Deletion of claim-like clauses: Board of Appeal finds legal basis for adaptation of the description from "the EPC as a whole" (T 0438/22)

The recent decision in T 0438/22 is a relatively lengthy justification of the EPO's idiosyncratic requirement for the description of a patent to be amended in line with the allowed claims. The decision is yet another contribution to the ongoing debate over whether the EPO's strict approach to amendment of the description has legal basis, In the meantime, we await a potential referral to the Enlarged Board of Appeal (EBA) on the issue (IPKat).  T 0438/22 specifically related to the requirement in the EPO Guidelines for Examination for claim-like clauses to be deleted from the description pre-grant (F-IV, 4.4). 

Amending the description and claim-like clauses

The EPO Guidelines for Examination currently require that, when adapting the description, an applicant must either delete subject matter not covered by the claims or explicitly state in the description that such subject matter is not part of the invention (F-IV, 4.3). It is argued that these amendments are necessary in view of the stipulation in the EPC that the description "shall be used to interpret the claims" (Article 69 EPC). As such, it is argued, the claims can only clearly define the invention (Article 84 EPC) if they are not contradicted by the description. However, amending the description is fraught with risk for patentees. Description amendments may add matter to the application as filed, and lead to the inescapable added matter trap post-grant. Amendments to the description may also adversely impact claim interpretation post-grant in both invalidity and infringement proceedings.

Amended semiconductors

Part of the Guidelines on adaptation of the description relate specifically to the deletion of "claim-like clauses". It is common practice when filing a divisional European patent application to add the complete set of original claims (e.g. PCT claims) to the description. These claim-like clauses can then be used as basis for future claim amendments if subject-matter in the original claims inadvertently lacked explicit basis in the original description. However, the EPO Guidelines for Examination currently state that, in order to bring the description in line with the claims: "claim-like clauses must also be deleted or amended to avoid claim-like language prior to grant since they otherwise may lead to unclarity on the subject-matter for which protection is sought" (F-IV, 4.4). However, claim-like clauses may also be used as basis for amendments post-grant (e.g. in Opposition proceedings). 

Legal basis for requiring deletion of claim-like clauses? (T 0438/22)

In T 0438/22, the Board of Appeal considered the legality of the provision in the Guidelines that claim-like clauses should be removed from the description. The case related to Intel's European patent application EP3511980 for semiconductors. The application was refused by the Examining Division because the description had not been amended to remove the claim-like clauses and, in the view of the Examining Division, thereby lead to unclarity as to the scope of protection of the claims (Article 84 EPC). 

Intel argued that there was no EBA case law justifying the stipulation in the Guidelines that claim-like clauses should be deleted. As such, Intel argued, unconditional application of the Guidelines forced the applicant to limit the disclosure of a patent application in an unjustified way. Intel cited T 1989/18, in which the Board of Appeal found that the requirement for mandatory adaptation of the description in line with the claim lacked legal basis (IPKat). Intel further argued that deletion of the claim-like clauses would lead to a loss of disclosed subject-matter that might be needed post-grant, such as basis for amendments during opposition. 

Intel also requested a referral to the EBA on the question of whether it was justified to force an applicant to delete parts of the description solely due to the wording being identified by the examining division as "claim-like". 

On appeal, the Board agreed with Intel that there was no risk of the claim-like clauses being misinterpreted as claims, and that "[t]here is no justification for deleting them just because they were drafted as claim-like clauses." (r. 3.4). The Board of Appeal also agreed with Intel that "there appears to be no decision of the boards or the Enlarged Board of Appeal that forms the basis of the relevant part of the Guidelines" relating to the required deletion of claim-like clauses (r. 3.2). 

However, on the general principle of adaptation of the description, the Board of Appeal did not agree with T 1989/18 that this requirement lacked legal basis. Instead, the Board of Appeal considered there to be "long established" case law that Article 84 EPC requires the entirety of the description to be consistent with the allowed claim, citing T 1024/18T 1808/06, T 2293/18, T 1399/17, T 2766/17, T 1516/20, T 0121/20, T 1968/18, and T 2685/19.

