Marta Giner Asins and Arnaud Sanz, Norton Rose Fulbright
This is an extract from the first edition of The Guide to Life Sciences published by Global Competition Review. The whole publication is available here.
Denigration cases in the life sciences area have been rare in the EU as well as worldwide, and in most cases the denigration behaviour is combined with other infringements, such as abuse of administrative procedures or product hopping (such as in the Suboxone case). It seems that to constitute an infringement by itself, denigration has to meet a strict standard that is rare in practice.
However, since the early 2010s, the French Competition Authority (FCA) has undertaken a number of cases that have progressively widened the definition of denigration to the point of becoming a major compliance burden for companies. This trend is manifest in various decisional aspects:
- as mentioned above, in early cases, denigration was a ‘support’ measure included as part of a wider anti-generic strategy (Subutex), whereas it was considered as a stand-alone practice in other decisions (Plavix);
- the definition of denigration itself has been extended to include expressly denigrating messages (Janssen) as well as ‘implied’ denigration (i.e., focusingon communications with healthcare professionals (HCPs) concerning the differences between the originator drug and generics, which could have a misleading effect); HCPs may interpret these differences as meaning the originator drug is superior to the generics. For example, in Plavix, the originator’s messages highlighted the fact that the generics used a different component; although not expressly mentioned, the FCA considered that the focus on this difference suggested the existence of a qualitative difference. This trend takes the focus away from the truthfulness of declarations as even truthful statements may be considered misleading;
- denigration was traditionally analysed under the prohibition of abuse of dominant position (AstraZeneca) but the FCA has also approached it as an anticompetitive agreement and even as an abuse of collective dominant position (Roche Novartis); it appears that if the FCA identifies a behaviour as being questionable, it will adapt its approach to catch it, particularly if it encounters difficulties in proving the existence of a dominant position. This implies that all pharma companies should carefully monitor their communications, and not only those holding a dominant position; and
- particularly noteworthy, denigration concerns not only commercial communications with HCPs but also contact with the administrative authorities. Whereas traditionally, lobbying towards administrative authorities was rarely considered as abusive, in the most recent French decisions the same strict standard was applied to communications with both HCPs and administrative authorities (Roche Novartis); communication should not only be truthful, but also objective, measured and not devoid of legal basis. This strict standard departs from the approach taken in other jurisdictions, such as the US in analysis of sham citizen petitions.
It is important to underline a key trend, considering its impact in terms of compliance: the narrow approach defined by the FCA seems to apply to all communications with authorities and not only to cases in which the parties used allegedly denigrating messages. The French cases have widened not only the definition of denigration, but also, more generally, the standard applicable to relations with authorities, by referring to cases such as AstraZeneca  in which there was no denigration and extending their scope. The French Supreme Court has clearly indicated that the simple fact of intervening in the decision process of a public authority may be considered as an abuse.
The French focus on denigration is not limited to life sciences: the FCA has adopted denigration decisions in other areas, such as electricity, telecommunication and phone services, and the French courts have developed a strong corpus of judgments in the unfair competition area, which have provided a supporting background for the FCA’s wide definition of denigration.
However, it appears that the approach adopted in the life sciences sector is particularly strict: in all the pharma decisions, the FCA has mentioned a recurring element of context (i.e., the fact that HCPs present a particularly high aversion to risk (Janssen, Plavix), which, according to the FCA, also exists for health authorities (Roche Novartis)). In the FCA’s view, this seems to justify the application of a specifically high standard of care in pharma companies’ communication practices, more so than in other industries. In addition, the FCA considers that public health authorities depend, to a large extent, on the information communicated by HCPs, as well as by pharma companies, to accomplish their missions, thus imposing additional responsibility on these companies.
The FCA’s increasingly wide definition of denigration in the life sciences area seems to have opened the door to cases initiated by other national competition authorities, such as the Falck case in Denmark and the Roche Novartis case in Italy, which targeted facts similar to those sanctioned in the French Roche Novartis decision. In this case, the Italian Supreme Administrative Court sent a request for a preliminary ruling to the European Court of Justice (ECJ), which took the view that:
the dissemination, in a context of scientific uncertainty, to the [European Medicines Agency], healthcare professionals and the general public of misleading information relating to adverse reactions resulting from the use of one of those products for the treatment of diseases not covered by the [marketing authorisation] for that product, with a view to reducing the competitive pressure resulting from such use on the use of the other medicinal product, constitutes a restriction of competition ‘by object’ for the purposes of that provision.
