The importance of genomics to psychopharmacology
Virtually every physical and psychiatric disorder has a genetic component.
However, the vast majority of these diseases have a complex pattern of
inheritance and there is no evidence that a single genetic locus is
responsible for any of the major psychiatric disorders. Rather it appears
that multiple alleles (gene products) occurring at multiple sites within the
genome interact to produce a vulnerability to the disorder. The enthusiastic
reception for the unravelling of the human genome rests largely on the
promise that it will soon lead to an understanding of the pathological basis
of most diseases which, in turn, will aid the development of more effective
therapeutic treatments.
Following the sequencing of the human genome it was found that there
were between 30 000 and 40 000 genes that code for proteins, only twice as
many as occur in the fruit fly or the nematode worm! However, it does
appear that human genes are more complex than those of flies and worms
in that they generate a large number of proteins due to the alternative waysof splicing the molecules. Hopefully knowledge of the human genome will
enable genes to be identified that convey a risk for psychiatric diseases in
addition to those genes which are linked to a therapeutic response to drug
treatment. Knowledge of the latter forms the basis of pharmacogenomics
which, hopefully, will eventually lead to the development of specific
treatments for the individual patient.
The potential value of pharmacogenomics can be illustrated by two
examples involving the response of individual patients to antidepressants.
In this approach, the potential importance of the cDNA microarray
technique for identifying changes in thousands of individual genes that
are expressed in the mouse brain is now widely accepted. Experimental
studies have indicated that different antidepressants exert distinct effects on
gene expression in the mouse brain, these differences becoming more
marked as the duration of the treatment increased. Such findings may
eventually lead to an individualized treatment strategy for depressed
patients based upon their cDNA analysis.
At the practical clinical level, individual differences in the pharmacokinetic
characteristics of antidepressant drugs have been more successful. It
is well established that the enzymatic activity of different allelic forms of the
cytochrome P450 oxidase system in the liver is particularly important in
the metabolism of many psychotropic and non-psychotropic drugs (see
pp. 91–94). Of the major forms of cytochrome P450 in man, the 2D6 isozyme
is particularly important in the metabolism of antidepressants and a
potential cause of drug interactions. Three of the five commonly available
SSRI antidepressants (fluoxetine, paroxetine and sertraline) undergo
autoinhibition of this isozyme and can therefore increase the tissue
concentration of a more toxic drug (for example, an antiarrhythmic or
beta-blocker) should it be given concurrently.
Over 50 allelic variants of the cytochrome P450 2D6 gene have been
identified, including individuals who lack the gene and others who have
multiple copies of the gene. This means that an individual (the functional
genotype) can either be normal, a slow or an ultra-fast metabolizer of a drug
that passes through the 2D6 pathway in the liver. Slow metabolizers will
therefore be at an increased risk for adverse effects while the rapid
metabolizers will have little benefit from the normal doses. Thus
genotyping the enzymes that metabolize the commonly used psychotropic
drugs could help to optimize the response, and to indicate the potential for
adverse drug effects, of the individual patient.
A new term has recently been introduced to cover the application of
pharmacogenomics to the design of drugs for the individual patient,
namely theranostics (from therapeutics+diagnostics). This approach
involves creating tests that can identify which patients are most suited to
a particular therapy and also to provide information on how effective thisdrug is in optimizing the treatment. Theranostics is said to adopt a broad
dynamic and integrated approach to therapeutics which may be of
practical relevance in differentiating diseases which are closely associated
diagnostically (for example, Alzheimer’s disease and Lewy body dementia)
by applying a combination of immunoassays that enhance the differential
diagnosis. Several biotechnological companies now specialize in designing
immunoassays for application to infectious diseases such as hepatitis by
genotyping the hepatitis C virus for example. There are six genotypes of
the virus known: genotype 1 is more resistant to standard therapy
(requiring at least one year of continuous therapy) whereas the other
genotypes usually respond to treatment within 6 months. Clearly a
knowledge of which viral genotype is present is important in determining
the duration of treatment in the individual patient and hopefully it will
soon be possible to extend such approaches to the drug treatment of
central nervous system disorders.
Applying pharmacogenomics to the pharmacodynamic aspects of
psychopharmacology is still at a very early stage of development, largely
because so little is known of the psychopathological basis of the major
psychiatric disorders or of the mechanisms whereby psychotropic drugs
work. In depression, for example, it is widely assumed that the inhibition of
the serotonin transporter on the neuronal membrane is ultimately
responsible for the enhanced serotonin function caused by the SSRI
antidepressants. The serotonin transporter is structurally complex. The
promoter region of the transporter, to which serotonin is linked before it is
transported back into the neuron following its release into the synaptic cleft,
exists in several polymorphic forms which are broadly categorized into the
long and short forms. It is known that when the polymorphic form occurs in
which an additional 44 pairs of nucleotide bases are inserted, there is a
higher transcription rate and a greater degree of binding of serotonin to the
promoter region. The practical importance of this finding is that depressed
patients with the long form of the transporter show a better response to
SSRIs than those with the short form. In bipolar patients, there is an
indication that the short form of the promoter is more likely to result in the
precipitation of a manic episode if given an SSRI during the depressive
phase of the disorder. There is also some evidence that the short and long
forms of the transporter may be correlated with the frequency of
extrapyramidal side effects and akathisia, which is sometimes caused by
SSRIs.
There are two caveats that should be taken into account with regard to
the application of pharmacogenomics. Drug response is as complex as the
underlying genetic basis of the disease due not only to the genotypic
variation taking place at mostly unknown chromosomal loci, but also from
variations in gene expression, post-translational modification of proteins,pharmacokinetic features of the drugs, the effect of diet, drug interactions,
etc. One would therefore anticipate that the effects of individual genes on
the drug response are relatively slight. Thus it has been shown in studies of
pharmacogenetic markers that they only confer a twofold increased
likelihood of predicting drug response. However, the widespread
application of microarray technology, whereby information on thousands
of genes can be determined simultaneously, may help to overcome the
limitations of the candidate gene approach, the method which until now
has been used to obtain information on a few genes presumed to be
involved in the underlying pathology of a disease or its response to drug
treatment.
Another aspect requiring attention concerns the statistical evaluation of
the results. For example, recently it has been shown that in a study of
asthma one genotype had a 100% positive predictive value for non-response
to a drug. However, because the susceptibility genotype only occurs in less
than 9% of patients, in practice less than 10% of the non-response to
treatment can be attributed to this abnormal genotype. In this case, it has
been calculated that avoidance of the drug as a result of pharmacogenomic
profiling would only improve its efficacy from 46% to 51%. Thus the
reliance on candidate gene variation, which ranges from 2% to 7%, is
currently not in the range for practical application.
No comments:
Post a Comment