Functional Features in Antibody Claims: The UPC vs. the EPO

Abstract

Functional features play a crucial role in securing robust patent protection for antibody inventions and have been the focus of extensive case law from the European Patent Office’s Technical Boards of Appeal. This article analyzes key patterns and emerging trends within this case law, explores their application in examination practice, and provides a comparative perspective on recent jurisprudence from the Unified Patent Court. A particular focus is on the EPO’s concerning shift towards demanding structural predictability in the context of functional features.

I. INTRODUCTION

European patents are granted by the European Patent Office (EPO) for currently 39 contracting states, making the EPO the largest of the regional patent offices worldwide. In addition, European patents can be extended to seven further extension/validation states, two of which are expected to accede to the European Patent Convention in the near future, increasing the number of contracting states beyond 40.¹

While European patents are granted in identical form for a large number of countries, the national case law on patent infringement remains persistently divergent, in particular with respect to the doctrine of equivalents. Long-standing efforts to remedy the disparate practices and promote pan-European harmonization have recently culminated in the establishment of the Unified Patent Court (UPC), a multinational specialized court currently having exclusive competence in relation to European patents in 18 of the Member States of the European Union. During a transitional period, the UPC operates alongside national courts in these 18 states, adding a further layer of complexity.

Absent a harmonized approach to infringement under the doctrine of equivalents, innovators seeking to secure effective patent protection enabling the tremendously costly development of biological agents for medical applications, such as therapeutic antibodies, can only rely on a reasonable interpretation of the literal scope of protection of their European patents. Consequently, it is often crucial to obtain patents with claims extending beyond the specific exemplified antibodies, encompassing variants that can be made without inventive effort once the specific antibody species contributed by the inventors has been made available by the publication of the corresponding patent application. This is where functional features come into play.

In biotechnology, the history of functional features in the EPO already began in the 1980s. One of the first cases was the polypeptide expression in bacterial hosts case, T 292/85 (January 27, 1988). The patent at issue contained claims directed to recombinant plasmids for the transformation of any bacterial host, comprising any heterologous DNA encoding any functional heterologous polypeptide under the control of any homologous regulon. The Examining Division rejected the application for lack of enablement because the claims covered bacteria, plasmids, etc., that had yet to be invented, concluding: “In general, no components should be defined in functional terms in this field of technology.” Thus, “components of the future,” frequently also called “embodiments of the future,” should not be allowed to be covered. The Board of Appeal disagreed and paved the way for relying on functional features for characterizing embodiments in biotech patent applications: 3.1.

Components of the future

3.1.1

… According to the Examining Division[,] this situation contradicts the suggested requirement that all embodiments within the claims should be reproducible at will by the skilled person without having to make an invention.

3.1.2

There is, however, in the opinion of the Board, no such requirement in the European Patent Convention, nor is such principle established in normal patent practice within the Contracting States. The suggested features in the claims are essentially functional terms in this particular context, in spite of structural connotations, and may cover an unlimited number of possibilities. It follows that the features may generically embrace the use of unknown or not yet envisaged possibilities, including specific variants which might be provided or invented in the future. This Board concurs with the decision of another Board (T 68/85-3.3.1., “Synergistic herbicides”, OJ EPO 1987, 228) in which the possibility of using functional terminology in claims was approved if “such features cannot otherwise be defined more precisely without restricting the scope of the invention” and their reduction to practice was not an undue burden. The Board sees no valid reason why this should not be equally true for the field of biotechnology as in other fields of technology.

…

3.1.4

The objection raised against the terms “plasmid” and “bacteria” that they are too broad since some of them rely on yet unavailable entities is untenable. The Board is of the opinion that this is quite normal practice in many technical fields where terms as “carriers”, “resilient means”, or “amplifying means” are commonplace and embrace new components, be they inventive or not. …

3.1.5

The above examples show that (“the need for a fair protection”) governs both the considerations of the scope of claims and of the requirements for sufficient disclosure. Unless variants of components are also embraced in the claims, which are, now or later on, equally suitable to achieve the same effect in a manner which could not have been envisaged without the invention, the protection provided by the patent would be ineffectual. Thus[,] it is the view of the Board that an invention is sufficiently disclosed if at least one way is clearly indicated enabling the skilled person to carry out the invention.

(emphasis added)

The same rationale was applied by the Board in another very early case, the α-interferon case, T 301/87 (February 16, 1989). The granted claims were directed to particular α-interferon-encoding DNA sequences as well as to DNA sequences hybridizing to these and encoding a polypeptide of the α-interferon type, i.e., having α-interferon activity. The decision stated: 4.4

The present case is somewhat different but nevertheless also relies on an open definition which relates to an unknown but probably finite number of human and animal interferons of the α-type. These materials would somewhat differ in constitution from each other but still represent some structural similarity in view of the affinity in hybridisation tests. Furthermore, as a class, the members provide end products with the same biological activity. As long as this is achieved by the invention there is no necessity to provide instructions in advance how each and every member of the class would have to be prepared. In view of the nature of the technique there is not even a guarantee that the same product is obtained from the same source after an identical repetition of the complicated and lengthy experiments. At this broad level, any one member of the class is an adequate representant of the invention.

(emphasis added)

Having referred to T 292/85, the Board concluded in T 301/87:

4.7… Unless claims with such functional connotations are allowable, no worthwhile protection is provided against a third party which faithfully repeats the process of the patent and obtains new but equally useful variants of the invention.

…

4.13 The requirement for sufficiency is not a matter of satisfying the perfectionist but to enable the skilled person to handle the invention in normal practice.

(emphasis added)

The notion of rewarding innovation by accepting claims that literally embrace equally useful variants in a claim scope defined by functional features was also applied—self-evidently—in the patenting of antibody inventions in the EPO. Early examples of cases where enablement was acknowledged for claims defining classes of antibodies in entirely functional terms include T 512/94 (June 23, 1998) and T 431/96 (February 23, 1999). The reasoning of T 292/85 and T 301/87 is still clearly recognizable in the later decision T 1300/05 (July 11, 2006), concluding in point 17 of the reasons that when antibodies are functionally defined in a claim, the wrong question to ask is whether all potential antibodies encompassed can be made. Rather, the right question is whether a person skilled in the art can obtain an antibody as claimed if desired; see also section IV.B., infra.

