Women present a special complexity in terms of tolerability because there are several issues to consider and the weighting of these issues may change over time. For example, for a young person – a teenager – going onto antiepileptic drugs (AEDs), compliance is important: a drug that is taken once or twice a day has an advantage over one that is taken many times a day.

Cognitive issues will also be important. However, as women get older it is important to consider interactions with the oral contraceptive pill, as well as fertility issues: for instance, there are some potential linkages with conditions such as polycystic ovary syndrome (PCOS). Some drugs have been shown to have a higher teratogenic potential than other drugs, and there is also evidence of a long-term risk to bone health with some AEDs. No drug as yet ticks all the boxes all the way through the various ages, so the emphasis changes over time as well as for particular individuals.

What are the risks of taking antiepilepsy medication while pregnant?
I help to run the UK Epilepsy and Pregnancy Register, which is one of four major epilepsy pregnancy registers worldwide. It has been running for 11 years now and useful information is just starting to emerge. One always has to accept the codicil that these registers are not randomised controlled trials and therefore there may be inherent biases. Nevertheless, they do seem to demonstrate that there are potential differences among the AEDs in terms of risk of major congenital malformations.

So far the spotlight has shone on sodium valproate. The UK register highlights the fact that sodium valproate is associated with a significantly higher risk of having a child with major malformation than, for instance, carbamazepine or lamotrigine.

However, one should always emphasise that, although there is a higher risk, that risk is only between 6 and 7%: in other words, a woman taking sodium valproate still has at least a 93% chance of having a perfectly normal child (better odds, incidentally, than those reported for valproate–lamotrigine polytherapy). Therefore, the results are generally reassuring, but there are differences, which raises the potential of being able to reduce risk.

The larger issues with major congenital malformations concern delay in cognitive and behaviour development. There are studies emerging from Gus Baker’s group in Liverpool, and we have a paper in the pipeline, that show that children exposed to sodium valproate in utero seem to have higher levels of neurodevelopmental, cognitive or behavioural delay than do other children. This is an issue that is going to emerge over the next few years and may be another reason for choosing one drug over another.