The European Medicines Agency's new policy on the proactive disclosure of clinical trial data

The European Medicines Agency (EMA) has, since its creation, launched several initiatives to increase transparency of, or expand access to, information relating to its decision-making concerning medicinal products authorised through the centralised procedure.  On 2 October, the EMA published its latest initiative - its proactive disclosure policy¹ - the product of many months of consultation and discussion with key stakeholders. The final policy provides that the clinical data submitted with an application for marketing authorisation will be available on the EMA's website after the grant of a centralised marketing authorisation (or refusal, or withdrawal of the application). The "clinical data" covers all clinical reports and individual patient data. However, use of the data will be subject to the EMA's terms of use policy, which allows controlled access with the intention that use can only be made of the data for non-commercial purposes.

The policy will enter into force on 1 January 2015 for new applications, and 1 July 2015 for extension applications, and will only apply to clinical study reports contained in applications submitted after those dates. A second phase of the policy initiative will consider the basis upon which patient level data, where provided to the EMA, will be made available to third parties.

The initiative reflects a policy on the disclosure of the information within an authorisation dossier that attempts to balance the interests of the various stakeholders, and considers how the data might be unfairly used by competitors. But the policy is based on the EMA's view that there is also a public interest in the detailed clinical trial data being released and considered by academics and healthcare professionals.

This advisory summarises the key points of the policy and some outstanding questions for industry.

 Context and policy development

The publication of the European Public Assessment Report ("EPAR") under Articles 12 and 13 of Regulation  726/2004/EC sought to provide transparency as to the basis for the decision to grant a marketing authorisation.  In addition,  Regulation 726/2004/EC governing the centralised procedure requires the Transparency Regulation 1049/2001/EC to be applied to documents held by the EMA to enable public access. The EMA has embraced this principle of transparency and access to documents, and in November 2010, published a policy defining how it would respond to requests for access to documents under the Regulation.²Under this reactive policy, the EMA has been releasing documents submitted as part of applications for marketing authorisation to any party that requests them, subject to fairly limited exceptions. These exemptions include a more narrowly interpreted exemption for commercially confidential information which contrasts starkly with the EMA's previous practice of treating all detailed pre-clinical and clinical data as commercially confidential.

More recently, the EMA committed to adopt a policy for the proactive publication of data from clinical trials supporting marketing authorisations following a workshop hosted by the Agency in November 2012.  This precipitated the publication of the first draft of the new policy on proactive access to clinical trial data in 2013, which was the subject of protracted public consultations given the level of interest, with over 1000 comments from nearly 170 stakeholders.

The draft policy originally stated that the data within the marketing authorisation would be made available on the EMA's website at the same time as publication of the EPAR. The draft policy "respects and will not divulge commercially confidential data or information", although it stated that "in general, however, clinical trial data cannot be considered commercially confidential information", and that the interests of public health outweigh considerations of commercial confidentiality.

The draft policy raised questions from industry about the uncontrolled use of clinical trial data, what data within the dossier could be considered as confidential information, and how third parties would be able to use the data. However, the calls from industry for protection of their interests have been met by criticisms from academics and patient groups who claim that the public has the right to access clinical trial data.Researchers in particular have raised a number of concerns that the EPAR "poorly and inconsistently" documents Phase III trials, and that the detailed data are important for doctors and researchers to analyse the data and carry out meta-analyses.

Published Proactive Policy

As a result of these discussions, the EMA adopted its proactive publication policy on 2 October 2014. The EMA acknowledges that the proactive policy has been developed in the absence of any clear legal provision mandating the EMA to proactively publish documents submitted by third parties. As a result, the policy is stated to be a balanced approach, taking into account the view of various stakeholders - "This compromise allows access to clinical data but, at the same time, aims to discourage unfair commercial use of the data."

The main aspects of the policy are as follows:

