Ablynx NV & Anor v VHsquared Ltd & Ors

This is a dispute between two licensees each of which holds a licence to exploit three European patents, each of which has a UK designation. The dispute mainly turns on the interpretation of EU Regulation 1215/2012 usually known as Brussels 1 Recast, or the Recast Regulation.

The licences derive from the same patentee, and the same European patents. Ablynx NV alleges that VHsquared Ltd has infringed the UK designations of the patents within the jurisdiction. The licence (in fact a sub-licence) granted to VHsquared is limited to what the licence calls “the Reserved Sector”. The licence granted to Ablynx applies to all fields except the Reserved Sector. The licence under which VHsquared’s sub-licence was granted (described below as the Unilever Licence) includes an exclusive choice of court clause conferring jurisdiction on the Belgian court. VHsquared says that its activities are covered by its licence and therefore there has been no infringement; and that any question about the scope of the licence is within the exclusive jurisdiction of the Belgian court. It also raises other defences to which I will come. Ablynx says that the exclusive jurisdiction clause is invalid because the Recast Regulation allocates exclusive jurisdiction over the validity of patents with a UK designation to the UK court.

Proceedings are on foot both in England and in Belgium. The UK court is the court first seised. The underlying question is which of the two courts has jurisdiction over the dispute. But there is a preliminary question: which court should decide who has jurisdiction over the substantive dispute?

HHJ Hacon decided that the UK court had exclusive jurisdiction; and refused to stay the proceedings in England. The reason for his decision was that the exclusive jurisdiction clause in the Unilever Licence had no legal force because it purported to exclude the UK court’s exclusive jurisdiction in an action concerned with the validity of the UK designations of the European patents. His judgment is at [2019] EWHC 792 (Pat).

I can take the facts from the judge’s judgment. I have simplified them somewhat. To aid simplicity, except where necessary to distinguish between them, I have referred to the current Appellants as VHsquared (even where some of the activities were carried out by other Appellants); and I have referred to the current Respondents as Ablynx. VHsquared is a company domiciled in the United Kingdom. Ablynx is a company domiciled in Belgium.

The invention relates to immunoglobulins derived from camelid antibodies. (Camelids are an animal family which includes camels, llamas and other even-toed ungulates). The patentee is a Belgian university abbreviated to VUB. Each of the patents is a European patent with a UK designation (and, I assume, designations in other contracting states). The patents expired in August 2013.

On 16 April 1997 the patentee, VUB, entered into an agreement with the third defendant ('Unilever NH') which included a worldwide licence granted to Unilever NH ('the Unilever Licence') under the patents for certain applications of the inventions. The licence was partly exclusive and partly non-exclusive. The exclusive licence was defined as follows ('the University' is VUB):

“3.1.a The University shall grant Unilever an exclusive and territorially unlimited licence to exploit [the patents] and corresponding foreign patents … for the following products and sectors:

(i)

packed food products;

(ii)

washing and cleaning agents;

(iii)

cosmetics with a non-medical orientation; and

(iv)

process auxiliary agents, more specifically the catalytic and separation process auxiliary agents, for use in fields (i), (ii) and (iii) above.

3.1.b The University shall grant Unilever a non-exclusive and territorially unlimited licence to exploit [the patents] and corresponding foreign patents, as detailed in Annexe 1 to this Agreement for OTC diagnostics for non-medically orientated cosmetics, and for the use of antibodies in cattle fodder, for process auxiliary agents, more specifically the catalytic and separation process auxiliary agents for use in both fields stated in this Article 3.1.b.”

Those sectors are called “the Reserved Sector”. Article 9 provided:

“9.

Applicable law and settlement of disputes

9.1

Belgian law shall govern this Agreement. The Court of Brussels shall alone be competent in case of dispute between the Parties or one of their (sub)licensee(s) concerning this Agreement.

9.2

In exemption from Article 9.1, the Parties hereby also agree that all disputes on the definition of the Reserved Sector, the scope of [the patents] and its delimitation from the Community Patents shall exclusively be settled by arbitral tribunal consisting of one jurist and two scientists, sitting in Brussels, according to the rules of the International Chamber of Commerce.”

On 23 September 1998 VUB granted a licence under the patents ('the VIB Licence') to an institute abbreviated to VIB. It was an exclusive worldwide licence for all fields except the Reserved Sector. The Reserved Sector was expressly defined to have the same meaning as in the Unilever Licence. On 14 November 2001 VIB granted to

It is common ground that the Ablynx Licence could not be broader than the exclusive VIB Licence, so its permitted field necessarily excluded the Reserved Sector. The Ablynx Licence also contained an acknowledgment by Ablynx that it had received a copy of the Unilever Licence in which the Reserved Sector had been defined.

By an agreement dated 2 June 2005 ('the Novation Agreement') between VUB, Unilever NH and BAC IP BV ('BAC'), the Unilever Licence was novated from Unilever NH to BAC. BAC was part of the Unilever group, under consideration for spinning off. Under the terms of the Novation Agreement, BAC granted back to Unilever NH a non-exclusive worldwide licence to exploit the Patents in the Reserved Sector. The Novation Agreement contained the following terms:

By an agreement dated 22 December 2010 Unilever NH granted VHsquared a sublicence under the patents for the Reserved Sector.

In summary, by 2011 Ablynx had the exclusive right to exploit the inventions claimed in the patents by using camelid antibodies for certain medical uses. For its part, VHsquared had a non-exclusive licence to exploit the inventions in the Reserved Sector.

There have been two rounds of litigation in the Netherlands between the parties. The judge set out the details. Only the first is directly relevant to this appeal, because it concerned the scope of the Unilever Licence. That round of that litigation went to the Dutch Court of Appeal. VHsquared’s case on the appeal was that it was undoubtedly licensed to produce packaged food products and that there was no limitation as regards the nature of those packaged products. They could be what VHsquared called 'functional foods', including foods containing rotavirus antibodies.

The Court recorded at [2.10] that Unilever had carried out research in Bangladesh and India between 2007 and 2012 into the use of antibodies directed against rotavirus in children.

During the course of the appeal, Ablynx narrowed their approach to the interpretation of the “Reserved Sector”. They contended that the patents would be infringed by “products with a therapeutic or prophylactic effect with regard to specific pathogens, whatever the nature of those products”.

The Court rejected VHsquared’s argument that they were entitled to make any sort of packaged foodstuff. It held (in agreed translation):

“4.10

The term (packaged) foodstuff is also unsuitable to serve as a limitation, now that the qualification as foodstuff, as Unilever et al. appears to fail to understand, does not exclude that such a product must also be deemed a medicinal product, while – as Unilever acknowledges – it is precisely medicinal products that are excluded from the licence. In addition, Unilever asserted that its licence did not extend to 'pills, syringes or powders (products that can be purchased from a pharmacy)'. The location where the products are offered cannot, however, be deemed a suitable delineation. After all, medicinal products are also offered in supermarkets, while foodstuffs (including functional foods) can also be sold in pharmacies and are not to be deemed a medicinal product.”

