This morning the Court of Appeal ([2019] EWCA Civ 2272) dismissed Gilead's appeal of Mr Justice Arnold's (as he then was) September 2018 decision in Teva v Gilead finding that Gilead's SPC protecting its combination HIV anti-retroviral drug, Truvada, was invalid.  The product subject to the SPC was a combination of tenofovir disoproxil and emtricitabine.  Gilead claimed that this product was protected by Claim 27 of European Patent No 0 915 894 (the '894 Patent).  Teva (and a merry band of other generic companies) argued that the '894 Patent did not protect this combination product under Article 3(a) of the SPC Regulation.  Article 3(a) requires that an SPC can only be granted if the product subject to the SPC is "protected by a basic patent in force".  

Back in December 2016, the parties first clashed in a two day trial before Arnold J (as he then was). It was in that judgment, delivered in January 2017, that the Judge referred the question that he previously referred in Actavis v Sanofi to the Court of Justice of the European Union (CJEU).  The reference asked:

The Judge, as he said at [96], was hoping that the CJEU would "provide a clear answer this time" as to what Article 3(a) required.  Did it require merely that the product fall within at least one claim of the basic patent applying the Extent of Protection Rules (i.e. Article 69 of the European Patent Convention and the accompanying Protocol), or did it require something stricter?  In his reference, the Judge trotted through the English court's and CJEU's case law Article 3(a) - Takeda, Farmitalia, Daiichi, Yeda, Medeva (and its progeny), Actavis v Sanofi, Eli Lilly v HGS, Actavis v Boehringer, - and found that it was clear that something more was required, but what that "something" was was not clear.  

His view, expressed at [97], was to embrace the "embodies the inventive advance" test.  He explained:  

Back on the terra firma of the Patents Court, Mr Justice Arnold (as he then was) was asked by Teva et al to provide his judgment in light of the CJEU's decision.  On 18 September 2018, the Judge held the SPC was invalid.  The first part of the CJEU's test, he said, was that:

Gilead's SPC failed this part.  The Judge held that the '894 Patent said nothing about tenofovir disoproxil and emtricitabine being combined to treat HIV; emtricitabine is not even mentioned.  There was thus no basis for the skilled person to understand that the combination embodies the technical contribution of the '894 Patent.  Tenofovir disoproxil might be the technical contribution, but not the combination.  

The second part of the CJEU's test was that: 

Gilead's SPC failed this second part of the test.  Although tenofovir disoproxil was specifically identifiable, emtricitabine was not.  Emtricitabine was not mentioned by the '894 Patent, was not a member of a specific class of  compounds mentioned by the '894 Patent (either by reference to structure or activity) and, although known by the skilled person at the priority date, there was no evidence that it was either known or CGK to be "effective" in the treatment of HIV.  

The Court concluded that this result accorded with the objective of the SPC Regulation in that Gilead has already obtained a marking authorization for the "mono" fumerate form of  tenfovir disoproxil less than 5 years after the '894 Patent was filed - thus not suffering any regulatory delay in exploiting the patent.  The final sentences of the decision referred to Gilead's patent for the combination having been revoked by the Opposition Division and their appeal dismissed.  "Thus Gilead made no invention in devising the combination which warranted the grant of a patent, let alone a SPC."

So, Gilead appealed.  The Court of Appeal hearing was last Tuesday and the decision was handed down this morning. 

Gilead argued that the test was an Extent of Protection test, subject only to the two provisos (i.e. the two part test) provided by the CJEU.  Gilead satisfied both parts of the test.  The first part was meant to exclude products where the claim was only to A, but was said to protect A+B.   The '894 Patent did't fall into this camp as the claim referred to other therapeutic ingredients.  The second  part of the test was meant to "stop the clock" in what knowledge the skilled person could use to determine whether the ingredient was identifiable.  Emtricitabine  was identifiable at the priority date, so Gilead satisfied that test.  Gilead argued that the judge reached these conclusions because he had actually applied the "inventive advance" or "technical contribution" test and had assessed the second part of the test by reference to CGK, despite the court having expressly ruled that all the prior art formed the relevant pool of  knowledge.  Teva argued that the judge using the "technical contribution" test was not objectionable - the AG found the test objectionable only insofar as it conflated SPC law with the tests for patentability and inventive step.  What the Judge was really asking was "is the product the actual subject matter of the patent".  

