A review of topical EPO Case Law

Plausibility and Post-Published Data

The concept of plausibility continues to be a hotly contested subject at the EPO, with the pending referral to the Enlarged ​Board of Appeal in G2/21 expected to be decided in early 2023. According to European case law, for a claim to meet the ​requirements of inventive step, the patent application as filed must make it plausible that the problem purportedly solved by ​the invention, is indeed solved across the entire scope of the claim. In the 2005 Board of Appeal decision T1329/04, in which ​the concept of plausibility has its origins, the Board held that data published after the filing date of the patent application ​(“post-published” or “post-filed” data) which confirms that the problem has been solved can be used in evidence, but only if ​the initial plausibility threshold is met based on the disclosure of the application. In other words, post-published evidence can ​be used to support plausibility, but cannot be used to remedy deficiencies that existed in the application from the outset. As ​such, the concept of plausibility is intertwined with the admissibility of post-published evidence to support inventive step.

The concept of whether a patent application makes the invention plausible also has some overlap with the requirement of ​sufficiency of disclosure. For an invention to be sufficiently disclosed, the application must contain sufficient information to ​allow the person skilled in the art, using common general knowledge, to perform the invention over the whole area claimed ​without undue burden and without needing inventive skill.An objection of lack of sufficient disclosure presupposes that there ​are “serious doubts, substantiated by verifiable facts”, that the skilled person would be able to perform the invention. If ​inadequate information relating to the invention is provided, in addition to causing a lack of sufficiency, this may also lead to a ​challenge to inventive step on the basis that the claim does not plausibly solve the problem across its entire scope.

But when does an invention have enough data to support a patent application, and when can post-filed data come into play? ​These issues have been considered by the Boards of Appeal in recent decisions.

The EPO’s approach to “sufficiency of disclosure” and ​“plausibility” was considered in decision T 2015/20, which ​highlights that the burden of proof for a lack of sufficiency ​objection lies with the Examining Division.

The application in question claimed a composition for ​providing a specific dose of aclidinium bromide for use by ​inhalation for the treatment of asthma. The application ​was initially refused in examination on the grounds of ​insufficient disclosure. The experimental evidence in the ​application (and a corresponding later-published scientific ​paper) only demonstrated the claimed treatment to be ​effective in the treatment of chronic obstructive pulmonary ​disease (“COPD”). The Examining Division of the EPO ​argued the available data relating to COPD did not make ​treatment of asthma plausible, in view of the different ​etiologies of the two conditions. The Examining Division ​also referred to a product insert for a medicinal product ​containing aclinidium bromide for the treatment of COPD, ​which warned that it was not to be used for treating ​asthma, as relevant clinical studies had not been ​conducted.

However, the Board of Appeal took a different view and ​overturned the Examining Division’s decision to refuse the ​application. The Board of Appeal referred to the ​established principle that “a convincing objection of lack ​of sufficiency of disclosure presupposes that there are ​serious doubts, substantiated by verifiable facts, on the ​possibility for a skilled person to carry out the invention as ​claimed”, and where no serious doubts are substantiated, ​sufficiency of disclosure must be acknowledged.

T2015/20 - the requirement for “serious doubt”

As such, T2015/20 sets a high bar for a lack of sufficiency ​objection to be successful. According to the Board, the ​product insert cannot be taken as evidence of “serious ​doubt” since it simply warns that no official authorisation ​for use in asthma had been granted, and that no clinical ​studies were conducted. The proposition that COPD is a ​distinct disease from asthma was also reconsidered. The ​Board noted that, in fact, the two diseases did share ​some common features, albeit that the prominence of ​certain pathways in asthma was reduced compared to ​the prominence of those pathways in COPD (and hence ​the efficacy of classes of drugs that target those ​pathways would also be affected). In spite of differences ​in the biology underpinning these diseases, it could not be ​concluded that aclidinium bromide would not be effective ​in treatment of asthma; at worst the benefits of treatment ​may be less. Hence, no “serious doubts, substantiated by ​verifiable facts” as to the claimed invention existed.

In T2015/20, the Board further noted a number of other ​cases regarding sufficiency of disclosure in second ​medical use inventions, and in particular noted that a ​simple assertion of therapeutic use is not sufficient; some ​further evidence is always needed, even if this evidence is ​not experimental data for the specific condition (referring ​to T 609/02). The limitation to a specific dose also helped ​the applicant’s case, with the Board noting that this ​dosage information would be considered a specific ​technical contribution, beyond simply a vague statement ​of effectiveness.

