Keeping up with Dutch patent litigation: second half-year case law review 2020

The DutchKat taking some time
in her box to focus on just
one thing - the latest
half-year patent review (and 
her pink jingle ball)

Last year the IPKat started a series recapping patent litigation in some of the major patent litigation jurisdictions in Europe (see reports from the Netherlands, Germany and France). Why?  Because we are all facing information overload especially in these times, so Merpel meowed at the AmeriKat to cut through all the noise and just get to the point.  So we continue with the series in this the second roundup of Dutch patent litigation addressing fascinating issues such as cross-border jurisdiction, health insurers' reimbursement following from a later discharged PI and an additional question to the equivalence test.  Over to the IPKat's friends at Brinkhof in the Netherlands in the form of Barbara Mooij and Alexander de Leeuw:

"In this second roundup of Dutch patent litigation, we dive into some of the noteworthy patent cases of the Dutch courts from July – December 2020. Although the Dutch courts have by now fully adjusted to conducting hearings via video-conference, they remain extremely busy. The amount of case law in the second half of 2020 was similar to the first half of 2020.

Highlighted cases include a case concerning unjust enrichment by the patentee at the expense of a health insurance company as well as an update on the Dutch pemetrexed saga. We start with two decisions concerning auxiliary requests in national litigation.

1.  Auxiliary requests and added matter

a.  Biogen and Samsung Bioepis / Richter Gedeon Nyrt, District Court of The Hague 29 August 2020 (Judges Kokke, Knijff and Schüller) [Dutch decision here]

In a case between pharmaceutical companies Biogen c.s. and Richter, the District Court of The Hague considered how an auxiliary request filed by Richter with proceedings pending affects the invalidity discussion. In a somewhat remarkable (interlocutory) decision, the District Court decided that by relying on the auxiliary request and no longer defending the claims as granted the patentee effectively gave up its main validity defence, and could no longer rely on that defence for the validity of the auxiliary requests.

The case concerns a formulation patent for the antibody adalimumab, which is well-known from the world’s best-selling drug Humira. Biogen c.s. markets the adalimumab biosimilar Imraldi in a so-called buffer. A buffer allows the pH of a formulation to remain dynamically stable. Richter’s patent covers a pharmaceutical formulation of adalimumab with a very particular buffer, i.e. with histidine and citrate, and without phosphate. Biogen c.s. started a nullity action against Richter in the Netherlands, who in turn in separate (but later joined) proceedings claimed infringement of their patent. Richter conditionally submitted two auxiliary requests in its statement of answer for the case in which the District Court would consider the patent as granted anticipated by the prior art.

The hearing took place in two stages (due to the COVID-19 restrictions): written pleading notes and later a video-conference in which the court can ask questions and the parties can orally make a reply and rejoinder (this seems to be customary now in patent cases). In the written pleading notes, Richter expressed its intention to limit the debate to the auxiliary requests. Moreover, Richter stated that all arguments it put forward with regard to the patent as granted applied a fortiori to the auxiliary requests. In spite of this explicit statement, the District Court concluded that a complete fall back on the auxiliary requests meant that Richter’s defence against the nullity arguments for the patent as granted were apparently no longer upheld by Richter. For that reason alone, the court assumed in the rest of the proceedings that “it is no longer a matter of dispute that the claimed formulation belonged to the state of the art on the priority date” (par. 5.22).

This approach of the District Court makes the fall back on auxiliary requests in fact meaningless. According to the District Court, the limiting features of the auxiliary requests were so minimal that they could not provide novelty or inventiveness. Since Richter was not allowed to rely on their arguments presented for the patent as granted – mostly relevant for inventive step – the auxiliary requests almost automatically shared the fate of the original claims. The decision was also met with some criticism, see for example the critical annotation of R. Kleemans in Dutch here.

It remains to be seen whether this case will be appealed and if so, how the Court of Appeal deals with this issue. A possible solution for the time being could be for the patentee to limit the debate to the auxiliary requests in an early phase in the proceedings in combination with a conditional claim defending the main request.

b.  MSD / Wyeth, District Court of The Hague 11 November 2020 (Judges Kokke, Aalbers and Schüller) [Dutch decision here]

In a second decision on auxiliary requests in a case between pharmaceutical companies MSD and Wyeth, the District Court – in a panel with two of the same judges as in the Biogen/Richter case – decided along the same lines as the Biogen/Richter case with regard to the scope of the validity debate after (unconditional) auxiliary requests.

