THE EUROPEAN FEDERATION OF PHARMACEUTICAL INDUSTRIES AND ASSOCIATIONS REPONSE TO THE PRELIMINARY REPORT IN THE PHARMACEUTICAL SECTOR INQUIRY

Transcription

1 THE EUROPEAN FEDERATION OF PHARMACEUTICAL INDUSTRIES AND ASSOCIATIONS REPONSE TO THE PRELIMINARY REPORT IN THE PHARMACEUTICAL SECTOR INQUIRY 30 January 2009 #

2 CONTENTS Sections Pages EXECUTIVE SUMMARY INTRODUCTION The Missed Opportunity: Analysis of Competition Impossible without Considering Member State Regulation Time to Generic Entry is Short and Getting Shorter: the Blocking Tactics Theory is Unsubstantiated Innovation is Robust: the Facts Do Not Support the Theory that Innovators Block the Development of New Products (Section C.3) The Regulatory Framework (Section D ) GENERIC ENTRY IS BY ANY MEASURE FAST AND THE FACTS DO NOT SUPPORT ANY THEORY THAT INNOVATORS CAUSE DELAYS The Premise of Day One Entry is Flawed (Section B ) Generic Entry Periods are Short and Decreasing, a Pattern of Conduct Incompatible with any Theory of Illegitimate Blocking Tactics ALLEGED BLOCKING TACTICS ARE LEGITIMATE ACTIVITIES, COMMON TO ALL INNOVATIVE INDUSTRIES Patent Portfolios (Section C.2.1) Patent Litigation (Section C.2.2) Settlements (Section C.2.4) Regulatory Complaints and Litigation (Section C.2.5.1) Marketing and Next Generation Products (Section C ) Accumulation of Practices (Section C.2.7) NO EVIDENCE THAT INNOVATOR PATENT STRATEGIES REDUCE INNOVATION (Section C.3) The Report Disregards All Other Possible Causes of Innovation Decline Patenting Strategies (Section C.3.1.2) Antitrust Intervention Should Not Add Still Greater Regulatory and Investment Uncertainty to the Process of Innovation POLICY RECOMMENDATIONS TO IMPROVE COMPETITION...35 ANNEXES 1. Technical Annex 2. Comments on DG Competition's Pharmaceutical Sector Inquiry: Preliminary Report, CRA International, January 2009 #

3 EXECUTIVE SUMMARY 1. Since January 2008, DG Competition has gathered unprecedented quantities of data from the pharmaceutical sector and other stakeholders at a very considerable cost. It has marshalled that data into a very readable Preliminary Report. However, the fact-finding presented therein is selective in its design to bolster the theories of harm that triggered the launch of the inquiry. 2. From the outset, EFPIA has sought to demonstrate that the unique challenges facing the sector have little to do with DG Competition's claim that innovator patenting strategies and other commercial conduct have the effect of inhibiting innovation or of delaying generic entry. Rather than seeking to objectively weigh the many factors that heavily influence competition in European patented and off-patent pharmaceutical markets, DG Competition continues to focus primarily on the conduct of innovator companies. 3. If carried over to the Final Report, the tone and approach of the Preliminary Report risk causing considerable damage to the innovative sector. The threat of antitrust intervention in relation to many common legitimate practices will undermine what is already a limited commercial window of opportunity in the quest to deliver tangible improvements in public health. 4. The months prior to the publication of the Final Report are an opportunity for the Commission to come to a more considered and balanced view of the challenges faced by the pharmaceutical sector and to contribute to the efforts underway elsewhere in the Commission to reduce undue regulatory burdens, generate efficiencies, and increase investment certainty in line with the Lisbon Agenda. 5. In a spirit of constructive engagement, this paper highlights those aspects of the Preliminary Report which EFPIA views as inaccurate, and proposes a number of policy recommendations that should be included in the Final Report to reap the benefits of low cost generic products and to ensure that innovative medicines are accessible more rapidly in the interest of public health. 1. Strong Patent Protection is the Foundation of R&D: There is No Evidence that Patents Hinder Innovation 6. Taking the number of new molecular entities as the sole indicator of innovation is a narrow approach that fails to reflect the value of advances in biomedical science, vaccines and incremental improvements. Notwithstanding this qualification, the pharmaceutical industry itself has expressed concern about a certain productivity decline in research. The welldocumented reasons for this decline are many. They include more complex scientific targets, and increased costs and higher attrition rates in late stage development due to both a greater risk aversion on the part of health authorities and negative signals from pricing and reimbursement authorities unwilling to pay for incremental innovation. 7. These factors have been the subject of much discussion between the industry and other Commission services. They were explained in great detail in earlier submissions made by EFPIA but have largely been ignored in the Preliminary Report. DG Competition instead continues to allege that patenting strategies have the effect of blocking competing R&D and dampening innovation. 8. This central thesis is overly simplistic and unfounded, and reveals a lack of understanding of how innovation occurs and how the patent system works in the pharmaceutical sector. #

4 Without a strong system of intellectual property protection and an ability to enforce and defend patents, it would be even more difficult to fund high-risk pharmaceutical research. 2. Delays to Generic Entry are Over-Stated and Wrongly Attributed 9. DG Competition's findings confirm that generic entry has accelerated over the time period investigated and is especially fast in relation to high value blockbuster medicines. It also notes in passing the relevance of regulatory factors marketing approval and reimbursement decisions in delaying market access of generic drugs. Nevertheless, the press announcements on the release of the Preliminary Report focused predominantly on the estimate that in the period 2000 to 2007, approximately 3 billion across the 17 countries sampled could have been saved had generics entered the market immediately upon the expiry of the innovator's exclusive rights. The inference that much of this expenditure was unduly pocketed by the innovative companies, through recourse to a so-called toolbox of tactics to unfairly impede generic market entry, is unfounded. The quicker entry in relation to high value products, coupled with the variations across countries and time, suggest that the commercial attractiveness of the product and the prevailing regulatory systems are the principal factors influencing the timing of generic entry. 10. The toolbox approach casts doubts on the lawfulness of common commercial practices (patent portfolios, patent litigation, settlement agreements, regulatory interventions and the promotion of next generation improved products) in certain, but unidentified, circumstances. The marketing spend of the research-based industry is portrayed as a barrier to generic entry implying that these funds could be better spent on research. This disregards the unique need to continually inform and obtain feedback from the medical community. The fact is that no other industry spends as much as the research-based pharmaceutical sector on R&D. Patents filed near the expiry of a "primary" product patent are portrayed as an attempt to deter competition from imitation products rather than to protect improvements during a product's lifecycle. This disregards the fact that modern technologies across all sectors rely on a matrix of extremely complex patented inventions that are upgraded on a constant basis. 3. The Potential for Savings from More Efficient Generic Markets is Largely Ignored 11. EFPIA has long advocated more competitive and efficient generic markets in Europe as the corollary to strong patent protection and reward for innovation and improved patient access to medicines. It is paradoxical that Europe, as shown in the Preliminary Report, pays significantly more for generics but less for innovative drugs than the United States. There is an increasing body of independent literature attesting to the potential savings to be gained from more efficient generic markets. The Preliminary Report cites the Dutch preference policy favouring the lowest price generics with a limited period of exclusivity which resulted in immediate price reductions in the region of 80%. If replicated across Europe, such mechanisms would result in savings that would vastly exceed the estimated benefits of immediate generic entry. 4. Policy Recommendations 12. The Final Report should contain policy recommendations on how to reduce costs and increase legal and commercial certainty for all parties. In particular, it should address the most significant market entry barrier, namely the sheer complexity and diversity of the applicable national regulatory regimes. Whereas the Preliminary Report has focused on alleged practices designed to delay the access of generic medicines, the reality is that entry delays are much more significant for innovative medicines. Streamlining processes to provide faster access to therapeutic advances (including through the better enforcement of the Transparency #

5 Directive) improves the quality of life/longevity of patients and is in the broader public interest. 13. Stimulating price competition amongst generics and ensuring that those savings are in large part passed on to the ultimate payors should be a major focus of the Final Report. The resultant societal benefits are acknowledged in the December 2008 Commission Communication on Safe, Innovative and Accessible Medicines. 14. The Final Report will contain policy recommendations on how the current patent system can be improved. These should include: adoption of the European Community Patent; the creation of a unified, specialised litigation system in Europe; a streamlining of the opposition procedure of the European Patent Office; and a mechanism to address patent disputes before generic launch. 15. The European Courts have established that intrusion into intellectual property rights can only be justified in the most "exceptional circumstances", and any attempt to expand this notion to challenge legitimate commercial practices will have a chilling effect on innovation. 16. A repetition of DG Competition's preliminary findings in the Final Report would not only damage the reputation of the research-based pharmaceutical industry, but create significant legal uncertainty affecting all innovative industries that rely on strong patent protection. The Commission s Final Report should take due account of the various parameters that shape competition and act as a stimulus to (1) reduce regulatory barriers and strengthen intellectual property protection, (2) generate efficiencies and savings for healthcare budgets, and (3) promote Europe's health and competitiveness. The implications of this ambitious agenda for patients are clearly illustrated by the following statistics 1 that are a stark reminder of the importance of fostering pharmaceutical innovation: Five-year Relative Survival (%)* during Three Time Periods By Cancer Site Site All sites Breast (female) Colon Leukemia Lung and bronchus Melanoma Non-Hodgkin lymphoma Ovary Pancreas Prostate Rectum Urinary bladder *5-year relative survival rates based on follow up of patients through Recent changes in classification of ovarian cancer have affected survival rates. Source: Surveillance, Epidemiology, and End Results Program, , Division of Cancer Control and Population Sciences, National Cancer Institute, Extract from presentation made during a site visit with DG Competition s sector inquiry task force in October 2008 by Paul Stoffels, Group Chairman Johnson & Johnson, Challenges for Pharmaceutical Research. #

6 1. INTRODUCTION 17. The European Federation of Pharmaceutical Industries and Associations ("EFPIA") welcomes the opportunity to contribute to the examination of competition and innovation in the pharmaceutical sector, a sector vital to the health and welfare of Europeans and a leading player in Europe's innovative economy. 18. The core of the preliminary report 2 ("report" or "PR") concerns the tension between patents and competition, healthcare cost containment and the right climate for pharmaceutical innovation. These have long been the subject of debate between industry and policy makers 3. The Communication on the Pharmaceutical Package 4 shows that much progress has been made, but much remains to be done in this complex policy area. 19. It is regrettable that subjects which merit careful and measured consideration are presented in a one-sided manner. Legitimate and essential business activities for all innovative sectors of the economy patent portfolios, patent litigation and settlement agreement agreements,, regulatory interventions, and the development and marketing of next generation products are portrayed as costly "delaying tactics," 5 "shocking facts" 6 and inherently suspect conduct The report fails to establish the reasons for fewer new medicines coming to market, and apparent delays in generics entering the market. The analysis presented does not support the report's conclusions. It paints a partial picture of competition in the pharmaceutical sector and makes no effort to assess or attribute causation. Key factors that affect the competitive dynamics of entry and innovation in European pharmaceutical markets are either omitted or barely treated. 21. The analysis is further undermined by misunderstandings as to the nature of the patent system, use of selective quotation rather than hard data, and over-reliance on agenda driven commentary. No independent scrutiny of the data used in the report has been possible. We seek to offer, nonetheless, comments on the report and suggest the appropriate antitrust framework to guide further analysis of the issues facing the pharmaceutical industry. 1.1 The Missed Opportunity: Analysis of Competition Impossible without Considering Member State Regulation (a) State regulation permeates every aspect of competition in the sector 22. The report's limited analysis of State regulation and its impact on competition is a serious defect. State regulation should be central to the report's consideration of: (1) generic entry, Case No Comp/D2/ Commission Decision of 15 January 2008 initiating an inquiry into the pharmaceutical sector pursuant to Article 17 of Council Regulation No 1/2003 [2008] OJ C59/06. High Level Pharmaceutical Forum , Final Conclusions and Recommendations, 2 October Communication from the Commission: Safe, Innovative and Accessible Medicines: a Renewed Vision for the Pharmaceutical Sector COM(2008)666 final, 10 December Commission Press Release of 28 November 2008 Preliminary report on pharmaceutical sector inquiry highlights cost of pharma companies' delaying tactics; see, also, paras. 232, 771 and 891. Deputy Director General Ungerer, Oral Hearing, 28 November Commissioner Kroes, Press Conference, 28 November 2008 ("[W]hen we put it on paper, what is wrong, what is absolutely to be changed, then the players in the game so to say are already changing their behaviour so I sincerely hope that our clear language is reason for the industry and certain companies to change their attitude already."). #

7 where state controls, not innovator conduct, are responsible for significant delays; and (2) the decline in innovative productivity which is heavily influenced by state controls The omission is surprising in light of the Commission's call for "better regulation for a more competitive industry" in the Communication on the Pharmaceutical Package which highlights the harm caused by regulatory distortions. 9 EFPIA urges that this deficit be addressed in the final report. (b) Substantial inefficiencies in the generics markets leave European healthcare systems overpaying billions for generic medicines 24. A wide range of stakeholders, the judiciary, industry and payors expressed concerns at the launch of the report that competition in the off-patent market is not part of the review. Commentators asked how prices in a commodity market could remain on average at 80% of the branded price one year after generic entry. In a genuinely competitive generics market, those margins should be competed away to the benefit of healthcare providers. 25. The London School of Economics has identified that potential savings of 44% on generic medicines could be achieved through better generic to generic competition. 10 US data show that competitive generic markets can offer prices significantly below the branded price. 11 Recent literature on international pricing also attests to the fact that non-us prices are 20-40% lower than US manufacturer prices but only 10-30% lower than US public prices (indicating that distribution margins absorb a larger share of total spending in certain regulated European markets 12 ), and outlines why generics are also cheaper in the US than in other countries. 13 Economic research commissioned by EFPIA for the inquiry 14 examines how the many different State incentives for generics can deliver more competitive outcomes. The report does not consider this research. 26. The experience of the Dutch health insurers provides the most compelling example of the savings available. The insurers successfully achieved 80% savings on generics by challenging pharmacists' incentives, driven by high margin retentions, to maintain high prices in a kind of silent trust between the pharmacist and the generic industry. 15 This suggests that examination of the generic supply chain would yield substantial savings for Member States The Current State of Innovation in the Pharmaceuticals Industry, CRA International, June 2008 ("CRA Innovation Report", Annex 4 to the EFPIA submission of 13 June 2008). Communication on the Pharmaceutical Package, supra, n 4, pp5-7 ("Different systems lead to disparities in pricing, time-to-market delays and access inequalities;" "Genuinely transparent and speedy pricing and reimbursement decisions should be made possible by enhancing the application of the Transparency Directive;" "Requirements that cause a high administrative burden without providing a clear public health benefit have a strong negative impact on the competitiveness of the EU industry;" "[A] high level of joint commitment is required from all actors in order to make pharmaceutical rules clearer, simpler and more flexible"). Competition in off-patent drug markets: Issues, regulation and evidence, Panos Kanavos, Joan Costa-Font and Elizabeth Seeley, Economic Policy July 2008 pp , at pp US Generics Pharmaceutical Association ("[A] generic medicine can cost 30% to 80% less than its equivalent brand name drug"); National Association of Chain Drug Stores (in 2007 the average generic drug price was USD compared to USD for the branded innovator drug). This is reaffirmed by the 35% difference between ex-factory and retail prices for pharmaceuticals (PR, para. 5 on page 21) which appears high by comparison with certain Member States and suggests that greater efficiencies in the supply chain could also be a source of substantial savings. International Prices and Availability of Pharmaceuticals in 2005, Patricia Danzon and Michael Furukawa, Health Affairs Volume 27, Number 1, January/February Competition in the off-patent market post generic entry, CRA International, September Edwin Velzel, CEO UVIT, Presentation to the Pharmaceutical Sector Inquiry Hearing, 28 November #

8 27. The Dutch authorities' focus on generic to generic competition achieved per capita savings of EUR 24 per annum, 16 many times higher than the EUR 0.90 per capita 17 which the report suggests could be saved by quicker generic entry. 18 Very substantial sums could therefore be saved if the final report encouraged other Member States to replicate the Dutch insurers example. "[T]hat problem we have tackled, the problem of rapid access to generic medicines, but we had another problem in place. That was overpayment [the innovator conduct in the report] it is not the whole problem, it is not even the largest part of the problem." Edwin Velzel, CEO of Dutch Health Insurer, UVIT at the 28 November 2008 hearing 28. A recent example in Italy, where it was discovered that generic companies had been affording pharmacies unofficial discounts of up to 80% off the official price listed in the Official Journal, further illustrates this point. In essence the pharmacies would invoice the Italian health insurer for 100% of the price but would only pay 20% to the generic companies with the pharmacies pocketing the resulting windfalls. It is estimated that this practice cost the national health insurance fund a total of 1 billion per annum With savings at this level, the much promised "financial headroom for innovative medicine" to be delivered by real generic savings and identified in the Communication on the Pharmaceutical Package, 20 could be genuinely achievable. 1.2 Time to Generic Entry is Short and Getting Shorter: the Blocking Tactics Theory is Unsubstantiated (a) The alleged EUR 3bn cost of delay cannot be substantiated 30. There is no evidence that innovators behaviour contributed to an alleged EUR 3bn cost of delayed generic entry. The evidence in fact shows that after an innovator's exclusive rights expire ("loss of exclusivity" or "LOE"), generic entry is fast: on average 6.6 months (weighted); under four months for highest value medicines; and as little as 2.5 months in some countries. Generics themselves state that they will not be on the market on day one after LOE, inter alia, because of regulatory delays. They say these delays, not innovator conduct, are "costing European healthcare systems as much as EUR 100m per medicine per year." Representing savings of EUR 400m (the maximum achieved) divided by the population of the Netherlands. Representing savings of EUR 3bn over eight years divided between the 415m population of the 17 Member States in question (Eurostat population data, 19 December 2008). The savings were up to EUR 400m in one Member State, in one year over 33 medicines, compared to the EUR 375m per year (EUR 3bn over 8 years) covering 75 products in 17 member states which the report claims is possible from quicker generic entry. La Repubblica newspaper of 24 May Communication on the Pharmaceutical Package, supra, n 4, p7 ("Many Member States recognise that generic medicines play an important role in helping to limit their healthcare expenditure in their reimbursement and prescribing practices. Competition with off-patent products enables sustainable treatment of more patients with less financial resources. The generated savings create financial headroom for innovative medicines."). Myth & Reality 08 Intellectual property protection EU vs USA ("Immediate generic competition upon patent expiration No Pricing & reimbursement decisions create delays in most Member States"); European Generics Association; Sustaining Generic Medicines Markets in Europe, (April 2006) Prof. Dr. Steven Simoens, Sandra De Coster, Research Centre for Pharmaceutical Care and Pharmaco-economics, p67 (Notes P&R delay post LOE and price regulation deterring entry); EGA Board of Directors urges early implementation of "Bolar provision" for Generics to ensure the future of European Research & Development, EGA Press Release, 25 March (Post-patent expiry regulatory delays cost European health authorities up to EUR 100m per medicine.) #

9 31. The report simply assumes each day of non-entry after LOE is caused in part by innovator conduct. It does not consider the many regulatory and commercial reasons why generics are likely to enter later than LOE. It does not examine the expert report on these matters provided by EFPIA. 32. In fact, the data show no connection between alleged innovator blocking tactics and the time of generic entry. Generics enter most quickly after LOE for drugs where alleged blocking tactics are claimed to be most prevalent. The speed of entry has accelerated over the period under investigation despite claims that blocking tactics have increased two, three or fourfold over that time. The report, therefore, offers no credible basis for assuming generics could enter any quicker than they do. The claimed EUR 3bn cost of delay is a distraction from the real issues affecting competition in the sector. (b) The "toolbox" lists only legitimate business practices essential to all innovative industry sectors 33. The so-called "toolbox" of originator strategies alleged to delay generic entry 22 is simply a description of lawful commercial activities common to all innovative industries: patent portfolios, patent litigation, settlements, regulatory interventions, and the patenting, development and marketing of next generation products. 23 Calling into question the legality of any of these activities is to invite technological stagnation. It would have a significant and far-reaching chilling effect on innovation, investment and employment across all research based industries on which Europe and the achievement of the Lisbon Agenda depend. 34. Patent portfolios (Section C.2.1.2) reflect the level of innovation in the pharmaceutical industry as in any high technology sector. The much quoted figure of 1,300 patents or patent applications misleadingly inflates the number of parallel patent "families" needed to obtain protection in each of the EU 27 states. The truth, as the US Federal Trade Commission ("FTC") has observed, is that a medicine is protected by relatively few patent families. 24 According to the PR, the average is 12 and there are generally no more than patent families. This may be compared to the hundreds or even thousands protecting other high technology products. Generic competitors are undeterred by patent portfolios. They have shown themselves adept at designing around, challenging patents and obtaining their own patents relating to an innovator's product. 35. The apparent criticism of the patenting of improvements to a product late in the original patent term disregards the importance of incremental improvements as a source of innovation. 25 Such improvements provide real benefits to patients in terms of potency, reduced side effects, broader therapeutic indications and simpler dosing regimens. Competition rules should not prevent or inhibit an innovator from using its superior knowledge of its own products to devise and protect new and valuable improvements for the benefit of patients and prescribers. The timing of new developments is irrelevant. There is See, eg, PR, para Patents and Pharmaceuticals, Presentation to the Pharmaceutical Sector Inquiry Hearing, 28 November 2008, Lord Justice Jacob, senior intellectual property judge of the Court of Appeal of England and Wales ("[A]ny experienced patent lawyer will tell you that clusters of improvement patents are a feature of nearly all industries. It is a bit worrying that the Commission seems to think that it has discovered something new and special to the pharma industry."). To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, Federal Trade Commission, October 2003, p32. Indeed, patent jurisprudence and reforms have actively sought to enable the protection of investments made by innovators in developing and testing new medical uses for existing medicines. #

10 no cut off point in the continuum of development beyond which inventions become suspect. That would be to deny to society the advantages of later inventions. 36. Patent litigation (Section C.2.2) is essential for intellectual property protection. The report cannot validly extrapolate from a handful of decided cases that the pharmaceutical sector is characterised by "weak secondary" patents. To do so commits the "self-selection error" of choosing the most contentious cases the minority of patents that are litigated to generalise about the rest. The reality is that once the primary patent has expired, imitators are free to copy it. If weak patents are identified, generic companies have demonstrated their ability and willingness to challenge or design around them. Suggesting that certain types of patent litigation may be suspect creates damaging legal uncertainty for innovators who rely on patent enforcement as the only effective means of protecting their intellectual property rights. 37. Patent and other settlements (Section C.2.4) are essential for businesses to draw a line under protracted and expensive litigation and move on, allowing the parties to invest in conditions of business certainty. The public interest in facilitating settlement of litigation, both for litigants and overworked courts, is universally recognised. Settlements cannot be categorised as suspect or legitimate based on certain commercial terms within them, as the FTC advocates and the report repeats. Just as the most recent US case law rejects the FTC's approach, the correct antitrust analysis is that there can be no presumptively good or bad settlements. Restrictions that go no further than the exclusionary effect of the patent generally do not restrict competition. 38. Recourse to regulators (Section C.2.5.1) is essential and in the public interest in any highly regulated industry. Indeed, it can be a regulatory duty to raise concerns. Companies must be free to come forward with bona fide concerns as to generic safety, quality or counterfeits without fear of ex-post antitrust scrutiny. The innovator is likely to be most knowledgeable about its products and the risks posed by non-compliant generics. It is regarded by regulators and the medical community as responsible for the molecule it originated, even in generic form. It must be permitted to raise any legitimate suspicions. The regulator can and will decide whether companies' concerns are justified. 39. Marketing and development of next generation products (Section C.2.6) are the essence of competition, particularly in innovative industries. It would be absurd to suggest that companies should not introduce next generation products because to do so unfairly disadvantages generics who can copy only old technologies. If next generation products are successful then they will add to the range and choice of products available to doctors and patients, including generic versions of the old products. If they are not, then the cheaper generic versions of the old technology will prevail. There can be no time or cut off date at which incremental improvements may be judged to be suspect conduct. Timing will depend on development of the improvement, obsolescence of the old product, and anticipated demand for the new one. 1.3 Innovation is Robust: the Facts Do Not Support the Theory that Innovators Block the Development of New Products (Section C.3) 40. The report criticises apparent "defensive" patenting strategies with no intent to use the patents to bring a new or improved medicine to market. In support of this theory it cites 1,100 instances across the EU where an innovator's patents might overlap with the R&D programme and/or patents held by another originator. This allegedly creates "significant potential for originators to find their research activities blocked" with an unquantified detrimental effect on the innovation process. #

11 41. The report fails to make its case. Competition amongst innovators is robust. Billions are spent in the race to bring medicines to market to address unmet and, by their nature, increasingly complex medical needs. The data show a steady increase in R&D spend and in the number of projects under development. In 2007 there were almost 4,400 products in clinical development, including 763 for cancer, 224 for cardiovascular diseases, 177 for respiratory diseases and 92 for Alzheimers. Amongst the literature examining the reasons for productivity decline in the sector, including the economic report submitted by EFPIA but not addressed in the report, 26 none has stated that patent protection is anything other than an essential precondition for successful innovation. 42. No quantifiable harm has been established to have resulted from the alleged patent blocks. Cross-licensing is common practice and the report finds only one instance of an R&D project being abandoned because of potential intellectual property concerns. All the indications are that late stage attrition is the culprit - particularly at phase II and III of clinical trials - at a time when the IP position of the candidate drugs will long since have been secured. Innovators face other challenges outlined in detail in EFPIA's submission of 13 June 2008 which the report has chosen not to address. 1.4 The Regulatory Framework (Section D ) 43. Finally, the report notes the shortcomings of the current regulatory framework. However, it does not make any assessment of its impact beyond the conclusion that a single Community patent and accompanying unified and specialised judiciary is needed. EFPIA supports the recommendations for an efficient and world-class system of patent grants and litigation. It objects to the methodology employed which focuses only on corporate conduct as the cause of concern, rather than the well documented competitive distortions caused by regulatory controls. 44. Undue competition law intervention in common commercial practices would undermine what is already a limited commercial window of opportunity. It would increase legal uncertainty in an industry that is already subject to an unprecedented level of regulatory intervention, and could ultimately limit patients access to innovative medicines. EFPIA urges the Commission to fully assess in the final report: (1) the impact of the fragmented regulatory framework; (2) the scope of savings available if there were more efficient generic markets; and (3) the policy adjustments that will be required to create a genuine climate for innovation in Europe. The Commission's far-sighted partnership with industry on the Innovative Medicines Initiative is just such an initiative with the real potential to promote innovation in the European healthcare sector. 26 See EFPIA submission of 13 June #

12 2. GENERIC ENTRY IS BY ANY MEASURE FAST AND THE FACTS DO NOT SUPPORT ANY THEORY THAT INNOVATORS CAUSE DELAYS 2.1 The Premise of Day One Entry is Flawed (Section B ) 45. The report seeks to establish that entry by generics "occurs later than could be expected" and that the innovator "practices under investigation contribute to this [delay]". 27 Each day of delay between LOE 28 and the date of generic entry is assumed to be caused at least in part by innovators and an estimated EUR 3bn lost to European healthcare budgets accordingly. 46. The premise of generic entry on day one after LOE has no factual or legal basis. 29 De novo entry into any industry sector takes substantial time and resources. 30 The substantial regulatory and commercial challenges involved in bringing a medicine to market including production, distribution, marketing authorisation, pricing and reimbursement approvals, make a general assumption of day one entry wholly implausible. Generic companies admit that they will not be on the market on day one after LOE. They state that, inter alia, that regulatory delays are the cause: Reason Pricing & Reimbursement Bolar Data Exclusivity Price Controls Post LOE Generic Delay "Immediate generic competition upon patent expiration No* Pricing & reimbursement decisions create delays in most member states." 31 "[G]eneric manufacturers must wait until the original patent has expired before beginning R&D work Waiting until patent expiry delays market entry for generic products, costing European patients and healthcare systems as much as 100 million per medicine per year." 32 "The effective period of market exclusivity gained by the originator company is the period of data exclusivity (currently 6 or 10 years) plus the time it takes to register and market the generic medicine a further 1 to 3 years." 33 "Pricing regulation drives down the originator price over the life cycle of the medicine. This lowers the potential profit margin for a generic medicine company and discourages market entry." In fact, the average times to entry after LOE are just 6.6 months (weighted average) for all molecules, under 4 months for high value medicines and as low as 2.5 months in some PR, p5 (executive summary). Assessment of the date of LOE can be complex and depends on many factors. Lack of precision in collating the data can create an inaccurate data set. That is not to say that entry on day one is impossible in all cases (see, eg, PR, para. 166.) It may be noted that the US, a much larger and less commercially regulated market, achieves day one entry, but it certainly cannot be an assumption in the smaller and highly price-regulated European markets. Compare Guidelines on Vertical Restraints [2000] OJ C291/1, para. 26 (A competitor's threatened entry within one year is sufficiently swift to exercise a competitive constraint on market incumbents.). Myths & Reality 08 Intellectual Property Protection, EU v USA, European Generics Association. EGA Board of Directors urges early implementation of 'Bolar provision' for generics to ensure the future of European R&D, European Generics Association, Press Release 25 March 2004 (Note that some generics did testing outside the EU in countries whose patent laws did not prohibit testing in order to circumvent this restriction. Testing was permitted by patent reforms effective from October 2005, but 6 out of the 7 years examined by the report will have been subject to the old law.). European Generics Association, at (viewed 16 January 2009). Sustaining Generic Medicines Markets in Europe, Research Centre for Pharmaceutical Care and Pharmaco-economics April 2006, Prof. Dr. Steven Simoens, Sandra De Coster, p67 (the report was funded by the European Generics Association). #

