This is the third part of our three-part article on EPO Enlarged Board of Appeal referral G 2/21. Here, we take a deep dive into the case law for those looking for the detail, and include insights that can be gleaned from historical decisions on this area of law. Part I provides a five-minute summary of what you need to know (in a hurry), followed by practical implications for attorneys and applicants to help ensure applications are (plausibly) fit-for-filing and prosecution before the EPO. Part II is an analysis of the technical/legal facts of the appeal behind G 2/21 and provides a digest of post-filed data/plausibility issues in practice.

Plausibility isn’t mentioned or defined in European statutory law

As mentioned in part I of this article, the European Patent Convention (EPC) contains certain safeguards against entirely speculative patent applications for so-called “armchair inventions”. These include:

  • Article 83 EPC – sufficiency of disclosure (a skilled person reading the application as filed should be able to carry out the invention without further help/information);

  • Article 57 EPC – industrial application (the application must relate to an invention that can be made/used in industry – this is particularly applicable to gene inventions);

  • Article 52 EPC – patentable inventions (some things cannot be classed as an invention, e.g. discoveries and scientific theories); and

  • Article 123(2) EPC – amendments (an application cannot be amended to add subject matter which goes beyond the application as filed, e.g. adding a scientific rationale for your proposed technical effect after filing).

Over time, the EPO has also adopted a stance that technical effects/advantages must be at least plausibly demonstrated in a patent application as originally filed. Otherwise, objections under Article 56 EPC (inventive step) or Article 83 EPC (sufficiency of disclosure) may be raised. However, the term “plausibility” does not actually appear in the Rules and Articles of the EPC, whether in Articles 56 or 83 EPC or otherwise. Instead, the term has emerged from the last few decades of case law. This section takes us back to where it all began and provides insights into its subtleties.

Three key early decisions

(i) The dawn of the plausibility hurdle in Europe: T 939/92 (AgrEvo, 1995)

Those familiar with this topic will know that the concept of plausibility is considered to originate from EPO Board of Appeal decision T 939/92, AgrEvo, although the word “plausible” doesn’t actually appear in this decision.

In AgrEvo, claim 1 of European patent application no. 87303463.1 related to broad range of triazole compounds which were allegedly useful as herbicides. At first instance, the EPO Examining Division found that the scope of the claims did not represent a reasonable generalisation of the examples provided in the description because only around 30 compounds were tested for their herbicidal activity, whereas the claims defined a genus protecting millions of compounds.

On appeal, this opinion was shared by the EPO Technical Board of Appeal (TBA). The decision made reference to the quid pro quo principle that the extent of a patent monopoly should correspond to its technical contribution to the art – here it was determined that the scope of protection granted was too broad. Although the TBA invited the applicant to either restrict the claims to compounds which were credibly herbicidal or provide further evidence of wider herbicidal activity (Reasons 2.6.7), the applicant did not oblige and the application was refused (finally, with no further option of appeal).

In Europe, inventive step is assessed using the “problem-solution” approach. As part of this test, a solution to an underlying technical problem must be achieved across the whole scope of the claims. In AgrEvo, this meant that the applicant needed to demonstrate that it was “credible that substantially all claimed compounds possessed this [(herbicidal)] activity” (Reasons 2.6). However, the TBA in this decision therefore concluded that, since the applicant could only demonstrate that some of the claimed compounds possessed herbicidal activity, inventive step could not be acknowledged across the whole scope of the claim.

The underlying result of this decision is that it must be credible that a technical effect is achieved across the whole scope of the claims. This result has been frequently invoked by European examiners, opponents and boards of appeal.

(ii) Plausibility in terms of sufficiency of disclosure: T 609/02 (Salk Institute, 2004)

In this seminal decision, a key claim of the patent at issue (EP0552202, claim 6) related to the use of an active agent for the preparation of a pharmaceutical for use in the treatment of arthritis and other disorders (i.e. a Swiss-type medical use claim). In Europe, the medical indication in such claims is interpreted as a functional feature.

However, the TBA determined that the application as filed provided no evidence that this functional feature was actually achieved, e.g. there were no data or information in the application as filed demonstrating the impact of the agent on the listed diseases.

The patentee filed post-filed data (also known as post-published data) in support of this effect. However, the TBA concluded that the application as filed provided only a “vague indication of a possible medical use” and that “later more detailed evidence cannot be used to remedy the fundamental insufficiency of disclosure of such subject-matter” (see decision Reasons 13).

