New EBA referral: When is prior use of a product excluded from the prior art for lack of enablement?
There has been a new referral to the Enlarged Board of Appeal (EBA). This first referral of 2023 seeks clarification on the enablement test for prior use of a product. The referral particularly asks whether the non-enabling prior use of a product excludes the composition of the product from the prior art or whether it also excludes the product per se (i.e. so that it can not be used as a starting point for inventive step). The referral also seeks clarification over the definition of enablement with respect to the prior use of a product, and particularly whether "enablement" requires that the skilled person be able to fully analyse and reproduce the exact same product.
Legal background
The EPC defines the prior art with respect to a claimed invention as "everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application" (Article 54(2) EPC). According to existing case law, in order to constitute prior art, a disclosure must also be enabled (EPO Guidelines for Examination, G-IV, 2). Further, according to the established case law from the EBA, if the invention is a product, prior disclosures of the product are only considered to have been "made available to the public" if a skilled person would have been able to analyse and reproduce the product, using common general knowledge (CGK) (G 1/92, headnote).
 |
| Soaking up the solar energy |
The recent referral arises from the appeal case T 0438/19, relating to the Opposition Division decision to maintain the Mitsui Chemical patent EP 2626911 as granted. The patent relates to polymers for coating and protecting solar cells in solar panels. Claim 1 as granted claims a material for encapsulating a solar cell, comprising various materials such as ethylene and aluminium in specified amounts and densities.
The inventive step of the claimed invention depended on whether prior use of the product ENGAGE 8400 polymer should be considered prior art. It was uncontested that ENGAGE 8400 was commercially available to the public prior to the priority date of the patent. The Opponent submitted data showing that ENGAGE 8400 was a polymer satisfying all the limitations of claim 1 of the granted patent, except for the specified aluminium content. Altering the aluminium content was considered by the Board of Appeal to constitute an obvious modification.
The Patentee argued that the sale of ENGAGE 8400 before the priority date did not constitute an enabling disclosure of the polymer's composition. Particularly, the Patentee argued that it would be an undue burden for the skilled person to reverse engineer ENGAGE 8400 and then to select the specific catalysts and reaction conditions necessary for manufacture of the product.
Diverging Board of Appeal interpretations of G 1/92 and Article 54(2) EPC
In G 1/92, the EBA concluded that the composition of a product is to be considered "made available to the public" when the product has been made available to the public and can be analysed and reproduced by the skilled person (Headnote). For 30 years, Boards of Appeal have mostly interpreted G 1/92 as requiring three criteria in order for a commercially availability of a product to be considered prior art with respect to a claimed invention: 1) a member of the public must be able to access the product; 2) the skilled person must be able analyse the composition or internal structure of the product; and 3) the skilled person must be able to reproduce the product using CGK without undue burden.
However, the Board of Appeal in the present case found no justification for the "enabling" requirement of "made available to the public" in the intention of the legislator, as evidenced by the historical documentation related to the genesis of Article 54(2) EPC (r. 10.3).
The Board of Appeal also pointed to a potentially confusing paragraph in G 1/92, in which the EBA appeared to suggest that there was a difference between the requirement of the disclosure of a product per se, and the disclosure of the composition of the product:
"An essential purpose of any technical teaching is to enable the person skilled in the art to manufacture or use a given product by applying such teaching. Where such teaching results from a product put on the market, the person skilled in the art will have to rely on his general technical knowledge to gather all information enabling him to prepare the said product. Where it is possible for the skilled person to discover the composition or the internal structure of the product and to reproduce it without undue burden, then both the product and its composition or internal structure become state of the art."
(G 1/92, r. 1.4, emphasis added)
The Board of Appeal in the present referring case identified divergence in how Boards of Appeal have interpreted G 1/92 in view of this paragraph. Particularly, the referring Board pointed to diverging decisions over how G 1/92 should be applied in cases where a product that was publicly available before the priority date of a patent could not be analysed or reproduced by a skilled person. According to one body of case law, in such a case, only the internal composition of the product should be excluded from the prior art (T 370/02, T 2045/09, T 1833/14, and T 0023/11). According to another body of case law both the product per se and its internal composition should be excluded from the prior art (T 946/04, T 1666/16).