Although this was a preliminary ruling, and therefore no factual analysis was carried out, the European Commission may have interpreted it as encouragement to open two recent denigration investigations:
- the Teva Copaxone investigation, which targeted a potential abuse of patent filing procedures, paired with a denigrating campaign; and
- the Vifor investigation, which appears to target denigration as a stand-alone practice.
The coming months and years will be enlightening in terms of the definition of denigration, and key elements will likely result from the above Commission investigations as well as the appeal in the French Roche Novartis case, in which the Paris Court of Appeal may reverse the FCA decision despite the strict approach recently adopted by the French Supreme Court in Janssen.
This chapter focuses on:
- what constitutes abusive denigration in communications with HCPs;
- the standard applied to relations with public authorities; and
- a few practical recommendations for pharma companies.
Abusive denigration in communications with HCPs
The most detailed presentation of what constitutes product denigration in communications with HCPs can be found in the French Roche Novartis case. Even though this decision is under appeal, it is the most recent benchmark and it summarises the principles contained in previous decisions.
The factual framework in this case was more complex than a straightforward anti-generic strategy such as those targeted in Subutex, Plavix and Janssen. In Roche Novartis, Genentech had developed two products: Avastin, marketed in the oncology area but found to have positive effects for age-related macular degeneration (AMD), although it had no marketing authorisation (MA) for the treatment of AMD, and Lucentis, used for ophthalmic diseases such as AMD. For various reasons, Lucentis was sold at a much higher price than Avastin, and due to this price difference, certain doctors used Avastin, off-label, in the ophthalmic area.
In the context of the scientific community being uncertain of the potential risks of this off-label use, Novartis developed a communication campaign aimed at reducing this usage, directed at HCPs, particularly key opinion leaders (KOLs). To conclude that this communication campaign was denigrating and abusive, the FCA focused on certain contextual elements, as well as the content and anticipated effects of the communication.
The following paragraphs: summarise these contextual elements as they provide a useful benchmark for the context in which a company needs to be particularly careful about its communications; and analyse the content of messages to identify what may constitute an abusive statement.
Main contextual elements
The Roche Novartis decision (as well as previous FCA rulings) mentions certain contextual elements that were not, strictly speaking, part of the infringement itself. However, they illustrate situations in which an abuse is more likely to be identified.
Commercial purpose of the campaign
The FCA analysed a large amount of internal documents that, in its opinion, showed that Novartis saw the off-label use of Avastin as a commercial threat and developed a self-defence strategy, in the same way that, in other cases, originator companies have tried to protect themselves from generic penetration. The FCA highlighted that Novartis’ campaign differed from mere criticism in that it was implemented by an economic operator aiming at obtaining a competitive advantage.
The commercial objective of the communication is only a preliminary element and not sufficient in itself: a dominant undertaking has, in principle, the right to try to preserve its position on the market using competition on the merits (e.g., if it bases its communication on the objective advantages of its product, without comparative or denigrating elements). But the FCA considered that the existence of a commercial purpose excludes any purely scientific or cautionary explanation provided by Novartis.
This reasoning is consistent with other FCA decisions, which all refer to the concept of competition on the merits. Internal documents presenting a global defence strategy are not problematic in themselves; however, they may be used to prove that all adopted measures had a commercial objective, and, therefore, that the communications could not be justified by a scientific reason.
Therefore, it is important for pharma companies to monitor the way in which defence strategies and communication campaigns are presented internally and to avoid any exaggerated language.
HCPs’ fear of risk
In all pharma product denigration decisions to date, the FCA has referred to HCPs’ high fear of risk. Because of this particularity of the life sciences sector, pharma companies must be particularly cautious in their communications.
This general statement appears questionable, particularly in the Roche Novartis case, in which the whole problem resulted from doctors’ practice of using Avastin off-label, which did not particularly reflect a high aversion to risk. In addition, it is difficult to understand why pharma companies are not considered to share the same aversion to risk, which may lead them to express legitimate concerns not incompatible with commercial purposes. However, despite these weaknesses, this element has consistently been retained by the FCA and other authorities in their decisions, and compliance policies need to take it into account.
