Pfizer ordered to pay damages to the National Health System for abusing a dominant position in managing the Xalatan patent lifecycle

1. Background

Pfizer owns and markets the product Xalatan, a medicine based on the active ingredient latanoprost. Xalatan is used to treat glaucoma, a common disease of the optic nerve. Xalatan is the most commonly prescribed medicine for the treatment of glaucoma in the world. This blockbuster drug was made eligible for NHS reimbursement, which means that in certain circumstances a portion of the price of the drug is covered by the NHS. When only the original producer’s product is on the market, the reimbursement price is usually higher, as the brand-name drug is typically more expensive than a generic and no alternative generic exists.^[1] Once a generic obtains marketing authorization (“MA”), its producer can apply for it to be included in the Transparency List of reimbursable drugs as an automatic substitute for the original at a lower reimbursement rate.^[2]

1.1 Pfizer patent strategy

Xalatan was originally covered by a European patent registration owned by Pfizer^[3] (“primary patent”). The patent application was filed in 1989, and the patent was granted in 1994 and scheduled to expire on September 6, 2009. In 1997, supplementary protection certificates^[4] (SPCs) to the primary patent were requested in almost every EU country of validation,^[5] thereby securing protection of Pfizer’s proprietary rights until July 17, 2011 in those countries. However, for unknown reasons (allegedly an oversight^[6] on the part of the patent owner) Italy was not the subject of an SPC request. Therefore, Pfizer’s exclusive rights to Xalatan were scheduled to expire in Italy on the primary patent expiration date—September 6, 2009—before they expired in the other countries where SPCs had been obtained.

Producers of generics had that patent expiration date on their calendars when they started filing MA applications in 2007. Obviously, they hoped to launch in Italy on September 7, 2009 (day 1). However, it later emerged that in January 2009 Pfizer had obtained a divisional European patent from the EPO^[7] (“divisional patent”) to further protect certain latanoprost formulation (which resulted in Xalatan being covered again), and that the divisional patent had been validated in Italy only. As a result, even though the primary patent and the divisional patent had the same expiration date, Pfizer, with a new patent in its portfolio, applied for and received (another) SPC for Italy. That entailed a period under protection in line with those of other countries, i.e., from September 6, 2009 to July 17, 2011. Pfizer also applied for a pediatric extension,^[8] which would have added an additional six months to the SPC expiration date. In July 2009 (after the divisional patent had been granted but before generic producers planned to place their generics on the market), Pfizer started to send warning letters to generic producers to discourage them from marketing their generic products, claiming that doing so would infringe the industrial property rights still covering Xalatan. It also started litigating against generic companies in civil courts (filing injunctions and seizures and counterclaims for infringement) and administrative courts (challenging inclusion in the Transparency List).

1.2 The AGCM ascertained that Pfizer patent strategy was an abuse of its dominant position and the Supreme Administrative Court upheld the decision

On January 10, 2012, by means of a complaint from Ratiopharm, which had received the warning letters and was in litigation with Pfizer, the AGCM learned that Pfizer had infringed Article 102 TFEU by deliberately pursuing a patent strategy with the sole aim of prolonging Pfizer’s intellectual property rights to Xalatan and creating a competitive disadvantage for generic producers. The AGCM deemed Pfizer’s patent strategy abusive pursuant to Article 102 of the Treaty on the Functioning of the European Union because:

• Pfizer filed the divisional patent 13 years after the primary patent^[9] and just as some competitors were entering the market.
• The divisional patent did not cover any new therapeutic use of Xalatan and no product was launched on the market covered by the divisional patent.
• The divisional patent was only validated in Italy and the SPC was only requested for Italy and not for other countries where it could have been requested as well.
• A pediatric extension for a drug that treats a disease like glaucoma that predominantly affects the elderly raised red flags.

The AGCM investigation was extensive and took into account multiple factors, including the warning letters sent to generic companies, the litigation underway (consisting not only of counterclaims for infringement in passive litigation, but also injunctions and seizure proceedings brought by Pfizer), and a wealth of internal correspondence showing that Pfizer’s parent company was aware of the Italian situation and was working strategically to achieve the same duration of protection in Italy.

According to the AGCM, despite the fact that the overall conduct was technically compliant with patent law and sector regulation, in substance it was designed to block entry and push generics out of the Italian market, since Pfizer had no other plausible economic interest or strategic rationale. According to the AGCM, this caused a delay^[10] in the plans of generic producers to launch their products on the market, because of the legal uncertainty surrounding the patent portfolio. The AGCM noted that although Ratiopharm had obtained MA in September 2009, it didn’t launch the products until May 2010—an unusual scenario in the pharmaceutical industry. Producers usually launch a generic swiftly after MA is granted. Therefore, it was considered likely that altering the scenario that would have played out absent the contested conduct was detrimental to competition and presumably to the NHS and taxpayers because it delayed entry of a less expensive alternative drug.