As the judge recorded, in broad terms both sides to the litigation agreed that Ablynx’s exclusive licence was for certain medicinal products whereas VHsquared’s licence was for non-medicinal products. The difficulty was being precise about where the line was to be drawn. The Dutch Court of Appeal said:

“4.12

… What keeps the parties divided is what 'medicinal product' is to be understood to mean in this respect. As this limitation, which is acknowledged by both parties, is not included in the licence agreement itself, the literal wording of the agreement itself cannot provide any clarity about the contents and meaning of that term – and thus of the scope of the licence. The Court of Appeal sees cause in these circumstances of this case to not only take account of the meaning of the literal wording of the agreement, but also to take note of the history of the making of the licence agreement, in order to determine what sense each of the parties could reasonably assign in the given circumstances to the provisions of the licence agreement and what they could reasonably reasonably expect of each other in this respect. This approach is also in accordance with Belgian law that applies to the licence agreement, which stipulates that account must be primarily taken of the common intent of the parties …”

The Court found that the new limited declaration sought by Ablynx was satisfactory:

“4.11

… This limitation particularly does justice to the reasonable expectation – acknowledged by Ablynx – that Unilever could have that it was allowed to develop and market functional foods under the licence, which have a 'general' health promoting effect, in the sense that this seeks to promote the intrinsic functioning of the body. This includes foodstuffs to increase resistance, to reduce the cholesterol level and/or blood pressure, to optimise the functioning of the intestines and of organs like the liver and kidneys.

…

… this limitation also does justice to VUB's expectation that the licence did not include products – whether or not they are to be deemed foodstuff – for curing (therapeutic effect) or prevention (prophylactic effect) of ailments caused by pathogens (pathogens of biological origin, of which viruses, bacteria and moulds are the most common).”

At [4.17] the Court recorded that Ablynx had submitted, unchallenged, that the nanobodies necessary for that research had been developed by Unilever in the Netherlands during the term of the patents; using technology protected by the patents; and that the research was aimed at the production of foodstuffs to prevent or cure an ailment caused by a pathogen. It went on to say that in view of the interpretation of the licence that it had adopted, Unilever had “thereby gone beyond it”. The fact that a foodstuff was not on the market did not detract from that conclusion. The Court of Appeal thus overruled the opinion of the District Court that the research to which the Court of Appeal referred in paragraph [2.10] fell within the terms of the Unilever Licence; and also held that the same conclusion applied to preparatory actions of Unilever in the Netherlands.

Ablynx issued the claim form on 4 September 2018. It alleged that VHsquared had infringed the patents. Infringement is defined by section 60 of the Patents Act 1977. A person infringes a patent “if, but only if, while the patent is in force, he does any of the following things in the United Kingdom in relation to the invention without the consent of the proprietor of the patent”. There then follows a list of what activities can qualify as infringements.

We are concerned with the UK designation of a European patent. This takes effect as if it were the grant of a national patent under the Patents Act 1977: Patents Act 1977 s. 77 (1). By virtue of section 67 (1) of the Act the holder of an exclusive licence under the patent has “the same right as the proprietor of the patent” to bring proceedings for infringements. In addition, section 67 (3) requires the patentee to be made a party to the proceedings.

Paragraph 5 of the Particulars of Claim defines the “Ablynx Exclusive Field of Use” by reference to the scope of the Ablynx Licence (i.e. excluding the Reserved Sector as defined by the Unilever Licence). Paragraph 13 of the Particulars of Claim alleges that VHsquared has obtained a wrongful advantage. The essence of the allegation is that during the lifetime of the patents it was conducting research into the use of camelid derived proteins for the treatment of gastrointestinal pathogens. After the expiry of the patents it changed its field of research into treatments for inflammatory bowel disease. That resulted in the creation of an antibody called V565. The creation of that antibody was the result of acts of infringement set out in the Particulars of Infringement and therefore amounts to a wrongful advantage. The Particulars of Infringement assert that VHsquared has performed a number of acts “without the consent” of Ablynx. Those particulars do not mention V565 in terms. They allege infringements of the patents by carrying out certain activities “all of which fall within the scope of the Ablynx Exclusive Field of Use”. The Particulars quote from posts on VHsquared’s website. These include:

“Our novel antibody platform will provide a portfolio of new and innovative functional food products to target gastrointestinal infectious and immune mediated diseases.” (about April 2010)

They also referred to a letter from VHsquared’s patent attorneys dated 16 October 2012 in which it was said:

“VHsquared is carrying our early experimental research relating to certain gastrointestinal pathogens. Antibodies to these pathogens, including rotavirus, may in due course be exploited commercially.

…

VHsquared’s research is for experimental purposes related to the subject matter of the Patents.”

On 30 January 2019 VHsquared issued an application contesting the jurisdiction of the UK court. The application was initially supported by a witness statement of Mr Ian Karet. He pointed out that the claim referred to a VHsquared antibody product V565. He described V565 as a genetically engineered camelid antibody. It neutralises a protein called TNFα which is produced naturally within a mammalian body; and which promotes inflammation in response to injury. TNFα is not a pathogen. The critical feature of V565 is its stability in the intestine, which facilitates oral delivery. It

The six italicised words are all that is said in the evidence about invalidity. I should, perhaps, add that the reference to experimental activities is a reference to the exception contained in section 60 (5) (b) of the Patents Act 1977, which qualifies the definition of infringement.

It must also be emphasised that if the action proceeds in the Patents Court that court will necessarily have to rule on the scope of the Unilever Licence, even though the choice of court clause allocates exclusive jurisdiction to do that to the Belgian court.

On 11 March 2019 VHsquared issued proceedings in the Brussels Enterprise Court. Those proceedings raise two issues:

The validity of the patents is not in issue in Belgium.

The relevant regulation is the Recast Regulation. It has predecessors both in regulation and in convention. The immediately preceding regulation was Regulation (EC) No 44/2110, commonly called Brussels I. Under that Regulation there was a problem known as “the Italian torpedo”. How that worked is explained in Hartley: Choice-of-Court Agreements and the New Brussels I Regulation (2013) 129 LQR 309. In short, it was a litigation device that relied on a combination of the EU rule of lis pendens and the lengthy delays in legal proceedings that occur in certain EU member states to allow unscrupulous litigants to block proceedings against them in other member states. The idea was that persons facing patent infringement actions could

In the light of this ruling it was decided that the Regulation had to be recast. In its proposal the European Commission said at paragraph 3.1.3:

“Where the parties have designated a particular court or courts to resolve their dispute, the proposal gives priority to the chosen court to decide on its jurisdiction, regardless of whether it is first or second seised. Any other court has to stay its proceedings until the chosen court has established or – in case the agreement is invalid - declined jurisdiction. This modification will increase the effectiveness of choice of court agreements and eliminate the incentives for abusive litigation in non-competent courts.”

The proposal was incorporated into the Recast Regulation.

A number of the recitals to the Recast Regulation are relevant:

“(15)

The rules of jurisdiction should be highly predictable and founded on the principle that jurisdiction is generally based on the defendant’s domicile. Jurisdiction should always be available on this ground save in a few well-defined situations in which the subject-matter of the dispute or the autonomy of the parties warrants a different connecting factor.

(19)

The autonomy of the parties to a contract, other than an insurance, consumer or employment contract, where only limited autonomy to determine the courts having jurisdiction is allowed, should be respected subject to the exclusive grounds of jurisdiction laid down in this Regulation.