The approach to enablement underlying the above-discussed decisions essentially concludes that classes of biological molecules can be claimed using functional limitations because there is no necessity to be able to predict which other “structure,” i.e., which specific molecules covered by the functional language, would be equally useful. This principle went hand in hand with the Technical Boards’ established position rejecting “structural non-obviousness,” i.e., repeatedly holding that a biological agent such as a protein, a nucleic acid (e.g., an allele), or an antibody is not rendered inventive where it (merely) has a novel structure but otherwise the same functions and properties as corresponding agents already known in the art. Examples of decisions applying this notion when assessing claims directed to antibodies include T 512/94 (June 23, 1998), T 735/00 (March 23, 2004), T 187/04 (January 11, 2007), and T 605/14 (June 7, 2018).²

In its decision Amgen v. Sanofi-Aventis and Regeneron Pharmaceuticals , the UPC Court of Appeal applied the Technical Boards’ historical approach to enablement, as we shall see in section II., infra. While this UPC jurisprudence appreciates and further solidifies the justification of functional characterizations/limitations for classes of biological molecules (such as antibodies in biotech cases), it seems that the Technical Boards may be gradually losing such appreciation. Their new tendency is to consider functional definitions as inadequate when there is no structural predictability or, in the alternative, as inadequate because they merely characterize a “result to be achieved.” Thus, it seems as if the Technical Boards, after decades of consistent case law, have now come to agree with a line of reasoning that countless opponents have put forward, largely in vain, since the 1980s. It seems questionable whether this new approach reflects a reasonable appreciation of innovation and the diligent science behind it, not to mention the vast cost of such innovations.

The opinion of the authors of this publication is that we should not return full circle to where we started before the advent of the seminal decisions T 292/85 and T 301/87. If innovators were not awarded patents protecting antibodies (or other biologics) in functional terms (and thereby conferring a reasonable scope of protection reflecting the actual contribution of the corresponding invention to the art), they would be left with patents narrowly claiming a limited number of structurally defined antibody species. While such narrow patents may still protect against biosimilar copycats, they fail to prevent competitors from developing “me-too” (or fast-follower) drugs.³ It may be tempting to assume that preventing broad patents could be desirable, as it would reduce freedom-to-operate risks and thereby boost the number of new therapeutic antibodies undergoing development and marketing approval.^4,5 Ultimately, however, this would come at considerable cost: it would reward incremental innovation while devaluing fundamental inventions and platform technologies, e.g., discoveries that make available entirely new classes of pharmaceuticals. The development of novel first-in-class drugs is typically a high-risk endeavor that is not made easier if broad patent protection for a corresponding novel class of drugs is denied as a matter of principle. In practice, it is the task of patent offices and courts to strike a fair balance. Nevertheless, the necessity to allow reasonable patents claiming a whole new antibody genus as a reward for a corresponding fundamental innovation cannot be forgotten.

II. UPC_COA_528/2024 AND UPC_COA_529/2024, AMGEN V. SANOFI-AVENTIS AND REGENERON PHARMACEUTICALS (NOVEMBER 25, 2025)

In its decision in the joint cases UPC_CoA_528/2024 and UPC_CoA_529/2024 issued November 25, 2025, the UPC Court of Appeal (CoA) dealt with European patent EP-B1 3 666 797 held by Amgen relating to antibodies directed to the proprotein convertase subtilisin kexin type 9 (PCSK9). The technical contribution of this patent was, in essence, the finding that such antibodies are capable of blocking the activity of PCSK9 to degrade the low-density lipoprotein receptor (LDLR). LDLR is expressed on the surface of liver cells and is important for removing low-density lipoprotein cholesterol (LDL-C) from the bloodstream. LDLR binds LDL-C on the cell surface and transports it into the cell, where it is metabolized. LDLR returns to the cell surface and begins another cycle of removing LDL-C from the bloodstream. Thus, blocking the activity of PCSK9 can result in an effective treatment or prevention of hypercholesterolemia or atherosclerotic disease. With this understanding, Amgen’s patent had the following second medical use claim for antagonistic anti-PCSK9 antibodies:

A monoclonal antibody or an antigen-binding fragment thereof for use in treating or preventing hypercholesterolemia or an atherosclerotic disease related to elevated serum cholesterol levels;

or for use in reducing the risk of a recurrent cardiovascular event related to elevated serum cholesterol levels;

wherein the monoclonal antibody or the antigen-binding fragment thereof binds to the catalytic domain of a PCSK9 protein of the amino acid sequence of SEQ ID NO: 1, and prevents or reduces the binding of PCSK9 to LDLR.

(emphasis added)

The dispute arose because each party marketed their own cholesterol-lowering antagonistic anti-PCSK9 antibody allegedly falling under the scope of the claim: Amgen’s drug was sold under the name “Repatha^®,” Sanofi and Regeneron sold their drug under the name “Praluent^®.”

The UPC CoA’s decision provides guiding principles in headnotes 1 to 22 for standards applicable to UPC litigation concerning claim interpretation and the assessment of added matter, sufficiency of disclosure and inventive step. These standards find basis in historically developed case law of the EPO’s Boards of Appeal, where the formal support for amendments was reasonably assessed and functional features were accepted for claim drafting. Most use terminology that is either literally identical to historically developed terminology of the Boards of Appeal or highly analogous to and recognizably originating from terminology coined by the Boards of Appeal over the years. These parallels will be illustrated in the following by connecting the UPC CoA’s headnotes on enablement to analogous passages in either the EPO’s Case Law Book, 11^th Edition, 2025 (“CLBA”) or corresponding Board of Appeal decisions (both in bold text). 5.

Sufficiency has to be examined on the basis of the patent as a whole, thus on the basis of the claims, description, and drawings, from the perspective of the skilled person with his common general knowledge at the filing or priority date. (Case Law Book, 11th ed. 2025 (CLBA), II.C.1, II.C.3.1, II.C.4).

The test to be applied is whether the skilled person is able to reproduce the claimed subject matter on the basis of the patent without any inventive effort and without undue burden. An invention is sufficiently disclosed if the patent specification shows the skilled person at least one way—and in case of functional features: one technical concept—of performing the claimed invention. (CLBA II.C.5.4, T 292/85, T 19/90).

Where a claim contains one or more functional features, it is not required that the disclosure includes specific instructions as to how each and every conceivable embodiment within the functional definition(s) should be obtained. A fair protection requires that variants of specifically disclosed embodiments that are equally suitable to achieve the same effect, which could not have been envisaged without the invention, should also be protected by the claim. Consequently, any non-availability of some embodiments of a functionally defined claim is immaterial to sufficiency, as long as the skilled person through the disclosure is able to obtain suitable embodiments within the scope of the claim. (T 636/97, T 1300/05, T 617/07, T 1856/16).

