Drug designers factor in gene type

Sarah Houlton reviews recent developments in genotyping that have paved the way for more effective drugs with reduced side-effects.

With all the hype that surrounded the decoding of the human genome, one might have been forgiven for thinking that massive advances in human healthcare were imminent.Of course, in the real world, instant cures for all ills were never going to be possible. But it has helped the science of pharmacogenetics immensely, and the growing understanding of disease states has also led to the discovery of numerous new drug targets that could have great potential, particularly in the field of cancer therapy.

Many of the older cancer drugs were non-specific and frequently associated with serious toxicity problems. But an insight into the genetic basis of the disease has revolutionised the science of searching for new cancer drugs: many of the pathways involved have been identified, and targeted interventions can thus be developed. Of the 30,000 genes in the human genome, it is believed that 10-15% have potential in treating disease in some form or other.

Pharmacogenetics is not a new science. However, in combination with the vast amount of information that has come out of the human genome project, it is now giving a much greater insight into how certain individuals respond to specific drugs. Variations in genetic make-up can have a huge bearing on how effective a drug is, the side-effects that may be experienced, or even whether it works at all in a particular individual. This throws up the fascinating possibility of personalised medicines, and these are already beginning to reach the market as several treatments now require an initial genetic test to establish whether the patient is likely to respond or not.

The pioneering drug in this field was Novartis' imatinib (Glivec), which was designed to treat chronic myelocytic leukaemia (CML). The previous standard treatment for this disease was a combination of interferon alpha with a low dose of the drug cytaribine - a regime that is expensive and causes side-effects. Imatinib is a tyrosine kinase inhibitor that interferes with internal cell signalling pathways involved in the development of the disease.

faulty gene

Around 95% of those with CML have the abnormal Philadelphia chromosome, which results from the transfer of a fragment of DNA between chromosome 9 and chromosome 22. This gene fragment fuses with part of another gene on the second chromosome, giving a hybrid gene that continuously manufactures tyrosine kinase, an enzyme that forms part of the cell division, survival and growth process.

Novartis scientists established that this was indeed implicated in leukaemia by inserting it into mouse DNA, and so they reasoned that if they could block the activity of tyrosine kinase, it should be active against CML in those people with the faulty gene.

And imatinib does, indeed, have a dramatic effect against CML, and also largely eliminates the Philadelphia chromosomes from both the blood and bone marrow. Because the therapy is so targeted, it was approved by the regulators in record time, and its potential in other forms of blood cancer is also being investigated. However, resistance has emerged in some patients, probably as a result of mutations in the rogue gene that makes it more difficult for the drug to bind. Second generation tyrosine kinase inhibitors are now under development.

Another pharmacogenetic medicine has been in the headlines recently. Roche's Herceptin (trastuzumab) is a monoclonal antibody that is targeted at the HER2 receptor, which is overexpressed on breast cancer cells. HER2 is believed to be activated by human epidermal growth factor (EGF), and when trastuzumab binds to the HER2 receptor, it prevents EGF activation. However, this rogue gene is present in only around a fifth of women with breast cancer, so there is little point in treating those who do not have the HER2 gene as the drug would have little effect. In some women with the gene, the drug shrinks the tumour; in others it merely arrests its growth.

serious side-effects

While at first sight trastuzumab may appear to be a 'magic bullet' for this particularly aggressive form of breast cancer, there are a number of potential significant side-effects, notably cardiovascular problems. The drug is currently licensed for use only in advanced cancer, but as a result of promising trials results in early stage breast cancer which indicated that women given the drug are 50% less likely to suffer recurrences, there has been a clamour for it to be made available for women in the early stages of breast cancer. This is despite the very real potential for serious side-effects - and completely disregards the fact that Roche has not yet even put in an application to the regulators for authorisation in early breast cancer.

The UK health secretary's decision that local health trusts should fund the drug flies in the face of normal licensing and safety procedures, and introduces the very real possibility of many women being adversely affected by its negative effects.

specified genotypes

Another controversial anticancer drug - ImClone's Erbitux (cetuximab) - also works better in people with certain specific genotypes. The chimeric mouse-human antibody inhibits ligand-dependent activation of the epidermal growth factor receptor (EGFR). This receptor is overexpressed in a number of solid tumours, including colorectal and non small cell lung cancers. Cetuximab attaches itself to this receptor, preventing ligand binding and thus halting ligand-mediated receptor tyrosine kinase phosphorylation - a process which is implicated in cell growth and tumour progression - and causing receptor internalisation.

Cetuximab is approved for use in EGFR-expressing metastatic colon cancer - if the patient does not have the relevant gene, the drug is unlikely to have any effect, and those with high EGFR gene copy numbers are more likely to have a poor prognosis. The controversy surrounding the drug hinges on its efficacy; if it is given only to those with the correct gene then it is far more likely to have an anticancer effect. It has recently been submitted to the FDA for use against squamous cell carcinoma of the head and neck, and has been granted priority review.

Genetics are not just helping the search for new drugs, however. There is also great potential for pinpointing those patients in whom certain drugs might be expected to work more effectively - and to exclude those who are more likely to suffer serious adverse events from treatment.

