Actavis UK Ltd & Ors v Eli Lilly & Company

This appeal is from the judgment of Arnold J dated 15 May 2014 and his consequent order in an action by companies in the Actavis group of companies (together “Actavis”) for declarations of non-infringement (“DNIs”) of European Patent (UK) No. 1 313 508 and the corresponding national designations in France, Italy and Spain. I will refer to European Patent 1 313 508 as “the 508 patent” or “the patent”. The patentee of the 508 patent and the appellant is Eli Lilly & Company (“Lilly”). The appeal first raises issues of substantive patent law, which I shall call “the patent law issues”. The first of the patent law issues concerns the correct approach under our law (and the law of certain other designated states) to the construction of the 508 patent claims, in particular the requirement under the European Patent Convention 2000 (“EPC 2000”) to take account of “equivalents” as well as the extent, if at all, to which it is permissible to make use of the prosecution history of the patent in reaching conclusions about construction. The second of the patent law issues concerns whether the application of our law of contributory infringement (as well as that of the other designated states) justifies a finding of infringement in this case.

Depending on the outcome of the patent law issues, there are further issues about whether European Parliament and Council Regulation 864/2007/EC of 31 July 2007 (“the Rome II Regulation” or “Rome II”) means that the English court must apply the corresponding foreign laws governing the conditions for applying for DNIs in each of the foreign jurisdictions, or whether English law, as the lex fori, applies. I will call this “the Rome II issue”. Finally, if the foreign laws apply, there would be further issues about precisely what the foreign laws provide about the conditions for applying for DNIs in those countries and whether Actavis have complied with those conditions. I will call these “the DNI factual issues” notwithstanding that, as is well known, issues of foreign law are a rather special type of issue of fact.

Pemetrexed disodium is a cancer treatment which Lilly has marketed under the brand name Alimta since 2004. Lilly had patent protection for pemetrexed disodium by way of European Patent No 0 432 677 (“the 677 patent”). Actavis sought the DNIs of the 508 patent in order to place themselves in a position to market a competing product to Alimta on the expiry of the 677 patent in December 2015. They did so in a single action in this jurisdiction. Actavis make no challenge to the validity of the 508 patent: to do so would have meant that this court did not have jurisdiction to deal with the DNIs in respect of the foreign designations, because there is exclusive jurisdiction in relation to issues of validity in the courts of the country where the patent is registered. In an earlier judgment Arnold J rejected Lilly’s challenge to the English court’s jurisdiction to hear and determine Actavis’ action for DNIs in respect of the foreign designations and his judgment was upheld in this court: [2013] EWCA Civ 517; [2013] RPC 37. This court considered that there was jurisdiction on the ground that Lilly had accepted service of the proceedings. It did not find it necessary to decide an alternative issue concerning whether Arnold J had been right to refuse to stay the proceedings on the ground of forum non-conveniens.

The 508 patent relates to the use of pemetrexed disodium in the manufacture of a medicament for use in combination therapy with vitamin B12 and, optionally, folic acid. The active ingredient in Actavis’ proposed product is one of (a) pemetrexed diacid, (b) pemetrexed ditromethamine or (c) pemetrexed dipotassium (“the Actavis AIs”). Actavis’ case is a simple one. It is that, because they intend to use the Actavis AIs and not pemetrexed disodium, the 508 patent's claims to the use of pemetrexed disodium are not infringed. Lilly does not agree and contends that there will be either direct or indirect infringement of the 508 patent if Actavis launch any of the Actavis AIs in the UK or in any of the other states in relation to which DNIs are sought.

Arnold J agreed with Actavis that their proposed uses of the Actavis AIs would not infringe any claim of the 508 patent as a matter of English law. Next, he had to consider whether the same result would follow under the laws of France, Italy and Spain. It was common ground that those foreign laws applied to the issue of whether the Actavis AIs would infringe in those countries. He concluded that applying the infringement laws of those countries would lead to the same result as the application of English law, and that, therefore, there was no infringement in those countries either.

This meant that the judge needed to go on and deal with the Rome II issue before he could decide whether it followed that his decisions thus far would lead to DNIs in respect of the French, Italian and Spanish designations of the 508 patent. He concluded that English law, as the lex fori, applied to the conditions for the admissibility of a DNI, rejecting Lilly’s argument that it was the lex loci protectionis or lex causae, that is to say the patent law of the contracting state in respect of which the declaration was sought. Because there is no difference on the facts of this case between these two Latin descriptors, I will use the term lex causae in this judgment to describe the laws of the designated states in respect of which the DNI is requested.

Those conclusions were sufficient to enable the judge to grant DNIs in respect of each of the designations sought by Actavis, but the judge went on to consider whether, if he was wrong as to the applicability of the lex fori, Actavis had in fact complied with the conditions for obtaining a DNI under the lex causae. He concluded that Actavis had so complied.

Lilly appeals with the judge’s permission. On the appeal Mr Henry Carr QC, Mr Thomas Mitcheson QC and Mr Stuart Baran presented the arguments for Lilly. Mr Richard Meade QC, Mr Thomas Raphael QC and Ms Isabel Jamal did the same for Actavis.

In summary, therefore, the issues we have to decide on this appeal are:

whether the judge was wrong to decide that there was no direct or indirect infringement applying English law by use of the Actavis AIs as a matter of English law;

whether the judge was wrong to decide that the same result followed under the laws of France, Italy and Spain;

if the judge was wrong on issue (ii), whether the judge was also wrong on the Rome II issue;

if the judge was wrong on the Rome II issue, whether he was also wrong on the DNI factual issues.

I turn first to deal with the patent law issues.

The 508 patent is entitled “Combination containing an antifolate and methylmalonic acid lowering agent”. It has an earliest claimed priority date of 30 June 2000. Pemetrexed is a member of the class known by the name antifolates. Vitamin B12 is a methylmalonic acid lowering agent.

The specification begins at [0001] by stating that "Potentially, life-threatening toxicity remains a major limitation to the optimal administration of antifolates". It is explained at [0002] that antifolates work by inhibiting anti-folate-requiring enzymes by competing with reduced folates for binding sites on those enzymes. The specification identifies several antifolate drugs as being in development, including Lilly’s Alimta.

The specification then explains at [0003] that a limitation to the development of these drugs is that they may be associated with substantial toxicity, including mortality, for some patients. These toxicity effects had led to the abandonment of the development of some antifolates. In [0004] the specification explains that previous work had been done on the use of folic acid as a treatment for toxicity in this area. It also records work on vitamin B12 as a predictor of cytotoxic events.

The specification then continues:

These early, general statements are made in relation to antifolates as a class. However at [0010] the specification says, in what patent lawyers refer to as “a consistory clause”, that the invention:

This paragraph, in referring only to pemetrexed disodium, is highly specific. However at [0016] the specification reverts to greater generality when it states:

Paragraph [0022] is in a section which contains a number of definitions:

This passage, with its use of “the antifolate” rather than “an antifolate” in the emphasised sentence, is another indication that the invention is not the use of antifolates as a class. The invention is then illustrated by reference to a number of examples, which relate to animal and human tests in which the only antifolate used is pemetrexed disodium. At paragraph [0035] the specification states that animals were treated with “pemetrexed disodium (ALIMTA^®) (100 mg/kg or 150 mg/kg) once daily … by intraperitoneal injection alone or along with folic acid”. This is an indication that pemetrexed disodium is administered in a solution. In a similar fashion, the specification indicates at [0044] that, in a typical clinical evaluation using cancer patients, the antifolate is to be administered in four doses over a two week period by rapid intravenous injection. This again can only be the case if the antifolate, in this case pemetrexed disodium, is in solution.

Claims 1-11 of the Patent are use claims in the general form “use of A in the manufacture of a medicament for treatment of disease B” (so-called “Swiss” claims). Claims 12-14 are in the general form “product A for use in a treatment Y” (so-called “purpose-bound product” claims). It is only necessary to refer to claims 1 and 12, which are as follows:

It is beyond argument, therefore, that the claims are limited by reference to the words “pemetrexed disodium”. The task for the court is to determine what the skilled reader would have understood that phrase to mean.

Lilly alleges that the use of the Actavis AIs will give rise to direct or indirect infringement of claim 1. It does not so allege in the case of claim 12 because of the limitation to a combined preparation. Before coming on to the questions of construction which arise, I will summarise the prosecution history of the 508 patent. This is necessary because the judge relied on it in support of his interpretation of the patent. I will discuss the relevance of the prosecution history to an exercise such as this at a later point in the judgment.

The 508 patent was applied for by an international application under the Patent Co-operation Treaty filed on 15 June 2001. The application contained claims directed to a method of treatment, claims in Swiss form and purpose-bound product claims.