(20)

Where a question arises as to whether a choice-of-court agreement in favour of a court or the courts of a Member State is null and void as to its substantive validity, that question should be decided in accordance with the law of the Member State of the court or courts designated in the agreement, including the conflict-of-laws rules of that Member State.

(21)

In the interests of the harmonious administration of justice it is necessary to minimise the possibility of concurrent proceedings and to ensure that irreconcilable judgments will not be given in different Member States. There should be a clear and effective mechanism for resolving cases of lis pendens and related actions, and for obviating problems flowing from national differences as to the determination of the time when a case is regarded as pending. For the purposes of this Regulation, that time should be defined autonomously.

(22)

However, in order to enhance the effectiveness of exclusive choice-of-court agreements and to avoid abusive litigation tactics, it is necessary to provide for an exception to the general lis pendens rule in order to deal satisfactorily with a particular situation in which concurrent proceedings may arise. This is the situation where a court not designated in an exclusive choice-of-court agreement has been seised of proceedings and the designated court is seised subsequently of proceedings involving the same cause of action and between the same parties. In such a case, the court first seised should be required to stay its proceedings as soon as the designated court has been seised and until such time as the latter court declares that it has no jurisdiction under the exclusive choice-of-court agreement. This is to ensure that, in such a situation, the designated court has priority to decide on the validity of the agreement and on the extent to which the agreement applies to the dispute pending before it. The designated court should be able to proceed irrespective of whether the non-designated court has already decided on the stay of proceedings.”

The basic jurisdictional rule is that a defendant must be sued in the member state in which he is domiciled. There is an exception to that rule in the case of tort. Under article 7 (2) a defendant to an action in tort may be sued in the member state in which the harmful act occurred. It is common ground that this gives the UK court jurisdiction over an action for infringements of a patent where the infringement takes place within the United Kingdom. Article 6 of the Recast Regulation provides for exclusive jurisdiction in some cases.

Article 24 provides, so far as relevant:

“The following courts of a Member State shall have exclusive jurisdiction, regardless of the domicile of the parties

(1)

in proceedings which have as their object rights in rem in immovable property or tenancies of immovable property, the courts of the Member State in which the property is situated….

(2)

in proceedings which have as their object the validity of the constitution, the nullity or the dissolution of companies or other legal persons or associations of natural or legal persons, or the validity of the decisions of their organs, the courts of the Member State in which the company, legal person or association has its seat. In order to determine that seat, the court shall apply its rules of private international law;

(3)

in proceedings which have as their object the validity of entries in public registers, the courts of the Member State in which the register is kept;

(4)

in proceedings concerned with the registration or validity of patents, trade marks, designs, or other similar rights required to be deposited or registered, irrespective of whether the issue is raised by way of an action or as a defence, the courts of the Member State in which the deposit or registration has been applied for, has taken place or is under the terms of an instrument of the Union or an international convention deemed to have taken place.

Without prejudice to the jurisdiction of the European Patent Office under the Convention on the Grant of European Patents, signed at Munich on 5 October 1973, the courts of each Member State shall have exclusive jurisdiction in proceedings concerned with the registration or validity of any European patent granted for that Member State.”

At this stage, I make two comments on article 24 (4). First, the words “irrespective of whether the issue is raised by way of an action or as a defence” were not in the previous regulation. They reflect the decision of the ECJ in (Case C-4/03) Gesellschaft für Antriebstechnik mbH & Co KG v Lamellen und Kupplungsbau Beteiligungs KG [2006] FSR 45 (“GAT”), to which I refer further below.

Second, the patents in suit all derive from European patents. A patent of that kind is granted by the EPO under the European Patent Convention. The applicant may designate the states in which he seeks patent protection. The effect of such a grant is that the patent takes effect as a national patent in each of the designated states. Once the patent has been granted, a person has a period in which to oppose the grant on the ground of invalidity: in effect to apply for its revocation. If the opposition succeeds in the EPO then the patent ceases to have effect in each of the designated states. But in parallel, the Convention allows each designated state to decide the question of invalidity. That question may arise, for example, in a subsequent action for infringement. But if a national court decides that a patent is invalid, its decision only applies to the designated state concerned. The patent will continue to have effect in all other designated states unless those states (or the EPO) decide that it is invalid. So a European patent may be valid in some states but invalid in others.

I mention this because the philosophy underlying article 24 is that only one member state should have exclusive jurisdiction in cases falling within that article; so that, where possible, a single authoritative judgment can be given by the single member

The consequence of this is that the Belgian courts would have exclusive jurisdiction over the Benelux designation of the patents; the English courts over the UK designation of the patents and so on. So multiple member states would have exclusive jurisdiction over what is, in effect, the same European patent, albeit that the effect of their decisions would only affect their own national territory. It seems, therefore, as though the desideratum of a single member state being able to pronounce on the validity of what is in effect the same European patent is unlikely to be achieved by article 24 (4).

In cases of infringement, the CJEU has justified this on the basis that the question of infringement is to be decided according to national law (which may differ from member state to member state); and on the further basis that the facts of the alleged infringement may also differ from member state to member state. If, therefore, the courts of two member states differed in their conclusions, that would not result in irreconcilable judgments. It also envisaged that where invalidity was in issue, a separate challenge had to be brought in each member state where the patent had effect: see (Case C-539/03) Roche Nederland BV v Primus [2007] FSR 5. Birss J discussed this question (amongst others) in Eli Lilly and Co v Genentech Inc [2017] EWHC 3104 (Pat), [2018] 1 WLR 1755.

Article 25 provides:

“1.

If the parties, regardless of their domicile, have agreed that a court or the courts of a Member State are to have jurisdiction to settle any disputes which have arisen or which may arise in connection with a particular legal relationship, that court or those courts shall have jurisdiction, unless the agreement is null and void as to its substantive validity under the law of that Member State. The agreement conferring jurisdiction shall be either:

(a)

in writing or evidenced in writing;

(b)

in a form which accords with practices which the parties have established between themselves; or

(c)

in international trade or commerce, in a form which accords with a usage of which the parties are or ought to have been aware and which in such trade or commerce is widely known to, and regularly observed by, parties to contracts of the type involved in the particular trade or commerce concerned.

…

4.

Agreements … conferring jurisdiction shall have no legal force … if the courts whose jurisdiction they purport to exclude have exclusive jurisdiction by virtue of Article 24.”

Article 26 (1) provides:

“Apart from jurisdiction derived from other provisions of this Regulation, a court of a Member State before which a defendant enters an appearance shall have jurisdiction. This rule shall not apply where appearance was entered to contest the jurisdiction, or where another court has exclusive jurisdiction by virtue of Article 24.”

Article 27 provides:

“Where a court of a Member State is seised of a claim which is principally concerned with a matter over which the courts of another Member State have exclusive jurisdiction by virtue of Article 24, it shall declare of its own motion that it has no jurisdiction.”

Article 29 provides:

“1.

Without prejudice to Article 31(2), where proceedings involving the same cause of action and between the same parties are brought in the courts of different Member States, any court other than the court first seised shall of its own motion stay its proceedings until such time as the jurisdiction of the court first seised is established.

2.

In cases referred to in paragraph 1, upon request by a court seised of the dispute, any other court seised shall without delay inform the former court of the date when it was seised in accordance with Article 32.