13 countries. Those times are fast when considering the regulatory and commercial challenges involved. 35 The report does not consider the extent to which these factors contribute to the alleged delay. It does not examine the expert economic evidence provided by EFPIA on these matters As to regulatory factors, marketing authorisation involves considerable timing uncertainties; as the report states, generics receive authorisation on average 5.3 months later than expected. 37 There are two key delays after LOE: (1) for most of the period under investigation 38 national patent laws generally prevented generics conducting bioequivalence studies or trials within the EU necessary to obtain market authorisation, 39 a process that could take as long as two years post LOE; 40 (2) for products where data exclusivity rights expire at or near the LOE date there would necessarily be a further delay of around 12 months or more as generics would only then be able to start the procedure for obtaining marketing authorisation based on innovator regulatory data. 41 Data exclusivity is particularly important in countries where patent or SPC protection is weak or non-existent, inter alia, in Spain, Greece and Portugal, as those countries formerly had limited patent protection. 42 The report notes entry periods of months in those countries. 49. Generics also identify P&R approval as causing significant delay after LOE. 43 Price regulation has a dissuasive effect on generic entry as generic companies and the report note Depending on the market or medicinal product, there can also be substantial commercial disincentives for generics to enter a market where: (i) there is competition from other therapies; (ii) the market is a small one and thus an unattractive commercial opportunity; (iii) the medicine goes off patent in only one market rather than multiple markets simultaneously, again making entry a less attractive commercial proposition; (iv) the medicine is difficult to PR, paras , Tables 13 and 14. Factors Affecting Generic Entry in Europe, CRA International, June 2008 ("CRA Generics Report") (Annex 5 to the EFPIA submission of 13 June 2008); Comments on DG Competition's Pharmaceutical Sector Inquiry: Preliminary Report, CRA International, January 2009 ("CRA Comments on the Preliminary Report"). PR, para Six out of the seven years of molecule expiries subject to the investigation (PR, paras , Figs and Table 11) were prior to introduction of the Bolar amendments (PR, para. 260) under Article 10(6), Directive 2001/83 on the Community code relating to medicinal products for human use (as amended) [2004] OJ L 311/67, as amended by Directive 2004/27/EC [2004] OJ L 136/34, the transposition of which was required by 30 October From that date, testing by generics prior to LOE was permitted for the purpose of obtaining marketing authorisation. CRA Generics Report, supra, n 36, pp Canada Patent Protection of Pharmaceutical Products WT/DS114/R (17 March 2000), Report of the Dispute Settlement Body of the World Trade Organisation, para. 4.7 (Commission cites EFPIA data estimating two year postpatent expiry period required to conduct such preparatory studies and trials). PR, paras The products under investigation will be subject to the data exclusivity rules applicable prior to amendments under Directive 2004/27. The amended regime applies prospectively only to medicinal products for which an authorisation application was filed after 20 November 2005 under the mutual recognition procedure and after 30 October 2005 for the centralised procedure (Article 89, Regulation 726/2004 and Article 2, Directive 2004/27). In Spain, Greece and Portugal, data exclusivity is particularly important because the patent system formerly offered protection only for processes. CRA Generics Report, supra, n 36, p3; Myth & Reality 08 Intellectual property protection EU vs USA, supra, n 32,("Immediate generic competition upon patent expiration No Pricing & reimbursement decisions create delays in most Member States"); at the hearing on 28 November 2008, the EGA stated that this can be 13 months in some Member States. PR, para. 147 ("The level of generic penetration in the EU is influenced by the different public policy choices made by the Member States"); CRA Generics Report, supra, n 37, p42-43; Sustaining Generic Medicines Markets in Europe, April 2006; Prof. Dr. Steven Simoens, Sandra De Coster, Research Centre for Pharmaceutical Care and Pharmacoeconomics, p67 ("Pricing regulation drives down the originator price over the life cycle of the medicine. This lowers the potential profit margin for a generic medicine company and discourages market entry." This report was financed by the EGA.). #

14 manufacture; (v) the medicine has an associated delivery device, such as an inhaler, syringe or patch which is costly or complex to manufacture; or (vi) loyalty to incumbent brands require costly marketing to displace. 51. Conventional economic analysis would examine whether these factors might be the cause of entry delays so as to establish the "counterfactual" by which other causes of entry delay might be judged. The report does not take this step and so cannot validly conclude that innovators contributed in any way to the time to generic entry or associated costs Generic Entry Periods are Short and Decreasing, a Pattern of Conduct Incompatible with any Theory of Illegitimate Blocking Tactics 52. Analysis of the data indicates that innovator conduct is not involved in alleged generic delays: (i) generic entry is fastest in relation to medicines where alleged innovator blocking conduct is claimed to be most prevalent; 46 (ii) generic entry has become faster over a time period 47 where the prevalence of the alleged innovator blocking conduct has doubled (patents), trebled (settlements) or quadrupled (litigation); 48 (iii) generic entry is fastest in countries having some of the highest incidence of such alleged innovator blocking conduct. It is almost three times slower in a country that has the lowest incidence. 49 On this basis, the expert economic consultancy, CRA International, concludes in its comments on the report that "[t]he Commission do not present any evidence demonstrating that the time from LOE until generic entry is resulting from the use of a 'toolbox' of measures i.e. there is no evidence demonstrating causality." The data are consistent with entry being influenced by commercial and regulatory factors and not by alleged innovator conduct CRA Comments on the Preliminary Report, p3. The report shows that generic entry is quickest in relation to high value medicines and entry occurs in the highest numbers (PR, para. 166 ("[I]t takes less time for high value products to be faced with generic entry."); PR, para. 172 and Figure 16; see, also, CRA Generics Report, pp27-28), even though it is alleged "companies resort to broader and more intensive use of lifecycle tools for INNs generating higher revenue" (PR, paras. 904 and 906, Figure 132) and that "the basic effect of multiple actions can be that entry will tend to occur later." (PR, para. 909). The report shows the speed of generic entry is increasing (PR, paras and 168, Figure 12); EFPIA's research shows that in the main five European markets for 50-76% of products (weighted) entry occurred within three months over , rising to 60-80% for the latter part of the period ( ) (CRA Generics Report, p26). Patent applications doubled (PR, p336), number of INNs covered by settlements trebled (PR, para. 582, Figure 91) and litigation quadrupled (PR, para. 471). Out of the big five markets, the UK has the quickest weighted entry (3 months post LOE entry) but the second highest number of patents, patent litigation, interim injunctions, BII settlements and regulatory interventions before marketing authorities. Conversely, France has 8 months weighted post-loe entry, almost three times the UK's period, but the lowest number of instances of litigation, interim injunctions and BII settlements out of the big five (slightly lower patent numbers than the UK and the same number of marketing authorisation interventions). CRA Comments on the Preliminary Report, pp3, 6-7. #

15 3. ALLEGED BLOCKING TACTICS ARE LEGITIMATE ACTIVITIES, COMMON TO ALL INNOVATIVE INDUSTRIES 53. The fact that innovators seek to protect and market their products by means of patent portfolios, litigation, settlements, regulatory intervention, marketing, and the introduction of next generation products, is common to all innovative industries. EFPIA strongly disagrees with any suggestion that these practices are of questionable legality, and with the use of terms such as "toolbox" or "delaying tactics" in a pejorative manner to describe them. EFPIA further rejects any suggestion that the lawfulness of these practices can be judged by the intent of the companies who use them; the conduct in question primarily concerns patents which, by their nature, are intentionally exclusionary of imitators. 3.1 Patent Portfolios (Section C.2.1) (a) One patent cannot extend another (Section C.2.2.1) 54. The report apparently proceeds on the incorrect assumption that one or multiple patents can extend the term or protection of another patent. 51 The terms "thicket," "cluster," "web" or other pejorative descriptions of patent portfolios adopted by the report have no meaning in patent law. They do not alter the well established legal position that no later patent can extend an earlier patent's term or protection. 55. The patent on the basic molecule is central to the medicinal product and tends to be the first to expire, after which imitators are free to copy it. Subsequent patent applications must be for different inventions and generally have narrower claims covering further developments of the medicinal product. Generic companies, as the FTC notes, can and do "'design around' patents, so that they can develop non-infringing generic versions of brand-name drug products" 52 and, indeed, commonly obtain their own patents inter alia on process or formulations relating to a successful innovator product. In relation to one well- known blockbuster, a total of 129 third parties, many of them generics, had filed for patents. The top five of those third parties had filed 78% in contrast to the innovator s 22% of patent families indexed to the product in all its forms That generics are successful in designing around any patent portfolio is clear from the data submitted on the commercial window of exclusivity available to innovators. For the top 10 medicines, the period between first innovator sale and first generic sale was in all but one case less than 15 years - the period identified by the EC legislator as an appropriate window of commercial exploitation 54 - in some cases significantly less PR, paras , , Tables 52 and 53 (eg, at para. 377 "[A] patent cluster, during and towards the end of the protection period of the base patent, with the aim of keeping generics off the market beyond expiry of the first patent."). To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, supra, n 24, Chapter 3, pp 1 and 14 ("Representatives from generic pharmaceutical firms discussed how patent disclosures guide their efforts to "design around" patents, so that they can develop non-infringing generic versions of brandname drug products"; "Patent disclosure requirements can enable brand-name and generic competitors to design around some patents covering brandname drug products in order to bring competing products to market."). Data taken from the DOLPHIN database of all pharmaceutical inventions on 28 January 2009, published by the Thomson Corporation but not independently verified by EFPIA. See, also, EFPIA submission of 13 June 2008, p86, Figure 22. Recital 8, Council Regulation 1768/92/EC concerning the creation of a supplementary protection certificate for medicinal products (as amended) [1992] OJ L 182/1. EFPIA submission of 13 June 2008, Figure 9 (Effective period of protection for top 10 medicines by 2007 value losing exclusivity between 2000 and 2006, biologics and combinations excluded, in the EU 15, Norway and Switzerland). See #

16 57. Thus, regardless of whether innovators entirely legitimately seek to patent further inventive developments of the original drug, there can be no assumption that such patents preclude generic competition. (b) Misstatements inflating patent numbers (Section C.2.1.2) 58. The report unsuccessfully seeks to present medicines as being "surrounded by a multitude of patents and patent applications" such as to preclude generic challenges via this "multilayered" protection. 56 It singles out a medicine with 1,300 patents or applications but it has incorrectly cumulated independent national patents to reach this total, an approach which disregards the national nature of patent law. In each country the same invention the basic compound, formulation or process for example is protected in parallel by national patents. A German patent does not provide "multilayered" protection for a French one. 59. The correct method is to count "families" of parallel patents 57 protecting the same invention in different States Thus the real number is therefore a fraction 59 of the 1,300 quoted throughout the report and DG Competition press statements. The truth is, as the FTC has noted, that pharmaceuticals are protected by a relatively limited number of patents, 60 in contrast to other industries where 100s or 1,000s can be the norm. 61 The report's theories of "webs" and "clusters" have no factual basis. (c) Poor patent quality cannot be inferred from litigation outcomes, stricter grant rates or higher opposition, indeed the reverse may be the case (Section C.2.1.2) also Patents and Pharmaceuticals, supra, n 23, para.7, where Jacob LJ comments that the period of exclusivity granted by a patent, even as extended by a SPC, may not be sufficient to promote innovation and investment. PR, paras ("In such a [web of innovator patents] any attempt to develop a generic version... would inevitably infringe a patent."). The numbers are correctly represented at page 147. There it notes that 700 patent applications were filed in respect of the most valuable 20 medicines in the E75 group or on average 35 per product or, to put it in the context of a 20 year product lifecycle, an average of 1.75 patents per year. This is overlooked in the remainder of the report. While the average number of countries in which protection is sought is 15 (PR, para. 334), most modern medicines are protected in all 27 countries (PR, para. 1085) and the number of countries in which protection is sought is highest for the most valuable medicines (PR, para. 957), as is likely to be the case with the product in question. The average is 12 patent families. PR, para. 334, (around 38,865 patents/applications cover the 219 INNs investigated divided by the average number of countries designated in each application, being 14.8). To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy,supra, n 24, p32 and Chapter 3, p14 ("In a simple economic model of innovation and patents, each invention requires access to only one or a few patents to commercialize the patented product. Certain industries, such as pharmaceuticals, have tended to follow this model. Some suggest, however, that more and more industries are moving toward the model in which, for commercialization, a product requires access to many patents dozens, hundreds, or even thousands."; "Fewer patent thicket issues arise in the pharmaceutical context than in industries where innovation is less discrete and individual products are covered by many patents."). 20 patents cover innovations on a wetshave razor ("What makes Gillette Fusion Power better than Mach3 is 8 years of shaving innovation and 20 more patents"); 200 for iphone innovations ("Selon Apple, l'appareil intègre des innovations protégées par plus de 200 brevets." Conseil de la Concurrence Décision n 08-MC-01 du 17 décembre 2008 relative à des pratiques mises en oeuvre dans la distribution des iphones); 1,000 for digital cameras ("Kodak has over 1,000 digital camera patents." Interview with Kodak engineer Stephen J Sasson, Rediff News, 7 August 2006); "thousands" for telecoms consumer products ("In the telecom kingdom, today's great wireless knights - AT&T, Sprint, T-Mobile and Verizon - often must license thousands of patents to provide any one consumer product" Gridlock Economy, Michael Heller, Basic Books 2008, p100); 90,000 for microprocessors and 420,000 for semiconductors and systems ("One panellist stated that more than "90,000 patents generally related to microprocessors are held by more than 10,000 parties." Likewise, he reported, there are approximately 420,000 semiconductor and systems patents held by more than 40,000 parties." To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, supra, n 24). #

17 61. The report commits a number of factual and economic errors in seeking to portray patent quality as weak. 62. Firstly, the criticism that the European Patent Office ("EPO") grants weak patents, based solely on statements attributed to generics, is regrettable and unwarranted. 62 The only fact said to support this finding, that the "majority of litigated patents were revoked," is simply wrong. Only 28.3% were invalidated. 63 The EPO is generally recognised as a benchmark of excellence for patent offices around the world. 64 It regularly reviews its practices so as to seek to ensure that patents issue only for genuine innovations satisfying patentability criteria Second, the report mistakenly seeks to generalise about patent quality from a small sample of litigated patents. Such sampling is well recognised as prone to "self-selection error." 66 The sample chosen is not a random one, but will rather comprise only the most contentious patents. The FTC warns against precisely this error, noting that it is accordingly "unlikely that patents litigated to final results are fully representative of patents as a whole." The 92 cases in which innovators have lost to generics in litigation (62% of the 149 litigated cases) 68 therefore provide no indication of the general quality of the 26,630 patents, 69 nor of the approximately 35,000 "secondary" patents/ applications referred to in the report. 70 Generics will only provoke litigation through market entry or seek non-infringement or invalidity declarations where they consider they have a prospect of successfully challenging relevant patents and the sums at stake make challenge worthwhile. The rest will never be challenged or the parties will prefer to resolve litigation through settlement. 65. With this in mind, the 38% of all cases won by innovators, or 51% of cases they initiated, are unsurprising and in line with other industries. 71 Indeed, bearing in mind that generics study carefully which patents to provoke litigation on, even lower win rates for innovators might PR, para 393. PR, footnote 257 (28.8% of patents are invalidated, being 27.5% invalid and 1.3% invalid and non-infringed, the remainder are non-infringement findings). It is also wrong if only "secondary" patents, as the report defines them, are considered. PR, footnote 262 (36% invalidated, being 34% invalid and 2% invalid and non-infringed). Future workload, EPO Administrative Council Board 28, November 2007, p. 2 ("[T]he European patent system is generally perceived as one of the best patent systems world-wide which provides for well balanced and high quality results."). With patent applications on the rise, European Patent Office puts quality before quantity, EPO press release of 1 April 2008; EPO and USPTO to Pilot Patent Prosecution Highway, EPO press release 28 April 2008 (EPO and USPTO streamline and improve procedures by sharing information); Future workload, EPO Administrative Council Board 28, November 2007 (New procedures to focus on patent quality, more effective use of third party information, strengthening international cooperation in searches, workload reduction, procedural and institutional reforms.). See, eg, Mathematical Statistics, 7th Edition, Freund, p264 ("[N]ot all samples lend themselves to valid generalizations about the populations from which they came."). To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, supra, n 24, Chapter 5, p8 ("The result [invalidity of litigated patents] should be interpreted with caution: self-selection in bringing and settling suits makes it unlikely that patents litigated to final results are fully representative of patents as a whole."). PR, para ; most litigants would regard a judgement on the merits followed by a settlement as to quantum as a "win" for the claimant. Adding as a "win" the approximately 50 settlements of this kind referred to at paras. 585 and 630 would significantly change this statistic. Dividing the 92 cases by the number of patents granted reveals that in only 0.35% of cases have innovator companies failed to successfully defend their intellectual property rights. Such a statistic can hardly be indicative of the "weakness" of pharmaceutical patents in general. PR, paras and 514. Global Patent Litigation: Win Rates and Strategies, FICPI Presentation, Michael Elmer, 31 May 2007 (an informal survey, in the absence of official data, suggesting win rates for innovators of between 26% and 55% depending on jurisdiction). #

18 have been expected. It may be noted that the report's data on judgments as to the validity of litigated pharmaceutical patents (72% held valid according to the report) are rather better than US studies of litigated patents across all industry sectors (54% held valid), 72 though allowances must be made for the different time periods and procedures involved. 66. Finally, stricter grant rates for pharmaceutical patents (33% compared to the 50% EPO average) 73 and higher levels of opposition (8% compared to the 6% average) are also no indication of poor quality. 74 Firstly, only 4% of non-grants are refusals, the rest are voluntary withdrawals which may well reflect lack of further commercial interest rather than patent office objections, bearing in mind the high numbers of compounds that get discarded early in the R&D process. Secondly, if anything, stricter grant and higher opposition rates suggest rigorous scrutiny of pharmaceutical patents ensuring the high quality of those patents that endure. (d) The patenting lifecycle (Section C.2.1.1) 67. The report finds examples of patents filed near to expiry of the "primary patent" protecting a drug, the suggestion being that this somehow extends patent protection. 75 There is no legal basis to this suggestion. 68. The trends found in the patent lifecycle illustrate clearly the continuum of inventive development that one would expect during a product's life. 76 More inventions are developed over time and those that meet the patentability requirements are patented. 77 After a drug enters phase III trials, any changes might require additional trials, such that it would be standard to wait for successful launch before examining, say, a new product formulation, process or indication. Equally, after a product is launched, companies routinely revisit the production process, potentially resulting in the development of a patentable second generation process. 69. That companies continue to explore the potential for new developments relating to the high value medicines within their portfolio well after product launch is a rational economic strategy. They focus their efforts on seeking to improve their most successful products where there tends to be the greatest unmet medical need rather than those for which there is less demand. No rational firm would behave otherwise, though the report does not consider this reason for the observed trend. 70. Far from showing a general spike in patents at the time of "primary" patent expiry, the report identifies only one such example, evidence in itself that there is no such general trend. 78 Rather, the report finds 84% of applications are filed during the research phase, 79 a "steady increase in the number of applications" over the product's lifetime for high value medicines, 80 and patented post-launch incremental improvements resulting from additional trials and PR, footnote 257; "Empirical Evidence on the Validity of Litigated Patents" John R. Allison, p205. PR, paras. 233 and 346. PR, paras. 552 and PR, para. 377, Figure 53 and para PR, para. 357, Figure 48. PR, para 392 (acknowledging the importance of incremental innovation during the product's lifecycle). PR, para. 360, Figures 49 and 50. PR, para PR, para #

19 R&D. 81 In any event, there is no suggestion that patents applied for at the time of "primary" patent expiry are anything other than legitimately applied for and/or obtained and no suspect conduct can be inferred from their timing. The occurrence of a new invention relating to a product, regardless of when it occurs, must meet the same tests of patentability as at any other time. Delaying patent applications for known inventions to later in the lifecycle would be counterintuitive. Companies have every incentive to file to protect an invention as soon as they can. Delaying risks losing any chance of patent protection if a rival discovers and applies to patent, or publishes, the same invention. 82 (e) Divisionals (Section C.2.1.3) 71. The stated concern that divisionals "serve to prevent or delay generic entry" though it seems the prevalence of this concern is very limited 83 again has no legal basis. Divisionals are a necessary part of the patent system, allowing a claim to be split off from a parent application as a separate application in its own right. 84 They can be filed for any number of wholly legitimate reasons, such as disunity objections or to allow rapid grant for a specific claim held patentable by the EPO (while the remaining claims are split off). 85 The exclusive rights resulting from a divisional can last no longer and be no wider than the parent, so the generics' ill-articulated complaints of delay or uncertainty are not borne out by the position in law. 86 (f) "Intended effects" cannot distinguish between legitimate and "suspect" patenting practices (Section C.2.1.4) 72. The suggestion that intended effects can somehow distinguish good and bad patenting practice must also be rejected. The report appears to take this line - though the sense of the paragraph is hard to follow - when it states "while [generic exclusion], during the period of exclusivity, is generally in line with the underlying objectives of the patent systems, it may in certain cases be aimed at excluding competition and not safeguarding a viable commercial development of own innovation." 87 Any such theory is meaningless in the patent context for obvious reasons. Patents exclude, and intend to exclude, competition by preventing copying of the protected inventions by generics or any third party. That is their function. 88 It cannot be a basis for critical appraisal. It is the essence of how they protect and stimulate innovation. (g) The chilling effect of any rule that seeks to constrain patent portfolios 73. Any suggestion that patent portfolios are constrained by competition law, contrary to prevailing precedent, would be hugely damaging to innovation: PR, para. 126 (recognising incremental innovations that occur post launch: "further R&D (incremental innovation) aimed at improving the medicine or finding new uses is frequently conducted by originator companies during this phase. New patent applications can be therefore submitted at this [post launch] stage"). PR, para. 389 (amply illustrated by the quotation ascribed to an innovator "Don't play games in patenting new salt forms too late, the generics are starting earlier and earlier."). PR, para PR, paras Divisionals can also be used: (i) to obtain claims of narrow scope so that broader scope has divided out; (ii) as a safety device ahead of an ex-parte appeal in case the case could be rectified in principle, but, due to procedural restrictions, can't be rectified at the oral proceedings; (iii) before grant in the expectation of an opposition so that there is scope for pursuing different subject-matter if issues unexpectedly arise during the opposition that can be or are most easily rectified pre-grant in the divisional application; (iv) for entitlement reasons, for example different subject matter may be owned by different parties. PR, para PR, para PR, para The ECJ has recognised this as the essence of the subject matter of a patent in case law since the 1960s. #

20 An enforced narrow patenting or enforcement strategy that limits an innovator's portfolio would invite free riding on the original, published, invention and deter innovation across all industry sectors; It would be antithetical to innovation for antitrust rules to prevent or inhibit an innovator from using its superior knowledge of its own products, and substantial sunk R&D costs, to create improvements upon them. The innovator will necessarily be best placed to capitalise on its invention and to devise new and valuable improvements. Denying this source of innovation to the market would, to the contrary, be an irrational competitive restraint, denying patients access to beneficial improvements Such a position is rightly not a target of EC competition law. 90 Any antitrust rule that mandated narrow patenting strategies or limited the ability to enforce patents would require the competition authority to substitute its view for the innovator and expert patent office of what constituted a "good" or "bad" invention, an approach rejected by the OECD as unworkable. 91 Such a role would usurp the patent office's function and the proper functioning of antitrust regulation, to preserve, but not to control, efficient markets. A requirement that a patented invention not only be new, inventive and capable of industrial application, but also that its exercise should not be found anticompetitive at some ill-defined future moment, is manifestly unworkable and undermines any meaningful investment protection. 3.2 Patent Litigation (Section C.2.2) 75. It is unclear why the report should wish to label the use of this essential and only means of protecting intellectual property pejoratively as "aggressive". Forthright statements in company documents as to the need to protect intellectual property are absolutely legitimate and to be expected in the face of a generics industry whose dedicated business model is to copy the successful products of innovators. 92 The report cites no evidential basis for asserting that innovators knowingly assert meritless claims. 93 (a) Win/loss litigation statistics (Section C ) 76. As noted above (section [3.1(d)]), statistics on litigated cases do not in any way substantiate that types of patents are generally "weak" 94 or that "the vast majority" of claims are unsuccessful for innovators. 95 A 40-50% win rate for innovators (38% overall and 51% when See Patents and Pharmaceuticals, supra n 23, where Lord Justice Jacob references T.A. Blanco White: "The original manufacturer's position need not deteriorate with further lapse of time: he should always be some years ahead in design so far as patentable improvements are concerned, while the longer he keeps the field to himself the greater the advantage he has in manufacturing experience." Case T-83/91 Tetra Pak International SA v Commission [1994] ECR II-755, para. 242; Automatic Radio Mfg. Co. v Hazeltine Research, Inc. 339 US 827, 834 (1950) ("Mere accumulation of patents, no matter how many, is not in and of itself illegal."). Competition, Patents and Innovation, OECD 2 January 2008, page 36. ("For several reasons, such as a lack of relevant technical expertise and limited resources, it does not appear to be prudent for competition authorities to assume responsibilities related to the initial review of IP applications."). PR, paras PR, paras. 432, 893 and p166 (the quote allegedly supporting this assertion relates to the prospects of obtaining interim relief, not to the underlying merits of the claim). PR, para. 514 (wrongly asserting weakness of process, first and second medical use patents). PR, para. 893 (wrongly stating that "although the majority of legal actions in the EU in recent years were initiated by originator companies, the generic companies won the vast majority of them." To the contrary, the majority of innovator initiated cases were won by innovators.). #

21 innovator initiated) is entirely to be expected when one considers that only the most contentious cases litigate and their outcome is a matter of huge legal and factual uncertainty. 77. Nor do the statistics show unwarranted generic delay since ongoing litigation is no bar to generic entry absent interim relief. The report does not say in how many cases, if any, an innovator who obtained injunctive relief, lost at trial. 96 Contrary to the report's characterisation of interim relief as having "very serious" consequences for the generic, 97 generics are typically entitled to claim damages for lost profits if the innovator loses at trial. The consequences of generic entry for the innovator can be far more serious, not least due to the impact on domestic and international reference pricing for its products, such that it may sustain incalculable, and irrecoverable, losses and the market for its products may never recover. 98 (b) Opposition (Section C ) 78. EFPIA agrees that the time taken by opposition procedures can be lengthy and this can be highly unsatisfactory. It objects to the allegation that lengthy opposition is due to innovator delay. The only "evidence" offered is generics' hearsay. 99 Filings in opposition proceedings must be made to a timetable. There is no analysis of what, if any, procedural positions taken by innovators were inappropriate, to whom procedural delays generic, innovator or EPO may be ascribed or, indeed, the tactical advantages to generics of bringing or delaying opposition, for example, as a means of persuading national courts that interim relief should not be granted during a pending opposition. (c) No analysis of path clearing strategies (Section C ) 79. The alleged impact of patents or patent enforcement cannot be established without analysis of path clearing strategies by generics. 100 The patent system, properly used, provides ample opportunity for generics to revoke or seek non-infringement declarations in respect of blocking patents. The high number of generic initiated suits (46%) suggests that generics are prudently using judicial techniques to clear the path prior to entry. 101 (d) Antitrust does not and should not curtail the proper use of litigation to protect rights 80. The EC courts have never found an infringement of the competition rules based on a claim of vexatious litigation. In ITT Promedia, the Court of First Instance left open the question of whether any limitation by competition law was appropriate, given that access to justice constitutes a fundamental right that can only be circumscribed in "exceptional circumstances." 102 It examined only whether the test advocated by the Commission has been properly applied, namely held that a claim made in litigation could not be illegal unless (1) undertaken by a dominant company, (2) where it "cannot reasonably be considered to be an It may be noted there are similar rates of success for innovators in obtaining interim relief (44%) and winning final judgment (38%). PR, para In passing, it may be noted that fear of damages denotes fear of losing in litigation, suggesting that the generics surveyed may have known the speculative nature of their patent challenges. See, eg, EFPIA submission of 13 June 2008, paras. 166 and 184. PR, para These were analysed in detail in the EFPIA submission of 13 June 2008, paras PR, para. 468 and Table 62. Case T-111/96 ITT Promedia NV v Commission [1998] ECR II-2937, para. 60. #

