Alchemy in the UK: the Supreme Court in Eli Lilly v Actavis transmutes sodium into potassium, but will it provide gold for patentees?

Brian Whitehead and Stuart Jackson


In July 2017, the Supreme Court overruled[1] the Patents Court and the Court of Appeal by holding that Eli Lilly’s patent EP(UK) 1313508 (“Patent”) concerning the use of pemetrexed disodium in the manufacture of a medicament in combination with vitamin B12 for the treatment of cancer, would be directly infringed by Actavis’s proposed formulation utilising one of three different forms of pemetrexed: the free acid, or the ditromethamine or dipotassium salts (non-chemist readers are referred to the background section below, in which an explanation of the various terms is given).

In handing down what is almost certainly the most important UK patent law decision since the House of Lords’ judgment in Kirin-Amgen, the Supreme Court has introduced, for the first time, an express “doctrine of equivalents” into UK patent law. As the authors explain below, this is likely to lead in future to a wider scope of protection for patents in the UK. The Supreme Court also gave some limited approval to the idea of using the patent prosecution history as a guide to patent construction, although it is debatable, as this article explains, whether this will have as much of an impact going forward as the decision on equivalents.


Pemetrexed, the chemical structure of which is shown below, is an antifolate i.e. it blocks the actions of folic acid (vitamin B9). Although it has an anti-cancer effect, it has too many side-effects to be used on its own as an anti-cancer treatment. Eli Lilly discovered that the damaging side-effects can be reduced by administering pemetrexed together with vitamin B12. That discovery underlies the Patent, and Eli Lilly has been marketing a pemetrexed/vitamin B12 medicament under the brand Alimta since 2004.

The two CO2H groups shown in the above structure are called carboxylic acid groups. The form shown above (in which the hydrogen atoms in the two CO2H groups are present) is known as the “free acid”. The hydrogens can be replaced, for example by sodium or potassium, resulting in species called the “disodium salt” and “dipotassium salt” respectively. A very large number of salts of pemetrexed can be prepared, and another salt which formed part of the dispute is the ditromethamine salt, which contains the unit shown below in place of each of the two hydrogens.

The Patent claims are in the so-called Swiss form (“use of substance X for the preparation of a medicament for treating indication Y”). Prior to the entry into force of the European Patent Convention 2000, this was the only means by which a second medical use of a known pharmaceutical could be patented. Although Swiss-form claims are process, not product, claims, and it is now possible to claim a second medical use as a product claim, this distinction does not impact on the scope of the Supreme Court’s decision, which is of general applicability as explained below.

History of approach to patent construction in the UK

Over the past 35 years there have been three seminal decisions in the English courts governing construction of patents. The first, Catnic[2], concerned a patent for a steel lintel for use in the building trade, in which the claim specified a particular component of the lintel “extending vertically” from a horizontal plate. In the defendant’s lintel the corresponding component was inclined between six and eight degrees from the vertical. In holding that the defendant’s lintel nonetheless infringed, Lord Diplock posited that it would be obvious to a builder that a small inclination from the vertical would make no material difference to the way the lintel worked when used in building operations. More generally, he held that the relevant question in determining whether a variant falls within the scope of a patent claim is:

whether persons with practical knowledge and experience of the kind of work in which the invention was intended to be used, would understand that strict compliance with a particular descriptive word or phrase appearing in a claim was intended by the patentee to be an essential requirement of the invention so that any variant would fall outside the monopoly claimed, even though it could have no material effect upon the way the invention worked.

Catnic concerned the interpretation of patents granted under the Patents Act 1949, but the reasoning in that decision has continued to be applied to patents granted under the Patents Act 1977 (which implemented the European Patent Convention (“EPC”)). Article 69(1) of the EPC (as amended in 2000, but the original version was substantially similar) provides that:

The extent of the protection conferred by a European patent or a European patent application shall be determined by the terms of the claims. Nevertheless, the description and drawings shall be used to interpret the claims.