3.

Where the jurisdiction of the court first seised is established, any court other than the court first seised shall decline jurisdiction in favour of that court.”

Article 31 provides:

“2.

Without prejudice to Article 26, where a court of a Member State on which an agreement as referred to in Article 25 confers exclusive jurisdiction is seised, any court of another Member State shall stay the proceedings until such time as the court seised on the basis of the agreement declares that it has no jurisdiction under the agreement.

3.

Where the court designated in the agreement has established jurisdiction in accordance with the agreement, any court of another Member State shall decline jurisdiction in favour of that court.”

There are three possible gateways by which the UK court may have jurisdiction over the claim against VHsquared (I leave out of account the other defendants). The first is under article 4, because of VHsquared’s domicile. The second is under article 7 (2), because the United Kingdom is said to be the place where the harmful event occurred. Both those are non-exclusive jurisdictions. The third is under article 24 (4). That is an exclusive jurisdiction. In so far as the Unilever Licence is concerned, the Belgian court has jurisdiction over the Belgian domiciled companies, including Ablynx. That is a non-exclusive jurisdiction. In addition, the Belgian court has jurisdiction under article 31 (2) because of the choice of court clause. That, too, is an exclusive jurisdiction. The contest, then, is in effect between two competing exclusive jurisdictions.

The judge approached the issues before him in the following order:

The argument for Ablynx is that the six words in Mr Karet’s witness statement mean that English action is “concerned with” the validity of patents. That has two consequences. The first is that the English court has exclusive jurisdiction. The second is that the choice of court clause in the Unilever Licence has no legal force as regards a claim that a UK designation of a European patent is invalid.

Although VHsquared disputes both those propositions, it raises a different and anterior point. That is that whether Ablynx’s argument is a good one or not should be decided by the Belgian court in accordance with article 31 (2).

Ms Lane QC submits that the order in which the judge approached the various issues was wrong. He should first have addressed the question whether article 31 (2) was engaged. He was not required to decide more than that there was an apparent or prima facie case that it was engaged. If he came to that conclusion (as he ultimately did) then he was required by article 31 to stay the proceedings.

Whether the jurisdiction agreement was or was not overridden by article 25 (4) was then a matter for the Belgian court to decide, because it bore directly on the legal force of the choice of court clause. If the Belgian court decided that article 25 (4) did override the jurisdiction agreement because the dispute concerned the validity of a patent with a UK designation, then the English action would be revived; and the stay lifted. If, on the other hand, the Belgian court decided that article 25 (4) did not override the choice of court clause, it would be for that court to determine the dispute in so far as it fell within the scope of the clause. That is the policy described in recital (22).

Accordingly, for the UK court to decide for itself whether article 25 (4) overrode the choice of court clause was to usurp the function of the Belgian court.

Ms Lane submits that article 31 (2) is engaged. She says:

The judge agreed that the test was as set out at [53] i) above. Although there is no authority on the point, the test that the judge applied is supported by many commentators: Cheshire, North & Fawcett Private International Law (15^th ed) p 453 (“apparent” or “prima facie” jurisdiction agreement); Dickinson and Lein: The Brussels I Regulation Recast para 11.50 (“Prima facie evidence of the conditions for

I did not understand that test to be in dispute. But the remainder of the Appellants’ analysis is incomplete. First, there is the question whether Ablynx is bound by the jurisdiction agreement, which is contained in a contract to which it is not a party. Second, there is the question of the scope of the jurisdiction clause. The judge approached both these questions as questions of Belgian law.

Ablynx’s first argument is that there is no prima facie or apparent case that article 31 (2) is engaged. There is no claim under the Unilever Licence: the claim is one in tort. Nor is there any dispute about the meaning of that licence; and VHsquared has not explained why it says that its activities during the lifetime of the patents were authorised by the licence. In broad terms, the judge agreed. He said:

“[85] Ms Lane took me to the Particulars of Infringement which sets out the history and alleged effect of the licences in some detail. She invited me to infer that inevitably the trial will for the most part be taken up with arguments about the scope of VHsquared's licence. I don't see why. The defendants have, entirely reasonably, avoided entering an appearance and so there is no pleaded Defence. But one real possibility is that little or none of Ablynx's argument on the scope of VHsquared's licence will be in dispute. It was again incumbent on the defendants to provide at least some reason for me to suppose that there will be arguments of substance on the scope of the licence.

[86] The defendants might seek to re-argue all their points on the scope of Unilever Licence which did not succeed in the first Dutch litigation. Neither side raised issue estoppel. The infringing product would be different: it is now V565 instead of the rotavirus antibody. I was not told whether, and if so why the defendants believe that the conclusion of the Dutch Court of Appeal on scope was wrong. I accept that the Dutch Court of Appeal (in the first Dutch Action) did not have to consider V565. This might make some difference. But Crohn's disease involves an abnormal reaction of the patient's immune system to certain bacteria in the intestines. The ailment is treated by V565. If the view on the limits of Unilever's licence by the Dutch Court of Appeal in its paragraph 4.11 (see above) is taken to be correct, the ways in which argument on the scope of that licence can be reheated might be limited.”

I agree with the judge that it is by no means clear that VHsquared does take issue with the interpretation of the Unilever licence by the Dutch court. But that still leaves the question whether the impugned activities fall within the scope of the licence as so interpreted. Based on both the Particulars of Infringement and Mr Karet’s evidence it seems likely (to put it no higher) that VHsquared will argue either that the impugned activities are within the scope of the licence as interpreted by the Dutch court; or that they were experimental and therefore outside the national definition of infringement. A claim that the impugned activities fall within the scope of the licence is also one of the claims made in the Belgian proceedings. In the quoted passage, however, the judge also concentrated on V565. But V565 was not referred to in the Particulars of Infringement. Contrary to the judge’s belief, it is not said to be an infringing product. It is referred to in a different context; namely that V565 takes unfair advantage of other activities outside the Reserved Sector that are alleged to have infringed the patents before their expiry. Since the hearing before the judge Ablynx have applied for permission to amend the Particulars of Infringement. But VHsquared has asserted that the application has been made outside the limitation period and therefore that it cannot be allowed. That is a live dispute; and I do not therefore consider that it would be right to treat the Particulars of Infringement as if they had already been amended.

It is only necessary for VHsquared to present a prima facie case that Ablynx is bound by the jurisdiction agreement, because ultimately it will be for the Belgian court to make a definitive ruling. Although, as Mr Turner QC correctly pointed out, Ablynx was not a party to the contract that contained the choice of court clause, that is not the end of the matter. If the choice of court clause was valid as between the original contracting parties as a matter of EU law it appears to be a question of national law whether a third party has succeeded to the original party’s rights and obligations: (Case 71/83) Partenreederei MS Tilly Russ v Haven & Vervoebedrijf Nova NV [1985] Q.B. 931 at [25] and [26].

On this question the judge was presented with expert evidence on Belgian law. The two experts disagreed about whether, as a matter of Belgian law, Ablynx was bound by the jurisdiction agreement. Having considered the evidence of both experts the judge concluded at [105]:

“But I accept that there is a prima facie case that VHsquared is not barred under Belgian law from relying on art.9.1 of the Unilever Licence, assuming that article has the meaning and effect which the defendants say it has.”

