1. PTC

PTC Therapeutics International Ltd (PTC) is the maker of Translarna, a drug used to treat Duchenne muscular dystrophy. In February 2014, the European Medicines Agency (EMA) granted PTC a conditional marketing authorisation (MA) for Translarna.

In July 2015, the EMA received a request pursuant to Regulation 1049/2001Footnote ³ seeking access to a clinical study report that PTC had submitted as part of its MA application. The report was an efficacy and safety study of the active ingredient of Translarna and was the main clinical study conducted prior to granting the conditional MA. In accordance with its obligation to consult PTC under Article 4(4) of the Regulation, the EMA invited PTC to propose redactions over commercially confidential information.Footnote ⁴

PTC requested that the EMA treat the report as confidential in its entirety.Footnote ⁵ The EMA rejected this request. In subsequent correspondence, PTC refused to recommend redactions and maintained its assertion of confidentiality over the whole report. On 25 November 2015, the EMA adopted a decision granting access to the requesting third party, subject to certain redactions (“the PTC decision”).Footnote ⁶ The redactions included protocol design discussions, batch numbers, materials and equipment, exploratory assays, description of the drug concentration measurement method and information that could identify patients.Footnote ⁷

PTC brought an application to the General Court (GC) seeking annulment of the PTC decision and arguing that the entire report should remain confidential. The GC dismissed all of PTC’s pleas and upheld the EMA’s decision.

2. MSD

MSD Animal Health Innovation GmbH and Intervet International BV (together “Merck”) are part of the Merck group. In February 2014, they were granted MA for Bravecto, an animal medicine used to treat tick and flea infestations in dogs.

In August 2015, the EMA notified Merck that it had received a request under Regulation 1049/2001 for access to five toxicology test reports in the Bravecto file.Footnote ⁸ The EMA stated that it planned to disclose three of the five reports and invited Merck to submit proposed redactions. Although initially Merck claimed that each report was protected by a general presumption of confidentiality, after communications with the EMA it proposed, strictly in the alternative, specific redactions.Footnote ⁹ The EMA accepted some proposed redactions but rejected most and decided to disclose the documents in November 2015.Footnote ¹⁰

Merck brought an action for annulment of the EMA’s decision to disclose the reports, advancing substantially similar pleas to those in PTC. The GC dismissed the application in similar terms to its PTC judgment.

1. Whether the reports were protected by a general presumption of confidentiality

General presumptions of confidentiality (GPs) are a judicially created device allowing institutions to presume certain categories of documents fall within an exception to access, without specifically and individually examining whether the exception actually applies. The scope of GPs and their relationship to the protection of commercial interests exception is unclear. GPs have normally been applied to protect ongoing administrative and judicial processes under the “space to think” exception. Whether they can be used to protect commercial interests has never been decided. Moreover, the criteria for recognising a GP over a new category are unclear. The difficulty of identifying these criteria and whether it is appropriate to recognise GPs on the basis of commercial confidentiality are demonstrated by the fact that the GC, Advocate General (AG) and ECJ all arrived at different conclusions.

The appellants argued that there was a GP that both the first indent of Article 4(2) (protection of commercial interests) and Article 4(3) (space to think) of Regulation 1049/2001 applied to clinical study reports. The GC, rejecting these pleas, held that the criteria for recognising a GP were not met. These criteria were that: first, access was “incompatible with the proper conduct” of the relevant procedure or there was “a risk that those procedures could be undermined”;Footnote ¹² second, documents relate to “ongoing administrative or judicial proceedings”;Footnote ¹³ and finally, the legislation relating to the procedure contain specific rules allowing a GP to be inferred.Footnote ¹⁴ The second criterion would close off the possibility that GPs of commercial confidentiality may ever apply.

Interestingly, AG Hogan disagreed with the GC’s analysis, proposing a far less stringent test for recognising a GP over a new category of documents. He cited a number of cases in which a GP was recognised where the procedure was neither ongoing nor were there specific rules on access.Footnote ¹⁵ In his view, the correct test as the ECJ enunciated in Clientearth Footnote ¹⁶ was simply whether it was “reasonably foreseeable that disclosure of the type of document would be liable to actually undermine the interest protected by the exception …”.Footnote ¹⁷ Applying that test, he concluded that the reports benefitted from a GP.Footnote ¹⁸

Accordingly, the GC and AG Hogan were at odds primarily on two matters: first, whether GPs were limited to ongoing administrative or judicial proceedings or whether they could apply to any exception to access (including commercial confidentiality); and second, whether GPs must have a basis in the procedure’s underlying legislation.