T0391/18 – How much data is enough data?

The question of whether the scope of a claim is supported by the data provided in the application as filed is assessed on a ​case by case basis. It can therefore be difficult to know in advance how much data needs to be included in the application as ​filed to ensure a claim is adequately supported. These concerns arise where the invention purports to achieve a particular ​technical effect, but the data available in the application only provides specific examples to support the claims. It is of course ​impractical for an applicant to provide experimental data for every single possible embodiment covered by the scope of a ​claim, but care should be taken to include sufficient data to credibly demonstrate the purported technical effect. There is a ​balance to be struck between the scope of the claim, and the supporting data. It is therefore imperative to ensure that the ​application as filed provides sufficient data, to avoid any scenario where post‑published data is relied upon.

These issues were considered in T0391/18, which related to a European Patent directed to a combination therapy for use in ​the treatment of HIV infection. The combination comprised a particular compound (TMC278) and a therapeutically effective ​nucleoside or nucleotide reverse transcriptase inhibitor to be administered once daily. The patent was opposed, and the ​opponent argued that it was not plausible that all combinations covered by the claim were suitable for the stated purpose of ​reduced side effects. The opponent argued the patent provided no data for efficacy or safety of the claimed combinations, that ​only pharmacokinetic data relating to a particular isomer of TMC278 was included, and that combinations covered by the claim ​scope may give rise to antagonistic interactions. The Board considered it plausible ab initio from the knowledge of the skilled ​person that such a combination would be effective in the treatment of HIV, but the patent itself did not demonstrate the ​claimed technical effect. The proprietor filed post-published clinical data to support the claim scope, but as the clinical data ​only related to specific combinations of TMC278 and reverse transcriptase inhibitors, the Board was not persuaded that the ​invention has been plausibly demonstrated across the entire scope of the general medical use claim.

This case highlights how deficiencies in data provided in the application as filed can cause serious issues which cannot later be ​remedied by post-filed clinical data. Given the high experimental burden on the applicants working the pharmaceutical field, ​these issues are perhaps unsurprisingly prevalent in medical use cases such as the one discussed above. It is often the case ​that application as filed may only exemplify particular substances or compositions in pre-clinical studies, and the cell models ​used in the pre‑clinical studies may not adequately represent the clinical reality. The cell models may even be based on ​particular cell lines associated with particular disease characteristics, which do not fully represent the intended medical ​indication to be treated in accordance with the claims.

In such instances, referring post-filed data may be useful. However, even if clinical trials have commenced, post-published data ​derived from those trials often relates to treating specific medical indications which patients recruited to the study present with, ​and so may not be useful for supporting a more general claim to a broader disease category. Similarly, as demonstrated by ​T0391/18, the clinical trial may use specific dosage regimens and specific substances or compositions. Relying on developing ​additional data after filing an application can be a risky strategy, as the specificity of the clinical trial data may mean its ​teachings are too narrow to support the entire scope of a claim.

T0116/18 – Plausible or not implausible?

As outlined in the cases discussed above, where doubt is cast on whether a claimed invention will achieve its claimed ​technical effect across its entire scope, the disclosure provided in the application needs to make that technical effect ​plausible. However, some existing EPO case law, such as T0578/06, suggests that plausibility is only relevant when the ​particular case supports doubts over whether the claimed invention is solved. In other words, the claimed solution must not ​be implausible.

In the 2021 decision T0116/18, the Enlarged Board of Appeal considered this difference in approach. The patent in question, ​belonging to Sumitomo Chemical Company Ltd, is directed to combinations of insecticides which are said to be synergistic, ​with claim 1 referring to a combination of thiamethoxam and one or more compounds represented by a general formula. The ​patent itself includes experimental data showing synergy, but only for two examples covered by the claims, against two moth ​species. During opposition, the opponent submitted data intended to show that a combination within the scope of claim 1 did ​not show synergy against one of these species or against a closely-related species. The patentee criticised this data, and ​submitted their own post-published data showing that the combination was indeed synergistic against various species. The ​patent survived opposition, with the Opposition Division of the EPO accepting that this data confirmed the presence of ​synergy, and hence an inventive step.