22 attempt to assert the right of the undertaking concerned and can therefore only serve to harass the opposing party", and (3) is part of a plan to eliminate the competitor National courts have their own methods for dealing with vexatious litigation. 104 This can be by way of summary judgment that there is no arguable case, by way of striking out a claim for disclosing no cause of action, by way of costs awards or a separate actionable claim. 82. In relation to patents, not only have they been subjected to the prior scrutiny of the patent office, but also generic manufacturers have a risk free opportunity to seek pre-emptively to challenge allegedly invalid patents by way of opposition procedure or to claim noninfringement or invalidity in declaratory actions in national courts. If they choose not to avail themselves of these mechanisms, they bring upon themselves the risk of interim or final injunctions and damages by launching "at risk." The application of competition law is not appropriate in circumstances where the patent system has already undertaken an assessment of validity, such that the patent holder clearly has a reasonable basis for asserting enforcement and, moreover, the patent system has ample remedies for pre-emptively addressing invalid or non-infringed patents. In such cases, fear of ex-post second guessing by antitrust regulators should not prevent innovators from legitimately asserting their intellectual property rights. Effective IP protection is an essential prerequisite to all innovative activity. 106 As the Commission has stated, "[t]he mere possession and enforcement of a patent against a competitor does not infringe competition law." Settlements (Section C.2.4) 84. The report suggests that certain "arrangements could fall within the scope of competition rules" and result in loss to consumers. 108 These patent settlements, categorised as BII in the report, involved payments (in money or other value transfer) to a generic and generic agreement to stay off the market. The implication is that such settlements could raise concerns because they mask an agreement to exclude a competitor rather than a bona fide settlement. However, the correct legal and factual analysis shows that no general antitrust prohibition can be based on this categorisation. (a) Categorisation of agreements neither aids antitrust analysis nor discloses prima facie antitrust concerns 85. The categorisation of settlement agreements by reference to specific terms, namely "limitation on entry" and "transfer of value", does not assist the antitrust analysis. Indeed, using the definition of "limitation on generic entry" prejudges the outcome. It presupposes the generic would otherwise be free to enter unfettered by patents. It assumes absolute knowledge of the Case T-111/96 ITT Promedia NV v Commission [1998] ECR II-2937 para. 55. ITT Promedia v Commission: Establishing an Abuse of Predatory Litigation? S. Preece [1999] ECLR 118. See, eg, SmithKline Beecham v Apotex [2003] FSR 524 ("The defendants have known for a long time about this patent. You would have to be very naïve in the pharmaceutical industry to think that the patentee, with a product as important as this would not, if it had anything other than a frivolous chance of success, take action. That is what the procedures for revocation and declaration of non-infringement are for."). Recital 3, Directive 2004/48/EC on the enforcement of intellectual property rights [2004] OJ L195/16 ("[W]ithout effective means of enforcing intellectual property rights, innovation and creativity are discouraged and investment diminished the means of enforcing intellectual property rights are of paramount importance."). Commission Decision COMP/A /F3 AstraZeneca (15 June 2005), para PR, para #

23 patent position which, by definition, simply does not exist in a settlement. Of necessity, settlements reflect a negotiated position. There is no decisive outcome. 86. In fact, none of the allegedly suspect BII agreements appear to exhibit prima facie antitrust concerns. Of the 23 BII agreements involving a generic payment and limiting generic entry, 13, by means of licensing or supply arrangements, in fact facilitated entry. 109 A further six limited entry only while a test case was brought to determine the merits of the case. 110 The remaining four involved sundry payments for legal costs, innovator's repurchasing of potentially infringing stock or marketing contributions (for example, where a supply/licence has been agreed). (b) The appropriate framework for antitrust analysis 87. Competition law analysis must start from the position that, absent final adjudication, it cannot be shown that the potentially infringing generic was an actual or potential competitor of the innovator. 111 The presumption must be that the generic is blocked from competing by the asserted patents. 112 Any restriction in a settlement agreement that goes no further than the exclusionary effect of the patent cannot therefore be anticompetitive There can be no per se condemnation of specific settlement terms. The report singles out terms involving non-entry and value transfer to the generic. The implication is that legitimate settlements should reflect a binary outcome. Either it favours the generic (providing for entry and/or damages to the generic) or the innovator (providing for exclusion of the generic and/or damages for the innovator). But such an approach is misplaced, for two reasons. 89. First, it misunderstands the dynamics of settlement. By definition each party must secure a benefit that makes settling more attractive than final judgement. In every case where the generic agrees to restricted entry it will, acting rationally, also have secured a benefit that makes restriction on entry through settlement more attractive than further litigation which could lead to the same result. In many cases, that benefit will be avoidance of the costs and risks of liability in continued litigation. However, it may also extend to other benefits such as transfers of value or other commercial dealings. None of these benefits are presumptively illegal. They are the bargain inherent in every settlement. 90. Secondly, in complex, uncertain cases, especially where the parties have very different views on the risks or merits of the action, settlements may involve any number of commercial arrangements designed to achieve mutually acceptable terms, including payments or other PR, paras If all "value transfer and limitation" cases are taken into account approximately 84% facilitated entry via licensing or supply (PR, para. 634). Where interim relief is not available or unlikely to be granted, a later damages award will often fail to compensate for the harm caused to the innovator's business by generic entry, even if the generic could afford to pay. Restoring the market to the pre-generic entry price will generally be impossible even if the generic loses. Agreements to await ruling on the merits, with interim compensation to the generic, serve the same purpose as interim relief, and may be reasonable to avoid potentially unquantifiable and irrecoverable losses in the market pending trial were the generic to enter pending trial. PR, para. 596 (explaining the counterfactual from the point of view of the generic company, "when deciding what type of settlement agreement to conclude, we [a generic company] aim at obtaining the earliest possible entry date with a reasonable degree of certainty, weighing the considerable risk that continuing the litigation would result in us being excluded from the market for the entire patent term."). TTBER Guidelines, para. 32. (where a party believes it has "good and valid reasons to believe that a blocking position exists," the presumption must be that there is a block to actual or potential competition). An Economic Assessment of Patent Settlements in the Pharmaceutical Industry, Brett Dickey, Jonathan Orszag & Laura Tyson, 2-29, p4. #

24 benefits to the other party. Again, these are not suspect practices, they are legitimate means to secure settlement in conditions of uncertainty and differing expectations Any other approach would be untenable. A requirement that parties litigate to final judgment (or must refuse settlement if benefits flow to both parties) for fear of being accused of using illegal settlement provisions would itself engender anticompetitive consequences. It would cause uncertainty in the parties' investment decisions in the months and years that the case is ongoing. It wastes resources and management time in pursuing litigation which would be better devoted to innovative activity. 115 It would be contrary to the public policy of encouraging settlement for these reasons and to make efficient use of court resources. 116 (c) US agency practice finds disfavour in the US courts and is an inappropriate model for the different regulatory regime in the EU (Section C ) 92. The report goes on to examine the FTC's enforcement practice. 117 Notwithstanding the FTC's Cephalon complaint by no means a "decision" as the report states the recent US judgments do not support the contention that "reverse payments" are presumptively anticompetitive. 120 Rejecting submissions, inter alia, from the FTC challenging "reverse payment" type provisions, the Court of Appeals for the Federal Circuit confirmed that settlements that go no further than the patent do not infringe competition law Incorrectly finding "more similarities than differences," 122 the report seeks to downplay the differences between the EU and US systems of patent litigation, in particular the step change in litigation incentives brought about by the Hatch-Waxman Act. 123 The report is mistaken, however, to state that there are parallels in relation to the frequency of settlement agreements. The US is one patent jurisdiction, the EU has 27. Compared on a per jurisdiction basis, the number of settlements in any EU country is a fraction of those in the US There are, to the contrary, significant differences with the US system. These include incentives to litigation for generics without being at risk of damages, the six months generic exclusivity offered to first filers and the far higher value of the single US market for drugs. 114 "Red Flag or Red Herring? Reverse Payments and the Settlement of Pharmaceutical Patent Litigation", Mark William Murphy, European Comp. Journal, Vol 4 No. 2, December 2008, , p See Supra, n 113, p27 ("To the extent that limits on patent settlements reduce incentives to invest in pharmaceutical R&D, consumers may suffer significant adverse effects in the long-run, in the form of a smaller number of new medicines that become available."). 116 See, eg, UK Civil Procedural Rules (which have "firmly recognised the public policy interest in avoidance of litigation."); Section 278 German Code of Civil Procedure (Court directed to invite consideration of settlement possibilities); French Report N 3696 on "Mediation in Europe", 13 February 2007 (Endorsing mediation and settlement options as alternatives to litigation). 117 EFPIA refers to its submission of 13 June 2008 for a fuller analysis of the position under US law. 118 Federal Trade Commission, v. Cephalon, Inc., Civil Action: 08-CV-2141-RBS. 119 PR, para In re Ciprofloxacin Hydrochloride Antitrust Litigation, No , 2008 WL (Fed. Cir. Oct. 15, 2008). 121 Id, at 14 ("[T]he essence of the Agreements was to exclude the defendants from profiting from the patented invention. This is well within Bayer's rights as the patentee. Furthermore, there is a long-standing policy in the law in favor of settlements Settlement of patent claims by agreement between the parties including exchange of consideration rather than by litigation is not precluded by the Sherman Act even though it may have some adverse effects on competition."). 122 PR, para Drug Price Competition and Patent Term Restoration Act of PR, para. 657 and Figure 111. #

25 The US courts themselves cite specificities of the Hatch-Waxman Act as the cause of much of the criticism of settlements in the US. 125 (d) Conclusion 95. Settlements, for financial or other benefits, are an essential part of business. Companies must be allowed to settle, move on and continue to invest in the conditions of legal certainty that settlement allows. Courts actively encourage settlement for these and other reasons. Second guessing of settlement terms by competition regulators undoes these business benefits. Blacklisting one or other term of a settlement is inappropriate, given the highly fact specific nature of litigation. Each needs to be considered on its merits. The FTC's approach is not appropriate in light of the very different EU regulatory environment and the FTC's lack of success before the US courts. 3.4 Regulatory Complaints and Litigation (Section C.2.5.1) 96. The report notes interventions before regulatory authorities and suggests that in some cases such interventions could not be substantiated before national courts and/or may have led to regulatory delays for generics. The report fails to assess or attribute causation. It does not examine the essential nature of companies' dialogue with regulators, nor, from the handful of cases considered, establish any basis for finding such conduct to be a "standard" blocking tactic or indeed anything other than companies' bona fide concerns. (a) The need for interventions before authorities (Section C.2.5.1) 97. As in any regulated industry, prompt recourse to the regulator in the event of concerns is essential and, in many cases, a duty for pharmaceutical companies. 126 Regulators and the medical community expect that the innovator company remains vigilant and responsible for the molecules it originates, even when the medicine is sold by generics. 98. The risks associated with substandard or counterfeit products are extremely serious as recent examples attest. The FDA recently banned imports of 30 Ranbaxy generic medicines into the US citing manufacturing concerns. 127 Counterfeit medicines now account for 1% of medicines in industrialised countries, according to WHO estimates, and there are now instances of counterfeits entering the legitimate supply chain. 2.5m counterfeit medicinal products were seized at EU borders in "Ghost" generics generics granted provisional authorisation without full review by new Member States just prior to accession so as to benefit from transitional provisions allowing them to be marketed appear to be In re Ciprofloxacin Hydrochloride Antitrust Litigation, No , 2008 WL (Fed. Cir. Oct. 15, 2008)., at 17, (Declining to follow In re Cardizem CD Antitrust Litigation, 332 F.3d 896 (6th Cir. 2003) which found a "reverse payment" per se unlawful, finding the case law turned not on the existence of the reverse settlement per se, but on the fact that "the settlement in that case included, in addition to a reverse payment, an agreement by the generic manufacturer to not relinquish its 180-day exclusivity period [a key feature of the Hatch-Waxman Act], thereby delaying the entry of other generic manufacturers." emphasis and annotations in square brackets supplied). See, eg, Articles 23 and 98, Directive 2001/83 on the Community code relating to medicinal products for human use [2004] OJ L 311/67. FDA imposes Ranbaxy import ban, Financial Times, 18 September Impact Assessment for the Proposal for a Directive amending Directive 2001/83/EC as regards the prevention of the entry into the legal supply chain of medicinal products which are falsified in relation to their identity, history or source, 10 December 2008, SEC(2008) 2674, p9. #

26 government policy in certain new Member States. 129 transposition of EU market exclusivity legislation. 130 Two new Member States are resisting 99. Though the report objects to what it calls "patent linkage", Member States have a right to know whether a product is at risk of disruption of supply because of patent protection. Some routinely ask for this information or their laws require it. 131 The report's statement that patent protection is irrelevant to pricing and reimbursement under EC law is misleading. 132 If national authorities wish to reward innovation based on patent status they are free to do so. National pricing and reimbursement rules are not harmonised by EC law except for the aspects governed by the Transparency Directive 89/ Whether the product at issue is on patent or generic is often highly relevant to the price setting process National patent laws and courts may also reach a different view from the report as to whether facilitation of sale by a marketing authorisation authority can be considered to violate the patentee's rights. This is particularly the case in countries where patent protection is a constitutional right and/or protected by criminal law. 135 National laws may therefore treat this as a separate and unharmonised legal question, quite different from the determination of whether marketing authorisation may be granted. EC law makes clear that abridged approval procedures for generics are "without prejudice to the law relating to the protection of industrial and commercial property." The innovator is likely to be best placed to identify these concerns and to alert the authorities. If innovators have bona fide concerns, they must be entitled to act. It is for the authorities to decide whether those concerns are justified and competition law should not inhibit this channel of communication which could have serious patient safety implications. (b) Complaints are exceptional not "a standard instrument" (Section C ) 102. Though the report alleges they are "a standard instrument in the strategic tool-box", 137 of necessity, regulatory complaints around substandard generics, lack of bioequivalence, counterfeit or patent rights are exceptional, 138 as the report goes on to admit. 139 They will Commission reasoned opinion of 18 July 2007 to Poland in case 2006/4822 (Ghost products/illegal copies of products); notice of proceedings in Case C-385/08, Commission v Poland [2008] OJ C 313/18. Hungary and Poland have failed to transpose the market and data exclusivity rules by the required deadline of November 2005 under Regulation 726/2004 and Directive 2004/27; Latvia initiated the necessary legislative process in PR, paras. 716 and 1157, Tables 22 and 36; Indirectly, Article 3(3)(b) Regulation 725/2004 also involves communication of the patent position to the marketing authorisation authorities, since it requires that the Summary of Product Characteristics for national authorisation is the consistent with that granted centrally, "except for those parts still covered by patent law." PR, para. 758 ("Under EU law, patent protection is not a criterion to be considered by the authorities when approving prices or granting reimbursement status."). Council Directive 89/105/EEC relating to the transparency of measures relating to the pricing of medicinal products for human use and their inclusion in the scope of national health insurance systems, OJ No L40,11/02/1989 p8 Report for the DG Competition: Surveying, Assessing and Analysing the Pharmaceutical Sector in the 25 EU Member States, Österreichisches Bundesinstitut für Gesundheitswesen, July 2006, see, eg Belgium (p71), Poland and Denmark (p151). See, eg, Recital 13, Regulation 726/2004 (allowing derogation from sole consideration of quality, safety and efficacy criteria so that "Member States should be able exceptionally to prohibit the use in their territory of medicinal products for human use which infringe objectively defined concepts of public policy and public morality."). Article 10.1, Directive 2001/83/EC on the Community code relating to medicinal products for human use (as amended) [2001] OJ L 311/67. PR, paras. 707 and interventions over 8 years in relation to 219 INNs in 27 Member States is an average of per INN, per Member State, per year. #

27 be highly case specific and the result of particular circumstances. This is well illustrated by the report's findings that 40% of interventions related to just four products To the contrary, the data show that far from being a "standard" blocking tactic used to protect high value medicines at LOE, the concerns expressed to authorities are evenly spread between products with LOE over the inquiry period (the "E75") and those that are not. 141 (c) Win/loss statistics provide no basis for antitrust analysis in relation to such legally and factually complex matters 104. Litigation statistics are no basis for asserting that concerns raised are generally unsubstantiated, 142 as the report suggests. It is as meaningless to say that such claims are unsubstantiated based on win/loss statistics favouring generics in a tiny selection of litigated cases (24 litigated cases on patent or effectiveness claims 143 or 18 cases on data exclusivity 144 ) as it is to say that they are always well founded based on statistics favouring innovators in P&R cases (5 cases suspending P&R status of generics) 145 given the heterogeneity of fact and law involved as well as complexity of standing and procedure involved in judicial review. 146 (d) The dangers inherent in restricting companies' access to the regulator or courts 105. Raising bona fide concerns with a regulator should not be subject to ex post antitrust scrutiny. Access to regulators and officials is essential to the democratic process. 147 Recourse to judicial review is a fundamental right. 148 Since the producer of the branded drug will often be in the best position to assess quality or safety issues, for example, relating to a generic drug, it is particularly important that they should not be discouraged from making such observations. Authorities themselves are able to assess the legality of the procedure at issue and guard against potential misuse, and this is the case regardless of the intent of the complainant PR, para. 709 ("Although the total number of such pre-litigation contacts compared to the overall number of applications for marketing authorisations submitted in Europe every year appears to be limited."). PR, para See also PR, para. 719 (80% of complaints about generic quality relate to 6 products), paras. 723 and 731 (data exclusivity complaints related to 6 products and related litigation to only 10) and (litigation on reference pricing relating to four products). PR, para PR, pp12 ("When the patent-related matters resulted in litigation, the claims of the originator companies were upheld in only 2% of the cases, suggesting that the arguments submitted against the generic medicine could not be substantiated") and 294, para Note the inconsistency between p12 and para. 727/Figure 118 with the former said to relate to patent matters, but the latter to both patents and safety concerns. PR, para. 727, Figure 118. PR, para. 734, Figure 119. PR, para. 765 (seven settlements, six pending, five suspensions and one injunction). The report notes 195 "interventions" (para. 710), 211 concerns about quality (para. 718), 28 about data exclusivity (paras ), 137 cases of litigation on patent infringement or quality (para. 726), 43 cases of litigation on data exclusivity (para. 729), 70 complaints to P&R authorities (para. 751). AG Jacobs Opinion, Case C-67/96 AlbanyInternational BV v Stichting Bedrijfspensioenfonds Textilindustrie ECR I ; AG Leger Opinion, Case C-35/99 Arduino ECR I -1529, para.74. Case T-111/96 ITT Promedia NV v Commission [1998] ECR II-2937; Article 6, Convention for the Protection of Human Rights and Fundamental Freedoms. Entirely lawful regulatory procedures often have as their intended result that barriers to entry are raised or competition restricted for important public interest reasons (Case T-83/91 Tetra Pak International SA v Commission [1994] ECR II- 755, para. 242 (patent applications necessarily, but wholly lawfully, exclude competition); Case T-111/96 ITT Promedia NV v Commission [1998] ECR II-2937, paras (Litigation aimed at excluding competition lawful if reasonably asserted) ;Case T-5/97 Industrie des Poudres Sphériques v Commission [2000] ECR II-3755 (antidumping complaint lawful regardless of the impact on foreign competitors) ; AG Jacobs Opinion, Case C-67/96 AlbanyInternational BV v Stichting Bedrijfspensioenfonds Textilindustrie ECR I ; AG Leger Opinion, Case C- 35/99 Arduino ECR I -1529, para.74 (lobbying lawful even with intent to exclude)). #

28 3.5 Marketing and Next Generation Products (Section C ) 106. The report's attempt to characterise the marketing or timing of the launch of next generation products as suspect is unsuccessful. The report offers selective data about marketing that may be intended to give the impression that the pharmaceutical industry does not invest in R&D. It states that more is spent on marketing (23% of sales) than R&D (17% of sales) and only 1.5% of sales is "basic research." Any such suggestion must be rejected. The industry is the R&D leader in Europe, spending almost twice as much as the software and computer services sector. 150 It may be compared with companies in the fast moving consumer goods sector who spend 36% of sales on marketing and only 2% on R&D. 151 Whatever the report may mean by "basic research", every Euro of the 17% of sales spend on R&D is essential to bring potential medicines through the year R&D cycle to market. Interestingly, the PR does not show the breakdown of turnover for generic companies although those data have been collected. 152 As was observed at the hearing on 28 November 2008: "The most important figure in the Preliminary Report is that current R&D costs are 17% of turnover. That is much higher than most if not all other industries. I am sorry that the interim report somewhat belittles the figure by adding that only 1.5% is "basic research". I am not sure I know what is meant here, but it does not matter. It is no good finding a potential suitable molecule. Unless you spend the money and time to find out if it really works and is safe, it is no use. Assuming the figure is indeed 1.5% it would come to nothing without the remaining 15.5%." Lord Justice Jacob, 28 November It is certainly true that the medical community must be made aware of the relative clinical benefits of the latest products. Communicating fully and accurately the risks and benefits of the latest and often complex medicines to doctors, and obtaining their feedback to inform future improvements, is a resource intensive process. This is part of ensuring that patients get the right medicines Neither marketing, in general, nor the promotion of next generation products is illegal under EC competition law. Marketing is an essential part of competition and better quality healthcare Next generation products, with greater potency, fewer side effects, broader therapeutic indications or reduced dosing regimens, offer competing and improved product choices. If successful, they will be prescribed and gain share. If unsuccessful, the market will prefer the first generation product. The next generation product will have to prove its additional merits on the market This is the essence of a competitive marketplace. Indeed, the report notes without adverse comment that exactly these marketing practices are undertaken by generics to patent new formulations and versions of existing product lines to differentiate themselves from European Commission 2007 EU Industrial R&D Investment Scoreboard, October 2007, p. 12. (Compare software and computer services 9.7%. The scoreboard considers the world's top 1,400 companies with R&D spend over a minimum threshold comprising the top 400 EU and 1,000 non-eu companies.). The top 10 fast moving consumer goods companies spent 36% of revenues on marketing (source: 2007 annual reports). PR, para. 81. PR, para #

29 competitors. 154 The attempt to criticise this practice when conducted by innovators, but not by generics indicates a lack of understanding of the nature of competition in this and other high technology industries: (i) (ii) (iii) (iv) (v) The theory that only old technologies should compete - It would be absurd to say that only the first version of software, cars, mobile phones or laptops could be launched, while any attempt to offer upgraded versions during the lifetime of the original IP protection would be considered an anticompetitive attempt to exclude rivals who offer only older versions. In the same way, there can be no criticism of pharmaceutical companies launching improved versions of their products regardless of timing; The allegedly non-innovative nature of products - There is no basis for the report's assertion that "[s]trategies for launching follow-on products can in some cases be rather aimed at preventing generic companies entering the market than actually protecting a new innovative product." 155 Antitrust does not regulate the type of products, whether original or next generation products, that may come to market. The market, in this case doctors and patients, decides whether or not they show value over competing or older technologies; 156 The allegedly suspect timing of introduction of new technologies when old technologies reach the end of their lifecycle - The statement that it is suspect for the launch of the new product to coincide with the obsolescence of the old product is contrary to economic principles. 157 Though many innovators will introduce next generation products at various stages of the first product lifecycle, the spur to competition through innovation is of course most acute when potential competitors are about to enter the market. The report recognises just this stimulus to competition created by the time limited term of the patent system; 158 The so-called evidence of suspect conduct shows only intense competition - The report casts doubt on statements showing the keenest possible awareness of impending competition and the efforts of the innovator to compete through new product lines: "the most effective strategic initiative to counter generic [versions of the first generation product]; [ ]" "a new formulation to maintain the growth momentum of the product." 159 It is difficult to understand how DG Competition could read these extracts in any other way or, indeed, whether it would prefer to see competitive complacency between innovators and generics; The allegedly suspect focus on high value molecules - Well established economic principles mean that, naturally, innovators will focus on creating new products to PR, para. 364 ("[G]eneric companies regularly file applications for secondary patents on many of the INNs covered by the inquiry. Moreover, these secondary patent applications relate, inter alia, to manufacturing processes, formulations, polymorphic crystalline forms, salts, particle sizes and tablet forms of INNs for which the primary patent is held by an originator company."). PR, para EFPIA refers to its detailed submission on the nature of innovation and incremental innovation in its submission of 13 June 2008, para. 49 and Annex 2. PR, paras. 832 ("[T]he circumstances typically associated with the introduction of follow-on products to the market suggest that the latter often form part of originator companies' broader life cycle strategy attempting to delay or prevent generic erosion due to the loss of exclusivity of the first product. Most notably, switches to follow-on products frequently occur when generic competition is imminent.") and , Table 128. PR, para PR, paras #

30 complement and extend the most successful ones in their portfolio. It would be irrational to focus on further development of unsuccessful lines. 160 Innovators will entirely legitimately want to protect investments in successful new lines through patent protection. Doing so is entirely rational and should not be pejoratively described as "patent clusters created to confront generic competition to the followon product;" 161 (vi) The alleged bridging tactics theory - The attempt to allege that other conduct relating to a next generation product is suspect "bridging tactics" 162 is equally unsuccessful. That innovators cease to promote or withdraw the old technology in favour of the new is standard economics. A company may rationally decide to discontinue promotion of old technologies when the next generation is available. This is particularly the case where generics may be expected to free ride on any promotion of the old product line. It is to be expected that litigation, complaints and settlements are most common around these first and next generation product lines, since they are typically the most valuable in the innovator's portfolio and, therefore, generics have the most to gain by seeking premature, unlawful entry. 163 The complaints by generics, listed verbatim in the report, 164 that introduction of new technology pre-empts their attempts to benefit from copying old products denote pure self-interest to the detriment of innovation, quality improvements and choice The attempt to portray next generation products as suspect is to invite economic stagnation. In every industry, innovation tends to proceed by increment, as the successive generations of cars, planes, trains, phones, computers, software, in fact almost every modern technology, well attest. There are very few "eureka" moments. It would cause immense harm to innovative activity across Europe if such conduct were considered to cause antitrust concerns. 3.6 Accumulation of Practices (Section C.2.7) 112. The observation that many of the practices described in this section of the report can occur more than once in relation to high value medicines is not grounds for suspicion. 165 As noted at the outset, generics seek to enter those markets where the value opportunity is the highest. There the incentive for premature unlawful entry is the most compelling and, for the same reason, innovators will seek to defend their rights most strenuously The report's allegation that such cumulative strategies unfairly delay generics is flatly contradicted by its own data. 166 It is precisely in relation to these highest value medicines where generics are entering the market most quickly. They do so more rapidly in those countries in which the alleged blocking conduct is more frequently applied. The concluding remarks on the "toolbox" that such conduct "can cost public health budgets and ultimately consumers significant amounts of money" fail to identify any plausible counterfactual and are simply not borne out by the report's own analysis of delays to generic entry PR, para PR, para. 845 and PR, paras PR, paras PR, paras. 884 and 886. PR, paras PR, paras #

31 4. NO EVIDENCE THAT INNOVATOR PATENT STRATEGIES REDUCE INNOVATION (Section C.3) 4.1 The Report Disregards All Other Possible Causes of Innovation Decline 114. The report focuses on patents alone as a possible cause of innovation decline. It fails to examine the extensive body of research on the many complex factors that impact upon pharmaceutical innovation presented to it in the course of the inquiry. 167 In that substantial literature, including reports for the Commission and FTC, patents are generally considered the spur for innovation rather than a cause of decline. 168 "The innovation that the patent system spurs for the discovery and commercialization of NCEs in the pharmaceutical industry in many ways showcases the patent system's benefits." The absence of any study of other factors deprives the report of analytical value. There is no basis for stating patent strategies are "possible causes" of decline 170 in the number of new medicines reaching the market, without any consideration of what other factors are in play. There is simply no causal link established between the two In fact, the evidence is inconsistent with such a theory. R&D spending and the number of pipeline projects in clinical trials rose year on year during the period under investigation. 171 In 2007 there were almost 4,400 products in clinical development, including 763 for cancer, 224 for cardiovascular diseases, 177 for respiratory diseases and 92 for Alzheimers There therefore appears to be no hold up in terms of projects during the early R&D stages, when the IP position would be first investigated. It is, to the contrary, during late stage trials and market launch where the reduction in projects under development is observed. 173 But at that stage patent blocks are highly unlikely to be the cause. Innovators would not commit substantial sums to commencing clinical trials if there was uncertainty as to whether the line of research was blocked by patent rights There is also no evidence of hold-up if one examines individual product areas. Research shows the time lag between first-in-class and followers is reducing. 175 It is extremely rare to find a therapy area in which there is only one product. There are more than 15 beta blockers, nine protease inhibitors, 15 non-steroidal anti-inflammatory drugs and more than ten statins available on the European market The report uses a partial data set by which to measure innovation and alleged innovative decline. The data exclude a range of innovative medicines, including vaccines, new salts, pro drugs, metabolites and esters of existing compounds and certain biologic compounds (such as antigens). 176 A complete data set from EMEA shows the output of new active substances The literature was examined in the CRA Innovation Report. Innovation in the Pharmaceutical Sector: A Study Undertaken for the European Commission, Charles River Associates, 8 November To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, supra, n 24, Chapter 3, p14. PR, p5. CRA Innovation Report, supra, n 8), p15. IMS LifeCycle R&D Focus (2007). CRA Innovation Report, supra, n 8, pp CRA Innovation Report, Id., pp (R&D directors denied IP caused pipeline attrition at this stage). The Economics of Follow-on Drug Research and Development: Trends in Entry Rates and Timing of Development, DiMasi and Paquette, Pharmacoeconomics 2004:22(2)1-14; EFPIA submission of 13 June 2008, p81. CRA Comments on the Preliminary Report, p8. #

