Decision G2/21 concerning the EPO Enlarged Board of Appeal (EBA) Plausibility Referral was published, together with a Press Release, on 23 March 2023.

The guidance is: in support of inventive step, yes, post-filed data can be used, if it relates to the same invention as disclosed in the original application.

Summary of the Decision

In summary, in support of inventive step in a claim in a European patent:
• yes, post-filed evidence can be used; and
• yes, it can be used to prove a technical effect.

But this is subject to a proviso:
• that the technical effect to be proved can be derived from the application as filed and that it is directed at the same invention as disclosed in the application as filed.

This last bit needs some unpicking, as I have (riskily?) paraphrased the decision; the precise wording of the proviso reads:
“ … if the skilled person, having the common general knowledge in mind, and based on the application as filed, would derive said effect as encompassed by the technical teaching and embodied by the same originally disclosed invention” (Order, Point 2).

The Message?
My reading is there is a message of huge importance to all of us drafting patent applications: there is a need to commit, in the application as filed, to the invention and its specific technical effect if we subsequently wish to rely on post-filed data demonstrating that the invention really has the alleged technical effect. Post-filed data showing a different technical effect, not originally disclosed, can not be relied upon, no matter how wonderful, efficient, beneficial, etc the invention turns out to be in a respect not mentioned on filing.

Discussion of the Decision and the Guidance it Provides

Commonality is needed between the new data and the original disclosure. A similar requirement for correspondence of disclosure is thus invoked to that required for a valid claim to priority, as per EBA Decision G2/98 (Requirement for claiming priority of the ‘same invention’). That is, the disclosure is read in the light of the common general knowledge (CGK) but without supplementing that disclosure.
From this CGK-informed reading the invention is determined and the evidence needs to support inventive step in the same invention. There must thus be common ground between the invention disclosed in the application as filed and the invention supported by the evidence. The post-filed data may not change the nature of the invention from that originally disclosed (G2/21, Reasons, 93).

Within this framework, for an invention (necessarily having a technical effect) derivable from the application as filed then post-filed data can be relied upon to support the inventive step of that invention.

We now have an answer for all the applications and opposed patents whose progress at the EPO has been stayed till now on the basis their fate lies entirely on whether post-filed evidence can be used to support inventive step. The EBA decision is the guidance required to assess whether post-filed data can be used. The guidance seems clear, and it is now for each competent board or other EPO deciding body to apply this to the facts of a given case (G2/21, Reasons, 95).

The decision can be accessed here.
The Press Release can be accessed here.

Facts of the referred case

As per the facts of the case underlying the referral, a combination of active agents was claimed and the combination was novel. It was also sufficiently disclosed. Synergy of the combination was alleged in the patent but not proven by data in the patent. This meant that patentability (inventive step) would turn solely on post-filed data showing synergy. And data had been filed. The object of this referral was thus determining whether the data can be relied upon.

Regarding the questions referred to the EBA

Without reciting the referral questions in full, a critical component was whether post-filed evidence is admissible when without that evidence the claim is not patentable for lack of inventive step. As noted above, the answer is yes, with the proviso.

EPO Case Law

There’s a lot of it, as the EBA acknowledged, but we can review the more recent cases as these are the culmination of the body of case law (G2/21, Reasons, 64). T31/18 held that the technical effect relied upon for inventive step must be explicitly described in the application as filed or at least derivable therefrom (G2/21, Reasons, 65). This allowed a parallel to be drawn with G2/98, as discussed above.

There is, in addition and in parallel, case law supporting the ab initio plausibility approach and the ab initio implausibility approach. Briefly, though without doing full justice to the minutiae of the criteria for these approaches: one is concerned with whether the disclosure is plausible based on some disclosure, preferably data; and the other with whether the disclosure, not being supported by data, is nevertheless not implausible.

In all of this the EBA found a common theme: was the technical effect derivable from the contents of the application as filed (G2/21, Reasons, 72 and 87)?

Free evaluation of evidence

It’s a basic procedural right to bring evidence in support of an inference in legal proceedings that is decisive for the final decision (G2/21, Reasons, 32). Whether that evidence is convincing is another matter (G2/21, Reasons, 34) though if it is relevant and offered as proving the key inference then it should be admitted and considered (G2/21, Reasons, 42 and 43).

Bringing the threads together

The combination of these strands of analysis yields the guidance: invention has technical character and is disclosed in the application as filed, taking CGK into account; evidence is in principle admissible if it is relevant; and it is relevant if it relates to the same invention and does not change the nature of the invention. The evidence must of course finally be assessed as to whether it is convincing.

