European Medicines Agency

• 1. Introduction

  ›

  The European Medicines Agency exists because, for decades, a pharmaceutical company trying to sell a single drug across Europe had to win separate approval from every single member state. That process cost the industry an estimated 350 million euros every year. It also exposed a problem that went beyond money: national regulators sometimes blocked new drugs not on scientific grounds, but to protect homegrown pharmaceutical industries from foreign competition. The question that haunted policymakers was whether Europe could build one agency trusted enough to speak for an entire continent. That question took more than seven years of negotiations among EU governments to answer. What emerged in 1995 was not a replacement for national regulators but something new: a decentralised scientific agency designed to harmonise their work, sitting above them without abolishing them. How the EMA was built, what it actually does, and how a cyberattack during the darkest days of the COVID-19 pandemic shook its foundations are the threads this documentary will follow.

• 2. Seven Years in the Making

  ›

  More than seven years of negotiations among EU governments preceded the EMA's founding in 1995. That is a long time to debate the design of a single agency, and it reflects how much was at stake. When the agency finally opened, it replaced two existing bodies: the Committee for Proprietary Medicinal Products and the Committee for Veterinary Medicinal Products. Neither was dissolved outright. Both were reborn as core scientific advisory committees within the new structure, carrying institutional knowledge into the fresh organisation. The founding funding model was deliberately mixed. The European Union contributed, the pharmaceutical industry contributed, and member states provided indirect subsidies. That mix was a signal that the agency was meant to serve multiple interests at once: public health, regulatory efficiency, and industrial competitiveness. The hope was that a single centralised evaluation would make drug approval faster and cheaper for companies while also stripping away the protectionist incentives that had distorted national decisions. London was chosen as the agency's first home. It would remain there until a referendum on a different continent-sized question forced a move that nobody had planned for.

• 3. How the Agency Works

  ›

  Around 4,500 experts drawn from across the European Union form the scientific network the EMA relies on to evaluate medicines. They are not EMA employees. The agency draws on resources from over 40 National Competent Authorities across EU member states, which is what makes it genuinely decentralised in practice. Inside the agency, seven scientific committees divide the territory: one each for human medicines, veterinary medicines, herbal medicines, orphan drugs, paediatrics, advanced therapies, and pharmacovigilance risk assessment. Each has a distinct mandate and was established at a different point in the agency's history. The Secretariat, which runs to approximately 600 staff, is organised into five units covering everything from human medicines development to information technology and administration. The management board sits above the Secretariat and holds real administrative power. It approves budgets and plans and selects the executive director. Its membership reflects the political architecture of the EU: one representative from each of the 27 member states, two from the European Commission, two from the European Parliament, two from patients' organisations, one from doctors' organisations, and one from veterinarians' organisations. A company seeking approval for a genuinely novel medicine submits a single application to the EMA. If the Committee for Medicinal Products for Human Use concludes that quality, safety, and efficacy are sufficiently proven, it issues a positive opinion. That opinion travels to the European Commission, which transforms it into a marketing authorisation valid across the entire EU, as well as in Iceland, Liechtenstein, and Norway. The committee is required by regulation to reach its decisions within 210 days, though the clock stops if the committee needs to ask the applicant for more information.

• 4. The Committees and Their Mandates

  ›

  Orphan diseases are defined precisely in EU law: conditions affecting not more than five in 10,000 persons in the European Union. The Committee on Orphan Medicinal Products has administered the granting of orphan drug status since 2000, giving companies that develop treatments for these rare conditions a recognised pathway and a formal designation from the European Commission. The Paediatric Committee, known as the PDCO, has operated since 2007 under Regulation (EC) No 1901/2006. Its mandate is direct: any application for marketing authorisation of a new medicine must either include data from paediatric studies agreed with the PDCO in advance, or obtain a PDCO waiver, or formally defer those studies. Children cannot be an afterthought. The Committee for Advanced Therapies was established under Regulation (EC) No 1394/2007 specifically to handle gene therapy, somatic cell therapy, and tissue-engineered products. These are medicines that do not fit the classical model of a chemical compound in a tablet. The committee assesses their quality, safety, and efficacy while tracking scientific developments in a field that moves quickly. The Pharmacovigilance Risk Assessment Committee, the seventh and newest of the major committees, came into function in 2012 following implementation of the new EU pharmacovigilance legislation under Directive 2010/84/EU. Its job is to monitor what happens to medicines after they are approved, not just before. The 210-day review clock that governs the human medicines committee compares favourably with the average approval timeline at the U.S. Food and Drug Administration, which stood at around 500 days in 2008.

