely used drugs, including b-blockers, antidepressants, antiarrhythmics,
loratadine and codeine. Warfarin is metabolised by CYP2C9.
• Transportation variations Differences in
transport of a drug from the gastrointestinal tract into the
systemic circulation or
across the blood-brain barrier can alter a drug’s efficacy
or toxicity. An example is anti-infective agents.
• Receptor variations Variations in a drug’s target receptor — such
as its abundance or function — can affect the drug’s
efficacy and toxicity. Drugs affected include astemizole, b-agonists,
clozapine, cyproterone, glucocorticoids, halothane, insulin, oestrogen,
sulphonylureas, terfenadine and warfarin. |
“We recommend that efforts should be made to encourage pharmacogenetic
research on existing medicines, where there is reason to believe that
such research could significantly improve efficacy or safety,” the
Nuffield Council report states. It acknowledges that in some cases
it might be quicker and easier to continue today’s practice of
giving a medicine and observing its effects instead of developing pharmacogenetic
tests.
However, a first step towards routine use of pharmacogenetics was made
this summer. A diagnostic test for cytochrome P450 was launched in
the United States in June. The AmpliChip CYP450 microarray is manufactured
by Roche Diagnostics. It enables genetic variations in two CYP enzymes
to be identified. Speaking at the time of the launch, a company representative
commented: “The launch of this product heralds the emergence of
pharmacogenetics as a medical and commercial reality.” Initially
it will only be available in specialist clinics.
The ability to determine more accurately which patients will benefit
from treatment could also have an impact on drugs that have been withdrawn.
Withdrawals tend to be on safety grounds and if a pharmacogenetic test
could identify patients at increased risk of an adverse reaction then
the drug could be avoided in these patients.
The report notes that an example of this is alosetron (Lotronex), a
treatment for irritable bowel syndrome approved in the United States
three years
ago but subsequently withdrawn because some patients experienced adverse
reactions. The Food and Drug Administration and the drug’s manufacturer
are currently re-examining the medicine, including using a pharmacogenetic
analysis.
What about new medicines? One drug that came to the market recently — and
represents one of the few current uses of pharmacogenetics in practice — is
trastuzumab (Herceptin). Used in the treatment of breast cancer, it
can only be given to patients who over-express the human epidermal
growth
factor receptor 2 (HER2) gene.
The Nuffield Council report says that it might be discovered that some
conditions currently thought of as single disorders might have different
causes. “It may turn out that the nature and efficacy of treatment
depends on which type of the disease is present. Such heterogeneity may
be behind some of the well-known variation in efficacy of medicines given
to people affected by what appears superficially to be the same disorder,” it
states.
Better targeting of drugs could make the market smaller and the report
warns that some medicines may not be developed if the number of patients
who would benefit is too small to be profitable. On the other hand, it
could also encourage drug development by preventing early halts to clinical
trials if a subgroup of patients with a particular genetic profile for
whom the drug appears to be effective can be distinguished.
Use in practice
Use of pharmacogenetics in practice will pose challenges for many aspects
of health care. Where tests will take place will have to be decided.
The Nuffield Council predicts that this will depend on the test itself
with some carried out in a pharmacy or GP surgery and others requiring
more specialised testing facilities. It warns that if testing takes
up too much time then it might impede the delivery of health care.
Some tests might become available to buy over the counter but only if
they provide clear, readily interpretable information. The Nuffield Council
report comments: “The majority of pharmacogenetic tests will be
more complex, providing less certain predictions. In these cases, professional
advice is likely to be needed both before and after taking the test which
means that the direct commercial provision of tests will be inappropriate.”
Pharmacogenetic testing will introduce a new raft of decisions about
a patient’s treatment: decisions around whether to take the test,
whether the medicine should be available after the test and what to do
if the patient refuses to take the test. If a patient refuses a test
that is part of a licence condition then it is unlikely that the patient
will be prescribed the medicine. In other cases, it might only form part
of the decision-making process. But the Nuffield Council warns: “We
note that advances in pharmacogenetics can be expected to lead to the
licensing of medicines that would not have been licensed had there been
no associated test, because of the serious danger those medicines pose
to a sub-population. To allow prescription without the test in such a
case would be wrong.”
Consent will become an increasingly important issue with the possibility
that some pharmacogenetic tests will require patient consent. However,
information generated by a pharmacogenetic test is likely to be less
predictive of health outcomes than genetic tests so consent might not
be such a thorny issue as has been previously imagined. The report recommends: “If
information about unrelated medicines or diseases is likely to be obtained,
or if the results of the test will have significant impact on the health
or lifestyle of the patient, written consent may be appropriate. We take
the view that, in most cases, written forms will not be required.”
Patients will certainly need access to information about pharmacogenetic
and health professionals will have to be able to provide it. |
