Kymab is famous as the Cambridge antibody mouse company that fought Regeneron all the way to the UK Supreme Court, and won, on sufficiency (no mean feat, given the very strict approach the UK courts have taken to the sufficiency test for pharmaceutical inventions in recent years). In T 0527/24, Kymab was on the other side, arguing as Opponent. In an argument reminiscent of the Kymab case, this Board of Appeal decision focused on how prophetic examples should be viewed in inventive step analysis. If the prior art just says it was going to make something, how much weight should be given to the disclosure?
Legal background: Starting points, suggestions and the reasonable expectation of success
A prophetic or hypothetical patent document may describe work that has not actually been done. US practice has a formal doctrine around prophetic examples, but EPO practice does not. In Europe, a prior art document is assessed for what it discloses and what it would have suggested to the skilled person, whether or not the authors ever got around to doing the experiment. The question in T 0527/24 was how much weight a Patentee can place on the argument that the prior art suggestion was “merely prophetic”.
Case background: The common light chain mouse
The patent in question in this appeal was EP 3456190 (Merus N.V.), relating to a transgenic mouse for producing antibodies that share a common light chain. The variable domain of an antibody light chain is assembled in developing B cells by recombination of one V and one J gene segment. Normal antibody diversity comes from both chains. The idea behind a common light chain platform is to fix the light chain, so that all of the diversity, and all of the specificity, comes from the heavy chain. Antibodies that share a light chain can then be readily combined into bispecific and multispecific formats.
The common light chain mouse is the front end of Merus’s antibody discovery engine, marketed as MeMo, and the bispecifics it feeds are the company’s Biclonics platform. Merus has long offered MeMo to other biopharmaceutical companies under non-exclusive licence, so the patents around the mouse are revenue-generating in their own right as well as protecting Merus’s own pipeline. That pipeline has a marketed product, zenocutuzumab (Bizengri), a HER2 x HER3 bispecific approved by the FDA in December 2024 for NRG1 fusion-positive pancreatic adenocarcinoma and non-small cell lung cancer, and the first approved medicine to come out of the Biclonics platform.
The estate has been fought over before, and until now Merus had the better of it. The Opposition Division upheld Merus’s earlier common light chain mouse patent, EP 2147594, without amendment in 2016, dismissing novelty and inventive step attacks brought by Regeneron. At the time, Merus said that the patent estate gave it protection through 2029. The patents are enforced as well as licensed. Merus has sued Xencor in Delaware, alleging that its common light chain patent was infringed in the course of work on antibodies derived from Xencor’s own RenLite mouse.
Claim 1 as granted was directed to a transgenic murine host comprising, integrated in its germline, “a rearranged nucleic acid comprising one human germline V and one human germline J gene segment encoding a rearranged human VL region“, wherein the light chain variable region “is capable of pairing with at least two different heavy chains encoded by the host, such that variety in specificity of antibodies comprising said human VL region is retained through rearrangements and hypermutations in the heavy chains“.
Kymab opposed the patent on all grounds. The OD rejected the opposition. On inventive step, the OD had accepted the Patentee’s argument that the prior art, a US patent application from the 1990s disclosing transgenic mouse platforms for human antibodies, followed a “more is more” philosophy of maximising light chain diversity, and that the skilled person would have had no reasonable expectation that the light chains it described would pair with different heavy chains. Kymab appealed the decision.
Does it matter that the prior art never made the mouse?
The key question for inventive step was how much weight should be given to the prior art prophetic example. The prior art described the PCR amplification of rearranged human kappa light chain genes from human spleen DNA, cloning of the 0.7 kb product and sequencing of five clones, two of which had functional VJ joints. The document then continued that “Transgenic mice generated with the rearranged light chain constructs can be bred with heavy chain minilocus transgenics to produce a strain of mice that express a spectrum of fully human antibodies in which all of the diversity of the primary repertoire is contributed by the heavy chain” (emphasis added).
The Patentee’s case was that this was all prophetic. The document did not disclose a rearranged nucleic acid at all, the figure showed a construct with five J segments, the experiments had never been put into practice, and, in any event, the skilled person reading a document with 1990s priority dates would have doubted that any of it had been achieved.
