Will EMA Rules Calm the Disclosure Debate?

It’s a long-running battle, and the fat lady still hasn’t sung, so it ain’t over yet. But Europe moved one step closer in March to quenching the conflagration over transparency and clinical trial data. A weighty document running to nearly a hundred pages appeared from the European Medicines Agency (EMA), aiming to squeeze much of the oxygen out of the debate that has been raging for years about how much information companies should disclose about their products and their trials of them. (view the report here).

The guidance for the publication of clinical data explains to everyone who wants to know just how the agency is going to operate its new system of mandatory publication of data, and just what requirements are to be imposed on industry on submission of clinical data for publication.

The new policy entered has been in force since Jan. 1, 2015, in that it covers the clinical reports contained in all marketing-authorization applications submitted from then onwards, but the first actual reports to appear are-because of the time-lag in processing applications-currently likely to appear only from this coming September.

One of the hot spots in the document is the guidance on how to anonymize clinical reports for publication, so as to prevent any re-identification of trial participants. Bowing to the inevitable, given the wide range of methods available, EMA recognizes that no one method can be imposed, and permits some freedom of approach. But it gives recommendations to companies on how to best balance data utility for researchers with a minimal risk of re-identification, and companies will need to provide a report explaining their approach. That report will in turn be reviewed and published by EMA.

But the real heat will be emitted by the document’s approach to commercially confidential information (CCI). Fierce arguments over the very concept of CCI have put drug companies (and regulators) at odds with healthcare campaigners in Europe, many of whom flatly refuse to recognize the merits of any methodology for identification and redaction of what companies perceive as sensitive material. In a scorched earth approach, the most vocal campaigners reject any such rights, and argue that the only ownership of such data lies with the trial participants whose involvement has generated the data.

EMA has trodden a middle path of reason on the issue. It knows how inflammable this discussion is, and how determined industry critics are to pursue the fullest access to information. So its guidance makes clear that we are now very much into open season for data hunters. 

“The vast majority of the information contained in clinical reports is not considered CCI,” it says. But “the vast majority” is not the totality, and EMA goes on to spell out its exceptions. It defines CCI as “any information contained in the clinical reports submitted to EMA by the applicant which is not in the public domain or publicly available and where disclosure may undermine the legitimate economic interest of the applicant.” There are, it says, “limited circumstances in which clinical reports might contain CCI,” and the data classified as CCI may indeed be redacted (or, to put it plainly, lest this column should be accused of unjustified redaction, simply blacked out). 

With masterful understatement, the guidance document remarks: “It is anticipated that the preparation and publication of the documents will raise some practical questions, such as on how to apply the aforementioned redaction principles, and on the presentation and justification of the proposed redactions.”

Where redaction takes place, companies will need to provide justification to EMA. The guidance clarifies which type of data EMA would typically refuse as being CCI and how the redaction of such data will be handled. Its broad framework starts from the principle that it will not accept redaction of any information in the public domain or that has no innovative features. It will also frown upon attempts to redact quality, non-clinical and clinical data which it believes to be necessary for the understanding of the rest of the clinical report, thus making its disclosure a matter of public interest. And the agency helpfully supplies some real-life examples of attempts that have already been rejected to justify redaction of material in reports.

So don’t waste any time with the following arguments: “Unpublished data-These study results have not been published in any peered-reviewed [sic] publication.” “Company confidential information-Disclosure of these elements will harm [the company]’s commercial interests because it may enable third-party access to business-critical information.” “This information can be interpreted out of context. Such interpretation could lead to a misleading image of the safety profile of the product.”

Clinical reports will be published at the conclusion of regulatory decision-making in the centralized marketing authorization procedures-irrespective of whether the decision on a marketing authorization application is positive or not. The reports-with anonymization and agreed redaction-will be published by EMA on its corporate website, within 60 days of the final decision for marketing authorization applications, line extension applications, and extension of indication applications. Where an application is withdrawn, the publication of the redacted/anonymized clinical reports will take place within 150 days after the receipt of the withdrawal

letter.

The hope within EMA is that its approach will damp down the debate and allow attention to return to the content of reports, rather than the processes for accessing the data they contain. The agency has been engaged in extensive consultation with all parties concerned throughout 2015, and is cautiously optimistic that its patient negotiation has permitted a well-balanced set of requirements to emerge that can satisfy all sides.

Now that D-Day is approaching for publishing the first reports under this new dispensation, the agency is planning to hold talks with the companies at the front of the firing line-those for which the decision-making process has been finalized since the policy entered into force, and whose reports will constitute the first wave of publication. It is also planning a webinar in the late spring for companies to raise outstanding practical questions. 

Further down the track, still more fiery exchanges can still be expected, because the scheme now starting to deliver real reports is only the first phase of the agency’s CT transparency bid. While this first phase deals only with publication of clinical reports, a second phase will deal with the still-more sensitive issue of publishing individual patient data. But this is still on the back burner; the agency says that although it is committed to moving in that direction, no dates have been set, and it “will be implemented at a later stage.”

Ultimately, the success-or failure-of this attempt to calm spirits on access to trial data may depend more on emotion than on logic. There are implacable views among the most earnest advocates of access to data, who recognize no claims for exemption and who, in the style of Wikileaks, demand full release of everything all the time. A lengthy and carefully-reasoned list of guidelines on the right and wrong way of doing things may not prove a sufficient response to that type of argument.

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