US patent attorneys wishing to understand certain peculiarities of European patent drafting need look no further than the recent Board of Appeal decision in T 2171/21. This case is a textbook example of the EPO's strict approach to added matter. The decision highlights the inadequacy of mere lists of various features for satisfying the basis requirements in Europe. The decision may further help elucidate to the confused US attorney why their European colleagues continue to fill patent drafts with so-called US "patent profanities". Striking the middle ground between satisfying the strict basis requirement in Europe and avoiding patent profanity in the US, renders the decision of when to file a patent application all the more critical.
Legal background: Basis for selections from lists
The EPO has a famously strict bar for basis. Common with other jurisdictions, a European patent application or patent may not be amended to introduce subject-matter that was not disclosed in the application as filed (Article 123(2) EPC). However, the EPO has a particularly narrow definition of what it means for subject matter to have been “disclosed in the application as filed”. The EPO, for example, takes a very strict stance against the "cherry-picking" and combining of features from separate embodiments or lists of features provided in the application as filed. As such, at the EPO, it is not permissible to select and combine features from multiple lists unless the singled-out combination is specifically disclosed or pointed to elsewhere in the specification.
Added matter or economical drafting?
The Novartis patent EP 3111954 in T 2171/21 related to the second medical use of the anti-IL-17 antibody secukinumab. Secukinumab is marketed by Novartis in Europe as Cosentyx for a number of inflammatory diseases, including psoriasis and ankylosing spondylitis (AS). The SPC on the composition of matter case for secukinumab expires in 2030. The patent in T 2171/21 claimed a loading dosage regime of secukinumab in AS, expiring 2031.
Profanity |
The patent was opposed for added matter. On appeal the Opponent argued that in order to arrive at the subject matter of claim 1, it was necessary to combine multiple selections from embodiments disclosed in the earlier application as filed (of which the patent in question was a divisional). Particularly, it was necessary to 1) select AS from a list of four inflammatory diseases, 2) select the claimed dosage regimen from a list, and 3) select a particular induction dose from a list of two induction dosages. The Opponent argued that there was no pointer to the selected combination in the application as filed and as such the selection lacked basis.
The Patentee argued that an applicant should not be forced to claim or describe as individual embodiments all the possible combinations of, in this case, the dosage regimens and diseases described in the application as filed. In their particular case, the Patentee argued that a summary table provided in the description represented a shorthand for stating all the disclosed combinations individually.
Given the established case, law, it is no surprise that the Board of Appeal was not convinced by the Patentee's defence. The Board of Appeal noted:
"a difference exists between the conceptual disclosure of a number of possible combinations and the individualised disclosure of specific combinations. While the former might be a more economical way of drafting a patent application, it does not necessarily allow the skilled person to derive each and every individual combination directly and unambiguously".
The Board of Appeal agreed with the Opponent that, in order to arrive at the claimed subject-matter, the skilled person had to select and combine elements from the three different lists. The Board of Appeal was not convinced that such a selection could be considered equivalent to the gold standard for basis: namely that the subject-matter was disclosed directly and unambiguously in the earlier application as filed (r. 5).
Profane pointers
In Europe, in order to provide basis for certain selections, pointers towards certain embodiments may often be given by using terms such as "preferably" (Case Law of the Board of Appeal, II-E-1.6.2c)). However, in the present case, the Board of Appeal found no pointer in the application as filed towards the selected subject matter. If anything, the application as filed appeared to point towards a different disease (rheumatoid arthritis) and different doses. Whilst AS was mentioned in the examples, this was only in the context of a different dosage regime. The Board of Appeal thus concluded:
"The earlier application as filed therefore does not contain a pointer to the chosen combination of features. The claimed subject-matter thus represents a single individualised combination in the absence of a preference for this combination or, at least, a preference for one of the combined features having been disclosed in the earlier application as filed. The claimed subject-matter thus extends beyond what a skilled person would derive directly and unambiguously, using common general knowledge, and seen objectively and relative to the date of filing, from the whole of the earlier application as filed."