Under cover of a letter dated 8 January 2003 Dr Ivan Burnside, Lilly's in-house European Patent Attorney, filed a revised set of claims which omitted the method of treatment claims. Claims 1 and 2 were as follows:

These claims are of course the other way round from the claims ultimately granted (i.e. they start with the use of the methylmalonic lowering agent rather than pemetrexed disodium), but that fact is of no significance here. The point to note is that these claims were entirely general as to the identity of the antifolate. On 9 March 2004 the EPO examiner issued an official communication which raised, amongst other things, objections under Article 83 EPC (disclosure) and Article 84 EPC (clarity). The clarity and lack of disclosure objections were that the claims related to too many possible combinations of compounds by using general expressions such as “antifolate”, “methylmalonic acid lowering agent” and “FBP binding agent”. Moreover the claims, by using those same terms covered all compounds having these characteristics or properties, whereas the application provided support and disclosure for only a very limited number of such compounds.

Dr Burnside replied to the official communication in a letter dated 23 December 2004, under cover of which he filed new claims 1 and 2, this time starting with the use of the antifolate, now limited to “pemetrexed” as follows:

As the patent did not define what was meant by “pemetrexed” as a free-standing descriptor, it was perhaps inevitable that Dr Burnside was going to meet further official objection. Although the specification of the patent contains one mention of the term “pemetrexed” at [0004], it is followed by a Lilly reference number which, so we were told, shows it to be pemetrexed disodium. However, in support of the new claims, Dr Burnside said that, “in order to expedite the application proceeding to grant” Lilly had elected to amend the claims so as to reflect more closely the specific examples provided. The amendments were made without prejudice to Lilly’s right to obtain protection for other patentable subject matter in one or more divisional applications.

Notwithstanding these amendments, on 17 May 2005 the EPO examiner issued a further official communication objecting to the admissibility of the new claims. He said that the amendments introduced subject matter beyond the content of the originally filed documents, contrary to Art. 123(2) EPC (“added matter”). Thus the inclusion in claim 1 of “use of pemetrexed...” and in claim 13 of "a product containing pemetrexed...” did not find basis in the application documents as filed. According to the EPO Examiner “pemetrexed” was a distinct compound from pemetrexed disodium in that it was accorded a different number in a Registry maintained by the Chemical Abstracts Service from that accorded to pemetrexed disodium. Although the prosecution history does not reveal this fact, examination of the Chemical Abstracts Service Registry reference for “pemetrexed” shows that it is the diacid. The specification of the patent does contain one mention of the term “pemetrexed” at [0004], but it is followed by a Lilly reference number which, so we were told, shows it to be pemetrexed disodium. It was therefore, at best, uncertain what the term “pemetrexed” on its own was intended to refer to.

Dr Burnside replied to the official communication in a letter dated 8 March 2006, under cover of which he again filed new claims. The new claims were limited to pemetrexed disodium. Dr Burnside said:

The EPO examiner accepted the new claims, and the application proceeded to grant.

A patent specification is to be taken to be addressed to those likely to have a practical interest in the subject matter of the invention, in other words to a person skilled in the art: see per Lord Diplock in Catnic v Hill & Smith [1982] RPC 242. It is well settled that the skilled addressee may be a notional team of people with different specialisations where more than one specialisation is involved in the invention: General Tire v Firestone [1972] RPC 457 at 485.

Lilly contends that the 508 patent is addressed to an oncologist, whereas Actavis contends that it is addressed to a team which, although it includes an oncologist, also includes a chemist. The judge preferred Actavis’ argument on this point. Lilly’s argument ran as follows:

The judge should have addressed the identity of the skilled team by reference to the problem which the patent had solved, namely the reduction in side effects caused by the administration of pemetrexed whilst retaining efficacy. The role of a chemist was peripheral.

Claim 1 is only framed by reference to “manufacture” (i.e. in the Swiss form) because of the legal fiction required to circumvent the restriction on patenting of methods of treatment. The skilled reader would appreciate this and therefore take no account of the reference to manufacture.

There are many peripheral roles which would be involved in implementing an invention such as the present: for example there might be questions as to the appropriate packaging for the finished product. Not every peripheral task involved in bringing an invention to market justified a representative on the skilled team.

The evidence showed that the oncologist was not concerned with the source of the pemetrexed anions, and the sodium ions present in the injectable solution of pemetrexed were not relevant to the efficacy of the invention.

Actavis, for its part pointed to claim 12, and submitted that it included within its scope a combined preparation of pemetrexed disodium and vitamin B12. Accordingly a formulation chemist would be needed to carry out the invention of claim 12. If that was so then claim 1, being of broadly similar scope, was addressed to the same set of persons.

I have no doubt that the judge was right to find that the patent is addressed to a team which includes a chemist in addition to the oncologist. Firstly, the invention requires the use of the pemetrexed disodium, and the specification of the 508 patent expects the skilled person to be able to obtain it and make it into an injectable solution. It is no answer to say, as Lilly does, that the claim is only framed in relation to manufacture because of a legal fiction concerned with the restriction on patentability of second medical use inventions. However the claim is formulated, the skilled addressee needs to manufacture the medicament and the manufacturing step is an essential requirement of the claim, necessary to prevent the claim falling foul of the method of treatment exclusion from patentability. In the absence of evidence that pemetrexed disodium was generally available, manufacture of the medicament includes making the active ingredient. Secondly, as the judge found, the teams who deal with developing and making medicaments for use in treatment in the real world comprise specialists in a range of disciplines, and in this context would comprise both a medical oncologist and a chemist. Finally, there is a fundamental inconsistency as the judge pointed out, arising out of Lilly’s case that the claim would be understood as extending to the use of active ingredients other than pemetrexed disodium. Assuming for a moment that Lilly is correct, the evidence showed that choice of an appropriate alternative salt would not be something that the medical oncologist could assist with.

I would prefer not to express any concluded view on whether any weight can be attached to Actavis’ argument based on claim 12, which attracted the judge. If the only reason for engaging the discipline in question were the existence of a separate set of claims, or a subsidiary claim, I would be disinclined to hold that the same addressee was necessarily required for all claims. But as I am of the view that claim 1 requires a chemist in any event, I need not explore that question further.

The judge found that the relevant common general knowledge of the oncologist in 2000/2001 included the following:

Antifolates were a class of drugs which were used in cancer chemotherapy. Some drugs in this class, such as methotrexate, had been used for a considerable period of time, but others were under development. There was some understanding of the mechanism of action of antifolates. It was well known that the use of antifolates in chemotherapy caused toxic side effects which it would be desirable to avoid or reduce if possible.

The antifolate pemetrexed was the subject of clinical trials for use in chemotherapy. It operated by targeting multiple enzymes and was administered intravenously.

The only form of pemetrexed which had been shown to be effective and safe, to the extent that this had been shown, was pemetrexed disodium, which was manufactured by Lilly under the trade mark Alimta.

Both vitamin B12 and folic acid had been well known for a considerable period of time, and their characteristics, structure and functions were well understood. It was well known that there were a number of different safe and effective forms of both vitamin B12 and folic acid available.

However the judge found that skilled oncologists did not think about drugs such as pemetrexed in their ionic form, nor did they consider issues regarding the choice of counter-ion or the effect, if any, of counter-ions on the efficacy, safety or other properties of the drug. Both parties’ experts agreed that the choice of salt form was really the province of the chemist and that the oncologist would not be involved in this. Professor Ferry, Lilly’s expert, agreed that the properties of salt forms and free acids were difficult to predict and that the chemist would need to address this problem by conducting experiments.

The judge also made detailed findings as to the common general knowledge of the chemist:

Drugs frequently contain one or more acidic or basic groups. Where this is the case, it is generally possible to form different salts of the parent molecule by reacting it with a complementary base or acid. The salt will typically have different properties from the parent molecule. For example, a salt may be a solid at room temperature, whereas the parent molecule is a liquid; a salt may be soluble in water, whereas the parent molecule is not; and so on. Furthermore, different salts will typically have different properties from each other. For these reasons, salt screening is a routine, but important, part of the process of determining the most suitable form of a drug for formulation.

Formation of a salt involves the transfer of one or more protons (hydrogen ions) from the acid to the base, resulting in a negatively charged species (an anion) and a positively charged species (a cation). The ion which is not derived from the parent molecule is generally referred to as the "counter-ion". Where the parent molecule is an acid, it will form an anion when reacted with a base. The base will provide the counter-ion. Thus pemetrexed diacid reacts with sodium hydroxide to form pemetrexed disodium salt. In this case the counter-ion is sodium, which is a cation.

Solid salts consist of the anions and cations regularly arranged in a fixed lattice structure. In a salt consisting of a single cation and a single anion, there are equal numbers of alternating cations and anions in the lattice. Where there are different ratios of cations and anions, this gives rise to different lattice structures. The different lattice structures in turn give rise to different crystal structures. Although lattices contain infinite numbers of cations and anions, the fact that the cations and anions are present not only in fixed proportions, but also fixed relative positions, means that it is possible to speak meaningfully of the salt as being present in solid form.

When a salt such as pemetrexed disodium is dissolved in a solvent such as water, the ions dissociate from each other and become surrounded by solvent molecules. The result is free cations and anions in solution. It follows that the salt does not exist as such in the solution, but rather there is a solution containing the separate constituent cations and anions. Thus a solution of sodium chloride does not contain sodium chloride, it contains sodium cations and chloride anions. It is commonplace to refer to "a salt solution" or "a salt in solution", but this is a convenient shorthand which is not technically entirely accurate.