I do not consider that this court can say that he was wrong.

The agreement referred to in article 25 is an agreement that the courts of a member state shall have jurisdiction “to settle any disputes which have arisen or which may arise in connection with a particular legal relationship”. In other words, the mere fact that an agreement contains an exclusive jurisdiction agreement is not enough. The agreement must cover the particular dispute that has arisen. The judge correctly identified the necessary elements at [51].

The Appellants assert that there is a prima facie case that the current dispute falls within the scope of clause 9.1. The Respondents assert that there is not.

The judge was once again referred to expert evidence on Belgian law adduced on behalf of each party. One of the issues that they addressed was whether clause 9.1 only applied to contractual claims or disagreements about the meaning and effect of the agreement itself. The two experts disagreed. The judge summarised the disagreement at [95]:

“Mr Troussel [for the Respondents] gave evidence in support of this second reason. He referred to articles of the Belgian patent statute and said that under Belgian law a choice of jurisdiction clause in a patent licence could not impede a licensor's right to assert the patent against a licensee acting outside the terms of the licence. On behalf of [the Appellants] Mr Van Den Broecke said that in Belgian law tortious claims for damages and other extracontractual claims which are connected with a contractual relationship are deemed to be covered by a jurisdiction clause in the contract unless it is stated to be limited to contractual claims. He referred to a commentary by U. Magnus and P. Mankowski.”

At [100] the judge said that this disagreement required the resolution of conflicting evidence of Belgian law. He was not in a position to do that, at least in a way that left the VHsquared without a prima facie case that the clause applied to this dispute.

Once again, I do not consider that this court can say that he was wrong.

Mr Turner sought to persuade us that VHsquared had no arguable defence to the claim. He criticised the paucity of the evidence adduced on behalf of VHsquared; and also criticised the lack of specificity in the Belgian statement of case. I do not regard these arguments as relevant to the narrow issue before us. I might also add that each of the two actions is at a very early stage; and that we do not know what the Belgian rules of pleading are.

Article 31 (2) is new in the Recast Regulation. It did not exist in its predecessors. The Recast Regulation applies to legal proceedings on or after 10 January 2015. The fact that article 31 (2) is new in the Recast Regulation (and was specifically designed to overrule a previous decision of the Court of Justice) means that cases decided under previous regimes must be approached with caution. Its purpose, as explained in recital (22), is “to enhance the effectiveness of exclusive choice-of-court agreements and to avoid abusive litigation tactics” so that “the designated court has priority to decide on the validity of the agreement and on the extent to which the agreement applies to the dispute pending before it.” The judge recognised this purpose. He said at [50]:

“The penultimate sentence indicates that the designated court, i.e. the court chosen to resolve disputes, has priority in deciding the validity and effect of the alleged agreement.”

But earlier in his judgment he had rejected precisely this proposition advanced by Ms Lane on behalf of VHsquared. His reasons for doing so were:

“[17] Ms Lane submitted that the position is clear: art.31(2) is engaged and therefore these proceedings must be stayed. Art.24 could never make a difference in this court because it cannot override art.31(2). That is because art.31(2) is expressly stated to be without prejudice to art.26 but not art.24. The consequence is that all issues arising in these proceedings must be ceded to the Brussels courts, including the question whether art.24(4) is engaged and if so, what should be done about it. It is not the concern of this court.

[18]

I disagree. To my mind art.25(4) explains why there is no mention of art.24 in art.31(2). Art.31(2) is necessarily without prejudice to art.24 since an agreement relied on for a stay under art.31(2) can carry no legal force if it purports to exclude the courts having exclusive jurisdiction under art.24. Even on the assumptions I have stated, art.31(2) cannot apply if art.24(4) is engaged. Art.24(4)'s engagement depends on whether these proceedings are 'concerned with' the validity of the Patents UK within the meaning of art.24(4). I must resolve this last question before I can decide whether the (assumed) agreement carries legal force and therefore whether art.31(2) is engaged.

[19]

I also note that art.26 is itself made subject to art.24. This reinforces my view that the recasting of Brussels I has not altered the hierarchy of provisions awarding jurisdiction, with art.24 at the top. Arts.24 and 25 both speak of 'exclusive jurisdiction', but that conferred by art.24 is the more exclusive.” (emphasis added)

I agree with Ms Lane that the judge’s reasoning is to some extent inconsistent. The question that the judge was addressing in this part of the judgment was which court actually has jurisdiction; not which court decides that question. The judge was right that article 24 is “at the top” in so far as substantive jurisdiction is concerned. As Lord Sales explained in Koza Ltd v Akçil (another case to which I will return) at [25]:

“For the cases falling within article 24, the principle of exclusive jurisdiction cuts across and takes priority over the other principles underlying the Recast Regulation, including the principle of jurisdiction for the courts of the member state where the defendant is domiciled and the principle of respect for party autonomy referred to in recital (19) and reflected in various provisions of the Regulation.” (emphasis added)

I should point out that recital (19) when dealing with overriding party autonomy is referring to the ability of a party to choose as between two (or more) non-exclusive jurisdictions. It is not dealing with a clash between two exclusive jurisdictions. Moreover, the underlying policy does not necessarily preclude the Belgian court from deciding which court has substantive jurisdiction. The Recast Regulation applies uniformly in both Belgium and the UK. The scheme of the Recast Regulation envisages that a court in one member state may have to rule on whether a court in a different member state has exclusive jurisdiction. That is expressly dealt with by article 27. It is clear from the decision of the CJEU in GAT at [19] that article 27 (or its predecessor) is not a jurisdictional rule, but a procedural one:

“Article [27], which, in certain language versions, refers to a claim being brought 'principally', does not provide further clarity. Apart from the fact that the degree of clarity of the wording of that provision varies according to the particular language version, that provision, as the Commission has observed, does not confer jurisdiction but merely requires the court seised to examine whether it has jurisdiction and in certain cases to declare of its own motion that it has none.”

In my judgment, the same is true of article 31 (2). It follows that, under the terms of the Recast Regulation, the Belgian court is empowered to decide whether the English court has exclusive jurisdiction; and that question will involve the question whether the choice of court agreement is overridden by article 25 (4). It is true that the validity of the UK designation of the patents is not, and will not be, directly in issue in the Belgian proceedings. But in my judgment that is beside the point. Ablynx’s goal is to have all issues decided in the English court despite the existence of the choice of court clause. For that purpose it argues that the Belgian court does not have jurisdiction on the ground that the choice of court clause is invalidated by the assertion of invalidity in the English proceedings. It is not required itself to put the validity of the patents in issue in Belgium (if, indeed, it could). All it is required to do is to demonstrate that VHsquared has put it in issue in England; and then argue (in Belgium) about what the consequences are for the effectiveness of the choice of court clause.

As Ms Lane put it, the judge mixed up articles 24 (4) and 25 (4) which are substantive rules about jurisdiction; and article 31 (2) which is a procedural rule about which court should take the lead in deciding the question of jurisdiction where there are parallel actions. I agree.

I consider, therefore, that the judge was wrong when he said: “I must resolve this last question.” In my judgment he was only required to decide whether there was a prima facie case that article 25 (4) did not invalidate the jurisdiction agreement. If the English court reaches that conclusion, then it is up to the Belgian court to decide definitively.