32 increased over the last three years. 177 Expert economic evidence as to the EMEA position and the many alternative indices available to DG Competition to measure innovation was submitted during the investigation. 178 These indices show a more balanced picture of innovation, with some showing that innovation is flat and some a slight rise The evidence shows a range of reasons for the difficulties that pharmaceutical companies face in bringing new products to market, none of which suggests that patents are of concern to the innovative process. Pharmaceutical companies face rising R&D costs and a consequent decline in productivity due to objective market factors: industry retooling; increased costs of regulation and trials; and, most significantly, the market environment. Buyers are signalling that they do not value or are not always prepared to pay for innovation The report does not examine alternative indices on innovation or analyse the expert evidence on innovation provided to it. Without such fuller analysis, the economic consultancy, CRA International in its comments on the report, concludes: "[w]e do not believe that the PR can conclude there is a decline in innovation on the basis of the analysis presented" and "associat[ing] the decline [in innovation] to issues related to patents contradicts a significant economic literature on this global phenomena and the evidence presented in CRA." Patenting Strategies (Section C.3.1.2) 122. The report unsuccessfully alleges that companies engage in defensive patenting strategies aimed primarily at blocking the development of a new competing medicine with no intention to work those patents itself. It is also alleged that once such patent applications are published, they become prior art and therefore their development may cease to be of commercial interest to companies. (a) Defensive patenting (Section C ) 123. Defensive patenting has no meaning under patent law, as all patents are a defensive right by nature excluding use by others. The conduct described in the report is, to the contrary, clearly pro-competitive. At the stage of filing the patent application, companies seek protection of potential medicines and to keep as many avenues of future research in the field open. In view of the high attrition rate of candidate compounds in early development researchers will often work on several related series of compounds at the same time in order to provide a back-up candidate in the likely event the lead compound has to be dropped from development. These series of compounds are generally protected by separate patent families. Some of these may indeed ultimately not be further developed or commercialised but the real value of a patent will only become known many years after filing. The quotations used in the report merely reflect the uncertainty that the commercial potential of patents may only become apparent at a later stage EFPIA submission of 13 June 2008, p65, Figure 18. CRA Innovation Report, supra, n 8. EFPIA submission of 13 June 2008, pp62-69; CRA Innovation Report; CRA Comments on the Preliminary Report, pp8-10. CRA Comments on the Preliminary Report,,p3. PR, paras. 966 and 969 ("Patents that are no longer of interest but that [our company] does not want to dedicate to the public (and to the competition) are kept "; "Even if it turns out later that the combination of [our company's molecule] with [other molecules] will not be of interest ". (Emphasis added)). #

33 124. This conduct is desirable and essential to spur innovation. If an innovator has spent hundreds of millions of Euros developing a new product, that investment must be protected. It is under no duty to lend a hand to its competitors by leaving the field clear for them to develop a similar product. The company's best judgment, based on the business and commercial risks at the time of filing, that patent protection was necessary to make a successful and secure investment in the project should not be subject to ex-post scrutiny. (b) Defensive publishing (Section C ) 125. The report also suggests that it is anti-competitive to allow publication of a patent application because it puts information into the public domain and thus reduces development interest for other companies. This disregards the public interest in having this knowledge available to the entire research community. It does not follow that it thereby blocks the development of competing inventions. 182 Rather, patent publication enables and stimulates follow on research, including the filing of third party patents relating to improvements to the disclosed technology. (c) Patent-related exchanges and litigation (Section C3.2) 126. The report goes on to analyse patent related exchanges and litigation between innovators. But, again, it does not substantiate that patents are a disincentive to R&D If there are overlaps and it is generally recognised that overlaps are far fewer in pharma than in other sectors recourse to licensing is frequent. In a highly competitive R&D driven market, and over a period of eight years, it is perhaps surprising that the report identifies " a total of more than 1,100 instances, across the 27 Member States, where patents of one originator company may be infringed by the INNs, R&D programmes and/or patents of another company". Though the report characterises this as creating "a significant potential for innovator compounds to find their research activities blocked", its factual findings contradict any such blocking theory: In , four companies reported eleven disputes concerning four products that did not end in litigation; 184 In , there were 99 originator to originator requests for licences concerning 84 products and R&D plans. In 77 cases, a licence was granted. 185 In only two cases was a licence refused without any reasons being provided. 186 In only one case did an originator state that it abandoned a project due to the patent holders' refusal to grant a licence; In , 16 companies reported being involved in at least one separate case of patent-related litigation. There were a reported 66 cases in total concerning PR, para To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy, supra, n 24, Chapter 3, p14 ("The innovation that the patent system spurs for the discovery and commercialization of NCEs in the pharmaceutical industry in many ways showcases the patent system's benefits. Such innovation entails the high fixed research costs, relative ease of imitation, and free riding problems that patent protection effectively manages. Fewer patent thicket issues arise in the pharmaceutical context than in industries where innovation is less discrete and individual products are covered by many patents."). PR, para PR, para PR, para #

34 products out of which 26 occurred in Germany, 9 in the Netherlands, 8 in the UK, and 3 of each Spain, Italy, Belgium and France and there were a total of 11 cases in the remaining 10 Member States in which litigation was reported. In seven cases, the patent was annulled. Based on these facts, there is absolutely no evidence of a hold up to innovation. (d) Opposition and appeals (Section C.3.3) 128. In , 58 opposition procedures were initiated with final decisions in 26 cases. In 18 of these, the patent was revoked. In five cases, the patent was reduced in scope. In two cases the patent was upheld. Only one of these cases concerned a primary patent. The evidence thus establishes that innovators challenge patents with success when they are perceived to be weak and that the incidence of such cases show that they are the exception rather than the rule a mere five cases compared to the 40,000 patents/applications relating to the investigated INNs. The existing checks and balances work. Patenting strategies do not hold up innovator competition. (e) Settlements (Section C.3.4) 129. According to the report, 27 settlement agreements were entered into between originators in the eight year reference period of Most of these were licensing agreements and many resolved several disputes and litigation cases. Again no evidence of hold up is established. 4.3 Antitrust Intervention Should Not Add Still Greater Regulatory and Investment Uncertainty to the Process of Innovation 130. The reality is that patents are the spur, not the impediment, to innovation. The patent system, and other exclusive rights, play an essential and central role in the pharmaceutical sector. It is the most research intensive industry in Europe. Only 2 out of 10 new molecular entities reaching the market cover their R&D expenses. 187 Ultimately, without strong intellectual property protection, no pharmaceutical company would make the considerable investments that are necessary to bring new drugs to market. Those rights must ensure that the potential to recoup and continue investment is secure Legal uncertainty is damaging to any business, but as the Commission's most recent communication makes clear, it can be immensely harmful to the innovative pharmaceutical industry that must spend on average USD 1bn or more over the course of a decade for each product launched before seeing any return. 188 Antitrust intervention risks undermining what is already a limited commercial window of opportunity by further increasing legal uncertainty in an industry that subject to burdensome and complex regulatory obligations. This would, in turn, undermine the ability of innovators to deliver improvements in public health Drug Development Costs when Financial Risk is Measured Using the Fama-French Three Factor Model, John A. Vernon, Joseph H. Golec, and Joseph A. DiMasi, unpublished working paper (submitted to the Journal of Health Economics for publication). Communication on the Pharmaceutical Package, supra, n 4, pp5-7 ("Different systems lead to disparities in pricing, time-to-market delays and access inequalities;" "Genuinely transparent and speedy pricing and reimbursement decisions should be made possible by enhancing the application of the Transparency Directive;" "Requirements that cause a high administrative burden without providing a clear public health benefit have a strong negative impact on the competitiveness of the EU industry;" "[A] high level of joint commitment is required from all actors in order to make pharmaceutical rules clearer, simpler and more flexible"). #

35 5. POLICY RECOMMENDATIONS TO IMPROVE COMPETITION 132. The principal conclusion offered by the report is on the patent system, recommending a single European patent court to streamline and reduce the costs of challenging patents and a single Community patent. This is an area where EFPIA continues to engage with the Commission and other stakeholders in efforts to achieve a single litigation forum that must be capable of producing high quality, cost-effective and timely judgments As to the other bottlenecks and regulatory concerns identified, the report offers little by way of constructive solutions. Industry and stakeholders have been working together for many years to reach a consensus on achieving the right climate for innovation in Europe. 189 EFPIA and its members are pleased that the Commission's latest Communication calls for better regulation and recognises the distortions and risks to innovation caused by the current fragmented regime of pharmaceutical regulation Creating an effective climate for innovation and a competitive pharmaceutical sector in Europe will require: (i) (ii) (iii) (iv) An efficient generics market that allows savings to be reinvested in innovative medicines; A holistic (versus silo-budget) approach to national healthcare budgets, with the focus shifted from the cost of new medicines to their overall value; Transparent, objective, fair, swift and predictable pricing and reimbursement processes (including through enforcement of the Transparency Directive) 191 that reinvigorate innovation in the pharmaceutical sector, ensuring that the right buyer signals stimulate greater innovative efforts through offering a fair reward for innovation, including incremental innovation; A pricing and reimbursement regime that ensures early and equal patient access to medicines; (v) A territorial limit to the impact of State price controls; 192 (vi) Improving the current patent system to reduce costs and increase legal and commercial certainty for all parties by means of: adoption of the European Community Patent and the creation of a unified, specialised litigation system in Europe, a streamlining of the opposition procedure of the European Patent Office, and a mechanism to address patent disputes before generic launch; (vii) Re-examining whether the current SPC system has been effective in creating a 15 year period of protected commercial exploitation sufficient to incentivise innovation; (viii) A reasonable risk-benefit balance in regulatory processes; and High Level Pharmaceutical Forum , Final Report, 2 October Communication on the Pharmaceutical Package, supra, n 4, pp5-7. The PR, para. 1145, recognises that the Transparency Directive is not correctly applied. See the High Level Pharmaceutical Forum , Final Report of 2 Oct 2008, page 85: Member States should abstain from fixing prices for products that will not be used within their territory and that will not impact on their national budgets (as outlined by Recommendation 6 of the G-10 Medicines Report). #

36 (ix) Removal of R&D bottlenecks and reduction of the costs of development through public-private partnerships exemplified by the IMI initiative, an initiative of real vision that brings together the industry and public sector to promote innovation. 30 January 2009 European Federation of Pharmaceutical Industries and Associations #

37 THE EUROPEAN FEDERATION OF PHARMACEUTICAL INDUSTRIES AND ASSOCIATIONS TECHNICAL ANNEX IN RESPONSE TO THE PHARMACEUTICAL SECTOR INQUIRY PRELIMINARY REPORT 30 January 2009 #

38 Content Page Number PURPOSE OF THE TECHNICAL ANNEX...4 EXECUTIVE SUMMARY...5 A. INTRODUCTION...7 B. MARKET FEATURES OF THE PHARMACEUTICAL SECTOR MAIN MARKET FEATURES MARKET STRUCTURE PRODUCT LIFE CYCLE IMPACT OF GENERIC ENTRY THE REGULATORY FRAMEWORK PATENTS...11 C. MAIN FINDINGS PRODUCTS AND PATENTS COMPETITION BETWEEN INNOVATOR AND GENERIC COMPANIES THE ISSUES PATENT FILING AND PATENT ENFORCEMENT STRATEGIES PATENT RELATED EXCHANGES AND LITIGATION OPPOSITIONS AND APPEALS SETTLEMENTS AND OTHER AGREEMENTS OTHER PRACTICES AFFECTING GENERIC ENTRY LIFE CYCLE STRATEGIES FOR FOLLOW-ON PRODUCTS CUMULATIVE USE OF PRACTICES COMPETITION BETWEEN INNOVATOR COMPANIES PATENT STRATEGIES PATENT RELATED EXCHANGES AND LITIGATION OPPOSITIONS AND APPEALS SETTLEMENTS AND OTHER AGREEMENTS...16 D. COMMENTS ON THE REGULATORY FRAMEWORK PATENTS MARKETING AUTHORISATION PRICING AND REIMBURSEMENT...17 #

39 A. INTRODUCTION...18 B. MARKET CHARACTERISTICS AND STRUCTURE OF THE PHARMACEUTICAL SECTOR MAIN MARKET FEATURES MAIN STRUCTURE PRODUCT LIFE CYCLE IMPACT OF GENERIC ENTRY THE REGULATORY FRAMEWORK PATENTS MARKETING AUTHORISATION...27 C. MAIN ISSUES INVESTIGATED INNS, PRODUCTS AND PATENTS PRODUCTS AND INNS PATENTS COMPETITION BETWEEN INNOVATOR AND GENERIC COMPANIES THE ISSUES PATENT FILING AND PATENT ENFORCEMENT STRATEGIES PATENT-RELATED EXCHANGES AND LITIGATION OPPOSITIONS AND APPEALS SETTLEMENT AND OTHER AGREEMENTS OTHER PRACTICES AFFECTING GENERIC ENTRY LIFE CYCLE STRATEGIES FOR FOLLOW-ON PRODUCTS CUMULATIVE USE OF PRACTICES AGAINST GENERIC COMPANIES COMPETITION BETWEEN INNOVATOR COMPANIES THE ISSUES PATENT STRATEGIES PATENT-RELATED EXCHANGES AND LITIGATION OPPOSITION AND APPEALS SETTLEMENTS AND OTHER AGREEMENTS...64 D. COMMENTS ON THE REGULATORY FRAMEWORK PATENTS MARKETING AUTHORISATION PRICING & REIMBURSEMENT...66 E. CONCLUSIONS AND LAUNCH OF PUBLIC CONSULTATION...67 ANNEXES TO CHAPTER B...68 #

40 PURPOSE OF THE TECHNICAL ANNEX DG Competition has presented its Preliminary Report as an objective fact-finding exercise. Against this background, EFPIA wishes to point out a number of factual inconsistencies and discrepancies between the facts as presented and the implications drawn from those facts. In this technical annex ( Technical Annex ), EFPIA provides a detailed response in respect of certain sections of the preliminary report 1 ( report or PR ) which call for further observations and/or which are not addressed by EFPIA s comments on the report (the Response ). Failure to address any matter set out in the report should not be taken as an admission or acceptance of that matter. For convenience we adopt the headings and subheadings of the report, separating major headings with a new page. At the outset of each major heading of the report, we identify any top level or recurring concerns or questions which, in EFPIA s view, need to be addressed in that section. 1 Case No Comp/D2/ Commission Decision of 15 January 2008 initiating an inquiry into the pharmaceutical sector pursuant to Article 17 of Council Regulation No 1/2003 [2008] OJ C59/06. #

41 EXECUTIVE SUMMARY We address certain recurring themes in the report first introduced by the executive summary. The terms toolbox and strategies are first referred to here and are widely used in the report to describe patents, litigation, settlement, regulatory intervention and next generation marketing of patented products. The terms in themselves are unobjectionable, but their usage in the report quickly becomes pejorative, suggesting that these separate, standard and entirely legitimate practices have, either individually or cumulatively, some suspicious or ulterior object or effect beyond their intended purpose.this insinuation colours much of the report and leads to the damaging impression that DG Competition considers such activities as inherently suspect. We submit that that view is incorrect and one which has raised concern in many quarters. Patents of course underpin most innovative industries. The unrestricted right to use and enforce them is essential. Regulation is pervasive too in many industries. The channel of free communication to and dialogue with regulators should not be constrained. No company should delay bringing a bona fide concern to a regulator for fear of antitrust challenge. Generic delay and innovative decline are first alleged here. We deal with both, noting that the generic delay analysis cannot be valid without establishing a counterfactual. The report needs to identify what factors might cause a generic to choose to enter into a market at a particular time. Only then can it determine whether innovator conduct examined is a material factor in play and what, if any, delay that causes. The analysis does not do this. It assumes immediate entry, the day after loss of exclusivity ( LOE ). It follows this with the unwarranted conclusion that each day of non-entry post- LOE is in part innovator driven. Statistical correlation of entry times and the contested activities points to innovators having no influence on generic entry: Generic entry is fastest in relation to medicines where innovator conduct is alleged to be most prevalent. Generic entry has become faster over a time period where the prevalence of the impugned conduct has doubled or quadrupled. Generic entry is fastest in a country having the highest incidence of such conduct, and slowest in a country that has some of the lowest prevalence. As to innovation, we examine the partial data set which DG Competition presents in its report, both at the outset of the inquiry and now, as the sole basis for alleging innovative decline. We identify why that data set, alone, is an inadequate basis for the report s conclusion on innovation. We note that commentators consider the patent system as an essential precondition for innovation in this industry. Patent blocks have not been considered a cause of decline in the substantial literature on the state of innovation in the sector. We express surprise that detailed economic analysis available to DG Competition on measures of innovation is not the subject of the report. We question certain of the factual statements in the report relevant to the R&D process. A critique of the economic analysis is provided by CRA International. 2 CRA is in the unique position of having produced reports on the state of innovation both for the Commission, in 2004, and an update to that report for EFPIA in 2008, for this inquiry. We invite DG Competition to consider the points CRA makes in this and the earlier expert reports provided during the investigation. 2 See Comments on DG Competition s Pharmaceutical Sector Inquiry: Preliminary Report, CRA International, January 2009 ( CRA Comments on the Preliminary Report ), p8, attached as Annex 2 to the EFPIA Response. #

42 We respond to selective comments in the report that might leave readers with the impression that the pharmaceutical industry does not invest in R&D. The industry is, of course, the European leader in R&D investment, far more than any other industry. Every Euro is essential to bring potential medicines through the year R&D cycle to market. Communicating fully and accurately the risks and benefits of the latest medicines to doctors is a resource intensive process, far removed from conventional advertising. This is a two way process that allows the industry to capture and react to feedback from the medical community to ensure that drug applications are constantly improved and updated. This is part of ensuring that patients get the earliest possible access to the most efficacious medicines. #

43 A. INTRODUCTION (Page 5, para. 1) (Page 5, para. 2) The surprisingly substantial costs of distribution, a 35% difference between ex-factory and retail prices, are not examined in the report. The margin captured by distribution and the State (VAT rates vary from 0 to 25% across the 27 Member States) is substantial. 3 This is consistent with the observations of other stakeholders, including Edwin Velzel, CEO of the Dutch Health Insurer, UVIT, at the hearing of 28 November 2008, that inefficiencies in the supply chain divert substantial resources from healthcare budgets (Response, section 1.1(b), para.10-12). A full examination of how more efficient distribution could reduce this average margin would potentially be of enormous benefit. It would yield savings far higher than the posited 5% of generic spend which the report wrongly attributes at least partially to innovators. The alleged triggering factors for the sector inquiry are unsubstantiated. First, as to innovation, DG Competition used a partial data set by which to measure any alleged decline in innovation measured by new medicines reaching the market. The CMR data to which it refers captures part only of the new active substances released on the market. CMR excludes vaccines, new salts, pro drugs, metabolites and esters of existing compounds and certain biologic compounds (such as antigens). 4 DG Competition did not use data from the European Medicines Agency ( EMEA ). Those data show the output of new active substances increased over the last three years. 5 Expert economic evidence as to the EMEA position and the many alternative indices available to DG Competition to measure innovation was submitted during the investigation. 6 These indices show a more balanced picture, with some showing that innovation is flat and some a slight rise. The report does not examine alternative indices or analyse the expert evidence on innovation provided to it. Second, the sector inquiry decision and the report do not identify the instances of delayed market entry which triggered the inquiry. In fact, the times to entry for generics appear very quick (for high value medicines where the commercial incentive is the strongest) by the standards of the usual time required for a business to gear up to enter into a new market. 7 Generic entry is becoming quicker over time. Bearing in mind that generic entry constitutes de novo entry into a market with substantial commercial and regulatory barriers to entry, entry times of three to four months are extremely fast. There is evidence from generics companies that regulatory delay alone can cause years of delay after an innovator's exclusive rights expire ( LOE ). The report does not show that generics can or would enter more quickly, absent the conduct examined. The use of the word delay is inaccurate without this counterfactual (Response, section 2.1, para ) Medicines Distribution, Office of Fair Trading Market Study, December Comments on DG Competition s Pharmaceutical Sector Inquiry: Preliminary Report, CRA International, January 2009 ( CRA Comments on the Preliminary Report ), p8. EFPIA submission of 13 June 2008, p65, Figure 18. The Current State of Innovation in the Pharmaceuticals Industry, CRA International, June 2008 ( CRA Innovation Report ) (Annex 4 to the EFPIA submission of 13 June 2008). Factors Affecting Generic Entry in Europe, CRA International, June 2008 ( CRA Generics Report ) (Annex 5 of the EFPIA submission of 13 June 2008). #

44 Competition between Innovator Companies and Generic Companies (Page 5, para. 6) This is the first use of the terms tool-box and strategies in the report to describe certain legitimate business activities: patents, litigation, settlements, regulation and marketing of patented second generation products. In the course of the report, these terms come to be used in a pejorative manner to suggest that these activities have some ulterior or suspect purpose. This is variously described as ensuring or maximising revenues. Here and throughout the report, however, there is no evidence that these activities have any objective other than for their legitimate and intended business purpose: Patents are intended to protect companies investments in innovation by excluding others from using the protected invention. Patent litigation, including, where appropriate, settlement, is the judicial means by which these rights are enforced. Patents and their enforcement are essential to the maintenance of competition and innovation across all innovative industries; Regulatory interventions are the means by which companies bring bona fide concerns before regulators. They are essential to the proper functioning of a highly regulated industry; Marketing of second generation products bring new and improved medicines to market, patented where they meet the appropriate level of inventiveness, increasing competition and the range of medicines to the benefit of patients. It would be extremely damaging to these important and legitimate activities that underpin all innovative industries if it were suggested they are suspect on antitrust grounds. The costs caused by delay posited by this paragraph presuppose that generics would enter earlier than they do, a counterfactual which the report does not establish (Response, section 2.1, paras ). (Page 5, para. 7, page 7, para. 1) The report does not establish that generic entry in many instances occurs later than could be expected (Response, section 2.1, paras.45-51, page 5, para 2 above). The claimed savings of EUR 3bn or 5% are accordingly also not established. #

45 Competition between Innovator Companies (Page 6, para. 2) All patents are defensive. They exclude others from using the patented invention. The blocking of rival inventions is a pro-competitive stimulus to innovate and the trade off inherent in intellectual property protection. The report does not establish obstacles to innovation. It does not consider what would have happened had the innovator not benefitted from the protection of the patent system. It is well recognised in the economic and legal literature that, particularly in the pharmaceutical industry, an innovator is unlikely to make the investment required to bring a pharmaceutical product to market without patent protection. At the outset of the research an innovator will not know which of the many promising patentable compounds it has developed will be commercially viable or successful. It is necessary to patent the most promising compounds to ensure that due protection is secured for further development, even if that development does not, ultimately, result in a commercialised product. At the time of the initial R&D, the innovator will not know this. It is unlikely to risk proceeding with the project without the necessary conditions of certainty afforded by patent protection. The better question is whether, judged at the time, the innovator would have proceeded with the investment having patented only one of the possible compounds under investigation with the potential to address the relevant biological target (Response, section 1.3, 4.2(a)). The Regulatory Framework (Page 6, para. 3) Failure to fully recognise that the regulatory framework is central to the analysis of competition in the sector, having a significant impact both upon innovation and generic time to entry (Response, sections 1.4, 5), undermines the conclusions of the report. #

46 B. MARKET FEATURES OF THE PHARMACEUTICAL SECTOR 1. MAIN MARKET FEATURES 1.1 Market Structure (Page 6, para. 5/page 7, para. 1) The accuracy of this paragraph will depend on the data set collected. EFPIA's members believe that the phrasing of the questions on R&D spend by R&D stage was open to different interpretations which could have resulted in inconsistencies in the data set. By way of example, the report finds that only 1.5% of turnover is spent on average on basic, mainly pre-clinical, research whilst the latest industry survey shows that spend on pre-clinical research is at least 6% (as evidenced in EFPIA s submission of 13 June 2008) 8 and this figure understates the total spend on basic research. It should also state that all of the R&D spending, 17% of turnover, is essential to bring a pharmaceutical to market. The report compares marketing and R&D as a percentage of sales (page 36 of the report contains conflicting figures) without consideration of the percentages applicable to conventional industry sectors. The pharmaceutical industry spends the most on R&D in Europe, almost twice as much as the software and computer services sector. 9 It may be compared with companies in the fast moving consumer goods sector who spend 36% of sales on marketing and only 2% on R&D. 10 Although DG Competition has chosen not to make the data public thus far, we suspect that this is in line with the level of generic activity. The report does not identify the resource intensive nature of marketing in the pharmaceutical sector. Marketing in the sense used by the report is not conventional advertising. It is the process by which companies inform and educate doctors in relation to medicines. Physicians have to be fully and accurately informed of all the risks and benefits of a product before they can make an informed choice as to whether their patients are likely to benefit from the new medicine. It is equally important that pharmaceutical companies receive regular feedback in order to insure that products are constantly improved and updated. This is an essential part of ensuring patients receive the right medicines. 1.2 Product Life Cycle (Page 7, para. 5) The report refers to innovations acquired from third parties. It does not make clear that these are in many cases unlikely to have been viable medicines without access to the acquiring innovator s resources in relation to testing, clinical trials and securing regulatory approvals (see PR, footnote 12 and para. 40). There is an inconsistency with See The Pharmaceutical Industry in Figures, (Annex 9 to the EFPIA submission of 13 June 2008). European Commission 2007 EU Industrial R&D Investment Scoreboard, October 2007, p. 12. (Compare software and computer services 9.7%. The scoreboard considers the world s top 1,400 companies with R&D spend over a minimum threshold comprising the top 400 EU and 1,000 non-eu companies.). The top 10 fast moving consumer goods companies spent 36% of revenues on marketing (source: 2007 annual reports). #

47 pages 31 and 56 which note different percentages of acquired or in-licensed molecules. (Page 7, paras. 5-7) Research and development is not confined to the first stage of the product lifecycle. The identification of improvements to medicines, which can entail, inter alia, a new medicine, new manufacturing process or new indication for an existing product, can occur at any time and requires new clinical trials. This is in line with the continuous research and development efforts of all innovative industries. 1.3 Impact of Generic Entry (Page 8, paras. 1-4) (Page 8, para. 4) The report does not examine why generic pricing should remain so high after entry in what should be, essentially, a commoditised market. Nor does it examine the reasons behind the significant variations between EU Member States. It is a major omission not to look into the sector of competition between generic companies where others have identified very substantial savings are possible (Response, section 1.1(b), paras.24-27). The statement if generic entry had taken place without delay implies that generic entry would be expected earlier. The savings are premised upon entry on day one after LOE. The report does not establish that day one entry is likely. The expert analysis provided to DG Competition during the inquiry shows that it is not likely. The expert economic analysis is not examined by the report. EFPIA has submitted a further expert report 11 on this matter which outlines the report s shortcomings and identifies the regulatory, technical and commercial factors in the earlier expert evidence which DG Competition has not addressed (Response, section 2.1, paras ). 2. THE REGULATORY FRAMEWORK 2.1 Patents (Page 8, para. 7) The report notes that once granted, the European patent turns into a bundle of national patents. The report does not carry forward that finding into the analysis of numbers of patents where it misleadingly cumulates parallel, substantively identical, national patents across the EU to posit a thicket or cluster where none exists (see paras [349] below, Report, section C.2.2.1). 11 CRA Comments on the Preliminary Report, pp5-8. #

48 C. MAIN FINDINGS 1. PRODUCTS AND PATENTS (Page 9, para. 5) The doubling of patent applications over the period has not had an effect on generic entry. Generic entry became faster over the same period (Report, page9). The pharmaceutical sector like all industries is characterised by continuous efforts to improve products. The report implicitly questions the innovative character of these further developments by the innovator after product launch. Yet the report identifies only one product exhibiting a spike in patent applications at the end of the first patent s term. On average, the pattern of patent applications reflects a continuum of further research and development (Response, section 3.1(d), paras.67-71, para. 360 below). 2. COMPETITION BETWEEN INNOVATOR AND GENERIC COMPANIES THE ISSUES 2.1 Patent Filing and Patent Enforcement Strategies (Page 9, paras. 7-8) Patents exclude, and intend to exclude, competition by preventing copying of the protected inventions by generics or by any third parties. The report s statements to that effect, characterised as strategies, are no grounds for criticism. The characterisation of a patent occurring very late in the life cycle of a product suggests that there must be a cut off point for the continuum of development that occurs in all innovative sectors. There is no cut off point for developments. The description of certain patents as late is therefore misconceived. The number of patents misleadingly cumulates parallel, substantively identical national patents across the EU to posit a thicket or cluster where none exists (Response, section 3.1(b) para.58, see paras. [349] below). (Page 10, para. 1) (Page 10, para. 2) (Page 10, para. 3) The report states some of these patents are not strong. Statements as to strength or weakness are shorthand for a complex legal and factual assessment. In some cases the validity of a patent cannot be conclusively determined. Indeed, different national courts themselves reach different conclusions on parallel national patents. It is therefore not grounds for criticism that the validity of some patents may not be definitively determined prior to judgment in litigation (Response, section 3.1(c), see paras. 393, 514 below). Divisionals can be of no longer duration or of wider application than the invention disclosed in the parent application. The report does not substantiate the claim of generic uncertainty (Response, Section 3.1(e), see paras below). The statement that litigation creates obstacles or is a deterrent for infringers simply denotes the enforcement function of litigation. The report does not substantiate the claim of unmeritorious litigation (Response, section 3.2, see paras below). #

49 2.2 Patent Related Exchanges and Litigation (Page 10, para. 5) (Page 10, para. 6) (Page 10, para. 8) The alleged four-fold increase in litigation does not appear to have impacted generic entry which became faster over the period. The win/loss statistics are in line with other industries. They do not form a basis for extrapolating that all patents will be invalid or non-infringed to the same degree. Generics choose which patents to challenge in litigation or by provoking litigation through premature market entry. They will naturally choose those they consider most vulnerable and where the highest sums are at stake. The sample of patents litigated to judgment will of necessity be only those which are most uncertain, since the remainder will be unchallenged or proceedings will settle. Without significant comparative data for other industries, and because the sample is such a small subset of all pharmaceutical patents, it is not possible to draw any conclusions as to the strength of pharmaceutical patents or litigations. The report does not identify the number of cases in which interim injunctions were obtained but final judgment went against the innovator. It is only in these cases that delay of generic entry caused by litigation is possible. The generic will generally be entitled to compensation in such cases, related to the actual period of delay (if any). However, any such delay is the result of a considered judicial decision applying applicable national law and procedure for such injunctions and in the light of evidence from both parties. There is no justification for ex post antitrust scrutiny of such cases. The number of injunctions requested is inconsistent with the figure quoted at para 521 of the report. 2.3 Oppositions and Appeals (Page 11, para. 1) Amendment of the patent in opposition will not necessarily be a success for the generic. The amended patent may well continue legitimately to exclude generics (see paras. 569, 1043 below). 2.4 Settlements and Other Agreements (Page 11, paras. 5-6) The report applies a methodology of categorising settlement agreements that does not aid antitrust assessment, misunderstands the dynamics of settlements and does not take into account the counterfactual of total generic exclusion. The Federal Trade Commission s approach has not found success in recent US judgments and is inapposite, in EFPIA s submission, in relation to the different regulatory regime of the EU (Response, section 3.3, see paras. 579). 2.5 Other Practices Affecting Generic Entry (Page 12, para. 1) The interventions show no pattern of being focussed on valuable products soon to go off patent, but appear to be focussed, as one would expect, on a limited number of #