Wyeth is the holder of a patent on a siliconized container containing a formulation for a pneumococcus vaccine. In its statement of answer, Wyeth indicated that it would primarily rely on its second auxiliary request, and would not defend the claims as granted. Accordingly, the District Court decided that the granted claims are invalid. According to the District Court, “the falling back on the auxiliary requests can, in terms of legal consequences, be equated with the abandonment of EP 679 NL as granted. To hold otherwise would mean that falling back on the auxiliary requests – which are more limited in scope of protection than the patent – would be devoid of any significance” (par. 4.5).

As to the auxiliary request, the District Court considered that the claimed applications were focused on two separate inventive concepts of two distinct embodiments: one in a formulation with a surfactant for stabilisation of the protein conjugate, and the other in a formulation with an aluminium salt to inhibit silicone-induced aggregation. Explicit instructions for the intended combination were lacking (and also not argued by Wyeth) and would also not follow from common general knowledge. Since the combination could not be deduced directly and unambiguously, implicitly or explicitly, from the original and divisional application, the claims were invalid due to added matter.

2.  Unjust enrichment of pharmaceutical company towards health insurer

Menzis / AstraZeneca, District Court of The Hague 14 October 2020 (Judge Knijff, Brinkman and Aalbers) [Dutch decision here]

The second half of 2020 brought a remarkable decision in a case between health insurance company Menzis and pharmaceutical company AstraZeneca. The District Court ruled that AstraZeneca was unjustifiably enriched at the expense of insurance company Menzis due to enforcement actions (against Sandoz) based on a patent that was subsequently invalidated.

AstraZeneca’s patent covered a sustained release formulation of quetiapine, a drug used in the treatment of i.a. schizophrenia and bipolar disorder. The validity of the patent was challenged by Sandoz in 2011, but the District Court found the patent to be valid. However, in parallel proceedings before the UK High Court the patent was found invalid (later confirmed by the UK Court of Appeal). Sandoz appealed the District Court decision, but in the meantime listed its product in the G-standard (the Dutch drug database). A listing in the G-standard is often met with PI proceedings, which is exactly what AstraZeneca did in 2013, and a PI was granted against Sandoz. On the 20th of August 2013 AstraZeneca served the decision on Sandoz, thereby forcing Sandoz off the market. About a year after the PI was granted it was overturned on appeal and the Dutch part of the patent was nullified. This led to an interesting liability debate.

How do we get from patent enforcement against a generic drug to liability against a health insurance company? Under the Dutch health insurance system a health insurance company can designate a preferred drug for compensation, meaning that only the preferred product will be reimbursed. Health insurance companies generally designate a generic drug as the preferred drug in order to minimize costs. With the PI in place, Sandoz was obviously barred from entering the Dutch market. However, the existence of the PI also deterred other companies from entering the market with generic quetiapine. This in turn forced Menzis to reimburse the relatively high costs of AstraZeneca’s product for all of its insured patients. As the patent later turned out to be invalid, Menzis sought compensation for the price difference with respect to the reimbursed drugs.

The District Court first reiterated that under Dutch case law the patentee is liable for damages caused by enforcement of an injunction. This means that AstraZeneca is liable for any damages caused with Sandoz due to the enforcement. The Court saw no need to also assess whether AstraZeneca could be held liable against Menzis (under general tort), because AstraZeneca was in any event unjustifiably enriched due to its enforcement actions. AstraZeneca enjoyed an exclusive market position as a result of the enforcement of an invalided patent, which allowed it to benefit from relatively high drug prices. Menzis was forced to reimburse the use of AstraZeneca’s drug, as there was no other variant available. The District Court therefore concludes that AstraZeneca has been unjustifiably enriched at the expense of Menzis.

A legitimate question to ask is whether Menzis suffered damages or their policy holders, as the costs of the drugs are (indirectly) recouped via a higher premium for the insurance. The District Court considered that any damage claims of persons insured with Menzis accrue to Menzis by way of subrogation. Moreover, a higher premium does not detract from the fact that the insured persons were charged with the unjustifiably high price for the drug, and Menzis fully reimbursed them. Hence, by virtue of the insured persons and the full compensation provided by Menzis, Menzis accrued the insured persons’ full damages claim against AstraZeneca. The amount of damages is to be calculated in subsequent proceedings.

It does not seem unlikely that we will see more of these cases in the future. That is, of course, if the decision is upheld on appeal.

