Eli Lilly & Co & Ors v Genentech Inc [2017] EWHC 3104 (Pat)

Case No HP-2017-000041[2017] EWHC 3104 (Pat)IN THE HIGH COURT OF JUSTICEBUSINESS AND PROPERTY COURTS OF ENGLAND AND WALESINTELLECTUAL PROPERTY LISTPATENTS COURTVenue Royal Courts of Justice, Rolls Building, Fetter Lane, London, EC4A 1NLDate 06/12/2017
Before
THE HON. MR JUSTICE BIRSS
Between
(1) Eli Lilly & Co.Claimants(2) Lilly France SAS.Claimant(3) Eli Lilly Deutschland GmbHClaimant(4) Eli Lilly Italia SpAClaimant(5) Eli Lilly and Co (Ireland) LimitedClaimant(6) Eli Lilly Kinsale LimitedClaimant(7) Lilly SAClaimant(8) Eli Lilly and Company LimitedClaimantGenentech Inc.Defendant
Alexander Layton QC, Thomas Hinchliffe QC and Oliver Caplin (instructed by Allen & Overy LLP) for ClaimantsDavid Lewis QC, Michael Ashcroft QC and William Duncan (instructed by Marks & Clerk Solicitors LLP) for DefendantHearing Hearing dates: 6th and 7th November 2017
JudgmentMr Justice Birss :
[1]This is a jurisdiction application. The action concerns various national designations of European patent EP 1 641 822, entitled “IL-17A/F Heterologous Polypeptides and Therapeutic Uses Thereof”. Recognising that there is no such thing as a European patent but rather a bundle of national patents I will refer to the whole bundle as EP 822. The claimants are all members of the Eli Lilly group of companies. The group is a major international pharmaceutical company. I will refer to the claimants as Lilly. The defendant, Genentech, is another major international pharmaceutical company. The patent belongs to Genentech and that company, which is a US corporation and therefore domiciled outside the area of the Brussels 1 (recast) Regulation or Lugano conventions, is the proprietor of all the relevant national designations of EP 822.[2]Lilly makes and sells a medicinal product marketed under the brand name Taltz with the non-proprietary name ixekizumab. The product is a monoclonal antibody and has a centralised European marketing authorisation to treat a form of the disease psoriasis. It is available on the NHS in England and NICE has recommended it. Lilly describes the product as an anti-IL17 antibody. IL17 or interleukin 17 is a human cytokine. Cytokines such as the interleukins are proteins which play important roles in the immune system. IL17 may play a role in stimulating the production of IL6 and IL8 in certain circumstances. The interleukins can exist in the form of a homodimer in which two identical molecules are bonded together.[3]EP 822 refers to IL17 as IL17A and acknowledges that it is a known cytokine. It also acknowledges that there are a number of known cytokines in the IL17 family (IL17A, IL17B, IL17C etc.). The invention described in EP 822 is based on the discovery of a new cytokine which the patent calls IL17A/F. The patent refers to this as a heterodimer, because it consists of an IL17A molecule bound to an IL17F molecule. Using this notation one could refer to the IL17 homodimer as IL17A/A.[4]There are opposition proceedings before the EPO in which a number of companies have challenged the validity of EP 822. Lilly is not a party to those proceedings. The Opposition Division has decided that EP 822 is invalid because the main request and three auxiliary requests involve added matter contrary to Art 123(2) EPC. The decision is now under appeal to the Technical Board of Appeal of the EPO. Stripped of its verbiage, claim 1 of the main request essentially claims an antibody which “specifically binds” to IL17A/F.[5]According to Lilly, Genentech does not have an IL17 product on the market or any product for psoriasis on the market or under development. Nevertheless Genentech has applied for a supplementary protection certificate under Regulation 469/2009 in various EU states (the UK, France, Germany, Spain, Italy and Ireland). The SPC application is based on Genentech’s patent EP 822 and on Lilly’s marketing authorisation for Taltz. Necessarily therefore Genentech’s position must be that Taltz is a product protected by EP 822. Lilly disputes that Genentech can acquire such an SPC in that way but that is not germane to the issues I have to decide on this application.[6]Lilly does not accept that Taltz infringes EP 822 and so has brought these proceedings in the Patents Court in London. The claims are in these proceedings can be seen in two parts. First Lilly brings claims relating to the UK designation of EP 822. These are a claim for a declaration of non-infringement (a “DNI”) in relation to the Taltz product, as well as a claim for revocation of that UK patent on various grounds. The DNI actually seeks two declarations, one is in conventional form drafted as a declaration that committing various acts such as making or selling the Taltz product does not infringe EP (UK) 1 641 822, the other is in a different form – a declaration that the product is not within the scope of the claims of the patent. This first part of the proceedings does not cause any difficulty. The English court’s jurisdiction to hear and determine these claims is not in dispute.[7]The second part of the proceedings consists of claims for declarations of non-infringement (and scope) in relation to the five national designations of EP 822 in France, Germany, Spain, Italy and Ireland. These have been referred to as the “Related Claims”. No claim for revocation or invalidity is made about those five foreign designations.[8]Leave to serve the proceedings out of the jurisdiction was given by Master Price on an ex parte application by Lilly. On this application Genentech applies to set aside service of the proceedings insofar as they include the Related Claims or for a stay on jurisdiction grounds. Three grounds are advanced by Genentech. The parties advanced them in argument in the opposite order, no doubt because, as good advocates, each side wanted to start with its best point. The three grounds are as follows. The first is that the Related Claims cannot be brought within one of the gateways for service out of the jurisdiction and so service should be set aside under CPR r6.36/6.37. The gateways are specified in CPR PD6B. The only relevant gateway is that in paragraph 3.1 (4A) of PD6B and Genentech submit the Related Claims are not within that gateway. The second is that the proceedings are “concerned with” the validity of the non-UK patents and thereby engage Art 24(4) of the Brussels I (recast) Regulation 1215/2012 and so, since under that provision the courts of the Member State in which the patent is registered (France, Germany, Spain, Italy and Ireland) have exclusive jurisdiction over validity, this court must decline jurisdiction over this aspect of these proceedings under Art 27 of the regulation. The third ground is forum non-conveniens. The argument is that England is not the proper forum for these claims. I have ranked the claims in this way because the gateway point is entirely distinct from the other two and because there is an overlap between the arguments on Art 24(4) and forum non-conveniens and it is convenient to resolve the former before dealing with the latter. Although it does not matter, I believe Art 24(4) is Genentech’s best point, as was indicated by their approach to arguing the issues.[9]There was no dispute that on all three grounds the burden is on Lilly. It is not on Genentech.[10]There was also no dispute that where, as here, the defendant is domiciled outside the Brussels-Lugano area the jurisdiction over that defendant is determined by national law save for certain exceptions. The exceptions do matter in this case because one of them is Art 24 of the Regulation.