The EPC also contains a Protocol on the Interpretation of article 69 (“the Protocol”), Article 1 of which (as amended in 2000, but again the original version was substantially similar) states:

Article 69 should not be interpreted as meaning that the extent of the protection conferred by a European patent is to be understood as that defined by the strict, literal meaning of the wording used in the claims, the description and drawings being employed only for the purpose of resolving an ambiguity found in the claims. Nor should it be taken to mean that the claims serve only as a guideline and that the actual protection conferred may extend to what, from a consideration of the description and drawings by a person skilled in the art, the patent proprietor has contemplated. On the contrary, it is to be interpreted as defining a position between these extremes which combines a fair protection for the patent proprietor with a reasonable degree of legal certainty for third parties.

In the second key decision, Improver[3], Hoffmann J (as he then was) set out a three-stage test to determine whether a variant falls within the scope of a patent claim, as follows:

(1)       Does the variant have a material effect upon the way the invention works? If yes, the variant is outside the claim. If no –

(2)       Would this (ie that the variant had no material effect) have been obvious at the date of publication of the patent to a reader skilled in the art? If no, the variant is outside the claim. If yes –

(3)       Would the reader skilled in the art nevertheless have understood from the language of the claim that the patentee intended that strict compliance with the primary meaning was an essential requirement of the invention? If yes, the variant is outside the claim.

Although the above test is purely a matter of UK case law and is not found in the EPC or the Patents Act 1977, it was very widely applied following Improver, to the extent that the three questions were often referred to as the “Protocol questions”.

In the third key decision, Kirin-Amgen[4], Lord Hoffmann re-visited his earlier approach in Improver. He brought the English law approach to patent construction in line with that used for contract construction, stating that:

the question is always what the person skilled in the art would have understood the patentee to be using the language of the claim to mean. And for this purpose, the language he has chosen is usually of critical importance.

Lord Hoffmann rejected the notion that there is any doctrine of equivalents in English patent law which would “allow the patentee to extend his monopoly beyond his claims”, stating that Article 69 of the EPC “firmly shuts the door on any doctrine which extends protection outside the claims”.

Finally, in affirming the purposive approach to patent construction as laid down in Catnic as “the bedrock of patent construction, universally applicable”, Lord Hoffman said that the Improver questions were “guidelines for applying that principle to equivalents … more useful in some cases than in others”. In practice, the Improver questions were rarely encountered after Kirin-Amgen, although as explained below they have now been given a new lease of life by the Supreme Court’s decision in this case.

Decisions of the Patents Court and Court of Appeal in this action

Actavis sought a declaration of non-infringement (“DNI”) in relation to proposed formulations utilising pemetrexed diacid, pemetrexed dipotassium or pemetrexed ditromethamine as the active ingredient. DNIs were in fact sought with respect to the French, German, Italian, Spanish and UK designations of the Patent. Although infringement in each jurisdiction has been considered by the English courts throughout this dispute, the authors focus in this article on the English courts’ rulings in relation to the UK only.

At first instance[5], Arnold J, unusually, applied the Improver questions (as stated above, the Improver questions largely fell into disuse after Kirin-Amgen in 2005). His answers to the questions were, respectively:

  1. The variants have no material effect upon the way the invention works;
  2. It would not be obvious to the skilled team that the variants would have no material effect on the way the invention works;
  3. There is nothing in the specification or the common general knowledge of the skilled team to suggest to the skilled team that Lilly intended to use the expression “pemetrexed disodium” in anything other than its conventional sense i.e. to refer to the sodium salt alone.

In particular, in relation to question 2 Arnold J took into account that the skilled team would not be able to say, without carrying out a considerable degree of testing, whether substituting any of the variant forms of pemetrexed would impact on the solubility or pharmaceutical acceptability of the source of pemetrexed. Arnold J rejected Lilly’s argument that question 2 should be approached on the basis the skilled team knows that the variant has received regulatory approval, and hence that they know it is pharmaceutically acceptable and sufficiently soluble, stating:

The skilled team must be able to ask and answer question 2 before investing what may be four years’ effort and considerable expense to obtain regulatory approval even on the basis of bioequivalence.

