Consequences of Asserting Rights Under a Patent Later Declared Invalid

Dutch Civil Code, Book 6, Art. 212; Dutch Patents Act, Art. 75(6) ‒ Menzis v AstraZeneca

Headnotes by the Editorial Office

1. The Court confirms the lower court’s finding that the holder of the patent later declared invalid did not know or ought to have realised that there was a serious, non-negligible chance that its patent would not stand up in opposition or invalidity proceedings.

2. This is due to the judgment confirming the validity by the court of first instance and several courts abroad that had reached a similar conclusion on validity at that time.

3. The Court confirms that the mere fact that there was no unlawful conduct does not automatically mean that the enrichment is justified.

Supreme Court (Hoge Raad), judgment of 3 November 2023 ‒ 22/01071

[…]

2 Assumptions and facts

2.1 For the purposes of the present appeal, the following can be assumed:

(i) Menzis is a health insurer within the meaning of Art. 1(b) of the Dutch Health Insurance Act, and is a non-profit organisation.

(ii) AstraZeneca AB and AstraZeneca B.V. belong to the AstraZeneca group, which pursues the development of pharmaceuticals.

(iii) Within the AstraZeneca group, a drug for treating, inter alia, schizophrenia and bipolar disorder was developed, with quetiapine as the active ingredient. This drug was brought onto the market in the form of tablets under the brand name Seroquel, initially only in an immediate-release (IR) formulation (Seroquel IR).

(iv) Until 23 March 2007, quetiapine was protected as an active ingredient by European patent EP 240 228, held by (a predecessor of) AstraZeneca Inc. On the basis of that patent, supplementary protection certificate 980022 (hereinafter: ‘the SPC’) was granted in the Netherlands. It expired on 23 March 2012.

(v) By European patent EP 0 907 364 (hereinafter: ‘the patent’), granted on 14 August 2002 on the basis of an application dated 27 May 1997, AstraZeneca AB obtained protection for the extended-release formulation of quetiapine. The patent is (was) valid in several European countries, including the Netherlands.

(vi) AstraZeneca B.V. obtained a Dutch marketing authorisation for an extended-release formulation of quetiapine in August 2007 and subsequently launched this medicine on the Dutch market under the brand name Seroquel XR.

(vii) From 2012 on, patients were in many cases prescribed Seroquel XR rather than Seroquel IR.

(viii) Almost immediately after the SPC expired on 23 March 2012, generic quetiapine tablets in an immediate-release formulation, such as the generic quetiapine IR from Accord Healthcare (hereinafter: ‘Accord’), came onto the Dutch market.

(ix) As of 1 May 2012, Menzis included in its preference policy Accord’s generic products for all strengths of quetiapine IR tablets.

(x) Although Accord and other competitors of AstraZeneca that were engaged in the sale and distribution of generic drugs had also obtained marketing authorisations for extended-release formulations of quetiapine, they did not enter the market with tablets in such formulation after the SPC expired on 23 March 2012.

(xi) A number of competitors, including Sandoz B.V. (hereinafter: ‘Sandoz’), initiated legal proceedings against AstraZeneca AB, in which they sought a declaration of invalidity in respect of the Dutch part of the patent. By ruling of 7 March 2012, the District Court of The Hague dismissed their claims.

(xii) By judgment of 22 March 2012,¹ the English High Court held that the English part of the patent was not inventive and was therefore invalid. Its judgment was upheld on appeal on 30 April 2013.

(xiii) AstraZeneca AB then sought interim relief to prohibit Sandoz from infringing the patent in the Netherlands. Sandoz argued in response that the district court ruling that had found the patent to be valid contained manifest errors and that AstraZeneca AB had brought the claims only to harm Sandoz. By ruling of 15 August 2013,² the interim relief judge rejected that defence and upheld AstraZeneca AB’s claims. AstraZeneca AB served the interlocutory ruling on Sandoz on 20 August 2013.

(xiv) By judgment of 10 June 2014,³ the Court of Appeal of The Hague set aside the ruling of 7 March 2012 referred to under (xi) above and declared the Dutch part of the patent invalid for lack of inventive step.

(xv) Competitors of AstraZeneca started bringing their extended-release quetiapine tablets onto the market in July 2014.