The salt form can have a significant impact on the effectiveness of a drug in that it can modify the solubility, therapeutic use, pharmaceutical dosage forms, pharmacokinetic properties (e.g. absorption, distribution, metabolism and excretion of the parent molecule in the body) and the chemical and physical stability of the drug, and its suitability for industrial processing.

In particular, in relation to solubility, if a salt form has poor solubility and dissolution, this can result in poor bioavailability, as good solubility and/or dissolution are indicators of how likely it is that the drug will be absorbed in the gut. When considering a drug for intravenous chemotherapy, the solubility of the salt form is crucial.

By contrast, if a salt is too soluble, then it may not result in direct crystallization or precipitation of the desired salt, and therefore the salt cannot be made in solid form in the first place.

In general, there can be many dead-ends and false leads when attempting to prepare salts of a parent molecule for the first time.

It was common ground between the experts that one could not predict in advance (a) whether one could make a particular salt form of a parent molecule, (b) what its properties would be once the salt form was made or (c) whether it would affect the efficacy of the drug.

Sodium salts. It was well known that sodium was the most preferred counter-ion. Thus sodium would be the chemist's first choice. It was known that sodium salts were generally not toxic. Sodium salts would generally be expected to be reasonably soluble, but they were not always easy to make.

Potassium salts. Potassium was known to be used in pharmaceutical compositions. Although it had a general tendency to be in the "more soluble" class of salt with sodium, there were exceptions to this tendency. It was known that there were some concerns about the potential toxicity of potassium salts in terms of cardiac side effects. This is something which would require particular consideration if large quantities of the drug (such as gram quantities) were to be administered.

Tromethamine salts. Tromethamine salts were very much in the minority in 2000/2001 and there were only a small handful of examples of its use. It is still not particularly high on the list to be used as a counter-ion even now. It was known that tromethamine salts might well be too soluble, such that one would not be able to make and harvest the solid form.

Free acids. In principle, an acidic parent molecule could be administered in the form of the free acid, and this is something that the chemist would consider. It was often the case, however, that there was a need or desire to change from the free acid to a salt form in order to improve kinetics, absorption or physicochemical properties. In particular, the free acid might not be adequately soluble, and a common way to try to address that was through salt formation.

The correct approach to the construction of a patent claim is that explained in the speech of Lord Hoffmann in Kirin-Amgen Inc v Hoechst Marion Roussel Limited [2004] UKHL 46. The claims of a patent specification are not to be approached on the basis that the literal meaning of the claims is to prevail unless it is possible to detect some ambiguity in them. As Lord Hoffmann explained at [29] such an approach, if strictly applied, could result in the court construing the specification in a sense which a reasonable reader, aware of the context and background, would not have thought the author intended. The purpose of the specification (to identify and communicate a novel idea) and the identity of the audience (the addressee of the specification equipped with the common general knowledge in the art) are essential to determining the meaning which would be conveyed by the words chosen in the claims of the patent.

On the other hand, as Lord Hoffmann explained at [34], such a purposive construction does not justify going beyond the definition of the technical subject matter for which the patentee seeks protection in the claims:

Lord Hoffmann also considered the approach which should be taken to the understandable concern, often expressed, that it should not be possible to avoid infringement by making an immaterial variant to that which is expressly claimed, when the variant would have no material effect on the way the invention worked. In the United States, and in some European countries, the courts apply a doctrine of equivalence which extends the scope of protection outside the scope of the claims. Lord Hoffmann firmly rejected the adoption of such an approach into our patent law. On his analysis, such a doctrine was born out of the application of “unsparing logic” and literalism to the interpretation of claims, which were constraining courts to hold that the scope of protection was narrower than the person skilled in the art would reasonably have understood it to be. Although one answer to this problem was to apply a doctrine of equivalence, Lord Hoffmann preferred the abandonment of “literalism” and the adoption of a principle of construction which actually gave effect to what the person skilled in the art would have understood the patentee to be claiming. That step had already been taken before the EPC came into effect in Catnic Components Ltd v Hill & Smith Ltd [1982] RPC 183, when the House of Lords had preferred a doctrine of purposive construction over one which allowed for infringement by “taking the pith and marrow of the invention”. Lord Hoffmann said at [44]:

The fact that English courts do not apply a general doctrine of equivalence to the construction of patent claims does not mean that the existence of equivalents which have no material effect on the way the invention works has no bearing on the proper, purposive interpretation of a patent claim. To the contrary, it has long been the law that such equivalents form part of the background of facts known to the skilled reader which would affect what he understands the claim to mean (see Kirin-Amgen at [49]). It is in this way that English law complies not only with the Protocol to Article 69 of the EPC in its original form, but also with the new article 2 added to the Protocol by the EPC 2000:

English law recognises the impact of equivalents in what became known as the Improver and subsequently the Protocol questions, after their exposition by Hoffmann J (as he then was) in Improver Corporation v Remington Consumer Products Ltd [1990] FSR 181. In that case he said, at page 189:

As Lord Hoffmann stressed in Kirin-Amgen at [52], the Protocol questions are not legal rules, but guides more useful in some cases than in others which in appropriate cases help to decide what the skilled person would have understood the patentee to mean. On the other hand, the basic principles of purposive construction, what Lord Hoffmann called “the bedrock of patent construction”, are applicable in every case. There was a tendency before the decision of the House of Lords in Kirin-Amgen for the Protocol questions to be deployed in every case, regardless of whether a sensible answer could be arrived at without reference to them: so that the tail had, in a sense, started to wag the dog.

In cases where it is useful to ask it, the second Protocol question involves asking whether it would be obvious to the skilled reader that the variant has no material effect on “the way the invention works”. At [75] in Kirin Amgen Lord Hoffmann suggested that in future it might be better to ask, as the German courts apparently do, whether it “solves the problem underlying the invention by means which have the same technical effect”. The advantage of posing the question in that way is that it avoids any assumption that the variant works at all. In both cases, however, it is axiomatic that one is concerned with what would be obvious, in the sense of immediately apparent, to the skilled reader from a reading of the specification informed by common general knowledge.

Thus, for the purposes of the assessment involved in the second Improver question, I would take it as clear that the notional addressee is not presented with information which he cannot derive from the patent or his common general knowledge about whether the variant will in fact have no material effect on the way the invention works. Quite apart from the inference to be drawn from Lord Hoffmann’s observation which I have referred to in the previous paragraph, the Improver questions are no more than an aid to construction in suitable cases. In arriving at the skilled person’s understanding of the language of the claim, it cannot possibly be right to provide him with information which he could not derive either from the specification or his common general knowledge.

Mr Carr submitted that there may be situations, of which the present case is an example, where the missing information is so easy for the skilled person to obtain that it should be taken to have been reasonably available to him when he reads the patent. It may be that the law allows for some flexibility in this respect by analogy with the material which, in the context of construing contracts, is taken to be reasonably available to the parties. However it is not clear to me that this is how the case was advanced before the judge, and as it is a matter on which some evidence might have been of assistance, I do not think it right to explore that way of putting the case any further on this appeal.

Mr Carr also reminded us of two cases, one before and one after the EPC and its transposition into domestic law by the Patents Act 1977. These, he said, illustrated the generous approach to construction which the courts had been inclined to adopt when what the defendant was doing or proposing to do was making use of the underlying invention, albeit that, in the defendant’s hands, the invention was "temporarily masked". The first of these cases was Beecham Group Limited v Bristol Laboratories Limited and others [1978] RPC 153. In that case the defendant imported the acetone derivative of the patented drug, the well known antibiotic ampicillin. The acetone derivative reverted, in the presence of water, to ampicillin itself, which was exclusively responsible for the clinical effectiveness of the defendant's product. The House of Lords, applying what was then called the "doctrine of infringement by taking the pith and marrow of the invention" held that the defendant's drug was an infringement. Although it was literally true that the defendant's drug did not conform to the formula in the claim at the time of importation, that fact ceased to be true as soon as the drug was put to the only use for which it was intended.

The second case was Pharmacia Corp v Merck & Co. Inc. [2001] EWCA Civ 1610; [2002] RPC 41. The patent in suit claimed a chemical species, I will call it X, which was an “enol”. In the presence of water X existed in equilibrium with the keto form of X, the keto form being by far the major form. The allegedly infringing product was the solid keto form. Direct and indirect infringement were both alleged. Even though the words used in the claim had a precise chemical meaning, i.e. the enol form, the Court of Appeal held that there was direct infringement. The court did not therefore need to (and did not) express a view on indirect infringement. Aldous LJ held (at [46] – [47]) that, in order to arrive at a scope of claim which gave fair protection to the patentee it was necessary to take account of what occurred to the substance in the body. There the enol would be in an inseparable equilibrium with the keto form, and it was the composite which would have the therapeutic effect. This would not prevent a reasonable degree of certainty for third parties, as third parties “ would therefore not be surprised if manufacture and sale of the other tautomers which would form the enol in solution, would infringe ”. Arden LJ held that there was infringement because (paragraph [175]) the skilled person would know that the keto form exists in equilibrium with the enol form in solution and that “ it cannot be divided up from the enol in that form ”. The words of the claim were capable of being read as the enol form and the keto form into which the enol form is “ constantly and ineluctably interconverting when in solution .”