As for question 3, Arnold J took the view that Lilly had deliberately limited the claims of the Patent to pemetrexed disodium. Lilly could, if it wished, have sought broader claims if it thought it was entitled to them. It had not done so, on the basis of professional advice, and in Arnold J’s view it was not unfair to limit Lilly’s scope of protection to the wording of the claim as granted.

Accordingly, Arnold J held that none of Actavis’s proposed formulations would constitute direct infringement of the Patent.

Arnold J then turned to the question of indirect infringement under section 60(2) of the Patents Act 1977. Lilly’s argument was that pemetrexed diacid, dipotassium and ditromethamine are means relating to an essential element of the invention since they provide a source of pemetrexed. Actavis’s formulation was intended to be dissolved by users in a saline (i.e. sodium chloride) solution, which would provide a source of sodium ions. Lilly contended that the diluted solution would put the invention into effect, as both pemetrexed and sodium would be present.

Arnold J rejected this argument, on the basis that at no point would pemetrexed disodium be used in the manufacture of a medicament by anyone, characterising the diluted solution as having nothing more than “sodium ions and pemetrexed ions floating around”. Accordingly, Arnold J held that indirect infringement was not made out either.

The Court of Appeal[6] agreed with the first instance decision in relation to direct infringement, but disagreed with Arnold J in relation to indirect infringement. Floyd LJ characterised Arnold J’s approach as being predicated upon an assumption (not expressly stated by Arnold J) that the reference to “pemetrexed disodium” in the claim is to the solid form of that substance. Floyd LJ disagreed with that underlying assumption, holding that the skilled team would understand that pemetrexed disodium is “also used to refer to solutions which contain pemetrexed ions and sodium ions in solution”. Applying that alternative construction, the excess of sodium ions present in the saline solution meant that that integer of the claim would be possessed by Actavis’s proposed formulations.

In the authors’ view, the Patents Court’s and Court of Appeal’s decisions on direct infringement were unsurprising, as they were a straightforward application of the approach which had been understood to be the law at the time. Whereas geometrical terms such as “vertical”, and even mathematical ranges such as “between 1 per cent and 25 per cent”[7] have a degree of “fuzziness” and hence are susceptible to interpretation, the same cannot be said of a chemical element such as sodium. Short of an alchemic transmutation, there is no construction of the word “sodium” that can encompass “potassium”, which is a completely different element.

On the issue of indirect infringement, the authors preferred Arnold J’s approach, considering that Floyd LJ’s approach, whilst entirely logical, leads to a rather arbitrary outcome. The saline solution is nothing more than an inert carrier, and if Actavis’s formulation was dissolved in another (non sodium-containing) carrier, infringement under s. 60(2) would not have been made out. In fact, following the Court of Appeal’s decision, Actavis applied for DNIs in respect of pemetrexed formulations containing express instructions that they were to be reconstituted in dextrose (i.e. a non sodium-containing) solution, and Arnold J granted those declarations following a subsequent hearing[8].

The Supreme Court Judgment – equivalents

All the above has now been turned on its head by the Supreme Court’s decision. Lord Neuberger, giving the only judgment, provided a detailed account of the history of patent construction in UK law. He focused in particular on article 2 of the Protocol, which states:

For the purpose of determining the extent of protection conferred by a European patent, due account shall be taken of any element which is equivalent to an element specified in the claims.

Article 2 was in fact mentioned by Lord Hoffmann in Kirin-Amgen, but was not in force at the time that decision was handed down. Lord Neuberger distilled the case law, in light of article 2, as follows:

a problem of infringement is best approached by addressing two issues, each of which is to be considered through the eyes of the notional addressee of the patent in suit, ie the person skilled in the relevant art. Those issues are: (i) does the variant infringe any of the claims as a matter of normal interpretation; and, if not, (ii) does the variant nonetheless infringe because it varies from the invention in a way or ways which is or are immaterial?