Lord Justice Floyd gave the decision of the Court of Appeal, with Lord Justice Dingemans and Lord Justice Lewison agreeing.  He kicks off [18] by declaring that there 

Floyd LJ then provided an excellent summary of the crescendo of Article 3(a) case law to date, including the AG and  CJEU decisions in the present case and pending cases.  This can be summarized as follows:

(1)  The first part of the Teva v Gilead test is just "a more elaborate exposition of the 'necessarily' part of the first part of the Lilly v HGS test, i.e "the claims relate...necessarily...to the active ingredient in question."  

(2)  For a combination product to be protected, the claim must require the presence of two compounds.  This is not a simple Extent of Protection test - it goes further in that it demands that the claim "requires" each component be present in the claim. 

(3) Claim 27 refers to "other therapeutic ingredients" as expressly being optional. The convenience in claim drafting of making it optional, so as to avoid writing out separate claims, was not a good enough reason to find that the combination was protected.    The wording of the claim making it optional means that it is not "required".  Claim drafting is important in this area (citing Actavis v Boehringer (C-577/13)- although the AmeriKat takes issue with over-reliance on that case as being an Article 3(a) case...).  

(4)  Although the CJEU did not hold that the word "optionally" was fatal to Gilead's case, it still drew attention to the word.  The role that "optionally" plays in this assessment is a matter for the national courts.  

(5)  There is nothing in the '894 Patent that suggests to the skilled person that Claim 27 "requires the presence of another ingredient.  Everything points the other way. The skilled person might well know from the common general knowledge that other anti-viral agents would be useful in practice in the treatment of HIV, but he would not therefore assume that the presence of such an agent was required by the claim."  

(6) The breadth of Claim 27 is a factor.  The claim is not limited to a pharmaceutical composition containing tenofovir disoproxil for the treatment of HIV, nor is the phrase "other therapeutic ingredients" limited to anti-viral agents either.  Claim breadth is another reason for not reading a requirement for the presence of a second anti-viral agent into Claim 27.  At [83] the Court of  Appeal continues:

(7) Wading into the second-part of the test presents difficult questions of who knew what, when and by what means (i.e. all prior art or CGK?) and is unnecessary, given the failure under the first-part of the test (see [85]). Floyd LJ explained that he "would prefer to leave those issues to a case in which their resolution affected the result".  However, importantly, he said that:

According to the Court of  Appeal, patent claims apparently cannot be the Harry Potter-style "Room of Requirement" when it comes to satisfying Article 3(a) of the SPC Regulation.  The second active ingredient cannot just appear in the room if you need it, if it wasn't there when you first walked in.  But is that or, more accurately, should that be right when given the constraints on claim drafting and ticking clocks in a first-to-file system? 

There still seems to be a tension between saying the SPC confers the "same rights" as under the patent under Article 5, but then saying they are also "limited".  The only form of statutory limitation is that under Article 4 (to which Article 5 is subject), but that limitation is by reference to the product subject to the SPC.  If the patent protects A+B, then the SPC should protect A+B if the product subject to the MA is A+B.  However, the case law says that although the patent might protect A+B, if B is not "required" by the claims then an SPC cannot protect  A+B.  So the SPC is not "conferring the same rights" as that of the patent (assuming "rights" and "protection" are the same thing...).  This is consistent with the "something more" that the case law of the CJEU and of the Patents Court keeps referring to.  The "something more" is actually a limitation, contrary to Article 5.  Of course, one would rightly argue that the logic of some of the above is essentially an infringement test by the backdoor, which has been expressly rejected.  Medeva at [25] also that states that it follows from Article 5 "that Article 3(a) of the regulation precludes the grant of a SPC relating to active ingredients which are not specified in the wording of the claims of the basic patent".  But, like Lord Justice Arnold, the AmeriKat has never understood how "it follows", unless you misunderstand patent law. But maybe that is the point...