A reasonable amount of trial and error does not prevent the invention from being enabled. (“concept fit for generalization,” “no research project,” CLBA, II.C.6.7, principles laid down in T 292/85, T 301/87 and T 412/93 (“hybridization language” encompassing embodiments yet to be found by screening), these principles are presently seen rather critically by the Technical Boards and the first-instance departments of the EPO).

The burden of presentation and proof lies with the party invoking invalidity of the patent. (T 19/90, T 843/04, T 2224/08 and T 518/10 on the standard “serious doubts substantiated by verifiable facts” and the onus of proof).

The reasons of the UPC CoA decision also provide some interesting statements concerning the sufficiency of a functionally defined antibody genus: 111.

… The skilled person would find sufficient guidance to use the entire PCSK9 protein to raise antibodies that bind the catalytic domain. A further screening step would enable the skilled person to identify whether the antibodies obtained bind to at least one amino acid of the catalytic domain.

114.

… A reasonable amount of trial and error does not prevent the invention from being enabled. Respondents have not substantiated that it would not be possible, or at least not without undue burden, to obtain co-crystals of PCSK9 antibodies with PCSK9 and to map its interaction site, … (comparable with T 1225/07).

115.

… Occasional failure is part of scientific work, … (comparable with T 292/85, G 1/03, T 891/07, T 103/09).

117.

… What Respondents brought forward only shows that Amgen has not produced such antibody itself. The fact that it would be tricky to produce one binding to any of the 15 amino acids in that region, as one of the inventors stated, does not constitute evidence that EGFa mimics cannot be produced at all.

120.

… Respondents have furthermore not brought forward anything that would justify the conclusion that the skilled person would seriously doubt that antibodies that satisfy the features of the claim … are therapeutically effective for the diseases mentioned in the claim. …

(emphasis added)

Consequently, mirroring the historical case law of the Technical Boards of Appeal of the EPO and the resulting established practice, the UPC Court of Appeal appreciates (i) that innovators need fair protection encompassing variants of specifically disclosed embodiments that are equally suitable to achieve the same effect and that could not have been envisaged without the invention, and (ii) that the non-availability of some embodiments of a functionally defined claim is immaterial to sufficiency as long as the skilled person in view of the disclosure is able to obtain suitable embodiments within the scope of the claim without inventive effort/undue burden.

In parallel EPO opposition proceedings, Amgen’s patent was maintained as granted in the first-instance decision of the Opposition Division. The subsequent appeal case T 716/25 should have been decided in oral proceedings scheduled for April 2026. Following a global settlement between the parties, however, the appeal was withdrawn on March 12, 2026, thereby terminating the appeal proceedings and depriving Board 3.3.04 of the chance of engaging with the UPC CoA’s reasoning.

III. THE 2026 EDITION OF THE EPO’S GUIDELINES FOR EXAMINATION ON ANTIBODIES

Since 2021, the Guidelines for Examination in the EPO have contained a chapter dealing specifically with the patentability of antibodies. This separate chapter attests to the importance of this specific group of biological molecules, the peculiarities of how they can be claimed, and the corresponding large body of pertinent case law. In the 2026 edition, this special chapter can be found in Part G-II, 6. As set forth therein, antibodies can still be defined by reference to the target antigen and further functional features; see Part G-II, 6.1.3. Functional features can be features such as binding affinity, neutralizing properties, induction of apoptosis, internalization, inhibition or activation of receptors, and reference to the recognized epitope. However, it is required that the skilled person is able to produce further antibodies having the claimed functional properties without undue burden. In addition, the functional definitions must allow the skilled person to easily and unambiguously verify whether they are working inside or outside the scope of the claim.

It is further indicated that antibodies can also be defined by a combination of structural and functional features; Part G-II, 6.1.4. For instance, the Guidelines indicate that it is possible to claim an antibody characterized by the sequences of both variable domains or complementarity-determining regions (CDRs) with less than 100% sequence identity when combined with a clear functional feature. However, in actual practice before the EPO, it is becoming increasingly difficult to have a claim allowed that recites less than 100% identity of specifically exemplified CDRs, if not the entire variable region.

Thus, the Guidelines expressly accept that antibodies can still be defined in the claims of European patents by relying on functional features. Unfortunately, as noted, this is not consistently followed by Examining Divisions in actual practice; see also section IV.C., infra.

IV. DEVELOPMENTS IN THE TECHNICAL BOARDS’ CASE LAW ON FUNCTIONAL FEATURES IN ANTIBODY CLAIMS: VARIABLE (V) Regions

The variable (V) regions of antibodies can be modified to control immunogenicity and antibody recycling. The development from rodent monoclonals to chimeric antibodies, to CDR-grafted, to humanized, and finally to fully human antibodies has significantly reduced immunogenicity. It has also been established, for instance, that amino acid substitutions in the V regions can modulate relative antigen binding at the physiological pH of 7.4 as compared to the endosomal pH of 5.5. This generally applicable approach influences the pharmacokinetics of antibodies by affecting their recycling and, thereby, their half-life in blood.

A. The relevance of framework sequences for the function of the CDRs

In T 2127/16 (Technical Board 3.3.04, June 9, 2020), claim 1 related to a genus of anti-Aß-derived diffusible ligand (ADDL) antibodies characterized by the 3 VH CDRs and the 3 VL CDRs. The Board was satisfied that the technical effect captured by recitation of the binding specificity and reference to the 6 specific CDRs sufficed to delimit the claims from the prior art. Thus, a further functional definition besides the anti-ADDL binding was not necessary. The claimed class of antibodies was acknowledged to be enabled and inventive.

An analogous position was taken by the Board in T 2589/16 of November 16, 2021 (Technical Board 3.3.04), noting in point 84 that the CDRs determine an antibody’s binding specificity to an antigen.

Board 3.3.04 in a different composition decided T 1628/16 on September 16, 2020, dealing with claims where the antibodies were defined by the 6 CDRs and the technical effect was increased binding affinity. The Board held that the identification of the 6 CDRs and the specificity in a claim are not enough to capture that technical effect due to the influence of the framework regions (FRs) on the binding affinity. They held that the FRs must be selected according to specific criteria in such a case; see points 11–18 of the reasons. Given these facts, this decision does not contradict the rationale relied on in T 2127/16 and T 2589/16.

Consequently, it appears as if specifying the FRs is not required when binding specificity is the basis for inventive step but is required when affinity is the basis.