50 INNs where it seems that particular concerns arose. (Page 12, para. 2) (Page 12, para. 3) (Page 12, para. 4) (Page 12, para. 5) (Page 12, para. 6) The report is selective in omitting the categories of regulatory intervention where innovators were generally successful. No general strategy can be discerned from win/loss rates in litigation since such sampling is prone to selection bias. Only the most uncertain cases will litigate, the rest will settle or never become contentious (Response, Section 3.1(b), see paras. 504 below). The use of interfering to characterise innovator interventions is pejorative. The statements as to innovator s revenues may seek to suggest that interventions are made solely for financial reasons. The report has no basis for concluding that interventions in question were not made on the basis of good faith concerns. There is a disparity with page 37 of the report as to the level of spend on marketing which shows a figure of 20%for 2007, whilst DG Competition opted for the average percentage for the period of 23% of turnover. The resource intensive and essential nature of educating physicians and receiving regular feedback on new medicines is set out above (see page [6, para. 5] above), but omitted in the report. The report describes legitimate warnings as to patent infringement in a pejorative manner as an attempt to exercise influence over the distribution channel (see also para.799 below). Investigation of direct to pharmacy distribution has not found a material adverse impact on wholesaling and has recognised the efficiencies of the model Life Cycle Strategies for Follow-On Products (Page 12, paras. 7-8, page 13 paras. 1-2) Next generation products are an essential form of competition common to all industries, in particular innovative industries. The launch of next generation products, their patent protection and the timing and marketing of such a launch are all consistent with well established principles of economics and indicative of robust competition. The report incorrectly implies this conduct is suspect and states that it is based on weak innovation by relying on the hearsay of generic companies (Response, Section 3.5, paras , see paras.827, 838, 842, below). 2.7 Cumulative Use of Practices (Page 13, para. 3) (Page 13, para. 4) The use of phrases in a pejorative manner ( toolbox, instruments and the phrase even resort to the cumulative use of all these instruments ) to describe common and legitimate business activities is inappropriate. It is economically rational that generics will focus on the most valuable medicines to attempt to enter the market prematurely. Innovators must respond by asserting their 12 Medicines Distribution, Office of Fair Trading Market Study, December 2007 ( We recognise that there may be efficiency benefits to DTP and take the view that manufacturers should be free to choose the distribution method they consider to be most efficient. ) #

51 intellectual property rights. (Page 13, para. 5) The report incorrectly states that the combination of activities leads to generic delay. Firstly, there is no counterfactual analysis by which delay might be measured (see page 5, para. 2 above). Secondly, the report notes that generic entry is faster for the high value medicines where the alleged incidence of innovator life cycle instruments is most intensive. This is inconsistent with a theory of generic delay caused by innovators. Accordingly there is no basis for ascribing costs to health budgets or consumers to the practices examined. 3. COMPETITION BETWEEN INNOVATOR COMPANIES 3.1 Patent Strategies (Page 13, para. 6) As noted above, defensive patenting does not have a meaning in relation to patents which, by definition, are defensive rights. They will self-evidently block the development of a new competing medicine, if it falls within the scope of the patent (Response, section 4.2(a), see paras ). In a long and complex R&D process, it is unsurprising that certain patented compounds or other inventions are not, or not immediately, further developed. That does not mean they should not be protected as the innovator s invention, since they may be useful for future R&D or their use by rivals might threaten the viability or willingness of innovator to undertake a research project. (Page 13, para. 7) (Page 14, para. 1) Publishing the patent application is pro-competitive, disseminates scientific knowledge and fosters follow on innovation. The suggestion that it frustrates innovation is misplaced and contrary to the essential functioning of the patent system (Response, section 4.2(b) see paras.125). Divisionals are a necessary part of the patent system. The parent application will disclose the breadth of the potential patent derived from any divisional so that other innovators can design around if need be. The use of the opposition procedure is, as noted in this paragraph, an alternative means of seeking to clear a patent block (see paras ). #

52 3.2 Patent Related Exchanges and Litigation (Page. 14, para. 2) Patents do not have detrimental effects on the innovation process. All economic and legal theory suggests that, particularly in the pharmaceutical sector, patents foster innovation. The report does not examine what would have happened if the patent blocks did not exist. In all likelihood the first innovator would not have invested in the R&D project at all if there was a substantial risk of rival R&D projects free riding on inventions the former had discovered (Response, section 4.1, see paras 371, 410). Expert economic evidence was submitted during the inquiry that sought to present DG Competition with a report on the assessment of innovation in the pharmaceutical sector. It focused on the likely causes of observed trends innovation, in particular declining productivity in innovation (producing fewer medicines per dollar). This economist report was intended to update the work commissioned by DG Enterprise, using the same consultants, in The economic report concluded that industry retooling, regulatory burdens and pricing and reimbursement were the main causes of productivity decline. 13 The report does not examine that evidence. It does not consider any factor other than patents for the cause of productivity decline. It shows no causation between alleged patent blocks and innovation decline. (Page 14, paras. 3-4) The very small numbers of refused licenses and litigated patents do not support a theory of harm to innovation. It suggests that innovation is robust. As noted above, selection bias means that no general extrapolation can be made as to the validity of other patents from the extremely small sample of litigated patents (Response, section 3.1(c), see paras ). 3.3 Oppositions and Appeals (Page 14, paras. 5-6) Again, selection bias prevents the small sample of oppositions being grounds for generalisation about patent validity. In addition, the inclusion of amended patents as being counted as a loss for the patent holder misstates the legal position, since the resulting amended claims may still be grounds for excluding a rival (see paras. 569, 1043). 3.4 Settlements and Other Agreements (Page 14, para. 8) Since establishing the PR, DG Competition has sent additional questionnaires relating to the 1,450 agreements between innovators. DG Competition is invited to share its resultant findings with EFPIA for comments prior to publication of the Final Report in the interests of due process 13 CRA Innovation Report (Annex 4 to the EFPIA submission of 13 June 2008). #

53 D. COMMENTS ON THE REGULATORY FRAMEWORK 1. PATENTS (Page 15, paras. 2-4) EFPIA supports the creation of high quality, efficient and cost effective single forum for patent litigation. The industry is already engaged with other stakeholders and the Commission on this initiative. Patent quality is currently the subject of a number of EPO initiatives. We reject any implicit suggestion that patents are currently of low quality. It is unclear what is meant by patent strategies that may result in unnecessary delays. 2. MARKETING AUTHORISATION (Page 15, para. 5) EFPIA refers to the Response in relation to patent linkage (section 3.4(a) para 99, see also paras. 278, below). 3. PRICING AND REIMBURSEMENT (Page 16, para. 1) The report does not substantiate that regulatory interventions by innovators are other than good faith concerns about generic medicines, and any suggestion that additional time spent by regulators considering such concerns was not reasonably incurred is unfounded. #

54 A. INTRODUCTION The principal concerns expressed in this section are as follows: Understanding the pharmaceutical sector without understanding the regulatory constraints necessarily provides an incomplete view of the industry. It is only half the story, perhaps less than half. So much of the report s central observations are influenced by regulation that its omission is crucial. Generic to generic competition as a key for savings is well highlighted by the Netherlands case study. We recommend further analysis of this sector, based, inter alia, on the CRA analysis submitted during the inquiry, but not analysed in the report. 14 We critique the attribution of generic entry delays to innovator conduct since the report does not provide any statistical correlation in that respect. We note certain difficulties in the calculation method adopted in the calculation of alleged generic savings. We identify and correct certain misconceptions as to the nature of patents and their use as well as key regulatory delays that would have affected generics during the time period investigated. Firstly, the prohibition in most Member States on testing the innovator drug while still under patent was only abolished when Member States amended their laws in October Secondly, data exclusivity rules prevent generics from relying on innovator data until data exclusivity periods of six or ten years expire. There will be a further regulatory delay after that period, likely to be at least a year or longer, while the generic company applies for marketing authorisation. (2) Analysis of the behaviour of companies is incomplete without knowing the regulatory context in which they operate. The statement that the report focuses solely on behaviour is incorrect. The report in fact analyses the interplay of law or regulation (patents, litigation, regulation) and conduct. It would be more accurate to say that the report focuses on selected legal/regulatory regimes, those which involve the conduct in Section C, to the exclusion of other regulatory regimes. This selectivity renders the report a partial picture of competition in the pharmaceutical sector and is a significant analytical defect. Regulation, and in particular complex and diverse pricing and reimbursement regimes, constrain generic entry or make it less attractive (Response, section 2.1, CRA Generics Report). Although the report contains data that show the effect of price controls on competition in a number of markets, it omits any analysis of this phenomenon and the fact that in the pharmaceutical sector, EU governments have a dual role of price regulators and monopsony buyers. Regulation also reduces incentives to innovate (Response, section 2.1, CRA Innovation Report). Without analysis of these factors, statements as to observed generic entry times or alleged decline in innovation have no probative force. Without knowing the full regulatory and factual picture, there can be no understanding as to which of the many factors impacting generic entry caused it to choose to enter at a particular time. 14 Competition in the off-patent market post generic entry, CRA International, September 2008 (the CRA Off-Patent Competition Report ). #

55 (3) The absence of legal analysis creates damaging uncertainty. The report casts suspicion on legitimate and desirable business conduct without articulating what antitrust standard should apply. (5) The substantial costs of distribution, a 35% difference between ex-factory and retail prices, suggest that inefficiencies in the supply chain merit further examination (Response, section 1.1(b) The statement that generic medicines can enter the market upon loss of (patent) expiry does not establish that day one entry after LOE is the relevant comparator the counterfactual for competition law analysis to establish whether delays exist. Many factors influence the decision to enter a market (Response, section 2.1). (7) The recognition of the competitive impact of regulation is welcome. However, this finding should have been incorporated into the assessment of factors affecting generic entry and innovation. Without identifying how competition is constrained by regulation, no adequate counterfactual is established by which to judge corporate conduct. (9) The report uses toolbox of instruments and life cycle strategies in a pejorative manner. The empirical analysis of the conditions under which generic entry can be expected to occur referred to in this paragraph is not apparent in the main body of the report. As EFPIA s research has demonstrated, there are many factors that determine whether a generic company is likely to decide to enter a market (CRA Generics Report, CRA Comments on Preliminary Report, p5-8). (10) The financial incentive in relation to high value products may well incite generic companies to attempt premature entry and innovator companies will defend their rights and investments vigorously as the report notes where that is the case. That vigorous defence does not, according to the report s findings, translate into a longer period of time between LOE and generic entry. The reverse is the case in relation to high value products. Generic entry is quickest in relation to high value products (Response, section 2.1). (11-12 and 15) Generic to generic competition has been identified by expert economists and commentators as exhibiting inefficiencies. Those inefficiencies suggest that generic to generic competition would merit close examination and could yield very substantial healthcare savings (Response, section 1.1(b)). It is unclear why DG Competition excludes this area from examination. There was substantial evidence before it during the inquiry suggesting that different market environments could produce very substantial savings to European healthcare budgets from improving off patent market efficiencies. #

56 B. MARKET CHARACTERISTICS AND STRUCTURE OF THE PHARMACEUTICAL SECTOR 1. MAIN MARKET FEATURES 1.1 Main Structure (40 and 132) (50 and page 49) DG Competition asserts that SMEs occupy a vital role in maintaining the innovative character of the pharmaceutical sector and reports that large pharmaceutical companies are increasingly in-licensing new products. This is an oversimplification even if it concedes that data on the percentages of molecules acquired or licensed-in by innovators at Table 8 provide only a snapshot that does not allow any inference of whether innovator companies are today more or less successful in innovating than in the past. While SMEs play an important role in niche areas of research, they lack the resources to do the development work and deal with multiple layers of regulatory approval involved in successfully bringing their results to market. Especially in today s economic climate, collaboration with major innovators is often their only means of survival. The reference to innovators efforts to prolong their protection periods incorrectly suggests that the legal maxima of the exclusivity rights examined by the report can be extended. This is not the case save in well defined circumstances (for example, application for an SPC). (56-60) We refer to [page [6, para 5] above in relation to the R&D and marketing data presented. Note that para. 57 and table 4 contain different figures on marketing and R&D possibly because of the different time periods in question. The report offers no comparison with the marketing and R&D breakdown of generics, which, as to R&D, is likely to be substantially lower. (62) These trends identify important factors likely to impact upon innovation by increasing costs ( increasing requirements relating to safety and efficacy for new medicines resulting in higher R&D costs and new advances in genomics, proteomics and personalised medicines ) and reduced funding ( increasing control over prices and reimbursement levels, as well as on prescribing practices. ). These, and other factors, are likely to impact upon innovation incentives. The report does not further consider them in relation to the analysis of alleged innovative decline. 15 (63 and table 5) The number of NMEs launches based on CMR data in Table 5 is a partial data set which does not properly capture innovative activity (see page [5, para. 2] above). (70) Generic companies may not be able to manufacture all types of medicine. They find that some medicines or associated delivery devices are too complex or costly to manufacture. This is a factor in their decision to enter a market (EFPIA submission of 13 June 2008, p90, CRA Generics Report). (74) Many of the generic companies listed should be considered multinationals. Teva is number 18, Ranbaxy is number 50 and most of the others appear in the list of top 100 international pharmaceutical companies in Further, the EU market value for generics is higher than the US which could also explain the higher turnover generated 15 CRA Comments on Preliminary Inquiry Report, p9, footnote 25 #

57 in the EU. (77-78) Generic companies create new formulations, dosage forms and methods of delivery for off patent products and seek, as appropriate, patent protection for generic processes, products, formulations and methods of delivery. The report proceeds to be critical of the same activities when carried out by innovators. It should be recognised that such conduct is an essential part of competition and legitimate to all actors in the industry. There is no basis for prohibiting for innovators the innovative conduct which is permitted for generics. To do so would harm innovation and competition. (81) The marketing costs of generics will be mitigated by their focus only on successful brands where the innovator has already established a successful product on the market. The statement that marketing expenditure constitutes the largest share of generics costs is not substantiated in the report. (100) The Joint Declaration of CPME and EFPIA on the Cooperation between the Medical Profession and the Pharmaceutical Industry, June 2005, is a statement of high level principles to promote and consolidate high ethical standards. The EFPIA Code on the Promotion of Prescription-Only Medicines to, and Interactions With, Healthcare Professionals, first adopted in January 1992 and regularly updated, most recently in 2007, contains the detailed rules on interactions with healthcare professionals. 1.2 Product Life Cycle (108) Each of the activities referred to as life cycle management strategies said to be put in place in order to maximise the revenue streams from a given product are separate legitimate and essential means of protecting companies intellectual property and developing and marketing improved medicines (see page [5, para. 6] above). (116) There is no basis in patent law for a distinction between primary and secondary patents nor for any value judgement as to their respective merits. There may be circumstances where the molecule is not patented (a well known existing compound, for example) or off patent, but, for example, an inventive process or formulation is developed for which patents are applied for. 16 (117) The statement that patents are mainly aimed at foreclosing particular R&D of a competitor has no meaning. Patents are an exclusionary right that prevent use of the patented invention. That is the stimulus to innovation that the patent system provides. The suggestion that this may be suspect challenges basic tenets of the patent system. It also disregards the reality of the R&D process as explained at page [6, para. 2] above. At the outset of the research an innovator will not know which of the many promising patentable compounds it has developed will be commercially viable or successful. It is necessary to patent the most promising compounds to ensure that due protection is secured for further development. That is the case even if that 16 By way of example, after a significant R&D effort of over 10 years, one innovator made a breakthrough in the treatment of schizophrenia, combining the legacy of substantial mental health research and a new drug-delivery technology. The result, an injectable anti-psychotic, provides significant advantages over the oral formulation in that it guarantees a constant delivery of the drug over a two week period and thus reduces the patient s risk of relapse. This new combination has been patented in order to protect the innovator s significant investment. #

58 development does not, ultimately, result in a commercialised product. At the time of the initial R&D, the innovator will not know which product will be successful. It is unlikely to risk proceeding with the project without the necessary conditions of certainty afforded by patent protection. Such patent protection may well block competing R&D. That is the essence of the race to innovate created by the system. The innovator can be under no obligation to leave the way clear for competitors to use potentially promising compounds which it ultimately does not decide to develop. This would jeopardise any incentive to invest in the costly R&D process. ( ) The rolling average of the period from first patent application to first product launch appears unrealistically short in relation to the later period under consideration in Table 8. For example, bearing in mind an average duration of clinical trials of 5 years, the rolling average for of less than six years seems likely to be inaccurate. EFPIA s research suggests a development period of years. 17 (140) The statement that commercial considerations are less significant in terminating projects during clinical development is inconsistent with the findings of the economic research commissioned by EFPIA. 18 Many of the factors considered decisive for R&D choices at para. 134 are likely to change over a 10 year R&D process. This includes the launch of rival therapies targeting the disease area, anticipated pricing and reimbursement decisions and the likely price to be achieved by the product based on relative effectiveness compared to the next best therapy (something that may be discernible only in late stage trials). It seems that commercial considerations are therefore likely to continue to play an important role during clinical trials. Since progression from each clinical stage to the next commits the company to very substantial additional costs, it seems implausible that progression would occur without reassessment of commercial factors. It is possible that the responses received by the case team did not clearly identify, for example, where efficacy failed to meet regulatory standards or where it was considered not to show sufficient advantages over rival therapies to achieve suitable pricing and reimbursement status. 19 Research by PharmaProjects, one of the most comprehensive sources of such data, shows that the highest cause of attrition in clinical trials was strategic, a categorisation that includes commercial considerations, rather than safety or efficacy. 20 ( ) Again, the terms tool-box and life cycle management strategies appear to be used in a pejorative manner here. (147) The report rightly notes ( generic penetration in the EU is influenced by the different public policy choices made by the Member States ) that regulatory choices have a EFPIA submission of 13 June 2008, p28, CRA Innovation Report, p9. EFPIA submission of 13 June 2008, p69; CRA Innovation Report, pp85-89; CRA Comments on Preliminary Report, pp9-10. CRA Comments on Preliminary Report, pp9-10. Scrip 100 (2008), p36 ("Why are drugs not getting through? As well as poor candidate selection, and the fact that drugs are spending more time in Phase II, attrition must be held accountable. Of the projects for which companies have disclosed the reason for discontinuation, a startling 44% of decisions were strategic, while just 28% of programmes were dropped because of lack of efficacy and 11% for poor side effects"). #

59 significant impact on generic decisions to enter a market. The report fails to explain why regulatory delay is not highlighted as a significant impediment to generic entry. In Poland there has been little or no market access for innovative products in the period under investigation which could explain the relatively high share of generics on the market. Figure 9 does not state whether the market share value percentages shown relate to total market at retail or ex factory prices. 1.3 Impact of Generic Entry ( ) It is unclear from the data presented in relation to time to entry how time to entry was calculated for molecules for which no entry was observed during the period (footnote 67). Were such molecules excluded from the sample or assumed to enter at the end of the period observed? The data in Figure 13 identify an average of less than four months time to entry for highest value molecules, but this masks periods of between 0 and 50 months. If there are many long outliers, potentially driving up the average, it would be important to determine whether factors relating to the outliers make them unrepresentative (for example a compound which is difficult to manufacture). Excluding unrepresentative outliers may be appropriate in such circumstances. The finding that generics enter high value markets most quickly is inconsistent with the theory that blocking tactics seek effectively to exclude generics. The report identifies alleged blocking tactics as most prevalent in relation to high value medicines where it seems entry occurs most quickly (Response, section 2.1, 2.2). (168) The finding that time to entry is reducing over time is also inconsistent with a theory that alleged blocking tactics, which the report finds allegedly increase over time, seek effectively to exclude generics. (167, Figure 14) (174, Figure 17) The different time periods for entry in different Member States have no correlation with the prevalence of alleged blocking tactics. Out of the big five markets, the UK has the quickest weighted entry (3 months post LOE entry) but the second highest number of patents, patent litigation, interim injunctions, BII settlements and regulatory interventions before marketing authorities. Conversely, France has eight months weighted post-loe entry, at least double the UK, but has the lowest number of instances of litigation, interim injunctions and BII settlements out of the big five (slightly lower patent numbers than the UK and the same number of marketing authorisation interventions). In the narrative at 174, the UK should be included in the countries with a high rate of generic penetration. (184) The assumption that generic entry would be immediate after LOE is incorrect. There are a range of regulatory and commercial hurdles for the generic which make day one entry after LOE implausible. The calculated saving of EUR 3bn over eight years or 5% of the value of the relevant medicines based on an assumed day one entry after LOE is therefore not substantiated (Response, section 2.1). As noted above, the report does not establish that any of the innovator conduct under investigation caused generic delay. The analysis does not currently take into account #

60 the commercial and regulatory factors identified in the expert evidence presented to DG Competition. 21 As CRA states therefore it is not possible for the PR to conclude that the strategies included in the tool-box have any significant impact of the length of time to generic entry. Only by incorporating the data collected through the PR (in particular differentiating between products where these strategies had been employed and where they had not) into an analysis including the other factors could we test if this accounts for any of the delay after allowing for the other factors. 22 There are also methodological shortcomings in the alleged savings identified in the report: 23 There are a number of commercial and regulatory factors that make entry less attractive for generics. Day one entry is implausible. These issues were examined in detail in the evidence presented to DG Competition. It is unclear why they have not been taken into account; The PR sets out that the recent pattern of entry is faster than at the beginning of the sample. This should be used in any quantification of how the market works today; The assumption that the potential price reduction is the same where entry is observed and where we don t see entry or entry occurs after a period of time is inconsistent with the evidence and economic commonsense. There will be a larger cost saving for higher value products. But it is likely that the markets where entry is slowest are those where prices are lowest reducing the potential for cost savings. It is inappropriate to use the price reduction occurring in markets where there is a commercial incentive and applying this to markets where there is not; No allowance has been made for regulatory delays. The estimate of potential cost savings, which is itself already small compared to cost savings that generics have achieved, is therefore likely to be a significant overestimate (rather than a lower bound as suggested by the Commission). 24 ( , Figures 24 & 25) The PR notes that EU averages hide considerable variation between the Member States. Figures 24 and 25 clearly illustrate to what extent pricing and reimbursement regulations affect price competition in the national markets. The data for the Netherlands show a price decrease of about 8% two years after generic entry (with prices of generics even higher than prices of the innovator product). With the exception of the Czech Republic, those countries which record the biggest generic price decreases are those often categorised as high price countries. The report does not attempt to explain why price drops vary considerably across the EU from about 10% to 70% CRA Generics Report; CRA Comments on the Preliminary Report, p6 ( The full list includes commercial attractiveness, complexity of manufacturing, competition from other therapies, the advantages of simultaneous LOE in multiple markets ; whether there is an associated delivery device, such as an inhaler, syringe or patch and loyalty to incumbent brands require costly marketing to displace. ). CRA Comments on the Preliminary Report, p6. CRA Comments on the Preliminary Report, p6-7. PR page 76. #

61 ( , Figures 27 & 28) The report notes considerable variation between Member States in terms of generic penetration but does not attempt to explain these variations in light of national pricing and reimbursement regulations. With the exception of the Czech Republic, the countries mentioned as having the highest level of generic penetration rate are the countries that have free pricing (see para. 288). #

62 2. THE REGULATORY FRAMEWORK 2.1 Patents (209) It is incorrect to say there is no mechanism for obtaining licenses of unworked patents. Patents laws make provision for compulsory licenses upon the appropriate grounds being satisfied. (211) Patent rights may prevent a rival from marketing products based on the latter s invention if that invention falls within the scope of the patent. It is true, however, as this paragraph goes on to state, that competition occurs between patent protected medicines (Report, section B2.1, para. 211). Rather than being just possible it is very common across many classes of drugs. On the European market, there are more than 15 beta blockers, nine protease inhibitors, 15 non-steroidal anti-inflammatory drugs ( NSAIDs ) and more than 10 statins available. 25 (234) There is no evidential basis for the statements that: [T]he applicant wants to maintain legal uncertainty for potential competitors around an application; which The applicant knows is weak and likely to be rejected; or That the filing of multiple divisional applications could, for instance, be used for this purpose ; or That these are delaying tactics by applicants. There are many perfectly legitimate reasons why proceedings are delayed or divisionals are necessary. Any legal uncertainly would be contrary to the innovator s interests in knowing whether its substantial investments are protected by IP rights. The statements also take no account of the resource constraints of the EPO to deal with applications or divisionals expeditiously. (239) For as long as a patent has not been revoked there is a presumption is that it is valid and the patent holder is fully entitled to defend its rights. It is wrong to suggest that there is no financial risk entailed in such enforcement actions. Even in interim injunction cases, the generic is generally entitled to damages under cross-undertakings as well as costs if the patentee loses at trial and, in many Member States, in order to get an injunction, the innovator must provide financial guarantees upfront (Response, section 3.2). (240) Again, the report does not refer to the possibility of the generic obtaining damages if it is the subject of an interim injunction pending trial in a case where final judgment is against the patentee (Response, section 3.2). (241) National courts are not the final arbiter of patent validity in most cases. If the EPO revokes the patent in opposition it is revoked across all EPO jurisdiction (PR, para. 242). #

63 2.2 Marketing Authorisation (254) Footnote 150 is incorrect to say that the possibility for generic products to use the centralised procedure was introduced by Regulation 726/2004. It was introduced by the prior Regulation 2309/93. Regulation 726/2004 provided for the possibility of nationally authorised generics on the basis of centrally approved reference products. ( ) The data exclusivity rules applicable to the E75 will be the rules prior to amendment by Directive 2004/27. In other words data exclusivity will apply for the appropriate period of six or ten years to those INNs. After the end of data exclusivity, generic entry would not be immediate because the end of data exclusivity merely permits the generic to start the process of applying for authorisation. That process can take around a year, and in some cases longer, to complete (Response, section 2.1, para. 48). (260) The Bolar provision permitting studies and trials necessary for marketing authorisation on the patented product was required to be transposed by Member States by 31 October 2005, six years into the seven year period of the inquiry. It is likely that the large majority of the E75 will fall into this category. For those expiries, generics would face the delay of waiting for the end of patent expiry to conduct such tests (Response, section 2.1, para.46). (278) It overstates the position to say that patent status can have no bearing on market authorisation or should not be considered by authorities. The intellectual property position is expressly left open by Article 10(1), Directive 2001/83/EC ( without prejudice to the law relating to the protection of industrial and commercial property ) when a generic relies on innovator regulatory data to obtain marketing authorisation. Member States have a right to know whether a product is at risk of disruption of supply because of patent protection. Some routinely ask for this information or their laws require it (paras. 716 and 1157, Tables 22 and 36). (294) The later approaches for P&R status to countries with smaller markets are likely to involve commercial considerations, rather than the alleged reluctance based on reference pricing. Launch will be relatively more costly in comparison with the commercial opportunity. 25 EFPIA submission of 13 June 2008, p80. #

64 C. MAIN ISSUES INVESTIGATED 1. INNS, PRODUCTS AND PATENTS 1.1 Products and INNs As to the methodology applied by DG Competition to determine the products and INNs that would be the focus of the Inquiry (see also Methodology Annex, p ), EFPIA notes that the PR appears to rely in part on IMS Health data for loss of exclusivity dates. EFPIA knows from experience that the information which can be obtained from IMS Health has its limitations. In the knowledge that, even for their own products, it is not a trivial task for innovator companies to accurately identify loss of exclusivity dates for every EU27 Member State, this is not surprising. The list of 219 INNs includes a significant number of combination products. Predictions as to the loss of exclusivity for such products are complex, and treating them on a par with single active ingredient products will likely distort any data set. The list of 219 INNs also includes a number of biologics. During the time period in question regulatory requirements for biosimilars were unclear and generic companies did not tend to pursue such products. Treating them on a par with small molecule active ingredient products will also likely distort any data set. DG Competition s use of average periods between loss of exclusivity and date of generic entry is highly questionable. For example: it does not take into consideration the extent to which outliers affect the average; the fact that where there is early generic entry this is treated as entry at the date of loss of exclusivity biases the data (see paragraph 48 of Annex: Methodology and footnote 457); and it is not clear whether, if a generic enters the market and is subsequently found to infringe a patent, this is regarded as early entry and taken into account in calculating the average. The limitations noted above mean that the data do not allow definitive conclusions to be drawn; only broad directional observations can be made. As explained below (see i.a. sub para (322)), the findings that generic companies brought fewer products per INN to market and that successful INNs show a higher degree of product differentiation is neither surprising nor suspicious. (314) EFPIA notes that, for the purpose this Section 1.1, different dosages and different forms are considered as different products by DG Competition. This is inconsistent with its INN-based approach to the pharmaceutical patent arena (in particular Section C, Chapters 1.2 and 2.1), where it is said that innovator companies file a multitude of patents (on process, reformulation, etc.) protecting the product (e.g. paras (376) and (377)) or for the same medicine (e.g. Section 2.1 Summary, p. 166). (315) EFPIA further notes that DG Competition generally selected the INNs to be investigated on the basis of their commercial value ( top-selling ). It may be stressed that innovator companies also invest in less successful products and medicines for rare diseases. It is correct that generic companies are usually not very interested in these kinds of products given that there is not enough money to be made from them. #