On appeal to the Technical Board of Appeal, the question of plausibility and post-published evidence was again central to the ​case. The Board considered, firstly, whether the patentee’s post-published evidence was relevant for the assessment of ​inventive step. The Board took the view that, considering only the prior art, the two examples in the patent, and the ​opponent’s evidence that some combinations are not synergistic, the objective technical problem to be solved would be ​formulated as “providing an alternative insecticide composition”, and the claims as granted would not be considered inventive. ​When the patentee’s evidence was considered, however, a more specific objective technical problem was formulated – ​namely provision of an insecticide composition which acts synergistically against a particular insect species. The Board ​considered that this problem was solved by the patent, and so the claims would be considered inventive. Hence, the question ​of whether this post-published evidence can be taken into account would be critical for deciding on inventive step.

In view of the conflicting “plausibility” versus “not-implausible” approaches both found in the case law, the Board considered ​that it is necessary for the Enlarged Board to clarify the correct approach. As the Technical Board of Appeal noted, the two ​tests can result in very different outcomes; particularly so when the technical problem, as was the case here, is reformulated in ​light of the prior art. The following questions have therefore been referred to the Enlarged Board:

• Must post-published evidence be disregarded if an inventive step is acknowledged on the basis of a technical effect and ​the proof of the effect rests exclusively on such post-published evidence?
  
• If the answer to the first question is yes, can the post-published evidence nevertheless be taken into consideration if the ​skilled person at the filing date of the patent application would have considered the effect plausible or, at least, not ​implausible?
  

A hearing was held before the Enlarged Board towards the end of 2022, and we expect the written decision to issue in early ​2023. The Enlarged Board’s preliminary opinion, issued before the hearing, suggests that the technical effect relied upon ​“needs to be encompassed” by the technical teaching of the patent application, and “embody the same invention”. That is, ​the patentee may not rely on a completely unrelated technical effect from that of the original disclosure. Once this threshold is ​met, then, the Board suggests, the question is whether “the skilled person … would have had any significant reason to doubt” ​the existence of that technical effect at the time of filing the application.

This is a clear endorsement of the “not implausible” test, rather than the somewhat stricter “plausible” test. Although the ​“same invention” concept is new in this specific area, it is generally in line with the EPO’s established approach to ​reformulation of the technical problem as new prior art comes to light. Of course, this is not the final opinion, and the Board’s ​view may well have changed in the interim. We will provide an update once the decision is issued.

T2466/17 – Using post-filed evidence to demonstrate ‘serious doubts’

T2015/20, T0116/18 and T0391/18 discussed above ​focussed on the use of post-published data to support ​the plausibility of an invention over the entire scope of the ​claim. However, post-published data may equally be used ​to attack a claim after grant in opposition proceedings. In ​T2456/17, following opposition proceedings, an ​interlocutory decision to maintain the patent in amended ​form was issued, and both the opponent (Syngenta Crop ​Protection AG) and proprietor (Monsanto Technology, ​LLC) filed appeals against the decision. The patent related ​to a transgenic plant having more than one transgene for ​reducing or eliminating a plant insect pest (Western corn ​rootworm) infestation, comprising a dsRNA for ​suppression of an essential gene of the pest, and a gene ​encoding an insecticidal protein. The patent suggested ​there was a synergy between the dsRNA and the ​insecticidal protein, in the sense that the protein made it ​easier for the dsRNA to enter the cells of the insect pest.

The opponent argued that the application lacked sufficient ​disclosure, in part because the disclosure of suitable ​dsRNA lengths (19-500 nucleotides) for use in ​accordance with the invention was in direct contradiction ​to the teachings of a post-published document, which ​suggested at least 60 nucleotides was required.

However, the Board of Appeal did not find these ​arguments to be persuasive, pointing to the skilled ​person’s understanding that, for a dsRNA to suppress a ​gene, it is necessary for the dsRNA to hybridise to that ​gene, and the length and sequence homology of the ​dsRNA must be so as to allow such hybridisation. As ​such, the skilled person would understand how to design ​suitable dsRNA, including selection of a suitable length. ​The opponent also attacked the sufficiency of the claim ​on the basis that a document, published after the ​publication date of the application, indicated that not all ​dsRNAs targeted to different genes had an effect on ​western corn rootworm, and indicated that less than 50% ​of dsRNAs targeted to different genes investigated had a ​significant effect on larval mortality. However, the Board ​did not find this argument to be compelling as the majority ​of dsRNAs investigated did show some activity, and so it ​would not have been an undue burden for the skilled ​person to identify a suitable gene to be suppressed.