The Board of Appeal went on to provide what they considered to be the legal basis for the description amendment requirement. The Board of Appeal accepted that "Article 84 EPC is primarily directed to the claims and refers only indirectly to the description". However, the Board of Appeal considered that, according to the principles of the Vienna Convention, "the object and purpose of a legal provision may be derived not only from its literal wording but also from its context within the statute". As such, the Board of Appeal concluded that legal basis for requiring the applicant to adapt the description in line with the claim might "be derived from the EPC as a whole".

The Board of Appeal went on to outline in detail its view of the legal basis for mandatory adaptation of the description. The Board of Appeal concluded that "[i]t is a general and overarching objective, and as such also a 'requirement' of the Convention, that authorities, courts and the public interpreting the claims at a later stage should, as far as possible, arrive at the same understanding of the claimed subject-matter as the EPO bodies deciding on the patentability of the same subject-matter. The only tool for achieving this objective is the patent specification as the expression of a unitary legal title. The description, as an integral part of the patent specification, should therefore also serve this overriding objective, i.e. it should provide a common understanding and interpretation of the claims. If the description contains subject-matter which manifestly impedes a common understanding, it is legitimate to insist on its removal".

The Board of Appeal's reasoning on this point encroaches once again on the related issue of whether the description should be consulted for the interpretation of otherwise clear claim-language. On this issue, there are also arguably conflicting Board of Appeal decisions (IPKat). The Board of Appeal in this case explicitly disagreed with an approach to claim interpretation that does not take account of the description: "The idea that the description is entirely unnecessary for claim interpretation is unrealistic and simply 'wishful thinking', just like the case law expecting claims to be clear in themselves" (r.63)

The Board of Appeal nonetheless did accept that deletion of subject-matter from the description might be harmful to the patentee from a sufficiency perspective, given that "[r]equiring the unconditional deletion of subject-matter would often result in the deletion of enabling disclosure." In such cases, the Board of Appeal considered it "settled practice that subject-matter not covered by the claims must be 'removed' through an appropriate statement of the applicant in the description which effectively does not remove the technical teaching of the patent". However, the inclusion of such statements may still be detrimental to the patentee given that such statements could adversely affect post-grant claim interpretation, e.g. under the doctrine of equivalents, or even in US litigation (IPKat). On the plus side, the Board of Appeal did at least accept that it is not necessary to amend the description every time the claims are amended during prosecution (r. 5.4.3). 

The Board of Appeal concluded by rejecting Intel's request lacking amendment of the description. 

Final thoughts

We are currently waiting on tenterhooks for confirmation of a referral to the EBA on description amendments (IPKat). A parallel referral on interpretation of the claims in view of the description will be discussed at oral proceedings in appeal case T 439/22 in April (IPKat). If the referral on description amendment goes ahead, and is accepted by the EBA, then this latest decision in T 0438/22 is a preview of the potential legal arguments that will be submitted in favour of maintaining the status quo stipulated by the Guidelines. 

Further reading

Claim interpretation

Description amendments

Image credit: MidJourney
Deletion of claim-like clauses: Board of Appeal finds legal basis for adaptation of the description from "the EPC as a whole" (T 0438/22) Deletion of claim-like clauses: Board of Appeal finds legal basis for adaptation of the description from "the EPC as a whole" (T 0438/22) Reviewed by Rose Hughes on Sunday, January 21, 2024 Rating: 5


  1. IMO R42 does not leave room for claim-like clauses at the end of a description

    1. R42 is not an exhaustive list of everything the description can contain, simply a minimum requirement

  2. How can we distinguish a claim-like clause? Let's assume the description ends like this:

    "The disclosed product may comprise one or more of the following features:
    1. A.
    2. B.
    3. etc."

    Is this a claim-like clause?

  3. This review reads to me like the "claim-like clauses" had to be deleted- they didn't, they had to be neither inconsistent with, nor a repetition of, the claims. They appear in the now issued 71(3) communication.