The ECJ, however, sidestepped the issue, ruling that it was not necessary to decide whether the document was covered by a GP. This was because an agency “may always carry out a specific examination of the documents” and the application of a GP is “merely an option”.Footnote ¹⁹ The ECJ took the view that the GC’s statement of criteria for recognising a GP was included for “completeness”.Footnote ²⁰

2. Conditional MAs, future MAs and final decisions

In both cases, the appellants sought to rely on the “space to think” exception by arguing for an expansive understanding of the phrase “a matter where the decision has not been taken” in Article 4(3) of Regulation 1049/2001. The appellants in MSD argued that as they intended to submit future applications for the same substance, disclosure would seriously undermine the EMA’s decision-making process in respect of those future applications.Footnote ²¹ A similar argument was made in PTC, where PTC relied on the fact that it had only been granted a conditional MA and that disclosure would undermine annual renewal applications or a future application for a “full” MA.Footnote ²²

The ECJ rejected PTC’s argument concerning renewal applications, but only because it had not been advanced at trial.Footnote ²³ However, the court clearly takes a restrictive view as to what qualifies as a “matter where the decision has not been taken”. In PTC, it rejected the contention that a decision-making process was pending throughout the period of data exclusivity.Footnote ²⁴ In both cases, it held that the procedure in question was closed at the date the request was made.Footnote ²⁵ The possibility of data being reused in future applications was “based on a hypothetical premise” and accordingly rejected.Footnote ²⁶ Unlike other aspects of these cases, these points were uncontroversial, with the GC, ECJ and AG reaching similar conclusions.

3. Misuse by competitors and the burden of proving commercial confidentiality

In both cases, the appellants essentially argued that the reports should be considered commercially confidential in their entirety.Footnote ²⁷ A peculiar aspect of PTC was that it concerned an orphan medicinal product. Orphan products’ makers receive market exclusivity, but they may lose this exclusivity if a similar product is shown to be safer, more effective or otherwise clinically superior.Footnote ²⁸ It is therefore possible that a competitor could end PTC’s market exclusivity by drawing on the report to design a superior product.Footnote ²⁹ Furthermore, both appellants argued that there was a prospect that competitors could use the data to obtain MAs outside of the EU.Footnote ³⁰

The ECJ rejected these arguments, stating that they were tantamount to invoking a GP over the whole report.Footnote ³¹ It held that risk of misuse was not itself a ground for considering that information commercially confidential,Footnote ³² and that the appellants could establish that disclosure of “certain passages, specifically identified” would harm their commercial interests if they could demonstrate a “concrete and reasonably foreseeable risk that certain unpublished data … which … [is not] common knowledge … could be used in one or more third States by a competitor” to obtain MA.Footnote ³³

It moreover held that it was for the appellants, both before the EMA and GC, to “specifically and precisely identify … which of those passages, if disclosed would harm its commercial interests”.Footnote ³⁴ The risk to commercial interests must be reasonably foreseeable and not purely hypothetical.Footnote ³⁵ The Court went on to give more detailed guidance as to the evidentiary burden of proving commercial confidentiality. It held that MA holders must adduce “additional details concerning the nature, purpose and scope of the data … enabling the Courts … to understand how the disclosure of those data would be likely concretely and reasonably foreseeably to undermine the commercial interests …”.Footnote ³⁶

Furthermore, the appellants argued that the GC and EMA had erroneously considered that the EMA had discretionary power when assessing the confidential nature of commercial information.Footnote ³⁷ The Court rejected this argument, ruling that instead the EMA carried out a specific and individual examination to determine which data fell within the exception.Footnote ³⁸

1. Continued haziness around GPs

Clientearth itself appears to reject that the existence of specific rules limiting access is a criterion when determining whether a GP applies.Footnote ³⁹ Therefore, although the ECJ in PTC and MSD did not overrule the GC on this point, it appears that the GC was incorrect.