It is true that both of these cases are examples of situations in which the courts eschewed literalism in favour of a purposive construction of chemical claims. It is, however, dangerous to draw factual analogies between those cases and the present case. The skilled reader of the patent in the present case would not, in the normal run of things, expect any of the Actavis AIs to convert into pemetrexed disodium when dissolved in aqueous solution, or vice versa. In so saying I put to one side the special factors which are relied upon in support of the allegation of indirect infringement. I therefore do not think that Mr Carr can derive any assistance on his direct infringement case from these two authorities.

It was common ground between the parties that the prosecution history was not inadmissible as a guide to construction.

That common approach was said to be consistent with what Lord Hoffmann had said on the topic in Kirin-Amgen at [35]:

In Rohm & Haas v Collag [2002] FSR 28 at [42] Walker LJ (as he then was) held, obiter, that it was permissible to refer to “objective information about and commentary on experiments which were conducted in response to official observations” in resolving an issue of construction. I note, however, that the material in question was referred to in a note in the specification itself, which means that it is matter which was expressly drawn to the attention of the reader. Beyond that, and beyond the useful summary of the problems associated with taking it into account in the judgment of Jacob J in Bristol-Myers Squibb Co v Baker Norton Inc. [1999] RPC 253 at 274-275, there is no direct English authority on the point. I have to say that a rule which merely discourages reference to material, as opposed to treating it as inadmissible, has obvious practical disadvantages, as in the absence of an exclusionary rule the cost and expense associated with its deployment will almost invariably be incurred. However, as we are not asked to decide that the material is altogether inadmissible, I will, somewhat reluctantly, leave it at that.

The essence of the judge’s reasoning as to the relevance of the file history was given at [111] of his judgment:

The difficulty I feel with endorsing this reasoning is as follows. Firstly it assumes that the skilled reader will always read the prosecution history. I do not see why this should be so, given the limited value which, at least before the judgment in this case, it was generally recognised to have. Secondly, and more importantly, it suggests that the story told by the prosecution history of how the claims came to be drafted as they were will assist the court in preventing abuse of the system. To my mind this will be a very rare case indeed. Unless the acceptance of a restriction in a claim is to operate as some kind of estoppel against the patentee arguing for wider claims (a proposition for which neither side contended and which Jacob J rejected, at least on the basis of domestic estoppel, in Bristol Myers), there will always remain an issue as to whether the applicant needed to accept the restriction notwithstanding that he did so. In those circumstances, the light which the prosecution history sheds on the ultimate question of construction is likely to be extremely limited.

I therefore do not regard it as useful to go to the prosecution history in order to discover that the patentee accepted a restriction to his claim against an objection of lack of support in the specification. It is always open to a party attacking the patent to argue that the claims as sought to be construed by the patentee lack support in the specification: see for example American Home Products v Novartis [2001] RPC 8 at [31]. What purpose does it serve to illustrate this point by showing that the patentee was faced with an official objection to that effect and amended his claims in the light of it? It is still open to the patentee to say that he need not have done so, and the apparent concession he made in prosecution was wrongly made. If it is not open to the patentee so to contend, then the prosecution history is indeed creating a form of estoppel.

In any event, patent offices are usually concerned with patentability, not scope of protection. If an applicant were to conclude every letter by saying that he did not accept that by accepting this or that limitation he was necessarily restricting the scope of protection, no inference could be drawn from his conduct in accepting it. I would be reluctant to put the patent attorneys’ profession to this unnecessary trouble.

There are two issues of construction which I must address. The first, which relates to direct infringement, is whether the claim is limited to use of pemetrexed disodium as Actavis contends, or whether it extends to cover the use of the Actavis AIs as Lilly contends. The second is whether the claim is limited to the use of pemetrexed disodium in solid form, as Actavis contends, or whether it extends to a solution which contains pemetrexed ions and sodium ions, as Lilly contends. This second issue is of relevance to the argument about contributory infringement.

I can take this point relatively shortly, as I agree with the judge. The parties have approached the trial and this appeal on the basis that it is a case where the answer is provided by asking the three Protocol questions.

It is common ground that the variants represented by the various Actavis AIs have no material effect on the way the invention works. The judge explained that this was for two reasons. Firstly, from the oncologist’s perspective, the active anti-cancer principle in an aqueous solution of pemetrexed disodium for intravenous administration is the pemetrexed anion. From the oncologist’s perspective the source of that anion is immaterial, since it will not affect the efficacy or safety of the medicine for the treatment of cancer. Secondly, from the chemist’s perspective, it did not lie in Actavis’ mouth to contend that the Actavis AIs were not all bioequivalent to Alimta, as this was the basis on which they would be granted marketing approval. Thus the first protocol question must be answered favourably to Lilly.

In order to succeed on the appeal Lilly must additionally persuade me that, it is right on the second and third Protocol questions. Thus I must first consider whether the judge was wrong to hold that it would not be obvious to the skilled team that the variants represented by the Actavis AIs would have no material effect on the way the invention worked.

The judge had the benefit of expert technical evidence on this subject. He concluded that the crucial member of the team for this purpose would be the chemist, as the oncologist would have no idea as to the effect of the substitutions in question. The view of the chemist would be that he would not be able to predict the effect of the substitution without testing at least the solubility of the AI in question. Although the chemist would be reasonably confident that he would come up with a substitute for the sodium counter-ion, predicting in advance whether any particular counter-ion would work was not possible. In those circumstances, given that getting an effective amount of pemetrexed anion into solution was essential, it could not be said that the variant had no material effect on the way the invention worked. The same answer was arrived at if one asked whether the variant solved the problem underlying the invention by means which had the same technical effect.

Mr Carr attacks the judge’s conclusion with three main arguments. The first argument was that the skilled team did not include a chemist, and that the only response of relevance was that of the oncologist. This argument turns on the composition of the skilled team. I have already concluded that the judge was right to say that the skilled team included a chemist.

Mr Carr’s second argument is that when asking the second Protocol question about the Actavis AIs it is permissible to assume that the skilled chemist knows that the AIs are soluble. In support of this first argument he further submits that Actavis have committed large sums of money already to its project (including large sums in litigation costs) on the basis of its assumption that its AIs will be satisfactory substitutes for pemetrexed disodium.

I cannot accept this argument. The skilled reader of the patent simply does not know, and cannot predict whether variants such as the Actavis AIs will be sufficiently soluble to work. It is not a legitimate exercise to feed information about the Actavis AIs to the skilled team to help them to deal with their technical concerns about the workability of variants. The answer to such concerns must, as I have explained above, be found in the specification or from the team’s common general knowledge.

Mr Carr’s third argument is that the judge focused on the wrong problem when considering whether it was obvious that the variant had a material effect on the way the invention worked. The problem underlying the invention was not a manufacturing one, but a therapeutic one. The oncologist would know that the source of pemetrexed anions would have no effect at all in this context.

The judge recognised the force of this argument but, at [128], rejected it in the following terms:

I agree with the judge’s reasoning. Once the claim includes the step of manufacturing a medicament for treating a disease, it necessarily includes a requirement that the manufactured medicament is to some extent effective for treating the disease. If the skilled chemist is unable to predict that a variant will be sufficiently soluble to deliver an effective amount of pemetrexed anions in solution, then he is unable to say that the variant would have no material effect on the way the invention works.

I also think that the judge was right to hold, in answering the third Protocol question, that the skilled team would understand that the patent was clearly limited to the disodium salt, and did not extend to the diacid, or the dipotassium or ditromethamine salts. I would give the following reasons, many of which were relied on by the judge:

As I have pointed out when summarising the disclosure of the patent, the specification contains certain passages where the invention is described in very general "class" terms and others where the invention is clearly limited to pemetrexed disodium. When the reader comes to the claims, therefore, he or she will appreciate readily that the patentee has chosen to claim narrowly and by reference to a single chemical, and not broadly by reference to any class.

Pemetrexed disodium is a highly specific chemical compound. Putting aside for present purposes the precise form in which the compound is present, there is no obvious leeway as a matter of language for giving it a broad as opposed to a narrow construction.

The only escape from the above would be to say that pemetrexed disodium would be understood by the skilled person to be used in a figurative sense, so as to denote the best known member of a class. It is true that sodium is a well known and perhaps the best known counter ion for use in circumstances such as these. But if the claim is not limited to the sodium salt, there are great difficulties in ascertaining what that class might be. Anti-folates were a known class, but the claim cannot be to all anti-folates, because that term is used in the specification, but not chosen in the claims. As to pemetrexed, the disodium salt was the only salt known to be efficacious and safe.

The only data contained in the specification are for pemetrexed disodium, and broader claims therefore lack support and might have been unacceptable to the EPO.