65 (319) It would appear from these data that only 3 INNs still had exclusivity across the 27 Member States. This number seems very low which raises the question whether ghost generics or other illegal copies were taken into account. In turn, this raises questions as to strength of IP enforcement in certain Member States since the T50 list covers a significant number of in-patent medicines in France, Germany and the UK. If these figures are correct, EFPIA notes that intra-generic competition (and especially the apparent lack of any real price-competition between generic companies, at least in the first 12 months) is worthy of further investigation as it may have a very important impact on the healthcare budgets of the respective Member States. In fact, it seems impossible to paint an objective picture of the pharmaceutical sector without such an analysis. (321) The conclusion that innovator companies generally offer a larger number of different products per INN confirms that generic companies focus on only a few formulations whilst innovators continue to invest in R&D and offer a broad choice to doctors, in some instances on formulations or variants which barely cover their costs because the level of sales is so low. The findings that a relatively high number of generic companies is active per INN shows that there are few barriers to enter the market (after loss of exclusivity) other than regulatory factors inherent in the law, as opposed to company behaviour. (322) The data presented by DG Competition does not differentiate older INNs from more recent ones. Some products will be less successful than others simply because they have not been on the market as long. For those more recent (and less successful) products it is only logical that fewer improvements or variants have yet been identified. In any event, the fact that the more commercially valuable INNs present a larger number of different products is not surprising. Indeed, the INNs are precisely more successful because more variants are available. Additionally, as in any industry, companies tend to invest more heavily in further research regarding their most valuable products than in commercially less attractive ones. R&D is further encouraged by the fact that, as successful products are sold in larger numbers and to a larger patient base, innovator companies will have more data available (e.g. through its pharmacovigilance obligations), enabling them to identify further opportunities and improvements more easily. Finally, EFPIA notes that, in general, generic companies are also able (and willing) to market a larger number of different products per INN in the more commercially valuable classes, as confirmed by DG Competition. The fact that they launch fewer different products (in the investigated INN universe and over the investigated timeframe) is logical as they can only launch generic versions of product variants which faced loss of exclusivity. (323) EFPIA notes that the numbers are calculated on the basis of INNs for which in at least one Member State a new product was launched. Thus, a single product launched in different Member States over different years will each year be counted as a new product. In other words, a product launched in a given Member State which was previously already on the market in another Member State, will qualify as new product. There may be double counting here. #

66 1.2 Patents The PR reflects misunderstandings of the patent system, its rationale and in particular its relevance to innovation in the pharmaceutical industry. The PR states that patents provide exclusive rights and allow a pharmaceutical company ( ) to reap financial reward for investment made in the development of new therapeutic agents. The exclusive rights granted by a patent do not consist of a positive exploitation right or monopoly but a set of negative rights. Patents essentially confer on the holder the right to prevent third parties from commercially exploiting an invention without authorisation. This is the heart of the patent system. Blocking unauthorised third parties from applying the claimed invention is precisely what a patent is aimed at. To criticise a patent holder for using a patent in this way is in effect a criticism of the patent system itself. Patents are not awarded to reward/recoup investment but to provide potential reward for and thus to provide incentive for invention/innovation. Moreover, patents do not only reward innovation with respect to new therapeutic agents but also any patentable inventions with respect to existing (i.e. known) therapeutic agents. It is true that the pharmaceutical sector makes use of the patent system to a very significant degree. The increase in numbers of applications may reflect the increase in innovative activity in the industry. EFPIA does not share the general assumption that patent activities occur primarily in the so-called R&D phase as development and innovation continue during the product life cycle. Finally, DG Competition fails to acknowledge that these patents also disclose to the world, including generic companies, how products can be made or processes applied. They therefore increase the stock of publicly available knowledge and in doing so provide a basis for others to innovate. (331) DG Competition notes that generic companies increasingly seek ( secondary ) patent protection (see also para (364)). Indeed for some INNs, third parties, including generics, may have more patents than the original innovator. (334) The figure of 14.8 national patents per INN is irrelevant and, to the extent that it is used to form the basis of any argument, misleading. In many Member States there were significant limitations on the types of patents that could be obtained during the period for which patents would have been applied for in respect of the 219 INNs. It is more appropriate to regard companies generally as applying for patents in all Member States for which they are available (as confirmed in paragraphs (1085) and (1087)). It is notable that the UK, which according to Figure 14 (p. 68) has one of the shortest delays in generic entry, is highest on the list in terms of number of patent applications filed. Conversely, in Luxembourg, where according to Figure 14 the longest delays in generic entry is shown, is second to last in terms of number of patent applications filed. These findings clearly contradict suggestions that increased numbers of patent applications contribute to delays in generic entry. (335) DG Competition states that the period saw a markedly greater increase in pharmaceutical-related patent applications, comparing the pharmaceutical data with overall patent filings and application in the field of organic chemistry. It is by no means clear that it is appropriate to use organic chemistry as a proxy for pharmaceutical applications. The mere fact that the technological fields may be closely related from a #

67 scientific perspective does not mean that they can be meaningfully compared when assessing such things as the number of patent applications and their success rates. The number of applications will be influenced by the importance of patents in the commercial context (and, for example, the importance of patents in the pharmaceutical and polymer markets is very different) and success rates will depend, to a significant degree, on the amount of prior art that is available (i.e., the research and publication intensity in a particular field). Moreover, the fact that there is an increasing number of patent applications in the pharmaceutical sector and a greater rate of increase of such applications may be indicative of the amount of innovation occurring in this sector. As such, these paragraphs appear to lead to the conclusion that the pharmaceuticals sector is more innovative than other organic chemistry sectors. Over the years, more and more information has become available to pharmaceutical companies. Indeed, additional data are for example, generated through the more stringent testing requirements for obtaining a marketing authorisation and increasingly rigorous pharmacovigilance obligations. Furthermore, the more patents are being filed, the more inventions are being published, which again spurs more new inventions. It is precisely the underlying ratio of the patent system to foster further innovation through publication. Additionally, the increase in pharmaceutical inventions may partly be attributed as more and more biotech companies, universities and academia are filing patents in this field. Finally, the increase in pharmaceutical patents may also be partly explained to the increasing activity of generic companies in the patent arena. (336) As opposed to para (349) and following, where the statistics on patents and patent applications are counted either per country and thus inflating the numbers, here the numbers are given per invention. (338) It is unclear in this paragraph precisely what the figure of 28,750 relates to. (340) Pending European patent applications are counted by DG Competition as an individual application for each of the designated States. However, usually, all (or most) EPCcountries are designated in the application, whereas upon grant the patents are not necessarily validated in each of those designated States (as i.a. shown in Figure 41 (e.g. LU)). The approach of counting each designation for applications overstates the numbers and distorts the data set generated. In view of the above, EFPIA submits that a more objective picture would be painted if DG Competition would, instead of or at least in addition to the data presented, give similar data on the basis of patent families rather than individual patents/applications. (342) The terms primary and secondary patents can be interpreted in different ways. Indeed, these are not terms of art and were not clearly defined in DG Competition s questionnaires. For example, a primary patent is defined by DG Competition as a patent protecting the innovator company s active ingredient. 26 It is unclear whether this 26 EFPIA notes that, in para (217) of the PR, DG Competition seems to be saying that the term primary patents means basic patents or compound patents protecting new active ingredients. At various stages of the PR DG Competition uses the terms base patents, basic patents, basic substance patents, compound patents or first compound #

68 is limited to compound patents per se or also includes e.g. patents claiming processes to manufacture such active ingredients, thus indirectly (but certainly) protecting the active ingredient as such. This, in itself, means that all the data relating to primary and secondary patents are inherently suspect as would be any conclusions which are drawn from them. In any event, the ratios given by DG Competition seem perfectly logical as a pharmaceutical product is a hugely more sophisticated thing than the mere compound. EFPIA has already explained the complexity of moving from a chemical/biological compound to a mass produced final product and post launch improvements. (346) The statistics presented by DG Competition are misleading as they purport to show that 34% of applications were granted and 66% were refused/withdrawn. All that can be said is that for all the cases ultimately decided by the EPO, approximately 90% of the applications were granted and only 10% were refused. Applications may be withdrawn for several reasons. (349) The number of 1,300 granted patents and pending applications for a particular INN is a misleading way of presenting the patent data. EFPIA submits that this issue should be approached on a patent family basis rather than counting all designated Member States as individual patents or applications. Figure 44 indicates that, if innovators file in 14.8 Member States as an average, on average there are approximately 12 patent families per INN. Of course, there will be a huge variance between high and low ends of the data which lead to the average but the PR rarely acknowledges this when it uses averages. Nevertheless, DG Competition focuses intensely on the single INN with allegedly 1,300 patents and pending applications without clarifying properly that this does not relate to 1,300 separate inventions or that this example does not represent a typical portfolio (if such a thing even exists). In doing so it ignores the more than 200 patent families relating to the same INN owned by third parties. What can be taken from Figure 44 is the total count of the first 10 INNs is above or close to 800 national patents and applications each on an EU-wide basis (i.e. counting each invention separately for each Member State designated). Absent detailed information, the only source of interpretation is the shape of the graph. The form of the curve is of hyperbolic shape and indicates that the average number of patent filings is well below these numbers, around 180. While DG Competition reports many average numbers in its reports, it fails to provide such a figure here and simply refers to the 1,300 figure throughout its report. The substantial drop of values in the first quintile demonstrates that the innovator companies patent filing strategies are not to build up large numbers of patent families around existing products. Otherwise, the curve would not substantially fit into a hyperbolic format. Finally, the findings presented in Figure 13 (p. 67) that the most valuable products do not suffer any substantial delay in market entry contradicts DG Competition s later allegations of obstructive patent filing strategies with respect to these most valuable products. Indeed, the myth of harmful patent clusters is flatly contradicted by the fact that generics enter markets more likely, more frequently and more quickly with high value products while those high value products benefit most from high numbers of patents, NCE/NME patents, or API patents. In para (948), reference is even made to primary patents concerning a secondary aspect. #

69 patents. Moreover, when comparing Figure 14 with Figure 41, it even appears that generic entry is precisely earlier in those countries where most patents are being applied for (see above sub para (334)). If alleged patent clusters were a successful strategy to delay generic entry, such an early market entry would not be possible. (350) While DG Competition explains that a handful of INNs account for the majority of patents and patent applications, it does not explain which of the compounds had already lost most of their patents prior to the year This may have a distorting effect on their conclusions. It is not unthinkable that the top 50% of best selling products would be those that were still enjoying exclusivity for a larger part of the period concerned, whilst the products with lower sales may have lost patent protection either before or in the earlier part of the period concerned. DG Competition does not seem to have weighted the results to take this effect into account. (351) It is economically logical that companies (in any industry) will invest and continue to invest the most in their most successful products. Accordingly, more patents are being filed with respect to these products. Innovator companies align R&D with patient need. Further, the increasingly rigorous pharmacovigilance obligations generate additional data. The medicines with greatest patient populations give rise to more research for a number of sub-populations requiring special dosages, special formulations, etc. This may to some extent further explain the higher percentage of ongoing research and ongoing patent applications. Finally, as noted above there are no data on the number of patents or applications concerning the INNs in the hands of companies other than the innovator. That data might show that, and the extent to which, other companies (including generics for example) work on further developing successful INNs and seek to protect their developments. Further, to the extent that there are such patents, their potential effect on generic entry is not considered. (353) When presenting the patent portfolio lifecycles, DG Competition does not take into account applications which have been refused or withdrawn since these no longer present obstacles to competitors. Including the applications which were withdrawn, abandoned or refused before product launch in Figures would likely present a greater number of early applications and thus different graphs. (354) There is no such thing as a conventional portfolio lifecycle. As stressed by EFPIA in the report referred to by DG Competition, it is impossible to generalise as to the typical portfolio (see EFPIA Intellectual Property and Pharmaceuticals June 2008, Section 2.11, footnote 7, p. 10). Accordingly, DG Competition misrepresents what is stated in the EFPIA submission. Figure 2 of the IP Annex to the EFPIA submission is stated to be a hypothetical illustration. It is only DG Competition who characterises Figure 47 as a conventional lifecycle portfolio. EFPIA stresses as it did in its June submission - that each portfolio lifecycle will reflect the concrete factual and technical circumstances with respect to the underlying molecules, formulations, processes, uses, etc. Innovation occurs throughout the product lifecycle and one may expect a steady increase in patent filings throughout at least the life of the primary patent, as is acknowledged in para 357 of the PR. After launch, patent activity will depend on the specifics of the compound. In many cases, problems encountered (or new opportunities raised) in the #

70 real world prompt further R&D and further innovation and patenting. (356) Rather than the alleged conventional life cycle depicted in Figure 47, the life cycle shown in Figure 48 it is probably more typical for blockbusters (showing a steady increase over the lifetime of the primary patent see below sub paras (356)-(357)). Figure 48 indicates that the top 20 INNs have a total of approximately 750 patents and applications (presumably on a patent family basis). On average, this shows that approximately 2 applications per INN are made per year over the first 20 years. This is hardly a high number for products that undergo significant development over time, particularly as they are successful products. Finally, EFPIA notes that, apparently, about 50 primary patents were filed related to the 20 selected INNs in year 0, which gives an average of 2.5 primary patents per INN. This shows that the above concerns regarding the ambiguity of the definitions used by DG Competition are genuine. (357) When comparing Figure 48 with Figure 49, it is obvious that the only conclusion that can be drawn is that there is no common pattern and that each INN-related patent life cycle is to be examined on its own merits. The lack of raw data provided in the PR does not allow for any conclusions as to what has caused the individual profiles. If anything, the steady increase presented in Figure 48 shows that innovator companies permanently keep on innovating and improving their products and processes for the reasons stated above. It would not make economic sense for companies to develop all incremental improvements before launch as this would lead to unacceptable delays in launch and therefore an even shorter period prior to loss of exclusivity. Moreover, some improvements are only made possible or spurred by the data generated from the findings in the real world. (358) Figure 49 shows the presumed patent life cycle of the 10 most valuable INNs. However, the number of patent applications which have been refused or withdrawn early in the lifecycle is not shown which would affect the shape of the graph. Moreover, DG Competition does not have any weighting with regard to the number of patents filed for each product, which may result in a different picture. Furthermore, Figures 49 and 50 are meaningless without knowledge of the specifics of each product and the problems it encountered in real life. If anything, the discrepancies in the respective portfolio life cycles clearly show that generalisations are not appropriate. (359) DG Competition asserts that Figure 49 depicts a clear trend towards late-filing strategies. A mixture of trends is shown in the Figure. DG Competition s allegation is wholly unsupported by the data. (360) Figure 50, showing the patent life cycle of one particular INN, seems to be an outlier. As appears from Figure 49, this "example" is not typical and takes no account of the specifics of the case. There is no basis for drawing a stark contrast with a conventional lifecycle which does not exist. #

71 In any event, there is nothing inherently wrong with late-stage filings. In addition EFPIA notes that for the INN of Figure 50, none of the applications filed from years 16 to 26 could possibly prevent competition in respect of products marketed from years 10 to 16 for the simple fact that, to the extent that they covered such products, they would be invalid for lack of novelty. (364) The fact that generic companies are becoming increasingly active and filing patent applications may partly explain Table 16 and Figure 42 (paras (335) to (337), p ). Third parties including generic companies may have more applications on top-selling INNs than innovators. 2. COMPETITION BETWEEN INNOVATOR AND GENERIC COMPANIES THE ISSUES 2.1 Patent Filing and Patent Enforcement Strategies Throughout this chapter, DG Competition makes statements which are unsupported by the data presented and/or show a misunderstanding of the patent system. It cannot be objectionable to obtain patent protection for any innovation (whether primary or secondary ) which has been found by a specialist and independent authority (such as the EPO) to present the required novelty, inventive step and industrial applicability. Accordingly, extending the breadth and duration of patent protection through innovation is perfectly legitimate, although one patent cannot of course extend another. As said above, such innovation arises throughout the product lifecycle and is not reserved to the innovator company holding the base patent. The data presented in the PR do not support the statement that numerous patents are being filed covering the same medicine with the effect to delay or block market entry of generic medicines. As appears from Figure 13, generics enter the market more likely, more frequently and more quickly with high value, densely patented products. Similarly, when comparing Figure 14 with Figure 41, it seems that generic entry is generally earlier in the most densely patented countries (compare the UK with Luxembourg for example). In any event, the reference in this respect to the overall number of national patents or designations on an EU-wide basis is misleading. It is equally inaccurate to state and there is no evidence to support the proposition that the so-called patent clusters would create uncertainty for generic competitors as to whether and when they can start to develop a generic medicine without infringing one of the many (new) patents. Indeed: Pharmaceutical patents do not block the development of competing products by generic companies. In view of the research exemption and the Bolar exemption (which was introduced in 2004), generic manufacturers can now start the development of their products in the EU and even apply for a marketing authorisation long before the expiry date of any patent that would cover that generic product. The Bolar exemption only came into effect late in the period examined by the PR and thus many of the generic INNs discussed in the PR would have been developed at a time when there was no such exemption. Consequently, unless generic companies developed products either unlawfully or offshore, the absence of the Bolar exemption would have inherently delayed #

72 launch after loss of exclusivity. 27 The difference between the pre-bolar and post-bolar era is not acknowledged when the PR talks about delay following loss of exclusivity; the PR simply takes the average alleged delays throughout the entire period. Patents are only granted if the claimed invention is novel, has an inventive step and is capable of industrial application. If a patented invention is believed not to meet these conditions, competitors are able to initiate steps to clear the path in the early stages of the development rather than waiting until they are ready to launch. Whilst generic companies cannot design around the basic compound patent, it is possible to design around patents claiming a secondary process, solid form or formulation. Generic companies have indeed proven to be very adept at doing so. It is not inevitable to apply the teachings of one of the secondary patents to bring a medicinal product with a given active ingredient to market. In short, the alleged uncertainty can be reduced or removed in various ways such as (i) clearing the path; (ii) designing around; (iii) submitting manufacturing and product details to the innovator company and seeking clarity; or (iv) obtaining a license. Next to the alleged patent clusters, divisional patent applications are said to be a second instrument used by innovator companies. According to DG Competition, the fact that the examination of divisional applications may continue even if the parent application is no longer in play (i.e. withdrawn, abandoned, refused or revoked), may add to the legal uncertainty for generic companies. First, EFPIA stresses that divisional patent applications cannot, by their very nature, expand their scope beyond the parent application. Nor can they extend the protection period beyond that of the parent application. As such, they cannot be held to have any additional blocking effect. Next, there are several (legitimate) reasons for filing divisional applications (unity of invention being the most important one): a disunity objection issued by the EPO 28 ; to obtain rapid grant for a part of the claims of particular interest to the applicant that has been held patentable by the EPO, a divisional application then has to be filed for the rest of the subject-matter of the parent application which is of interest to the applicant; as a safety device ahead of an ex parte appeal in case the application could be rectified in principle but, due to procedural restrictions, cannot be rectified during oral proceedings; before grant in the expectation of an opposition so that there is scope for pursuing different subject-matter if issues unexpectedly arise during the opposition that can be or are most easily rectified pre-grant in the divisional application; for entitlement reasons, for example different subject matter may be owned by different parties At the time of introduction of the Bolar exemption, the EGA confirmed this by saying that "Waiting until patent expiry [to commence R&D] delays market entry for generic products, costing European patients and healthcare systems as much as 100 million per medicine per year (see For example in the field of heterocyclic chemistry, if there is a novelty objection against one ring system covered by the claims, the EPO will often require separate applications to be filed for the other ring systems claimed. #

73 When discussing the ways in which these divisional applications might deter generic entry, DG Competition particularly refers to Article 67 EPC (see e.g. para (380)), stipulating that provisional protection, including damage claims, can be conferred upon the applicant if so stipulated under national law. If this happens at all, in the experience of EFPIA members it is extremely rare. Finally, DG Competition concludes that enforcement of patents would be somehow inappropriate as litigation can be an efficient means of creating obstacles in particular for smaller generic companies due to its cost. There is no evidence to support this. Further, even if cost can be a factor in determining how a small company responds to litigation, this should not restrict bona fide enforcement of patent rights. If cost is an issue it should be addressed by the patent enforcement system and not by limiting the right to enforce the rights per se. Any such weakening of IP rights would have a severely detrimental effect not only on the pharmaceutical sector as a whole but also on any innovative industry and, ultimately, European consumers and patients. (371) All patents are intended to provide exclusionary rights for a limited period of time, and do so by protecting the patented invention from being copied. They inhibit imitation and promote innovation. Such is the essence of the patent system. (375) DG Competition seems to take the view that patent protection for pharmaceutical products should only last until the expiry date of the so-called base patent (see for example Figures 52 and 53 where expiry of the base patent is taken as loss of exclusivity). This makes no sense from a patent perspective as all inventions which meet the conditions of novelty, inventive step and industrial application, are worthy of patent protection. (376) Each patented invention is to be seen on its own merits and separate from any actual product on the market. Once a patent expires or is revoked, its subject-matter can be freely copied by others, subject only to the existence of other validly granted, prior existing patents. No amount of subsequent patenting can change that fact. (377) Patents create valid, enforceable rights so long as they involve a novel and inventive subject which is capable of industrial application, whatever the intent may be prior to applying. (388) The fact that innovator companies have successfully enforced secondary patents does not mean that they were obtained or used illegitimately; indeed it indicates the opposite, nor does it prove that such patents are not capable of being designed around by generic companies. (390) The statement that the relevant (sic) patent applications are typically filed in anticipation of imminent generic entry is incorrect: see Figure 48, showing a continuous R&D/innovation effort by innovative companies. (391) There is no such thing as a traditional patent portfolio lifecycle. (392) An increase in secondary patents is a sign of incremental innovation. (393) In all sectors, some patents are stronger than others. It is misleading to imply that secondary patents are, by their nature, weaker. Secondary patents are neither inherently weaker nor inherently stronger than primary patents. The quotes following para (393) merely acknowledge that: #

74 Some patents (primary and secondary) are stronger than others; Secondary patents are narrower than primary patents and thus the breadth of a secondary patent portfolio will be narrower than the breadth of the primary patent. Because they are narrower, they are easier to design around. It is simply wrong to state that the majority of litigated patents were revoked and that this could be a sign of an increase in weak patents. Footnote 257 discloses that in less than 30% of the litigations that proceeded to judgment the patent was annulled. (397) The mere fact that generic companies say that they see patent thickets as an obstacle is not evidence of anything. The PR shows that generic companies are able to enter the market more quickly for densely patented INNs than for others. (400) In the absence of absolute numbers and further specifics, it is impossible to assess the general statement of DG Competition. It also fails to mention how many of the divisional patent applications of which the parent had been refused or withdrawn, were eventually granted. Moreover, the facts presented are meaningless unless the relationship between the claims of the parent and the divisional claims is given, together with reasons for the withdrawal/refusal of the parent. If anything, this paragraph illustrates the robustness of the EPO process. (401) It is perfectly normal for broad (parent) applications to be refused while more narrow (divisional) applications can be granted. Again, this is not specific to the pharmaceutical industry. No data are given on how many of the divisional applications were granted in cases where the parent application was refused or withdrawn. (402) It is perfectly legitimate to file a divisional application if it is desired to obtain a quick patent grant in relation to core subject-matter of importance and to preserve rights in the broader subject matter (to which the applicant feels it is also entitled to patent protection) in a divisional application. (408) The objective evidence in the PR concerning the speed of generic entry (rather than generic company hearsay evidence) does not support the conclusion that divisional applications are causing significant problems. (410) All patents are filed to prevent others from applying the claimed invention and thus can only prevent or delay entry from any competitor who chooses to copy this invention. This applies not only to basic inventions but also to patented improvements or alternatives, and is not specific to the pharmaceutical industry. In fact, it is the very heart of the patent system. Patents foster competition through independent innovation by third parties. (412) It is certainly not impossible for a generic company to obtain legal certainty prior to launch. There are several methods of clearing the path to determine whether launch will be lawful or not. (415) DG Competition seems to object to late filed secondary patents as well as early filed secondary patents. #

75 (422) This correctly summarises the nature of patent rights and the legitimacy of enforcing them and should be at the forefront of the PR. (423) The suggestion that divisional applications are an indispensable asset for innovator companies implementation of their procedural enforcement strategies is unsupported by the data presented. (425) As paragraph 422 notes, enforcement of patents is legitimate procedural dimension of the material right granted. (428) Generic companies have a strong incentive to do everything they can to be the first entrant in a particular market. This includes attempts to launch before the loss of exclusivity date, in which they sometimes succeed. Being strong on patent enforcement discourages infringement generally. Innovator companies who are seen as being soft on enforcement may become targets for premature generic entry. Companies should not be criticised for defending their legal rights. (430) Contrary to what is implied, it is certainly not easy for an innovator company to obtain an interim injunction. Moreover, the potential impact on the innovator company if no injunction is obtained is enormous. Failure to obtain an interlocutory injunction may in many cases mean that even a good case settles as the market is effectively destroyed in the period until trial and no realistic compensation can be obtained (see IP annex to EFPIA s submission of 13 June 2008, p. 25). In any event, there is no evidence that the threat of incurring litigation costs or interim injunctions actually deters generic entry. 2.2 Patent-Related Exchanges and Litigation As for other sections in the PR, the data reported are questionable because respondents will have answered questions in different ways, using different categorisations and different applications of definitions. These will include the use of terms primary and secondary patents, which are not terms with any precise meaning in the patent field. Some EFPIA members consider the terms to be arbitrary. The section also seeks to categorise patents as product, process, use and other. Each category of patent may have claims to another category, and process claims cover products also (because any direct product of the claimed process will be covered). Therefore conclusions based on patent category may be subject to question because of the inaccurate use of terminology. The data sample is in fact very small, in particular for the win/loss statistics. These statistics also do not appear to take into account the substantial number of cases which went to judgment in favour of the patent holder but were settled before appeal. Having said that, the data which are reported show the overall win/loss ratio is close. The same data reveal, contrary to what the PR states, that the majority of patents are held valid in litigation. In addition, unless all respondents applied the definition of Loss of Exclusivity in the same way, this also calls the data, its analysis and any conclusions based thereon, into question. #

76 Patentees must be free to enforce their rights. Patents provide a right to sue for infringement and if this is watered down or fettered, the very value of patents is called into question. Such an approach is contrary to the rationale for the system and to the Enforcement Directive. As a general principle, the sending of warning letters is entirely appropriate. They are often either required or actively encouraged by the judicial systems in Member States failure to do so can have adverse consequences for patentees. Also as a general principle, patentees/parties to litigation should raise all issues at the same time, and not on a sequential basis. Failure to do so will mean they cannot raise these issues later, as this would merely drag out a dispute. Therefore, it is entirely appropriate that secondary patents be raised at the same time as primary patents. However these terms are applied, it will mean that older patents will be raised along with newer ones, and therefore secondary patents are likely to be raised early in their life by comparison with primary ones. This reflects good practice. There is no evidence to support an assertion that contacts and disputes create delay for generic entry. It is unsurprising that the most valuable INNs at a given time will attract the greatest number of contacts, disputes, litigation and settlement. These INNs will have attracted the greatest interest from generics who will generally only go after, or at least will favour, the most valuable products. Everything else will follow. Interim injunctions are a fundamental part of the patent and indeed intellectual property enforcement regime. They are now enshrined in the Enforcement Directive, which requires all Member States to have in place procedures to make these available. They are particularly important for innovators since without them early entry into the market, in violation of patent and other rights, could not be stopped and would result in catastrophic loss of market value. This key impact for innovators, which was addressed in EFPIA s submission of June 2008 is not discussed in the PR, which focuses only on the impact of such injunctions on generics. The courts have their own procedures for assessing requests for interim injunctions under national law and should be allowed to do so. Further, they also have in place safeguards to protect parties who, it turns out after the event and after trial investigating all the facts and law, should not have been injuncted. While many of the problems and uncertainties reported in the PR could be avoided if innovators were able to sue earlier than is currently possible, there is no evidence that generics are in fact put off by contacts, disputes and litigation. Also, it is largely in the hands of generics to clear their path in advance of product launch. While the report itself indicates that many do so (following encouragement from the courts), many more could also do so to remove or reduce the uncertainty reported. There is no clear evidence that pharmaceutical patent litigation is increasing per Member State. Expansion of the EU has led to more disputes overall, and will probably continue to do so. The data on inconsistent decisions are questionable yet encouraging. They do not for example consider whether the claims in issue were the same, whether the same prior art was raised or whether the same evidence was available. Even so, a 10% figure for inconsistent decisions is not all significant. Having said that, a good and usable single patent enforcement system would reduce inconsistent decisions. There is no correlation between the number of contacts, disputes and litigations by country and delay. In fact, the data show that in many countries with a large amount of litigation time to entry is short. #

