Statistics and Takeaways from Continued Application of G 2/21 by the EPO’s Boards of Appeal and First Impressions from the UPC on “Plausibility”

G 2/21 and its interpretation in the underlying case T 116/18 – a short recapReference is also made to the authors’ previous article in epi information issue 1|2024

In G 2/21 of March 23, 2023, the Enlarged Board of Appeal (EBA) decided on the requirements for reliance on a purported technical effect for supporting inventive step, which is often the decisive factor for acknowledging inventive step according to the established evaluation following the EPO’s problem-solution approach. According to Headnote II of G 2/21, a patent applicant or proprietor may rely upon a technical effect for inventive step if the skilled person, having the common general knowledge in mind, and based on the application as originally filed, would derive said effect as being (i) encompassed by the technical teaching“Requirement (i)” in the following and (ii) embodied by the same originally disclosed invention“Requirement (ii)” in the following. If these requirements are fulfilled, also post-filed data can support the technical effect in the absence of such support in the application as filed.

In the underlying case T 116/18, inventive step hinged on whether a specific synergistic effect shown by post-filed data could be taken into account or not. In its final decision, Board 3.3.02 conducts an in-depth evaluation of the meaning of the rather abstract Headnote II of G 2/21 and derives still general but much more concrete test criteria:

• Both requirements (i) and (ii) must be cumulatively met.Reason 11.4 of T 116/18

• For requirement (i) to be met, the purported technical effect together with the claimed subject-matter need only be conceptually comprised by the broadest technical teaching of the application as filed, meaning that said effect need not be literally disclosed in it by way of a positive verbal statement, which “would correspond to, or at least come very close to, the very strict gold standard of a direct and unambiguous disclosure”Reason 11.10, also referring to reason 11.13.1 of T 116/18. This criterion evaluates whether the purported technical effect in connection with the claimed subject-matter is within or outside the scope of the teaching of the application as filed.

• Requirement (ii) is met unless the skilled person, having the common general knowledge on the filing date in mind, and based on the application as filed, would have legitimate reason to doubt that the purported technical effect can be achieved with the claimed subject-matter. Experimental proof of the effect or a positive verbal statement is not necessarily required in the application as filed.Reason 11.13 and 11.14 of T 116/18 There might be a legitimate reason to doubt if the effect is irreconcilable with the common general knowledge.Reason 17.4.2 of T 116/18 This criterion is phrased as an exception to the rule of an effect being accepted. An effect, despite falling within the scope of the teaching, would only be excluded in case there is evidence of a “very basic disbelieve”. In case an opponent would like to rely on this exception, it carries the burden of proof. The standard for showing such a disbelieve is set rather high (“irreconcilable with the common general knowledge”). The mere argument that the effect is surprising is not sufficient.Reason 17.4.3 of T 116/18

The Board in T 116/18 concluded based on this evaluation that the proprietor may rely on the purported technical effect.Reason 18 of T 116/18 A petition for review (Art. 112a EPC) is still pending.Case R 4/24

Overview and statistics – The majority of decisions allowed reliance on the technical effect in question

As of March 2025, we are aware of 34 decisions issued by the EPO’s Boards of Appeal that apply the test established by the EBA in G 2/21 (Headnote II) for assessing whether a certain purported technical effect can be relied on in the context of inventive step (see Timeline).

In most decisions, the Boards took the purported technical effect into account (see Figure 1). Until now, the 3.3-Boards (Chemistry) and particularly Boards 3.3.02The Board in T 116/18, the underlying case to G 2/21 and 3.3.07 dominate the case law, while there is no decision yet from Board 3.3.08 (see Figures 2 to 4).No members allocated to Board 3.3.01 since G 2/21 published There are only few decisions from the 3.2-Boards (Mechanics) and 3.4-Boards (Physics), while there is no decision yet from the 3.5-Boards (Electricity).

Subsequent decisions referring to the underlying case T 116/18 and applying one or more of its test criteria

As outlined above, T 116/18 provides a very detailed analysis of G 2/21 and derives from the abstract guidance more concrete – however, still general – legal test criteria to be applied for accepting a purported technical effect. The question arises whether these more concrete test criteria are now generally applicable and accepted among the other Boards. Some decisions expressly refer to T 116/18 and use one or more test criteria elaborated therein:Decision colour code in the following: green = technical effect taken into account; red = technical effect not taken into account, orange = certain effects taken into account, others not

• Board 3.3.03 in T 364/22 evaluated post-filed evidence and agreed with the test criteria established by T 116/18 in that, regarding requirement (i), a positive verbal statement of the effect in the application is not required (reference to reason 11.10 of T 116/18) and that, regarding requirement (ii), the question is whether the skilled person would have legitimate reason to doubt that the technical effect can be achieved with the claimed subject matter (reference to reason 11.11 of T 116/18).Reason 1.4.4. of T 364/22