Importantly, there is a striking contrast between this very specific language and the general terms used in the claim for the methyl malonic acid lowering agent (any “pharmaceutical derivative”) and the folic acid components (any “physiologically available salt or ester thereof”) which the skilled reader could not fail to notice.

The skilled reader would have understood that there are plausible reasons why the patentee might have wished to limit to the disodium salt: for example he might have been content with a claim limited to his commercial embodiment ALIMTA.

For the reasons I have given I would not have found it necessary to go to the prosecution history to arrive at this conclusion when one can get there, straightforwardly, by reading the claims in the light of the specification.

It will be recalled that the judge held, when dealing with the common general knowledge, that it is not strictly technically accurate to say that a salt can exist in solution. Once a salt is dissolved, it dissociates into its constituent ions and any organised lattice structure which exists in the solid form disappears as the ions are dispersed in the solvent medium. The judge nevertheless recognised that it was commonplace to refer, for example, to a solution containing sodium ions and chloride ions as being sodium chloride in solution.

Subject to an irrelevant dispute about whether pemetrexed disodium is ever in a lattice structure in the solid form (as opposed to being amorphous) Mr Carr accepts, indeed embraces, these propositions. He goes on to submit that there are places in the specification of the 508 patent where it uses the term “pemetrexed disodium” not in its strict, technically accurate sense, but in the looser sense to which the judge refers. In particular the term is also used to refer to a solution in which there are sodium ions and pemetrexed ions in at least the same proportions as in pemetrexed disodium. This is made plain, amongst other places, at [35] in the specification, where the patentee refers to treating patients with an intraperitoneal injection of pemetrexed disodium. This can only refer to an injection of a solution containing pemetrexed ions and sodium ions in dissociated form, as the solid disodium salt will necessarily have dissociated. This is something which the skilled team would immediately have recognised.

Thus far I think Mr Carr is obviously right. His construction is supported by a consideration of what the term “medicament” in the claims of the patent would be understood to include. Although one form of the medicament covered by the claims of the 508 patent would be a solid dosage form, it is obvious that another form is an injectable solution, arrived at by dissolving pemetrexed disodium in water. The skilled person reading claim 1 of the patent would consider that pemetrexed disodium was present in the medicament, even though, as a matter of strict (one might even say literal) chemical nomenclature that is not a technically correct way of putting it. To put it another way, if one asks what, for the patentee’s purposes, constitutes the presence of pemetrexed disodium, the skilled reader would understand that it included a solution containing pemetrexed ions and sodium ions.

I recognise of course that the solution of pemetrexed ions and sodium ions that the skilled person would have in mind as one version of pemetrexed disodium covered by the claim could be said to arise in particular and limited circumstances. Thus, firstly, the solution contemplated by the examples of the specification will presumably (although this is not spelled out) have been made by taking solid pemetrexed disodium and dissolving it. One question is therefore whether an identical solution containing pemetrexed and sodium ions, but arrived at in a different way, would also be thought by the skilled reader to fall within the ambit of the language. I consider that it would be. I can see no reason why the skilled reader who understands from the specification that a solution containing pemetrexed ions and sodium ions is regarded as pemetrexed disodium should assume that the method by which that solution is prepared was of concern to the patentee. That is a matter of history, and has no bearing on the essential characteristics of what is called for by the term.

A second question is whether the skilled person would understand that the ratio of sodium ions to pemetrexed ions in the solid state (i.e. 2:1) should be maintained in the solution. I am fully prepared to accept that the ratio must reach the threshold of 2:1: anything less would not meet the description “disodium”. However I cannot see why the presence of more sodium ions, i.e. in excess of the 2:1 ratio should be regarded as material. Pemetrexed disodium in the sense called for by the claim would still be present, despite the presence of more sodium ions.

A final question is whether the presence of ions other than sodium could take the solution out of the claim. Again, I do not see why this should be so. A patent claim is normally to be construed as setting a set of requirements which must be present in the alleged infringement. Infringement is not generally avoided if the alleged infringement has all those features and adds something more. As Bowen LJ colourfully put it in Wenham Gas Co Ltd v Champion Gas Lamp Co (1892) 9 RPC 49 at 56 “the super adding of ingenuity to robbery does not make the operation justifiable”.

It follows from what I have said about the proper construction of the claim that it does not extend to pemetrexed diacid or any other pemetrexed salts other than disodium. On this basis there can be no direct infringement by Actavis by dealing in the Actavis AIs.

Section 60(2) of the Patents Act 1977 provides, so far as relevant, as follows:

The section therefore has potential application where the invention is not put into effect by the defendant himself, but, for example, by someone who is supplied with means relating to an essential element of the invention (“essential means”). Lilly’s case is that by supplying the Actavis AIs, Actavis are supplying an essential means for the doctor or pharmacist who makes up the solution to put the invention into effect.

If all that the doctor or pharmacist were doing was making an aqueous solution of one of the Actavis AIs, Actavis’ case of indirect infringement would add nothing to its case of direct infringement. A solution made by dissolving pemetrexed dipotassium in plain water results in a solution which contains only potassium ions and pemetrexed ions, and is therefore not within the claims because sodium ions are missing. However in the present case the facts may give rise to indirect infringement because, in the ordinary course, each of Actavis’ AIs will be dissolved and/or diluted in saline, which is of course a source of abundant sodium ions.

The judge summarised the relevant facts in the following way at [194] to [198] of his judgment, starting with the instructions for reconstituting Lilly’s product Alimta (for reasons which appear):

On the basis of these facts Lilly submits that Actavis would be indirectly infringing the patent. First, Lilly claims that each of the AIs is an essential means. Then, to take the example of pemetrexed dipotassium, they contend that when the solid is reconstituted in saline there will be present pemetrexed ions and sodium ions just as there are pemetrexed ions and sodium ions present when pemetrexed disodium is reconstituted. The fact that there will be potassium ions and chloride ions in addition is of no relevance, as is the fact that the sodium and pemetrexed ions have a different history. In the course of the procedure outlined in the statement of product characteristics pemetrexed disodium (in the loose, non-technical sense that that expression is used in the 508 patent) is formed. Moreover it will be used in the final stages of manufacture of the medicament.

The judge dealt with this issue in the following way:

I cannot agree with the argument advanced by Actavis which the judge accepted. The construction of pemetrexed disodium which that argument advances appears to assume that it is a solid. I say this because it proceeds from the finding that pemetrexed disodium “means pemetrexed disodium and not any form of pemetrexed that is pharmaceutically acceptable and sufficiently soluble”. It is thus relatively easy for the judge to accept the submission that “at no point is pemetrexed disodium used in the manufacture of a medicament by anyone”. It is undoubtedly the case that solid pemetrexed disodium is not used. However, as I have explained, the skilled team, particularly the chemist, would understand from the 508 patent that pemetrexed disodium is also used to refer to solutions which contain pemetrexed ions and sodium ions in solution.

The judge plainly regarded the fact that no solid pemetrexed disodium was used as conclusive of the indirect infringement argument. The judge then accepted Actavis’ characterisation of Lilly’s case as being that pemetrexed ions on their own constituted an essential element of the invention, and dismissed this on the basis that it was just another way of saying that the claim does not require pemetrexed disodium, but merely requires any form of pemetrexed which makes pemetrexed ions available. I do not think that this analysis does justice to Lilly’s argument. Lilly’s argument does not involve saying that the claim does not require pemetrexed disodium. It asserts that it does require it, and that pemetrexed disodium is used in the process of making up the medicament.

It is possible, although I doubt, that the judge thought that pemetrexed dipotassium and the other Actavis AIs could not be a “means relating to an essential element of the invention” because they did not themselves represent a free standing feature or element of the claim. If this was the judge’s reasoning I cannot accept it. The language of section 60(2) does not require the supply of an element of the claim, but a means relating to an essential element. A means for releasing pemetrexed ions into solution relates to an essential element of the invention where the invention calls for pemetrexed ions and sodium ions in solution, particularly as it is the presence of the pemetrexed ions in the manufactured medicament which is essential for its efficacy as a medicament. The invention is then put into effect when the pharmacist makes up the solution using pemetrexed dipotassium (or the other Actavis AIs), because there comes a stage in the course of that activity when pemetrexed disodium is present and is used.

In Grimme Maschinenfabrik v Scott [2010] EWCA Civ 1110 this court recognised that the essential means of section 60(2) did not have to be something which could be used without alteration by the buyer. In that case the defendant was selling whole machines with steel rollers which the buyer could adapt by incorporation of rubber rollers. The argument that this could not be an essential means was rejected. As Jacob LJ put it at [103]:

Applying this reasoning, pemetrexed dipotassium is a means relating to an essential element of the invention, notwithstanding that potassium is dissociated and sodium caused to be present in order to put the invention into effect.

I would therefore disagree with the judge on the issue of contributory infringement. I would refuse Actavis their DNIs as a matter of English law on these grounds.

It was common ground both before the judge and before us that there was no detectable difference in the laws of France, Italy and Spain on the approach to contributory infringement. It follows that the declarations should also be refused in respect of those countries.