The CPR Gateway in PD 6B para 3.1(4A)

[11]The basic principles are common ground. Lilly must show that it has a properly arguable case on the merits of the relevant claim and it must also show that it has much the better argument that the gateway is applicable. “Much the better argument” is a higher hurdle than “properly arguable”.[12]Genentech submitted that where there is a doubt about the construction of the gateway provision, that doubt should be resolved in favour of the foreign defendant, i.e. Genentech. Lilly did not agree and submitted that the decision of the Supreme Court in Abela v Baadarani [2013] UKSC 44 in which Lord Sumption held (at paragraph 53) that the traditional characterisation of service out of the jurisdiction using “muscular presumptions” and language such as “exorbitant” was no longer correct because the introduction into English law of the distinct forum non-conveniens doctrine in Spiliada v Cansulex [1987] AC 460 provided a proper and pragmatic safeguard for foreign defendants. Therefore Lilly argues that the previous line of authority starting from In re Hagen [1908] P 189, which is the basis for Genentech’s submission, is no longer good law. I must say I see the force in Lilly’s submission however in Cruz City 1 Mauritius Holdings v Unitech Ltd and Ors[2014] EWHC 3704 (Comm) Males J recently rejected that very same submission at paragraph 16 and since I am not convinced the point is central to the issues I have to decide I will not get into it any further.[13]Turning to the merits of the cause of action itself, the point is this. Lilly contends that the debate turns on the true construction of the claim and in particular the words “specifically binds”. Lilly’s case is that as a matter of fact Lilly’s antibody has a binding affinity to the IL17A/A homodimer of less than 3 pMol (pico moles) and also has a binding affinity of less than 3 pMol (pico moles) to IL 17A/F.[14]Lilly contends that the patent explains what “specifically binds” means and in summary the point is that the claimed antibody must not only bind to the target but must do so specifically, in other words it must not bind to other things. Put another way the claimed antibody must not cross-react with other proteins. Some small degree of cross-reactivity would be acceptable but it must be low, for example a figure of 10% of the level of target binding is mentioned. There is no question that IL17A/A is not the target of the invention. The sequence referred to in the claim is for IL17A/F and in fact IL17A/A is acknowledged as a known protein. So Lilly argues that the fact that ixekizumab binds to both IL17A/A and IL17A/F with such a high affinity (low pMol) means it is not within the scope of protection provided by the claim and does not infringe.[15]Genentech accepts that Lilly has a properly arguable case. I agree.[16]Although not relevant to the gateway it is convenient to take the merits one step further at this stage just to see the whole picture. Lilly has invalidity arguments which are independent of this claim construction issue but a feature of this case is that Lilly contends that if the claim is construed in such a way that it does cover ixekizumab, then it must be invalid for a number of reasons, including obviousness and insufficiency. So there is what is often referred to as a squeeze, or so called Gillette defence. I say “so called” because in English law there is not a distinct defence based on the famous Gillette principle. Rather it is a label given to kinds of argument. In any event the point is that it is not in dispute that certain grounds on which the UK patent is alleged to be invalid only apply on a claim construction which covers ixekizumab and fall away if the claim does not. This relationship between infringement and validity is a key aspect of the Art 24(4) debate.[17]Finally on the merits, I raised with Lilly the recent decision on equivalents in the Supreme Court (Actavis v Lilly [2017] UKSC 48). Lilly submitted it did not make any difference on this application and Genentech did not dissent from that. I will return briefly to this below.[18]Lilly makes ixekizumab in Ireland, carries out QC testing in Italy and sells the product in all the states covered by the Related Claims. The claimants in these proceedings include the relevant operating companies for the various jurisdictions of the Related Claims. The argument of non-infringement based on claim construction applies to the Related Claims as much to the claim against the UK patent I find that Lilly has established that it has a properly arguable case for the declarations of non-infringement (and scope) in all the Related Claims.[19]Turning to the gateway, paragraph 3.1 (4A) of PD 6B provides as follows:
“3.1 The claimant may serve the claim form out of the jurisdiction, with the permission of the court under r6.36 where-… (4A) A claim is made against the defendant in reliance on one or more of paragraphs (2), (6) to (16), (19) or (21) and a further claim is made against the same defendant which arises out of the same or closely connected facts.”
[20]Lilly submits this gateway applies in the following way. The action includes a claim which is made against Genentech for a DNI in relation to the UK patent (the UK designation of EP 822). There is no dispute this UK claim falls within relevant paragraphs of paragraph 3.1 of PD 6B i.e. (9) (tort) and (11) (property within the jurisdiction). The Related Claims are made against the same defendant, Genentech. The submission is that the Related Claims arise out of the same or closely connected facts as the UK DNI. Those facts include, at least, the characteristics of the antibody product itself. They include the acts the subject of the DNIs, such as sale and manufacture of the product. Lilly says that in a DNI those are a hypothetical set of facts on which the declarations are sought and are the same hypothetical facts for the UK patent and all the Related Claims. The facts may also include at least some of the facts relevant to claim construction (such as common general knowledge) although neither party got into that. So Lilly submits the gateway applies on the plain meaning of the words.[21]Genentech does not agree. It contends that the Related Claims do not involve the same or closely connected facts. That is for a number of reasons. First the facts. Each claim involves a distinct patent albeit they arise from the same bundle of rights granted by the EPO. They are territorially and legally distinct. Analytically the DNI claims have different claimants because the true claimant is the operating company in the relevant state. Each Related Claim will involve the application of foreign law, namely the local law of infringement, particularly about the determination of claim scope. Genentech submit that there are differences between these local laws despite their common origin in the EPC. Since under domestic law, foreign law is a question of fact, each Related Claim will involve facts which are unique to it and are not the same or closely connected to the UK DNI.[22]Genentech also suggested that there might be further facts relating to infringing acts which may differ in each jurisdiction but I was not persuaded that was a significant point. For one thing the declarations are all formulated on the same, hypothetical factual basis. For the same reason the fact that there are multiple Lilly companies as claimants does not assist Genentech here. A DNI could be brought by a single group company. Having the local operating companies as co-claimants is appropriate because it ensures the right entities are bound by the result and that works for the benefit of both sides. A final reason why I do not believe the infringing acts point assists Genentech is because it seems to be premised on the argument that the laws of infringement in different states were quite distinct, I think because the EU decisions in this area recognise that infringement is a matter of national law. So it is, and it is not (much) governed by the EPC, but in practice the infringement laws of the various states are closely aligned for a number of reasons including the Community Patent Convention (CPC).[23]However Genentech is clearly correct in its submission that each Related Claim will involve a legally distinct patent and a decision on foreign law which is a question of fact in this court.[24]Genentech’s next point is that the right way to look at gateway (4A) is to consider it alongside the other gateways concerned with joinder of parties and also with the joinder of parties under EU law in Art 8(1) of Brussels I (recast). This is the necessary and proper party test. That is said to be the right way to apply the gateway or at least arguably so such that the gateway is ambiguous and should be construed in Genentech’s favour that way. Applying that approach, says Genentech, the Related Claims fall outside the gateway.[25]Gateway (4A) is a relatively new provision which was introduced as part of a suite of amendments to PD 6B in 2015. Lilly sought to rely on certain documents which went to the Civil Procedure Rules Committee when this proposal was accepted as relevant to construction. They were a report of the Lord Chancellor’s advisory committee on private international law (the Mance Committee) and a final report of Chancery working group chaired by Henderson J. What these documents do is show that the proposal was supported (perhaps not surprisingly since it was implemented) and that alternative formulations of the words were discussed. I do not find this material of significant assistance. The exercise of looking at preparatory materials of this kind could help identify what the purpose of a provision was and that in turn would be relevant to construction but this material does not do that for this gateway.