The referring Board argued that the distinction over whether the composition of the product, or both the product and its composition, should be excluded from the prior art is not just a theoretical question, but has practical consequences for inventive step analysis. If a product per se is excluded from the prior art it cannot then be used as a starting point for inventive step in combination with other prior art documents, e.g. documents about the product that may assist in overcoming the enablement issues. Importantly, the enablement test permits the skilled person to use common general knowledge but not prior art documents that do not form part of the CGK.
The referring Board of Appeal also identified diverging decisions on the appropriate criteria for assessing the ability of a skilled person to analyse and reproduce the disclosed product, particularly over the level of analysis required and the degree of variability permitted in reproduction (as discussed in Takeda v Roche [2019] EWHC 1911 (IPKat)).
The Board of Appeal concluded by referring the following questions to the EBA (emphasis added):
1. Is a product put on the market before the date of filing of a European patent application to be excluded from the state of the art within the meaning of Article 54(2) EPC for the sole reason that its composition or internal structure could not be analysed and reproduced without undue burden by the skilled person before that date?
2. If the answer to question 1 is no, is technical information about said product which was made available to the public before the filing date (e.g. by publication of technical brochure, non-patent or patent literature) state of the art within the meaning of Article 54(2) EPC, irrespective of whether the composition or internal structure of the product could be analysed and reproduced without undue burden by the skilled person before that date?
3. If the answer to question 1 is yes or the answer to question 2 is no, which criteria are to be applied in order to determine whether or not the composition or internal structure of the product could be analysed and reproduced without undue burden within the meaning of opinion G 1/92? In particular, is it required that the composition and internal structure of the product be fully analysable and identically reproducible?
Final thoughts
The Board of Appeal in the referring decision believes it has identified a divergence in the case law on prior use with practical consequences for inventive step. As specifically noted in the headnote of G 1/92, the issue of enablement in prior use is relevant to all types of products, not just chemical compositions. The outcome of the referral, if accepted as admissible by the EBA, may therefore have consequences across the technical fields. Stay tuned to IPKat for further analysis!
Reviewed by Rose Hughes
on
Friday, June 30, 2023
Rating:
Follow the IPKat on LinkedIn ![[Guest post] Can AI be considered a PHOSITA? Policy debates in the US and the EU](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEggxDO6mW5r4n3-06Af5ckmIaMIhzgPJBoDP8AUsSYXY2zajUQt1ObGVn_GhCgidbG_YDVnybQuJ5XoAjjBG9Ws2xJWDQHNPMrLkn526SWIG371X_Kjt1E8tJmn8Ae_20Phyphenhyphen09JeuUOhhTR3aZE9lMRQlqHGQGcAWGxlo91rrKcLk0AfUjWCvw6RQ/s72-c/Two-cats-eating.jpg)
Posts
This should not have needed a referral, but I suppose there will be benefit from whatever answer the Enlarged Board. Essentially the issue is whether 'any' info can be derived from the marketed product as being public, where full analysis and reproducibility is not possible. The Board should have said that whilst this is not a classical G1/92 situation, getting some info from the marketed is clearly possible and that is enough to destroy inventive step. The Enlarged Board is going to say (I predict) that if partial info can be derived from the product (not full analysis) then that must be part of the prior art.
ReplyDeleteI agree that if partial info can be derived, then that info is part of the prior art.
Delete(And in general the EPO concept that a "non-enabling" disclosure is not part of the prior art under Art. 54(2) at all is broken. The disclosure may still contain useful information which can be combined with other disclosures. The "enablement" criterion should be read into Art. 54(1), not (2).)
What Q1 seems to ask is whether the product itself is part of the prior art.
Art. 54(2): "The state of the art shall be held to comprise everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the European patent application."
Does this refer only to INFORMATION that is made available to the public, or does it also include physical objects?