It is thus unnecessary for me to decide whether the courts of France, Italy and Spain would arrive at the same result as me on the issue of direct infringement, applying their respective laws. I would nevertheless record my conclusion, having heard Lilly’s arguments, that the judge was right to hold that there would be no direct infringement applying the laws of those countries either.

Actavis had originally included the German designation of the 508 patent within its action for DNIs. Despite the fact that the court first seised was England, Lilly brought proceedings against Actavis for threatened infringement of the German designation in the Landgericht Düsseldorf. However, in its judgment of 3 April 2014 delivered a few days before the start of the trial before Arnold J, the Landgericht Düsseldorf rejected Actavis’ jurisdictional challenge and proceeded to give judgment on the merits in favour of Lilly. Actavis launched an appeal in Germany against the Landgericht Düsseldorf’s judgment, but decided to discontinue the part of its claim in this jurisdiction which related to the German designation. At the trial, Lilly pressed Arnold J with the reasoning of the Düsseldorf court, but at paragraphs [150] to [158] of his judgment, the judge explained why he was not convinced by it.

In its written skeleton argument for this appeal Lilly renewed its reliance on the judgment of the Landgericht Düsseldorf. However, shortly before this appeal was heard, the appeal court, the Oberlandesgericht Düsseldorf, allowed Actavis’ appeal.

On the issue of direct infringement, therefore, the Oberlandesgricht and I are in agreement. No purpose would be served by a comparison of the routes by which we have arrived at this result.

By comparison with its reasoning on direct infringement, the issue of indirect infringement is covered shortly in the judgment of the Oberlandesgericht. The issue is dealt with at pages 35 to 36 of the translation. The court holds, firstly, that the “essential means” must be configured in such a way that a direct use of the invention is possible. However that was not the case in the present circumstances. Although the court recognised that a solution formed from pemetrexed dipotassium results in dissociated ions of potassium and pemetrexed, and that normal saline will include sodium ions in addition, it concluded that pemetrexed disodium was at no time used by either Actavis or its buyers for preparing a medicament. Arnold J’s similar conclusion was mentioned by the court in this context.

Whilst the decision of the Oberlandesgricht is entitled to great respect, including as it does distinguished patent judges, I am not persuaded by its reasoning to change the view which I have formed on the issue of contributory infringement. Firstly, if by its requirement that “the means must be configured in such a way that a direct use of the invention is possible” it is indicating that pemetrexed dipotassium cannot be an essential means, then that result is inconsistent with Grimme and it is not open to us to follow it. Secondly, it seems to me that the Oberlandesgericht has failed to give sufficient recognition to the fact that, applying the teaching of the patent, it is sufficient if one finds in the medicament in question sodium ions and pemetrexed ions in solution in a ratio of at least 2:1. Rather, as I read the judgment, the court appears to understand “pemetrexed disodium” as describing only that substance in solid form. I have explained why I do not agree with that construction.

As I have concluded that the proposed use of the Actavis AIs would amount to contributory infringement of the 508 patent, Actavis are not entitled to the DNIs sought and it is not necessary for us to decide the Rome II issue. However, as we have heard full argument on what is potentially a point of some importance, I will deal with it.

A party who applies for a DNI of a UK national patent or a European patent (UK) may do so by following the statutory procedure prescribed in section 71 of the Patents Act 1977. That section requires it to be shown that the party seeking the DNI has applied in writing to the proprietor for a written acknowledgement to the effect of the declaration claimed, provided full particulars in writing of the act in question, and that the proprietor has refused or failed to give any such acknowledgement.

Section 71 is expressly without prejudice to the court’s power to make a declaration apart from that section. As this court made clear in Messier-Dowty Ltd v Sabena SA [2000] 1 WLR 2040 at [41] the court has a general jurisdiction to make declarations of non-liability where such a declaration would serve a useful purpose. Actavis sought its DNIs under this more general jurisdiction, as the jurisdiction under the Patents Act does not appear to extend to foreign patents.

It is common ground that Actavis is entitled to bring its proceedings for a DNI in respect of the UK designation of the 508 patent under the court’s general jurisdiction as it has a sufficient commercial interest in knowing whether the use of the Actavis AIs would infringe, and the declarations would therefore serve a useful purpose. The Rome II issue is whether Actavis, by satisfying the lex fori (i.e. English law) test for the availability of a declaration, can obtain DNIs in respect of the French, Italian and Spanish designations of the 508 patent as well. Actavis contend that they can, and that this is the result of a correct application of Rome II. Lilly contends that Actavis must satisfy the requirements of the foreign law in the designated states, the lex causae .

The starting point is the Rome II regulation itself, so I set out the material provisions below. Recital (6) states:

Article 1 provides:

Articles 21 and 22, which are referred to in Article 1(3), make specific provision for formal validity of unilateral acts intended to have legal effect, and burden of proof. Article 8 provides:

Article 15 provides:

Actavis contend that the rules for obtaining DNIs are matters of procedure within Article 1(3) and hence fall outside the scope of the Regulation. Lilly contends that those rules are not procedural. The rules are either within Article 15, or are closely analogous to the matters specified in Article 15. In either case the rules are those found in the law applicable to non-contractual obligations which, under Rome II, is the lex causae.

If Actavis are right that the matter falls outside the scope of the Rome II Regulation, the applicable law will be determined by the English common law rules of private international law. In that event it is common ground that the applicable law is the lex fori, because English law regards the rules for obtaining negative declaratory relief as being procedural: see Messier-Dowty Ltd v Sabena SA [2000] 1 WLR 2040.

It would appear from the materials before us that English law takes the most relaxed attitude amongst the relevant designated states of what must be shown before a party can apply to the court for a DNI. The laws of Spain, France and Italy lay down more specific requirements as follows.

In Spain, Article 127 of the Spanish Patents Act 1986 provides as follows:

In France, Article L 615-9 of the Intellectual Property Code provides as follows:

In Italy, Article 100 of the Code of Civil Procedure provides:

It is common ground that this general provision of the Code of Civil Procedure applies to claims for DNIs of patents. It is also common ground that the test for a legitimate interest in relation to a claim for a DNI is that there is “an objective uncertainty giving rise to a present, concrete prejudice to the claimant which the judgment of the court is capable of curing”. It is also common ground that there is no need for a cease-and-desist letter to have been sent.

The judge dealt with the Rome II issue at [204] to [236] of his judgment. He concluded, in accordance with Actavis' submissions, that the national rules which specify the conditions which must be satisfied by a claimant in order to obtain a DNI fell into two categories. Firstly there were those rules which provided for a fact sensitive requirement of interest or purpose, as did the requirement under the general jurisdiction in English law. Secondly there were rules which required pre-action notification together with details of the proposed acts of the claimant, as in section 71 of the Patents Act 1977. The first type of rule was concerned with ensuring that the claimant had sufficient practical justification for seeking the adjudication of the court, and thus with avoiding the time of the court being taken up with academic questions. The second type of rule had two purposes. The first was to allow the patentee to give an acknowledgment of non-infringement, and thus avoid unnecessary litigation. The second was to ensure that the dispute, if it is to go forward, is sufficiently well defined for the court to adjudicate on it.

At [220] the judge said:

The judge next focused on Article 15 of Rome II, and in particular on Lilly's argument that the DNI was a remedy within Article 15(c). He rejected Actavis' submission that “remedy” in that article was limited to damages, holding that it extended to all financial remedies. He also rejected Lilly's argument that “remedy” extended to anything which might be described as a remedy, thus including DNIs. In this connection he doubted whether a DNI was a remedy in the relevant sense at all. He concluded, in any event, that there was a distinction to be drawn between the availability in principle of a remedy and the steps which must be taken in order to obtain that remedy. It was only the former which was caught by Article 15 and made subject to the lex causae .

The judge thus concluded that the rules in issue here were matters of procedure and thus not subject to the lex causae .

Mr Mitcheson QC, who argued this part of the case on behalf of Lilly, submitted, firstly, that it was no longer appropriate to give weight to the convenience involved in applying the lex fori as opposed to the lex causae . Even if that were appropriate under English common law conflict rules, it had been all but swept away under Rome II. Rather the approach should be that the court should apply the lex causae unless it was inconvenient or inefficient to do so. The lex causae should apply except in limited matters concerned with the constitution of the courts and the method of trial which were deeply embedded in the procedural system of the domestic court, in the sense that they were integral and indispensable features of that country's legal system such that they cannot be replaced by corresponding features of the lex causae .

Mr Mitcheson went on to submit that a number of paragraphs of Article 15 either covered or were closely analogous to the rules in issue here:

There was an “echo” of the DNI in paragraph (a), when it refers to “the basis and extent of liability, including the determination of persons who may be held liable for acts performed by them”. By a DNI a party sought to be determined to be a person who could not be held liable for acts performed by it.

The rules in paragraph (c) are about “the existence .. of a remedy”. A DNI was a remedy, but even if that is wrong and paragraph (c) is limited to monetary compensation, the rules were still concerned with the existence of a remedy for Lilly, because if a declaration is granted any monetary relief which would otherwise be available to Lilly is ruled out.