Link between denigration and dominance
Although the FCA consistently states that denigration may only be found abusive if it is related to a company’s dominant position, the demonstration of this link is rather weak and seems to be presumed. The FCA takes the view that if a dominant company implements a denigration campaign, this will necessarily have an impact on the market due to the reputation and privileged relationship of the company with stakeholders. From a practical point of view, companies should be cautious of how they present themselves in certain communications, and should avoid describing themselves as enjoying a unique or privileged position, or reputation, in the market.
Global and structured nature of the campaign
All FCA denigration decisions emphasise the fact that the sanctioned companies had not implemented isolated communications but had carried out wide and structured communication campaigns. A global communication campaign does not constitute an infringement if its content is not denigrating, but the wide nature of the campaign seems to be taken into account as an aggravating element supporting a violation.
It is also interesting to note that the analysis of communication campaigns has evolved: in Subutex, the FCA made a relatively clear distinction between the measures that were abusive and those that were not (such as the switch to direct sales that accompanied the communication efforts). In Janssen and Roche Novartis, this distinction is less clear or sometimes non-existent. In the latter two cases, the FCA mixes denigrating messages and legitimate measures, blurring the line between the two, and making it difficult to use these decisions as practical guidance.
The timing element is frequently taken into account to demonstrate the anticompetitive nature of conduct, and FCA denigration cases are no exception. In Subutex, Janssen and Plavix, the FCA noted that the communication campaigns were implemented just before the arrival of generics. This timing element is also taken into account in the Roche Novartis decision, although it is not an anti-generic case: the FCA considers that any threat may trigger anticompetitive conduct, and it paid particular attention to communications launched when the French government announced that it would study the possibility of allowing the off-label use of Avastin.
Legitimacy to communicate
In Roche Novartis, communications with HCPs were essentially implemented by Novartis (contrary to communications with the authorities, which, according to the FCA, were implemented by both Roche and Novartis). The decision points to the fact that Novartis was not the MA holder, and therefore it was not its role to carry out pharmacovigilance tasks, which were the sole responsibility of Roche. This element, which is interpreted as proof that Novartis had no reason to communicate with HCPs other than to protect its position, is also mentioned by the ECJ in its preliminary ruling in the parallel Italian case. Therefore, before implementing a communication campaign, a company should reflect on its role and whether it is entitled to communicate.
It is interesting to note that the ECJ seems to consider that scientific uncertainty, which was referred to by the defendants as justifying communication actions, is in fact an aggravating circumstance. The FCA considers that, in an uncertain context, pharma companies need to be particularly careful, thus confirming the reinforced standard applicable to these companies.
The general principle, as the ECJ pointed out, seems quite simple: companies should avoid misleading messages. However, the FCA applies a very strict standard to the application of this principle.
The first step of the analysis is to check whether statements are objective or whether they are based on unchecked assertions. In Roche Novartis, the FCA maintained that available scientific elements were insufficient to identify a risk resulting from the off-label use of Avastin.
This may be interpreted as meaning that a factually correct statement could not be considered as abusive, but that is not the case. In addition, the message should not be capable of raising doubt, particularly by focusing on differences, even if these are real:
the Authority considers that, if it is perfectly permissible for a pharma company to underline the objective qualities of its product, the fact of underlining not only its qualities, but also certain differences, which in the context of the language used and the conditions in which it is received, can only be understood as essential differences, capable of raising an objective doubt on the qualities of competing generic products, or on the risks associated with substitution, can evidence the will to mislead the practitioner.
According to the FCA, Novartis had tried to suggest a link between the differences in the products and the potential risks, a link that had not been scientifically demonstrated.
This focus on differences was also considered problematic in two earlier cases:
- in Janssen, a table comparing certain objective elements, such as the quantities of active pharmaceutical ingredient in the different generic products, was considered to be capable of raising doubts about the effectiveness of the generics. Also, the fact that specific warnings had been introduced in the generics registry by health authorities, and insisting on the fact that it was the first time that these warnings were introduced for a generic in France, was considered denigrating; and
- in Plavix, the references to the use of a different excipient by the generics was considered as establishing a link between two independent elements: on one hand, qualitative and quantitative differences between Janssen’s product and generic products; and on the other hand, supposed consequences for patient security arising from these differences. This link allegedly aimed at raising doubts as to safety issues related to generic substitution.
