New Times,
New Thinking.

  1. Politics
  2. Health
5 February 2025

The Do No Harm dilemma

What happens when a drug that can save lives could also ruin them?

By Hannah Barnes

The most fundamental maxim of medical practice is: first, do no harm. But sometimes, this aim is not so straightforward. “Young men will die, and completely unnecessarily,” consultant neurologist, Dominic Heaney, told me. He was talking about valproate – also known as sodium valproate, or valproic acid. It is a drug prescribed to around 65,000 men and boys in England, most often to manage epilepsy.

But it has also become a byword for scandal. Valproate is a teratogenic medicine, meaning it can negatively affect the foetus during pregnancy. An estimated 11 per cent of babies born to mothers exposed to valproate during pregnancy are affected by physical defects, including spina bifida, cleft palate, or malformations of the limbs or organs; an estimated 30-40 per cent have developmental difficulties or autism. But for decades, women were unaware of these risks.

When the drug was licensed for the treatment of epilepsy in the 1970s, doctors knew that animal tests showed valproate could potentially be harmful to foetuses. A decade later, scientific papers suggested these risks were also present in humans. The public was not informed. Women who reported problems were ignored. Valproate’s side effects weren’t added to some packaging until as late as 2016. And it wasn’t until 2018 that it became highly restricted for women of childbearing age.

Up to 20,000 babies may have been affected by exposure to valproate since it was introduced. Some, like the government’s patient safety commissioner, Dr Henrietta Hughes, have branded this “a scandal bigger than thalidomide”, referring to the severe birth defects caused by the morning-sickness drug licensed for UK use in the 1950s.

Last year, the UK placed restrictions on the prescription of valproate for men, too. The evidence base for this decision is patchy at best. Yet, Heaney fears thousands of men will come off the drug as a result – and put their lives at risk. “You have to remember that epilepsy is not just about the seizures,” he said. “It’s a potentially lethal disease. The risk of death for every convulsive seizure you have is about one in 500 each time.”

Treatment decisions always come down to balancing risk: allowing patients, in conversation with their doctor, to make informed decisions about the best course of action. Providing or withholding a medication may each bring about harm. The story of valproate epitomises this challenge, and raises questions that doctors across the medical world are grappling with: where does the balance of power lie between doctors and patients? How do we weigh bodily autonomy against the duty to protect? And how do we make sure that, in seeking not to repeat the mistakes of the past, we don’t inflict other, unintended harms?

Harry was diagnosed with juvenile myoclonic epilepsy (JME) when he was 14. He’d been experiencing absent seizures – “lapses in awareness and consciousness” – and myoclonic seizures, which cause jerks of the muscles. His first full-body seizure came a year later. JME is part of a group of “genetic generalised epilepsies”, and, like others with it, Harry (not his real name) was put on valproate.

Subscribe to The New Statesman today for only £1 per week

The drug is widely acknowledged to be the most effective treatment for these epilepsies. “It’s not that the other drugs don’t work; it’s just that they’re not as effective,” Tony Marson, professor of neurology at Liverpool University, told me. Marson led randomised controlled trials comparing valproate with other epilepsy drugs. If one drug has a 50 per cent chance of controlling seizures and another a 70 per cent chance, he said, “it’s a no-brainer”. That is, unless the more effective drug comes at a cost. This, Marson said, is the issue.

Harry, now 23, is scared. “The picture for my whole life has changed,” he said. A few months ago, out of the blue, he received a text from his GP about the risks of valproate. Harry, who has always wanted a family, was left fearing that any children he fathered would be at risk of birth defects or developmental disorders. “It was a heart-drop moment,” he said. The text directed him to guidance from the Medicines and Healthcare products Regulatory Agency (MHRA), which, Harry said, was “talking about one in 20” in terms of risk to the foetus. Harry decided that if this was true, “I’m not risking having my own kids. I don’t want to force that on the people that I love.” Heaney has talked through the evidence with him, but Harry said he feels “like a kite in a hurricane” due to the “completely contrasting bits of advice”. He has decided to come off valproate, and will carefully move to another medication. Experts insist that no one should stop taking valproate without seeking advice from their specialist.

Because of the known harm it causes to foetuses, there are stringent restrictions on valproate. The MHRA, which regulates medicines on behalf of the government, has ruled valproate must not be used by women and girls of childbearing age unless they are on a long-term contraceptive, whether or not they are sexually active. They must complete an annual risk form, too. (Lesbian women are also expected to comply.) They must also be counselled on the risks. Valproate can only be used during pregnancy if no other effective treatment is available.

Despite these restrictions, according to the MHRA, the drug was still being prescribed to pregnant women. So in January 2024 the body toughened its regulations and extended some of them to men. No patient younger than 55 can now be started on valproate unless two epilepsy specialists independently agree that there is no other effective or tolerated treatment, or there are compelling reasons that the reproductive risks do not apply. Nine months later, the MHRA issued further guidance saying men on valproate should also use contraception, citing a study that showed a “possible increased risk of neurodevelopmental disorders in children born to men treated with valproate”. The MHRA told the New Statesman this “robust regulatory action” was taken because there was evidence that patients still weren’t being informed of the risks.