[26]Genentech submitted that the considerations explained by Lord Collins in Altimo v Kyrgyz [2011] UKPC 7 at paragraph 87 in relation to the question of bringing a further foreign party into the proceedings under the necessary and proper party gateway, should be applied to gateway (4A). Lord Collins asked if, had the two defendants been within the jurisdiction, would they both have been proper parties and explained that this would be so if the claims against them both involved the “same investigation”, referring also to the matters being “closely bound up” and to a “common thread”.[27]This is related to Genentech’s submission that guidance can be had from Art 8(1) of the Brussels I (recast). Bearing in mind that the scheme of the regulation is that, subject to exceptions, defendants should be sued in the state in which they are domiciled, Art 8(1) is the provision which allows defendants domiciled in one state to be sued in another state, when they are one of a number of defendants, as long as one of the other defendants is domiciled in that other state. The regulation refers to claims which are “so closely connected that it is expedient to hear and determine them together to avoid a risk of irreconcilable judgments”. As Genentech point out in Roche v Primus (Case C-539/03) the ECJ held that different companies in the same corporate group in various European states could not be sued in one proceeding using the different designations of a European patent because the factual situation was different. Such defendants could only be joined when the proceedings are so closely connected that it is expedient to hear and determine then together to avoid the risk of irreconcilable judgments resulting from separate proceedings. There was no such risk in that case because for the CJEU a different outcome on two patents both of which are designations of the same EP is not what that court regarded as an irreconcilable judgment.[28]If the matter was free from authority I would not hold that the interpretation of gateway (4A) should be determined by the considerations about parties because those questions, both in English law and EU law, are all concerned with the circumstances in which a foreign party is to be brought before a jurisdiction they are not otherwise subject to. There are good reasons why a court in one state should take care not to be too ready to force a foreign party to litigate before it and that explains why a test of necessity (such as a need to avoid irreconcilable judgments) or propriety (necessary and proper party) are used in that context. However gateway (4A) is different in kind because it presupposes that the foreign defendant is already subject to the court’s jurisdiction. Since the defendant will be before the court anyway, there is no reason to read limitations into the words based on ideas of necessity or propriety.[29]The only judgment the parties have found on this gateway is the decision of Edis J in Eurasia Sports v Tsai [2016] EWHC 2207. Genentech relied on paragraph 55 of the judgment in which the judge considered gateway (4A) and said that it seemed likely to him that the words “same or closely connected facts” were likely to have a similar meaning to the “proper party” test in Altimo and that if one test is met, the other is likely met also. I entirely agree with Edis J that if the test in Altimo is satisfied then the gateway (4A) test is also likely to be met. The judge decided that that test was satisfied in the case before him and so that is the ratio of the decision. I do not read the judgment as going further than that and holding that the two tests have exactly the same scope but if and to the extent it does then I respectfully disagree. Gateway (4A) is not concerned with whether someone is a proper party and I believe it raises different considerations.[30]In my judgment the words as a whole and the phrase “the same or closely connected” contemplate that the facts relevant to the second claim need not be the same as the facts on which the first or “anchor” claim is based, such that differences can be tolerated as long as the facts are closely connected. The connection between the facts relevant to the second claim and the facts relevant to the anchor claim must be a close one, but I do not read the wording as ruling out a situation in which the second claim involves at least some facts which are unique to it and which are distinct from those in issue in the anchor claim. The question is - how close a connection will satisfy the test and what are the considerations which apply to that? This takes one to the purpose of the provision.[31]I have rejected the idea that the purpose of the provision can be derived from the rules on joinder of parties, because the gateway does not do that. Nor has gateway (4A) been enacted to fulfil any EU obligations or other treaty obligations. So what is its purpose? It often makes good practical sense to try claims together if they involve common issues of fact. If a defendant was within the jurisdiction, the most important reasons to handle what are legally distinct claims in the same action are practical considerations such as procedural economy and the avoidance of inconsistent results. By inconsistent results I include results which are formally irreconcilable in the narrow sense used by the CJEU in Roche v Primus but also on a wider less technical basis. A finding that a given product infringes the UK designation of an EP is inconsistent with a finding of non-infringement relating to the same product but in relation to the German designation. They are not formally irreconcilable but that does not make the outcome a desirable result. It is not. The CJEU took a narrow formal view of irreconcilability in part because Art 8 is an exception to the primary European rule that EU defendants are to be sued in their member state of domicile. That consideration does not apply to gateway (4A).[32]Gateway (4A) is certainly not as wide as the ambit of the court’s power to try distinct domestic cases together, where no jurisdictional question arises, but it seems to me that essentially the same practical considerations indicate what the provision is aimed at. Given that the defendant is already properly subject to the court’s jurisdiction, pragmatic factors are appropriate matters to take into account in deciding whether the connection between the facts is sufficiently close to justify service out having regard to the overall justice of the circumstances. The purpose of gateway (4A) is to allow the joinder of a further claim against the same defendant based on the same or closely connected facts so as to further the interests of justice, including taking into account practical considerations such as procedural economy and an avoidance of inconsistent results.[33]If facts unique to the second claim include disputed matters of primary fact and would require extensive cross-examination and disclosure resolve, then that would support a conclusion that the connection between the facts is not sufficiently close to justify service out. But if the key factual issues in the second claim are facts it has in common with the anchor claim and the difference in the relevant facts is only on matters which are undisputed or which can readily be dealt with without substantial disclosure and little or no cross-examination, then the factual connection may be sufficiently close to justify service out.[34]A specific point arises on the applicable law being foreign law and therefore a matter of fact. If the applicable law is the only difference between two sets of facts (because foreign law is a question of fact) then I am not sure I would accept this was a relevant difference between the facts, however I do not have to decide that issue because that is not the only difference in this case. I will assume without deciding that foreign applicable law counts in the relevant analysis.[35]In this case the anchor claim is a claim for a declaration of non-infringement by a product relating to a UK patent. The Related Claims are claims for the same kind of relief relating to the same product on the same assumed facts. It is true that the patents are legally distinct but they are all designations of the same European bundle patent as the UK patent, with the same claims and the same specification. Although the applicable law for each related claim is the national law of the relevant state (Rome II) those laws are all based on the same key legislative materials, i.e. Art 69 EPC and the Protocol. Furthermore they are all laws which the Patents Court and the courts hearing appeals from the Patents Court frequently have regard to in striving as best we can to come to a consistent approach to infringement across all EP states. Experience has shown that the Patents Court is well able to deal with that sort of issue.[36]So although it can be said that there will not be a single investigation, in the sense that each Related Claim is based on a legally different patent and involves (slightly) different applicable law, I find that the claimants have much the better argument that gateway (4A) is applicable to the Related Claims. I dismiss this part of Genentech’s application.