3.  Cross-border competence for non-Dutch defendants

Novartis / Mylan c.s., PI Judge of the District Court of The Hague 29 September 2020 (PI Judge Brinkman) [Dutch decision here]

The Dutch patent courts generally are not afraid to accept cross-border jurisdiction, which was confirmed once again in the preliminary relief proceedings between pharmaceutical companies Novartis and Mylan c.s.

This case concerned Novartis’ patent on the active ingredient deferasirox, which is used to reduce chronic iron overload. The patent expired in June 2017, the SPC will expire in August 2021, and the paediatric extension will expire in February 2022. Deferasirox was also registered as an orphan drug. Mylan applied for a centralised marketing authorisation for its generic deferasirox product, which was granted in September 2019. Subsequently, Novartis sent a warning letter to Mylan detailing the patent, SPC and paediatric extension protection on deferasirox. Mylan responded that it did not plan to launch its generic product in the Netherlands prior to August 2021 – which is the date the SPC would expire, and not the date the pediatric extension would expire –, and not the date the pediatric extension would expire – and explained that it did not believe the paediatric extension was valid.

In response, Novartis initiated preliminary relief proceedings against three Mylan entities: Mylan SAS (non-Dutch entity and the holder of the market authorisation), Mylan N.V. (the parent company of the Mylan group) and Mylan B.V. (the Dutch executive entity of Mylan). Novartis argued that it had an interest in obtaining a PI against the non-Dutch entity Mylan SAS, because that entity actively makes the central marketing authorisation available to the local entities of the Mylan group, with which it knowingly facilitates infringement of the (paediatric extended) SPC. The PI Judge accepted cross-border jurisdiction with respect to all three entities on the basis of Art. 4 and 8 Brussels I-bis Regulation.

The PI Judge granted a preliminary injunction based on (a serious threat of) patent infringement against Mylan B.V. as it was clear that it intended to enter the Dutch market after August 2021, upon expiry of the SPC. With regard to Mylan SAS and Mylan N.V., the PI Judge considered that merely holding a central marketing authorisation does not constitute an infringing act under patent law. However, the PI Judge accepted that making available the central marketing authorisation for the marketing of generic deferasirox tort in all countries in which the SPC is granted can constitute a tort. The PI Judge accepted, however, that making available the central marketing authorisation for the marketing of generic deferasirox in violation of a patent can be considered to facilitate patent infringement, and therefore constitute a tort.Moreover, the PI judge accepted that the potential exercise of control by Mylan N.V. leads to a serious threat of facilitating infringement. Accordingly, a cross-border injunction was granted.

4.  Update on pemetrexed in The Netherlands

Eli Lilly / Fresenius, Court of Appeal of The Hague 27 October 2020 (Judges Blok, Frieling and Van der Burg) [Dutch decision here]

Late 2020 the Court of Appeal delivered a decision in the Dutch pemetrexed case on the merits between Eli Lilly and Fresenius. Most IPKat readers are familiar with the worldwide pemetrexed saga in which Eli Lilly won in most instances. In the Netherlands, however, the District Court rendered a somewhat surprising decision finding against infringement by equivalence. The Court of Appeal now overturned that decision.

The Court of Appeal applied the two-step approach as also adopted in Germany, France and the UK. The first step includes determining if there is ‘literal infringement’, i.e. interpreting the claims in light of the description and figures in line with article 69(1) EPC and article 1 of the Protocol. The second step is to determine whether the feature(s) that are not fulfilled literally are equivalent, and whether it is appropriate to bring them within the scope of the patent at issue (while balancing the fair protection for the patentee with the reasonable legal certainty for third parties). For this second step the Court of Appeal formulated four questions, wherein the fourth question could be seen as a Dutch addition to the UK equivalence test:

1.  Is the deviating feature technically equivalent to the claimed feature? This would be the case if the product with the deviating feature solves the problem of the patent, and performs the same function as the claimed feature (‘technical equivalence’).

2.  Is it appropriate to take equivalents into account in the scope of protection from the perspective of the patentee – at the time of the infringement? This question is aimed at establishing if the protection offered by the patent is proportionate to the contribution made by the patent.

3.  Is equivalence under the circumstances of the case appropriate from the point of view of reasonable legal certainty for third parties? There would be reasonable legal certainty if the skilled person would understand that the teaching of the patent is broader than the wording of the claims, and when there are no good grounds for limiting the scope of protection to the features of the claims (including but not limited to a waiver).

4.  Is the equivalent variant new and inventive in light of the prior art? This question aims to ensure that the scope of protection of the patent is not broadened so much that also obvious variants (Gilette/Formstein) would fall under the claims.