T2456/17 provides insights on how an opponent may use ​post-published data to attack the sufficiency of a claim ​after grant. In this case, the data pointed to by the ​opponent was not found to be compelling, seemingly ​because the majority of the experiments performed ​supported the claim. If only a minority of dsRNAs were ​shown to have the desired activity, then the Board may ​well have arrived at a different decision.

T1742/15 - Synergy

Synergy between compounds is frequently encountered in patents relating to treatments, and evidence of synergy can ​provide compelling basis for inventive step. T0116/18, discussed above, considered whether the claimed combinations of ​insecticides provided a synergistic effect, and whether the claim is considered to involve an inventive step hangs in the ​balance depending on the Enlarged Board of Appeal’s decision on their interpretation of post-published data. Similarly, ​T2456/17, also discussed above, described synergy between dsRNAs and the expression of an insecticidal protein. The ​patent in question provided mechanistic explanations for the synergy observed (specifically that the insecticidal proteins create ​entry pores through which the dsRNA molecules are able to penetrate into the cells), and the Board of Appeal saw no reason ​to question those explanations or their plausibility.

In T1742/15 the purported synergistic effect related to a mixture comprising two compounds, mesotrione and s-metolachlor, ​for controlling the growth of crabgrass or whiteclover. Synergistic effects of the claimed combinations were demonstrated in ​the examples, but the experimental data provided in annexes to the application revealed that, in certain experiments, synergy ​was not always observed. As such, when analysing inventive step, the Board did not consider synergy as part of the technical ​effect of claim 1, as the technical effect was not achieved across the entire scope of the claim. The Board ultimately arrived at ​the decision the claims lacked an inventive step, and the patent was revoked. This decision serves as a reminder of the ​importance of providing adequate supporting data and, when relying on synergy as a technical effect, ensuring that the scope ​of the claims extends only to those combinations where synergy is in fact achieved.

T0420/19 – Using modern methods to assist the skilled person in carrying out the invention

Increasingly, high-throughput methods are being used to screen large databases, ranging from compound databases to cell-​based assays. T0420/19 suggests that the availability of high-throughput methods may provide some support in overcoming ​sufficiency objections where large numbers of examples require screening. In T0420/19, during opposition proceedings, the ​Opposition Division had, on their own initiative, introduced a new ground of opposition under Article 100(b) EPC (sufficiency). ​The claims were directed to a barley plant carrying a mutation in the gene encoding methionine-S-methyltransferase (MMT) ​that causes a total loss of MMT function. Use of this barley in beer production was said to afford improved taste.

The Opposition Division considered that the only method given in the patent to obtain the claimed mutants was mutagenesis ​and subsequent screening, thus relying on identification of a chance event. In view of the number of plants that would have to ​be screened to arrive at the claimed invention, the Opposition Division believed undue burden was placed on the skilled ​person. The Board of Appeal disagreed, on the basis the skilled person would be able to mutagenise barley seed kernels and ​subsequently screen the resulting mutants using the high-throughput screening assays described in the application as filed. ​The Board found that the mere fact that mutagenesis is a random process and that large numbers of mutant barley plants ​might have to be screened did not, in this case, represent an undue burden on the skilled person.

T0580/19 – Conflicting data and support for claimed technical effects

Caution should be exercised as to whether the data ​provided in the application actually supports the claimed ​technical effect. If a significant proportion of the examples ​provided do not achieve the technical effect, then the ​patent application may run into trouble. A decision arising ​from opposition appeal proceedings, T0580/19 related to ​a patent in the name of B.R.A.H.M.S GmbH, which was ​opposed by Radiometer Medical ApS. The patent was ​directed to an in vitro method for diagnosing bacterial ​infections in a patient based on determining the level of ​procalcitonin (or a fragment thereof) in a patient sample ​shortly after the patient suffered a stroke or transient ​ischemic attack, and comparing this level to a threshold ​value. The opponent argued that the experimental data ​provided in the application as filed demonstrated that the ​method resulted in a significant number of false negatives, ​leading to misinformation being provided to medical ​practitioners. There were, the opponent argued, therefore ​serious doubts regarding the suitability of the claimed ​method for clinical diagnosis.

