Valproate and reproductive risks in women and men with epilepsy: Dr. Tony Marson

In use since the 1970s, valproate is highly effective for genetic generalized epilepsies and some forms of pediatric epilepsy. The medication also is associated with a 10% risk of birth defects in children exposed to it in utero, as well as a 30% to 40% risk of neurodevelopmental issues, including autism. Sharp Waves spoke with Dr. Tony Marson about the reproductive risks for women and men taking valproate, discussing UK policies meant to ensure appropriate prescription of valproate and communication about its risks. The recent policy regarding men who take valproate, and the evidence used to support that policy, is discussed near the end of the episode.

Since the episode was recorded, the UK has changed its policy regarding men taking valproate; see this link for more information: February 2025 update on valproate policy in the UK

UK Drug Safety Update page - June 2025

February 2025 update on valproate policy in the UK

Regulatory safety and educational materials on valproate - for health care professionals and patients

Public assessment report: Managing reproductive risks in males under 55 taking valproate

Sharp Waves episodes are meant for informational purposes only, and not as clinical or medical advice.

Let us know how we're doing: podcast@ilae.org.

The International League Against Epilepsy is the world's preeminent association of health professionals and scientists, working toward a world where no person's life is limited by epilepsy. Visit us on Facebook, Instagram, and LinkedIn.

Transcript

Nancy Volkers: This episode discusses UK policy on prescribing valproate to women and men under age 55.  The policy originally required two experts to endorse both new and existing prescriptions for both women and men. In February 2025, the UK Commission on Human Medicines stated that current valproate prescriptions for men will not need review by two experts.

New prescriptions for men and both new and existing prescriptions for women will still require two expert review. A link to the UK drug safety update page is available in the show notes.

Welcome to another episode of Sharp Waves, and today we are discussing sodium valproate or valproic acid. We have Dr. Tony Marson with us, and I would like him to introduce himself first, and then we'll get into the topic. Dr. Marson. 

Dr. Tony Marson: Hi. Yeah, so I'm Tony Marson. I'm an academic neurologist in Liverpool in the UK. I spent nearly half my life being a clinical neurologist, mainly seeing people with epilepsy in our huge epilepsy service in Liverpool. And the other half as a researcher, and the focus of my research has really been on clinical trials and deriving evidence which informs everyday decision making, so trying to inform the decisions that we make in the clinic.

Nancy Volkers: Excellent, thank you. So could you start by giving maybe some background on valproate? How long has it been in use for epilepsy When is it most likely to be used? 

Dr. Tony Marson: Yeah, so it's a medication that's been used since the 1970s. It was discovered by accident in some animal models. It was effectively the solvent being used to try a whole lot of potential anesthesia medications in an animal model. And it turned out that most things were working and it turned out that it was valproate that was the effective agent. 

And over the years, we've discovered that it's a really effective treatment for a group of epilepsies that we call the idiopathic or the genetic generalized epilepsies. And it's also effective in some of the rarer and more difficult pediatric epilepsies, although it's often thought of as a drug with broad spectrum of efficacy. So it has efficacy in most seizure types, but it is most effective in the generalized epilepsies. And for a long time it was used as a first line treatment, but we've increasingly discovered that it's not a safe drug in pregnancy. 

So when I was at medical school back in the 1980s, we were told about spina bifida and neural tube defects associated with sodium valproate, and that was something to worry about. But then over the course of my career, we've learned that it causes a whole range of malformations and it's particularly toxic at high doses. And it causes significant reduction in IQ in about a third of children that are exposed to it in utero. So we try hard not to prescribe it in women of childbearing age. 

Nancy Volkers: Thank you. Yeah. There is sort of a history of guidelines or policies that limit its use in women of childbearing age. Can you talk a little bit about those and kind of where they stand now and how, whether they've been effective?

Dr. Tony Marson: Yes. As our knowledge is accumulated around the harm when it's taken during pregnancy, the regulations and the policies have changed and they've also started to change for men more recently, I think we'll come on to men later during the podcast. In Europe, we refer to a pregnancy prevention scheme. And I'm aware that there are similar policies in the States and in other bits of the world. 