One of the main differences between these two cases is that in Janssen the bioequivalence seemed to be expressly challenged (‘generics unlike the others’), whereas in Plavix, the standard was pushed one step further, by considering that even an implied challenge of bioequivalence was abusive. This raises particular difficulties for compliance in practice, as sales teams cannot refer to objective differences even if they believe them to be legitimate.
The use of any exaggeration or aggressive language is also proscribed: the FCA’s Roche Novartis decision underlines that all messages need not only be objective but also measured:
the presentation by Novartis in its communication of the potential causes of the risks associated with the use of Avastin, opposed to the certainty about the safety of Lucentis, was not expressed with sufficient measure, given the scientific context in which the communication occurred.
This notion of ‘measured’ communication, which is key in the FCA reasoning, goes far beyond the truthfulness of information and may raise interpretation difficulties in practice, considering it is not entirely devoid of subjectivity.
The communication of incomplete or partial information is also identified as a denigrating characteristic. For example, in Roche Novartis, the companies made a broad reference to a change of summary of product characteristics when the authorities had not changed the ‘adverse reactions’ section, but had only issued ‘special warnings and precautions for use’. According to the FCA, this proved that Novartis had selected the information that served its arguments.
A very important factor taken into account by the FCA in the identification of denigrating messages is the fact that they focus on patient risk and HCPs’ professional and criminal liability. The Roche Novartis decision provides numerous examples of this: for example, a warning included in a Q&A document for use by the sales force, mentioning that ‘the doctor may be found liable from a civil, criminal, administrative and disciplinary point of view’.
Certain decisions have also condemned references to the geographical origin of generics when linked to a fear message, particularly where potential issues in the manufacturing conditions were implied (e.g., sources of supply in countries outside the European Union that were suggested as failing to offer quality guarantees, ‘such as Asian countries’).
As regards communications referring to warnings or recommendations by authorities, the FCA has sanctioned communications that went beyond the strict scope of these.
These different factors result in a wide definition of denigration, which contrasts with the approach taken in the US Suboxone case, in which Reckitt Benckiser was accused of providing ‘false’ information to HCPs. The complex analysis applied by the French FCA goes far beyond the mere truthfulness of the disseminated information.
Real or expected effects of communications
In all decisions, the FCA analyses the potential effects of communication campaigns, which may seem surprising given that denigration is considered to be a by-object infringement, meaning that the demonstration of effects is not strictly necessary. However, the FCA appears to review the possible or expected effects as part of the by-object definition and seeks to prove that the disseminated messages were capable of having an effect. In this respect, several elements are taken into account, such as the wide nature of the campaign or the fact that it was focused on strategic HCPs, such as KOLs.
However, this analysis of effects remains cursory, as reflected in the Roche Novartis decision; for example, the FCA mentions the impact of the actions on competing products’ sales volumes although this analysis is limited to Novartis’ internal documents referring to successful communication actions (‘situation solved’ in a number of hospitals) – there is no economic analysis of the real impact on sales values. Similarly, a market share analysis is only based on Novartis’ internal estimates. HCP statements against the use of Avastin are considered as proof that Novartis’ communication measures were effective, even if no direct link is demonstrated, only presumed. The FCA does not seem to take into account the possibility that HCPs may have expressed their own convictions, and disregards HCPs’ statements declaring that they were not influenced by Novartis’ communication.
Relations with authorities
The standard used to identify denigrating messages in relations with authorities has also seen an important evolution.
Initial approach: pre-MA lobbying is rarely abusive
The EU AstraZeneca case is not strictly considered as a denigration case because it did not feature a communication campaign as such. However, it is a useful reference as it analysed the deliberate provision of misleading information to public authorities to obtain extended protection for intellectual property rights. It contains two elements that are of particular interest for subsequent denigration cases:
- the fact that providing misleading information to authorities may constitute an abuse, which is the principle subsequently developed in denigration cases; and
- the decision insists on the fact that the misleading nature of the information must be analysed in concreto. In that respect, it refers to the importance of analysing the scope of the targeted authorities’ mission and powers: ‘the limited discretion of public authorities or the absence of any obligation on their part to verify the accuracy or veracity of the information provided may be relevant factors to be taken into consideration for the purposes of determining whether the practice in question is liable to raise regulatory obstacles to competition’.