However, the Board of Appeal noted that the relevant date is always the priority date of the patent, not that of the prior art: “The skilled person reading [the prior art] has all the common general knowledge available at that date” (r. 29). Additionally, for the Board of Appeal, the Patentee’s argument “could be understood as suggesting that the disclosure in [the prior art] is not sufficiently disclosed, in analogy with Article 83 EPC. However, no relevant evidence or technical argument was submitted in support of such a position” (r. 29). The Board of Appeal did concede that the mice themselves were only suggested, the document stating that they “can be bred” rather than that they were bred (r. 31). However, for the Board of Appeal, a suggestion was all that was needed. This is in line with the rather low sufficiency bar at the EPO. It is often lower than you might think.
Asserting a prejudice is not the same as proving one
The OD’s decision to uphold the patent had turned on the functional pairing feature of the claim, namely the requirement that the rearranged VL region be “capable of pairing with at least two different heavy chains encoded by the host”. The prior art acknowledged that “not all light chains will be equivalent with respect to their ability to combine with a variety of different heavy chains” and suggested a programme of generating and testing different strains of mice. The OD read this as meaning that there was no reasonable expectation that any given rearranged light chain would pair with different heavy chains.
The Board of Appeal’s view differed on this point. For the Board of Appeal, the passage merely offered a second mechanism for diversity and a screening suggestion and did not imply that the previously disclosed light chain could not pair. More fundamentally, “The board [had] not seen any evidence to suggest that the known difference in pairing promiscuity of VL regions means that VL regions exist that can only pair with a single VH region” (r. 35). For the Board of Appeal, the OD’s conclusion went “against what is explicitly taught in [the prior art] without a technical explanation based on common general knowledge or a document from the state of the art” (r. 38).
The technical difference over the prior art was therefore found to be that the prior art did not disclose the mouse itself, only nucleic acids suggested as suitable for making one (r. 40). The technical effect was thus characterised as “to actually put into practice what is suggested” (r. 41), and the objective technical problem was formulated accordingly as “putting into practice the suggestion in paragraphs 481 and 482 of [the prior art] to produce immunoglobulins having a common light chain” (r. 42).
The Board of Appeal concluded that a skilled person reading the suggestion to breed the mice was motivated to do so, and “had a reasonable expectation of success because no evidence has been put forward that the skilled person would have had doubts about the feasibility of carrying out the suggestions in [the prior art] at the relevant date” (r. 45, emphasis added).
Analysis
In this case, the Board of Appeal did not treat the prior art as having actually disclosed the relevant mouse. It treated the prior art as suggesting that the skilled person make the mouse and asked whether anything would have prevented or dissuaded the skilled person from doing so. The decision is in line with the general EPO approach that antibody generation is generally routine, and that a prior art statement that a strategy is “promising” is enough for a reasonable expectation of success.
Prophetic and hypothetical disclosures are potentially about to become even more common and problematic. Generative AI can produce a plausible paper example, a prospective protocol or a whole speculative specification in seconds, and both patent applications and defensive research disclosures may already be being drafted that way. T 0527/24 is a reminder that the EPO may very well take such suggestions at face value unless there is evidence that the skilled person could not have carried them out. A machine-generated suggestion published today may be considered perfectly good prior art against an actualised, evidenced and supported later invention. However, there is also a difference between the evidence standard for prior art successfully citable against a later invention, and the evidence needed to support the invention itself for patentability. These standards are not the same. An invention must have evidence that supports both sufficiency and inventive step (i.e. it solves the technical problem to be solved over the whole scope of the invention). To be cited successfully in an inventive step attack, the prior art just needs to make a suggestion a skilled person can act upon. Things are not going to get any easier for Patentees within this framework.
Author: Rose Hughes
Rose is a biotech and pharmaceutical patent specialist with over a decade of experience in intellectual property. Rose is a Patent Director at Evolve, where she leverages our unique fractional in-house model to provide clients with deep patent law expertise combined with the strategic commercial oversight typically associated with senior in-house counsel.
With a PhD in Immunology from UCL, Rose applies her technical background to complex innovations in biologics, cell and gene therapies, and the rapidly emerging field of AI-assisted drug development. Previously, Rose held the role of Director, Patents at AstraZeneca, where she was responsible for global IP portfolios and IP strategy at every stage of the pharmaceutical pipeline, from platform development and on-market commercialisation to SPCs and patent term extensions.
A recognised thought leader in the field, Rose has been a regular contributor to IPKat since 2018, offering practical insights into European patent law developments. She is also a frequent speaker on the epi podcast, a guest lecturer for the Brunel University IP law Postgrad Certificate, and a contributing author to published books A User’s Guide to Intellectual Property in Life Sciences (2021) and Developments and Directions in Intellectual Property Law (2023).