In Lord Neuberger’s view, the approach in Kirin-Amgen wrongly conflated the above two issues into a single question of interpretation. Following the Supreme Court’s decision, it is now clear that a two-step approach is required, first applying normal (purposive) rules of construction to determine the meaning of the wording used in the claims, and secondly considering whether the scope of protection afforded by the claim should extend beyond that meaning.

Clearly, any protection extending beyond the meaning of the words actually used in the claim can give rise to uncertainty, so Lord Neuberger sought to contain such uncertainty by limiting the extended scope to those variants which contain immaterial variations from the invention. In that regard, Lord Neuberger considered that the Improver questions provide helpful guidance, and chose to reformulate them as follows:

  • Notwithstanding that it is not within the literal meaning of the relevant claim(s) of the patent, does the variant achieve substantially the same result in substantially the same way as the invention, ie the inventive concept revealed by the patent?
  • Would it be obvious to the person skilled in the art, reading the patent at the priority date, but knowing that the variant achieves substantially the same result as the invention, that it does so in substantially the same way as the invention?
  • Would such a reader of the patent have concluded that the patentee nonetheless intended that strict compliance with the literal meaning of the relevant claim(s) of the patent was an essential requirement of the invention?

In order to establish infringement in a case where there is no literal infringement, a patentee would have to establish that the answer to the first two questions was “yes” and that the answer to the third question was “no”.

Applying those reformulated questions, Lord Neuberger held that the skilled team, upon being told that any of Actavis’s variants worked, would have appreciated that they work in precisely the same way as a formulation including pemetrexed disodium (because the variants are all simply means of introducing the active ingredient into the medicament). Question 2 was therefore answered “yes”.

On question 3, Lord Neuberger was of the view that the Patents Court / Court of Appeal’s approach placed insufficient weight on article 2 of the Protocol. In light of the skilled team’s knowledge of the properties of acids and salts, and their impact on solubility, the skilled team would not have concluded that the patentee intended to exclude anything other than pemetrexed disodium from the scope of the claim. Actavis’s proposed formulations would therefore, subject to the prosecution history of the Patent, directly infringe the Patent.

The Supreme Court Judgment – relevance of prosecution history

Actavis contended that the prosecution history of the Patent – particularly the patentee’s responses to objections raised by the EPO examiner – made it clear that the claims of the Patent should be limited to pemetrexed disodium. Lord Neuberger considered previous English case law as to the admissibility of prosecution history to the issue of patent construction (so-called “file wrapper estoppel”), and considered also the approach of other European countries’ courts to the issue. He concluded that

it is appropriate for the UK courts to adopt a sceptical, but not absolutist, attitude to a suggestion that the contents of the prosecution file of a patent should be referred to when considering a question of interpretation or infringement, along substantially the same lines as the German and Dutch courts.

Specifically, and without ruling out the possibility that there may be other circumstances justifying referring to the prosecution history, Lord Neuberger said that such reference would be appropriate where:

  • the point at issue is truly unclear if one confines oneself to the specification and claims of the patent, and the contents of the file unambiguously resolve the point, or
  • it would be contrary to the public interest for the contents of the file to be ignored.

The first category was said to be self-explanatory, the second was said to be exemplified by a case where the patentee had made it clear to the EPO that he was not seeking to contend that his patent, if granted, would extend its scope to the sort of variant which he now claims infringes.

On the facts before him, Lord Neuberger held that nothing in the prosecution history changed his initial view that the patentee had not intended to exclude anything other than pemetrexed disodium from the scope of the claim. Direct infringement of the Patent was therefore made out in respect of each of Actavis’s formulations.