Therefore, this decision seems to suggest a different standard when the targeted authorities are in charge of checking the information communicated to them. This may explain the fact that lobbying activities have rarely in the past been considered as constituting a competition infringement as such.
This approach was also initially followed by the FCA in the Janssen interim measures decision, which established a clear distinction between the pre-MA lobbying activities and the communication campaign targeting HCPs:
these contacts have been made before the product was marketed, during the course of the regulatory phase with the [French Agency for the Safety of Health Products], which is a public authority and has the competence and expertise necessary to evaluate the bioequivalence of a generic, and the risks that may result from the substitution of the originator drug by a generic. Therefore, the [FCA] is not competent to assess the terms of implementation by this health product safety authority, of its public power prerogatives.
This logic is in line with US decisions concerning the abuse of citizen petitions. US law allows individuals to express to the Federal Drugs Agency concerns about the safety of a product any time before, or after, its market entry: on this basis, any person or entity, including a pharmaceutical company, may file a citizen petition expressing concerns about the approval of a generic. It has been frequently highlighted that these citizen petitions have the potential to delay or impede competition (e.g., if a company implements a strategy of filing baseless petitions with the intent of delaying generic entry).
In principle, this process is protected by the Noerr-Pennington doctrine, which ‘generally immunizes efforts to petition the government from antitrust liability. The doctrine is based on the premise that third parties should be able to exercise their First Amendment right to petition the government without penalty.’
However, there are exceptions to this doctrine, particularly when petitions are a sham, clearly intended to interfere with the activities of a competitor. This appears to be the case in the Suboxone case, in which Reckitt Benckiser was accused of filing a sham citizen petition, based on ‘false data’.
This approach, which targets clearly false allegations, seems to be confirmed by the US Food and Drug Administration’s guidance on citizen petitions, which has listed a number of considerations that may suggest that a petition is a sham. This includes the fact that it has been submitted ‘with little or no data or information in support of the scientific positions set forth in the petition’.
A clear line seems to emerge from these elements: when authorities are not responsible for checking the information submitted to them, the companies providing the information must be subject to the highest standard, and any attempt to divert the process may be abusive. This does not mean that when authorities have the duty and the means to check the information, false or misleading statements are allowed: but the standard is lower, and only clearly baseless actions may be sanctioned as an abuse.
However, this distinction was blurred in the Janssen decision on the merits and entirely disappeared in Roche Novartis.
Janssen and Roche Novartis: a step too far?
The ECJ’s preliminary ruling in Roche Novartis confirmed the principle according to which information provided to authorities must not be misleading. The decision does not refer to national authorities, only to the European Medicines Agency, but it seems reasonable to extend this principle and consider that companies should not disseminate misleading information in their relations with stakeholders or authorities.
However, the Janssen and Roche Novartis decisions might go too far in that they apply the same standard for what constitutes misleading information no matter who is on the receiving end: HCPs or public authorities entrusted with the mission of verifying the information submitted to them. In this respect, it is significant to note that both decisions refer, in the same terms, to the fear of risk of both HCPs and public authorities.
In Janssen, to demonstrate the existence of an abuse, the FCA took into account the fact that the originator company contacted an authority that was not competent, as it did not have the power to refuse an MA. The FCA relies on the fact that the European Commission had already granted the MA and that the margin of intervention of the French authorities was very limited, although, de facto, it could result in the generic entry being delayed. Therefore, Janssen’s intervention was ‘legally unfounded’, in that it aimed at convincing an authority to adopt a decision it could not and ought not to adopt.
The decision even insists on the fact that there was ‘no doubt’ that the French authorities were obliged to grant an MA once it had been granted by the European Commission. This highlights one of the most unsettling elements of the decision: if there was no doubt, it is difficult to understand how Janssen’s behaviour could have any effect on the authorities’ decision. However, this point has been confirmed by the French Supreme Court: the mere fact of raising an unfounded debate with an authority may constitute an abuse.