But while the AmeriKat rustles around with this ball of yarn until she tires herself out, in-house patent attorneys will - subject to an appeal to the Supreme Court - likely be considering how to carve around the Court of Appeal's directions on claim drafting, "optionally" language, unity of invention and filing in time before the clinical trials are public.  

So there you have it.  Happy holidays to all you SPC-lovers out there - a meaty little judgment for you to get your teeth into with your holiday champagne - the AmeriKat hopes it is a big bottle...

This morning, the Court of Appeal handed down its judgment in Conversant Wireless Licensing S.A.R.L v Huawei Technologies Co. Ltd, ZTE Corporation and Ors [2019] EWCA Civ 38.  As regular Kat readers will know, the appeal concerns the English court's jurisdiction to grant a global FRAND licence where UK sales only account for 1% or less of the sales on which royalties are claimed.  The Kat's correspondents, Jane Mutimear and Richard Vary, send this report:

On 30 January 2019 the Court of Appeal handed down its judgment in Huawei and ZTE's appeal against Carr J's refusal to dismiss the case brought against them by Conversant on jurisdictional grounds.  The Court of Appeal dismissed the appeal, with Lord Justice Floyd setting out the reasons with which Lord Justices Flaux and Patten agreed. 


Background

Conversant sued both companies on patents it claims are standard essential, seeking injunctive relief and damages, together with a declaration that it had made FRAND offers or a determination of the FRAND terms.  Huawei and ZTE challenged jurisdiction on two grounds: 1) that the case would raise validity and infringement of foreign patents which is not justiciable subject matter in the English Court and 2) that the Chinese court is the natural and appropriate forum so the Court should refuse permission to serve the Chinese entities out of the jurisdiction on forum non conveniens grounds.  Following the CoA judgment in Unwired Planet v Huawei, the appellants accepted that the justiciability point was no longer arguable before the Court of Appeal, so the Court only had to deal with the forum non conveniens ground (which Huawei had not argued in Unwired Planet).  The appellants also proposed that if the appeal failed on this point, questions should be referred to the ECJ.  The judgment goes through the background of licensing in the SEP area, quoting from the Court of Appeal's judgment in Unwired Planet, and explained briefly the wider fight between the companies, which includes proceedings commenced in China by Huawei China and ZTE China against Conversant seeking to invalidate Conversant's Chinese patents and seeking a FRAND determination.

Jurisdiction over UK defendants

The Court addressed the position as against the two UK domiciled appellants, turning to the Brussels Recast Regulation Article 4, which provides that "persons domiciled in a Member State shall, whatever their nationality, be sued in the courts of that Member State" and relied on the CJEU judgement in Owusu v Jackson [2005] for authority that the English court could not apply forum non conveniens to decline jurisdiction over a claim against a person domiciled in a contracting state on the ground that the natural forum for the claims was the courts of a non-contracting state.  The question was whether this reasoning applied to cases involving Article 24 (giving Member States exclusive jurisdiction over proceedings involving registration or validity of patents).  The Court concluded that it was a matter of discretion for the court whether to assume jurisdiction under Article 4 when the subject matter fell within Article 24 (relying on the judgment of Andrew Smith J in Ferrexpo v Gilson [2012]).

Service out of the jurisdiction on Chinese companies

Turning to service on the Chinese entities, there was no dispute that the first two hurdles were met: 1) the claim falls within a gateway (CPR 6 PDB) and 2) there is a serious issue to be tried.  The dispute related to the third hurdle i.e. whether England and Wales is the proper forum to hear the case.  The "case" for this analysis is not just what the Claimant has pleaded; the court must have regard to the totality of the dispute, including the likely defence and the case must be characterised in a way which does not pre-judge the appropriate forum.