If it includes physical objects, then the commercial product for sale before the filing date, including all its (potentially hidden) properties, is part of the prior art (reproducibility is not a problem: you can buy it). Its properties would then not have to be derivable already at the filing date to be novelty destroying (they would still need to be proved somehow before lack of novelty can be concluded, obviously).
If the skilled person would need information about the composition of the product in order to modify it into something that falls within the terms of the claims, then that information must have been derivable from the product or otherwise have been available at the filing date. T 438/19 does not seem to explain whether this is the case.
Many commercial products, like foods or detergents are acompanied by a declaration of ingredients. These products cannot be identically reproduced already because the “perfum” or “flavour” is a generic designation for a trade secret mixture of fragrances or taste giving ingredients. The answer to the third question has fat reaching implications.
ReplyDeletePrior use as prior art requires to show reproducibility. No reproducibility, no prior art. I thought it was crystal clear. Which partial information can be derived from what does not belong to the state of the art? Some boards appear to be desirous of misunderstanding.
ReplyDeleteA one-time prior use may not be reproducible, but if a product is for sale, then you can reproduce it by buying it.
DeleteOtherwise, where do you stop requiring reproducibility? If you are not allowed to shop for ingredients (or end products, but I see no difference there), do you have to be able to mine the minerals yourselves?
Nobody but Intel is able to (re-)produce a modern Intel chip. Are these widely available chips not prior art?
So, Niko, how often does a product need to be sold in order to become prior art?
DeleteThe argument that something is reproducible because you can buy it is flawed. In this logic, the product is no longer reproducbile (and disappears from the prior art?) once the manufacturer decides to take it off the market. By the same token you could argue that a patent application is sufficiently disclosed if the applicant offers the protected product for sale?
If someone trialling a ploughshare in the middle of nowhere counts as prior art because there was a public footpath within sight then a product placed on the market is also prior art.
DeleteI like the analogy used by the opponent in the referring decision (also adopted by the Board at reasons paragraph 20). The recipe of Coca Cola is secret and it is not exactly reproducible without that secret recipe. But we know that it is a fizzy drink (i.e. carbonated and sugary) and that it contains a caramel colour.
DeleteIf I understand the logic of the first anonymous, this does not count as prior art against someone trying to claim a carbonated, sugary, caramel-coloured drink, because Coca Cola is not reproducible. Clearly this is an absurd result. Surely what matters is that the relevant properties of the product are known, even if not every detail is known (i.e. including details which are irrelevant to the claim).
Anonymous, thanks for your comment. I was wrong to suggest that a product as prior art would need to be reproducible.
DeleteLet me first be clear that I do not know if the EBA will decide that a product itself can be prior art under Art. 54(2) (as opposed to only the information derivable from the product). In my view Art. 54(2) leaves room for accepting a product itself as prior art, and I believe this is desirable (since otherwise one could potentially patent a product that has been widely for sale for years).
If a product itself can be prior art under Art. 54(2), then all that is needed is for the product to have been made available to the public. A product that was for sale some time before the priority date was available to the public, even if no member of the public actually bought it. (Compare with a book that was put on sale: its content was thereby made available to the public even if nobody bought it.) Criteria such as "reproducibility" or "enablement" play no role here (but see below for Art. 54(1)).
A prior use such as the demonstration of a product at an exhibition normally does not make the product itself available to the public but only the information that can be derived from the demonstration. Reproducibility or enablement again should play no role for the question whether this information is part of the prior art under Art. 54(2).
However, when testing for lack of novelty under Art. 54(1), we do not compare sets of words but technical objects. A product as part of the prior art under Art. 54(2) is a technical object and it is novelty destroying if it can be shown (today) that it satisfies all the limitations included in the claim. Information as part of the prior art under Art. 54(2) is novelty destroying only if it allows the skilled person to produce a technical object that satisfies all claim limitations. So it is here where reproducibility/enablement comes in.
I believe this position is coherent, but I am open to criticism.
G 1/92 may or may not fully support my view, but with G 1/23 the EBA has the chance to overrule/refine/clarify G 1/92.