(xvi) Menzis included Accord’s generic extended-release quetiapine tablets in its preference policy on 1 January 2015.

(xvii) By letters dated 22 December 2016, Menzis held AstraZeneca liable for the damage suffered by Menzis as a consequence of AstraZeneca’s enforcement of the patent, or at least as a consequence of their keeping (potential) generic competitors out of the market.

2.2 In these proceedings, Menzis seeks a ruling to the effect that AstraZeneca acted unlawfully towards Menzis, or at least unjustly enriched itself at the expense of Menzis, and is obliged to make good Menzis’s loss. Menzis requests that AstraZeneca be ordered, jointly and severally, to make good the damage suffered by Menzis, estimated at [EUR] 4,168,935.67 or, at any rate, another amount to be determined.

2.3 The district court ruled that AstraZeneca had unjustly enriched itself at the expense of Menzis, via the latter’s policyholders, and was obliged to pay the damages subrogated by those policyholders to Menzis.⁴

2.4 The court of appeal set aside the district court’s ruling, deeming that there had been no unlawful conduct or unjust enrichment.⁵ To reach this decision, the court of appeal took the following into account (footnotes of cited judgment not reproduced):

No unlawful conduct

No strict liability

5.1 This court rejects Menzis’s argument that AstraZeneca acted unlawfully towards Menzis by invoking the patent from 24 March 2012 (date of expiry of the SPC) to 10 June 2014 (date of the judgment declaring the Dutch part of the patent invalid), while this patent was subsequently and with retroactive effect declared invalid for lack of inventive step. Menzis’s argument that the patent’s cancellation was at AstraZeneca’s risk is not supported by the law. The court will explain this below.

5.2 It is not disputed that, in the relationship between a patent holder and its competitors, no strict liability applies to invoking a patent that is later cancelled. The judgment in CFS Bakel v Stork shows that Dutch law takes the view, which prevails also in its neighbouring countries, that some form of culpability on the part of the patent holder is required for the assumption of liability after the cancellation of a patent. It does not accept that a patent holder that invoked a patent is liable for damage suffered by its competitors or others as a consequence of such conduct on the sole ground that the patent was subsequently cancelled.

5.3 Contrary to what Menzis argues, there is no reason to assume any strict liability in a relationship between a patent holder and non-competitors such as Menzis. Rather, the fact that Menzis does not itself manufacture or market pharmaceuticals and that AstraZeneca therefore never invoked the patent against Menzis, argues precisely against strict liability in AstraZeneca’s relationship with Menzis. Nor does the fact that Menzis did not have the opportunity to disregard the patent make Menzis’s position substantially different from that of the competitor in the case that led to the judgment in CFS Bakel v. Stork. In that case, the patent holder, CFS Bakel, had invoked its patent against (potential) customers of its competitor, Stork. The damage incurred by Stork as a result could not have been prevented by Stork’s simply disregarding CFS Bakel’s patent claim.

5.4 Nor does the fact that this case involves a patent on a medicinal product mean that strict liability must be applied. In itself, Menzis is correct to point to the public interest of good and affordable care. Strict liability can contribute to this in the sense that the prices of medicines will go down if patent holders no longer assert their patent rights, and damages are paid if the same patent holders do assert their patent rights and the patents are subsequently cancelled. As the Supreme Court held in CFS Bakel v Stork, strict liability can, however, also reduce the incentive to develop innovative medicines. That is also not in the interest of healthcare. The fact that, in this case, it has been established that the medicine claimed by AstraZeneca in its patent is not inventive does not alter this. Accepting strict liability would also affect holders of valid patents. Indeed, strict liability implies that a patent holder should never rely on its assessment that the patent will survive invalidity proceedings.

5.5 Nor does the fact that AstraZeneca has had 25 years of protection for the active ingredient quetiapine necessarily mean that strict liability must be applied as claimed by Menzis. Contrary to what Menzis argues, AstraZeneca’s assertion of its patent rights did not extend its market exclusivity. That is because the patent in question does not concern quetiapine as an active ingredient as such, but only the extended-release formulation thereof. During the period covered by Menzis’s claim, competitors were free to offer generic versions of Seroquel IR, and it is not disputed that competitors did actually do that.