Art 24(4) of Brussels I (recast)

[37]The question is whether these proceedings are “concerned with” the validity of the non-UK patents and thereby engage Art 24(4) of the regulation. If they do then the court must decline jurisdiction. The parties are fundamentally at odds about how the question is to be answered. Lilly contends that as the proceedings are presently constituted, they are not concerned with validity because validity is not in issue in any formal sense either here or anywhere else that matters such as any of the home states of the non-UK patents. It is common ground that the EPO proceedings do not count for Art 24(4) purposes. Lilly submits that the fact squeeze arguments and suchlike are being deployed does not make any difference. Only if validity is formally put in issue either here, or in the state with exclusive jurisdiction, has what Mr Layton called the GAT trigger been pulled. Once the trigger is pulled then Art 24(4) is engaged but until then it cannot be. Genentech disagrees. Briefly put Genentech’s point is that one must look at the substance of what is being alleged and not only the form. Genentech point to judgments which say or can be read as saying exactly that. Lilly says that no other case has gone as far as Genentech’s submission needs to go for Genentech to be correct and that there is more to the cases than Genentech’s submission allows. In order to resolve this fundamental disagreement about the authorities I need to go through the cases with some care.

The law

[38]Article 24(4) provides that regardless of the domicile of the parties, the courts of a member state have exclusive jurisdiction in proceedings concerned with the registration or validity of patents registered in that state. The Jenard report makes clear that the provision exists because of sovereignty considerations concerning the grant of a monopoly in that state. That in itself has echoes of the reason why, in Germany, infringement is tried in the local courts while validity is tried in a different federal court.[39]On the other hand it is also clear that under the regulation infringement is not subject to this exclusive jurisdiction. Over the years various ingenious patent lawyers, mostly Dutch, have tried to find a way to do cross border pan-European patent litigation at least relating to infringement. After all the corresponding EP designations of a European patent, albeit they are legally distinct, are closely related and considerations of procedural economy suggest it would make sense. Moreover, having regard to the scheme of the regulation whereby the primary jurisdiction rule is that defendants should be sued in the state in which they are domiciled, one might have expected to see claims for infringement of a patent granted in one member state being litigated in the courts of another state in which the defendant is domiciled. However none of this has taken place very much, if at all. There are various reasons why this is the case but a critical one is the effect of Art 24(4) and its predecessors.[40]One of the key decisions of the CJEU is GAT v LUK (Case C-4/03) of 13th July 2006, but before getting there I need to look at the position prior to it. Since the original Brussels Convention came into force (of which the Regulation is the current successor), it was thought that in principle under EU law it might be possible to commence a single patent infringement proceeding against a defendant domiciled in a member state in which claims for infringement were made based on a number of EP designations. So for example a claimant might in theory sue a UK company in the UK for infringements of the UK, French and German designations of an EP patent. Under EU law the UK court could entertain such a claim. However there was little point in doing this because, assuming the defendant did not want to have all these claims decided in the same court, the defendant could simply counterclaim for revocation. That would trigger what is now Art 24(4) and Art 27. The claimant might try and argue that validity and infringement are distinct (after all the inclusion only of validity in Art 24(4) does not otherwise make sense) but it was settled law that in the circumstance described the UK court would have to cede jurisdiction of both validity and infringement to the French and German courts.[41]So in Coin Controls v Suzo [1999] Ch 33 at pp51-52 Laddie J held that before revocation proceedings were commenced, the UK could entertain the infringement claim; however the subsequent commencement of the revocation proceedings triggered what are now Art 24 and Art 27. The UK court was then obliged to decline jurisdiction in favour of the court with exclusive jurisdiction over validity. What is now Art 27 required the UK court to cede not only jurisdiction over the validity claim on the foreign patent but also the infringement claim because they are so closely interrelated they should be treated as a single claim. Moreover having infringement decided in one court when another court was deciding validity was an unhelpful proliferation of proceedings. Some of Laddie J’s reasoning involved emphasis on the word principally in the phrase “principally connected with” in what is now Art 27 and that reliance has later been shown to be misplaced because the word does not appear in other language texts, but nothing turns on that.[42]1. Note that Laddie J accurately observed that the UK court could hear a pure infringement claim (my words) citing Pearce v Ove Arup [1997] Ch 293. That was a significant observation at the time because another part of the explanation why one did not see cross-border patent cases in the UK was that traditionally the idea that the UK court should (even if it could) hear cases about foreign intellectual property rights was, at least, controversial. Arnold J made essentially the latter point in Actavis v Lilly [2012] EWHC 3316 (Pat) at paragraph 86 citing Aldous J in Plastus Creative v 3M [1995] RPC 438. However as Arnold J went on to show in the following paragraphs of Actavis, the judgment of the Supreme Court in Lucasfilm v Ainsworth [2011] UKSC 39 represented a significant shift in favour of the UK courts hearing cases of the infringement of foreign intellectual property rights including patents. The subject matter is justiciable and not only that, the reasons in Lucasfilm formed a key part of Arnold’s judgment in Actavis declining to grant a stay on forum non-conveniens grounds of claims for DNIs relating to French, Italian and Spanish patents which were EP designations of a UK patent in suit.[43]The reasoning in Coin Controls was endorsed in the Court of Appeal in Fort Dodge v Akzo [1998] FSR 222. Some of the reasoning in Fort Dodge has overtones of the pre-Lucasfilm view of the justiciability of foreign patent infringement claims but no matter. The reasoning in Coin Controls was also approved in Prudential v Prudential [2003] 1 WLR 2295 (CA), a trade mark case.[44]The problem in GAT v LUK was as follows. GAT brought an action in Germany against a German domiciled patentee for a DNI in relation to two French patents and for a declaration that the patents were invalid. The argument was that this did not violate Art 16(4) (the predecessor of Art 24(4)) because that only applied to cases in which invalidity was raised by way of action and did not apply if invalidity was raised in the way it had been. The CJEU decided that the Article applied in both circumstances. The court explained in paragraph 25 that:
“… the view must be taken that the exclusive jurisdiction provided for by that provision [now Art 24(4)] should apply whatever form of proceedings in which the issue of a patent’s validity is raised, be it by way of an action or plea in objection, at the time the case is brought or at a later stage.”
