Valproate Raises Questions Again About the Quality of European Drug Regulation

Have the regulatory authorities in Europe dropped the ball again? After the Mediator debacle a decade ago, in which thousands of patients were exposed to dangerous adverse effects, European rules on drug safety monitoring were tightened up and the watchword was “never again.” But now the teratogenic effects of sodium valproate are being revealed in all their appalling severity and scale-50 years after its launch, and in a chilling echo of the thalidomide tragedy that was the genesis of modern drug regulation more than half a century ago. The European Medicines Agency (EMA) hosted a public hearing on Sept. 26 as part of a new review of valproate use. The agency is considering whether to restrict use of valproate-containing medicines by women of childbearing age more severely than it already did in 2014. 

Did patients know?

The warnings and restrictions then were designed to ensure that patients were aware of the risks of malformations and developmental problems in babies who are exposed to valproate in the womb. But plenty of evidence, anecdotal and systematic, suggests that many patients were not made aware. EMA admits that “concerns have been raised about how effective the measures have been in increasing awareness and reducing valproate use appropriately in its various indications.” In France, in particular, where a class action is now underway against Sanofi, the principal manufacturer there, the national medicines regulator has asked EMA to consider whether further EU-wide action should be recommended.

The agency’s pharmacovigilance risk assessment committee is examining the available evidence and consulting with stakeholders. Nearly 100 members of the public applied to take part in the September hearing, to talk about their experiences with valproate.

Long history of questions

This episode has, at first sight, all the appearance of a gigantic error by regulators. The risks associated with valproate in pregnancy have been well known for years, with early questions raised as far back as the 1980s.

In 2006, the U.K.’s National Institute for Health and Care Excellence (NICE) said, while assessing Abbott’s version of the drug, Depakote, that it should not be routinely prescribed for women with child-bearing potential. Already then, studies were suggesting that many doctors and patients were not aware of the risks related to its use. Over the following years, concerns proliferated in a range of published studies, and by 2011, NICE was having to defend itself against accusations that its evaluation of the drug was ignoring the risks. In 2015, the U.K.’s medicines agency urged healthcare professionals to give better information to women about the risks, and in France, where 80,000 women were taking the drug, the national agency limited valproate prescribing to specialists. As anxieties mounted and lawsuits started to fly, the French health minister instigated an inquiry into whether sufficient warnings had been given by doctors.

 By early 2016, the state investigators concluded that the health authorities and drug firms in France had shown “a lack of responsiveness” about the need to provide adequate patient information. Sanofi said that it had “always been proactive” on the subject and had respected its “obligations to inform health professionals and patients,” and had asked the authorities to update product information as far back as 2003. But for generic versions, leaflets and product information summaries were updated only in mid-2015-and elsewhere in Europe, other authorities were also accused of displaying “a degree of inertia.” 

Answers needed

So what has gone wrong? Have regulators not acted with sufficient energy? Have drug firms been slow to modify their drug information? Have prescribing physicians not taken account of advice to act with caution, or to inform patients of risks? Or have patients not listened, or not understood? It is not, of course, quite so simple. For all its hazards, there are cases in which valproate is the only drug which can adequately control bipolar disease or epilepsy. EMA states in its invitation to the September hearing that “sometimes there may be no alternative to using valproate.” The risks of dropping treatment before or during pregnancy also have to be weighed against the risks of fetal damage. And the incidence of adverse effects is not universal: valproate has been shown only to increase the risks significantly. Defenders of valproate point out that uncontrolled epilepsy also poses a risk for both the fetus and the mother, and claim that switching treatments once the patient is pregnant can also carry risks.

Nevertheless, across Europe there are now hundreds, even thousands, of parents of children with developmental problems, and they are now complaining vociferously that mothers were unaware of the risks when they took valproate. The official reviews and investigations now underway seem rather late and leisurely in the circumstances. And the questions raised-and that should be answered promptly-go much wider than the specifics of valproate. They need to cover a range of issues relating to information to patients, communication between regulatory authorities and prescribers, and regulatory focus at national and European levels when signals accumulate about risk.

Business (almost) as usual

 Separately, but interestingly, the EMA is erring on the side of caution in accepting candidates for its PRIME scheme of enhanced support for the development of medicines that target an unmet medical need. At its September meeting, the agency's chief scientific body, the Committee for Medicinal Products for Human Use (CHMP) rejected all four applications-a flu vaccine, a treatment for oesophagitis, a product to increase fistula survival, and a therapy for stage four esophageal carcinoma. 

 This brings to 97 the total of eligibility requests denied, against just 28 accepted, with five more ruled out of scope. Oncology is the category with most requests granted, with nine out of 43 applications,  followed by hematology, with seven out of 12. Two requests have been granted for neurology, gastroenterology-hepatology, immunology, metabolism, and psychiatry. So far, one has been granted for infectious diseases and one for vaccines. The smaller companies that the scheme was largely designed to assist have not done very well out of it so far, with just 12 applications excepted out of 64.

But there has certainly been no slowdown in EMA’s  activity on scientific advice and protocol assistance. During its September meeting, the CHMP approved 26 scientific advice letters, eight protocol persistence letters, 11 follow-up scientific advice, seven follow-up protocol assistance, seven qualifications of novel methodologies, and one health technology assessment (HTA) parallel advice. It also accepted 59 new requests, mainly for initial scientific advice, but including two requests for HTA parallel advice.

Peter O'Donnell is a freelance journalist who specializes in European health affairs and is based in Brussels, Belgium