  4. Goodness knows how the courts and the EPO can reach a "common understanding" of what the words of the claim actually mean when the courts are required to factor Art 69 EPC into their textual analysis whereas the Examining and Opposition Divisions are expressly required not to. It occurs to me that this TBA, by invoking the importance of a "common understanding, has found a tactful way to draw attention to the idiosyncracy of the Guidelines without, in doing so, annoying EPO management. Very political!

    Claim-like clauses are not as big an issue, I would say, as illustrated embodiments at the edge of the literal wording of the claim which might or might fall within the Art 69 ambit of the claim.

    1. Dear Maxdrei
      Is the answer not given in the decision? A common understanding is reached by adaptating the description to the granted claims so that inconsistencies between them are avoided and so that the claims are (fully) supported by the description.

    2. Dear MaxDrei,

      Three years ago, you posted the following comment on the Kluwer Patent Blog post reporting the release of the 2021 Guidelines setting out the new policy regarding adaptation of the description :
      « These days, I have the feeling that, inside the EPO, management doesn’t give a toss about the law or the extent of protection given to the customers. All that interests them is “efficiency”, driving down cost and driving up profit. The writer of this useful blog post surmises that EPO management saw the high rate of rejection of Examiners’ scribbled amendments in the Druckexemplar as a drag on efficiency and profit, so resolved to do something about it. »

      In hindsight, your comment seem full of insight.

      What we have seen in the string of BOA decisions supporting the requirement to adapt the description is that the legal basis they put forward has shifted from an expanded interpretation of the support prong of Art 84 in the initial decisions to the first line of Art 84 alleged to require a « precise determination of the scope of protection » and to clarity alone.

      The global picture which emerges is that the adaptation of the description has to be completed, irrespective of the rationale which is put forward to justify it.This fully corroborates your "feeling".

      Practically speaking, based on feedback from colleagues, the current situation is that description amendments entered by the ED at the 71(3) stage are inconsistent across EDs and frequently lack explanation. A colleague told me yesterday that he had to challenge the amendments in one out of two cases. And he is obliged to charge his time for that, whereas his usual practice is to charge nothing for the 71(3) stage, considered purely administrative.

      Some opinions now regarding a referral to the EBA are that it could be held inadmissible because the grant procedure should be stayed until the EBA delivers a decision (compromising the "production" objectives of the management. What an odd situation ! Has anybody an idea on how to get out of this stalemate ?

  5. The present decision has been commented in another blog.

    The question of claim-like clauses could be solved, either by a decision of a BA or the EBA, which would be taken up in the Guidelines. Another way to solve the problem would be by amendment of the Implementing Rules simply banning claim-like clauses in general. After all, they are not foreseen in the EPC and find ther origin in US law and prosecution.

    The second part of the decision on the adaptation of the description is more important, as the board confirms the long lasting line of case law in this matter.

    What is even more important is the referral about the interpretation of the claims under Art 69 and the Protocol.

    A referral on this question is much more useful as a referral on the adaptation of the description, as the case law on application of Art 69 is clearly divergent, which is not the case for the adaptation of the description.

    1. By "another blog", I assume you mean

      Claim-like clauses are simply one (highly efficient and concise) way of presenting a large number of embodiments of an invention. It would therefore be rather counter-productive to ban them, not least because alternative ways of presenting the same number of embodiments could be much harder going for both examiners and the public (eg assigning numbers or codes to key features and then providing long lists of combinations of numbers and/or codes). Is that really what you want?

    2. @ Proof of the pudding alias Doubting Thomas

      Thanks for advertising my blog.

      Claim-like clauses are an import from the US, and I am not sure at all that they have their place in EPO procedures. Why on Earth is not possible to state those different items in plain language and present them as proper embodiments as it should be the case for what I call a “normal” description. It looks rather odd to have on the one side a “normal” description, which is then complemented by claim like-clauses.

      Assigning to key features of a claim a number or codes is a way of breaking down the features of a claim and is actually the result of what is done when a prior art matrix is established in order to assess the prior art submitted by the opponent with respect to the claim at stake. This way of working has first been presented in EPO procedures by German representatives and has progressively be used by other representatives.