B. Cases where the characterization of V regions by functional features was accepted

T 1300/05 by Board 3.3.08, decided on July 11, 2006, was and is—since its rationale was fully adopted in the UPC Court of Appeal’s decision Amgen v. Sanofi-Aventis and Regeneron Pharmaceuticals of November 25, 2025—a landmark case; loc. cit. The claims were directed to a genus of antibodies defined solely by functional limitations: 1.

A monoclonal antibody generated against a cell-line susceptible to infection by macrophage-tropic HIV-1 isolates and derived from the HuT 78 T lymphoblastoid cell-line, wherein said antibody inhibits HIV-1 envelope glycoprotein mediated membrane fusion between Hela-envJR-FL and said cell line, but does not inhibit HIV-1 envelope glycoprotein-mediated membrane fusion between Hela-envLAI and Sup-T1 cells or between Hela-envLAI and Hela-CD4⁺ cells.

The Examining Division denied enablement but the Board disagreed: 17.

The examining division denied sufficiency of disclosure for the reason that it would be undue burden to obtain all antibodies falling within the scope of the claims. As explained above, it is the board’s opinion that all necessary information for doing so is contained within the application. Thus, assuming for the sake of discussion that the skilled person would ever want to isolate all of the antibodies falling within the scope of the claims, the possibly undue amount of work involved would not stem from deficiencies in the way the invention was described but rather from the task which he/she chose to accomplish.

(emphasis added)

The rationale of T 1300/05 is entirely in line with that underlying T 292/85 and T 301/87 discussed in section I., supra, using the same or analogous terminology.

T 617/07 confirmed the approach taken in T 1300/05. It was decided by Board 3.3.04 on August 4, 2009. Claim 20 related to antibodies binding the high-affinity tyrosine kinase receptor of the nerve growth factor, which were structurally defined by at least one of the 3 VH CDRs or the 3 VL CDRs and further functional features. The Board held that it was not an undue burden to identify antibodies having the combination of structural and functional features of the claim; see points 22–26, 28, and 32.

T 835/21, a decision rendered by Board 3.3.08 on March 28, 2023, is still consistent with the analysis expressed in the above-cited cases. The claims related to a genus of functionally defined antibodies that were antagonistic and directed to an epitope (characterized by a specific amino acid sequence) of the low-density-lipoprotein receptor-related protein 6 polypeptide (LRP6; also characterized by an amino acid sequence). The Board acknowledged enablement, noting that the provision of antibodies binding to a specific epitope was a routine task: 23.

… the skilled person must be able to prepare the claimed monoclonal antibody. This antibody is essentially defined by two functional features, …

24.

The preparation of a monoclonal antibody that binds to an epitope within a defined amino acid sequence, i.e. a known target, … is a routine task for the skilled person and does not require any inventive activity. …

…

27.

… It is true that the patent does not disclose a reproducible example [apparently there were only a few Fab fragments without any sequence data] and comprises no figures, and that Table III does not refer to an identifier of the tested Fab fragments and therefore cannot be linked to Table II. However, Article 83 EPC does not require an application to contain a reproducible example. …

(emphasis added)

While the opponent argued that it would require a research project to find an antibody as claimed, the Board disagreed in points 34 and 36, concluding that the necessity for routine screening may be tedious but does not amount to an undue burden: 34.

The claim concerns a well-defined class of molecules, namely monoclonal antibodies binding to the propeller 3 domain of LRP6. … the provision of monoclonal antibodies binding to the propeller 3 domain of human LRP6 is routine for the skilled person. Methods for screening these antibodies for inhibition of Wnt3- and Wnt3a-specific signalling activity are also commonly known … . The application and the skilled person’s common general knowledge thus provide sufficient information for the skilled person to produce monoclonal antibodies that specifically bind to an epitope of human LRP6 within the propeller 3 domain and to screen these antibodies for an antagonistic activity against Wnt3 and Wnt3a.

…

36.

… the tools for identifying antibodies having the characteristics recited in the claims are well known and only require routine steps. It may be tedious to screen candidate antibodies binding to the propeller 3 domain of LRP6 for inhibition of Wnt3- and Wnt3a-specific signalling activity, but this does not necessarily amount to an undue burden if the screening results in the desired product, …

(emphasis added)

Thus, it was again decided that time and effort for finding antibodies falling within a claim to an antibody genus were irrelevant as long as only routine experimentation was required.

T 326/22 by Board 3.3.08, issued on August 1, 2024, continued this line of jurisprudence and acknowledged enablement for a fully functionally defined genus of antagonistic anti-human CD47 antibodies: 8.1

The relevant issue is whether or not the skilled person[,] based on the teaching of the patent application taking common general knowledge into account[,] is able to find further antibodies with the claimed functional properties across substantially the whole breadth of the claim in a reliable manner with a reasonable amount of trial and error.

8.2

… as long as the generation of these further CD47 antibodies requires nothing but routine work that may be tedious and time consuming, the method of generating these antibodies cannot be regarded per se as being based on undue burden (T 431/96, Reasons 6).

…

9.3

… the patent application provides the skilled person with the antigen, the epitope, the assays needed for selecting antibodies with the claimed properties and for assessing the antibodies’ binding to the claimed epitope. Moreover, structural information of several exemplary antibodies is provided. [The Board stated in 9.2 that varying these sequences and obtaining further antibodies was routine and would impose no undue burden].

…

9.6

There are … no indications … that the generation of further antibodies … requires specific conditions or circumstances.

(emphasis added)

The above-discussed decisions T 1300/05, T 617/07, T 835/21, and T 326/22 are illustrative examples of the long-standing practice of the Technical Boards accepting functional characterization(s) to define classes of antibodies. Nevertheless, as we shall see below, not all compositions of Boards 3.3.04 and 3.3.08 continue to agree with these principles.

C. Cases where the characterization of V regions by functional features was rejected

T 1466/05 (Board 3.3.08, July 27, 2007) deals with claims directed to a class of fully functionally defined antibodies reactive with pyridinoline in peptide-linked pyridinoline and not free pyridinoline. The Board accepted enablement of a concretely disclosed antibody as claimed since it was deposited in the form of a hybridoma. However, the Board denied enablement of the remainder of the claimed antibodies in the absence of any directions or a suitable protocol for the preparation of further antibodies as defined in the claim. In particular, the Board noted that the application at issue provided no guidance with respect to any specific antigen suitable for raising and selecting antibodies with the desired properties; see T 1466/05 at points 11 to 16.

