OJ EPO SE 5/2015, p120 - Rian KALDEN

Senior Judge, Court of Appeal The Hague

Discussion of recent CJEU case law on SPCs: the three 12 December 2013 cases

I. Introduction

A supplementary protection certificate (SPC) is a sui generis intellectual property right that commences at the expiry of a basic patent. An SPC may last for up to 5½ years. SPCs were created by, and are presently governed by, EU Regulation 469/2009 (replacing 1768/92). SPCs are very important. Up to 80% of the total revenue for a major pharmaceutical product can be generated in the SPC term. It is said that as much as 80 percent of the revenue for a pharmaceutical company could be generated in the term of the SPC.

The Regulation contains definitions, inter alia on "basic patent" (a patent protecting: (i) a product as such; (ii) a process or (iii) an application for a product); "product" (the active ingredient or combination of active ingredients of a medicinal product) and "medicinal product" (the substance presented for treating or preventing disease in humans). The definition that has been focused on in the Court of Justice of the European Union (CJEU) cases that I am going to discuss is what a "product" means. As I will point out, "what you see" (in the Regulation) is not always "what you get" (from the CJEU).

Article 3 of the Regulation contains the conditions for obtaining a certificate. The term "product" features in every single paragraph of Article 3. Over the course of the last two or three years, the CJEU has looked at all of subsections (a), (b) and (d) of Article 3. The term "product" was very tightly interpreted, following the MIT jurisprudence. Articles 4/5 were given a broad interpretation. An SPC for A would cover any medicine containing A, even if the MA on which it was based was for A + B and even if the medicine attacked contained other ingredients than just A. In 2013, there were Reasoned Orders which seem to continue this jurisprudence.

II. The state of the law – more precise, of the case law – on Article 3 at the end of 2012

Subsection (a) requires that the product is "protected by a basic patent in force".

Until 2011, there was controversy whether the test was the "disclosure" test introduced by LJ Jacob in Takeda or the infringement test. We (the judges and practitioners alike) all hoped that it was the infringement test because at least that would be clear. The product would be protected if it would infringe the basic patent. But we were told that was wrong in Medeva and that the product had to be specified in the wording of the claims or, in later cases, identified in the wording of the claims. I was told it doesn't mean anything different but that it is just a translation issue. But still: what did "specified" mean? This was an outstanding issue. Many questions arose:

  • What about Markush formulae? If you've got a Markush formula contemplating 2 billion compounds, are they all specified?
  • What about classes of products? If you've got acid addition salt, is that good enough to specify the oxalate salts?
  • What about diuretics? Diuretic could cover anything from hydrochlorothiazide to strong lager. Would a claim to A plus a diuretic specify HCTZ? This question arose in Actavis.
  • What about functional definitions? If you had, like in Eli Lilly, an antibody that binds to this particular protein, was that good enough to specify all substances performing that function?

All these were still outstanding questions. Some light was shed on this during the course of 2013.

Subsection (b) requires that a valid authorisation to place the product on the market as a medicinal product has been granted.

The world was turned upside down in November 2011 in Georgetown, when it was said that a marketing authorisation (MA) for A plus B was a valid basis to apply for an SPC for A, B, and A plus B. No regulatory lawyer expected that outcome as it meant a 180 degree turn from practice in most if not all countries. It seems difficult to reconcile this approach with the definition of "product".

Another issue that arose under 3(b) was: who did the MA have to belong to? Did it have to belong to the patentee or the patentee's licensee or could it be the patentee's arch-enemy? Again, this was not at all clear.

Subsection (c) requires that the product may not already have been the subject of a certificate.

It was clear already from Biogen that there could be more than one SPC for any given product where you had two patents, both covering the products, in the hands of different patentees. It was common ground that you could have one SPC per patent per product. But a little throwaway comment in Medeva suggested it was one SPC per patent, which attracted much attention due to more elaborate comments made by Advocate General Trstenjak on the issue. Did the Court of Justice mean to deviate from this generally accepted understanding of Biogen or not?

Subsection (d) prescribes that the authorisation referred to under (b) must be the first authorisation to place the product on the market as a medicinal product.