Paragraph (d) was also relevant because what Actavis were seeking to do was to insulate themselves against a remedy in damages, or alternatively to terminate the prejudice caused by the uncertainty of not knowing whether or not they infringed.

Finally, there was an analogy with limitation rules under paragraph (h).

Next, Mr Mitcheson submitted that a helpful test is "is the rule concerned with a decision on the merits?" All the matters here are “concerned with” a decision on the merits, as the foreign law questions were concerned with determining issues of scope of claim and infringement. If the court is applying foreign law to those issues, it is sensible and convenient to apply foreign law to the question of whether Actavis would have a cause of action at all for the DNI which it seeks. Actions for infringement by the patentee and actions for DNIs are, for the purposes of Article 21 of the Brussels Convention, recognised as actions between the same parties in respect of similar or identical subject matter. Viewed in that light the law which governed them should be the same. Mr Mitcheson’s submission that an issue is governed by the lex causae if it is “concerned with the decision on the merits” is derived from an article by Dr Martin Illmer in Civil Justice Quarterly Volume 28 Issue 2 pages 237-260. Dr Illmer proposes a criterion of “neutrality” as the autonomous criterion for allocating rules to the lex fori. Applying this criterion, rules which are regarded by national law as procedural, but which are nonetheless so closely intertwined with the rules governing the material dispute that their non-application would frustrate the rights and remedies under the applicable law, may be governed by the lex causae. Dr Illmer recognises that it is not every rule which might affect the outcome of the case which is to be made subject to the lex causae, only those which are “directed at” the decision on the merits. Dr Illmer suggests that:

Viewed in this way, one would ask whether the rules concerning the availability of a DNI are directed at the decision on the merits, and so intertwined with the merits as to affect the outcome as to require the application of the lex causae, or whether they are concerned on the other hand with the commencement of proceedings and the formalities of bringing a claim, so as to fall under the lex fori.

Mr Mitcheson also relied on what the authors of Dicey, Morris & Collins The Conflict of Laws, 15^th Edn (“Dicey”) say at 34-036. They suggest that the Article 1(3) exclusion of evidence and procedure should be construed narrowly as covering only matters such as the constitution and powers of courts and the mode of trial, that are an integral and indispensable feature of the forum’s legal framework for resolving disputes, such that they cannot satisfactorily be replaced by corresponding rules of the lex causae.

Further, Mr Mitcheson submitted that the judge had been wrong in his characterisation of the rules. The rules in issue on the appeal were about whether the party had done sufficient acts to raise the issue of infringement at all, and were at the heart of the substantive rights given by the patent. Without formally conceding that the rules of the second type identified by the judge (written acknowledgments, full particulars in writing etc) were procedural, he submitted that the rules of the first type, which required a fact sensitive inquiry into the state of the claimant's actual preparedness to exploit the invention, were not. Such rules went to the availability of the remedy and not to procedural steps necessary to obtain it. Thus Mr Mitcheson submitted that, even if the judge was right to distinguish between the availability of a remedy in principle and the steps necessary to obtain it, the rules in question fell into the former category.

Mr Raphael QC, who argued this part of the case for Actavis, drew our attention to the relevant parts of the legislative history of Rome II. The Commission’s original proposal for Rome II did not include the exclusion now contained in Article 1(3) for evidence and procedure, which it regarded as implicit in the original draft of Article 15 (then Article 11). The commentary in the proposal for what was then Article 11 explains that the approach taken in Member States to the listed questions was not entirely uniform, and so the step was therefore taken, as was done in Rome I, of listing the questions to be settled by the law actually designated.

The Commission’s commentary on the various paragraphs of Article 11 (now 15) is informative. Thus Article 11(a) was said to be particularly concerned with “intrinsic factors of liability”, i.e. nature of liability (strict or fault-based); the definition of fault; causation; persons potentially liable, and division of liability between joint perpetrators.

It is also worth noting that, in relation to what is now Article 15(c) (then 11(c)) the language of the proposal was simply “the existence and kinds of injury or damage for which compensation may be due”. There was no use of the more general term “remedy” until later. The word “redress” was first used (and said to be a “technical, grammatical and terminological correction and distinction”), and then substituted by “remedy” without further explanation.

A report on the proceedings in the Parliament on the Commission Proposal records the proposal for the inclusion of Article 1(3), as it now is, in the following terms:

Mr Raphael submitted that the position had been correctly stated by Professor Halfmeier of the Frankfurt School of Finance & Management, Department of Law and Ethics in his contribution to Calliess Rome Regulations – Commentary on the European Rules on the Conflict of Laws (Kluwer 2011) when he said at page 391:

Article 1(3) of Rome II is a rule about what is sometimes called the “vertical scope” of the Regulation. Evidence and procedure are excluded from the scope of the Regulation. Although it does not automatically follow that these issues will be subject to the lex fori, the private international law principle that such matters are for the law of the forum is well recognised. It is enough to quote Dicey at paragraph 7.002:

Article 15 of Rome II is not itself directly concerned with clarifying the distinction between substance on the one hand and evidence and procedure on the other. It simply contains a list of matters which are “in particular” to fall under the designated law. Included in the list are matters, such as limitation periods, which were traditionally the subject of some debate as to whether they were substance or procedure. Article 15 does not answer that question, but merely declares that they will be subject to the law which governs non-contractual obligations under Rome II. I therefore do not regard Article 15 as a safe guide to whether matters which do not fall within its scope are procedural or substantive.

The distinction between substance and procedure is a fundamental one. The principle underlying it is said to be that a litigant resorting to a domestic court cannot expect to occupy a different procedural position from that of a domestic litigant. Thus, that litigant cannot expect to take advantage of some procedural rule of his own country to enjoy greater advantage than other litigants here. Equally he should not be deprived of some procedural advantage enjoyed by domestic litigants merely because such an advantage is not available to him at home. Thus, at common law, every remedy was regarded as procedure: see for example Don v Lippmann (1837) 2 Sh. & MacL. 682 at 724-5.

Whether a rule is to be classified as one of substance or one of procedure or evidence under Rome II is a matter of EU law: the fact that a rule is classified as one or the other under domestic law is of no relevance. There is therefore a need for an autonomous EU criterion for allocating rules into one or the other category.

In Wall v Mutelle de Poitiers Assurances [2014] EWCA Civ 138; [2014] 1 WLR 4263, the claimant motorcyclist was injured in a motor accident in France. He claimed damages against the other driver’s insurers in England. Liability, which was governed by French law, was admitted. The question arose as to whether the claimant should be permitted to adduce expert evidence in accordance with English practice, or whether a single joint expert should be instructed, as would be the practice in France. This court held that the issue of which expert evidence the court should order was one of “evidence and procedure” within Article 1(3) and not an issue relating to “the existence, the nature and the assessment of damage” within Article 15(c) of Rome II. It was argued that the objective of the Regulation was to ensure uniformity of outcome, and that the English court should do its best to ensure that uniformity by adopting all the rules of the foreign court which might affect outcome. The court rejected that argument (see Longmore LJ at [11] to [14] and Jackson LJ at [40] to [43]), holding that it was inevitable that the same facts tried in different countries might achieve different outcomes. The words “evidence and procedure” were thus given what Jackson LJ called their “natural meaning”.

In my judgment, subject to any impact on the question which Rome II may have had, the rules with which we are concerned are conditions of admissibility of actions, rather than rules concerned with the substance or content of parties’ rights. They are all concerned with whether the court should hear a dispute about substance. They are not concerned directly with the substance itself. Thus:

a rule about the need to seek an acknowledgement from the patentee will avoid the dispute coming to court if the acknowledgment is given;

a rule requiring the giving of particulars will ensure that the proposed act is sufficiently formulated for the court to be able to adjudicate on whether it infringes;

a rule requiring some form of interest, or degree of preparation, will avoid cases coming to court if the party seeking the DNI has not reached a stage where it has sufficiently formulated its plans;

a rule requiring that the party seeking the DNI can show that it would serve a useful purpose avoids the court adjudicating on pointless disputes.

Such rules would traditionally, for private international law purposes, be classified as procedural and not substantive. In my judgment, therefore, they should continue to be so treated unless Rome II requires a different outcome.

I do not think Mr Mitcheson’s argument based on Dr Illmer’s illuminating article displaces this view. As Dr Illmer himself recognises, “matters of procedure concerned with the commencement of the proceedings” will continue to be governed by the lex fori. I consider that the rules with which we are concerned fall within that description. They are not so intertwined with matters of substance as to require them to be dealt with under the lex causae. Whilst the passage from Dicey on which Mr Mitcheson relies suggests a very narrow interpretation of “evidence and procedure”, the authors nevertheless say at 7-072:

Whilst rules which require an interest, or effective preparations, are different, I can see no reason in principle why they should not be categorised in the same way. As the judge observed, they have the same broad purpose as the more formalistic rules to which Dicey expressly refers, and are quite distinct from the rules which govern the parties’ substantive rights.