      Arguing only on the basis of the numbers or codes can be very dangerous as after a while, one loses track. When using number and/or codes it is best to always actually state the feature at stake next to the number and/or code. This avoids big mistakes in the argumentation. I have seen many of those in my long career as examiner.

      Providing long lists of combinations of numbers and/or codes is the worse what an applicant/proprietor could do. It is to me equivalent to a matrix presentation of AR. Only in exceptional cases, such a matrix has been accepted by boards. What is required by the EPC are written-out claims.

      Checking claim-like clauses, either to accept or to object to them, requires a lot of work. With the aim to render the procedure more efficient, I would not be surprised in order to end any discussion on claim-like clauses, which are clearly not foreseen under the EPC, that the EPO will eventually amend the Implementing Regulations so as to ban anything which looks like a claim, but is not a claim but for instance merely a claim-like clause. To take over your wording: “Is that really what you want?”.

      R 43(2) has as primary object avoiding multiplicity of independent claims of the same category, with on top overlapping scopes of protection. This is US practice, but not EPO practice, hence the ban.

      Claim-like clauses are a constant source of headache for EDs/ODs: why are they there, what is their purpose, can it not be said in another way, what is hidden behind them?

      As representative you want a positive answer to your requests from the deciding body. Insisting on claim-like clauses is a good way to antagonise a deciding body. Then do not be surprised that, within the boundaries of the valid legal rules, the discretion of a deciding body tends asymptotically to 0. Taking over once again your wording: “Is that really what you want?”.

    3. Yes indeed, Pudding, but isn't the point that any one patent application must be directed to just one invention and that the claim-like clauses are directed to an invention that is not the same one as is defined in claim 1 intended for grant? Isn't that the reason why the claim-likes have to go?

    4. Max, I believe that you are envisaging only one (quite unusual) type of claim-like clause.

      Claim-like clauses are nothing other than one possible way of presenting embodiments in a description. Like anything, they can be used for both "good" and "bad" purposes.

      The "good" purposes include:
      - ensuring that the description of a divisional application captures all of the subject-matter that is (clearly and unambiguously) disclosed by way of the original claims of the parent application as filed; and
      - (clearly and unambiguously) disclosing specific embodiments of the invention.

      What you and Mr Thomas seem to be envisaging, however, is the "bad" purpose of retaining in the description subject matter that is not encompassed by the claims.

      It is pointless to tar all claim-like clauses with the same brush. That is because doing so will end up with throwing the baby out with the bathwater. I mean, do we really want to "ban" claim-like clauses solely because we are suspicious of the format they use for disclosing what could simply be embodiments of the invention that fall within the scope of the claims? Would that not represent a triumph of form over substance?

    5. Mr Thomas, I refer you to my reply to Max Drei. I have presented you with perfectly good reasons for using claim-like clauses. Your experience may make you suspicious of those clauses. However, that it not a valid reason for banning what boils down to one possible format for (clearly and unambiguously) disclosing embodiments of an invention. As you seem to agree, other formats could be much worse. Do you propose to ban them all?

      At the end of the day, the EPO only has itself to blame for the rise of claim-like clauses. This is because, when used for a "good" purpose, they represent an attempt by the author to ensure that the embodiments that they (concisely) describe are viewed by the EPO as being clearly and unambiguously disclosed in the sense of case law on Art 123(2).

      It is of course important to avoid antagonising a deciding body. But what choice does one have when that body develops practices for which one can see no legal justification, especially when those practices are actively harmful to the legitimate interests of applicants?

    6. @ Proof of the pudding alias Doubting Thomas,

      In my opinion, claim-like clauses have to be banned.

      I gave given good reasons for this in my last reply of today to Mr Hagel under the link you have given above. I will not repeat them here.

      In your opinion, deleting claim-like clauses and adapting the description to the claims have no legal justification. You will not be surprised that I do not share this view as I rather consider that applied the right way they cannot actively harm the legitimate interests of applicants.

      Our exchanges under the same link are showing our different approaches, and do not need to be repeated here. For me the topic of adaptation of the description is exhausted. The deletion of claim-like clauses as well.