The situation was similar in T 435/20 (Board 3.3.08, July 14, 2022) where the claimed class of antibodies was directed to a conformational epitope of human IL-23p19. The conformational epitope was characterized in the claim as comprising two short non-contiguous amino acid sequences. The Board criticized the claim at least in that (1) the spatial extent of the epitope was not unambiguously delimited in the claim (point 18 of the reasons), (2) the claimed antibodies did not need to bind exactly the amino acid residues recited in the claim as long as the epitope comprised the amino acid residues (point 19 of the reasons), and (3) the claim encompassed antibodies with the same specificity as the exemplified antibody but which were not structurally related to it (point 22 of the reasons).

It was accepted that peptides consisting of the primary sequence of the claimed conformational epitope were unsuitable for raising or screening for the claimed antibodies (point 30 of the reasons). The Board saw problems in that the patent did not disclose how the exemplified antibody was prepared, i.e., which antigen/immunogen was used for its generation or selection process. Thus, the Board considered that the patent contained no guidance regarding a suitable antigen or screening process for the generation and selection of antibodies that were structurally unrelated to the specific exemplified antibody (point 31 of the reasons).

In view of the above and further considerations, the Board finally concluded that there was a lack of enablement because the functional definition of the claimed antibodies amounted to an invitation to perform a research program without any guarantee of success. Such a situation was considered to constitute an undue burden for the skilled person (point 60 of the reasons).

A similar situation as in T 1466/05 and T 435/20 also arose in T 1345/20 (Board 3.3.08, September 26, 2023). The claims related to an immunoassay for diagnosing Gaucher’s disease. The biomarker to be detected was a small molecule called “free lyso-Gb1.” Neither the patent nor the prior art disclosed how to generate and obtain an anti-free lyso-Gb1 antibody. Thus, the Board stated: 3.

While the provision of antibodies against known targets is usually a matter of routine, as argued by the respondent and supported by case law (e.g. T 431/96), this is only the case if the skilled person knows from the disclosure in the patent or from common general knowledge (i) which antigens are suitable for raising antibodies having the desired properties and (ii) which screening process should be used to select these antibodies without undue burden (see decision T 435/20, reasons 28). …

(emphasis added)

The Board went on to conclude: 4.

The target molecule in the present case … can be considered an unconventional target in the sense of decision T 435/20. There is no indication, either in the prior art or in the patent whether glucosylsphingosine is a suitable antigen, but even if it were, there is still no teaching which screening process would ensure a reliable selection of antibodies specifically detecting only free lyso-Gb1 at low concentration, knowing that there are many other structurally related small molecules, which could even be present at higher concentration in the sample from the subject to be diagnosed. Accordingly, the two criteria of T 435/20 (supra) for enablement of antibodies against an unconventional target are not fulfilled.

(emphasis added)

T 1103/22 by Board 3.3.04, dated June 20, 2024, is another case in line with T 435/20. Here, a second medical use of a completely functionally defined class of antagonistic anti-factor XI antibodies, inhibiting the activity of factor XIa, was at stake. The Board denied enablement, concluding that it was not possible to determine the active site of factor XIa, that there was no concrete information about the amino acids involved or about the structure, that it was common general knowledge that antibodies against some epitopes are not routine to make and can require considerable inventive skill (reference was made to T 435/20), and that no starting point was provided in the form of an actual antibody to generate and compare further antibodies meeting the requirements of the claim.

While these cases deny enablement of classes of functionally defined antibodies, they do so for specific reasons that are based on a lack of disclosure of suitable antigens and suitable screening methods. Such situations amounted to an invitation to perform a research program without any guarantee of success. Consequently, inventive effort was required because applying routine procedures was not sufficient. Notably, none of these cases denies enablement of completely functionally or both structurally and functionally defined classes of antibodies as a matter of principle.

The rationale of T 2164/21 of Board 3.3.08, dated September 9, 2022, however, is entirely irreconcilable with the case law that has been evaluated so far. Claim 1 read: 1.

A method of producing an antibody with reduced susceptibility to deamidation, wherein the method produces a mutant antibody, wherein the method comprises the step of substituting a glycine that is located adjacent to the C-terminal side of an asparagine of the original antibody with another amino acid, wherein the mutant antibody has a reduced susceptibility to deamidation relative to the original antibody before the amino acid substitution, wherein the mutant antibody has the same binding specificity as the original antibody, and wherein the asparagine exists in the complementar[it]y determining region 2 (CDR2) of the heavy chain as determined by Kabat numbering.

(emphasis added)

The Board stated: 39.

The application discloses one way of carrying out the claimed invention (see Example 2), in which one specific amino acid in HCDR2 of an anti-TF antibody, namely glycine at position 55 (Gly55), is substituted with 18 different amino acids and the binding activity of the resultant mutants is measured. …

40.

It is evident from Figure 11 that not all the replacements result in binding activity which is “70% or higher than the antigen-binding activity of the antibody before the amino acid substitution.” In this particular example, the binding activity of 3 out of 18 mutants was significantly decreased compared with the non-substituted antibody. Consequently, the nature of the amino acid replacing the glycine residue in the HCDR2 seems to be critical.

41.

What is at issue is whether the application, when considered alone or in combination with common general knowledge, provides guidance which allows the skilled person to obtain substantially all the embodiments falling within the ambit of the claim without undue burden.

(emphasis added)

The Board denied enablement because it was not willing to accept the necessity of routine screening, meaning that structural predictability would have been required: 46.

In sum, neither the application as filed nor common general knowledge provides the skilled person with any information that would reliably guide them to the amino acid substitutions which result in an antibody fulfilling the functional requirements of the claim.

47.

Accordingly, for each and every antibody comprising an Asn-Gly sequence in HCDR2, in the absence of any guidance as to which amino acid should replace the glycine in any given circumstance, other than to use all possible amino acids and to screen the substituted antibodies for those with the desired characteristics, the skilled person has to identify each time, by trial and error, which amino acid replacement will provide an antibody with the desired characteristics, without any guarantee that such a substitution will be found at all.

48.

The fact that methods for generating an antibody with a specific mutation, methods for assessing deamidation[,] and methods for [testing] antigen-binding activity were described in the application, were well known to the skilled person of the application and were routine on the priority date of the application does not mean that the invention can be put into practice without undue burden.

(emphasis added)

The approach adopted by Board 3.3.08 in T 2164/21 did not include any assessment whether inventive effort was required to find a method resulting in the production of antibodies with reduced susceptibility to deamidation. This would have been hard to deny given that the common general knowledge provided routine means to produce phage display libraries with substitutions by proteinogenic amino acids at a single position of a given antibody at the relevant priority date. In fact, the decision even accepts this as can be taken from its point 48, cited above. Nevertheless, it came to a negative conclusion in the absence of structural predictability. Consequently, the rationale of this decision is indeed entirely irreconcilable with the case law that has been evaluated so far.