The AGCM estimated the damage to the NHS to be in the region of 14,000,0000 euros.^[11] The estimate was made not with a view to awarding damages (as the AGCM is not a forum for damages), but with a view to assessing the gravity of the infringement under competition law.

The AGCM decision was overruled by the Administrative Court of Lazio (“TAR Lazio”), and then upheld by the Supreme Administrative Court (“Consiglio di Stato”) as abuse of rights, i.e., exploitation of rights to obtain an unfair advantage that goes beyond the intended objective of those rights. Significantly, the Supreme Administrative Court refuted Pfizer’s arguments^[12] by claiming that the final validity of the patent and the fact that the overall strategy was allowed by patent law were irrelevant when it came to ascertaining abuse of dominant position. The Supreme Administrative Court stated that in order to abuse a right, one must have that right in the first place—which Pfizer did.

2. Claim for damages suffered by the NHS

2.1 The NHS claim for damages before the Court of Rome was dismissed

Once the Supreme Administrative Court confirmed the fine, the Ministry of Health and the Ministry of Economy and Finance filed an action before the Court of Rome to recover damages suffered by the NHS as a result of having paid a higher reimbursement price for Xalatan during the months when the generic producer delayed its inclusion in the Transparency List and its market entry. The NHS requested the amount of damages determined by the AGCM (i.e., approximately 14,000,0000 euros). However, the Court of Rome dismissed the claim, stating that the NHS failed to prove that the generic producer’s late entry into the market was caused by the abuse of dominant position (i.e., a causal link between the abuse and the delay was insufficiently substantiated). The logic was as follows:

• The AGCM decision is only evidence that a breach of competition law occurred, not that such breach caused a delay and thus damage (to the NHS). Indeed, according to the court, the value of the AGCM decision as “stronger evidence” of competition infringement—which was temporally applicable to that infringement—was limited to the elements of the decision relevant to the finding of abuse and does not impact quantification of anti-competitive effects or harm, as those are not necessary for that finding (in line with the scope of the binding effect of national authorities’ decisions set out in Article 9 of Directive 2014/104/EU).^[13]
• Ratiopharm was not ready to go to market before May 2010 anyway, as it obtained MA in April 2010.
• Ratiopharm could not manufacture the active substance before Pfizer’s industrial property rights expired without breaching Article 66.2 (a) Industrial Property Code, which states, “The patent specifically grants the owner the following exclusive rights: (a) if the subject of the patent is a product, the right to prohibit third parties, subject to the consent of the owner, from producing, using, putting on the market, selling, or importing the product in question for such purposes.” Therefore, it could not have launched on day 1.
• Pfizer’s conduct could not be deemed abusive as divisional patent validity had been confirmed by the Board of Appeals a few months after the AGCM decision.

Eventually, the court found that when it came to setting damages, the NHS did not prove that each package sold by Pfizer during the delay was reimbursed by the NHS.

2.2 Court of Appeals of Rome overruled first-instance decision and awarded damages

The claimants appealed the first-instance decision before the Court of Appeals of Rome, which overturned the lower court’s judgement and awarded damages in the amount of 13,360,464 euros. The Court of Appeals found that:

• Under Article 9 of Directive 2014/104/EU, an AGCM decision cannot per se establish a causal link between a breach of competition law and alleged damages; however, this principle does not prevent such a causal link from being proven by presumptions based on the facts ascertained by the AGCM and gathered during AGCM investigations.
• Indications of a causal link could be established based on Pfizer’s dominant position, the anti-competitive effects that substantiate Pfizer’s abuse of its dominant position, and other factual elements collected by the AGCM (e.g., the fact that producers of generics usually enter the market quickly after obtaining MA, but in this case they did not because Pfizer’s conduct created too much uncertainty about the consequences of entering the market).
• Ratiopharm’s first MA application was filed in September 2007, but it was withdrawn due to pressure from Pfizer.
• It is untrue that generic producers could not start manufacturing generics before Pfizer IP rights expired and therefore would have not been able to enter the market earlier absent the conduct, as Article 66 IP Code prohibits only sale of the patented product, not manufacturing.^[14]
• The fact that the divisional patent was found valid by the Board of Appeals had already been taken into account by the Supreme Administrative Court that confirmed the AGCM decision and was found to be irrelevant to ascertaining abuse.