The alternative is to decide that Art. 54(2) is about information only and to then carefully deduce the consequences of this position. I would still argue that reproducability/enablement comes in at Art. 54(1), not 54(2).
The idea that a "partial" disclosure does not exist at all under Art. 54(2) cannot be right. If a document from 1980 does not disclose in sufficient detail a key aspect of a product or method, but the skilled person at the priority date in 2010 would have found a document published in 1990 presenting a solution to the technical problem of how to carry out that key aspect (in a way that gets you to something within the terms of the claim), then there should be lack of inventive step.
Your conclusion is wrong. You can use as prior art any published documents sufficiently disclosing (in reproducible technical terms) carbonated sugary drinks and there properties. For inventive step, you do not need to start from a prior use of CocaCola. This was precisely the case in the present decision. The grounds for the referral complicate things in an unreasonable manner. Cheers. First Anonymous.
DeleteNiko. I agree that what you say makes sense. But G1/92 was a very balanced decision because the 'incentive' to investigate properties is the real problem in the situation. A leaf on a tree in a park is available to all for analysis, but we accept in patent law that all the substances in the leaf are not publically disclosed and so are not prior art. The fact they could be isolated and analysed still does not make them prior art. So we need to keep the correct balance with marketed products, and saying that partial info from a marketed product can become prior art is changing the balance in quite a big way. This is quite a difficult decision for the Enlarged Board
ReplyDeleteI agree that G 1/92 was about the "incentive" to investigate, and it should be read from this perspective.
DeleteBut I'm not sure I agree with what you say about the tree leaf. The answer of G 1/92 is that a particular incentive is not necessary ("irrespective of whether or not particular reasons can be identified for analysing the composition"). All the information that could be derived from the leaf by the available methods of analysis were made available to the public, with the proviso that the analysis may not require "specifically chosen outside conditions".
An interesting question is at what date those analysis methods should have been available. I believe the prevailing view is still that novelty is to be assessed at the date that the relevant piece of prior art was made available to the public. It probably makes more sense to take the filing/priority date as the reference point. G 1/92 is strangely silent on this.
Anyway, I suppose we agree that even if we may not know everything about the internal composition of tree leaves, this does not remove leaves of known trees from the prior art under Art. 54(2) for the purpose of assessing novelty or inventive step of, say, a method of filling a bag with tree leaves. It should be enough that trees make their leaves available to the public (as products).
Thanks Niko. So according to G1/92 for a marketed product an incentive is not necessary. But for products in nature, practically speaking, the approach is very different at the EPO. Much of biotech practice, and the validity of the biotech patents that are granted, relies on the fact that nature is not disclosed (or prior art) in any analysed or characterised way. For example the individual substances in the leaf are not disclosed.
DeleteThis referral is interesting because it concerns disclosure of a polymer and polymers are very difficult to fully analyse and replicate exactly, but you can often come pretty close to replicating something similar. So essentially we need a polymer-specific decision to solve the legal issue here which takes into account the technical limitations (and vagueness) inherent in the art
What is of note in this case is that a data sheet for the commercial product ENGAGE 4800 has been available on the website of its manufacturer since 2011 and is prior art as such. This is quite different from typical "prior use" such as a one-time presentation in a conference in which the information disclosed to the public may be in doubt.
DeleteFor a commercial polymer, I assume a sample can be procured and reverse engineered. This is no "undue burden" for the skilled person, reverse engineering techniques are part of the common general knowledge.
I agree with you that the relevant date for assessing the skilled person's common general knowledge should be the filing or priority date, the same as for any prior public disclosure. This is important because reverse engineering techniques and common general knowledge are constantly improving. For example, the Al content of a polymer could be derived with current techniques whereas it could not be so easy 10 years ago.
On the other hand, taking as the relevant date the date of a prior use - quite unclear anyway in the case of a product which has been commercial for 10 years - would entail serious difficulty as to the assessment of the common general knowledge at that date.
By the way, the burden of proof as to the enabling character of a public prior disclosure does not seem to be mentioned in the referring decisions.