A similar expectation of structural predictability as to which variants would have a certain functionality can be taken from the preliminary opinion of Board 3.3.04 issued on June 21, 2022 in T 1511/18. The claim at stake was: 1.

A method for improving the pharmacokinetics of an antibody, said method comprising substituting at least one amino acid of a CDR of the antibody with a histidine, or inserting at least one histidine into a CDR of the antibody, wherein the histidine substitution or insertion increases the KD(pH5.8)/KD(pH7.4) value as compared to the KD(pH5.8)/KD(pH7.4) value before the histidine substitution or insertion, …, and wherein said histidine substitution or insertion

(a)

prolongs the half-life in plasma or the mean plasma retention time of said antibody; or

(b)

increases the number of times of antigen-binding for said antibody, …

(emphasis added)

The Opposition Division did not have any problems with the requirements of Article 83 EPC because of the functional features characterizing the claimed method, solidly relying on the established practice of the historic jurisprudence of the Technical Boards. They stated in point 12.3 of their decision that the burden of proof for showing a lack of enablement fell on the opponent, that serious doubts substantiated by verifiable facts would have been needed, that the examples and the data on file showed that the KD ratio can be increased such that the pharmacokinetic properties of an antibody can be improved without abolishing antibody binding, that it was plausible that the claimed method can be generally applied in the absence of any evidence to the contrary, that no absolute certainty of success was required, that even occasional failure would have been acceptable as long as no inventive effort was required to arrive at the desired result, and that it is irrelevant that the claimed method might be time consuming as long as it can be performed based on the teachings of the patent and routine measures.

However, Board 3.3.04 disagreed in their preliminary opinion, concluding that the claim was not limited by the functional requirements but required a priori knowledge of which modifications would necessarily result in the claimed functional effects. Thus, relying on a structural predictability requirement, the Board arrived at a diametrically opposed preliminary opinion in comparison to the Opposition Division’s decision: 33.

The board currently has doubts whether the skilled person with the disclosure of the patent and in view of common general knowledge was able to carry out the claimed method over its whole breadth without undue burden. In particular, the board is not convinced that antibodies against any target having the properties mentioned in the claim could be obtained by applying the method steps, namely histidine substitution or insertion into the CDR. … Furthermore, since the CDRs define the binding specificity of an antibody and thus differ between different targets or epitopes[,] it appears a priori not possible to define particular sites for histidine substitution or insertion in the CDR which result in the required functional properties in all antibodies … . The patent therefore also provides no teaching which could be generalised from the specific exemplified antibodies to antibodies against other targets or epitopes. The board currently considers the effort to identify relevant positions for histidine substitution/insertion in antibodies for any target to amount to an undue burden.

(emphasis added)

The Board finally revoked the patent on formal grounds relating to added matter and, thus, did not render a final decision on enablement. Nevertheless, their negative attitude was again based on the lack of structural predictability and not on denying that routine procedures would have been sufficient to carry out the claimed method.

Unfortunately, the tendency of the Technical Boards has already been adopted by some Examining Divisions.

In one example, a class of antibodies was structurally defined by alternative VH and VL CDR sequences in combination with defined alternative H-chain constant regions. The class of antibodies was functionally defined by a binding activity. The overall functional limitation of the claim thereby excluded frameworks affecting the binding activity such that it is lower than required. Thus, this situation was different from T 1628/16 discussed in section IV.A., supra. The application demonstrated that the promised technical effect was indeed achieved by recited, specific sets of CDR sequences in combination with recited Fc sequences. Nevertheless, the Examining Division saw a problem under Article 84 EPC (clarity) that was raised as a kind of enablement objection. The functional limitation in the claim was seen as a result to be achieved. Consequently, so the objection, such functional language amounted to a reformulation of the underlying problem without providing the technical features to achieve this effect.

Such an objection effectively means that functional features, even though explicitly allowed in principle by the Guidelines for antibody inventions, might now be seen as a mere (and objectionable) “result to be achieved.” However, basically any functional feature can be characterized as a “result to be achieved” if this position is taken.

Interestingly, the objection was made in the absence of any assertion that the test for verifying whether an antibody having the desired property as claimed required inventive effort. Thus, the objection ultimately amounts to a structural predictability objection, i.e., seemingly requiring the recitation of framework sequences.⁶

In another example, the position was taken that second medical use claims relating to a dosage regimen should be limited to the antibody and the dosage schedule expressly exemplified in the application. A functional generalization of the antibody and a generalization of the actual dosage tested were objected to, although providing alternatives/variations would only have required routine testing but no inventive effort. This amounts to rejecting the principle that antibodies and dosages that are not pharmaceutically effective within the dosage range of the claim cannot be covered because second medical use claims inherently embody and, therefore, require the effect of the treatment; see, e.g., T 1031/06, point 23, T 1107/06, points 54 to 56, T 1859/08, point 13, and T 1394/21, point 3.3.1. In this context, it is interesting to note that T 2347/19, decided by Board 3.3.07 on July 4, 2023, concluded that a dosage regimen claim does not need to specify the doses of a functionally defined antibody if the application as filed provides sufficient orientation for determining these; see T 2347/19 at points 3.2.1 to 3.2.3. Similarly, Board 3.3.01 held in T 263/16 at points 6.4 to 6.5 that dosage determination for a medical use is normally a routine matter and concluded that the failure to disclose any specific effective dosage does not, by itself, suffice to substantiate an enablement objection.

In summary, T 1466/05, T 435/20, T 1345/20, and T 1103/22 denied enablement of classes of antibodies defined functionally or both structurally and functionally but did not deny the enablement of such classes as a matter of principle. In each one of these decisions, there were particular technical hurdles that had to be overcome by measures considered as requiring inventive effort. Consequently, none is incompatible with the long-standing practice of the Technical Boards in which functional features or the combination of structural features with functional features for defining classes of antibodies were accepted in principle.

However, the approach taken in T 2164/21 and T 709/23 reflects an erosion of the previously established principles. These decisions do not see the limit of enablement in whether inventive effort is required to provide an equally suitable antibody. They require predictability of the structure, i.e., an amino acid sequence of the antibody that is equally suitable. Again, this is not in line with the prevailing previous jurisprudence of the biotech Boards⁷; see sections I. and IV.B., supra.