• 5. Transparency Under Scrutiny

  ›

  Louise Brinth, a Danish physician, went public with a pointed critique of the EMA's review process after the agency issued a rebuttal of an assessment that included her work. Her objections were specific. The experts reviewing data, she noted, remain unnamed and appear bound to secrecy. Meeting minutes are not released to the public. Diverging opinions among committee members are not reported, creating the impression of unanimous agreement where none may exist. In her view, the process was both unscientific and undemocratic. These are serious charges against an agency whose authority rests entirely on public trust in its scientific rigour. The EMA's defenders point to the practical logic of confidentiality: companies submitting commercially sensitive data need assurance that competitors cannot access it through the review process. But Brinth's critique cuts at something deeper, the question of whether regulators can credibly claim to serve the public while keeping the mechanics of their judgements hidden from that same public. The tension has not been resolved. It sits at the centre of any honest account of what the EMA is and what it is not.

• 6. Brexit, Amsterdam, and a London Lease

  ›

  The United Kingdom voted to leave the European Union in 2016, and the EMA had no choice but to find a new home inside EU territory. EU ministers voted on successor cities. The EU's Health Commissioner Vytenis Andriukaitis had framed the preferred destination as one offering easy setup and a guarantee of smooth operations. Nearly every remaining EU member state except the Baltic States and Luxembourg submitted a proposal. The relocation decision was made on the 20th of November 2017 at an EU General Affairs Council meeting, but reaching it required three rounds of voting and, in the end, a drawing of lots. After the first round, Milan held 25 votes, Amsterdam held 20, and Copenhagen held 20. Milan and Amsterdam survived to the second round, with Milan taking twelve votes and Amsterdam nine. The subsequent vote ended in a tie: thirteen votes each. A drawing of lots broke the deadlock in Amsterdam's favour. EMA staff left London in March 2019 for a temporary building in Amsterdam. By January 2020 the permanent building in Amsterdam's Zuidas district was ready. The London chapter was not clean. The EMA attempted to argue that the doctrine of legal "frustration" should release it from its 25-year lease, since EU law required the agency to be located within EU territory. The High Court ruled against this argument, leaving the EMA bound to a lease with much of its original 25-year term still remaining.

• 7. COVID-19 and the Cyberattack

  ›

  In December 2020, while the EMA was in the middle of evaluating the Pfizer-BioNTech COVID-19 vaccine, the agency suffered a cyberattack. Classified regulatory documents were stolen and leaked to journalists, academics, and the public via the dark web. The leaked materials revealed that regulators had internal concerns about low production quality in the mRNA vaccine candidate. They also showed that the EMA had pressed Pfizer and BioNTech to fix those deficiencies before authorisation. The EMA authorised the vaccine on the 21st of December 2020, stating that product quality was "sufficiently consistent and acceptable." Earlier in the pandemic, the agency had joined the Accelerating COVID-19 Therapeutic Interventions and Vaccines partnership, known as ACTIV, hosted by the Foundation for the National Institutes of Health. That collaboration brought together international government agencies and private companies to coordinate a shared research strategy for prioritising and accelerating vaccine and treatment development. The cyberattack raised questions that extended far beyond any single vaccine: how secure are the systems that hold the most sensitive data in pharmaceutical regulation, and what happens when those systems fail at the precise moment global health depends on them. The EMA's leak of internal quality concerns also gave the public a rare, unfiltered view of exactly how much negotiation happens inside an approval process before a product reaches a pharmacy shelf.