Thanks Francis Hagel, I believe in this particular case the polymer could be not be fully analysed and could not be replicated exactly. However it could be analysed to the extent that it was possible to tell it would fall within the definition in the claim, and a similar (but different) polymer could be made using common general knowledge, but exact replication was not possible. So the G1/92 conditions in this respect could not be met which is what makes the referral very interesting
DeleteThe referring decision mixes up things. Some properties (composition) of the allegedly prior used material are disclosed in D1, D2 and D5/5a, so for inventive step one could start from any of these documents as such, provided that they belong to the prior art, not fro the prior use of the material, which would require analysis/reproducibility according to G1/92. This could have been clarified by the board and did not need a referral.
ReplyDeleteThe last time I looked, EPC did not say that public prior use requires reproduecability. Allowing that a commercially available product might not constitute part of the prior art opens up the possibility of "forever patents". Even simple mechnical devices could be made tamper-proof so that there would be "undue burden" for the skilled person to figure out how they work and the producer could every 20 years or so patent a different aspect of a by then decades old product.
ReplyDeleteNo product is perfectly reproducible in all its details. Quantum mechanics sees to that. And even putting aside the quantum, at some levels all products are irreproducible. Compare one pencil to the next, the wood grain will be different from one to the next. Compare two diamonds cut alike, each has a different set of inclusions. Compare two injection-moulded parts, on the level of the tolerance of the process the parts will differ. All that ought to matter is "what is disclosed in an enabling manner" from any particular event, to whom, in what circumstances.
ReplyDeleteAnonymous of Wednesday, 12 July 2023 at 12:44:00 GMT+1, that is a very good point. But here it is not about 'enablement' in the sense of a claim being enabled. It is about a marketed product in the prior art, and so the test for 'enablement' of that is difficult to define. For example, can the teaching of the relevant patent specification be used to assist in enablement of the marketed product. To what extent must it be enabled (thinking of your wood grain point)? To the extent it falls within the claims? What happens if the claims are narrowed? Does replication of the marketed product stop being enabled below a certain claim scope because then the common general knowledge is no longer good enough, and you need the teaching in the spec?
ReplyDeleteSo my argument is that, in a novelty analysis, the "enablement" criterion only rears its head when you assess, under Art. 54(1), whether prior art information is sufficient to allow the skilled person to carry out a method or to create a product that falls within the terms of the claim.
DeleteSo if the prior art under Art. 54(2) includes not only information that was made available to the public but also products that were made available to the public, then "enablement" plays a role for information but not for products. A prior-art product already is a product, and under Art. 54(1) you just need to verify whether the product factually satisfies the limitations listed in the claim. Under Art. 54(2), the only question is whether the product was "made available", which is the case if it was put on sale (even if nobody ever bought it).
Ultimately patent law is not about words but about the processes and objects that are protected. A claim outlines a collection of such processes and objects, and if one of them is in the prior art, then the claim is not new. A prior-art document is just words, but it makes available to the public the products and objects which the skilled person obtains by following the instructions given in the document. Here is where one naturally needs enablement.
I agree with Niko’s comments that G 1/92 is about the « incentive to investigate ». We may call it the ratio decidendi. In that sense, the reference to reproductivity in G 1/92 could be considered obiter dicta.
ReplyDeleteI have another take from G 1/92, which is very general and from which today’s Boards should take inspiration. A broader teaching of G 1/92 is to interpret the EPC as it is written and to keep away from interpretations which amount to amendments, be it the addition (the incentive to investigate in G 1/92) or the deletion of a condition.
As to Anonymous’s reference to « tamper-proof » products, I beg to differ. Practical protection devices of a commercial product could always be dismantled by the buyer to reverse engineer it. Protection against reverse engineering could only be secured by having the buyer agree on confidentiality restrictions but this is not part of a regular sales agreement and could raise competition law issues. If it is possible for software, this is because software is covered by copyright and copyright law applicable to software does include a protection against reverse compiling, the software equivalent of reverse engineering.