[45]The reference to a later stage is not to the possibility that validity might be put in issue in future but to the fact that validity was not part of the proceedings at the outset but only came in at a later stage. Since GAT v LUK the terms of what is now Art 24(4) were amended to include an express reference to the issue being raised “as a defence”.[46]A debate before me was about the characterisation of the court’s interpretation of Art 24(4) as “wide”. The answer is as follows. There were three interpretations before the court in GAT and the court chose the middle one, rejected a wider one and a narrower one. At paragraphs 16-21 and 36-47 of his opinion the Advocate General in GAT notes that the German government proposed a narrow interpretation of Art 24(4) in which the validity challenge only mattered if it was the “main claim” and the French and British governments proposed a wide interpretation on the footing that questions of validity and infringement cannot be disassociated. That wide interpretation would apply what is now Art 24(4) to infringement actions in the interests of legal certainty (see AG para 19). That wide interpretation also has echoes of some of the widest statements in Coin Controls but not the core reasoning nor the result in that case. The Commission proposed a “half way hypothesis” (AG para 20) in which Art 24(4) applied to validity however it was raised, whereas other questions about a patent were outside the article. The half way hypothesis was the one supported by the AG (AG para 37). It would mean that what the AG calls a “pure breach” case would not engage Art 24(4) (see AG para 43). This half way hypothesis is the approach which was adopted by the court. At paragraph 16 of the court’s judgment, although the term “pure breach” is not used, the court there expressly holds that an infringement action in which validity is not called into question would not engage the article. That disposes of the very wide approach of the French and British governments. The German government’s approach is expressly rejected at paragraph 30.[47]On the same day judgment was delivered in GAT v LUK, the CJEU gave also gave judgment in Roche v Primus which I have mentioned already. That judgment put paid to another aspect of attempts to conduct cross-border patent litigation in Europe, the pulling together into one action in Holland of different companies in a group and therefore different EP designations.[48]GAT v LUK was applied in the UK in Knorr-Bremse v Haldex [2008] EWHC 156 (Pat). The claimant KBS and the defendant Haldex had a dispute about whether certain KBS products infringed German patents of Haldex. The proceedings settled by agreement. The agreement had an exclusive jurisdiction clause in favour of the Düsseldorf court. KBS developed a new product and brought a DNI in the UK in relation to two UK patents held by Haldex which were designation of EPs which were counterparts of the German patents. Haldex applied for a stay in favour of the Düsseldorf court. Lewison J decided that the relevant KBS company which was the claimant was not bound by the settlement. Since they had been argued he went on obiter to address other issues which would have mattered if it had been bound. One was whether the UK court had exclusive jurisdiction under what is now Art 24(4). At paragraph 40 the judge noted that “the pleaded case contained in the Particulars of Claim claims only a declaration of non-infringement. It does not assert that either patent is invalid” however he went on to record that KBS had made clear in evidence that if the patentee advanced a construction which was wide enough to cover KBS’s product then KBS would “have no option but to put the validity of the patents in issue”. At the hearing counsel offered an undertaking by KBS either to counterclaim for invalidity or to amend its claim form to claim that the patents were invalid.[49]Haldex submitted that there was in fact no plea of invalidity in the proceedings and it would be wrong to treat the case as if there were. Lewison J referred to GAT and to the approved passage in Coin Controls and, following Laddie J, held that it was not necessary for validity to have been formally pleaded as long as it was clear that validity “is to be” put in issue. Since it was clear that KBS were going to raise the issue, the court had exclusive jurisdiction under what is now Art 24(4).[50]In 2012 the CJEU decided Solvay v Honeywell (Case C-616/10). That judgment dealt with whether raising validity as a defence not in a main claim but in a claim for a provisional measure engaged what is now Art 24(4) and Art 27. The court held it did not. In its judgment the CJEU referred to the court in GAT giving the article a “wide” interpretation. I have explained GAT above.[51]Then came Actavis v Lilly. In this case Actavis brought DNIs about the same product relating to the UK patent and some other corresponding EP designations. The claims were brought against the patentee which was a US corporation. In his judgment on jurisdiction (cited above) Arnold J held that the proceedings were properly served within this jurisdiction either because the patentee consented to service or because they were served under CPR r6.9(2). In the latter case no stay of the non-UK designations on forum non-conveniens grounds should be made.[52]As far as validity is concerned note that like GAT v LUK and Knorr-Bremse, in Actavis v Lilly the patentee was the defendant not the claimant. And also unlike GAT v LUK the claimant was not challenging validity in the UK proceedings. There was no claim for revocation along with the claim for a DNI. Actavis also offered an undertaking, which was accepted by the court, not to challenge validity in those proceedings. Until I heard this application I had not needed to examine Actavis v Lilly closely and I had misread the judgment as recording that Actavis had undertaken not to challenge validity anywhere (i.e. either before the UK court or before the relevant foreign court or anywhere else) but that is not accurate. Counsel for Genentech sought to persuade me that the effect of the undertakings recorded in the order actually made in Actavis v Lilly meant that Actavis’s undertaking had that wider effect. I do not agree. Paragraph 28 of the judgment makes clear that Actavis’ undertaking was not to challenge validity or contend the patent is invalid either by way of claim or defence to any counterclaim for infringement “in these proceedings”, i.e. the UK proceedings.[53]The judge recorded in paragraph 28 that in the circumstances the patentee’s counsel accepted that the court was not required to decline jurisdiction by virtue of what is now Art 24(4) of the regulation, citing GAT v LUK. This is significant because, as of course the judge will have well understood, the jurisdiction issue under Art 24(4) is something the court must be satisfied about of its own motion (Art 27).[54]Actavis v Lilly went to trial and ended up in the Supreme Court. At all three levels the UK courts made decisions on both the UK designation and the foreign designations and in doing so decided foreign law. Not only is the final outcome of Actavis v Lilly in the Supreme Court an important case given the judgment on equivalents but the proceedings as a whole and the Supreme Court’s decision on the foreign patents represent a significant shift in practice in that it is now clear that there can be situations in which the UK court will hear and determine cross-border patent cases. In particular the UK court can and will decide questions of infringement of foreign patents which are other EP designations of a UK patent in suit.[55]Coming back to the chronology of jurisdiction cases, the next relevant case is Anan Kasei v Molycorp [2016] EWHC 1722 (Pat). Here the claimant brought a claim in the UK for infringement of a UK and German designation of an EP patent against a defendant domiciled in England. After the action began the defendant issued nullity proceedings before the German Federal Patent Court. In the UK the defendant challenged jurisdiction (in fact the challenge started before the nullity suit but nothing turns on that). In the defence and counterclaim the defendant challenged the validity of the UK patent. After the draft of this judgment was circulated it was pointed out that in addition, in its pleading, the defendant referred to the German nullity proceedings and stated that if it was necessary to do so it would challenge the validity of the German patent on the same grounds as it had in Germany. In his judgment at paragraph 11 Arnold J noted that the claimant accepted that the court had no jurisdiction over the claim for infringement of the German patent as originally formulated but proposed an amendment to the Particulars of Claim which alleged that if the German designation was not invalid (which was to be determined by the German courts) the defendant had infringed. A declaration to that effect was also sought.[56]Arnold J noted that the claim could not be stayed on forum non-conveniens grounds (paragraph 27). He held that Art 24(4) was engaged and concluded therefore that the court had no jurisdiction over the claim in respect of the German patent. I will say at the outside that I would have reached the very same conclusion. However the way the argument before me developed suggested that the reasoning in Anan Kasei could be interpreted in a manner which I respectfully would doubt. Although they never put it quite this way Genentech’s argument seemed to be that the judge’s emphasis on the “single issue of whether Molycorp has infringed a valid claim” would mean that even if the nullity proceedings had not been commenced in that case, the claim in the UK still engaged Art 24(4). Arnold J does not say that. I will return to this below after dealing with the last cases in the chronology.[57]At paragraph 21 of Anan Kasei Arnold J noted that part of Laddie J’s reasoning in Coin Controls was doubted by Aikens LJ in JP Morgan Chase v Berliner [2012] QB 176 (nothing turns on that part) and also that Aikens LJ also held that Laddie J’s analysis of the relationship between validity and infringement for the purposes of what is now Art 24(4) and 27 appear to have been implicitly approved in GAT v LUK.[58]After Anan Kasei, Arnold J decided Fujifilm v Abbvie [2016] EWHC 2204 (Pat). This was a claim for an Arrow declaration in relation to the UK. The issue was service out of the jurisdiction. The claimant submitted that permission to serve out was not required because the UK court had exclusive jurisdiction over the claim because, as an Arrow declaration, the proceedings were concerned with validity of a European Patent (UK) and so Art 24(4) was engaged. Arnold J accepted that submission. He summarised his decision in Anan Kasei in paragraph 74 in the following way:
“… the Court of Justice of the European Union has held that Article 24(4) should be given a wide interpretation, and it has implicitly endorsed the reasoning of Laddie J in Coin Controls Ltd v Suzo International (UK) Ltd [1999] Ch 33 to the effect that the issues of validity and infringement are inseparable in patent cases because there is in the end only one question: has the defendant infringed a valid claim? Accordingly, I held that this court had no jurisdiction over Rhodia's claim in respect of the German designation of the Patent because, although framed exclusively as a claim for infringement, it implicitly asserted that the German designation was valid when, as the claim itself acknowledged, the validity of the German designation had been challenged by Molycorp and had to be determined by the German courts.”