Neurim threw out all the earlier case law, really without reference to it, and basically said you can ignore an earlier MA if it falls outside of the scope of the later basic patent relied on. It therewith opened the door to second-medical-use SPCs. This is in my view the only correct approach, but certainly not the CJEU's view up till then. An outstanding issue is whether this also overturns MIT.

The three cases decided by the CJEU on 12 December 2013 all deal with Article 3 of the Regulation.

Human Genome Sciences v Eli Lilly (C-493/12)

The relevant facts of the case were as follows. HGS have a patent which claims the protein neutrokine-alpha and antibodies which bind to neutrokine-alpha. Eli Lilly have an antibody in development called tabalumab. That was accepted to infringe the patent.

Eli Lilly were concerned that HGS would apply for an SPC, based on HGS's patent and Eli Lilly's MA. They applied to the UK court for a declaration that any such SPC would be invalid. They raised two grounds for that. First of all, tabalumab is not specified by a claim that simply says "any antibody that binds to neutrokine-alpha", as such functional definition contains insufficient structural information. The second ground was that it would be illegal for HGS to get an SPC on Eli Lilly's MA when Eli Lilly does not consent to that.

Mr Justice Warren's provisional view was that you could get an SPC based on a third party's MA. He thought it was clear and didn't need a referral. He did agree that the question of specification was suitable for referral. He referred the question of whether it was sufficient that the antibody or antibodies were defined in terms of their binding characteristics to a target protein, or whether it was necessary to provide a structural definition for the antibody or antibodies. From the CJEU decision one thing is clear. The court continues to reject the infringement test (par. 33). Otherwise the decision is less clear. The CJEU said it is not necessary for an active ingredient to be identified by a structural formula and that you can have an SPC for a product that is defined only by a functional formula, but on the condition that it is possible to reach the conclusion, on the basis of Article 69 EPC and its protocol, that the claims relate implicitly, but necessarily and specifically, to the active ingredient, and that is a matter for the referring court (par. 39/40).

There is quite some discussion as to what that means. My guess is as follows.

In Medeva, much emphasis is laid on the claims. This is reiterated in paragraph 34. The court however now accepts a functional definition, which means that the specification of the actual active ingredient is not to be found in the claim itself. Moreover, a functional claim may cover many compounds. But when a functional claim is used, as I understand the CJEU, the "specified in the claim" requirement must be satisfied implicitlty, i.e. by means of interpretation of the claims. An active ingredient will be 'specified in the claim' if by proper claim interpretation under Article 69 EPC, which takes into account the patent specification, the claim is necessarily and specifically understood (by the skilled person) to relate to that specific active ingredient.

This means that the "specified in the claim" criterion can be satisfied in two ways: either the product (active ingredient) is mentioned in the claims by its structural formula, or the claim is necessarily understood, through proper Article 69 EPC interpretation, to specifically (also) relate to that product. This suggests, in my opinion, that the product would at least satisfy the 3(a) test if it is specifically identified in the specification.

The CJEU seems to distinguish between an infringement test and a "falling within the scope of the claims" test. In paragraph 37 it said: Tabalumab is infringing, but that is not a crucial factor for determining whether that active ingredient is protected by that patent as meant by Article 3(a).

I think what the CJEU means is that if you have a functional formula which includes A, then A infringes the patent. Whether A is "protected" by that patent as meant in Article 3(a) is quite another matter. In my view, that would only be the case if upon claim interpretation the functional claim is understood to relate specifically to A. If A is mentioned in the description, e.g. as an example, that would presumably be sufficient. But what level of specificity is required? I will come back to that.

First I turn to yet another issue arising from Eli Lilly. In paragraph 43, when discussing the purpose of the Regulation, importance is attributed to the effort made by the patentee and the research done to identify the specific active ingredient of the medicinal product: "the refusal of an SPC application for an active ingredient which is not specifically referred to by a patent ... may be justified in circumstances such as those in the main proceedings, where the holder of the patent ... has failed to take any steps to carry out more in-depth research and identify his invention specifically, making it possible to ascertain clearly the active ingredient which may be commercially exploited in a medicinal product corresponding to the needs of certain patients. In such a situation, if an SPC were granted to the patent holder, even though – since he was not the holder of the MA granted for the medicinal product developed from the specifications of the source patent – that patent holder had not made any investment in research relating to that aspect of his original invention, that would undermine the objective of (the) Regulation".