So we've got this really challenging kind of risk benefit ratio to get right, because on the one hand we've got a treatment which we know is the most effective. And in randomized controlled trials, we know that your chance of having your seizures controlled on valproate are about 20% higher than they are on say, levetiracetam or lamotrigine. So on the one hand, we know that valproate's more effective at controlling seizures. But on the other hand, it is more harmful in pregnancy. And of course we're starting treatment in sodium valproate quite often in children and teenagers where pregnancy isn't on the top of their list of things to do at that time.

But we know that if we've started people on treatment, then they may be still on it in a number of years when pregnancy is much more of an issue. And there's a really important kind of question here around women's and females’ rights and autonomy. What right have they got to the most effective treatment? And again, what responsibilities do they have and do we have as clinicians to and ensure that any future pregnancy is safe? 

The way that that's worked out in policy is that when a diagnosis is made of one of the generalized epilepsies, women and girls are generally offered an alternative to valproate, like lamotrigine or levetiracetam, and valproate is kept in reserve for people whose seizures aren't controlled on the first line treatment or maybe there’s a second or even a third line treatment before valproate is truly considered.

The risk there is that on the one hand, women don't get their epilepsy treated properly because they have to wait for one, two, or three drug failures before they finally get a treatment which is likely to control their seizures. But on the other hand, if we use valproate too early, there's a risk that more pregnancies might be exposed to valproate.

The pregnancy prevention schemes around the world will typically involve a requirement that women are counseled, adequately counseled about the risks, and also that they sign an annual risk acknowledgement form to say that they understand the risks and that they will avoid pregnancy whilst on sodium valproate. 

In the UK, and I'm sure in many other countries, this hasn't worked as well as it should have done. There are some women on valproate that weren't achieving the counseling, and there was an inquiry in the UK led by Baroness Cumberlege, who highlighted that the fact that the processes and the systems weren't working—women weren't receiving the counseling they should have been receiving. And some pregnancies have been exposed to valproate where it could have been avoided. So the regulations are increasingly tightened up in the UK and in many other countries.

Nancy Volkers: So when you talk about counseling women about the risks, I guess my question would be, is there training for this? Are there protocols? Do you have a sense of what the obstacles are there?

Dr. Tony Marson: I think there are two main obstacles. One, one is manpower or woman power. Person power. And the provision of services. It tends to vary quite significantly geographically, both within countries and between countries. If you live in a city, there's a pretty high chance you've got access to a neurologist with expertise. If you live in a more remote place or where the resources are more scarce, you may have less access to a neurologist and expertise. But increasingly it's very clear that women on valproate should be accessing specialist expertise. 

The other problem is that epilepsy is a long-term condition and many people will have well controlled epilepsy and may have been started on valproate before we were aware of all of the problems. And if people have well controlled epilepsy, they might not be very keen to go and see their neurologist because they're not having seizures and they're not aware of problems with their treatment. And also health services or health systems might not have the information available to readily pull those people back into clinical services. 

I think one of the problems is the kind of a prevalent population of people with epilepsy that are on sodium valproate that aren't actively engaged in services, that aren't necessarily getting the counseling. One of the challenges that we're getting on top of in our service is making sure that we know exactly who all the people are being prescribed sodium valproate in our bit of England, be that for a neurological or a mental health problem. And we are making sure that they are accessing services and receiving the counseling. 

In terms of information provision, there are booklets and there is information provided for both clinicians and patients which is readily available and provides the information about the risks and information about effective contraception.

I don't think people need specific training to provide that guidance. I think most clinicians should have the skillset to communicate with patients, their parents, and significant others to explain about the risks and help with decision making.

Nancy Volkers: So we've talked about risks and you mentioned spina bifida and maybe a couple of other ones earlier. I'd like you to talk about what risks are known for women of childbearing age, if they are taking valproate and they become pregnant, what sort of risks are we talking about and what is the evidence for those risks?