At least some Examining Divisions follow this restrictive approach, as discussed above. Their rationale can be taken as a denial of the possibility of functionally characterizing antibodies in the claims—despite what the Guidelines explicitly permit. The Guidelines do not require structural predictability and do not qualify functional features as a “result to be achieved.”

V. THE TECHNICAL BOARDS’ CASE LAW ON FUNCTIONAL FEATURES IN ANTIBODY CLAIMS: CONSTANT (C) REGIONS

The fragment crystallizable region (Fc region) of antibodies is the tail region of an antibody formed by the C-terminal domains of the heavy chain constant regions. It is capable of interacting with cell surface receptors called Fc receptors and with some proteins of the complement system. It can be modified to control receptor binding, immunogenicity, and antibody recycling: •

Amino acid substitutions or changes in the glycosylation pattern can modulate the binding of the Fc region to, e.g., the Fcγ receptors: this can influence cytolytic activity, antibody-dependent cellular cytotoxicity (ADCC) activity, and/or complement-dependent cytotoxicity (CDC) activity, as well as modulate antibody immunogenicty.

•

Amino acid substitutions can also modulate the binding of the Fc region to the FcRn receptor at the physiological pH of 7.4 versus its binding at the endosomal pH of 5.5: this influences the pharmacokinetics of the antibody by influencing its recycling and, thereby, its half-life in blood.

A claim to an antibody having an Fc modification was addressed in T 1856/16, decided by Board 3.3.04 on August 24, 2020, where claim 1 related to:

An antibody comprising a human IgG1 variant Fc region,

wherein said variant Fc region comprises at least one amino acid modification relative to a wild-type Fc region, such that said antibody specifically binds FcγRIIIA with a greater affinity than a comparable antibody comprising the wild-type Fc region;

and wherein said at least one amino acid modification comprises a substitution at position 396 with leucine, wherein said numbering is according to the EU index as in Kabat.

(the effect of the greater affinity to FcγRIIIA is enhanced ADCC).

While the claim required one particular amino acid modification, the remaining part of the Fc region was not further limited, except by the functional feature of specifically binding FcγRIIIA with greater affinity. It is evident from the decision that the Fc regions had the substitutions at position 396 in combination with other substitutions. Even though mutating and testing with an unpredictable result was required, the Board acknowledged enablement: 33.

The application discloses how to produce antibodies as claimed (see pages 75 to 78). Furthermore, the application provides experimental evidence of increased binding of Fc regions carrying the P396L substitution to the FcγRIIIA receptor. This includes single substitutions and combinations with other substitutions (see Table 5 on pages 65 ff). The application also discloses a number of known cancer antigens and therapeutic antibodies targeted to those (see paragraph [0017] and Table 6 on pages 119 ff).

34.

The board therefore finds that the disclosure in the application not only teaches the skilled person how to produce antibodies as claimed but also discloses tests by which the functional characteristics as required by the claim can be ascertained.

(emphasis added)

It can be concluded from this decision that a need for screening was considered to be acceptable, and that the predictability of which specific mutation would achieve the desired effect was unnecessary. This is consistent with the approach taken by the Technical Boards in the above-discussed decisions T 1466/05, T 435/20, T 1345/20, and T 1103/22.

VI. CONCLUSIONS

A. The UPC and functional features in antibody claims

According to the UPC Court of Appeal decision Amgen v. Sanofi-Aventis and Regeneron Pharmaceuticals , loc. cit., claiming a genus of functionally defined antibodies is not a problem if there is a technical concept allowing the practice of the claimed invention. No instruction is required for how to provide each and every conceivable embodiment. The necessity of screening for equally useful variants or alternative antibodies is not an issue, even if such screening is tedious and tricky, as long as the skilled person would not have serious doubts that further antibodies as claimed can be obtained without inventive effort. A reasonable amount of trial and error as well as occasional failures are acceptable, whereas predictability of the structure of the functional variant or alternative is unnecessary; see section II., supra. This is consistent with the historical approach of the EPO’s Technical Boards of Appeal dealing with biotech cases.

B. The EPO and its Technical Boards of Appeal

The approach taken in some more recent cases diverges from what the Technical Boards have historically decided in the biotech field regarding the acceptance of functional features and the necessity of screening. A particularly striking example is T 2164/21, where Board 3.3.08 denied enablement of a method of producing a structurally and functionally defined class of antibodies. The only effort that was required to develop the desired, modified antibodies was the substitution of a single, defined amino acid position in CDR2 of the heavy chain variable region. This could have been done by a routine exercise at the time of the priority date, as the Board even admitted in point 48 of their decision. Nevertheless, they denied enablement because there was no predictability as to which specific amino acid at the defined position would be functional; see section IV.C., supra. Consequently, whether the technical contribution underlying a functionally defined claim allows providing alternative embodiments encompassed by the corresponding functional definition by routine screening procedures without inventive effort may become irrelevant in the eroding case law and may be replaced by a requirement of structural predictability.

Of concern are also cases in substantive examination in the first instance in the EPO where functional features are no longer accepted, even though they are specifically foreseen as an option for claim drafting in the Guidelines for Examination applicable to antibody inventions. We increasingly encounter such objections.

It is also interesting to see that some Examining Divisions now make objections to functionally defined classes of antibodies as an unacceptable case of defining a mere “result to be achieved” while, at the same time, inventive step objections are raised by Examining Divisions based on the notion that making an antibody as claimed would not require any inventive skill. Does this mean that structural predictability is required for enablement and, when consistently applying this rationale, that a lack of structural predictability should be an indicator for inventive step?

Some authors have expressed their sympathy for the introduction of structural non-obviousness as a basis for judging the patentability of antibody inventions in the EPO.

T. Bucher, loc. cit., derives the basis for her proposal in this direction from how the CLBA, 10^th edition, 2022, section I.D.9.9.2 deals with the patenting of small molecule chemical compounds. The article concludes from this practice that a structural approach should also be used in the assessment of patentability of antibodies. There would be a lack of predictability in finding antibodies with meaningful affinity and biological activity. However, first of all, it has to be borne in mind that patents for low molecular weight chemical compounds frequently contain general formulae which cover a whole host of derivatives (Markush type of claims), thus building a fortification wall around the most important compound that was invented. Why should this not be allowed for biological agents? After all, allowing a reasonable degree of generalization, including by functional definitions, has been acknowledged as a prerequisite for meaningful patent protection; see, e.g., T 484/92, point 3.2, and the decisions presented in section I., supra. Moreover, importantly, the rationale of patenting small molecules can technically not be compared with the rationale that should be applied to the patenting of antibodies. As opposed to the former, the latter are molecules with a modular structure that can normally be routinely made against any given antigen. Thus, the Guidelines and the case law of the Boards took and still predominantly take the position that structural non-obviousness is not enough but that a surprising technical effect or the absence of a reasonable expectation of success in obtaining antibodies having the functional properties is required. This is a principle that has generally been applied in the field of biotechnology; see, e.g., T 111/00, dealing with inventive step of a nucleic acid and denying inventive step for the cloning of a human orthologue of the known mouse gene.