It is difficult to understand what the CJEU was really trying to say with this. I think it is clear from the judgment that the CJEU feels that tabalumab was not identified in the patent (par. 36) and therefore HGS did not invent it, nor invest in it (as that was Eli Lilly as the holder of the MA) and therefore it's not fair and not right for HGS to get an SPC based on its opponent's MA.

There are two aspects about this paragraph. Firstly it seems to deal with the third party issue, even though that was not referred – but the CJEU generally pays very little regard to what's actually referred. It is not uncommon for it to not answer questions that are referred, so it might as well answer questions that are not referred. Insofar as it does deal with the third party issue, the question remains whether the CJEU really wanted to overturn Biogen and the recent Georgetown case, where SPCs were applied for based on third-party MAs. The difference may be that in Biogen and Georgetown the holders of the MA did not oppose. Secondly, it seems to relate to Article 3(a). You can't get an SPC for tabalumab, because it's not specified, which means that the patentee did not really invent it (it is not the core of the invention) nor invest in research and it should not be compensated for research not done, because that is outside the policy of the Regulation.

The Eli Lilly case has been back to the High Court in England, to Mr Justice Warren. He had the unenviable task of trying to work out what the CJEU meant. Mr Justice Warren concluded that the CJEU meant as follows:

  • the national court should determine the scope of the claims, on the basis of Article 69 EPC. If a product is within the scope of the claims, it is protected as meant in Article 3(a), subject to the proviso;
  • but the proviso "that the claims relate implicitly but necessarily and specifically to the actual active ingredient" relates only to combinations based on 'extending wording' in a claim.

I respectfully disagree with his analysis, as it completely ignores the clear guidance in Medeva that the active ingredient must be specified/identified in the claims. The proviso in paragraph 39 of the judgment read as a whole makes it clear that an SPC may be based on a functional claim – which by nature does not contain a specification of a single active ingredient – on the condition that the skilled person necessarily understands such claim (thus implicitly) to relate specifically (also) to the actual active ingredient. The question that has to be answered is: "does the skilled person read this specific active ingredient automatically into the claim". This is to be determined by the national court on the basis of proper claim construction. I submit that this is another exercise than the determination of the scope of the claim because that, as Mr Justice Warren accepts, boils down to the infringement test which is expressly – and regretfully – rejected by the CJEU.

In the Eli Lilly case, tabalumab was not expressly named in the specification. The question remains whether the proviso means that the description should contain an actual specification of the active ingredient by a structural formula, as is suggested by Mr Justice Warren. The CJEU did not say so and it seems to me that the reference to Article 69 EPC means that such level of detailed description is not necessarily required. Applying the Article 69 EPC test, in this case the question to be answered is whether the skilled person, taking into account the disclosure of the patent as a whole, would consider tabalumab to be specifically disclosed – could the skilled person as it were already envisage it when reading the patent? In his judgment Mr Justice Warren states that it is common ground that once a target protein is identified, antibodies that bind to that target may be produced using standard techniques. The question of course remains whether that would be sufficiently specific to satisfy the proviso-test.

Leave to appeal from Mr Justice Warren's judgment has been granted so we will have to wait and see how it develops on appeal. I am quite sure there will be further references to the CJEU on this point in order to obtain the required clarity in the near future.

Actavis v Sanofi (C-443/12)

The relevant facts of this case were as follows. Sanofi had a compound patent which claimed irbesartan, but also claimed that compound in association with a diuretic. Sanofi had obtained an MA for both the single active ingredient (Aprovel) and the combination with HCTZ, which is a diuretic (CoAprovel), and also obtained two SPCs, the first one issued for irbesartan and the second one issued for the combination, expiring about one year later.

There were several proceedings pending in Europe. In the Netherlands, it was a preliminary injunction case. The parallel proceedings in the UK were clear-the-way proceedings, Actavis requesting revocation of the second SPC, and therefore there was more time to refer questions, which actually happened. There were two arguments in both jurisdictions. The first one was based on Article 3(a), that irbesartan-HCTZ was not specified in the claims as a combination. The second one was that Sanofi had already been granted an SPC for irbesartan, based on its first marketing authorisation, and that this was the product of the basic patent, so additional SPCs should not be granted pursuant to Article 3(c) because there could only be one SPC per patent (as interpreted from Medeva). In the Netherlands, Teva also ran the teleological argument that the combination SPC should be nullified as being contrary to the aim and purpose of the Regulation, since the invention was irbesartan and not HCTZ or the combination.