On a first reading of the judgment, the express introduction into English patent law of a “doctrine of equivalents” and “file wrapper estoppel” are its most striking aspects. In the authors’ view, however, those are not in fact the most profound changes made by the decision. Certainly, it appears that Lord Neuberger is of the view that the English courts have always, even if not expressly, been prepared to apply a doctrine of equivalents. At paragraph 65, he points out that there is no construction of the term “helical metal spring” that could encompass the slotted rubber rod which was held to infringe in Improver. Furthermore, at paragraph 57 he cites two patent decisions from as long ago as the 19th century in which a doctrine of equivalents appears to have been applied. Arguably, holding that potassium falls within the scope of a claim specifying sodium is no greater a leap than holding that a slotted rubber rod falls within the scope of a claim specifying a helical metal spring. It may be the case, therefore, that the courts have previously applied what is, in reality, a doctrine of equivalents, but under the guise of an interpretative approach.

Similarly, with regard to “file wrapper estoppel”, Lord Hoffmann in Kirin-Amgen had said “The courts of the United Kingdom … certainly discourage, if they do not actually prohibit, use of the patent office file in aid of construction”. In the authors’ view, even applying that very restrictive approach, it is unlikely that prior to this decision a patentee would have been permitted to argue for a construction of a patent that he had expressly disclaimed during prosecution. Lord Neuberger’s second category of circumstances justifying reference to prosecution history therefore probably introduces nothing new into English law. His first category may herald a new approach, but it is clearly restricted in scope (the requirements that the position is “truly unclear” and the contents of the file will “unambiguously resolve” the issue are likely to be met only in a small number of circumstances). Whereas the authors would not be surprised to see a number of test cases in the next few years further exploring the scope of file wrapper estoppel, they expect that the lower courts will fairly quickly provide their further clarification as to its scope.

In the authors’ view, the most radical aspect of the Supreme Court’s decision is in the reformulated second Improver question. The key wording, underlined below, is:

Would it be obvious to the person skilled in the art, reading the patent at the priority date, but knowing that the variant achieves substantially the same result as the invention, that it does so in substantially the same way as the invention?

The reason that Arnold J answered the second Improver question in the way he did was the considerable effort which would be required in order for the skilled team to satisfy itself that use of a different salt (or the free acid itself) would have no material effect on the invention. Arnold J summarised the expert evidence on salt forms in paragraph 80 as follows:

  1. i) the salt form can have a significant impact on the effectiveness of a drug;
  2. ii) salt forms 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;

iii) 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;

  1. iv) 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;
  2. v) there can be other problems trying to make salts before one can consider testing them further, including solvents or other impurities being trapped in the lattice. In such cases, although a salt has been made, the solid form would rarely be robust or commercialisable;
  3. vi) in general, there can be many dead-ends and false leads when attempting to prepare salts of a parent molecule for the first time.

At paragraphs 81-86, he then provided details as to the steps which would be required to test different salt forms, noting that it was common ground that “one cannot predict the solubility of any particular salt form in advance of making and testing it”, and at paragraph 126 suggested that the work required might involve “four years’ effort and considerable expense”. More generally, a person skilled in the relevant art might be able to tell more-or-less immediately upon inspection that a slotted rubber rod would work in the same way as a helical spring, but a similar assessment cannot typically be made in the case of medicines without years of work. No doubt the same also applies in many other fields of technology.

The effect of the Supreme Court’s reformulation of the second Improver question is to sidestep altogether all these issues. The second question is now asked on the basis that the variant does actually work, and the skilled person knows that it does. The reformulated second question imputes knowledge to the skilled person that he does not in fact possess.  In the case under consideration, for example, the skilled team did not know at the priority date whether or not potassium salt or the free acid would work in the same way as the claimed invention: they may have hoped, guessed or even expected that the alternatives, or one of them would work, but they did not know. In the authors’ view, this is a dramatic change, and is very favourable indeed to patentees. Patentees are now placed in the fortunate position of being able to say that whereas they do not know if the invention will still work if integer X is replaced by Y (which is why it was not claimed), nevertheless, if someone else spends years of work and millions of pounds to find out if Y works in the same way, that will be within the scope of the patent.