The genetics of breast cancer is the subject of a big study that is being carried out in at the Indiana School of Medicine in the US. A total of 500 women being treated for breast cancer with aromatase inhibitors such as anastrozole, exemestane or letrozole will be followed for five years to try to isolate genetic variances in them that might predict the effects these anti-oestrogen drugs will have on breast cancers, and to monitor how they are metabolised. The researchers also hope it will help predict which women are most likely to experience menopause-like hot flushes when they are taking these medicines.

drug effectiveness

This is just one example of a study that is being carried out to investigate the genetic background to the effectiveness of cancer drugs and their side-effects. Pharmacogenomics is being used to probe inherited genetic variations that may enable accurate predictions to be made of how individual patients are likely to respond to their chemotherapeutic regime. Variations such as nucleotide insertions, deletions and repeats - not to mention single nucleotide polymorphisms (SNPs) which change the amino acid sequence of the proteins the gene codes for - can all have an effect, as can gene transcription and RNA splicing. When these occur in those enzymes that metabolise medicines, such as the cytochromes P450, or in molecular targets or transporters, this has the potential to affect the way a drug works or is metabolised.

There are marked genetic variations between different ethnic groups, and studying haplotypes - a combination of polymorphisms that are inherited together - can be more effective than looking at single polymorphisms on their own.

Polymorphisms in drug targets can have a dramatic effect on the effectiveness of medicines. One example is 5-fluorouracil (5-FU). Despite the fact that it has been in clinical use for half a century, 5-FU remains one of the most widely used anticancer agents. It is active against a variety of tumours, and its synergistic interactions with other drugs mean it is commonly used in combination therapies. A major mechanism of action is the inhibition of the enzyme thimidylate synthase, which is involved in the synthesis of thimidylate. This is used to make thymidine triphosphate, which is involved in DNA synthesis and repair.

The development of clinical resistance to 5-FU has been associated with the overexpression of thimidylate synthase (TS) in tumours. The levels of enzyme expression seem to be regulated by the number of polymorphic tandem repeats in the TS enhancer region (TSER), which appears to increase expression of TS. People with three tandem repeats have been shown to have significantly higher TS mRNA expression than those with just two, and also had a lower response rate to 5-FU.¹ This, and other studies, indicate that carrying out TSER genotyping could be of great benefit when selecting those patients who are likely to respond best to 5-FU and similar drugs.

metabolic effect

A number of examples have been found of polymorphisms in drug metabolising enzymes that have relevance in cancer treatment. A good example is the purine antimetabolite 6-mercaptopurine, which is used to treat leukaemia. It works by inhibiting the formation of nucleotides used in DNA and RNA synthesis. The enzyme thiopurine methyltransferase (TPMT) S-methylates the drug, inactivating it. However, around one in 300 people have an inherited TMPT deficiency that means they have a high risk of severe toxic effects, as their bodies metabolise the drug much more slowly. The problems can be overcome by reducing the dose levels to compensate, and children being treated for acute lymphatic leukaemia are now routinely tested for this mutation before dosing with 6-mercaptopurine begins.²

5-Fluorouracil is also affected by metabolising enzyme polymorphisms. Up to 95% of the 5-FU administered to the patient is catabolised into inactive metabolites that are excreted, with just 5% anabolised into the cytotoxic nucleotides that are active against tumours. As a result, the activity of the relevant enzymes in the body has an enormous bearing on the systemic exposure to fluorodeoxyuridine monophosphate and the incidence of adverse effects.

The rate-limiting step in the catabolism process is catalysed by the enzyme dihydro-pyrimidine dehydrogenase (DPD), which is completely deficient in around 0.1% of the general population and partially deficient in a further 3-5%. Severe toxicity and fatalities have been associated with 5-FU treatment in patients with DPD deficiencies.³ The deficiency is thought to be a result of multiple polymorphisms in the DPYD gene, the result being a decrease in enzyme activity, but the genetic basis of DPD deficiency and the resulting raised toxic effects of 5-FU are still being investigated. There is not always an obvious mutation in the gene, and other markers will be needed if high-risk patients are to be identified in advance of treatment.

One of the most widely prescribed anticancer agents, Pfizer's Camptosar (irinotecan), is a prodrug that is activated in the body to give an active metabolite, 7-ethyl-10-hydroxycampto-thecin, that inhibits topoisomerase-1. While it has a wide activity profile against numerous forms of tumour, dose-limiting diarrhoea and myelo-suppression can develop. These adverse effects are commonly associated with increased levels of the active metabolite, and genetic variations in an enzyme that glucuronidates the active metabolite (UDP-glucuronosyltransferase 1A1) are believed to be a major causative factor.

One mutation of this enzyme is associated with reduced glucuronidation of the active, resulting in increased toxic reactions to irinotecan in patients. The relevant mutated genes are present in around 35% of Caucasians and African Americans, but the frequency is much lower in Asians. There are, as yet, no clinical guidelines, and trials are being carried out to investigate the impact of dose on the safety of those with the mutation.⁴

drugs revisited

This handful of examples will be joined by many more over the coming years as our understanding of the genetic basis of disease and drug metabolism increases. It may even see some drugs that were thought to be dead because of side-effect profiles returning to the market with box warnings against their use in certain diseases. Merck's Vioxx (rofecoxib) was withdrawn because of the cardiovascular problems that developed in a small proportion of patients. What if these could be put down to a specific genetic variation in a small percentage of the population and a simple test for the variation were developed?

It was a very good antiarthritic drug in the majority of patients who were not adversely affected. It could be back, and with those most likely to suffer cardiovascular complications excluded, it would once again be a safe, effective treatment for arthritis.