77 As regards costs of litigation, the PR does not mention in this context at all the potential to recover costs. In the higher cost countries like the UK, recovery is expected to a large extent. In addition, the Enforcement Directive requires Member States to allow for this. In any event, the number of cases initiated by generics suggests that generics are not generally put off by the cost of litigation. (433) It is a fundamental part of the patent system that patent owners are able to enforce their patents. To have any suggestion otherwise undermines the whole system and value of patent rights. DG Competition states that it has not drawn any conclusions as to the compatibility of patent related exchanges and litigations with EC competition law but nevertheless proceeds to imply in many instances that it has. (436) Paragraphs 436 and 438 suggest disputes and contacts have a dissuasive effect on planned generic entry. This is speculation and there is no evidence that this occurred; far less is there any evidence that where it did occur, there was not a valid patent in issue. A concern that there is a risk of damages is a reflection that the rights asserted were considered valid. (438) 1,337 disputes and contacts with regard to 219 INNs over a period of eight years does not amount to a large number (0.77 per INN per year). Furthermore, only 457 of this total were actual disputes as opposed to mere contacts (0.26 per INN per year). (440) Paragraph 440 shows that the disputes and contacts related to only 80 of 219 INNs. This is an average of 17 disputes per INN. Given that there may be multiple potential generics in multiple EU markets, this is not a high figure. (453) If primary patents are in dispute then the majority of disputes arise prior to, or at least in relation to matters that arose prior to, loss of exclusivity (however interpreted). This may be an indication that generics are prepared to take the risk of launching during the term of the primary patent because of the advantage of being first to market. (454) The conclusion here simply cannot be drawn. It is necessary to consider the nature of the disputes, how many were withdrawn, how many settled, the timing and the outcome. As a general principle, raising potential disputes before commencing litigation is actively encouraged by the courts so as to minimize, or narrow the matters involved in, litigation. It is legitimate and appropriate for patent holders to challenge infringement of their rights. (456) Without hard data to support it, the suggestion that disputes and contacts can have had any widespread significant effect on generic entry is speculation. (457) The example given is typical of warning letters sent not only in the pharma industry but for any patent dispute. National law will influence and can dictate the content of such letters. It is clear the inquiry did not appreciate this. It should be noted that national law (e.g. Section 70 of the UK Patents Act) provides remedies for groundless threats of infringement proceedings so as to guard against ultimately groundless threats of litigation. (459) The data given in figure 61 is unsurprising. It is inevitable that secondary patents will generally be raised at an earlier stage in their life. It is again a good thing that patentees do not raise their various rights on a sequential basis and indeed courts generally do not allow them to do so all issues should be raised at the same time and not later on. #

78 (466) Reference is made to interim injunctions having a very serious effect on generic companies but the report does not take into account the effect of an absence of such an injunction on innovators. If generics consider patents to be invalid and/or not infringed, they are free to raise this with the patentee prior to a product launch. In practice, few do so save in specific Member States such as the UK where this is required. Therefore, the risk and costs of patent litigation are frequently the result of a failure of generics to take steps to limit risk. If production is carried out outside the EU, it will not cease in these circumstances. Commercialization of the product outside the EU may continue and some generic companies operate internationally. In the event that the generic is correct and the patent is invalid and/or not infringed, any injunction restraining commercialization in the EU will be accompanied by a bond and/or cross-undertaking protecting the generic from damages. In addition, in most countries, and in particular following the introduction of the EU Enforcement Directive, any parties to litigation can recover costs if they are successful. None of these matters are discussed in the report. (468, footnote 247) Statistics based on unique case numbers are meaningless since courts in different jurisdictions assign case numbers in different ways. For example, some courts will assign one case number for a litigation involving several patents while other courts will assign a unique case number for each patent in dispute. In Germany, infringement and validity disputes involving the same patent will be assigned different case numbers. In the UK for example, if 5 generics challenge the same patent, 5 different action numbers will be assigned. This will both increase the number of cases reported and, if the generics are successful, inflate the success ratio in favour of the generics. The converse is true but it does raise questions as to any conclusions that may be drawn regarding the number of patents successfully attacked. It is to be expected that the most successful INNs will be subject to the most litigation as innovators will be faced with entry by more generic companies in more countries than for other products. This is especially the case as it is acknowledged elsewhere in the report (para 595) that being first to the market is important for generic companies who will, therefore, be seeking to enter as soon as they can. Interestingly, there is no comprehensive data as to how many cases were litigated in several jurisdictions at the same time (Fig 67 has some data but does not distinguish between types of action). There are no data on how many of those cases were fought in more than one jurisdiction when judgment had already been given in one jurisdiction. (469) Contrary to the suggestion in (466), this paragraph acknowledges that generic companies are proactive in initiating litigation to clear the path. Generic companies initiated almost as many lawsuits as innovators. As they know what they will be doing before innovators, this could have been done more often. The suggestion that generic companies are intimidated by the mere threat of litigation is not supported by the facts. (470) Absent data prior to 2000, and given the up and down trend year on year over the period of the Inquiry, it is not possible to draw the conclusion that there is a substantial increase in patent litigation. If there is an increase, there may be many explanations for this, including EU expansion. The variation in the number of cases from year to year makes it impossible in the data period to identify any particular trend or assess the causes of #

79 any such trend. (474) The report does not give data on how many patent cases were started in all industries. Without such comparative data it is difficult to assess whether the amount of litigation in the pharma sector is high compared to others. (475) Closer examination of the data shows that in 10 EU countries, the innovators have initiated the majority of the litigation. In eight countries, generics had done so. Although 10 is greater than eight, the difference is not significant. Some countries had so few cases that the data cannot be statistically significant. It is notable that a significant number of cases were commenced by generics in Italy. This could have been a result of the so-called torpedo tactic which was intended to have the effect of halting litigation under the same patent family elsewhere in the EU. This practice, commonly referred to as the Italian torpedo was only declared ineffective by the ECJ during the period covered by the inquiry. (476) One of the reasons why there is a high number of cases in the UK is because the courts there have encouraged generic companies to clear the path before launching a product. A reason for the high number of cases in Italy is given above. (481) Figure 67 confirms that an average of 10 litigations were brought for each of the 68 INNs. This number for a period of eight years is not big. Even INN 1 is the subject of only 100 litigations which over a period of eight years and with respect to a possible 27 Member States is not significant. DG Competition does not indicate how many generic companies were involved, which might explain the high number for INN 1. For the last 10 INNs the number of litigations is less than 18 over an 8-year period and in relation to 27 Member States. (482) Of the 219 INNs investigated, these data show that 151 were not subject to litigation (69%). (491) Given that the PR says there are seven times as many secondary patents as primary ones, it is not surprising that the majority of all litigated patents were secondary. However, the fact that over one-third of litigated patents are primary patents indicates the willingness of generics to launch at risk. (493) The contrast reported is not surprising. If generics are starting almost as much litigation as innovators, this will increase the number of secondary patents in issue because they will challenge more secondary patents than primary ones. In addition, when it comes to litigation by innovators, this is likely to occur once a generic has launched and it may be clearer which secondary patents have been infringed. (496) Primary patents are litigated later in their lifespan in part because, although generics can bring actions for declarations of non-infringement and revocation at any stage, innovators can only sue when there is launch or imminent threat of launch. This is likely to be late in the life of a primary patent. (500) The data show that approximately one-fifth of the reported litigations resulted in a final judgment. Without collating the settlement terms and/or the final outcome of the remaining four-fifths, a 20% sample of all litigations reported cannot lead to any clear conclusions as to the overall result of all litigation in the inquiry period. Because of the small dataset, if there is only a small number of cases in which the same #

80 patent has been challenged by a number of generics, it could have a significant effect on the ratios of both generic initiated and overall generic wins. Para 630 indicates that in about 50 cases that settled, the innovator had won at trial. It seems (but is not clear) that these figures are not included in the win/loss ratios. If they were, it would make a large difference. (There is no data on how many cases settled after the innovator lost at trial). In any case, the basic win/loss ratio (38/62) does not seem unreasonable. However, DG Competition gives no data concerning other sectors so no comparison can be made. In addition, statistics based around averages are unreliable. Figure 75 shows that the average duration of litigations ending with a final judgement in the Netherlands is approximately 30 months. Figure 79 shows that the average duration of an interim injunction in the Netherlands is 44 months. The apparent contradiction of having interlocutory injunctions lasting longer than the litigation is explained by the fact that such averages are inconsistently arrived at, either through data gathering or analysis differences. (502) Ignoring the validity of the data, it shows that generics won 92 cases, out of 698 litigations (or just 13%). The footnote to this paragraph shows that 28.8% of patents were annulled/revoked. This contradicts the statement in paragraph 393 that in the patent litigation cases between innovator and generic companies reaching a final judgment, the majority of litigated patents were revoked. That this latter statement appears to be a misrepresentation of the data is unfortunate. (504) Bearing in mind the inherently uncertain nature of patent litigation in any industry, in any case a fifty-fifty win: loss ratio on cases innovators start is perfectly fair. There is an inherent selection bias in the statistics relating to generic initiated litigation. (508) A non-infringed patent is still a valid patent. On this limited data, about 36-38% of secondary patents were annulled. It is logical that secondary patents may be less successfully litigated by innovators on the basis that secondary patents are narrower and therefore it is easier for generic companies to design around (34% found to be not infringed). (509) As noted above, process patents may involve cases concerning the products covered by process claims. It is difficult therefore to know what the real numbers are in relation to success and/or otherwise of either innovators or generics in relation to product and/or process claims. Respondents may have considered litigation under a process patent to be either product patent litigation or process patent litigation when responding. As regards second medical use patents, it should be noted that until relatively recently in some Member States such use claims were not allowable under national law. They are now. Further, some Member States courts have altered their opinion on method of treatment-type claims for example dosing regimen claims. These are now allowed by the EPO and the courts accept such claims when in the past they might not have done because they are seen as good and beneficial to the public. (514) Since DG Competition does not state whether the litigation was successful on grounds of non-infringement or invalidity, it is not possible to make the statement that the strength of these patents is more limited. It is an unsupported conclusion. If an #

81 innovator loses a process patent case, it may very well be due to the burden of proof or scope of the claim (thus the patent found not infringed) rather than the patent being weak. In some instances, a similar point could be made regarding second medical use patents. (516) The timescale for French litigation appears surprising in the light of EFPIA members experience. It may be because many cases have not been resolved despite a lengthy period of litigation. The data may therefore be of no significance. (Reference is made to footnote 274). (520) If innovators were able to address the question of infringement sooner (by having a right to sue before market launch and with sufficient information to do so), the need for interim injunctions would be much less. (521) The data show that interim injunctions were only sought in approximately one-third of cases, and approximately one-half of these requests were granted. So injunctions are only granted in one-sixth of cases. This does not suggest that interim injunctions are in reality a major source of difficulty for generics which may be why many generics will launch at risk. Significantly, there are no data reporting the number of cases in which win/loss follows an interim injunction. In those cases if an interim injunction is granted and the innovator succeeds at trial, there can be no questioning of the delay in entry. If no interim injunction is granted, there is no delay. There appears to be an inconsistency in this paragraph between the reference to 255 requests for interim injunctions in paragraph 521 and the reference to 225 requests in Figure 77. (522) These data do not take into account the number of cases in which an undertaking not to launch is voluntarily given by the generic. Such undertakings are increasingly common in the UK. (524) The period of an interim injunction is determined by the time that it takes a court to reach at least a first instance decision. It is not clear from the data which of the interim injunction cases turned into final injunction cases following a decision on the merits, which cases took longer to get to trial because of national procedures, and which took longer because of the complexity of the case. (527) The only thing that can be said is that the reasons for granting interim injunctions and the length of those injunctions, vary across the EU. Note also that footnote 274 shows a disparity between the length of patent litigation in France and the duration of interim injunctions. The data on the length of patent litigation in France appear to be flawed. 2.3 Oppositions and Appeals The same questions are raised concerning categorisation, definitions and consistency of response as have been raised above in relation to section 2.2 and elsewhere. The number of opposition proceedings (whether in totality or by INN) is not particularly large. #

82 The PR suggests there is a high percentage of patents subject to opposition in the closest proxy to the sector. While this proxy is questionable, it is just 1-2% above the average for the EPO. In addition, the inference that seems to be drawn that this somehow is an indication that patents in the sector may be weak cannot be drawn. All that can be said is that in this sector (or at least, the proxy) a slightly higher percentage of patents are subject to opposition. The most logical explanation is that there is more commercial interest in patents in the sector than in some others, and this attracts more opposition. No more can be inferred from this information. It is wrong to categorise an amendment to a patent in opposition as a success for an opponent. Often amended claims still cover the opponent s product; thus these are really a success for the patentee/innovator. This fundamental misconception indicates a lack of understanding of the patent system. DG Competition reports that so-called secondary patents are more often opposed than primary ones. Given that an opposition must be commenced within nine months of grant and in the case of primary patents, the likelihood is that the patentee will not have launched a product until long after the opposition period has expired. A successful product may be more likely to attract oppositions. While it is accepted that opposition proceedings take too long, as is the case in all sectors, there is no evidence that opposition delays generic entry. On the contrary, slow opposition proceedings can facilitate generic entry by affecting the innovators ability to pursue national enforcement and obtain injunctions. There is also no evidence to support an assertion by a generic company that innovators seek to delay opposition. Most of the delay in opposition is in fixing the date for oral proceedings. While national courts approaches to opposition vary, to the advantage or disadvantage of both parties, many national courts will accept national validity challenges while opposition is pending. If generics are concerned about speed of opposition, these procedures can be used. (539) A patent will not be invalid if its claim scope is reduced. The purpose of a limitation is to provide valid claims to a patent. Limitations may or may not mean that products no longer infringe the patent. In many instances, the limitations made in opposition do not lead to products escaping infringement as a general matter, innovators would not propose claims that generic products would not infringe. This is a point which appears to be missed in the whole section on oppositions. (541) Oppositions and appeal are not separate and distinct procedures. There is first an opposition and, following a decision of the opposition, there may be an appeal. Appeals sometimes remit cases back to the opposition division. This is exactly the same as is the case in litigation. The final sentence reveals the reality opposition and appeal, just as first instance litigation and appeal, need to be taken as a whole. (542) It is true that some national courts stay proceedings pending an EPO opposition. It is also true that the same courts generally accept a decision of the EPO and will not decide otherwise on validity unless the opponent and/or defendant raises new prior art that has not been considered by the EPO. Courts that do not stay proceedings tend not to treat themselves as bound by the outcome of an opposition (except where opposition or appeal revokes the patent entirely). (544) 170 opposition procedures over a period of eight years and 73 INNs. It works out at an average of 0.30 oppositions per INN per year. #

83 (545) The data show that opposition procedures increased from 5 in 2000 to 34 in 2005 and have since decreased. The number of oppositions in 2007 is almost exactly the same as that in Without data prior to 2000, nothing if anything can be said about any particular patterns or trends. (547) A straw man is typically used where a generic/opponent wishes to avoid being identified. This further increases the uncertainty for innovators because they are unable to identify where there may be infringement. (551) A61K is questionable as the closest proxy. Without examining the data for all industry sectors, sector by sector, it is impossible to draw any conclusions about the opposition rate for A61K in any case. Opposition rate does not (as noted below) say anything about the validity of patents in the sector it merely shows there are more parties interested in opposing patents in A61K than the average across the EPO. This suggests if anything that patents in the sector are more vigorously analysed than in some others. (553) This paragraph gives unsurprising data regarding the duration of opposition. There is nothing unusual about the pharmaceutical sector in this respect. (555) There are good reasons for allowing a patent holder additional time to respond to an opposition (or multiple oppositions) filed at the end of the opposition period. (558) In practice, where opposition has been commenced, many Member States do not look at validity again. Those that do are often those that have speedy procedures for resolving validity. (564) There are a number of reasons why primary patents (however defined) are less likely to be opposed. An obvious one is that if, as is likely, a primary patent is filed early in the life of a product, or before a product is launched, a generic will not know whether or not the product is of any interest to them. The opposition period will therefore expire. By the time secondary patents, or later primary patents, are filed it would be known whether the product is one that is worth a generic seeking to copy. (565) This comment is pure prejudice and not based on fact. The profile (in terms of outcome) is not peculiar to the pharmaceutical industry. It is inappropriate and purely prejudicial to include a quote that says there is heavy abuse of the patent system. Further, this shows that at least one generic company believes the opposition procedure to be effective. (569) The conclusions in this report reveal a basic misunderstanding of the outcome of opposition. The report concludes that where a patent is amended, as opposed to revoked, it is nevertheless always considered a success for the generic company. This is misplaced. There is no consideration as to whether the amended claims of the innovator company s patent will still cover the generic product and/or process. This is never considered by the European Patent Office. In many cases, patents amended in opposition proceedings will still cover the products of the opponents the patentee would not knowingly propose an amendment that had a different effect. Thus, the conclusion as regards overall success rate is fundamentally flawed. #

84 (571) Settlement of oppositions is commonplace in all industries. Settlements would be entered into on a case-by-case basis and the incentive for doing so on both sides is to remove the need to spend further time and money on pursuing the process, and to achieve certainty as to what the opponent may or may not do. 2.4 Settlement and other Agreements As a general matter, settlement of disputes and litigation is encouraged by the courts and dispute resolution organisations in Member States. It reduces litigation costs, provides certainty and frees up time for both the parties and court services. The most significant finding in the PR is that in paragraph 576 which correctly states that patent settlements are fact-specific, depending on the dispute at issue, and reflect the negotiated positions of the parties. It is not surprising that the number of settlements by INN mirrors the trend for disputes, contacts and litigation. The reasons are the same. These are the most successful and valuable products covered by the inquiry, and will attract most interest from generics who may decide to proceed at risk and infringe the patents of the innovators. There is no evidence that settlement led to delays to generic entry. It is incorrect to suggest that the factors which innovators consider when deciding on settlement differ from those considered by generics. All parties to all litigation will consider the same basic factors when deciding to settle, and one of these will always be prospects of success. Other factors identified are common to settlement of all litigation. The way in which agreements have been considered to have a value transfer or are restrictive is inappropriately broad. Reimbursement of costs or payment in lieu of damages are natural and commonplace terms of settlement, as is acknowledgement of patent validity and the taking of a licence (which the PR considers restrictive rather than permissive, without analysing the nature of the terms in the licence concerned). It is also unsurprising to find ancillary agreements entered into on settlement parties entering into a commercial arrangement way well consider entering into other arrangements of mutual benefit at the same time. It is not correct to say there are more similarities than differences between US and EU settlement agreements. While basic reasoning for settlement will be the same in all litigation (certainty, costs etc), the circumstances are very different in the US and the EU for the pharmaceutical sector. Simple comparisons cannot be made. Action taken by the Federal Trade Commission has been overturned by the courts in the United States. (574) Whilst DG Competition states that its aim is not to provide guidance on certain types of settlement agreements, the way the section is written suggests that certain undefined types of settlement agreement may not be acceptable. (575) Settlement of disputes is strongly encouraged by the various European courts and dispute resolution bodies. #

85 (576) The statement that patent settlements are fact-specific, depending on the dispute at issue and reflect the negotiated positions of the parties correctly states the position. Much of the remainder of this lengthy section includes facts and commentary which simply reflects this. (578) The concluding sentence in this paragraph is close to the correct position. Settlements are an accepted way of ending disputes. (579) It is noted that the Federal Trade Commission in the USA has had its decisions reversed by the US courts. (581) It is difficult to identify any trend from Figure 90, far less the causes of any trend. (582) It is difficult to identify any trend from Figure 91, far less the causes of any trend. (586) This shows the INNs for which there were more settlement agreements were the more valuable ones. This is logical. (588) As noted above, settlements will typically include more than one geographical area. This is good because it reduces the number of litigations not just in a single Member State but in multiple Member States. Courts would view settlement of all disputes relating to a particular matter, product and/or patent as sensible. The fact that there may be more settlement agreements concluded in countries with a high pharmaceutical market value is not at all surprising. There are more disputes. In all industries, markets where there is a high value for patented products produce a greater number of disputes because the incentives for those wishing to sell copy and/or competing products are greater and the risk of infringing patents may be worth running. This trend should, if the suggestion that litigation and settlement lead to delay has any basis, correspond geographically in terms of delay. There is no correlation revealed in the PR. (591) Without knowing precisely how each respondent considered how to answer the questions, it is difficult if not impossible to draw any conclusions as to which factors are more important than others for any company settling litigation. In reality, all companies will consider costs, prospects of success and commercial risk/reward when deciding whether or not to settle litigation. Also, patentees will frequently decide to take a stand when dealing with an infringer or persistent infringer, where it wishes to make it clear that it will not let infringement of its rights go unchallenged. This is true for all industries. The factors involved in settling litigation include costs, complexity, strength of case, commercial necessity and risk. All will be considered by all parties to litigation. It is noted that in paragraph 592, the risk of large damages claims from a patent holder is perceived as a reason to settle. Whilst in many cases such a risk is misconceived (damages claims for patent litigation in Europe are not large), if such a risk is genuinely considered by a generic company, it indicates that the patent which is being asserted against them is perceived to be both valid and infringed. This further illustrates that prospects of success are likely to be equally in consideration from the generic side as they are from the innovator s side. (593) This paragraph reports inter alia that for generic companies the risk/success evaluation can be complex. The same evaluation is made by the innovators and has the #

86 same complexity. Complexity and risk could be reduced if generic companies identify their processes and products and details thereof at an early stage. It is noted that the quote in the box directly supports the point made above that risk analysis and uncertainty of outcome is precisely the same as evaluating the strength of a case. (594) For an innovator company the failure to obtain an interim injunction can be very damaging because it can lead to a drop in price which is irrecoverable, even if, ultimately they succeed in the final outcome of the litigation. This provides a strong commercial incentive to settle a case rather than pursue it to its final outcome. The quote in the box merely reflects this reality. Also, it is possible, where an interim injunction is granted, that the patent may expire before the litigation is concluded. Settlement then makes sense. (595) It is noted that in paragraph 595 it is recognized by the generic company that there is a considerable risk that continuing the litigation would result in being excluded from the market for the entire patent term. This is recognition that the chances of being injuncted, because the patent asserted against them is both valid and infringed, are high. Settlement could therefore lead to an earlier generic entry than might otherwise have been the case if the litigation were continued because the patent is valid and infringed. (596) Considering the commercial value of any patented product will always be a factor when discussing settlement of patent litigation. (597) Duration of litigation is merely another factor that is taken into account in settling all patent litigation. If a patent expires, then the possibility to obtain an injunction, save in limited circumstances, is removed. This may provide a greater incentive to settle the case. (598) The discussion in paragraph 598 is true of all litigation (not just patents). Different approaches can be found even within Member States different judges may take a different view and/or may be presented with different evidence, depending on the case. This is the case in all litigation, which can never be a perfect resolution of a dispute. All litigants take part in forum-shopping for litigation, whether contract, tort or intellectual property cases. (599) It is not correct that whenever a settlement agreement is concluded there is no decision on the patent s validity. For example, settlement may take place after a first instance decision but before appeal. Furthermore, to suggest that the patents remain as potential obstacles to further generic competition ignores the fact that many settlement agreements are entered into in which the decision has been taken, or the determination has been made by the generic, that the patent is in fact valid and infringed. The assumption in this paragraph is that settlement takes place only where patents are perceived to be invalid and/or not infringed. This is not the case. There is nothing to stop other generics raising similar issues on the patent s validity and/or noninfringement. Where no court decision has been made, there is no decision which would have any sort of adverse impact on such a course of action. (605) This section purports to place, in order of importance, certain factors taken into consideration by either (a) innovator companies or (b) generic companies. The reality is that all of these factors will be considered by all companies when considering whether or not to settle litigation. It is impossible to rank such factors in order of general importance by party and different companies will have answered questions in different #

87 ways. (612, Figure 98) The categories used seem to be fairly artificial. Category B includes those which have a limitation on generic entry with or without a value transfer. Without full knowledge of the terms of the agreements and the economic and legal situation surrounding them, it is not possible to determine whether they have been correctly categorised in any event. (615) The section from paragraph 615 to 646 reviews settlement agreement types for pharmaceutical patent litigation cases in the EU. As a general comment, all that this reveals is again reflected in the summary conclusion in paragraph 576 that patent settlements are entered into on a case-by-case basis, being fact-specific depending on the issue at hand. (623) This highlights a situation in which an interim injunction has been granted but later it transpires that, as a result of the decision in the main action, it should, with the benefit of hindsight, not have been granted. Such an event is typically provided for by a crossundertaking in damages or a bond, from the innovator. Going through a further exhaustive process of fixing the level of damages is expensive and time consuming. It is often expedient simply to settle this. (629) It would appear that all of the agreements set out in this paragraph represent a total success for the innovator company. This needs to be taken into consideration in the context of the overall success/failure ratio for litigation between innovators and generics generally as it would appear that these were in effect successes for the innovator without final judicial resolution of the dispute. Indeed it would appear from paragraph 630 that there had been a first instance decision in favour of the innovator in almost all the 54 B1 cases. (631) The two examples given in the boxes are situations in which the parties consider that the innovator was going to succeed in the litigation. The generic company was restrained from entering the market and/or was obliged to pay damages to the innovator. In other words, the parties thought that the innovator was going to win. They decide to settle the litigation rather incur further costs, and further damages from the defendant. Accordingly, the monies paid to the innovator in these circumstances represent compensation for the expenses and loss incurred by the innovator. The notion of value transfer does not sit well with this situation. (633) The fact that different types of value transfer were involved indicates again that settlement is fact and issue-specific. 22 agreements allowed the generic companies to launch a competing product. (634) A generic company which agrees not to enter until the court has given judgment on infringement would appear to be one which recognizes the merit of the innovator s case. (638) The granting of licences and the waiving of claims are common provisions in settlement agreements in all patent litigation. (642) Supply of product itself or the active pharmaceutical ingredient can improve the position of the generic product. (647) It is observed that the regulatory and patent litigation regimes in the United States are very different to those in the EU. #

88 (657) On a very limited sample, it is difficult to draw any conclusions as to the number of patent settlement agreements being concluded in either the US or the EU. The reason for the greater number of settlement agreements in the EU over the time period may be explained by the fact that with a large number of Member States involved, and a greater number of local generic companies involved in the EU than in the US, it is unsurprising that the overall number of settlement agreements is higher. (661) Although some of the factors taken into account in determining whether and on what terms to settle will be common as between the US and EU, the regulatory and patent regimes are so significantly different that it is not legitimate to conclude that there are more similarities than differences. (683) The fact that a product is patented and has no generic copies does not mean that it dominates the market. That product in almost all cases will be competing with several other products in the same therapeutic class. Companies might launch their own generic version because of the consequence for international reference pricing of reducing the price of the branded product. By launching a generic version, the company would be able to compete on price, in the country concerned, with generics. 2.5 Other Practices Affecting Generic Entry Innovator companies can and do communicate with regulatory authorities concerning abridged applications for marketing authorisations. Contrary to what the PR asserts in the Summary to Chapter 2.5 on page 294, such communications do not occur in a significant number of cases, but in a relatively small number of cases in which there can be genuine questions concerning public health, substitutability and legal rights granted by Community law. There is nothing inherently suspect or wrong if innovators bring such concerns to the attention of regulators. On the contrary: such communications (labelled in the PR as interventions ) can prevent problems and situations that should not have occurred in the first place. If such concerns are shared by regulators, and lead to scientific and health questions to applicants, the time spent on answering those questions prevents problems, and cannot be labelled as a delay that is relevant for the purposes of any antitrust analysis or even attributed to innovator companies. The PR does not fully describe the regulatory framework. However, since most if not all of the innovator company interventions can be understood better if one takes into account how the procedures work, EFPIA hopes that the Final Report will assess the findings on company conduct taking full account of the regulatory framework. As regards court cases brought by innovator companies regarding generic marketing authorisations or reimbursement decisions affecting them (the Preliminary Report describes a limited number of such cases), EFPIA notes that access to the courts is a fundamental right, which also exists at Community level: decisions on marketing authorisations of the European Commission can be challenged before the Court of First Instance of the EC. With regard to patent status, the PR does not fully reflect that some national laws provide that patent status is taken into account. If innovator companies are involved in communications on patent status in these Member States, or merely point out legal rights under national law to national regulators, that cannot be subject to ex post antitrust scrutiny. Such communications should therefore not be described in a pejorative tone. #

89 With regard to promotion, it is certainly true that the medical community needs to be made aware of the relative clinical benefits of the latest products. Communicating fully and accurately the risks and benefits of the latest and often complex medicines to doctors, and obtaining their feedback to inform future improvements, is a resource intensive process. This is part of ensuring that patients get the right medicines. The PR does not mention that Community law, national laws and self-regulatory advertising codes strictly circumscribe all communications from innovator companies to doctors. The PR alleges that innovator companies influence wholesalers. The fact that they are put on notice about patent infringement is normal practice in patent enforcement. Sending notice letters on patent infringement is actively encouraged by courts with a view to preventing needless litigation. In addition, in some Member States, such notice letters are a procedural precondition for enforcing patents. (707) As the PR recognizes (footnote 315), many regulatory authorities have no systematic record of contacts by innovators. If questions are asked by the authorities, it indicates that they see merit in them. Regulators are perfectly capable of judging the merit of issues brought to their attention and will take action where they consider it necessary. The generic authorization process does not provide for any type of innovator consultation. Therefore, where generic entry is delayed, this is likely to be the consequence of national regulatory authorities asking scientific and regulatory questions of the generic applicant and the marketing authorization procedure which allows for several rounds of questions and clock stops for responses. This process serves a clear public health purpose and in addition it avoids the grant of generic marketing authorizations where the requirements of the legislation have not been met, or public health issues may not have been addressed. The public health aspects and communications from innovators to regulators relating to data exclusivity must be distinguished from cases where authorities plan to grant authorizations that directly infringe the innovator s data exclusivity, which is a right specifically granted to innovators by Community law. Innovators have every right to defend their data exclusivity rights, and there is no doubt that authorities will proceed to grant marketing authorization applications if they do not share the innovator s view. Furthermore, it is important to note that in the period covered by the Report, the current regime did not apply. In other words, reference to the innovator data could, in any event, not be made until the expiration of the six or ten year data exclusivity period. (710) Interventions are not common. The numbers quoted ( 118 interventions ) inflate the issue because they count each representation about each dose and each version of the generic product in each Member State as a separate representation, which is a skewed approach. The figures should also be viewed in the context of the very unsettled regulatory legislation at the time covered by the report (see para 720 below and the fact that many of the newer Member States sought derogations from the requirements of the legislation at the time of accession). (711) Since generic companies target the more successful products rather than less successful products, it is not surprising that the level of interventions would be higher for these. It is also quite possible that the 40% of interventions, concerning 4 INNs, referred to by #