[59]In deciding whether the Arrow declaration engaged Art 24(4) Arnold J held that it was vital to focus on the substance rather than the form of the declaration sought and held that in substance an Arrow declaration did engage Art 24(4). I agree. Arrow declarations are different in substance (and form) from declarations of non-infringement. If the Court grants the Arrow declaration, then as Arnold J explained in paragraph 78 it will mean that the claims will either be invalid or not infringed, and it does not matter which. The absence of a choice between invalidity and non-infringement is inherent in the Arrow declaration because it looks prospectively to say that based on a given application no valid claim could be granted which would be infringed by the applicant’s product or process.[60]In Chugai v UCB [2017] EWHC 1216 (Pat) the claim was for declarations about the payment of royalties due under a licence. Chugai claimed no royalty was due because its product fell outside the licensed patent, which was a US patent. The defendant contended that the claim was not justiciable in the English court because in substance the issue to be decided concerned not only the scope of the US patent but its validity and validity of foreign patents was not justiciable. Chugai’s primary argument was about claim construction (para 12) but it had two further arguments, one based on file wrapper estoppel which the judge held did not involve invalidity even though it entailed looking at prior art (paragraph 14). However the second further argument was based on a submission that the claim should be construed to preserve its validity having regard to an item of prior art which, it was argued, would otherwise deprive the claim of novelty (paragraph 17).[61]UCB submitted that validity and infringement were inseparable relying on Anan Kasei v Molycorp. Carr J concluded at paragraph 17 that Anan Kasei showed that the court should carefully examine the substance of the dispute in the context of challenges to jurisdiction under Art 24(4) of the regulation but that not every infringement dispute is concerned with a challenge to validity “for example, if a party has undertaken not to challenge validity, and only to pursue a case of non-infringement.” Carr J then referred to Actavis v Lilly as an example of this principle since it involved an undertaking not to challenge validity. He noted that in the trial judgment in Actavis v Lilly Arnold J had accepted as an aid to construction an argument that if the defendant’s construction was correct the patent would have been rejected by the EPO as invalid for added matter. Carr J noted in paragraph 30 that in paragraph 148 of his trial judgment Arnold J had noted that the patentee pointed out that validity was not in issue and said “but it does not follow that the court cannot consider what the consequences of Lilly’s construction would be for the validity of the patent”, citing the well known decision on that point in AHP v Novartis [2001] RPC 159. So as Carr J said in paragraph 31, even though validity was not and could not be raised, in the trial judgment in Actavis v Lilly there was a holding that it was appropriate to consider the consequences of the patentee’s construction for validity of the patent. Carr J explained that Actavis did not claim the patent was invalid, they proceeded on the basis the patent was valid, and contended that this supported their construction of the claims. He held this was permissible.[62]Having then dealt with justiciability Carr J held (paragraph 57-60) that the court should not decline jurisdiction. Amongst other things this was because Chugai did not contend the patent is invalid nor did Chugai seek any relief as to invalidity, insofar as it refers to validity it is to support its construction of the claims.[63]That completes a chronological run through of the relevant cases. For the purpose of analysis I will distinguish between the proceedings in the form in which they are actually constituted and the problem of what may or will happen in future.[64]Lilly contends that in Anan Kasei the fact that a nullity action on the German patent had been commenced was critical. Once that had happened one can see that there are live issues of validity and infringement. Now a difficulty is that the nullity proceedings were in Germany and not the UK. So it could be said that although validity was in issue in the German proceedings, the nature of the UK proceedings had not changed and if the English proceedings were not “concerned with” validity before the German nullity suit began, how can they have become “concerned with” validity afterwards? Although this concerned me at one stage, as mentioned above it was pointed out after the draft judgment was circulated that in Anan Kasei the defendant confirmed in its UK counterclaim that it would challenge the validity of the German patent if necessary. On that basis Anan Kasei is in the line of cases based on Coin Controls and Knorr Bremse in which validity “is to be” in issue.[65]On the other hand if the German nullity action had not been commenced and validity was not going to be in issue in the UK in Anan Kasei then I cannot see a convincing distinction between that case and Actavis v Lilly. Genentech’s argument seemed to suggest that the difference was that Actavis v Lilly an undertaking had been given not to challenge validity, however that undertaking did not cover a revocation claim in the patent’s home member state and moreover it was not an undertaking about the arguments Actavis could deploy. The undertaking meant that the English court was not going to be asked to make a finding about validity. It did not mean Actavis were fettered in the arguments they could deploy.[66]So if the distinction between Actavis and Anan Kasei is not the German nullity suit (including the fact the validity was to be in issue in the UK) nor before undertaking in Actavis, the only other possible distinction would have to be in the nature of the particular arguments which could actually be deployed. That is Genentech’s core point and it has a resonance in references in some of the judgments I have mentioned to looking at the substance of the arguments. But I am not convinced for two reasons.[67]First there is no clear distinction between the kinds of arguments a person arguing non-infringement may deploy. The arguments range from a submission that the claim must be construed in such a way that their product does not infringe, without any express reference to possible validity consequences, to an express deployment of a validity squeeze of the sort which Carr J identified was actually used by Actavis at trial. If you assert that the claim must be construed in one way because on the alternative construction it would be invalid, to then go on and state “but I do not say it is invalid, I say it is valid but I do not infringe” adds nothing of substance and adds nothing in terms of the issues before the court. The court would have to grapple with the submission and make findings about claim scope. In my judgment there is no clear line which can be drawn in relation to such a squeeze argument. A different case is if the person arguing non-infringement asserts in argument that the patent is invalid on the construction he contends is correct. However this does not cause any difficulty. If validity is not in issue then that submission is just not relevant to anything the court has to decide. It is hot air.[68]The Patents Act contains provisions which place restrictions on when validity can be put in issue (s74). This section does not prevent a party in a case in which validity has not been put in issue from making any submissions it wishes about claim scope, including validity squeeze arguments. Such an argument may not have the same impact if it is not accompanied by a revocation claim but that is not the point.[69]At this stage I will mention the Supreme Court in Actavis v Lilly. The parties did not use this occasion to argue about one of the major questions left undecided by that case - whether a thing which would fall within the scope of protection on an equivalents analysis but not otherwise would also be capable of depriving a claim of novelty (or making it obvious). The question is, on the premise that thing was prior art and therefore one way or another the patentee should lose, would the claim be invalid or is the correct analysis that the thing’s presence in the prior art just precludes an equivalents analysis? And if the latter is correct does it preclude an equivalents analysis altogether or in relation to the relevant claim feature? And does the answer depend on whether the prior art thing would be known to the skilled reader as a matter of common general knowledge?[70]The reason I mention this is because it is at least possible that one set of answers to those questions is capable of having a bearing on the issues on this application because they might tie together, in a formal legal sense, the issues of validity and infringement. But neither party submitted that Actavis v Lilly led to that result. Since this issue is likely to require a number of decisions to resolve both in the Patents Court and the Court of Appeal and perhaps another decision of the Supreme Court, I will not take it into account.[71]The second reason why I am not convinced by Genentech’s core point is that it does not make sense to me to decide something as fundamental as the court’s jurisdiction to determine a case based on the way the arguments are put. Many arguments about construction have unspoken consequences for validity. A party may be able to deploy two rhetorical points in support of a given construction, one based on the language and the other based on consequences such as validity. I do not see how the court’s jurisdiction can depend on whether the party chooses to deploy one or other or both pieces of rhetoric.