Application of the decision to issues of sufficiency

The decision relates to evidence in support of inventive step. In my own practice, oppositions where the proprietor has tried to file evidence in support of sufficiency, claiming this aspect is alone determinative for the outcome of the oppositions, have been refused a stay: the G2/21 referral relates only to inventive step, not more broadly. Here, the EBA explicitly concluded they were not permitted to extend the questions so as to relate to sufficiency (G2/21, Reasons, 11). In the next sentence, with a nod to exactly the opposite, the EBA noted it is aware of the respective EPO case law on sufficiency (G2/21, Reasons, 12).

The doors are at first sight flung wide to all those who will seek to cite the principles of G2/21 to permit filing of post-filed evidence to support sufficiency. The caveat is the decision also notes there is less scope at the EPO for reliance on post-filed data to support sufficiency (G2/21, Reasons, 77).

Farewell, ‘plausibility’?

Is the concept of plausibility now gone? The EPO President and others argued that plausibility criteria arise when a technical effect is referenced, and it is necessary to assess whether the skilled person believes, in the sense of reasonably expecting it to be true, that there is a technical effect (which can be objectively determined).

I once argued that when the emeritus Professor of Biochemistry tells you there is a technical effect then it is plausible and you believe so. This failed, alleged to be a non-objective analysis. In the current decision, the EBA noted that ‘plausibility’ arises, e.g. as a criterion for whether one should rely on an alleged technical effect (G2/21, Reasons, 58) and more explicitly that:
“plausibility […] does not amount to a specific legal concept or a specific patent law requirement under the EPC” (G2/21, Reasons, 92).

A worked example?

Let’s imagine claims directed to an aircraft wing, with particular profile, shape, size, etc. No indication of specific function is given, other than as an aircraft wing. Post-filed data shows improved anti-stall properties during slow flight in wet weather. So, the wing has beneficial properties. But this was not disclosed in the original application; the invention demonstrated in the data is not same as in the original application. The proviso in the guidance, commonality of invention between data and disclosure, is missing.

And another?

In the chemical field, compounds are disclosed, defined by a formula in which a series of embodiments are defined. Preferred compounds have -OH substituents, i.e. are alcohols. Post-filed data shows these preferred alcohols have improved solubility and are, as a result, more bioavailable and more effective therapeutically. The original disclosure says the alcohols are preferred but without saying why; to me a nexus between the choice of -OH substituent and the result obtained is missing, and the data risks being held not directed at the same invention, because the original disclosure does not disclose the invention relates to solubility.

Now we rewrite the original disclosure, with the hypothetical benefit of hindsight, to state these preferred -OH containing compounds have better solubility. The proviso is now met and the evidence can be relied upon.

Other points

Law of national courts

The EBA deliberations relied on case law of the national courts of the contracting states: 11 decisions from the combined courts of Switzerland, Germany, France and the Netherlands, and 16 decisions from the UK courts.

Regarding the EBA Priority Referrals

The EBA notes it is not bound by the wording of the questions referred and can deviate to better reflect the true object and focus of the referral (G2/21, Reasons, 4). Time will tell whether the questions referred to in the EBA Priority Referrals (G1/22, G2/22) might be reworded so as to relate solely to its real object, i.e. the PCT Joint Applicants approach, rather than to priority issues already decided in T844/18.

Regarding the referring case

A next step in the referring case is to act on the guidance, and it is a reasonable conclusion that the data, if it relates to the same invention and demonstrates the alleged synergy, can indeed be relied upon.


The requirement, from the proviso in the guidance, for a same invention test linking data and disclosure is fundamental. We will need to commit, in the application as filed, to the invention and its specific technical effect if we subsequently wish to rely on post-filed data demonstrating that the invention really has the alleged technical effect.

Of course, the decision provides guidance and we will be glad to apply it to a given case upon request!

Our articles are for general information only. They should not be considered specific legal advice, which is available upon request. All information in our articles is considered to be accurate at the date of publishing.

Latest Firm News

Schlich Walks for Mental Health Awareness Week

Schlich Walks for Mental Health Awareness Week

With the sun making an appearance and this week being Mental Health Awareness Week we have swapped our treadmill team effort and decided to join the ‘2024 Mental Health – Moving More Challenge’ by taking ourselves away from our desks and walking along the seafront during our lunch breaks. 

read more