At [38] the judge referred to the decision of the CJEU in (Case C-616/10) Solvay SA v Honeywell Fluorine Products Europe BV [2012] All ER (D) 127 (Jul). In that case Solvay brought infringement proceedings against Honeywell in the Netherlands. The infringements alleged all concerned the same products; but the allegations involved alleged infringements of designations of a European patent in a number of different member states. During the course of the proceedings Solvay applied for the equivalent of an interim injunction restraining the alleged infringements in those member states. Honeywell raised the question of the invalidity of the national parts of the patent concerned. One of the questions referred to the CJEU was whether the Dutch court had jurisdiction to grant the interim injunction; or whether the equivalent of article 24

removed its jurisdiction in favour of each of the member states in which the invalidity of the patent was alleged. The CJEU held that the Dutch court retained its jurisdiction. The nub of the reasoning was that the Dutch court was only being asked to grant interim relief. In considering whether to grant that relief, the court explained:

“[49] According to the referring court, the court before which the interim proceedings have been brought does not make a final decision on the validity of the patent invoked but makes an assessment as to how the court having jurisdiction under Article [24(4)] of the regulation would rule in that regard, and will refuse to adopt the provisional measure sought if it considers that there is a reasonable, non-negligible possibility that the patent invoked would be declared invalid by the competent court.

[50] In those circumstances, it is apparent that there is no risk of conflicting decisions as mentioned in paragraph 47 above, since the provisional decision taken by the court before which the interim proceedings have been brought will not in any way prejudice the decision to be taken on the substance by the court having jurisdiction under Article [24(4)]…. Thus, the reasons which led the Court to interpret widely the jurisdiction provided for in Article [24(4)]… do not require that, in a case such as that in the main proceedings, Article [35] of that regulation should be disapplied.”

What this case shows is that the mere fact that there is a whisper of invalidity does not automatically bring proceedings in a different member state to a juddering halt. If this approach were to be applied to article 31 (2) it would enable the Belgian court to decide, on a provisional basis, whether there was a non-negligible possibility that the UK court would declare the UK designation of the patents invalid. If it came to that conclusion, it would then have to decide to what extent that invalidated the exclusive jurisdiction agreement. Since the Recast Regulation applies in Belgium just as it does in the UK, the Belgian court is in as good a position as the UK court to decide the effect of article 25 (4).

This coincides with the policy underlying article 31 (2), explained in recital (22). The court first seised (in England) is required to stay its proceedings as soon as the designated court has been seised (in Belgium) and until such time as the latter court declares that it has no jurisdiction under the exclusive choice of court agreement.

I therefore agree with Ms Lane that it is for the court designated in the exclusive jurisdiction agreement (i.e. the Belgian court) to decide whether (and, if so, to what extent) it is deprived of its jurisdiction as a result of article 25 (4).

Ms Lane argued that the English court should not even consider the question whether article 25 (4) was engaged. It should leave that question entirely to the Belgian court. That seems to me to be an extreme and unjustified position. If the English court must at least consider (on a prima facie basis) whether there is an effective choice of court clause, why should it not also consider (on the same basis) whether the clause is

To repeat the relevant part of article 25 (4): an exclusive jurisdiction agreement has no legal force “if the courts whose jurisdiction [it purports] to exclude have exclusive jurisdiction by virtue of Article 24”. That leads to the question: what is the extent of the UK court’s exclusive jurisdiction under article 24?

It is clear from the jurisprudence of the CJEU and its predecessor that not every dispute about a patent falls within article 24. Thus, a dispute about ownership of a patent; or even an action for infringement (where the validity of the patent is not in issue) falls outside article 24: GAT at [16]. Likewise, the CJEU has held that where an action consists of a claim that someone has been wrongly registered as the proprietor of a trade mark, the action does not concern “the registration or validity” of the mark: (Case C-341/16) Hanssen Beleggingen BV v Prast-Knipping [2018] Bus LR 1384.

GAT was a case in which GAT, a German company, offered a motor vehicle manufacturer also established in Germany a vehicle part for sale. It was alleged that that offer infringed two French patents of which LUK, a third German company, was the proprietor. How GAT’s commercial activities in Germany infringed a French patent is not explained. GAT brought an action in the German court seeking (a) a declaration of non-infringement and (b) a determination that the French patents were void or invalid. It appears that the German court had jurisdiction (a) because of the defendant’s domicile and (b) because Germany was where the harmful event took place. The contest, then, was between non-exclusive jurisdiction on the one hand; and exclusive jurisdiction under article 24 (4) on the other.

In the course of his opinion Advocate-General Geelhoed noted that the English language version of the predecessor to article 24 (4) used the phrase “proceedings concerned with” rather than “proceedings which have as their object.” None of the other language versions made that distinction; and the Advocate-General, in agreement with the Commission, thought that that was a distinction without a difference. The Court appears to have agreed with that, albeit obliquely, at [19].

The CJEU endorsed that position in (Case C-144/10) Berliner Verkehrsbetriebe v JP Morgan Chase Bank NA [2011] 1 WLR 2087 (“BVG”) at [26] to [28]. That case concerned a swap contract between a German transport undertaking and an American Bank. The various contracts contained an exclusive jurisdiction clause conferring jurisdiction on the English court. The bank sued in England for payment of amounts due under the contracts. It was met by a number of defences, including allegations of misrepresentation, non-disclosure and breach of a consultancy contract. By way of further defence BVG alleged that the subject matter of the contracts was ultra vires its own statutes. It then raised the question of vires in separate proceedings in Germany. Thus there were parallel proceedings both in England and in Germany. Both the German Court of Appeal and the UK Supreme Court referred questions to the CJEU (although the Supreme Court later withdrew its reference: see [2012] QB 176 at 209 (note)).

The issue before the CJEU was whether the fact that BVG had raised the question of the validity of its decision in Germany meant that the German court had exclusive jurisdiction under what is now article 24 (2). At [32] the CJEU reaffirmed the principle that even though article 24 had primacy, that did not justify a broad interpretation of its provisions; not least because it would deny the parties to a contract all autonomy to choose another forum. Moreover, it held at [33] that a broad interpretation would be contrary, first, to the general aims of the regulation, namely to seek to attain rules of jurisdiction that are highly predicable, and second, to legal certainty.

At [38] it held:

“However, in a dispute of a contractual nature, questions relating to the contract's validity, interpretation or enforceability are at the heart of the dispute and form its subject matter. Any question concerning the validity of the decision to conclude the contract, taken previously by the organs of one of the companies party to it, must be considered ancillary. While it may form part of the analysis required to be carried out in that regard, it nevertheless does not constitute the sole, or even the principal, subject of the analysis.” (Emphasis added) 86. It expressed its conclusion at [44] as follows:

“Thus, the divergence noted in para 26 of the present judgment between the language versions of article 22(2) of Regulation No 44/2001 is to be resolved by interpreting that provision as covering only proceedings whose principal subject matter comprises the validity of the constitution, the nullity or the dissolution of the company, legal person or association or the validity of the decisions of its organs.” (Emphasis added)

The court then considered the GAT case. As to that it said at [46]:

“That case law cannot be applied to proceedings in which a question concerning the validity of a decision of a company's organs is raised. Since the validity of the patent concerned is an essential premiss, in particular in any infringement action, it is in the interests of the sound administration of justice that exclusive jurisdiction to adjudicate upon any dispute in which the patent's validity is contested is accorded to the courts of the member state in which deposit or registration of the patent has been applied for or has taken place, they being best placed to adjudicate upon the dispute. As has been pointed out in paras 37 to 39 of the present judgment, that is not so in the case of the courts where a company party to a contractual dispute has its seat if it pleads that the decision to enter into the contract taken by its own organs is invalid.”