[72]The court’s jurisdiction must depend on what is properly in issue. If a party applies to revoke the patent or seeks a declaration relating to validity (to the extent such things are possible) then validity will be in issue. The court would then be required to decide on the validity of the patent. However if the only issue the court has to decide is infringement then that must be a pure breach case (per GAT v LUK). And that will remain so even if the court has to evaluate arguments which are explicitly couched in terms of the consequences for validity of a particular claim construction.[73]Even leaving aside the fact that validity was to be in issue Anan Kasei, the relevant difference between Anan Kasei case and the present situation is really that once validity was in issue before the German court, the fact that infringement and validity are inseparable questions means that the issue of infringement before the UK court could not be separated from it. So the UK case was concerned with validity and Art 24(4) was engaged.[74]Since infringement and validity both depend on claim scope whatever submissions are deployed in argument, a “pure breach” infringement decision will reach a finding about claim scope which is capable of being inconsistent with a decision in a separate validity suit. That is an inevitable consequence of GAT v LUK. That kind of inconsistency is inherent in the scheme of Art 24(4) and its limitation to validity. The only way to avoid it would have been for the CJEU to have accepted the very wide submissions of the French and British governments, but the court did not.[75]Mr Layton QC submitted that under Art 45(2) of Brussels I (recast) the finding of a UK court on the scope of (say) a German patent made pursuant to the DNI would be binding on a later German court making a finding on validity if, for example, Lilly lost on its DNI and so after those proceedings were over started a nullity suit in Germany. He deployed a sophisticated argument based on the principle of mutual trust with reference to Krombach v Bamberski (Case C-7/98), Hoffmann v Krieg, (Case 145/86), Art 45(1)(e)(ii), Art 52, Art 54 and a number of academic articles. Mr Lewis QC did not agree and deployed sophisticated arguments, academic articles and Gothaer Allgemeine Versicherung AG v Samskip GmbH (Case C-456/11) in response. If Mr Layton is right this may have profound implications for patent cases under the Regulation since, as I understood his submission, it would extend to findings of fact – such as whether a given item of prior art was published – and would be applicable to any EP designation. In my judgment Mr Layton’s submission is not one which I need to resolve because even if it is wrong, GAT v LUK tolerates the inconsistency I have referred to.[76]The two issues – validity and infringement – are indeed inseparable once they are both in issue. But if validity is not in issue then there is no issue for infringement to be inseparable from.[77]I am well aware that in Anan Kasei (and the summary of it in Fujifilm) reliance was placed on the concept that a claim for infringement implicitly asserts that the patent is valid. I do not believe it is a necessary part of the reasoning. I doubt that such an implicit assertion could have the effect that validity is in issue in the UK since the whole point of s74 of the Patents Act 1977 is to place limits on the circumstances in which validity can be put in issue.[78]In the terms used by the Advocate General in GAT v LUK, it is possible to say that Actavis v Lilly was a pure breach case without reading the skeleton arguments advanced by Actavis. You can tell because validity has not been put in issue. Anan Kasei was a pure breach case until the German nullity suit was started and cross-referred to.[79]If I apply this approach to the Related Claims in this case then there is only one answer. Lilly’s non-infringement case is based fundamentally on a submission about claim scope. The arguments being deployed to say why the submission is correct include points which have nothing to do with validity. It is clear from the challenge to the validity of the UK patent that Lilly believes that all designations of EP 822 would be invalid if the claims are wide enough to catch ixekizumab. But no doubt Actavis believed the same sort of thing about the pemetrexed patent. These beliefs are irrelevant.[80]The challenge to validity of the UK patent is not, either in form or substance, a challenge to the validity of the foreign patents. If the court finds that the UK patent is invalid, only the UK patent will be revoked. The foreign patents will remain in the register. A finding of invalidity of one EP designation does not bind another court with exclusive jurisdiction to revoke the patent in its own state.[81]So I find that as presently constituted these proceedings would not engage Art 24(4). However that is not the end of the matters I need to decide. The expression “would not” is used because cases like Coin Controls and Knorr make it clear that the English court will not just look and see if validity is in issue but see if is to be in issue. I now turn to that.

Is validity going to be in issue in future?

[82]Lilly’s undertaking recorded in the order of Master Price was not to put validity in issue in the claim for a DNI. Since then Genentech has made clear that if the court does not decline jurisdiction then it intends to counterclaim for infringement. Lilly’s response was to state that its undertaking was not concerned with a counterclaim and it was not prepared to undertake not to challenge validity of the foreign designations until it had seen Genentech’s counterclaim.[83]Genentech submitted this was an inappropriate change in position as compared to the unqualified undertaking given to the Master and submitted in argument that the Master’s order should be discharged for that reason (no application for such an order was made). I agree with Genentech that Lilly’s position could and should have been explained to the Master more fully since the possibility of an infringement counterclaim from Genentech was foreseeable to the experienced legal team representing Lilly. However Genentech make too much of this. On its face the undertaking applies to the proceedings in the form they were constituted and that is what mattered. There is no justification in discharging the order on that ground.[84]However Lilly’s stance in relation to the infringement counterclaim means that there is no guarantee that validity of the foreign patents will not be put in issue in this action. If that happens Art 24(4) will plainly be engaged at that point and this entire exercise will have been something of a charade. I was concerned about this until two points were raised.[85]The first point is that Lilly says that they wish to see what exactly is said in any infringement counterclaim before they decide what to do. It may be that Genentech will not seek an injunction against Lilly to take Taltz off the market even if it does infringe. Maybe a royalty will be sought instead. This is not as fanciful as all that, it has been done in other cases recently in the Patents Court (cf Merck v Ono [2015] EWHC 2973 (Pat)). If that is what Genentech does then Lilly may take the view that the stakes are not so high that it wishes to challenge validity of the non-UK designations. So the court should not assume that a counterclaim for by disputing the validity of any of the non-UK patents will inevitably follow as a defence to a counterclaim for infringement of those patents.[86]The second point is that Lilly submit it is not inevitable that there will actually be a counterclaim for infringement. There is no undertaking to that effect from Genentech, intentions can change and, crucially, there is what could be a good reason why Genentech may wish not actually to counterclaim for infringement if it can effectively get the benefit of doing so without actually issuing a counterclaim. The issue has to do with the rules applicable to the parallel EPO opposition. Lilly is not a party to those proceedings and it cannot join them now because it did not bring an opposition within 9 months of grant. However under the EPO’s rules, if while opposition proceedings are pending the patentee brings a claim for infringement against someone who is not a party to those proceedings, then that person has the standing to join the EPO proceedings. Genentech may not want to have another opponent in the EPO and Lilly may feel its position on validity of the non-UK designations would be sufficiently protected by being in the EPO proceedings (c.f. Actavis in Lilly v Actavis). A challenge in the EPO does not engage Art 24(4).[87]There is no evidence about this from either side (to which I have had my attention drawn) but to be fair it all arose late in the day. If it was not for these unusual features then I would have no hesitation in finding that whatever the position is now, in all probability Genentech will counterclaim for infringement of the non-UK patents and Lilly would put validity in issue in response. In that case the approach of the English courts is to decline jurisdiction now rather than wait. However I am satisfied that these unusual circumstances mean that it would not be fair to pre-empt what each party may decide to do. There are sufficient uncertainties that the right thing to do is wait and see what happens. However in my firm but necessarily provisional view that wait should be at Lilly’s risk as to costs. If Genentech does counterclaim for infringement, and validity of the non-UK patents is put in issue (here or abroad) in response, then it is very likely that Lilly should bear the whole costs of this application even if they win it in its form today.