The proprietor countered these arguments on the basis the ​claimed method provided an effective tool for early diagnosis ​of bacterial infections, and while that came at the price of ​some false negatives occurring, false negatives are common ​to any diagnostic method. The opponent’s arguments were ​based on the extremely high false negative rate, suggesting ​that the method was not suitable for diagnosis. The data ​provided in the application as filed were based on clinical ​trial data, and indicated the best sensitivity achieved was ​around 41% (41 / 100 infected patients correctly identified). ​This meant that more than 50% of patients who were ​infected were not diagnosed. As a result, the Board of ​Appeal concluded that the claimed in vitro method was not ​suitable for the claimed purpose. Although the Board ​acknowledged that the invention provided an improvement ​over the prior art methods, the purpose of the claimed ​method was detection of infection in a patient, and the ​Board’s opinion was that the claim did not achieve this ​purpose. The Board of Appeal ultimately revoked the patent, ​overturning the decision of the Opposition Division to ​maintain the patent as granted.

T0555/17 & T1989/18 – Claim interpretation encompassing non-working embodiments

In T0555/17 the Board arrived at the conclusion that the claim should be read as to not encompass non-working ​embodiments, and declined to refer a question to the Enlarged Board of Appeal that had been submitted by the opponent ​during oral proceedings, namely “Is a claim sufficiently disclosed when it encompasses embodiments which the inventors ​explicitly state in the patent do not lead to success?”. The patent in question was granted with a claim directed to a method ​of diagnosing an anti-NMDA receptor encephalitis associated with dystonic movement, dyskinesias or autonomic instability in ​a subject, comprising the step of testing a sample for an antibody to an NR subunit of an NMDA receptor. The contribution of ​the invention was the establishment of a link between autoimmune encephalitis and the presence of autoantibodies, thereby ​defining a new patient group, namely those patients with auto-NMDAR encephalitis. The patent was opposed, with the ​opponent alleging that the claim encompassed embodiments that were disclosed in the patent not to work. The opponent ​was of the opinion that the claimed method should be interpreted to require the use of a detection antigen consisting solely ​of an NR subunit and, as the patent disclosed that individual NR subunits expressed in HEK cells could not be used ​successfully to detect the relevant antibodies, argued that the claim was insufficient.

In its decision, the Technical Board of Appeal reiterated that “For the purposes of Article 83 EPC / 100(b) EPC, the patent ​application must contain an enabling disclosure of testing for an antibody to an NR subunit of an NMDA receptor and it must ​be rendered at least plausible that such testing is suitable for diagnosing an anti-NMDA receptor encephalitis in a subject“. ​The Board disagreed with the opponent’s interpretation of the claims, and agreed with the patentee’s position that the claim ​is directed to a method of diagnosis requiring testing a sample for the presence of an (auto)-antibody against the NMDA ​receptor and that, as the NMDA receptor is exclusively formed of NR subunits (which associate to form a tetramer/heteromer ​in the functional receptor), any antibody which binds to the NMDA receptor must bind to one or more NR subunits.

The Board further reasoned that “non-working test formats are not part of the claim scope, because the claim is not ​restricted to any particular test and does not define which detector antigen should be used”. This case demonstrates that a ​sensible interpretation of the claim scope should be applied when construing a claim, in accordance with the principles set ​out in Article 69 EPC.

Final Remarks

The case law relating to meeting the plausibility threshold continues to evolve, and the decision of the Enlarged Board of ​Appeal (G2/21, on a referral from T0116/18) on the use of post-published data will be eagerly awaited. Of course, if possible, ​it is desirable to avoid the need to rely on post-published data at all. A fine balance must be struck between filing a patent ​application when enough data is available, but not waiting so long that a disclosure is made which precludes obtaining patent ​protection. What is apparent from recent case law is that when it comes to purported effects (particularly on living organisms - ​therapeutic or otherwise), and purported reports of synergy, examples and data are critical. Providing as much supporting ​data in an application from the outset will inevitably help to strengthen a patent application against future attacks.

