Pharmacogenetics in assisted reproduction: optimising response to ovarian stimulation

Pharmacogenetics is not new. One of our authors famously can drink a bottle of wine with little effect. The other falls asleep after just one glass. Indeed, the speed at which alcohol is broken down in the body is partly genetically determined. This principle applies to most, if not all, foods and drugs. Pharmacogenetics is the study of how a person's genes can influence their response to medication. In the latter half of the last century, researchers discovered that genetic variation could explain why different people respond to the same medicine in different ways. Since then, progress in pharmacogenetic research in parallel with the Human Genome Project has promised real potential for translation from bench to bedside.

A classic example of the importance of pharmacogenetics in medicine is in the treatment of childhood leukaemia. Leukaemia, perceived for decades as a death sentence for children, was transformed by the discovery of miracle drug, 6-mercaptopurine, which effectively cured over 80 per cent of affected children. Unfortunately, some children still died despite treatment. Some because of the leukaemia - the drug didn't work. Worse still, some were killed by the drug. More recently, a simple blood test was used to identify individuals with a genetic variant of the enzyme responsible for breaking down the highly cytotoxic 6-mercaptopurine. This test identifies children most at risk of dying because of the drug and those who might not respond at all.

Clearly, pharmacogenetic testing is an attractive prospect considering the estimated 10,000 deaths every year in the UK as a result of adverse drug reactions. An even greater number suffer detrimental side effects. For others, their medicine has no effect, positive or otherwise. In many cases, differences in drug response are caused by a change in only a single letter out of the six billion letters of the entire DNA code. We have millions of these changes, or single nucleotide polymorphisms (SNPs), in our genomes but the specific effects of each and every SNP are largely unknown. What we do know is that SNPs make each of us different and these differences extend beyond physical appearance into differential response to drugs.

What has 6-mercaptopurine got to do with assisted reproduction? We are not involved in prescribing life-threatening drugs - or are we? In 2005, the BBC reported the death of a woman from ovarian hyperstimulation syndrome (OHSS). Severe OHSS affects about one per cent of women undergoing IVF, although a less severe form is more common. Fatal adverse reactions to stimulation are extremely rare, but any avoidable death in a fit and healthy women is a death too many.

OHSS is the direct result of an over-exuberant ovarian response to the administration of Follicle Stimulating Hormone (FSH), the aim of which is to produce large numbers of oocytes and thereby improve IVF success. Fertility specialists currently consider age, body mass index, basal FSH, ovarian reserve assessment by ultrasonography, hormonal profile (including FSH) and previous response to FSH to select a drug dose that will limit the chances of OHSS but still maximise the chances of success. If this appears to be a little like guesswork, well, it is. A growing body of evidence is supporting the production of a patient-specific gene profile that identifies relevant SNPs to aid in fertility drug dosing. One particular gene, FSHR, encodes the receptor for FSH. Mutations in this gene are rare, but functional SNPs are common. One SNP - a change from the letter G to A - appears to affect the way women respond to FSH. Recent reports suggest women with two copies of SNP A don't respond to FSH as effectively as those with two copies of G, as measured by basal FSH levels (1), clinical pregnancy rate (2) and FSH dose required to achieve the same serum ovarian response in IVF cycles (3).

The goal of any pharmacogenetic profile is to avoid adverse drug reactions. In the world of assisted conception adverse drug reactions includes both over-response (with the possibility of life threatening OHSS) or under-response (leading to treatment cancellation or poorer outcomes). Indeed, the possibility of adjusting the dose based on pharmacogenetic test results in conjunction with existing parameters such as age, weight, previous response, etc, offers an improvement over 'one-size fits all' prescribing. A simple blood test can help identify individuals who may have a lower or excessive response to FSH. It really is that simple, and the beauty of this test (unlike basal serum FSH for example) is that it doesn't change from month to month. How would such tests work in practice? Patients with the A/A SNP who may respond poorly to FSH could benefit from a higher initial dose. Conversely, G/G patients might be able to avoid side effects such as OHSS by taking a lower dose.

As we expand the pharmacogenetic profile for patients undergoing assisted conception procedures to include SNPs in oestrogen receptor, aromatase, progesterone receptor and eventually the cytochrome P450 genes (responsible for metabolizing nearly half of all drugs in common usage), we may find that this 'profile' is also beneficial for other later non-fertility medical treatments. Furthermore, such profiling need not be restricted to IVF patients. Any women being treated with drugs for fertility problems could benefit. It might also predict who is more likely to succeed following minimal stimulation IVF or even natural cycle IVF (in which no drugs are used). We believe that any test that can help to improve the safety and efficacy of what is potentially the most dangerous part of the IVF process is worth doing. The ultimate goal for all of us should be to prescribe the right drug, at the right dose for each individual patient.