However, what remains unclear is whether documents can benefit from a GP on the basis that disclosure would ordinarily undermine commercial interests. In the context of the European Chemicals Agency (ECHA), there was already conflicting GC jurisprudence on this point.Footnote ⁴⁰

AG Hogan relied on the principles set out at Paragraph 80 of Clientearth to propose a broad test for identifying GPs that could apply to commercially confidential categories of document. However, Clientearth was not concerned with the protection of commercial interests. Rather, it related to an ongoing procedure of an EU institution, a fact not adverted to by AG Hogan. Thus, the ECJ did not have occasion in that case to consider whether a GP could protect commercially sensitive documents.

To muddy matters further, in Clientearth the ECJ specifically stated that all cases in which the court had recognised a GP “were clearly defined by the fact that they all belonged to a file relating to ongoing administrative or judicial proceedings”,Footnote ⁴¹ something also not mentioned by AG Hogan. This point bears acknowledging and clarifying, not least because the statement in Clientearth appears in the paragraph directly after the test adopted by the AG. Accordingly, whether a GP can protect commercially confidential documents is still an open question. Unfortunately, in PTC and MSD, the Court declined to clarify that question, and there remains considerable haziness around the relationship between protection of commercial interests and GPs.

If the Court had adopted the AG’s proposed test, it likely would have received criticism. The practical implication of a GP is that it places the burden of proof on those seeking information, who must prove an overriding public interest in disclosure.Footnote ⁴² However, proving an overriding public interest in documents one has not had sight of has been described as an “impossibly high” hurdle.Footnote ⁴³ GPs have been criticised for eradicating access to documents’ core principles of individual examination and widest possible access without a clear basis in legislation, creating “access free zones” within institutions.Footnote ⁴⁴ The broad test proposed by AG Hogan might greatly expand the range of circumstances in which new GPs could be recognised beyond the current narrower number of categories. This would be problematic from the perspective of both the Regulation and the Treaties. As Curtin and Leino write, “The general presumption is of openness in line with the Treaty. The Treaty recognises no general presumptions of secrecy …”.Footnote ⁴⁵

2. Misuse by competitors and the burden of proving commercial confidentiality

These cases highlight tensions between intellectual property rights and access to documents, which are not easily addressed.Footnote ⁴⁶ In both PTC and MSD, those seeking access were competitor companies. Although access to scientific information pursues democratic aims by allowing citizens to scrutinise the data underlying agency authorisations, nothing prevents competitors seeking access to that same data in hopes of gaining commercial advantages. In 2017, half of all access requests to the ECHA came from industry.Footnote ⁴⁷

While the appellants in both cases sought to keep the entirety of the relevant reports confidential, interestingly they adopted contrasting legal strategies before both the EMA and the Courts. On the one hand, PTC refused to engage in the EMA’s consultation processFootnote ⁴⁸ and simply asserted that the entire report was confidential, prompting the EMA to substitute its own redactions.Footnote ⁴⁹ On the other hand, Merck did engage with the EMA, who accepted some of their proposed redactions but rejected many others.Footnote ⁵⁰

It is difficult to know whether Merck’s strategy brought any greater success before the EMA than PTC’s strategy would have. Nonetheless, in both cases, the Court declined to overturn any part of the EMA’s decision. It appears from PTC and MSD that the Court requires that applicants specifically explain how disclosure of individual passages of a report could reasonably and foreseeably undermine its commercial interests. From the point of view of legal strategy, authorisation holders would therefore be advised to be as specific as possible in explaining how disclosure of individual passages could undermine their interests. A blanket assertion of confidentiality over the whole report backed up only by generalised claims of jeopardy to commercial interests is unlikely to find favour with the EU judiciary.

3. The consultation process and MA holder control

While the appellants sought to impugn the GC’s judgment for permitting the EMA to exercise discretion as to whether information fell within the commercial interests exception, the ECJ countered that in fact the EMA had conducted a specific and individual examination of all data to determine whether an exception applied in each case.

Concern had previously been raised that the practical operation of the consultation process might result in MA holders exercising control over the disclosure process, contrary to Regulation 1049/2001.Footnote ⁵¹ For example, if in practice the EMA accepted MA holders’ proposed redactions without itself evaluating whether the proposed redactions corresponded to the expectation under the legislation, this would grant MA holders an unacceptable degree of input into the disclosure process.Footnote ⁵² However, the facts of these cases demonstrate that in fact the agency itself specifically and individually examines requested information to determine whether an exception applies. Unless a GP applies, the agency must always conduct such examinations.