Finally, the Court of Appeals held that the NHS did not necessarily have to prove that every pack of the medication sold by Pfizer during the delay period was reimbursed by the NHS. The court was satisfied with the data demonstrating the volume of sales during the delay period and the AGCM estimate how much the NHS would have saved on reimbursement (approximately 14,000,0000 euros). The court deducted 5% from that amount to reflect the fact that some of the sales likely included packs that were not reimbursed by the NHS.

2.3. Supreme Court final order upheld damages award

Pfizer challenged the Court of Appeals decision before the Supreme Court. Pfizer’s arguments were as follows:

• There was no abuse of dominant position: The breach of competition law should have been reconsidered, as the Board of Appeals confirmed the validity of the divisional patent and this was not duly taken into consideration by the courts. According to Pfizer, the Court of Appeals did not mention this, as it erroneously considered the AGCM decisions “stronger evidence” and binding. Moreover, Pfizer’s strategy did not violate competition law as it was simply enforcing its legitimate rights under patent law.
• There was no evidence of a causal link between the abuse and the alleged damages, as the Court of Appeals based its reasoning on the AGCM decision, which was not “stronger evidence” to establish a causal link; moreover, manufacturing and packaging prior to expiration of Pfizer’s patent rights was not allowed under Articles 66 and 68 IP Code, so Ratiopharm could not enter the market on day 1.
• The burden of proof was not properly allocated, because the court was satisfied with Xalatan sales volume as ascertained by the AGCM and did not require the NHS to prove that every package sold by Pfizer during the delay was reimbursed by the NHS.
• The amount of damage could not be determined equitably based on principles of fairness (referring to the 5% reduction applied by the Court of Appeals), as there was no evidence of the damages suffered.

The Supreme Court rejected all of these arguments, stating that:

• On the abuse of dominant position: The Court of Appeals never stated that the finding of abuse by a national competition authority could not be challenged using contingent elements, e.g., subsequent confirmation of validity of the patent concerned; indeed, the AGCM never considered invalidity of Pfizer’s patent a prerequisite for breach of competition law. Instead, it was Pfizer’s overall strategy that was considered abusive, regardless of patent validity—the right must exist in order to be abused. Citing CJEU case law on abuse of dominant position, the court ruled that Pfizer had no legitimate or plausible economic interest in the contested patent strategy other than to exclude competition (by means of meritless conduct) that would have otherwise existed. In support of this reasoning, the court recalled the key elements that the AGCM decision relied on to characterize the conduct as abusive (in particular the timing of the divisional patent application and the fact that there was no actual new product).
• On the causal link: The Court of Appeals did not base its finding that the causal link was proven solely on the fact that the AGCM decision was deemed “stronger evidence.” Indeed, the court assessed the reliability of various elements (i.e., the AGCM decision and the investigation carried out during the proceedings before the authority) to establish a rebuttable presumption. On the basis of that presumption, the Supreme Court found it “more likely than not” that Pfizer was responsible for the delay, as generic producers took into account the risk of patent infringement litigation when deciding to delay entry until the legal situation was clarified. The Supreme Court also found it irrelevant that Articles 66 and 68 of the IP Code would have prevented generic manufacturers from manufacturing and packaging the validly patented drug and hence launching it on day 1, since prolonged exclusivity was the result of an abuse of rights, and therefore Pfizer should not have had prolonged exclusive rights in the first place.
• On the burden of proof of damages: The Supreme Court found no error in allocating the burden of proof and stated that Pfizer was mistaken in the matter. Pfizer complained about how the Court of Appeals appraised the value of pieces of evidence, but its appraisal was carried out correctly by autonomously considering factual elements (e.g., sales volume) as a foundation for a presumption, which is entirely lawful.
• On quantification of damages: The Supreme Court ruled that Pfizer misunderstood the reasoning of the Court of Appeals. The court did not apply equitable quantification in lieu of evidence; it combined the evidence at hand to arrive at a margin of probability that some packages might not have been reimbursed.