Even more surprisingly, in Roche Novartis, the FCA concluded that the companies’ actions affected the pricing decisions of the French public authorities. On the basis of a summary analysis, the FCA explained that the communication campaign had an indirect effect on price because the pricing authorities could not take Avastin as a comparison product to negotiate a price reduction for Lucentis.
In the FCA’s view, this justified the application of the ‘objective and measured’ standard to any information communicated to authorities, thus departing from the approach taken in the AstraZeneca case and in the Janssen interim measures case. This approach is also much stricter than the one followed by the US authorities in which lobbying is only considered as an abuse of dominant position in cases in which an abuse is manifest (e.g., where false information has been willingly provided). Elements such as the tone used may be relevant in communications with HCPs but it seems surprising to take these into consideration when assessing communications with public authorities endowed with extensive powers.
This highlights one of the main weaknesses of the Roche Novartis case: the whole issue arose from a significant price differential between two products, and in that respect it is important to remember that prices are regulated. The case seems to underline a deficiency in pricing rules or in the way in which they were applied by authorities, rather than purely anticompetitive behaviour. The FCA analyses the behaviour of pricing authorities as if they were any other buyer, when in reality they have the regulatory power to define the price. To assume that these powers are reduced in the presence of a monopoly is not realistic. In particular, pricing authorities could act against a monopolist that endeavoured to compel them to grant a certain price by threatening to refuse to supply its products as refusal to supply is also considered as an abuse of dominant position. It will be interesting to see whether the Court of Appeal takes these considerations into account and reverts to a more realistic standard of communication with authorities although the strict approach taken recently by the French Supreme Court in Janssen may limit these chances.
In Janssen, the Supreme Court ruled, as a general principle, that even if health authorities enjoy exclusive decisional power, pharma companies should not go beyond making scientific recommendations on the modalities of originator-generic substitution. This standard appears quite strict to meet in practice (e.g., pharma companies are allowed to discuss the risks of substitution with health authorities, but in doing so, cannot raise any doubts as to the generic status of a drug, even where the health authorities themselves have reserved their position on whether to add the drug to the generics register).
A few practical recommendations
Given the very strict standards developed by the FCA, which have recently been adopted by other national authorities, and which may also be followed by the Commission in its ongoing investigations, it may be difficult to define a safe course of action from a practical point of view. A few key elements should be kept in mind:
- keep messages objective and the tone neutral;
- the main recommendation is to avoid any ‘fear message’ in communications with all stakeholders and authorities. Any element related to risk must be presented in an objective and measured manner;
- be careful about timing: certain actions are to be reviewed in a particularly strict manner if, for example, a new competitor is about to enter the market;
- beware of ‘implied’ messages and suggested links between two independent pieces of information; and
- when addressing the authorities, check which authority is competent in light of the legal framework and contact the competent authorities only. Limit the requests to what is in the authority’s power.
Because authorities consider communication campaigns as a whole, it is important to review all types of support. Very often the difficulty lies with the number and diversity of documents and supporting material that may contain denigrating messages, such as:
- external communications, including press releases and letters to HCPs (such as dear doctor letters), which can be very sensitive;
- internal documents, particularly those used to train and prepare sales forces, such as Q&As, case studies, internal training materials and oral messages expressed during training sessions;
- messages posted on social networks (e.g., Twitter communications were mentioned in Roche Novartis);
- comparative tables or documents, including those for internal use and even those based on objective elements;
- warning messages in prescription software: pop-ups appearing when pharmacists deliver (these are not problematic by themselves, but care should be taken to avoid any confusion-inducing element);
- screen savers in prescription software for doctors;
- material used in e-training for pharmacists;
- material provided at events such as dinners during a conference;
- any material issued by, or with the assistance of, external providers, as the content would be more difficult to control; and
- internal messaging services or internal web applications.
Subscribe here for related content, breaking news and market analysis from Global Competition Review.
Global Competition Review covers the most crucial developments in competition law and enforcement worldwide with breaking news commentary and analysis five days a week, as well as original in-depth reports, interviews and features on antitrust enforcement worldwide.