As an aside, the description of the invention given in the patent explains that the invention consists of the administration of vitamin B12 in conjunction with an antifolate agent in order to mitigate the adverse side effects of the latter. Although pemetrexed disodium is quoted as an example of an antifolate agent, and that is the only such agent included in the claims, the description, as Lord Neuberger notes, continues to explain the invention almost exclusively in terms of the general term “antifolate agents”. It is interesting to speculate what the result would be if the three questions were applied to an alleged infringing product in which an equivalent product was obtained not by use of a different counter ion, but by substituting another antifolate agent for pemetrexed. That question is touched upon by Lord Neuberger in his consideration of question 3, and he airs it again at paragraph 89 when considering the issue of reference to the prosecution history, where he suggests that the examiner was wrong to limit the scope of Lilly’s patent to pemetrexed rather than to a broader class of antifolate agents. Lord Neuberger did not need to decide the point, and he does not do so. However, if it is indeed the case that a different antifolate, in conjunction with vitamin B12, works in the same way as pemetrexed in conjunction with vitamin B12, it seems likely that the issue of infringement would revolve only around the answer to the third Improver question, the requirements of the first two questions being satisfied.

The authors wonder whether this decision, which (it might be suggested) goes beyond a straightforward doctrine of equivalents, has been made not to assist in interpretation of the law, but for policy reasons. One such reason is clearly stated by Lord Neuberger, namely the commendable one of bringing UK patent law into line with that of other European countries. A second policy reason, perhaps, is to avoid the situation that a patentee who has invested heavily in a commercially and medically important invention should be effectively denied a monopoly because a simple modification that has equivalent effect, albeit one that could not have been predicted to work, is found to circumvent the claims of a patent.

Patentees will doubtless now be anxious to establish the new limits to patent protection afforded by the Supreme Court’s decision. Of course, it remains to be seen whether the Patents Court will apply the extended doctrine of equivalents and the reformulated Improver questions with vigour or with reluctance. The authors consider that English first instance judges have often adopted a fairly conservative approach to patent construction, favouring certainty and predictability over a more expansive approach. Whereas the Supreme Court has now handed patentees a new “tool”, patentees will still need to overcome the hurdles in the reformulated Improver questions, which will require expert evidence and/or experiments to satisfy the requirements. For example, the third Improver question (answered “yes” by both the Patents Court and Court of Appeal in this action) has not been reformulated by the Supreme Court, and may in future, depending on the precise wording of a patent’s specification and claims, continue to take many immaterial variants outside the scope of a patent’s claim.

In any event, whereas the boundary of the extent of protection afforded by a patent will henceforth be less certain in many cases, it is likely that the decision will be welcomed by patentees, both in the pharmaceutical sector and elsewhere, who will generally benefit by virtue of the chilling effect that the new rules on construction are likely to have on competitors unwilling to risk a costly development of something that unwittingly turns out to be an infringement.


About the authors

Dr. Brian Whitehead, senior solicitor with Kempner & Partners, holds an MA in chemistry (Oxford) and a PhD in biochemistry (Sheffield).  Prior to becoming a solicitor, he worked as a research scientist and academic publisher in the Netherlands.

Stuart Jackson, Consultant to Kempner & Partners, holds BSc (Imperial College) and MSc (Oxford) degrees in chemistry.  Prior to becoming a solicitor, he worked in industry as a chemist in R&D for 17 years, and is a Chartered Chemist.

[1] Eli Lilly and Company v Actavis UK Limited, [2017] UKSC 48

[2] Catnic Components Ltd v Hill & Smith Ltd [1982] RPC 183

[3] Improver Corpn v Remington Consumer Products Ltd [1990] FSR 181

[4] Kirin-Amgen Inc v Hoechst Marion Roussel Ltd [2005] RPC 9

[5] [2014] EWHC 1511 (Pat)

[6] [2015] EWCA Civ 555

[7] See for example Smith & Nephew Plc v ConvaTec Technologies Inc. [2013] EWHC 3955 (Pat) and [2015] EWCA Civ 607

[8] [2016] EWHC 234 (Pat)