5.6 Nor does strict liability follow from the Dutch Health Insurance Act and the interpretation thereof by the Supreme Court in VGZ v Nutricia and CZ v Momentum. The Dutch Health Insurance Act gives health insurers such as Menzis the task of ensuring good and efficient care and of controlling healthcare costs wherever possible. In the aforementioned judgments, the Supreme Court clarified that the violation of that statutory duty by health insurers could also be unlawful vis-à-vis healthcare providers. It does not follow that the holder of a patent that has been cancelled must bear the risk of having incorrectly assumed that the patent was valid.

5.7 Menzis’s reference to the judgment in Ciba Geigy v. [… 1] cannot result in any other verdict either. In that judgment, the Supreme Court held that, in principle, it had to be assumed that a person who, by threatening enforcement measures, forced another party to comply with an injunction issued in interlocutory proceedings acted unlawfully if it subsequently transpired from the judgment on the merits that the person was not entitled to require the other party to refrain from the conduct in question. Whether that judgment is in compliance with the rules of the EU’s Enforcement Directive as interpreted in the judgment by the Court of Justice of the European Union (CJEU) in Bayer v Richter is disputed by AstraZeneca. However, irrespective of whether or not it is in compliance, it does not follow from Ciba Geigy v [… 1] that the patent holder incurs any strict liability vis-à-vis any party other than the one that, threatened with enforcement measures, has been forced to comply with an interlocutory injunction. The reasons given for this judgment focus exclusively on the special position of that other party and is based partly on the consideration that penalty payments forfeited for failure to comply with an interlocutory injunction remain forfeit after a judgment to the contrary in the proceedings on the merits. This brings about a certain balance between the conflicting interests. There is no such balance when strict liability is extended to parties such as Menzis, which do not have to pay penalties for non-compliance with the injunction.

5.8 Incidentally, Menzis also does not take the position that AstraZeneca is liable because it has served an interlocutory ruling, believing rather that the strict liability assumed in Ciba Geigy v [… 1] should apply by analogy, because the granting of a patent, just like an interlocutory judgment, is a provisional measure. That reasoning cannot succeed for the very reason that, as the Court considered above, the judgment in Ciba Geigy v [… 1] was not based solely on the provisional nature of a decision in interim proceedings, but also on (the precisely final nature of) the obligation to pay forfeited penalty payments. In addition, a decision by the European Patent Office (EPO) to grant a patent cannot be equated with an interlocutory ruling. The fact that a decision of the EPO can be revoked or annulled does not make it a provisional measure by nature like an interlocutory ruling. Moreover, the validity of (the Dutch part of) the patent in this case was examined not only by the EPO but also by the Dutch court of first instance.

5.9 The fact that the patent was declared invalid on the ground that the invention claimed was obvious to the average person skilled in the art also does not imply that AstraZeneca knew that the invention was not inventive. The fact that it is now established that the claimed invention was obvious is not the same as establishing that the lack of inventive step was obvious. As the court will explain below in the context of assessing culpability, the latter is not the case.

No culpability

5.10 The question can be left open whether AstraZeneca’s invocation of its patent during the period between the expiry of the SPC and this court’s judgment in the invalidity cases was unlawful towards Menzis if, during that period, AstraZeneca knew, or ought to have realised, that there was a serious, non-negligible chance that the patent would not stand up in opposition or invalidity proceedings. The court finds that it cannot be concluded that AstraZeneca knew or ought to have realised that at the time.

5.11 In this regard, Menzis argues, first, that there could not reasonably have been any debate about the invalidity of the patent during the relevant period and that its invalidity was even obvious. That cannot be upheld in the light of the ruling of the District Court of The Hague of 7 March 2012, which precisely concluded that the patent was valid. In support of its argument on the foreseeability of the patent’s invalidity, Menzis refers mainly to the findings of this Court’s judgment of 10 June 2014. However, that judgment had not yet been made available during the relevant period. The fact that this Court, an appeal court, reached a different verdict on the issue of inventive step than the district court does not imply that AstraZeneca already knew or ought to have realised, prior to that judgment, that there was a serious, non-negligible chance that the patent would still be declared invalid on appeal.