The exclusive choice of court clause therefore prevailed over the exclusive jurisdiction under article 24 (2).

Mr Turner QC placed some reliance on the decision of the CJEU in (Case- C-438/12) Weber v Weber [2015] Ch 140. The case concerned the effectiveness of a right of preemption over real property in Munich. The nature of the dispute fell within the German court’s exclusive jurisdiction under article 24 (1). But before an action was begun in Germany, one of the parties began proceedings in Italy, challenging the effectiveness of the right. It was, therefore, a classic case of an Italian torpedo. The question for the CJEU was whether the German court had to stay its proceedings under article 29 until such time as the Italian court had determined its own jurisdiction. The CJEU held that it did not. The reason for the decision was that the German court had exclusive jurisdiction over the dispute and that a decision of the Italian court on the substance of the dispute would not be recognised in Germany. Accordingly, the German court, although second seised, was not entitled to decline jurisdiction and was therefore obliged to rule on the substance of the dispute. The key point, however, is that it was a contest between the court first seised which had a nonexclusive jurisdiction (Italy); and the court second seised which had an exclusive jurisdiction (Germany). It does not bear on a case in which both courts have exclusive jurisdiction, albeit by different routes.

At this point it is necessary to return to some of the domestic cases. As mentioned, BVG had been litigated in the domestic courts. In the Court of Appeal ([2010] EWCA Civ 390, [2012] QB 176) Aikens LJ comprehensively reviewed the then state of the law. In the course of his review he considered two earlier cases (Coin Controls Ltd v Suzo International (UK) Ltd [1999] Ch 33 and Fort Dodge Animal Health Ltd v Akzo Nobel NV [1998] FSR 222). In Coin Controls the claimant in England sued for infringement of (amongst others) German and Spanish designations of European patents. The jurisdiction of the English court was non-exclusive. As far as I can tell from the report, the only prospective defence was that the patents were invalid. Laddie J held that once invalidity was raised as a defence, both claim and defence, in so far as they concerned a particular designation of a European patent, fell within the jurisdiction of the court where the patent in question was registered. He therefore declined jurisdiction over the allegations of infringement of the German and Spanish designations. In Fort Dodge a patentee began proceedings in the Netherlands alleging infringement of Dutch and UK designations of the same European patent. It applied to the Dutch court for the equivalent of interim injunctions restraining infringement in both countries. The alleged infringer then applied in England to revoke the patent for invalidity; and sought an anti-suit injunction restraining further prosecution of the Dutch proceedings in relation to infringements of the UK designated patent. Both Laddie J and the Court of Appeal refused the injunction. Although both considered that, because the patent was alleged to be invalid, the question of infringement of the UK designation fell within the exclusive jurisdiction of the UK courts, they nevertheless considered that it was not acte clair that the Dutch court could not grant interim relief. They therefore referred questions to the CJEU. We now know, as a result of Solvay,that the Dutch court did have that jurisdiction, despite the allegation of invalidity.

As I have said, one of the issues raised in BVG was whether BVG’s action in entering into the swap contract was ultra vires. Aikens LJ said at [84]:

“I reject the submission … that the opening phrase of article [24(2)] must be read in such a way that if proceedings raise any issue within article [24(2)], ie, if the proceedings are concerned with the validity of the constitution of a company (etc) or the validity of the decisions of its organs (etc), then that is sufficient to make those proceedings “principally concerned with” that issue. … The validity issue may be one of many other issues which have nothing to do with the validity of the company or the validity of decisions of its organs, and those other issues may have to be decided by a different applicable law and may concern facts which are unrelated to the state where the company has its seat. In such a case the sound administration of justice could well require the courts of another state to determine the issues. That is even more so when the parties have agreed a jurisdiction for the resolution of disputes.”

At [87] he added:

“… given my view that the Court of Appeal …did endorse the approach of Mance J that a court has to undertake an exercise in “overall classification” and make an “overall judgment” to see whether the proceedings are “principally concerned” with one of the matters set out in article [24(2)], we are bound to follow that interpretation unless there has been a subsequent decision of the ECJ (or House of Lords) which has stated a contrary interpretation. There is none.”

In Koza the claimant was an English company; and part of a group of companies ultimately controlled in Turkey. The Turkish government replaced the boards of other companies in the group with court appointed trustees. One of the companies was KA, which was the immediate parent company of the claimant. The trustees served notice under the English Companies Act on behalf of KA requiring the requisitioning of a meeting of the claimant to pass resolutions replacing the directors. The claimant began proceedings claiming a declaration that the notices were invalid; and a declaration that the English court should not recognise the authority of the trustees to cause KA to do anything as a shareholder of an English company. In the Court of Appeal ([2017] EWCA Civ 1609, [2018] 1 BCLC 591) it was held that the English court had jurisdiction over the whole action. Floyd LJ said at [29]:

“In order to answer the question whether proceedings which raise multiple issues are “principally concerned “with one of the matters within art.24(2) the court has to undertake “an exercise in “overall classification” and make an “overall judgment” to see whether the proceedings are “principally concerned” with one of the matters set out in” the article: JP Morgan Chase at [87].”

At [44] he said that the flaw in the trustees’ argument was to treat the two issues as severable, and to treat the argument about the trustees’ authority as a free-standing claim. He continued at [46]:

“I do not take from the English or European authorities which were cited to us any suggestion that one is required in all cases to disentangle issues which are interlinked in this way and apply art.24(2) to each issue separately. On the contrary, faced with such proceedings, the court is required to form an overall evaluative judgment as to what the proceedings are principally concerned with. The position is obviously different from a case where two quite independent claims are made in the same proceedings. Exclusive jurisdiction in relation to each claim would, in those circumstances, have to be determined separately.”

The case then proceeded to the Supreme Court, where Lord Sales gave the only judgment ([2019] UKSC 40, [2019] 1 WLR 4830). At [28] Lord Sales contrasted the exclusive jurisdiction conferred by article 24 with the general rule of jurisdiction based on domicile. He continued:

“Under article 4 and those provisions, it is quite possible that the courts of two or more member states might have jurisdiction in relation to the same claim. This causes no difficulty under the scheme of the Recast Regulation. In all such cases it is the priority rules in Section 9 of Chapter II which determine the jurisdiction where the claim should proceed, which generally depends on which court is first seised. But as noted above, those rules are disapplied where a claim falls within the exclusive jurisdiction provision in article 24. Accordingly, it is clear from the scheme of the Regulation that the interpretation and application of that provision cannot depend on the type of evaluative judgment in relation to which different courts could reasonably take different views. In principle, there should be only one correct application of article 24 in relation to a given claim. This tells strongly against the broad evaluative approach to the interpretation and application of article 24(2) adopted by the courts below.”