Every neurologist I have spoken with believes the rules are now too stringent for women; that the balance between patient bodily autonomy and state protection has been upset. But they recognise the well-established and significant risk posed to the foetus when valproate is taken during pregnancy. What they cannot understand is why valproate is being restricted in men at all.

The MHRA’s 2024 changes are based on two main pieces of evidence: a retrospective reanalysis of data on men from Sweden, Denmark and Norway, carried out at the request of the European Medicines Agency (EMA), and studies on rats and mice. The latter show a potentially toxic effect on the testes (valproate is known to impact male fertility) and a transgenerational effect, whereby valproate-associated autism passes from parent to child.

Epilepsy clinicians have mixed views on the strength of the animal data. But none I spoke to believe it warrants the current restrictions. They all point to weaknesses in the human data, however. Professionals have made their opposition known in medical journals; others have publicly urged the MHRA to reconsider.

“For men, the evidence is really poor. We’ve got evidence of benefit, but we’ve got really poor evidence of harm,” said Tony Marson. “I’m not persuaded,” agreed Dominic Heaney. One of the world leaders in epilepsy research, Kimford Meador, clinical director of Stanford’s Comprehensive Epilepsy Centre, told me the risk to men was “theoretical”.

The UK’s restrictions on use of the drug for men are tighter than in other countries. The EMA study suggested “there may be an increased risk of neurodevelopmental disorders in children born to men taking valproate in the three months before conception”. The data showed around five in every 100 children born to fathers treated with valproate had a neurodevelopmental disorder compared with around three in 100 when born to fathers treated with alternative epilepsy drugs.

But the study had several limitations. The EMA acknowledged it could “not establish whether the increased occurrence of these disorders… was due to valproate use”. As a result, the European regulator recommended “precautionary measures”: doctors must inform men of the risks and suggest they use contraception. Regulators in New Zealand and Singapore have issued similar advice.

The study also appears not to have gone through the usual process of quality assurance. While the EMA insists it has been published – directing me towards a 635-page report, a 242-page correction to that report, and a 208-page addendum – it has not been in a peer-reviewed journal. “This is not the way it’s usually done,” Torbjörn Tomson, senior professor of neurology and epileptology at Sweden’s Karolinska Institute, told me.

Other studies have not replicated the EMA’s findings, either. In 2020 Tomson analysed some of the Swedish data and found no link between valproate use in men and an adverse impact on their babies. When Danish researchers tried to reproduce the EMA study’s findings in 2024 using Danish data, they too found no link. (The MHRA acknowledged that an association was only observed when countries’ data were pooled.) A systematic review of the evidence followed by researchers in Australia. They too concluded that paternal exposure to valproate at conception is “unlikely to pose any major risk of adverse outcomes for the offspring”.

Even if we cannot be certain of harm, surely it is best to be cautious? “As long as the precautionary measure doesn’t cause more harm,” Tomson said. “And I think it probably does in this case.” The UK’s restrictions on valproate for men are dangerous, he said. “In order to know if a patient fulfils the criteria of ‘no other medication is effective or tolerated’, the patient has to be exposed to seizures. So, on the one hand, you have weak and uncertain evidence of the risk to the child, and on the other, you have definite risks to the males.”

The neurologists accept that the MHRA has a responsibility to bring our attention to early-warning signs. But, Heaney added, it also has “a responsibility to provide a balanced view that… considers the clinical impact of what might be happening as a result of their advice – men coming off their valproate”.

If the drug’s risk to children is “uncertain”, the risk to men of coming off valproate is undeniable. Professor Marson thinks it is “inevitable” that people will be harmed as a consequence of the MHRA measures. He calls this a “one-dimensional” view of risk. “[The MHRA is] very interested in the harm caused by a drug, but they seem less interested in the harm through not having a drug.”

His research has shown that valproate withdrawal is “associated with significantly increased risks of emergency department attendance, hospital admission… and new-onset depression”. The MHRA cites this study to support its assertion that “there are no data to prove any association between regulatory action” and higher rates of sudden unexpected death in epilepsy (SUDEP). But Marson’s study looked at those who came off valproate long before the latest restrictions.

The neurologist Dominic Heaney told me about one of his patients, a cyclist who, having switched from valproate to another epilepsy drug, had a seizure while riding – his first in six years. He broke his hip and shoulder. “My fear is that this is going to happen again because there are thousands of young men in this situation,” said Heaney. Every week, he sees men who are “switching to drugs which aren’t… as effective at controlling their epilepsy, taking risks with their own health”.

There is also data that suggests women with epilepsy who stop taking valproate have a greater risk of death. The 2023 MBRRACE-UK report on maternal deaths in the UK and Ireland noted that the rate of SUDEP nearly doubled between 2013-15 and 2019-21: “This represents the period during which guidance on prescribing valproate for women and girls changed dramatically… None of the women who died were taking sodium valproate.”