5.12 More substantively, Menzis argues that AstraZeneca defended the patent’s inventive step by arguing that the average person skilled in the art would not have been motivated to develop an XR formulation of quetiapine and would not have had reasonable expectations of success. According to Menzis, such defence was obviously untenable because a publication by Gefvert et al., identified by the court as the closest prior art, concluded that ‘a more convenient dosage regimen would be beneficial’. However, it does not automatically follow from that conclusion by Gefvert et al. that the average person skilled in the art would have been motivated to develop an XR formulation of quetiapine and would have had reasonable expectations of success in doing so, especially considering what AstraZeneca has argued regarding motivation and expectations of success. Nor did this court, in its judgment of 10 June 2014, allow the lack of inventive step to be based solely on the conclusion reached by Gefvert et al. Menzis did not allege that this argument from AstraZeneca was obviously untenable for other substantive reasons, or at least did not provide sufficient reasons for such an allegation.

5.13 Secondly, Menzis argues that the English High Court ruling of 22 March 2012 had found the claimed invention to have no inventive step, and that other foreign courts had come to the same conclusion. However, even this does not necessarily mean that, in this case, AstraZeneca knew, or ought to have realised, that there was a serious, non-negligible chance that the patent would not stand up in opposition or invalidity proceedings. The present case involves invocation of the Dutch part of the European patent. In that context, the patent holder should, in principle, have been able to rely on the ruling on validity given on 7 March 2012 by the Dutch court of first instance. In addition, the majority of the foreign courts that ruled on the validity of the patent during the relevant period reached the same verdict as the Dutch court.

[…]

No unjust enrichment

5.16 Menzis’s claim of unjust enrichment within the meaning of Book 6, Art. 212, of the Dutch Civil Code cannot succeed either. The fact that the enrichment alleged by Menzis is not unjust and the reasons why it is not unjust are implicit in the aforementioned finding that AstraZeneca did not act unlawfully by asserting its patent rights vis-à-vis third parties during the relevant period. The fact that the cancellation of the patent has retroactive effect does not alter this.

5.17 The aforementioned finding that the alleged enrichment was not unjust follows from the judgment in CFS Bakel v Stork. It is also in compliance with the law. Menzis argues that AstraZeneca was enriched by the profits it made from the sale of Seroquel XR. However, that enrichment is in principle justified by the agreements under which AstraZeneca sold those products. The retroactive effect of the patent’s cancellation does not alter this subsequently. Article 75(6) of the Dutch Patents Act provides that such retroactivity does not affect agreements entered into before a cancellation insofar as they were performed before that cancellation. That rule applies here and serves legal certainty.

5.18 Article 75(6) of the Dutch Patents Act provides that, for reasons of equity, parties may claim the reimbursement of sums paid under agreements insofar as justified by the circumstances. The history of the law shows that the legislator had thereby considered, inter alia, the situation where a licensee has paid a licence fee in advance for a lengthy period and the patent is declared invalid shortly afterwards. There is no such situation here. In this case, there are no other reasons of equity that could justify a deviation from the guiding principle chosen by the legislator. On the contrary, the fact that, as the court has previously ruled, AstraZeneca cannot be reproached for asserting its patent rights during the relevant period argues against deviating from that principle.

3 Assessment of the grounds of appeal

3.1 Subsection 1.1 of the appeal complains that the court of appeal, in paras. 5.16-5.17 of its judgment, was wrong to hold that the fact that the enrichment alleged by Menzis was not unjust and the reasons why it was not unjust were implicit in the finding that AstraZeneca did not act unlawfully by asserting its patent rights vis-à-vis third parties during the relevant period. According to this subsection, the mere fact that there was no unlawful conduct does not automatically mean that the enrichment is justified. Subsection 1.2 complains that the court of appeal in any event failed to recognise that the enrichment of AstraZeneca in this case was unjust, or at the very least gave insufficient reasons for why such enrichment at the expense of Menzis was nevertheless justified in this case.