Forum non-conveniens

[88]Genentech contend that even if the previous arguments fail the Related Claims should be stayed on forum non-conveniens grounds. The test to be applied comes from the well known Spiliada v Cansulex case. The question is whether England is clearly and distinctly the most appropriate forum for trying the related claims in the interests of all the parties and the purposes of securing the ends of justice. The burden is on Lilly.[89]Subject to the point on the risk of irreconcilable judgments, this is a fairly straightforward matter. If the Related Claims are heard together with the issues concerning the UK patent, the basic facts would only have to be gone into and decided once. Those facts are about the nature of the ixekizumab product and what its properties are. That has the potential to provide a very substantial saving in costs. One legal team will be needed to get into the detailed issues of fact and science into issue. This applies to both sides and includes the court. The fact that each side may have engaged legal teams in other states to begin working on this dispute does not undermine that finding since the case is a long way from trial.[90]The fact that foreign law is involved is not a significant factor. The legislative basis is the same for all the patents. The differences are tangible but they are matters of nuance. Recent experience has shown that while foreign law experts will be needed, they are unlikely to be cross-examined. It is also now settled that claims such as the Related Claims are justiciable in this court.[91]The Supreme Court has just heard and endorsed a case which involved examination of the issue of infringement relating to three non-UK designations of an EP for which the UK designation is also one of the patents in suit. That does not mean that the Patents Court is necessarily the most appropriate forum for this case but it is relevant background.[92]No doubt it is correct as Genentech’s representatives submit that Lilly’s approach in this case will put pressure on Genentech to try and adopt a uniform construction of EP 822 which preserves validity and leads to a conclusion that ixekizumab infringes. But they would be trying to do this anyway even if the case is heard in different parallel jurisdictions as any rational patentee would. I reject the suggestion that this somehow makes these proceedings unfair.[93]Genentech submits there are no or no sufficient facts to link these matters to the UK. I disagree. It is true that the product is made in Ireland and that QC testing is in Italy, but the UK is a major market and the evidence shows that it may well be largest market in Europe well within the lifetime of the patents. The Patents Court in London is well equipped to handle cases of this kind including (I refer again to Actavis v Lilly) cases which bring together a group of non-UK patents from the same European bundle as the UK patent. Trying so many aspects of this overall dispute together in one place on one occasion and in a timely fashion is also likely to promote settlement.[94]I have dealt with these factors briefly because I believe that, subject to the irreconcilable judgments issue, they present a solid basis on which to reject a stay of this case on forum non-conveniens grounds. I turn to consider that issue.[95]The problem is easy to understand. In order to determine whether a product infringes one of the patents, the court will be required to reach a decision on the true interpretation of the claims and decide if the product falls within the scope of protection conferred by that patent. The same kind of decision is necessary when deciding validity. So a decision that the claim has a wide scope such that ixekizumab infringes would be inconsistent with a separate decision that the claim is valid because it has a narrow scope which would exclude ixekizumab. So if the Patents Court in London decides that ixekizumab infringes the UK patent, it will then go ahead and apply the same approach to the claims when looking at the validity of the UK patent. That may lead to the result that the wide claim renders the UK patent invalid. There would be no inconsistency in that instance. However if the Patents Court also decides that the patents in France, Germany, Spain, Italy and Ireland are also infringed by ixekizumab it will make no corresponding decision about validity. Lilly might then wish to go to the French, German, Spanish, Italian and Irish courts and bring a claim for revocation. But if the French, German, Spanish, Italian and Irish courts are not bound by the decision of the English court on claim scope, they may reach an inconsistent conclusion holding that the patent is valid because the claim is narrow. Lilly would then be left in a bind.[96]As mentioned above already Mr Layton for Lilly advances a sophisticated argument based on mutual trust in support of a submission that such a finding by the English court in the example I have given would be binding on the foreign court. Mr Lewis did not agree. I declined to decide it before and I decline to decide it again for similar reasons.[97]If Lilly is right the problem does not exist, but assuming Genentech is correct that the decision of the Patents Court on infringement of a foreign patent in a Brussels-Lugano state is not binding as to claim scope on the courts of that state trying the issue of validity, all that means is that the problem described is inherent in GAT v LUK. It is nothing to do with the substance of the particular arguments being run by one party or another in the infringement case and everything to do with the simple fact that the scope of the claims is as much an issue to be decided when considering infringement as it is an issue to be decided when considering validity. Since at worst it is inherent in the arrangement whereby the English court can hear a “pure breach” case, I do not regard it as a factor of significant weight in the overall analysis under Spiliada.[98]Also it seems to me that the problems which occur if Genentech is correct are much more likely to be problems for Lilly than problems for Genentech. The English Court’s decision will be binding on the issue of infringement between these parties. If Lilly wins in the English Court such that ixekizumab does not infringe then that’s that. It does not matter if the foreign patents are valid or invalid. Even if the foreign courts decided validity on a different claim construction, ixekizumab still does not infringe those patents. If Lilly loses in the English Court such that ixekizumab does infringe the foreign patents then that issue is still finally decided and that is irrespective of the issue of validity of the UK patent. Lilly may want to try and attack validity of the foreign patents and it may succeed or it may fail but the risk which arises from the foreign court accepting a narrower construction when considering validity rather than the wide one which led to the finding of infringement is all with Lilly.[99]Standing back and looking at the matter overall, in my judgment trying the Related Claims together with the issues relating to the UK patent is in the interests of both the Lilly parties and Genentech for the purposes of securing the ends of justice.

Conclusion

[100]I will dismiss Genentech’s applications, subject to the point on costs relating to Lilly challenging the validity of the foreign patents in response to a counterclaim for infringement.

Cited in 4 later judgments