In the English proceedings Mr Justice Arnold decided to refer again on the meaning of Article 3(a), but this time offering his own solution: whether a combination product is protected by the basic patent should be determined by reference to the inventive advance or technical contribution of the patent. If the technical contribution lies in ingredient A only, whichever combination with that active ingredient would not be protected by the patent relied on.

The CJEU didn't answer the question whether HCTZ is specified by a diuretic, because it wasn't necessary. The CJEU dealt with the case on the basis of Article 3(c). But can we have some clue out of the judgment as to what they would think about it? There are two paragraphs in this judgment which are relevant in this respect, paras. 30 and 38, with conflicting suggestions, but both also seem to be rather hypothetical. My conclusion is that there is nothing in this judgment which sheds any light on whether HCTZ is indeed specified or not in the claims. The Dutch court decided it was, on the basis of it being within the scope of the patent on the basis of Article 69 EPC interpretation and that may – looking at my theory on Eli Lilly – very well have been the correct approach.

What did the CJEU rule on the question whether, in view of Article 3(c), you can have a second SPC under the same patent? The CJEU said yes in principle (par. 29: "it is possible, on the basis of a patent which protects several different 'products' to obtain several SPCs in relation to each of those different products, provided each of those products is 'protected' as such by the 'basic patent'"). But it said no in this case. Why is that? First, if you look at paragraph 35, the CJEU says the first SPC permitted Sanofi to oppose the marketing of a medicinal product containing irbesartan in combination with HCTZ for a similar therapeutic use to that of Aprovel – basically saying that the scope of protection of the single SPC also covers the combination (remember the broad interpretation of Articles 4/5). So you had your go; you can't have a second one. But that's not all there is to it. On the hypothetical situation, if Sanofi had first chosen an SPC on the combination, and wouldn't have had the full protection because that wouldn't cover irbesartan alone, they still would not be entitled to a single SPC, because that would be precluded by Article 3(c), according to the CJEU in paragraph 38. So apparently the scope of protection of the first SPC is not the decisive factor. Then what is the decisive factor? The relevant issue is that the combination – irbesartan and HCTZ – on the one hand and irbesartan alone on the other hand are to be considered the same product. The invention lies in irbesartan, not in HCTZ or in the combination. This is made quite clear in paragraph 42: "Article 3(c) precludes a patent holder from obtaining, on the basis of one and the same patent, more than one SPC in connection with irbesartan since such SPCs would in fact be connected, wholly or in part with the same product".

I think the new approach taken is that the CJEU redefines the definition of "product". "Product" is now to be understood to mean "an innovative active ingredient". In the actual ruling, the CJEU says: "on the basis of a patent protecting an innovative active ingredient." Also in paragraph 41, the CJEU refers to "the core inventive advance that is the subject of the basic patent", or "the principal active ingredient protected as such by the basic patent".

What is the rationale behind this? In paragraph 30 of the judgment, the CJEU mentions that the second SPC would possibly be for a longer period of time than the first one, which they perceive to be "evergreening". The argument that the second SPC would be more limited in scope than the first was rejected in a – quite curious – paragraph, saying that if you had your combination SPC, you could thereby possibly, on the basis of indirect infringement, also prevent the marketing of the single active ingredient, thereby prolonging the term of protection. The argument that there is additional research or trials on combinations required in order for them to get to the market was also considered to be not relevant. The CJEU stated that the object of the regulation is not to compensate the patent holder fully for the delay to the marketing of his invention "in all its possible forms". These paragraphs not only make it quite clear that the CJEU takes a very purpose-driven approach to SPCs, but also raise the question what the CJEU would have decided if the single SPC had not expired later than the combination SPC. That leads to the next case.