3.2 Subsection 1.1 correctly assumes that the fact that AstraZeneca did not act unlawfully does not automatically mean that the enrichment alleged by Menzis is justified. However, it does not provide any factual basis for overturning the court of appeal’s decision. The court of appeal did not dismiss the claim of unjust enrichment solely on the ground that AstraZeneca had not acted unlawfully. The facts on the basis of which the court of appeal held that AstraZeneca had not acted unlawfully (see the Advocate General’s Opinion at 3.21) also support its finding that AstraZeneca was not unjustly enriched.⁶ That finding does not demonstrate any error of law and did not require any further reasoning. This is where also the complaints of Subsection 1.2 falter.

3.3 Section 2 is based on the assumption that the yardstick for assessing whether AstraZeneca acted unlawfully towards Menzis is that AstraZeneca knew or ought to have realised that there was a serious, non-negligible chance that its patent would not stand up in opposition or invalidity proceedings. Sections 2.4 and 2.5 complain, inter alia, that the court of appeal, by its finding in para. 5.13, set the threshold for the assumption of culpability too high by assuming that the holder of a European patent that was valid in various European countries, including the Netherlands, should, in principle, be able to rely on the finding of the Dutch court of first instance regarding the validity of the Dutch part of the patent. According to Secs. 2.4 and 2.5, when assessing whether AstraZeneca at any time during the period from 24 March 2012 (the expiry of the SPC) to 10 June 2014 (the cancellation of the Dutch part of the patent) knew or ought to have realised that there was a serious, non-negligible chance of the patent being declared invalid, the opinion of the Dutch court of first instance does not carry any more weight than the judgments of other European courts. Nor is the majority opinion decisive. Moreover, the complaint is made that the court of appeal did not give sufficiently comprehensible reasons for its decision because, without further explanation, no specific moment in the period from 24 March 2012 to 10 June 2014 can be identified when AstraZeneca knew or ought to have realised that there was a serious, non-negligible chance of the patent being declared invalid.

3.4 In para. 5.10, the court of appeal held that it could not be concluded that, in the period from 24 March 2012 to 10 June 2014, AstraZeneca knew or ought to have realised that there was a serious, non-negligible chance that the patent would not stand up in opposition or invalidity proceedings. The court of appeal based its decision on the following.

(1) In light of the ruling of the District Court of The Hague of 7 March 2012, which deemed the patent valid, Menzis’s contention that there could not reasonably have been any debate about the invalidity of the patent in the relevant period and that its invalidity was even obvious cannot be upheld.

(2) It does not automatically follow from Gefvert et al.’s conclusion that the average person skilled in the art would have been motivated to develop an XR formulation of quetiapine and would have had reasonable expectations of success in doing so, especially considering what AstraZeneca has argued regarding motivation and expectations of success. Therefore, given Gefvert et al.’s conclusion, Menzis’s argument that AstraZeneca’s assertion was obviously untenable does not stand up.

(3) Also the fact that the English High Court (see 2.1(xii) above) did not find the claimed invention to be inventive and that other foreign courts came to the same conclusion does not mean that AstraZeneca knew, or ought to have realised, that there was a serious, non-negligible chance that the patent would not stand up in opposition or invalidity proceedings. The patent holder should, in principle, have been able to rely on the ruling on validity by the Dutch court of first instance. In addition, the majority of the foreign courts that had ruled on the validity of the patent during the relevant period had reached the same verdict as the Dutch court.

3.5 The court of appeal took into account all the facts mentioned in 3.4 above in combination and did not base its judgment on only one of those facts. Insofar as Secs. 2.4 and 2.5 disregard this, they lack any factual basis. Moreover, the court of appeal did not set the threshold for assuming culpability too high. It was entitled to rule that AstraZeneca should in principle have been able to rely on a judgment on validity by the Dutch court of first instance, particularly since several courts abroad had reached a similar conclusion at that time. The court of appeal therefore did not commit an error of law. Nor is its decision incomprehensible. The complaints set out in 3.3 above therefore fail.

3.6 Nor can the remaining grounds of appeal overturn the court of appeal’s decision. The Supreme Court does not need to give reasons for reaching its decision. This is because, when assessing such complaints, it is not necessary to answer questions that are in the interests of the uniform application or development of the law (see Art. 81(1) of the Dutch Judiciary (Organisation) Act).

4 Decision

The Supreme Court:

• ‒ dismisses the appeal;

[…]

Translated from the Dutch by Gwyneth Little, Perth, Scotland.

Footnotes