T0941/16 – Disclosure of antibodies

When it comes to sufficient disclosure of antibodies defined by sequence, the EPO follows longstanding practice in that all six ​CDR sequences must normally be defined in the claims. However, this default position can be overturned if suitable evidence ​is provided to show this is unnecessary. In decision T0941/16, the Boards of Appeal went to great lengths to allow an ​applicant to submit such evidence. Claim 1 of the patent application on file defined an anti-PSMA antibody by “at least three” ​specific CDR sequences from the six present in the exemplified murine parent antibody, while an auxiliary request narrowed ​this to “at least five”, so leaving only one CDR undefined. The applicant’s position was that routine experimental techniques ​were capable of identifying antibodies with modified CDRs which retained the binding properties of the parent. Unusually, the ​Board did not give a decision at the hearing, but permitted the applicant to submit further data in support of this position. (As ​an aside, the applicant requested another hearing after submitting the data, but the Board considered this was unnecessary).

The applicant used several lines of argument. One, the existence of camelid single chain antibodies (having three CDRs) ​showed that specificity could be retained from a minimal set of CDRs. However, the Board dismissed this on the basis that ​the claims did not require all three CDRs to come from the same chain, which was a feature of “camelised” antibodies. Two, ​humanised antibodies may routinely incorporate additional modifications to the CDRs. The Board considered that the ​evidence on file showed that – in general – at least CDR L3 and H3 are necessary to retain binding specificity, and that other ​modifications may alter the recognised epitope. Three, specific experimental reports were provided showing that humanised ​versions of the exemplified antibody may include modifications to some of the CDRs while retaining specificity; again, though, ​the Board noted that each of these examples included unmodified CDR L3 and H3, which was not required by the claims.

Hence, despite being given ample opportunity to prove that antibodies as defined in the claims were possible to obtain, the ​applicant failed to do so, and the application was refused.

A further ambitious argument made was that, in the chemical field, it is permissible to recite a generic chemical formula in a ​claim, despite the existence of some non-functional embodiments within that claim. The applicant suggested that the same ​should apply to biotech patents. The Board noted that the same requirements for sufficiency apply in all technical fields, and ​that – following the case law – the claims on file were not allowable.

Thus, where a patent is sought for an antibody defined by the CDR sequences, the applicant should be prepared to provide ​evidence showing that variant antibodies can be obtained, or else should expect to have the claims limited to all six defined ​CDRs.

T0424/21 – Antibodies defined functionally, and sufficiency of broad medical use claims

Despite the above case, the EPO have shown in T 0424/21 ​that they are prepared to allow a claim to an antibody largely ​defined by its properties. The patent in question claimed an ​antibody having specific amino acid modifications to the Fc ​region, said to show reduced antibody-dependent cellular ​cytotoxicity (ADCC); thrombocyte aggregation; and effector ​function. The variable region of the claimed antibodies – that ​is, the region which binds to the target antigen – was ​undefined. The claim therefore covered a wide range of ​potential antibodies, yet the patent only included examples for ​specific antibodies against specific targets.

A further claim covered the use of these antibodies as a ​medicament; again, leaving specific diseases or specific ​targets undefined. The opponent, naturally, attacked the ​claims on the basis of insufficiency, suggesting that the Patent ​did not enable all antibodies covered by the claim to be ​produced, nor that there was any evidence that the antibodies ​could be used in therapy.

The patentee, on the other hand, argued that there was no ​reason to believe that the recited Fc mutations could not be ​incorporated into any antibody, or that they would interfere ​with target binding. Further, given that the patent disclosed ​the antibody for the first time, it was permissible to claim all ​therapeutic uses. The Board of Appeal agreed with the ​patentee, and, after reviewing EPO case law in relation to ​therapeutic uses, concluded that “This board cannot derive ​any requirement from the EPC whereby a patent would have ​to show that a compound is suitable for each and every ​disease in order for a first medical use to be sufficiently ​disclosed. Instead, it is sufficient to show that the compound ​is suitable for at least one particular medical use, as is the ​case in the patent at issue”.

Hence the patent was maintained with claims directed to the ​use of any antibody having the specific Fc mutations, for use ​in any therapy.

T 0032/17 – Biological deposits

For biotechnological inventions, such as a particular microbial strain or an antibody, where the invention may not have been ​completely structurally or genetically characterised, it may be possible to submit a biological deposit to meet sufficiency ​requirements by the EPO. The deposit is usually taken as disclosing the invention in sufficient detail to allow the skilled person to ​make and use it.

However, in T 0032/17, a claim to a monoclonal antibody defined in part by a deposit was found to be invalid. In particular, the ​claim was to an antibody defined by two features – the ability to recognise a specific vitamin (i.e. a functional feature) and that the ​antibody in question was produced by one of a specific list of deposited hybridomas (i.e. a product by process feature).