Georgetown II (C-484/12)

The facts of this case were as follows. The patent included claims to HPV-16, HPV-18, HPV-16 and HPV-18 together and various combinations. There were MAs for Gardasil® in 2006 (HPV-6, HPV-11, HPV-16 and HPV-18) and Cervirax® in 2007 (HPV-16 and HPV-18). SPCs were granted in respect of those combinations. Georgetown now applied for an SPC for HPV-16 alone.

In paragraph 32 the CJEU said it would appear to be common ground that the basic patent protects, at the very least, both the combinations and HPV-16 as marketed in Gardasil. Other than in Actavis, in this case the products were considered separate innovative products, and therefore, according to the CJEU, Georgetown could have separate SPCs for each and anyone of them, including HPV-16 alone.

But there were more facts different in Actavis v Sanofi than in Georgetown which may have been of relevance and may be in future cases. First of all, here, unlike in Actavis v Sanofi, Georgetown applied for a new SPC on the basis of the same marketing authorisation, i.e. the marketing authorisation for the first combination. So both SPCs would expire at the same time. In Georgetown there was therefore no (perceived) evergreening issue. Georgetown could, in theory, also have asked for an SPC on the basis of the second marketing authorisation. That would have given it a longer period of protection. Would the court have decided the same way had they chosen the second route? I'm not quite sure about that, because that would, at least in their view, lead to evergreening (and we will never know since the CJEU noted that such application would fail on the basis of Article 3(c) since the individual HPV-16 "product" was the same in both combinations, par. 38). Also, had Georgetown obtained the individual SPC first and then asked for a combination, under the Actavis v Sanofi reasoning of Articles 4/5, they would already have had their go with the single SPC and no need for separate and maybe even longer protection under a combination SPC. So although it seems clear at the moment, I am not quite sure, if there is a case like Georgetown, with separate products but different terms of protection, whether the CJEU would come up with the same solution. That remains to be seen in the future.

What conclusions can be drawn from these two decisions?

Clearly, you can have one SPC per product per patent, as long as each product is innovative in itself. The existence of an SPC for A prevents an SPC for A+B on the basis of Article 3(c) if the technical contribution was A; then A is regarded by the CJEU as the "product" of the patent; A+B is not another 'product' and therefore the 'product' is already the subject of an earlier SPC.

So Arnold J's suggestion has been adopted, but transplanted to the definition of "product" – which features in every paragraph of Article 3 – or maybe even to the objective of the Regulation – reformulated as "to compensate for the delay to the marketing of what constitutes the core inventive advance of the basic patent" (par. 41 Actavis v Sanofi) – the compliance with which seems to have become an additional overriding requirement for obtaining an SPC.

Where are we now?

On Article 3(a): The product must fall within the scope of the claims – either expressly when "specified/identified in the wording of the claims" itself or implicitly, but then the claim must necessarily and specifically be understood to relate to the active ingredient as prescribed by the Article 69 EPC test. This certainly applies to functional claims, but in my view might equally be applied to claims to Markush formulae, classes of product and the like.

On Article 3(b): It is still uncertain whether you can have an SPC on someone else's MA. It seems difficult to reconcile Biogen v Georgetown and Eli Lilly. Maybe the outcome is different depending on whether or not there is consent (or a licence). Further decisions have to be awaited to get more clarity on this.

On Article 3(c): If an SPC for A is granted, a further SPC for B or A+B will only be granted where B or A+B is "protected as such" by the basic patent – so one SPC per product protected "as such" per patent. An outstanding issue remains how to determine whether a combination – next to the single active ingredient – is 'protected as such (i.e.: innovative). Paragraph 42 Actavis suggests the combination should be the subject of a new basic patent. But that doesn't seem right and is also difficult to reconcile with Georgetown, whose patent contained both single active ingredients and various combinations, which were all held to be innovative.

On Article 3(d): Second-medical-use SPCs are allowed (Neurim). That at least is now clear. Another question for the future will be whether MIT still stands and for how long. Logically, following Neurim, it should be overturned in favour of new inventive formulations. GSK was not the right the case for that, but another MIT case with the right questions asked could be.

What will the future bring?

SPCs are still applied for and granted by national patent offices. It is unclear how this will be affected by the UPC system.

Interestingly, I was told that at the HGS v Eli Lilly hearing even the CJEU urged for reform of the SPC regulation, but it is very uncertain whether that will ever happen, with so many countries and so many different interests involved.