3. Comments

Pfizer was the first case of abuse of patent rights in the pharmaceutical sector decided by a national competition authority after the notorious AstraZeneca case^[15] in which the CJEU ruled for the first time on the misuse of patent regulation as abusive conduct under Article 102. Despite the differences in the conduct concerned, in both cases conduct was deemed lawful and compliant with patent rules and sector regulations, though this did not prevent the finding of abuse under competition law. Since the Pfizer decision was upheld by the Italian Supreme Administrative Court, the AGCM has issued other fines for abuse of dominance based on the concept of “abuse of rights” or “misuse of regulation” in the pharmaceutical sector, hitting on the unfair and exploitative tactics of dominant companies in negotiating reimbursement rates with the NHS to extract excessively high prices.^[16]

With this judgement from the Italian Supreme Court, enforcement of competition law with regard to this type of conduct in the Italian pharmaceutical sector has grown stronger and become a greater deterrent. The Supreme Court has confirmed that the NHS is entitled to be compensated for the higher prices paid to the dominant company because of its abusive delay of generic entry; it also confirmed that it is lawful for a court to rely on factual elements provided in an AGCM decision about the damages suffered by the NHS to establish a (rebuttable) presumption of a causal link and the amount to be compensated. More specifically, the Supreme Court clarified that a court cannot simply refer to elements written in an AGCM decision as stronger evidence that is per se sufficient to award damages. However, it can autonomously (and at its own discretion) appraise the merit and adequacy of such elements to constitute a presumption and an equitable proxy to calculate the compensation to be awarded to the NHS when no further evidence is available from either party.

Many features of the Pfizer patent strategy are now outdated, as they have been addressed by legislation. For example, pursuant to Rule 36.1 of the European Patent Convention, divisional patent applications now must be filed on a strict timeline.^[17] Nevertheless, this complex and protracted litigation demonstrates that the patent strategies of pharmaceutical companies must always comply with competition law, and that competition and IP professionals ought to work together to manage the patent lifecycle and establish litigation strategy to protect market exclusivity.

[1] Primarily because originators invest in research and development and carry out trials, while the producers of generics do not.

[2] Originator products automatically may be replaced by generics once listed in the Transparency List under Art. 11.12, Decree Law No. 1 of January 24, 2012, converted by Law No. 27 of Marcy 24, 2012.

[3] The primary patent application was filed by Pharmacia AB in 1989, and the requests for SPCs to the primary patent were filed in 1997. In 2003, Pfizer acquired Pharmacia AB and its IP portfolio, including the primary patent and the corresponding SPCs.

[4] SPCs are industrial property rights that apply to patents in the pharmaceutical field. They are intended to “compensate” patent owners for the time spent on lengthy testing and clinical trials before authorization to be placed on the market by extending protection up to five years after the patent expiration date. SPCs are not patent term extensions, but separate (or sui generis) rights that come into effect when a patent expires.

[5] Such as countries where the European patent application owner requests the patent to be in force: this designation is not automatic for European patents, as the patent owner has to select the countries in which to pursue the patents.

[6] Pfizer claimed to the AGCM that Pharmacia AB “forgot” (“dimenticanza”) to request SPC for Italy.

[7] A divisional patent is a patent that stems from a parent or primary patent. The purpose of a divisional patent should be protecting one or more inventions contained in a parent patent that cannot be covered by the primary patent as a patent can only cover one invention at a time. Divisional patents have the same expiration dates as their parent patents. Generic producers often see filing multiple divisional patents as a method of creating legal uncertainty to dissuade them from entering the market.

[8] Namely, an additional six months of protection, in addition to the SPC, that patent owners can obtain by proving that the medicinal product is related to children’s diseases.

[9] While this was lawful at that time, it is no longer permitted.

[10] Namely, an approx. seven-month delay from September 11, 2009 (MA grant) to May 17, 2010 (listing in the Transparency List).

[11] Obtained by calculating the difference between the reimbursement price of latanoprost drugs before manufacturers placed generic drugs on the market and after the commercialization of equivalent drugs, multiplied by the number of packages sold between October 2009 and May 2010.

[12] At the time of the AGCM decision, the divisional patent had been revoked by the EPO Opposition Division and appeal was pending. In 2012, a few months after the AGCM decision, the Board of Appeals confirmed that the divisional patent was valid.

[13] Which states that “infringement of competition law found by a final decision of a national competition authority or by a review court is deemed to be established for the purposes of an action for damages.”

[14] Under industrial property law, this is not completely accurate, as manufacturing and packaging are activities capable of infringing exclusive patent rights under Article 66 of the IP Code.

[15] Court of Justice of the European Union, AstraZeneca v. European Commission, December 6, 2012, C-457/10.

[16] See Case A480, Aspen (see our previous report here), and Case A524, Leadiant (see our previous report here), both ultimately upheld by the Italian Supreme Administrative Court.

[17] Rule 36(1)(a) of the EPC provides that a divisional patent application may be filed with the EPO on the basis of a pending patent application (not granted) within 24 months of the first notice issued by the EPO Examination Division with respect to that patent application.