Dr. Tony Marson: So, so the evidence for those risks comes mainly from pregnancy registers. There are a number of pregnancy registers that have been undertaken in Europe and in the States and other places. They provide an estimate of risk.

 The risk on a fairly high dose of valproate might be around a 10% risk of a major malformation, be that spina bifida or a problem with the face, or a cardiac defect. So there are the whole, you know, range of abnormalities that people might have. But if you add them all together, that's sort of a ballpark. Whereas it's a kind of a 1% or 2% risk on some of our safer anti-seizure medications.

Nancy Volkers: Thank you. And then there's also some neurodevelopmental risks you mentioned, I think you mentioned IQ before. There may also be an autism risk. 

Dr. Tony Marson: Yeah, so, so this kind of ballpark risk is about a 30 to 40% risk of a developmental syndrome if you are exposed to valproate in utero and that includes kind of significant reduction in IQ as well as potentially autistic traits or autistic spectrum disorder. The nature of those disorders continues to be investigated by researchers and the precise nature of those disabilities are being defined as cohorts of children are getting older. We’re learning more about the impacts on later life and functioning in society.

Nancy Volkers: And when you say 30 to 40%, this is an absolute number. It's not 30 to 40% higher than women taking lamotrigine, for example. 

Dr. Tony Marson: It's an absolute number. 

Nancy Volkers: Okay. 

Dr. Tony Marson: So it's, it's a lot. It's a lot. And hence the concern and the desire to make sure that virtually no pregnancies are exposed to sodium valproate.

And of course, it's completely impossible to prevent all pregnancies being exposed to valproate. You know, accidents happen and some women choose to have a pregnancy despite being on valproate and knowing the risks, and some people's epilepsy can only be controlled by sodium valproate, 

Nancy Volkers: What's known about discontinuing valproate or switching to another anti-seizure medication? I believe you have a clinical trial coming up, or you're working on one on this topic. 

Dr. Tony Marson: Yeah, sure. So, the clinical trials that we've done have been comparing one anti-seizure medication with another. And that's what provides us with the evidence about how much better sodium valproate is compared to the other treatments. What's more difficult to run a clinical trial on is the effect of withdrawal of sodium valproate. It can be done, but if we wanted to do it now, we'd have to wait 10 years for the results. So we need to use other methods to get an estimate now that we might want to test in a clinical trial moving forward. 

And that's really important for sodium valproate because one of the concerns is that the people might think, well, the policy is that we just take everybody off sodium valproate. And that poses a risk because from the clinical trial data, you’d estimate that of the people seizure free on sodium valproate, probably around 20% of them are going to start having seizures again if you swap them from sodium valproate to an alternative like levetiracetam. 

One of the ways that you can investigate this is in routine clinical data. So there are huge databases of routine clinical data that could be used to interrogate and you can develop algorithms whereby you can identify all the people with epilepsy in those data. Then you can set up a bit of an experiment, which looks a bit like a clinical trial or a controlled study. So you can identify people that were on valproate and stayed on it and were on valproate but came off it and then compare what happens to those two groups of people.

And one of our concerns when we did this, you're referring to a paper that we had published in Brain last year—one of our concerns is that coming off valproate might be associated with increased mortality, but we didn't find that. We didn't find an increase in death rates, but what we did find was increase in many other adverse health outcomes. Increase in injury, increase in fractures, increase in emergency department attendance, increase in depression and poorer mental health outcomes. So what was very clear is that coming off valproate in this kind of experiment in routine data seemed to be associated with adverse health outcomes.

And I think what you would predict, given what we know about the higher efficacy of sodium valproate. And that message is really important. And I think of the concerns is that regulators have a very kind of blinkered view of risk. They're very concerned about the risk, the adverse effects, risk of a treatment, but they're not necessarily as focused on the risk of not actually having the treatments.

So the risk is the regulators say, “Well, this drug is harmful, therefore we stop prescribing it.” And that's clearly not in the best interest of people with epilepsy because you would be denying them a treatment which is actually the most effective for their condition. 