• After the grant of a marketing authorisation (or refusal, or withdrawal of the application), the clinical data within the application dossier will be available on the EMA's website.
• The data will be available in two forms: (i) available to view by all those who register for the website, and (ii) available to download and re-use by academics and for other non-commercial research purposes.
• All use is subject to the EMA's terms of use, whereby the use of the data must be for non-commercial purposes only. Commercial purposes include using the data to support an application to obtain a marketing authorisation and any extensions or variations thereof for a product anywhere in the world, or selling, trading or supplying the data to a third party that has not agreed to the terms of use.
• Commercially confidential information will be redacted.  Commercially confidential information means any information contained in the clinical reports submitted to the EMA by the applicant that is not in the public domain or publicly available and where disclosure may undermine the legitimate economic interest of the applicant.  Examples of what is, in principle, accepted as commercially confidential information appear in Annex 3 to the policy statement. However, the policy retains the assertion that "in general", data within the application is not commercially confidential.  Where redaction of commercially confidential information is proposed by the applicant, the EMA will conduct a consultation with the applicant, and the EMA will decide if the information should be redacted.
• The terms of use are also for the benefit of any and all applicants or authorisation holders and, accordingly, it is said that each such applicant or authorisation holder may, in its own right, enforce the terms of use directly. The EMA also states that "Courts may require the EMA to disclose the identity of the users who do not comply with the [terms of use] to the marketing authorisation holders/applicants."
• The policy will come into effect for any new applications made after 1 January 2015, and 1 July 2015 for extension of indication and line extension applications relating to centrally authorised medicinal products.  The policy does not apply to clinical data already held by the EMA, or submitted before these dates.
• In the second phase of implementation of the policy, the Agency will review various aspects in relation to personal data (such as patient level data), including finding the most appropriate way to make patient level data available, in compliance with privacy and data protection laws.
• Persons can continue to submit a request for access to documents to the EMA under the current reactive policy.
• The policy is said to be "a useful complementary tool" ahead of the implementation of the new Clinical Trials Regulation 536/2014/EU, which will require publication of clinical study reports for trials commenced from 2016 onwards. Note that under the Clinical Trials Regulation, disclosure will apply to all trials with a study site in the EU, regardless of whether the ultimate authorisation is submitted via the centralised or national procedure. In contrast, the EMA's proactive policy applies to all data within a centralised marketing authorisation application, regardless of where that study was conducted.

There are some outstanding questions for industry as to how the policy will operate and apply in practice. For example, where an application has been withdrawn, perhaps because of some deficiencies in the data, and the applicant is intending to resubmit the data, how will publication of the data be dealt with?  The policy is stated to cover such data, but there is a question as to whether this may undermine a company's strategy for further research and development, and intellectual property interests with regard to the re-submission. In relation to the reactive policy, the EMA has been open to discussion on the confidential nature of information contained in dossiers that are to be resubmitted. It is unclear how this will operate under the proactive policy.

A key question, and one that has been strongly argued by stakeholders on both sides, is what data will be considered as commercially confidential information, and will the EMA be open to further submissions from companies about the redaction of information within the dossier?  Under the current reactive policy, there are significant differences of opinion on what data within the marketing authorisation dossier, other than manufacturing information, should be kept confidential. There appears to be a lack of appreciation of how disclosure can affect the commercial interests of companies, such as (a) the possibility that the disclosed information could be reformatted and used outside the EU, and (b) the effect of disclosure on potential patent rights. However, the proactive policy uses similar wording as under the reactive policy, and it seems that similar differences of opinion will arise.

While the policy applies to centrally assessed products, it is as yet unclear how this policy would apply for products authorised through the decentralised or mutual recognition procedures, and how (if at all) it would be implemented by national competent authorities. It would appear that several Member State authorities believe they are constrained by national law and precedent from releasing the contents of dossiers submitted by pharmaceutical companies to obtain marketing authorisations.

The uncertainties surrounding how the policy will be applied may affect the planning for EU regulatory filings by companies. For example, companies will, no doubt, wish to re-assess the extent to which certain categories of information that they consider commercially confidential but have hitherto referenced in the documents underlying their applications for marketing authorisation, including clinical study reports, really need to appear in the documents submitted to the authorities. Even if such information might well be redacted on request, it is simpler and less burdensome in terms both of the diversion of resources to undertake scrutiny and negotiation, and the potential consequences of errors in redaction occurring, to exclude such information from the outset, if possible.

This consideration also applies to patient level data. Unlike the FDA, the EMA does not ordinarily ask for "base data" (although it can request such data for re-analysis, where necessary), as it is not organised in a way that gives it the capability to conduct primary analysis. The EMA has indicated that it will not ask companies to submit patient level data for a trial solely for the purpose of making it public for others to conduct such primary analyses, but that it will be consulting with stakeholders on when the EMA should request such data and will publish a "Reflection Paper" by the end of 2014. This will include consideration of what standard measures should be adopted to avoid identification of patients, so as to protect privacy rights.

Conclusion

The EMA's new policy has clearly taken some of industry's concerns into account, by recognising that unfair commercial use could be made of the data within an application dossier.  The fact that the majority of applications for access to data seem to have been from pharmaceutical companies, or those acting on their behalf, emphasises the need for proper protection. This form of controlled access seems, in principle, a fairer approach to disclosing the information within applications for academic purposes, while protecting the rights of the authorisation holders. However, questions remain about how the policy will be applied, how it can effectively be enforced, and whether similar questions will arise as have done so under the current reactive policy, which itself has given rise to litigation in the General Court.