I do not accept that Article 15 should be given a wider effect than its language suggests, treating the listed matters as no more than examples of a class of analogous matters regarded as procedural in private international law, but now to be brought within the designated law. Mr Raphael is right that the legislative history shows that the Regulation was intended to respect the private international law principle that the ‘lex fori’ is applicable to procedural questions.

Although Article 15 applies the lex causae to a number of matters which at least the English common law would have treated as procedural, none of them, as it seems to me, is apt to encompass the rules for admissibility of a DNI. I take these in turn.

Paragraph (a) is concerned with the basic conditions and extent of liability under a non-contractual obligation, and the persons who may potentially be held liable. Whilst Mr Mitcheson’s attempt to fit the negative declaration into the wording of the paragraph is ingenious, it does not seem to me that, even if correct, it gets him home. That is because the conditions of admissibility of a a positive claim are not caught by the section. If that is so, then I fail to see how the conditions of admissibility of a DNI can be caught either.

The problem with reading paragraph (c) as widely as Mr Mitcheson contends is that it covers any remedy, when the legislative history shows it was concerned with financial remedies alone. Moreover, as the judge pointed out, other language versions of paragraph (c) use words which translate as “compensation”, “indemnity” or “reparation”. Reading it more broadly, the domestic court could find itself having to apply remedies of a nature unknown to its law. This would be in stark contrast to paragraph (d) which covers specific remedies aimed at preventing or terminating injury or damage, but which are limited by the opening words “within the limits of the powers conferred on the court by its procedural law.” To my mind, the negative declaration, whilst no doubt a remedy, is not a remedy which falls within (c).

The negative declaration is also not within (d), because it is not a measure which the court takes to prevent or terminate injury or damage, or provide compensation. Unlike an injunction to prevent infringement, it cannot be said that a characteristic of a DNI is that it prevents injury or damage. Moreover paragraph (d) is again concerned with the availability of such remedies, not the conditions which must be satisfied for their admissibility.

Finally, the mention of limitation periods in paragraph (h) is not a basis for suggesting that the conditions of applying for a DNI should be brought within the lex causae.

It follows that, had we needed to decide the point, I would have agreed with the judge that Rome II does not result in the conclusion that the lex causae applies to the conditions for applying for a DNI. Those conditions are procedural, and subject to the lex fori.

It is therefore, as it was for the judge, unnecessary for us to decide the factual issues which arose in relation to the application of the lex causae conditions for admissibility of the DNIs. Nevertheless, I should record that I was not persuaded that the judge had made any error in his assessment of the evidence of foreign law or its application to the facts.

For the reasons I have given relating to contributory infringement, I would allow Lilly’s appeal and set aside the declarations of non-infringement of the 508 patent by the Actavis AIs.

After this judgment was circulated to the parties for suggested editorial corrections, and very shortly before it was due to be formally handed down, counsel for Actavis wrote to the court with a request that we amend or qualify the draft judgment before it was handed down. The letter said:

The letter went on to say that the court “ought to” make a modified declaration which reflects the court’s finding “so that it is clear that it would be lawful for Actavis to sell its AIs to the extent that they are not intended for reconstitution with any source of sodium ions”, and in particular if they are for reconstitution with dextrose instead of saline. The letter also ventilated other possible scenarios, including one in which the AI was for reconstitution with saline in a ratio of less than 2:1.

Actavis went on to submit that, given the supposed need for alternative declarations, the court should consider revising its judgment so as formally to decide not only the Rome II issue but also the issue of whether there is direct infringement of any of the other foreign designations in issue (as opposed to expressing obiter opinions). This was on the ground that if the court were to make a declaration that there was no direct infringement in respect of the UK designation, it could also do so in respect of the foreign designations as well.

In the light of these suggestions for significant changes to the draft judgment it was no longer possible to hand the judgment down on the originally intended date of 28 May 2015. On 3 June 2015 Actavis filed voluminous “Post Appeal Submissions”. In its draft order Actavis proposed a declaration for the UK in two parts. The first part declares that acts done in relation to the Actavis AIs would not be an infringement of the 508 patent under s. 60(1) of the Patents Act 1977. The second part of the declaration contains a positive declaration that supplying or offering to supply the Actavis AIs would be an infringement of the 508 Patent under section 60(2) of the Act “if and only if Actavis knows, or it is obvious to a reasonable person in the circumstances, that such Actavis AI is suitable and intended for dilution or reconstitution in the United Kingdom so as to achieve a molar ratio of sodium to pemetrexed ions of 2:1 or more”.

Actavis also propose additional separate declarations for each of the foreign designations, referring in each case to the relevant national designation and the relevant national statutory provisions, again split into two parts, namely (a) a negative declaration for direct infringement and (b) a positive declaration for indirect infringement, “if and only if” Actavis have the relevant knowledge.

In its Post Appeal Submissions, however, Actavis have slightly amended all the proposed declarations for indirect infringement by changing “if and only if” to “if”.

Lilly opposes the making of these or any modified declarations. I will call this issue “the declaration issue”.

Actavis also ask the court to remit for trial in the Patents Court an issue concerned with their proposed new product for reconstitution with dextrose. The issue is (a) under what circumstances and with what frequency, if any, persons administering Actavis AIs to patients would dilute or reconstitute them with saline if directed to do so with dextrose and not with saline, (b) to what extent the same would be known or foreseeable to Actavis, and (c) any appropriate relief as a result of its findings on (a) and (b).

Lilly opposes this order for remission. I will call this issue the “dextrose remission issue”.

I would not make any of the alternative declarations proposed by Actavis. So far as the declaration on direct infringement is concerned I do not think this would serve any useful purpose. Actavis sought DNIs in order to clear the way as against Lilly’s 508 patent for the launch of specific products. They sought declarations in order to have protection from an action for infringement brought by Lilly if they launched those products. The direct infringement declaration sought does not serve that purpose, but merely recites one of the court’s conclusions. It provides Actavis with no protection for their products. In addition the proposed declaration presents the result of the litigation in a way which could be misleading to readers who have not read the judgment, and who are not alerted to the limited nature of the conclusion. That is undesirable.

No doubt in order to meet the second of these reasons, and to provide a more balanced view of the outcome, Actavis proffer the indirect infringement declaration. In its original form (including the words “if and only if”) this was plainly objectionable, as it could have been used to give the impression that, provided that Actavis did not have the particular state of knowledge identified, this court had decided positively that there was no indirect infringement either. Whilst the alternative wording meets that objection to some extent, I do not believe it is appropriate either. By continuing to make the declaration conditional on Actavis’ state of mind, it fails to record what the court has actually decided, namely that the supply of Actavis’ product for reconstitution in saline in their admitted state of knowledge will be an infringement of the patent.

That is sufficient to deal with the declaration issue. I would add that Actavis ought to have come forward with the wording of its proposed declaration at a much earlier stage. I accept that it is often not possible for a litigant to predict all the possible outcomes. In this case however the possibility that the court might find indirect but not direct infringement has been ventilated from an early stage: see for example the transcript of the hearing before Arnold J on 7 February 2014 at page 23. Indeed in the 15^th witness statement of Mark Hilton dated 3 June 2015 he says that Actavis had been aware of the potential issue of indirect infringement for some time and had been looking at alternative ways to reconstitute or dilute its products “for some time”. There would be no point in looking at these alternatives except in the context of a finding of no direct infringement. In his 16^th witness statement Mr Hilton explains that “for some time” means since November 2013. There has been ample time to formulate satisfactory declarations for the eventuality in question.

I am more sympathetic to Actavis’ position on the dextrose remission issue. At the hearing before Arnold J which I have just referred to, Actavis applied for an amendment to raise an issue concerned with reconstitution in dextrose. In the end Actavis did not press the amendment, making it clear that they wanted simply to avoid any finding in these proceedings that the Actavis AIs could only be reconstituted in saline. On the other hand counsel for Lilly, Mr Thorley QC, made it clear that Lilly reserved its right to contend in the future that to raise a claim at that stage for a declaration based on non-infringement with a dextrose product would be an abuse of process. We are not in a position to determine ourselves whether Lilly is right in its contention.

The court has power under CPR 52.10(2)(b) to refer claims or issues for determination by the lower court. Given that the parties disagree over whether acts done in relation to a product recommended for reconstitution in dextrose infringe the 508 patent, there are obvious procedural advantages in that issue being decided in the presently constituted proceedings, if it can be, with the minimum of additional formality and expense. At the same time the fact that this court has taken the step of remitting this issue should be entirely without prejudice to Lilly’s right to contend that to raise the issue at this stage, either within this action or by way of a separate action, is an abuse of process. I would therefore be prepared to exercise the power under CPR 52.10(2)(b) to remit an issue for determination in the Patents Court.

I am not however content with Actavis’ formulation of the issue. I would remit the issue in the broadest possible terms, namely whether the supply or offer to supply of the Actavis AIs when recommended for reconstitution in dextrose would infringe the 508 patent. The patents judge will then be in a position to give the necessary procedural directions for the proper identification of the issue and its trial.

I agree.

I agree also.