However, the Board of Appeal decided that despite the deposit of hybridomas (and hence the assumed availability of the ​antibodies to the skilled person) this feature could not be used to argue novelty. That is, the feature that the antibodies were ​produced by a particular hybridoma could not in itself distinguish the antibodies as such from prior art antibodies with the same ​functional property of recognising the specific vitamin, in part because there was no structural or sequence information for these ​antibodies.

Despite the patentee’s argument that the hybridoma deposit refers to a specific antibody having a specific sequence which ​would differ from other antibodies, the Board made the point that the bare deposit information does not in itself give any ​sequence or other information about the antibody as a product. In particular, the Board referred to the general case law ​requirements for product-by-process claims, and clarified that it is the burden of the patentee to demonstrate that a process ​feature of a product-by-process claim gives a “distinct and identifiable characteristic” to the resulting product. In short, this meant ​that this particular claim was only limited by the functional feature, and so lacked novelty over prior art antibodies against the ​vitamin.

The use of hybridoma deposits to define an antibody is less common than it once was, thanks to the relative ease of determining ​sequence information. However, this decision has broader implications for product-by-process claims where the process ​references a biological deposit (e.g. a specific cell line or microorganism). It appears that this process feature may not confer ​novelty, unless there is information in the application as filed that characterises the particular biological or chemical profile of the ​end product. For example, a claim to a beverage or food product produced from fermentation from a specific deposited yeast ​may not be novel over a product produced by an alternative fermentation source, unless there are specific chemical or functional ​characteristics of the resulting product that are included in the application as filed. Such information may not be strictly necessary ​for the skilled person to carry out the invention and therefore, in the normal course, the patentee would prefer to limit its ​disclosure. This decision may put weight behind an argument for retaining certain inventions as trade secrets rather than filing ​patent applications, where appropriate.

Alternatively, this decision gives further weight to the argument that relying on biological deposits to define an invention may be ​suboptimal if the invention can be defined in any other way. This is because the law dictates that a biological deposit can be ​inspected by any interested party after the patent is published, and can be freely accessed for any purpose after the patent ​expires. Therefore, rather than giving competitors access to a biological deposit that may directly lead to the commercial end ​product (e.g.. the disclosure of a transgenic cell line that expresses a mutant enzyme with a useful commercial property), it is ​likely to be more beneficial to explore broader ways of describing an invention. For example, characterising and claiming a ​specific mutation in a transgenic cell line, and claiming the cell line with reference to this mutation, rather than relying on a deposit ​of said cell line; or indeed trying to characterise the end product.

Accordingly, although biological deposits can be a precise and simple way to define an invention, and can be helpful where an ​invention cannot be defined in any other way, patent applicants may be well advised to seek to avoid relying on such deposits for ​both legal and commercial reasons.

Double Patenting

In G4/19, the Enlarged Board held that a European patent application can be refused if it claims the same subject-matter as ​an existing European patent (i.e. not just a co-pending EP application) which has been granted to the same applicant and has ​the same effective date. Double patenting is the claiming of the same subject matter in two applications by the same applicant ​which have the same effective date. Double patenting is prohibited in many patent systems around the world. For example, ​the UK patents Act at S73(2) specifies that a granted UK patent and a granted European patent which is validated in the UK, ​may not remain in force if both patents have been granted for the same invention having the same priority date, and that the ​applications for the patents were filed by the same applicant. The UK patents act at S18(5) also includes a double patenting ​provision against the patenting of two or more UK applications for the same invention having the same effective date by the ​same applicant.

However, the European Patent Convention (EPC) does not deal explicitly with double patenting, and simply states at Article ​125 EPC that, in the absence of procedural provisions, the EPO shall take into account the principles of procedural law ​generally recognised in the Contracting States. As such, it was argued that the EPC is not clear as to whether a European ​patent application can actually be refused on the ground of double patenting, and led to the following questions being referred ​to the Enlarged Board of Appeal:

1. Can a European patent application be refused under Article 97(2) EPC if it claims the same subject-matter as a European ​patent which was granted to the same applicant and does not form part of the state of the art pursuant to Article 54(2) and (3) ​EPC?

2.1 If the answer to the first question is yes, what are the conditions for such a refusal, and are different conditions to be ​applied depending on whether the European patent application under examination was filed,

1.a) on the same date as, or,

2.b) as a European divisional application (Article 76(1) EPC) in respect of, or,

3.c) claiming the priority (Article 88 EPC) in respect of a European patent application on the basis of which a European patent ​was granted to the same applicant?