The Court of Appeal answered the all questions in the affirmative for this case and concluded that Fresenius infringed the patent by equivalence, resulting in an injunction. This judgment gives some well awaited guidance on equivalence under Dutch patent law.

For a more comprehensive report on the decision of the District Court see: Eli Lilly v Fresenius Kabi: a decision far from equivalent to what has been decided before.

5.  Determining the closest prior art

a.  Shire-NPS / Accord, Court of Appeal of The Hague 18 August 2020 (Judges Kalden, Blok and Speyart van Woerden) [Dutch decision here]

In a case between pharmaceutical companies Shire-NPS (backed by its licensee Amgen) and Accord, the Court of Appeal overturned a decision of the District Court and upheld NPS’ patent and SPC. The Court of Appeal based its decision on a relatively strict selection of the closest prior art for assessing inventive step.

NPS’ patent relates to the compound cinacalcet, used for treating conditions associated with a malfunctioning regulation of the calcium content of blood and tissue fluids. NPS markets cinacalcet in Europe under the name Mimpara. The relevant prior art discloses Markush formulas which encompass many calcimimetic compounds including cinacalcet and individualizes other compounds such as NPS R-568, NPS 467, NPS 646/12Z, 18N and 13S. Accord argued that the selection of cinacalcet was arbitrary in light of the prior art, and therefore the patent lacks inventive step.

The Court of Appeal’s decision to uphold the patent and SPC is essentially based on the finding that NPS-568, and only NPS-568, should be taken as the closest prior art as it was disclosed in the prior art as having the best calcimimetic activity and the “lead calcimimetic compound”. Compared to NPS-568, the Court of Appeal was convinced that cinacalcet has a significantly improved calcimimetic activity. The Court of Appeal therefore held that cinacalcet’s activity is better than what the prior art promised for the broader group to which cinacalcet belongs and that this was not obvious for the skilled person.

According to the Court of Appeal, because NPS-568 was considered to be the lead compound at the time, and in the absence of any reported concrete values for the other compounds disclosed in the prior art, the skilled person would not consider the other individualized compounds to be realistic starting points. The Court of Appeal decided that “a starting point is only realistic if it would actually be considered by the skilled person” and “only if the skilled person would have several alternative starting points (i.e. if more compounds can be considered as closest prior art), the invention will have to be inventive starting from each of those alternative starting points”. The Court therefore only considered the ‘most’ realistic starting point. Nonetheless, in an obiter dictum the Court of Appeal explained that the patent would still be inventive over other compounds as starting point. In the absence of any reported concrete values for those other compounds and because NPS-568 was considered to be the lead compound at the time, the skilled person at the priority date would assume that the other compounds would at best have a similar calcimimetic activity as NPS-568. Cinacalcet’s improved activity would thus be unexpected either way.

An appeal to the Supreme Court is pending, so there is probably more to come from this case next year.

b.  Digital Revolution / Samsung Electronics, Court of Appeal of The Hague 21 April 2020 (Judges Kalden, Bonneur and Pinckaers) [Dutch decision here]

A similar issue concerning the selection of closest prior art arose in a case between e-commerce cartridge shop Digital Revolution and the well-known electronics company Samsung in a patent dispute about leaking and deforming cartridges. Samsung held two patents relating to cartridges for laser printers, and claimed that Digital Revolution infringed those patents. In response, Digital Revolution claimed the nullity of both patents. The District Court agreed with Samsung and granted an injunction against Digital Revolution. The Court of Appeal upheld the decision for one of the patents (patent valid and infringed), and overturned it for the other patent (patent invalid).

With regard to the former patent, for the inventive step assessment Digital Revolution relied on multiple documents about cartridges. All of these seem to belong to the same field as the invention. However, the Court of Appeal considered: “that it is impossible to see, and DR c.s. has not sufficiently substantiated, why one of the documents mentioned by DR c.s. could be regarded as the closest prior art. After all, the point is not that this document has the most structural features in common with the invention, but that the document is the most appropriate starting point for the skilled person in his search for a solution to his problem. No document relates to (the solution of the problem of) waste toner leakage or discloses (an indication of) the insights mentioned above in (a)-(c).” Again, the Court is looking for the ‘most’ appropriate starting point, and in doing so seems to further require that that the starting point must already relate to solving the technical problem that the patent claims to have overcome. The authors question whether a document concerning printer cartridges with – apparently – many structural features in common can be disregarded as closest prior art because it does not contain a disclosure concerning the problem solved by the patent in dispute. The authors understand that also in this case an appeal to the Supreme Court is pending."