  6. So is this clause allowable?

    "The disclosed product may comprise one or more of the following features:
    1. A.
    2. B.
    3. etc."

    Moreover, what if I shift a feature X from a dependent claim "Product according to one of the previous claims, wherein X" to the description ("The disclosed product may comprise X")?

    Is this allowed? If not, why?
    Clarity? It is crystal clear.
    Added-matter? X already disclosed as optional from the beginning.
    Support? Feature X falls within the scope of the claims.

    It seems to me that the EPO would like to get claims fees also for deleted claims...

  7. @DXThomas

    The suggestion in T 439/22 for a referral about the interpretation of the claims under Art 69 is ill-founded, there is no divide of the BOAs on this issue.

    Article 69 is specifically aimed at the determination of the extent of protection and not at the examination of applications. During the examination, the settled case law regarding the interpretation of claims is that their terms must be given their “broadest technically sensible meaning” : see CLB I.C.4.1 and CLB II.A.6.1.

    This implies that the description cannot be used to “read” into the claims unclaimed features disclosed in the description. If the applicant relies on an unclaimed feature to overcome an objection and the ED finds the argument persuasive, the unclaimed feature becomes “essential” and the ED must require the applicant to enter the unclaimed feature as a limiting feature into the claim. This implements the principle of the primacy of the claims. The general rule of claim interpretation of the case law referring to the broadest sensible meaning thus relies on the wording of the claims and must not take into account the description.

    In contrast, the interpretation under Article 69 by a court may depart from the literal wording of the claims, whether equivalents are considered or not, and takes into account the description. In short, the rule of claim interpretation is heavily context-dependent.

    The general rule of interpretation during examination may have an exception though, when the description provides a dictionary for the terms of the claims. As stated in CLB II.A.6.3.3,: « In several decisions the boards have stated that terms used in patent documents should be given their normal meaning in the relevant art, unless the description gives them a special meaning (see e.g. T 1321/04, T 1089/11, T 25/15, T 1844/15). According to the established case law, the patent document may be its own dictionary (see e.g. T 311/93, T 523/00, T 1192/02, T 61/03, T 1321/04, T 1388/09, T 2480/11, T 1817/14). »

    This is precisely the case for the patent in suit in T 439/22, EP3076804, in which the description contains a broad definition for the « gathered sheet », and the question is whether the interpretation should take into account the definition of the description or the wording of the claim.

    It is very striking that the Board’s suggestion for a referral to the EBA makes no mention whatsoever of the settled case law defining the general rule of interpretation. What the Board does is to cherry pick the debate in the case at hand around the applicability of the « description as a dictionary » exception, and extrapolate from this debate a general conclusion that there is a divide between BOAs about the use of the description in the interpretation of claims, citing just a couple of decisions, so as to raise the issue of the application of Art 69 to the examination of applications.

    But there is no more divide between the BOAs about the « description as a dictionary » exception, as clear from the quote of CBL II.A.6.3.3 above, than about the general rule. Both the general rule and the exception are settled case law. The only question is the applicability of this exception, which can only be assessed on a case-by-case basis.

    Absent a divide between the BOAs as to the general rule of interpretation during examination and as to the « description as a dictionary » exception, there is absolutely no justification for a referral to the EBA. My guess is that the Board’s suggestion is an attempt to create a diversion to block the referral on the adaptation of description suggested in T 56/21.

    1. Mr Hagel.

      I will repeat here what I have said in my blog in reply to your last entry on the same decision:

      When you claim that there is no divide between the boards on the application of Art 69, I beg to disagree.

      I have repeatedly shown that there is a clear divide between the boards on the application of Art 69. It is thus high time for this problem to be tackled and a satisfactory solution given by the EBA. Simply stating that it is case dependent will not be enough.

      The problem of the claim-like clauses is a minor problem compared with that of the application of Art 69. I maintain that claim-like clauses are not foreseen in the EPC. An amendment of the Implementing regulations in order to ban those can be expected.

      Not later than in my blog on EP3076804 - T 439/22 I have shown the that the following lines of case law can be distinguished.