90 DG Competition related to safety issues in which case interventions would have to be made to each and every Member State for the purposes of safeguarding patients. DG Competition does not reveal this level of detail to be able to comment satisfactorily. (720) In para. 720, DG Competition briefly describes the rules on data exclusivity but does not fully reflect the key rules and their implications. Under Directive 2001/83 and Regulation 726/2004, the expiry of data exclusivity is one of the conditions which must be satisfied before the abridged procedure can be used. In the period covered by the Report, the 6/10 year data exclusivity regime applied and not the new periods introduced under amendments introduced by Directive 2004/27 (8+2+1 years). Second, the PR does not reflect how much legal uncertainty there was (and is) concerning data exclusivity. Prior to the amendments introduced by Directive 2004/27, the law was silent as to what innovator product data would obtain their own period of data exclusivity and hence the need for clarification in a string of ECJ referrals (see case C-368/96 Generics, case C-106/01, Novartis, case C-36/03, APS). It was also unclear as to what generic products could use the abridged procedure hence the reference in the Synthon case (C-74/03, SmithKline Beecham). Since the amendments came into effect, the law has remained unsettled as regards whether innovator data submitted in support of a number of products, for example those not expressly stated in the definition of the socalled Global Marketing Authorization or second generation biologics, are entitled to its own period of exclusivity. Given this lack of clarity, it is not surprising there are disputes regarding data exclusivity. In such circumstances, the innovator was entitled to make representations to the regulatory authority as to why the data exclusivity period had not expired. Contrary to what the report suggests, the statistics (26 data exclusivity disputes over eight years in all Member States) suggest that despite the law being unsettled, disputes were kept to a minimum. (724) In para. 724 et seq., DG Competition comments on litigation concerning generic marketing authorizations. There are many reasons why a generic marketing authorization might be non-compliant with legal requirements, e.g. because of physical differences between that product and the reference product. These differences may have an impact upon the quality, safety and efficacy of a product. Therefore litigation challenging the grant of a marketing authorization where there are qualitative shortcomings or other safety issues with the generic medicine is justified. We do not read the report as suggesting that third parties should not be given an opportunity to challenge the decision to grant a marketing authorization and Figure 118, which shows an example of a court supporting an innovator, reflects the importance of such challenges being made. In fact, the same right exists in Community law: marketing authorization decisions of the European Commission in centralized procedures may be challenged before the Court of First Instance of the European Communities. (734) A success rate of 19% regarding the outcome of data exclusivity litigation, as shown in Figure 119 is significant and not limited as the PR states. In the context of assessing why innovators take action, the PR does not reflect the #

91 economic impact and irrecoverable damage to the innovator arising from the launch of a generic in circumstances where there are legitimate grounds for opposing the grant of the generic authorization. This has to be contrasted with the financial loss to the generic which can be readily compensated. (738) Table 28 shows that there are only a few cases each year across all Member States. It is difficult to reconcile the numbers with the statement that these effectively delay generic entry and impact health systems and patients. (741) Figure 121 is based on generic companies own views as to when it ought to have obtained its marketing authorisation and is therefore not made on an objective basis. Innovator companies almost always find that expected launch dates of new products are delayed due to questions raised by regulatory authorities, which stop the clock. There is a significant degree of uncertainty in timing of the Decentralised and Mutual Recognition Procedures, in addition to the capacity restraints and waiting times for slots available at national level. The PR does not reflect that the key factor determining time to market is the gap between expiry of data exclusivity and the expiry of the last patent effectively protecting the reference product, whether that gap is sufficient to complete all the regulatory procedures and how quick and efficiently generic companies act. Without an analysis of these factors, no meaningful conclusions can be drawn. (750) DG Competition briefly describes the 90- and 180-day deadlines set out in the Transparency Directive. The PR, however, does not reflect that, irrespective of the provisions of the Transparency Directive, in practice, decisions on pricing can take over a year in some countries. The EGA President stated at the public presentation of the Report that generics faced regulatory delays and gave the example of delays of 13 months in Belgium for generics. (755) Damages caused by generic sales while the reference product is still patent-protected are irreversible. Prices, once eroded by premature generic entry, will not go up again. (758) The claim that under EU law, patent protection is not a criterion to be considered by the authorities when approving prices or granting reimbursement status is open to question. As confirmed in the jurisprudence of the ECJ, the scope of patent protection is a matter for national law, as is the functioning of national social security and health systems within the boundaries set by the general principles of non-discrimination, freedom of movement, and the EC Transparency Directive (Directive 89/105/EEC). A Member State may therefore legitimately take the patent status of the innovator product into account as an objective and verifiable criterion in its pricing and reimbursement decisions, that is perfectly legitimate. (772) It is illegal to promote medicines to consumers except for OTC products. The promotion and information given to healthcare providers with regard to the merits of innovators own products is entirely legitimate. (773) This paragraph does not take due account of the fact that the provision of information to health professionals is strictly regulated by Directive 2001/83, the EFPIA Code of #

92 Practice and national advertising codes which implement the EFPIA Code. (777, 779) The PR addresses the awareness of some innovative companies of the Joint Declaration of the CPME and EFPIA concerning advertising rules. This Declaration is indeed not universally known. It is an agreement to further promote and consolidate high ethical standards but is secondary to the EFPIA Code of Practice which all companies are obliged to follow. Each year EFPIA publishes a report on the enforcement of and compliance with the EFPIA Code of Conduct which has existed since 1992 and has been revised and updated regularly, most recently in The PR also fails to take account of the advertising codes and self-regulatory bodies which operate at national level, which implement the EFPIA Code and which are strictly enforced. (784) Patent laws contain provisions to protect third parties against unlawful threats of patent infringement and innovator companies will ensure that their warning letters are compliant. Moreover, in some Member States, sending pre-litigation notice letters is a procedural pre-condition for bringing a patent enforcement case. (793) The report misrepresents the FarmaIndustria advertisement. Spanish law states that where generics and branded products are equally priced, the generic must be dispensed. The association rightly thought this discriminatory. If both medicines are priced competitively, why should the State enforce prescription of the generic? The advertisement was intended to capture that. Promoting the brand is not unlawful. The generic version was not denigrated. (799) The first sales will normally follow a number of months following the grant of a marketing authorization. Therefore it is not surprising that the grant of a generic marketing authorization is taken as an indication that a product launch is approaching, which is a factor to be taken into account in establishing the threat of patent infringement not only by the generic manufacturer but also others involved in the supply chain. The sending of warning letters in regulated and may even be required (see (784)). (810) The PR quotes the Office of Fair Trading (OFT) report on Direct to Pharmacy (DTP) arrangements. This report is inconclusive, and merely states that the OFT will monitor developments in the market to assess the consequences of DTP arrangements in future. The PR implies that by concentrating supplies, DTP can impact other wholesalers in a negative manner. However, that argument could equally be applied to short-line wholesalers and parallel traders. (814) If there are maximum prices for prices at patient level (as there are in most Member States), a DTP arrangement cannot result in higher prices for consumers. 2.6 Life Cycle Strategies for Follow-on Products In Chapter 2.6, the PR discusses second generation or follow on products launched by innovator companies (labelled as evergreening ). In the Summary on page 311, the PR states that the launch of such products occurred on average one year and five months before loss of exclusivity. This implies that innovator companies generally time improved products to hit the market a year and a half before patent expiry. However, such #

93 conclusions cannot be drawn from the data. The sample is very small (16 INNs); the variance is very large (from seven years before loss of exclusivity to one year after LoE, as defined by the Commission); and no account is taken of all the factors that determine the time to market of improved products, notably the marketing authorisation procedure which may require extensive clinical trials. Without taking account of all of these factors (and notably the gap between expiry of data exclusivity and patent expiry) it is difficult if not impossible to draw any conclusions on timing. Generally, Chapter 2.6 suggests that there is something implicitly wrong or inappropriate about innovator companies creating and marketing improved versions of their products. The PR seems to suggest that these products main or only goal is to prevent generic competition. However, a proper assessment of the drivers of improved products and incremental innovation would show that these are the result of competition on the merits: Incremental innovation is driven by patient needs. Innovator products are launched in a specific pharmaceutical form, which at the time of authorization was considered to be the most appropriate form. Feedback on clinical practice from doctors, pharmacists, patient associations and opinion leaders on how patients experience their product, adverse drug reactions, side effects and problems with patient s compliance with the dose regimen set out in the Patient Information Leaflet, can highlight the need for specific improvements. Competition from other innovator companies also drives incremental innovation. The main business of innovator companies is to compete with other innovative companies in the relevant therapeutic class. As noted in the PR, a first in class product is often followed quickly by other innovative products targeted at the same disease. This is why there are over 15 NSAIDs, 15 beta blockers, 9 protease inhibitors and over 10 statins 29, oral products instead of injectable products; subcutaneous injection forms instead of intravenous injection forms; once-a-day dosing instead of twice-a-day dosing, combinations with other active ingredients or medical devices, etc. The improved product is only as good as the therapeutic improvement that it brings. This fact of life is perfectly summarized by the quote included by DG Competition at para.853 of the PR. If the new product is indeed better, customers will want to use it. The situation is consistent with normal competition on the merits. If competitor A introduces a new product, the other competitors have to choose either to introduce new versions of their own, or to cut prices for the older models. EFPIA recalls that generics are very active in making improved versions of existing products, and that for some successful INNs generic and innovator competitors submit more patent applications concerning improvements than the innovator itself. 30 Chapter 2.6 is inconsistent with Chapter C.3, which recognizes the value of incremental innovation and the fierceness of competition between innovators. Only doctors take prescription decisions. Doctors receive information from several sources that report on new products and their relative advantages. In every Member State there are journals of pharmacists and doctors associations which are independent from pharmaceutical companies and which will ensure that the relative advantages are properly reflected. Then there are the national pricing and reimbursement authorities which in most countries evaluate whether an improved form or version of an innovator product actually brings something new in terms of therapeutic value. If it does not, it will be priced or reimbursed at the same level as the old product in which case the generic company can compete fully with it. In many Member States, doctors are encouraged or even obliged by the payors to take price differences into account See EFPIA submission to DG Competition of 13 June 2008, page 80, para.153. EFPIA Submission to DG Competition, 13 June 2008, p.86. #

94 Advertising and promotion are strictly regulated. In describing marketing and promotional efforts (para.874 et seq.), DG Competition ignores the fact that communications from innovative companies to doctors are strictly governed by the rules on advertising set out in the EFPIA Code and corresponding national codes on advertising. All these rules prevent an innovator company from exaggerating the qualities or improvements of its improved product. Given these competitive drivers, there is no ground to assume or allege that so-called evergreening is anything but competition on the merits. (827) There is no basis for the assumption that launches of next generation products often occur shortly before loss of exclusivity. As set out in the comments to para.868 below, DG Competition s small data set (16 products including six non-responders) do not allow it to draw such conclusions. (830) EFPIA questions whether consumer associations (as opposed to patient associations) are a good proxy for assessing the relative advantages of improved products. (832) DG Competition suggests that the introduction of follow-on products is largely or wholly motivated by blocking generic entry. However, for the reasons set out in the Executive Summary to Chapter 2.6, there are many drivers of incremental innovation, including fierce competition with other innovative companies that also improve products in the same therapeutic class. (838) DG Competition notes a discrepancy between innovators and generics view on what a second generation product is, and notes that such products were not reported by innovator companies. This is likely to reflect that the notion is open to interpretation. DG Competition itself does not seem to be clear as to the boundaries of follow-on products, noting in para. 860 that some fall within the definition of a line extension and other are sufficiently different to obtain a different INN name and/or data exclusivity. (842) The fact that second generation products relate primarily to first generation products, which constitute a big part of the company's turnover, is self-evident. EFPIA notes that the three graphs depicting sales of innovator products and improved products (specifically chosen by the Commission from a set of 88 INNs) show divergent patterns. The third graph at Month 61 shows a 60% drop in sales of the improved product three years after market entry (with sales of the first generation product having been near zero for over two years). (855) EFPIA regrets that the PR quotes consumer associations but not patient associations. (859) Because no extension to the data exclusivity period is granted for most line extensions, generic companies have the advantage of being able to refer to the extensive clinical trials and other development activities which will have been carried out by the innovator company with respect to the improvements. Since generic companies often have little difficulty in getting around formulation, process and other patents, the barriers for generic companies launching their own versions of newer products under abridged procedures should not be overstated. #

95 (864) DG Competition quotes the inertia of doctors to argue that switching is particularly difficult after generics have been introduced. This confirms that doctors need to be firmly convinced of the merits of an improved product and its benefit for their patients. If not, they will simply not prescribe it. (868) Figure 128 does not support DG Competition s conclusions on timing of the introduction of improved products. Figure 128 shows 22 INNS. For six of these, there are no data on improved versions. For the other 16 products, the date of introduction of improved product ranges from seven years before loss of exclusivity to almost a year after loss of exclusivity. Drawing statistically meaningful conclusions from such a small sample with such large variations is highly speculative. 2.7 Cumulative Use of Practices Against Generic Companies If particular steps or actions are legitimate when used alone, they do not become illegitimate simply because they may be taken at or around the same time as other legitimate practices. On the PR s own data, there is no correlation between the use of the tools and delay. DG Competition assumes the various life cycle tools are one, and indeed the only cause of delay. There is no analysis of the impact of other regulatory factors. The analysis of welfare effect is based on a hypothetical scenario of a case where an interim injunction is granted and the innovator loses at trial. There is no evidence as to the number of times this occurs or indeed when the innovator succeeds at trial. (887) Each of the measures in the so-called tool box is per se legitimate. Nowhere is it suggested otherwise. (890) Dominant companies are allowed to meet competition and may reduce prices to meet competition. The fact that the so called tools may be deployed in the period around expiry of a primary patent, underlines the fact that the innovator companies are responding to competitive pressures. There is no evidence to support the allegation that secondary patent clusters may be efficient means to deter or prevent generic entry. (891) If there is sufficient innovation to satisfy the tests of patentability, a patent covering that innovation may be granted. (892) The fact that disputes and contacts are initiated almost exclusively when a generic competitor launches or is about to launch a product is not surprising. The innovator will not know of the generic plans until this time and cannot bring proceedings in most, if not all, Member States beforehand. (893) It is a misrepresentation to say that generic companies won the vast majority of legal actions in the EU. They won 71% of those they initiated but lost 49% of those primarily infringement actions initiated by innovators. (901) The thresholds for when any particular tool is said to be employed appear to be random. For example, it is not understood why secondary patenting is said to be #

96 employed (in this section of the report) only when the number of secondary patents per INN exceeded the median number of secondary patents per INN within the E75 group, whereas disputes/contacts are said to be employed when there is at least one contact or dispute related to a specific INN. (903) Having patents is a precondition to patent litigation which is a precondition to settlement, so these three tools are connected. Counting them individually inflates the numbers of tools being used. (912) This paragraph of the report notes that the exercise of innovators statutory and other rights, be it on a stand-alone basis or by combining several instruments, could be legitimate. It would be more accurate to say that they will be legitimate in the absence of exceptional circumstances. (913) Whilst there are price reductions following launch of generics these should be compared to the level of price reductions in the USA, which are much greater. DG Competition appears not to have considered the CRA paper 31 on the reasons why greater price reductions do not occur. (918) In sub-paragraph (c), DG Competition is referring to jumbo referencing, which is inherently unfair to innovator products which are still patented. (920) The whole of the welfare analysis is based on a scenario where there has been an interim injunction. Interim injunctions are only granted in approximately one-sixth of all litigations reported, and one-third of those brought by innovators. There are no data on how many cases involved an interim injunction being granted and the innovator losing at trial. It is only in these cases that it can be shown that a generic was prevented by litigation from reaching the market. It is only in these cases that delay of generic entry caused by litigation is possible. The generic will generally be entitled to compensation in such cases related to the actual period of delay (if any). However, any such delay is the result of a considered judicial decision applying applicable national law and procedure for such injunctions and in the light of evidence from both parties. There is no justification for ex post antitrust scrutiny of such cases. These paragraphs of the report discuss a simplified diagram showing the welfare effects of delayed market entry. However, as described at (a) in the second paragraph of footnote 368 to paragraph 920, what appears to be an erroneous assumption has been made on which the entirety of this data is based. The erroneous assumption is that the expected price developments if generic entry could have occurred earlier take the shape of the actual price developments (in other words, the two lines defining the edge of the grey area follow each others path). However, there appears to be no factual basis for this assumption. In fact, it is most likely (and has been the case in practice) that at the outset of patent litigation, there are fewer generic companies (often only one) that have the means to produce a generic version of the product in question. Further, it is only towards the end of the period that the number of generic companies who have a product available, reach the number required to produce the extremely steep price decline which this particular illustration employs. Thus, a more realistic picture would be that the expected price development reduces slightly (to give the first generic company a slight price advantage) and then stabilises 31 Competition in the off-patent market post generic entry, CRA International, September #

97 until the next generic entrant had a product (when a second small step-down occurs, etc.). The overall picture is likely to produce a considerably different shape from the grey area depicted in Figure 133, the expected price development following the horizontal much more closely than is depicted. Footnote (b) states that innovators cannot expect to be able systematically to extend the protection period over the period prescribed in the law and that they should be allowed super-normal profits only during the standard protection period. This risks having a severe chilling effect on incremental innovation. It is a commercial reality that no company would be prepared to carry out the clinical trials to provide a once-a-day version of a product in place of a 3-times-a-day product or to develop an oral version of an injectable product etc. on the grounds that the original molecule patent would be the only one that could enjoy a period of monopoly. #

98 3. COMPETITION BETWEEN INNOVATOR COMPANIES THE ISSUES There is healthy competition between innovators as noted at para. 928, although the PR does not refer to the substantial evidence of this submitted by EFPIA. 3.1 Patent Strategies Often the only way to guarantee protection for investment in innovation is to exclude others exploiting a different but very similar invention. Thus it is economically rational for a company which invents products A and B which would compete in the market and which only wishes to develop one of these to patent both. In that sense it is impossible to dissociate the negative, excluding character of patents and the protection of the underlying inventions and investment. In any event the PR fails to establish that innovators undertake blocking strategies which have the effect of hindering or preventing the development of competing new medicines. To the contrary the data seem to indicate that with one exception every innovator company enjoyed sufficient freedom to operate to bring new products to the market despite patents and patent strategies. The PR suggests that publication per se is bad for innovation because it may create prior art thus preventing product development by others because they will not be able to obtain a patent for that development. Such a suggestion is contrary to the generally accepted proposition that publication of knowledge stimulates innovation. (935) It is noted that the PR accepts that patents have an information function beneficial to society. (949) Freedom to operate is not the only or not the most determining factor for an innovator to carry on R&D and investment in a programme. Commercial considerations are crucially important. (953) The Commission refers to filing patent applications with broad claims or to filing a number of patent applications relating to similar subject matter as a form of smoke screen which arguably would be a defensive mechanism for hiding the true candidate product under research. There is nothing wrong with such a practice. More fundamentally it reflects the fact that early broad patents or numerous patent applications related to the same subject matter are filed in order to acquire priority while further data will be gathered during the R&D phase in order to strengthen the applications. Doing otherwise might jeopardise priority rights in a system based on first to file and where several innovators might be focussing at the same time in the same therapeutic area because of unmet patients needs. (954) This paragraph indicates that divisionals and so-called secondary patents are strengthened on the basis of experimental data gathered during the R&D process. This tends to undermine other comments about the alleged weak nature of those divisionals and secondary patents. (958) The PR refers to so-called defensive patenting. While the PR accepts that a patent inherently and justifiably restricts the freedom to operate of competitors because it protects the investment in innovation, the PR suggests it is problematic when the patent is not used to protect a product that may proceed to development but to block competition exploiting the same or a similar invention. The PR fails to recognise that the exclusion of another product which would infringe a patent may be necessary to #

99 protect the investment in the original invention. (961) Whatever the alleged purpose suggested in this quotation, innovator companies are not prevented from bringing competing products onto the market. For example there are now 15 beta blockers, 9 protease inhibitors, 15 NSAIDs and 10 statins on the market in Europe. Two competing compounds to Viagra were launched within two years of the launch of Viagra. In its submission of June 2008, EFPIA demonstrated that the time to market of a second in class product had reduced to less than two years over time. (962) The fact that only 9 out of 43 companies identified some of their patents calls into question the meaning of the expression defensive patents. (967) The paragraph implicitly acknowledges the importance of having strong patent protection as a precondition to development of a new product. (971) The PR suggests that publication per se is bad for innovation because it may create prior art thus preventing product development by others because they will not be able to obtain a patent for that development. Such a suggestion is contrary to the generally accepted proposition that publication of knowledge stimulates innovation. The final report should clarify whether it is the policy perspective of DG Competition that companies should refrain from making technical knowledge public. 3.2 Patent-Related Exchanges and Litigation The data indicate that there were few allegedly blocking situations over the time period and that in such cases, innovators resorted to exchanges in order to solve the issue. Only one case is reported where a refusal to license was considered as impeding the development of a new product. This is not consistent with the proposition that patents are impeding innovator innovation. Litigation is far from common and when it occurs it tends to lead to a settlement which often includes a license. (984) This alleges that there are certain categories of patents where there is an increased risk of conflict. However, Figure 140 shows only 15 cases of process patent infringement over eight years. (989) 200 contacts is small given the 219 INNs, the eight year time period and the number of innovators. Very few of these contacts led to disputes. (996) The data show that in a large majority of cases where a license was asked for it was granted (78%). In 4 % of the cases negotiations were still ongoing. This is not consistent with a blocking effect. (997) Of the 18% of cases when a license was refused it appears that only in two cases a licence was refused without any reasons being given. (998) In only one case did a company say it closed a project on the basis of a refusal to grant a licence. #

100 (1005) In the light of the conclusion that there is often overlap in the areas of R&D between innovators, the fact that only 16 companies out of 43 (38%) have been involved in at least one litigation with another innovator over a period of eight years is not consistent with a general blocking effect. (1028) Interim injunctions between innovators are less likely because the harm done to one innovator by another s infringement is less likely to be irreparable than infringement by a generic. It does not have the price effect infringement by a generic would have. 3.3 Opposition and Appeals Oppositions between innovators are not common. (1036) 58 oppositions is not significant given the number of patents (around 40,000), the 219 INNs, the eight year period and the fact that at a conservative estimate there were 1100 overlaps. (1043) As noted previously, amendment is not to be regarded as success for an opponent as the amended claim may cover the opponent s product. 3.4 Settlements and Other Agreements The PR refers to 27 settlement agreements having been concluded over the eight years out of 66 litigations. In 67% of those cases settlements involved a license. These figures are not significant nor indicative of an industry or sector where there is blocking. (1048) 27 settlement agreements between innovator companies with respect to 40,000 patents and patent applications is an insignificant number. (1081) The PR states that the vast majority of agreements relate to the Italian and Spanish markets. This is hardly surprising as both markets are typical co-marketing markets because the number of doctors that exist in those countries makes it difficult for a single company to promote to each and every one of those doctors. Figure 152 also shows that behind Spain and Italy come Portugal, Germany & France, which also have a considerable number of doctors and co-marketing or co-promotion is therefore advisable. #

101 D. COMMENTS ON THE REGULATORY FRAMEWORK EFPIA believes there is substantial scope for agreement on the regulatory framework, but many issues remain unaddressed or misstated: The need for a single patent court that produces high-quality, cost effective and efficient judgments is generally recognised. A single Community patent would assist in preventing some of the uncertainties that arise in the national system, though is of less immediate importance in the pharmaceutical sector. Part of that package is the need for the patent system to increase the likelihood of resolution of uncertainties or disputes as to patent status in advance of product launch; Certain statements as to patent quality, divisionals and patent linkage are incorrect. But EFPIA does and continues to support maintenance of patent quality. All users of the patent system have an interest in patent quality to ensure that costly R&D is conducted in conditions of legal certainty as to the IP position. The same is true of providing the EPO with the resources necessary to conduct an efficient opposition procedure; P&R reform is also a matter where innovator and generics are concerned by the length of time for P&R status to be resolved. Efficient, speedy and transparent P&R rules (through the proper enforcement of the Transparency Directive in all Member States), would offer a fair reward for innovation and is a key reform which will stimulate innovation. 1. PATENTS (1111) Poor quality patents are not confirmed by stricter grant rates at the EPO. Stricter grant rates would, if anything, suggest rigorous scrutiny of pharmaceutical patents (Response, section 3.1(c)). No general statement as to patent quality can be made from the selected quotations cited (page 10, para. 1 and paras. 393, 514 above). (1116) No evidence that use of divisionals is intended to cause unwarranted delay to generics has been adduced in the report (paras above). The citation to B1.3 does not refer to divisionals (footnote 425 of the report). (1117) Third party observations allow companies to comment on or object to a patent grant. It is not, and cannot be, an alternative to an effective opposition procedure. ( and page 389, second paragraph) The report overstates the legal position in relation to patent linkage (para. 278 above). 2. MARKETING AUTHORISATION (1133) New medical uses are a significant source of innovation in the pharmaceutical sector and the subject of extremely costly R&D investment. For this reason they should not be labelled as a strategy in the apparently pejorative sense intended in this paragraph. (page 400, second The report shows no data supporting suggestions that innovator caused delay extends #

102 paragraph) pricing and reimbursement for generics beyond LOE. 3. PRICING & REIMBURSEMENT 1144 DG Competition notes that several Member States take significantly longer periods of time than those stipulated by the Transparency Directive. #

103 E. CONCLUSIONS AND LAUNCH OF PUBLIC CONSULTATION The primary conclusions of the report are not substantiated: The analysis of generic entry has not been able to establish that any part of observed delay is due to innovator conduct. There are many factors not analysed by the report which evidence shows are important considerations affecting generics decisions on time of entry. The conclusion on the purported decline in innovation leads with the damaging and counterintuitive suggestion that patents might be the cause of alleged innovative decline. Economic literature and theory teaches that patents are the essential pre-condition for innovation. In reaching the conclusion on generic entry, the report casts suspicion upon business activities, patents, litigation, settlements, regulatory interventions and second generation patented products, which are legitimate and essential for innovation to prosper and continue to deliver life changing improvements to society. Creating legal uncertainty in relation to these essential business activities is highly undesirable and retards the cause of innovation in Europe. The report identifies the consensus on a high quality single patent forum and Community patent. This conclusion is certainly valid, but misses opportunities to examine and consider the best way forward for far more significant policy issues affecting competition and innovation in the pharmaceutical sector, including: o o o o Improving the current patent system by means of inter alia, a streamlining of the European Patent Office opposition procedures and a mechanism to address patent disputes before generic launch; An efficient generics market that allows savings to be reinvested in innovative medicines; A holistic (versus silo-budget) approach to national healthcare budgets, with the focus shifted from the cost of new medicines to their overall value; Transparent, objective, fair, swift and predictable pricing and reimbursement processes that ensure that the right buyer signals stimulate greater innovative efforts through offering a fair reward for innovation, whilst ensuring early and equal patient access to medicines; o A territorial limit to the impact of State controls (implementing G10 Recommendation 6); o o o Re-examining whether the current SPC system has been effective in creating a 15 year period of protected commercial exploration sufficient to incentivize innovation; A reasonable risk-benefit balance in regulatory processes; and Removal of R&D bottlenecks and reduction of the costs of development through publicprivate partnerships exemplified by the IMI initiative. #

104 ANNEXES TO CHAPTER B Annex to Chapter B.1.2: Further Product Life Cycle Strategies during Patent Protection (39-41) The Report does not explain why the number of INNs on the E75 list relevant to each Member State decreases for France and Germany when adding companies to the IMS data set (Tables 37 and 38) (52) Price competition as a result of generic entry is consistent with established economic principles and central to healthy price competition. The assertion that economies of scale are used to drive out generics is not evidenced or substantiated. The relevant scale economies for generics or innovators are not examined in the report nor is the relationship between price and cost. (53) In those instances where it occurs, regardless of the timing before or after launch, the launch of an own generic denotes healthy competition in anticipation of or reaction to generic entry. (58) The statement that OTC medicines do not compete with generics is not substantiated. Marketing of the OTC medicines is an important part of competition. The extended data protection period provides an important incentive to undertake the additional market authorisation requirements necessary to switch to OTC. 30 January 2009 European Federation of Pharmaceutical Industries and Associations #

105 FINAL REPORT Prepared For: Baker & McKenzie CVBA/SCRL European & Competition Law Practice Avenue Louise 149 B-1050 Brussels Comments on DG Competition s Pharmaceutical Sector Inquiry: Preliminary Report Prepared By: Tim Wilsdon and Jim Attridge CRA International 99 Bishopsgate London, EC2M 3XD Date: January 2009 CRA Project No. D #