Final thoughts
In the present case, the Board of Appeal noted the absence of any "preference or pointer" towards the claimed subject matter. The selected subject matter could not therefore be said to have basis in the application as filed. The case law relating to pointers and selection inventions is well established CLBA, II-E-1.6.2c)). In T 68/99, for example the fact that certain features were disclosed as preferred in the application as filed acted as a pointer for a selected combination of features.
In the US, however, including terms such as "preferably" in the specification may unduly limit claim scope during litigation (see e.g. 212 F.3d 1241 (Fed. Cir. 2000)). "Preferably" and related terms are therefore considered US patent profanities that should be strictly avoided when drafting. However, as the EPO case law demonstrates, many US "patent profanities" are the exact same phrases which might help a selected combination of features satisfy the strict basis requirements in Europe.
How then should patent be drafted? In order to avoid both US patent profanity and ensure adequate basis, careful decisions need to be made with regards to the timing of filing and what to include in the patent application. The applicant in the present case, for example, may have had better success by a) avoiding any mention of doses that were not exemplified, and b) avoiding any mention of AS in an application primarily directed to rheumatoid arthritis. A later application to the selected disease and doses may then have been pursued after it became clear that this subject matter was commercially relevant. The danger of this approach, however, is that the novel subject matter of the selected disease and dose would also have to be demonstrably non-obvious in view of the prior art, e.g. by possessing an unexpected technical effect. Regardless, the lesson for US practitioners from T 2171/21 is that lists of features will not provide basis for a combination of features in the absence of a specific embodiment of the combination or patent profanities directing the skilled person towards the combination.
Further reading
- Does the requirement for a "technical contribution" in a novel selection survive? (if selecting from convergent lists) (T 1621/16) (March 2020)
- Adding matter by cherry-picking from separate embodiments: Philip Morris v BAT ([2023] EWHC 2616 (Pat)) (Oct 2023)
- Defining what the invention is not can be as important as defining what the invention is (T 0273/22, Chimeric antibodies/REGENERON) (Nov 2023)
- The criteria for the novelty and inventive step of pharmaceutical selection inventions (T 1356/21) (Dec 2023)
As sympathetic as I am towards the needs of US practitioners securing the broadest possible protection, I always feel like I'm talking to a brick wall when trying to reason with them about this stuff.
ReplyDelete"But if I put in that language, then it might be taken as an indication that it's what I'm trying to cover in the claims!"
Well... yes. That is, in fact, the point.
This divergence between US and European patent practice is likely to attract wider attention beyond the narrow realm of patent practitioners in the near future: in the tug-of-war between Moderna and BioNTech about parents on mRNA vaccines, BioNTech has reportedly just managed to invalidate a Moderna patent (in first instance opposition proceedings) on just this ground (the written decision is not yet out, though), so I guess it's just a matter of time until all the usual suspects in US politics and media start accusing BioNTech and the EPO of "stealing America's IP".
ReplyDeleteThis one? https://register.epo.org/application?number=EP20164728&lng=en&tab=doclist
DeleteThanks for that Link to EP-B-3718565. I see that the OD's written Decision is out. I see that Moderna in January 2023 moved away from the EPO representative that prosecuted the patent application at the EPO. I guess that happens a ,lot, when an OD revokes a patent on the ground that matter was added during prosecution.
DeleteWasn't it that the attorney moved firm?
DeleteAh! Didn't think of that. But now you mention it....yes...it seems so. Thanks for that observation.
DeleteIt must be remembered that added matter can be used by the EPO to revoke a patent when it is clear it will fail subsequently on enablement or inventive step, as added matter is normally discussed first at a hearing. This shouldn't happen, but it is human nature to get of cases as quickly as possible when it is clear a patent is not deserved. In this case the word 'pointer' probably took on a very complex meaning in that the data in the specification did not adequately reflect the claimed matter, and so the references to multiple selections and lack of pointer in the decision actually means that the data in the spec do not provide a good enough reason to justify limiting to the claimed combinations.