Others fear the NHS does not have the capacity to implement the MHRA regulations. The NHS in Oxfordshire has warned they “cannot be fulfilled or sustained without either additional staffing or cancellation of large numbers of outpatient clinics”. In this region alone, 2,769 men on valproate “now need additional explanation both verbally and in writing”. In 2024 the local integrated care board added the restrictions to its risk register, saying that implementing them could result in “harm to patients and [an] impact on services”.

Given the strength of feeling against the restrictions, and the seemingly weak data underpinning them, why has the MHRA ploughed on? The most likely explanation, I’m told, is that it’s an overreaction to the intense criticism it (rightly) received over its mishandling of the dangers of valproate to women.

In her damning 2020 review into the valproate scandal, the life peer Julia Cumberlege said women “had not been properly informed as to the risk they were taking and the options open to them”. Adverse outcomes reported weren’t monitored; women’s concerns were dismissed. The failure was systemic, not just confined to the MHRA, but Cumberlege argued that the MHRA needed “substantial revision”. System-wide mistakes had been “perpetuated through a culture of denial, a resistance to no-blame learning, and an absence of overall effective accountability”.

The MHRA declined to be interviewed. Its chief safety officer, Alison Cave, said in a statement that patient safety is its highest priority: valproate is a “highly teratogenic medicine and there is evolving evidence for harms in males”. For the majority, Cave said, “there are other effective medicines for epilepsy”. Where this isn’t the case, valproate remains an option.

The MHRA is in a difficult position: once criticised for acting too slowly for women, it is now accused of acting too quickly for men. The neuropsychologist Rebecca Bromley told me that while “it feels like there’s been an over-correction”, it is clear that women were exposed to valproate for too long. The current debate reminds her of the one 20 years ago, when the risks to women were discussed but dismissed: “I’m very struck by some of the comments that are almost identical to what we had with pregnancy. That doesn’t mean to say that I think there’s a risk [in males’ use of valproate]… But I’m frustrated that we’re not doing more to find out as fast as possible.”

Others say the decision has highlighted the MHRA’s flaws. “We’ve had so many scandals and tragedies arise from institutional groupthink in the NHS,” Heaney told me. “This is how institutions make mistakes: they’re closed; they don’t listen to criticism… In their quest for abundance of caution,” the MHRA failed to adequately assess that men would be “at risk of death” if they came off the drug.

Marson told me MHRA processes are “not fit for purpose”. He claims the consultation – which he was part of – was unsatisfactory and that the MHRA wanted “advice on how to implement what they decided, not to get advice as to whether what they wanted to implement was appropriate”. The regulator said that in considering significant regulatory change, it asks the government’s Commission on Human Medicines for advice, and invites experts to inform the discussion as needed.

The New Statesman has seen multiple documents that show epilepsy charities warning the MHRA of the dangers of increased restrictions when they were first mooted in December 2022. Nine charities, including the Epilepsy Society and SUDEP Action, wrote to the MHRA in January 2023. “We do not believe it is right that our community of epilepsy organisations has had no opportunity to influence the new guidance and restrictions.”

In her 2020 report, Cumberlege recommended that the MHRA be required to “demonstrate how patient views have been taken into account”. But what do you do when patient views clash? Families with children adversely affected by valproate have, rightly, now been listened to. But what about those whose loved ones have died as a consequence of their epilepsy, without access to valproate? “Surely, what we want to know is… what do men and their partners think?” Marson said. “The regulatory decision should be informed by what they want – not imposing a diktat.”

“I suppose the issue is: whose choice is it?” asked Jane Hanna, director of policy at SUDEP Action. “Is it the state? Or is it about women and men having their eyes wide open?” The patient, she believes, has become lost in epilepsy care. “The view now is that valproate is thalidomide. It’s really not. Thalidomide was used for nausea in pregnancy. It wasn’t a life-saving medication for people with epilepsy.” Valproate means “they can be free of the worry of dying suddenly. Quite a big freedom”.

It’s clear that both women and men with epilepsy are being poorly served. Women who were so badly let down by not being told of the risks of valproate are still being failed. Women who continue to take valproate through pregnancy are not routinely followed up with unless they are part of a medical study. Research into the effects on men is even further behind. Little is known about the other anti-seizure medicines, and whether they too might prove harmful.

Medicine always involves a trade-off between benefits and harms, professional expertise and a patient’s wishes. There are good reasons for caution. Scandals such as those over thalidomide and valproate have damaged trust in doctors and medical authorities. And politicians and government regulators are human: moved by stories of tragedy and wrongdoing, and fearful of repercussions if their decisions later come to seem deeply flawed.

But an emphasis on individual experience and choice can cloud our vision, and stop us learning from systematically gathered evidence. It can result in a medical establishment that grows too cautious, losing its ability to innovate and to make difficult decisions to protect those who most need it.

In a world governed by risk, the valproate story reveals a further danger: that in trying to avoid the mistakes of the past, we overcorrect, and instead inflict further wrongs.

Hannah Barnes discusses this story on the New Statesman podcast.

[See more: The cost of net zero in the town steel built]

Content from our partners
The future of exams
Skills are the key to economic growth
Skills Transition is investing in UK skills and jobs

Topics in this article : , ,

This article appears in the 05 Feb 2025 issue of the New Statesman, The New Gods of AI