B. Carion-Taravella/C. Ribard, loc. cit., held the same opinion given the need for patents on pharmaceutically important antibody drugs in the interest of patients. What we agree with is that it is urgent to find the right balance. However, the issues that need to be balanced are inventive step on the one hand and enablement on the other. This is what the present article also is about. Principles applied in the assessment of one of these patentability requirements have to be applied in a congruent fashion in the assessment of the other.

T. Bucher, loc. cit., suggests that the structural non-obviousness approach of the U.S. Patent and Trademark Office (USPTO) should be adopted because it would have been applied for many years without controversy. However, we question the “without controversy” and the suggested analogy is not directly applicable in the EPO. Enablement and obviousness are not intertwined pursuant to U.S. law and practice in the same fashion as are sufficiency and inventive step in the EPO. Thus, applying the concept of structural non-obviousness does not directly lead to the conclusion in United States practice that a class of functionally identified antibodies cannot be enabled because providing each covered functionally characterized molecule would be a surprise. Rather, the law practically limiting the scope of such antibody claims before the USPTO has been the additional requirement of 35 U.S.C. §112(a) that the specification “contain a written description of the invention.” This requirement has no parallel in the EPO. However, what has made the United States situation more complicated recently was that apparently enablement of a functionally defined class of antibodies was denied in the Amgen vs. Sanofi case; see Amgen Inc. vs. Sanofi, 598 U.S. 594, 143 S.Ct. 1243 (2023) (which was the case parallel to the Amgen v. Sanofi-Aventis and Regeneron Pharmaceuticals case decided by the UPC CoA, discussed in section II., supra).

The previous articles by T. Bucher and by B. Carion-Taravella/C. Ribard, loc. cit., apparently consider that the EPO can take a leading role in controlling whether there may be patents and dependent patents by adjusting allowable scopes of protection. We do not agree with proposing such a policy. Suffice it to mention that compound patents normally dominate any second medical use patents. This does not amount to unfair protection, and the parties concerned have been able to deal with such situations. Many second medical uses have been developed nevertheless. What is the task of the EPO, in our opinion, is to consistently apply the EPC and the existing case law so as to create legal certainty about what can be patented under certain circumstances and what cannot be patented. With respect to such considerations, T 1213/05 referred to G 1/98 at point 53 and stated that the EPO has not been vested with the task of taking into account the economic effects of the grant of patents in specific areas and restricting the field of patentable subject-matter accordingly.

C. Future perspective

If the Technical Boards of the EPO dealing with biotech inventions would indeed begin to generally require that claims can only encompass specifically exemplified embodiments or embodiments that are structurally predictable based on the disclosure of the patent/application, then the logical consequence should be that “structural non-obviousness” becomes an effective argument for establishing inventive step, resembling the practice of the USPTO. Otherwise, innovators would be punished twice. There is no rational justification for adopting a strict written description/enablement doctrine requiring structural predictability as it is applied by the USPTO but simultaneously refusing to adopt the USPTO’s structural non-obviousness doctrine. If the stakeholders in the area of patenting antibodies, i.e., users of the system and the authorities, would wish this to be the situation, any adjustment should be discussed among them and should be done consciously and transparently. Inconsistent and therefore unpredictable case law cannot be a reasonable basis for adjusting this important field of medical innovation. Here we agree with T. Bucher and B. Carion-Taravella/C. Ribard in principle. Nevertheless, a purely structure-based inventive step approach in the EPO without requiring any unexpected technical effect bears the danger of increasing the number of “me-too patents” which potentially devaluate the contribution made by the original innovator; see also M. Kawczynska/J. Meier, loc. cit.

Furthermore, the driving force for denying enablement of fully functionally or both structurally and functionally defined classes of antibodies and to rather insist on solely claiming fully structurally defined species should not be convenience and efficiency for patent offices, i.e., that the latter type of claims can be examined more easily and more quickly and that such narrow claims reduce the likelihood of oppositions that could be of limited profitability. The focus should rather be on fair—and balanced—protection, as has been emphasized by the Technical Boards already in 1988 and by the UPC CoA. The UPC CoA’s case law is of profound relevance because they do not only deal with questions of validity but also decide on questions of infringement, unlike the EPO.

The emerging new trend to increasingly consider functional features as an indication of a mere “result to be achieved,” rather than as an attempt to define a fair scope of protection, is exacerbated for innovators by the devastating recent statistics of the Technical Boards dealing with biotech cases. From 2000 to 2019, Board 3.3.04 apparently had a revocation rate of patents in opposition appeal cases of 40.5%.⁸ In the interval between 2020 and early February 2026, Board 3.3.04 increased this rate of revoking patents in opposition appeals to 67.9%, corresponding to an increase by about two-thirds (67.6%). Board 3.3.08 has apparently shown a similar increase, having a revocation rate of 39.0% between 2000 and 2019 that rose to 64.6% for the period from 2020 to early February 2026, an increase of 65.6%. When the statistics of the two biotech Boards in the interval from 2020 to February 2026 are analyzed for specific compositions of these Boards, the revocation rate in opposition appeals apparently even rises beyond 75% for Board 3.3.08. In contrast, in the same time intervals, the revocation rate in opposition appeals by the chemistry Boards even slightly decreased.

All considered, the new restrictive approach to functional features in antibody cases observed at the EPO, combined with the recent dramatic increase in the revocation rate of the biotech Boards in opposition appeal cases, certainly gives applicants and patent proprietors reason for concern.

At the same time, it is remarkable that the UPC Court of Appeal has chosen to adopt the historic case law of the EPO’s Boards of Appeal allowing functional features for antibody inventions rather than the EPO’s more recent restrictive approach. Looking ahead, it remains to be seen how the interaction between UPC jurisprudence and EPO case law will shape future patent practice to reasonably reward innovation concerning antibodies. Quo vadis iurisdictio?

Footnotes

Acknowledgment

The authors are grateful to Dr. Jürgen Meier for reviewing their article.