2.2 In particular, in the last of these cases, does an applicant have a legitimate interest in the grant of a patent on the ​(subsequent) European patent application in view of the fact that the filing date and not the priority date is the relevant date for ​calculating the term of the European patent under Article 63(1) EPC?

In answering the above questions, the Enlarged Board of Appeal held that double patenting was a principle in procedural law, ​and thus Article 125 EPC provides the source for the EPO to refuse applications for double patenting by taking into account ​national law of Contracting States.

The Enlarged Board further held that the prohibition on double patenting is not limited to applications directed to the same ​subject-matter which were filed on the same day. It also extends to parent and divisional applications, and to applications ​claiming the same priority. The Enlarged Board also confirmed that the prohibition only applies where the application under ​examination and the patent already granted have common designated Contracting States.

It is worth noting that some Contracting States only consider two patents to relate to the same subject-matter if the ​independent claims are identical, whereas other States, such as the UK, take a broader interpretation and consider two ​patents to relate to the same subject matter if they relate to the “same invention”. As noted in the decision, the EPO takes the ​narrower interpretation as to what constitutes the “same subject matter”, such that, where there are any differences between ​the claims of the pending patent application and the earlier patent or patent application, a refusal on the grounds of double ​patenting should not be issued. Despite this, a national office may still apply the broader interpretation once the patent in ​question is validated in that country.

Further, there are instances when applicants have legitimate reasons for wanting a second patent application to grant which is ​directed towards the same subject matter as an already granted patent. For example, the second application may have been ​filed later and therefore provide a longer term of protection. However, it is clear from this decision, that whether or not an ​applicant has a legitimate interest in obtaining a second patent directed towards the same subject matter as an already ​granted European patent, the EPO might reject the pending application for double patenting. Therefore, when filing a later ​application, such as a priority claiming application or a divisional application, care should be taken to ensure that the claims ​are not identical to an earlier application, or if the claims are identical ensure that only the desired application proceeds to ​grant.

It is also important to note, as we move nearer to the launch of the Unitary Patent, that some contracting states, notably ​Germany, apply different rules on double patenting depending on whether the situation relates to a national patent together ​with a “classical” European patent, or a national patent together with a Unitary Patent. Thus, in some circumstances it may be ​possible to obtain simultaneous coverage for the same subject matter from two distinct patents.

Selection Inventions

In T 2635/16, the Board considered what constitutes a selection from a list in the prior art. This case emphasises that a ​feature that is claimed in a list of alternative features cannot be considered to represent a preferred feature unless it is ​disclosed as such.

In this case, Claim 1 of the patent was directed to a specific salt of a specific compound. During prosecution, the Examiner ​cited document D1 as novelty destroying for Claim 1. D1 related to compounds of a general formula or salts thereof and ​disclosed the specific compound in one of its claims, amongst a list of 121 other compounds. Separately, D1 disclosed the ​specific salt in a list of possible salts, referring however to the general formula and not the individually claimed compounds. ​During examination, the Examiner considered that the disclosure of the specific compound in the claims of D1 meant it was a ​highly preferred embodiment and thus identifying that compound from the claimed list was not a selection per se. Therefore, ​the Examiner considered that a selection of the specific salt from the list combined with this compound would allow you to ​arrive at the claimed invention. As a result, the Examiner considered that only a single selection was required from D1 to arrive ​at the claimed salt, holding (in line with established EPO practice regarding lists) that the claim was not novel.

The Board disagreed with the Examiner’s approach and highlighted that both the disclosure of the compound and that of the ​salt are in distinct and extensive lists in D1. In particular, the Board commented that while a claimed compound may be ​viewed as a highly preferred embodiment of the prior art, it cannot be considered an isolated embodiment unless it is ​disclosed in a narrower list or as an example. No specific examples or narrower lists containing the relevant compound were ​disclosed in D1. As such, the Board considered that isolating both the relevant compound and the relevant salt from D1 ​required a selection to be made from a list. In addition, the Board considered that a further selection had to be made to arrive ​at the claimed invention: choosing for the compound to be a salt thereof rather than the free compound. Hence, three ​selections from distinct lists were required to arrive at the claimed invention. The Board determined that Claim 1 was novel ​with respect to D1.

Liam

O'Connor