ReplyDeleteBut there are also worrying aspects here to how the EPO is operating. I think we need to consider whether or not added matter should be available as such a powerful ground for attack in opposition. It may be better to adopt a test similar to national courts which rarely revoke for added matter (for example the 'new information' test of the UK Courts). I think post-grant must be recognised as different to pre-grant, and added matter at the EPO is too strict to use as a tool to revoke patents.
Also there is a further issue now of how the UPC will consider added matter, and one hopes it does not emulate the EPO on this, which unfortunately is looking possible given the decisions that have come out so far from the UPC. EPO case law and procedures will be helpful to the UPC, and in view of this the Boards in particular must become more judicial and court-like. Otherwise the UPC risks becoming too much like a patent office. We might think that the UPC and Boards are different and can therefore operate different sets of case law, but given that the UPC does not have any 'history' it is unlikely to remain free of EPO influence.
Santa, could it be that the counter-balance to strict enforcement of Art 123(2) at the EPO is sufficient freedom to the courts under Art 69 EPC to decide (with justice to both the inventor and the public) what the Applicant was using the language of the claim and the specification to mean? Is it not a feature of the European system, and not a bug, that those reading WO publications can opine with confidence on whatever FTO issues might arise over the life of any patent (parent or divisional) that might get to grant over the ensuing 19 years or so? Would it really help the public, to let Applicants re-write their patent applications years after their filing date? Me, I see signs that the USA is progressively ratcheting up the strictness of its WD requirement, in the direction of the EPO's Gold Standard, as it comes to grips with deciding fairly priority contests that arise in a First to File environment (still a novelty in the USA).
DeleteThanks Max Drei. Hope you had a nice Xmas. I'm going to make a prediction that now that the EPO is getting rid of purposive selection for novelty and also increasingly relying on 'pointers' for establishing which combinations have basis, other aspects of added matter/priority are also going to get reevaluated in the coming years. I think we'll essentially see a backlash against the present strictness, and further I believe UPC case law, supported by the influence of the EU, will become intertwined with EPO case law, leading to much more leniency at the EPO. The leniency will come from the UPC realising it needs to be less formalistic than the EPO when acting judicially and deciding on huge amounts of money to be given in damages. At the moment the EPO doesn't really see see the financial implications of its decisions and therefore has become too abstract and academic. The UPC powered by the EU will force the change
DeleteGood Christmas yes thanks, Santa. Until now, the EPO has, uniquely, been free from the iron fist and heavy boot of a supervisory non-patent-specialist supreme court. To my mind, this has been a good thing for the development of a solid body of jurisprudence on patentability and validity. Now though, after 40 years, comes the UPC, weighing in on matters of validity just as much as on what amounts to an infringement of the rights granted by the EPO. I have no idea whether the EPO and the UPC will see themselves as partners, rivals or adversaries but I hope that their interaction is going to be another good thing for the development and health of the patent system in Europe.
DeleteIn 212 F.3d 1241 (Fed. Cir. 2000)), "The Federal Circuit agreed with the district court, noting that throughout the prosecution history, Bayer asserted that they were claiming “a special form of nifedipine, namely having a specific surface area of 1.0 to 4 m2/g,” which was a “superior inventive range.” (the accused product used a value above 5 m2/g). I'm curious about the case that made 'preferably' into USA patent profanity.
ReplyDeleteAs a newly-qualified European Patent Attorney I was advised by a well-experienced in-house EPA never to use "preferably" unless the feature concerned was truly preferred, but I understood at the time that that was because they were worried about the effect the word might have, no case law was mentioned. Is there a US (or other) decision on the effect of the word "preferably"?
Delete