In other words, applying the principles of AgrEvo, it wasn’t made credible in the application as filed that the active agent could treat the diseases in the claim. Since the technical effect (the medical use) appeared in the claim, plausibility was raised here in the context of sufficiency of disclosure. As with AgrEvo, this result has been frequently invoked by European examiners, opponents and boards of appeal.

(iii) Plausibility in terms of inventive step: T 1329/04 (Johns Hopkins, 2005)

This decision is credited for entering the term “plausibility” into European patent law (and, incidentally, was handled by Kilburn & Strode).

The patent application in question (European application no. 94907259.9) related to a novel polypeptide which, according to the application as filed, belonged to a superfamily of compounds known to be important in embryonic development (the compounds behave as growth differentiation factors). No data relating to the functional role of the novel polypeptide were included, although it was suggested that the polypeptide could have either a stimulatory or an inhibitory role in the ovary, where it was shown to be located.

However, the TBA highlighted that the novel polypeptide did not exhibit what the TBA described as “the most striking structural feature” of the compound superfamily (a specific pattern of cysteine residues) and raised doubts that the novel polypeptide would truly be a growth differentiation factor in the same way as previously-known members of the superfamily.

The TBA concluded that inventive step could not be acknowledged because there was not enough evidence in the application as filed to “make at least plausible that a solution was found to the problem which was purportedly solved” (Reasons 11, emphasis added). Here, the problem to be solved was the provision of a further member of the superfamily.

The applicant submitted post-filed data showing that the polypeptide played a functional role similar to others in the superfamily and was indeed a growth differentiation factor, despite lacking the characteristic structural pattern with stimulatory effects on the ovary. However, these data were not considered by the TBA since they were deemed to be “the first disclosures going beyond speculation” that the polypeptide belonged to the superfamily. The application was refused (finally, with no further option of appeal).

The TBA in this decision mirrored the T 609/02 TBA’s adoption of the ab initio plausibility standard, stating that post-filed data “may not serve as the sole basis to establish that the application solves indeed the problem it purports to solve” (Reasons 12).

As with AgrEvo, the Johns Hopkins test is also often cited when determining whether or not post-filed data can be admitted into proceedings.

Back to G 2/21: EPO case law since Johns Hopkins

G 2/21 stems from EPO TBA opposition T 116/18. Here, the referring TBA considers there to be three diverging branches of case law, corresponding to the three possible standards discussed in the first and second parts of this article (the “ab initio plausibility” standard, the “ab initio implausibility” standard and the no plausibility standard).

The “ab initio plausibility” branch centres around T 1329/04 (Johns Hopkins) and T 609/02 (Salk Institute) described above. Other examples given are T 488/16 (part of the Dasatinib trilogy), T 415/11, T 895/13 and T 1791/11. The referring TBA notes that these cases did not allow post-filed data into proceedings because the respective applications as filed did not make the technical effect/advantage plausible.

In the “ab initio implausibility” branch, the referring TBA cites T 536/07, T 1437/07, T 578/06, T 266/10, T 863/12, T 919/15, T 184/16 and T 2015/20 as examples. According to the referring TBA, these cases did allow post-filed data into proceedings because, although there was arguably not enough information that demonstrated the technical effect/advantage was plausible, there were no legitimate reasons at the time of filing to doubt that the technical effect/advantage was plausible. In other words, there were no legitimate reasons to consider that the technical effect/advantage was implausible.

Under the “no plausibility” branch, the referring TBA cites only two decisions: T 2371/13 and T 31/18. In these cases, the notion of plausibility was rejected because this was considered to be incompatible with the problem-solution approach.

The examples cited by the TBA in these three branches include cases of both inventive step plausibility and sufficiency plausibility (the distinction between these is discussed in part II of this article). However, the preliminary opinion of the EBA indicates that the G 2/21 decision will not comment on sufficiency plausibility. We therefore expect EPO practice towards cases of sufficiency plausibility to continue on a case-by-case basis, using precedent set by the seminal T 609/02 (Salk Institute) case and later cases.

It remains to be seen which standard for inventive step plausibility the EBA will adopt in G 2/21. For practical tips on ensuring applications are (plausibly) fit-for-filing and prosecution before the EPO, please see part I. For a full digest of the facts at issue in G 2/21 and how the EBA’s decision could determine the validity of the patent in question, please see part II.

We’re closely monitoring developments: we’ll report further as G 2/21 progresses and the legal landscape becomes clearer.