At [33], having considered the BVG case in the CJEU, he referred specifically to what the Court of Appeal had said in Koza at [46] of its judgment. He said of that paragraph:

“But this approach had the effect of expanding the application of article 24(2) …contrary to the guidance in the Hassett case and the BVG case, rather than narrowing its application, as the Court of Justice had been at pains to do in its judgments in those cases. According to the Court of Appeal, article 24(2) of the Recast Regulation is to be read as having the effect of allowing a party which is able to bring one claim within that article (the English company law claim) to add on another claim (the authority claim) which is conceptually distinct and is not inextricably bound up with the former claim, so that the latter claim is to be taken to fall within the scope of article 24(2) as regards the jurisdiction of the English courts as well.

In my view, Mr Crow QC for Koza Altin and the trustees was right to criticise this step in the Court of Appeal's analysis as an illegitimate reversal of the approach indicated in the judgment of the Court of Justice in the BVG case.”

The Supreme Court thus concluded that although the English Court had jurisdiction over the Companies Act point, it had no jurisdiction under the Recast Regulation to adjudicate on the authority of the Turkish trustees.

Koza establishes three points that are relevant to this appeal:

I think that it must also follow from the Supreme Court’s rejection of the “overall evaluation” test as applied by the Court of Appeal, that the test proposed by BVG in the Court of Appeal, which the Court of Appeal followed in Koza, no longer represents the law. On the other hand, the Supreme Court, as it had to, followed BVG in the CJEU where that court said that the question of jurisdiction turned on whether the “principal subject matter” of the action concerned the validity of the German company’s actions. The explanation for the divergence seems to be that where the Court of Appeal in Koza (and necessarily in BVG) went wrong was in conducting an overall evaluation of the proceedings, rather than an evaluation of the particular claim.

The judge gave judgment in our case on 29 March 2019; after Koza had been argued in the Supreme Court, but before its decision. Not surprisingly he followed the approach of Aikens LJ in BVG. First, he applied the “overall evaluation” test at [41]. Second, he held at [43] that, following BVG, severance was “not an option”. In the light of the decision of the Supreme Court in Koza, both those aspects of the judgment can now be seen to be wrong.

If, as we are told, article 24 is to be narrowly interpreted, it is confined to an allegation (which has not yet been formally made) that the patents are invalid.

But there are, in addition, a number of other defences to the claim which do not involve attacking the validity of the patents. They are, in summary:

None of these potential defences engages article 24 (4). None of them seems to me to be inextricably linked with an allegation that the patents are invalid. If, therefore, the test in BVG is applied I do not think that it can be said that the validity of the patent is the “sole or even the principal” subject matter of the dispute. On the other hand, the court in BVG saw no conflict between that test, and its previous decision in GAT. It is therefore necessary to examine more closely exactly what GAT decided.

In GAT the German court referred a question to the CJEU which, as usual, the CJEU reformulated. The question reformulated at [13] was:

“It asks whether that rule concerns all proceedings concerned with the registration or validity of a patent, irrespective of whether the question is raised by way of an action or a plea in objection, or whether its application is limited solely to those cases in which the question of a patent's registration or validity is raised by way of an action.”

Thus the question addressed by the CJEU was, in essence, the procedural question whether the allegation of invalidity had to be raised in the originating process to fall within article 24 (4); or whether it was sufficient for it to be raised by way of defence. It answered that question at [25]:

“… the view must be taken that the exclusive jurisdiction provided for by that provision should apply whatever the form of proceedings in which the issue of a patent's validity is raised, be it by way of an action or a plea in objection, at the time the case is brought or at a later stage in the proceedings.”

It is instructive to consider the reasons that led the court to that conclusion:

“[26] First, to allow a court seised of an action for infringement or for a declaration that there has been no infringement to establish, indirectly, the invalidity of the patent at issue would undermine the binding nature of the rule of jurisdiction laid down in [article 24 (4)].

[27]

While the parties cannot rely on [article 24 (4)], the claimant would be able, simply by the way it formulates its claims, to circumvent the mandatory nature of the rule of jurisdiction laid down in that article.

[28]

Secondly, the possibility which this offers of circumventing [article 24 (4)] would have the effect of multiplying the heads of jurisdiction and would be liable to undermine the predictability of the rules of jurisdiction laid down by the [Regulation], and consequently to undermine the principle of legal certainty, which is the basis of the

[Regulation] …

[29]

Thirdly, to allow, within the scheme of the [Regulation], decisions in which courts other than those of a state in which a particular patent is issued rule indirectly on the validity of that patent would also multiply the risk of conflicting decisions which the [Regulation] seeks specifically to avoid… ”

These reasons no doubt lead to the conclusion that the Belgian court does not have jurisdiction to decide definitively the validity of the UK designation of the European patents (although it would have jurisdiction to decide the validity of the Benelux designations of the same patents). But it does not follow either from the narrow answer to the question asked and answered, or from the reasons given for that answer that, to adapt Lord Sales’ phrase in Koza, the whole of the action is pulled into the exclusive jurisdiction of the UK court. It is also to be noted that in GAT there was no contractual choice of court clause in play (as there was in BVG); and that GAT was decided before the appearance of article 31 (2) in the Regulation, which was deliberately designed to enhance party autonomy. As far as one can tell from the report the only defence to the infringement claim in GAT was the allegation of invalidity. Thus, as Aikens LJ observed in BVG at [73], distinguishing GAT:

“… in the present case there are also other issues which have no equivalent in the patent case that raises only the issues of validity and infringement.” 108.He returned to the point at [85]:

“I reject the submission … that if proceedings raise a number of issues and one of them is within the terms of article 22(2) and the resolution of that issue may be dispositive of the proceedings as a whole, that must mean that the proceedings are “principally concerned with” an issue within article 22(2) . Again, that is not what the article states. It is inconsistent with Mr Jenard's commentary. If the test is: what are the proceedings “principally concerned with”, as the English Court of Appeal has held, then one issue which may be dispositive may be what the proceedings are principally concerned with, but it is not necessarily the case. No ECJ decision has given the article that interpretation. The GAT case does not support this proposition because that was a patent case and, as the ECJ makes clear in its judgment, when an infringement claim raises the issue of validity of the patent they are really two aspects of one issue. The Coin Controls and Fort Dodge cases are to the same effect.”

Moreover, as Solvay shows, there is nothing objectionable about a Belgian court expressing provisional views about a UK designated patent.

In my judgment, there is at least an arguable case that in an action raising multiple issues, including the scope of an express licence to exploit the patented invention, article 24 (4) does not pull all the issues into the exclusive jurisdiction of the English court in the face of an exclusive choice of court clause. It is also arguable that article 25 (4) does not invalidate the choice of court clause in so far as it relates to disputes falling within its scope. It follows, in my judgment, that those questions are for the Belgian court to decide.

Accordingly, in my judgment article 31 (2) requires this court to stay “the proceedings.” I take that to mean the whole of the proceedings. Since the scope of the licence is an integral part of the action as whole, it makes sense in case management terms to stay the whole of the action until such time as the Belgian court (a) decides whether it has jurisdiction and (b) if it decides that it has, rules on the scope of the licence.

I would allow the appeal; and stay the action until the Belgian court has made its ruling.

I agree.

I also agree.