      1) No recourse to the description to interpret a claim
      T 175/98, T 515/19, T 2370/16, T 1778/17, T 1583/17, T 21/16, T 1853/19, T 19/18 or T 1989/18.

      2) Interpretation of the claims in the light of the description under Art 69
      T 16/87, T 416/87, T 684/95, T 587/95, T 942/96, T 275/10, T 1473/19, T 450/20.

      2a) Recourse to the description only in case of ambiguity
      T 481/95, T 169/20.

      3) Neither one nor the other – Use of Art 69 in opposition
      T 274/94, T 1271/05, T 2309/11, T 42/22
      T 42/22 A bit of both

      4) Use of Art 69 only in opposition when assessing Art 123(3)
      T 1271/05, T 131/15 or T 42/22.

      5) Use of Art 69 only when assessing IS
      T 274/94.

      The present list does not claim to be exhaustive and only serves to give examples of the diverging case law. The list also shows that the problem is not recent. It has simply been amplified by recent decisions.

  8. The EPO does like to have its cake and eat it. On the one hand it relies on precedents when it suits it to do so, while on the other hand, I have read several case reports where the attorney citing case law has been told that, unlike anglo-saxon jurisdictions that are based on precedent case law, the EPO is not, being code-driven by the provisions of the convention and its implementing regulations.

    1. I am also an ex-examiner, and your comment “The EPO does like to have its cake and eat it” is difficult to understand.

      As an ex-examiner, you should you be aware that only decisions of the EBA are binding for examiners and the boards of appeal, but decisions of the boards of appeal are only binding for examiners provided they are taken over in the Guidelines. Boards make regularly clear that they are not bound by other decisions of the boards, even of the same board.

      It is thus no surprise that on the one hand the EPO relies on precedents, EBA decisions, while on the other hand, it does not. To claim that it only relies on precedents when it suits it, appears thus not to be correct. There is no discretion for deciding bodies of the EPO not to follow decisions of the EBA, and the only limits for boards are those found in Art 20 and 21 RPBA.

      It is clear that the approach in common law countries, e.g. the UK and Ireland, where case law plays a major role in the interpretation of the law, this is not the case for the other contracting states of the EPC. This difference in approach can cause some tensions, but it is clear that the EPC/EPO system is not primarily based on common law.

      Anglo-Saxon representatives might thus come to different conclusions as to the influence of case law on the development of judicial/administrative practice. This is however part of life’s rich tapestry and has to be accepted.

    2. Ex-examiner, I would qualify the approach adopted by the EPO as being "code-driven by the provisions of the convention and its implementing regulations and the interpretations of those provisions as set out in the Guidelines for Examination".

      It is really the latter that represents the biggest problem. In essence, the authors of the Guidelines have greater power than any Board of Appeal (except, perhaps, the EBA) with regard to determining the interpretations of the provisions of the EPC that examiners will apply in first instance proceedings.

      The example of "claim-like" clauses illustrates this point well: the GL say that they are not acceptable, whereas there are no Board of Appeal decisions that reach that conclusion (and at least two that reach the opposite conclusion). In practice, this means that (most) examiners will follow the views of the authors of the GL, and will not be persuaded by contrary conclusions reached in the only on-point decisions of Boards of Appeal.

      One would have thought that the authors of the GL would be aware of the inherent unfairness of such situations and would therefore endeavour, as much as possible, to ensure that all practices set out in the GL have robust support in the case law of the Boards of Appeal. Apparently, however, such a common-sense perspective of view does not find much favour at the EPO, at least not when it comes to certain topics, including Art 84.

  9. When is a clause not "claim-like"?
    What are the statements of invention if not claim like clauses (as some drafting reiterate the claims). Should these be deleted from the description?
    It is said the claims must be fully supported by the description.
    ...and the description must be amended to be fully supported by the claims.
    Perhaps we should just delete the description and replace it with the claims at the 71(3) stage.
    Although there is a risk these may need to be deleted as "claim-like" clauses.


All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here:

Powered by Blogger.