The evidence before the judge at first instance had caused him to conclude that the Chinese courts did not have jurisdiction to determine essentiality or infringement of non-Chinese patents, nor did they have jurisdiction to determine FRAND rates without agreement from both parties.  Conversant had not agreed to accept the Chinese court's jurisdiction for such determination and the judge had concluded that their refusal was reasonable.  It was on this basis that the judge had concluded that China was not an alternative available forum.

The appellants sought to adduce new evidence before the Court of Appeal to introduce the Guangdong Court Guidelines, which had been issued 10 days after Carr J had handed down judgment, but before the order was sealed.  These guidelines stated that where the territorial scope of the SEPs in question exceeded the territorial scope of the court and the other party did not object (or the objection was deemed unreasonable), the court could determine the royalty.  The Appellants sought to rely on this as evidence that the Chinese court would now exercise jurisdiction to determine essentiality, infringement and FRAND for a global portfolio in circumstances where both parties consented and their expert claimed that if the facts were analogous to this case, the court would take the view that the refusal to agree was unreasonable. Conversant's expert disagreed and pointed out that there was no case law on the circumstances in which it would be "unreasonable" for one party to object. The Court of Appeal agreed with Conversant's expert.  The Guidelines had been introduced for a trial period and it was speculative to interpret them in the way the appellants' expert was seeking to do.

In reaching its decision the Court of Appeal stated that it was clear that you could get different answers to the forum conveniens questions, depending on the level of generality at which the dispute is characterised.  If you just look at the relief which would be granted in the English court, that would be too specific.  But if you ended up making the claimant pursue a claim on a different property right with different underlying facts in the foreign forum, that meant it had been defined too broadly. Conversant's claim was closely analogous to the claim advanced in Unwired Planet and accepted by the Court – (i) the UK patents are essential to the standard; (ii) that the claimant has complied with its ETSI undertaking in making FRAND offers; (iii) that the defendants have not so complied without any reasonable ground for so doing; and (iv) that it is entitled to enforce its UK SEPs and obtain the usual relief for infringement, including a FRAND injunction and damages, and a determination of the FRAND terms.  If Conversant were made to seek a remedy in China, they would have to advance their case based on different patents, relying on different facts.  Floyd LJ pointed out that that it was not legitimate to characterise the claim as one for the enforcement of a global portfolio right, as no such right exists.  It is a dispute involving determination of infringement, validity and essentiality of UK patents.  The FRAND element is embedded in the determination of whether Conversant is entitled to the relief for infringement of their UK SEPs.  The FRAND element cannot be considered separately for the purpose of the forum conveniens analysis.

Relevance of the new evidence

Although the new evidence in relation to the Chinese Court Guidelines would be admissible under the first Ladd v Marshall test (couldn't have been submitted at first instance trial) it did not meet the second criteria of having an influence on the outcome of the case.  There was no guidance on how the court would apply the test of when refusal by one party was reasonable, and even Huawei's letter to the Supreme Court in relation to its petition to that Court for leave to appeal in Unwired Planet stated "[T]here is no decided case in which the Chinese courts have gone further, as the English courts did in these proceedings, by conducting a global royalty-setting exercise without the consent of the parties and then imposing the result thereof as a condition of avoiding a territorial injunction."

Rejecting reference to CJEU

Finally the Court refused the CJEU reference on the basis that first, the Court of Appeal in Unwired Planet was able to reach a decision on justiciability without a reference. Secondly, Huawei had never suggested that a reference was necessary in that case, and now accepts that the Unwired CoA decision precludes any justiciability argument before the Court of Appeal in this case. Thirdly, despite a lengthy draft produced by Huawei's counsel, Floyd LJ was not persuaded that there is any lack of clarity in the relevant EU law.