Nancy Volkers: There was a recent UK policy that required two specialists to confirm that valproate was necessary as a treatment for epilepsy in both women and men. And there have been quite a few news articles about this on both sides. Some saying that, you know, these restrictions are infringing on people's right to be seizure free on the medication. Others saying that they are necessary. And then there's a particular flurry of articles, I guess on the paternal side of things. So men who may be fathering children while they're on valproate, these restrictions also apply to them. Can you talk a little bit about that policy and maybe where it came from? 

Dr. Tony Marson: Yes. So the UK regulator came up with a policy that there needed to be a double signature if valproate was being started in both men and women. And it also wanted a double signature for the prevalent population as a one-off. What the regulator is asking there for is scrutiny and assurance that valproate is being prescribed to people with generalized epilepsy and being prescribed appropriately. 

My view is that that step was unnecessary. You might suggest that there is additional scrutiny that could be done with audits or a less heavy-handed way. But the double signature approach is an expensive thing to do because it requires additional clinical time to review cases or to have a conversation about cases. And it also delays people getting access to treatments because you can rarely get that double signature there and then when you've made a decision to start treatment, you get a double signature probably a few days later, maybe later that day. But inevitably the patient's gone home and then they may need to come back to pick up a prescription or go some other route. So the whole thing is clunky and expensive and probably unnecessary in my view, but it wasn't my decision. And it's a decision that we have to live with and we're working hard to find solutions to make this as efficient as possible.

You can understand why the regulators were concerned about that and in particular that the families of children that have been exposed to valproate in utero were making a case that women aren't being counseled and we're not sure they're being assessed properly.

And the response in the regulator was to say, well, we need external scrutiny. We need another signature, and we just have to live with that. I think it's more difficult to justify for men because. We have data from a Scandinavian registry study, which suggests that there might be an increase in risk of neurodevelopmental disorders in children born to men taking sodium valproate. But it is possible that there is confounding there and it's difficult to be sure whether valproate is actually the cause.

But the regulator has considered the data and is concerned. And I think it's tough for the regulator because the regulator gets criticized when it doesn't act, and it got very severely criticized for not acting on sodium valproate for women, and now it's getting criticized for acting when it's seen signal of potential risk in men.

I think they've probably over overdone it from the male point of view, but I guess we'll have the data in a few years’ time to help us be sure whether the regulator have jumped too soon or whether in fact they were right all along. But our policy for many years has been to avoid valproate as a first line treatment for women, but to use valproate as a first line for men. But increasingly, I think the regulations are going to result in fewer men getting access to the most effective treatment for their epilepsy so that's kind of where we're at.

We've known for a long time that that valproate is associated with abnormal sperm count. So reduced sperm count, reduced sperm motility, and, and altered morphology. And there's the potential risk of reduced fertility for men. So whilst we've got the data that says the sperm abnormal and the testes might be smaller. We've got actually no evidence to support the fact that there is an association with valproate and reduced fertility, although it's kind of, it may be obvious that there might be a link there that that link, as far as I'm aware, hasn't, been proven.

 If you look at the literature, most of literature is about male fertility. The evidence for valproate causing neurodevelopmental defects when it's the farther exposed is pretty sparse.

 One of the concerns is that you can make animals autistic by exposing them to high doses of valproate in utero. That's one of the animal models for autism. And what is observed is that the offspring of those rodents also have some autistic traits, even though they weren't exposed in utero. So one of the concerns that needs, addressing is that there might be a risk, but also a transgenerational effect.

So there are two big concerns in the background which remain unproven. One, are children born to men taking valproate at risk? And also is there a transgenerational effect? With those concerns, the regulators made a judgment that men have to be counseled and we now need to do the science to find out whether it's true or not. This is a really challenging area and we must try hard to get the balance right between benefit.

Having controlling seizures, getting people back to work and education versus risk in pregnancy. What is really important is that the opinion of the woman and the rights of the woman is not lost in all of those decisions. And one of the risks is that the regulators are taking the rights of women away.

It’s really important that we have conversations with women, we make sure we're making balanced risk decisions. And it's also really important that we contribute to developing the evidence base so that we can make sure that our decisions are as best informed as possible in the future.