• Board 3.3.07 in T 228/23 concluded that the purported effect “constitutes a purpose of the invention” and is thus encompassed by the technical teaching of the original application as required by G 2/21.Reason 4.3.6 of T 228/23 As regards requirement (ii), the Board found that “[i]n this specific case of a very limited number of embodiments defined in the original application, one of these being also the subject-matter of the closest prior art and the other one being generally preferred, and in the absence of any legitimate reason at the effective date to doubt that the alleged effect could be achieved with the claimed subject-matter (see T 116/18, reasons 11.14),…the appellant - patent proprietor should be entitled to specify a preference for one of said originally disclosed embodiments over the other in connection with said effect”.Reason 4.3.7 of T 228/23

• Board 3.3.02 (i.e., the same Board as in T 116/18, but in a different composition) in T 1994/22 also agreed with the findings of T 116/18 under reason 11.10 regarding requirement (i) (effect together with the claimed subject-matter need only be conceptually comprised by the broadest technical teaching of the application) and reason 11.14 regarding requirement (ii) (no positive verbal statement and no data regarding the technical effect required in the application). However, in this case the Board concluded based on these test criteria that the purported effect of “improved photostability” cannot be relied upon. A “sweeping statement” in the application as filed referring to “high quality” and “easy industrial handleability” was considered too vague and remote with respect to the argued photostability effect.Reason 1.3 of T 1994/22

Further practical guidelines and considerations by the Boards in the context of the G 2/21 test

No application of the “gold standard”: Board 3.3.10 in T 1602/21 considered that whether the skilled person would derive an effect as being “encompassed” according to requirement (i) “is not equivalent to the “gold standard” disclosure required normally for Article 123(2) EPC”.Reason 36 of T 1602/21 See also T 116/18 in the context of both requirements of G 2/21Reason 11.10, 11.13.1, and 11.13.2 of T 116/18.

Preference for an embodiment may be helpful: Board 3.3.07 in T 873/21 considered a synergistic effect “embodied by the present combination since it was clearly the preferred combination in the original application”,Reason 3.3.3 of T 873/21 and Board 3.3.06 in T 681/21 regarding an Auxiliary Request considered that an improved effect “is especially obtained by using a combination with the components indicated as preferred”Reason 5.2.1 of T 681/21, but regarding the Main Request a preference seemed not sufficientReason 1.2.3 of T 681/21.

• If an effect is derivable, a specific improvement of the effect is likewise: From many decisions, it may be derived that once a certain technical effect is considered “encompassed” and “embodied” in line with G 2/21, an improvement, optimization, increase, or synergy of such effect over a prior art is accepted. Board 3.3.02 in T 840/22, for example, concluded that “it must be possible… to rely on an improvement in the derivable effect, at least in the case of an improvement over a piece of prior art… The fact… that one of several embodiments disclosed in the application as filed turns out to be better after the effective date with respect to the technical effect set out in the application as filed…, does not change this conclusion.”Reason 15.2 of T 840/22 Many decisions seem to follow this principle: T 873/21Reason 3.3.2 of T 837/21, T 1989/19Reason 3.3.16 of T 1989/19, T 885/21Reason 6.4.1 of T 885/21, T 1329/21Reason 3.2.4 of T 1329/21, T 728/21Reason 2.2 of T 728/21, T 681/21Reason 5.2.1, but different in reason 1.2.3 of T  681/21, T 2716/19Reason 8.3.2 of T 2716/19, T 1551/22Reasons 2.5 and 2.6 of T 1551/22 T 2046/21Reason 2.2.3 to 2.2.8 of T 2046/21, T 1525/19Reason 7.3.5 of T 1525/19, T 1602/21Reason 35 and 36 of T 1602/21 and T 318/22Reasons 4.5.3 to 4.5.9 of T 318/22.

Limits of accepting an effect: Board 3.3.02 in T 852/20Reason 3.5.3 of T 852/20 and Board 3.3.07 in T 1698/21Reason 4.3.3 of T 1698/21 expressed that an effect which is not associated with the claimed subject matter cannot be accepted. Furthermore, Board 3.3.07 in T 2046/21Reason 2.2.9 of T 2046/21 and T 258/21Reason 1.3.3 of T 258/21 rejected an effect which had no connection at all to the original disclosure. And in T 887/21Reason 2.15 of T 887/21, Board 3.3.04 applied a stricter approach in that an improvement of the effect seems not automatically covered. Additionally, the effect was considered not to be based on the originally disclosed hypothesis.

Interim summary

All decisions apply the EPO’s problem-solution approach, which includes by way of a “retrospective” comparison of the claims with the closest prior art the determination of the objective technical problem based on the effect provided by the difference. If no superior effect can be shown, the problem is only an alternative solution often considered not inventive. The effect is therefore regularly crucial but often not literally part of the original disclosure and therefore very strictly speaking a “new” effect, which may be because the later determined closest prior art was not necessarily known at the time of filing the application (see T 2465/19Reason 5.3.1 of T 2465/19) and/or an effect later proves not to be present over the whole scope of the claims (see T 840/22Reason 15.2 of T 840/22). Such a “new” effect could regularly not be accepted if a verbal statement of this effect and a specific disclosure rendering this effect plausible would be required. Most decisions including T 116/18 therefore seem to apply a more lenient standard. However, providing certain flexibility for applicants/proprietors to adapt an effect depending on the specific prior art cited may risk opening the door to mere speculation, which cannot always easily be distinguished. Board 3.3.04 in T 314/20 voices potential issues with the T 116/18 approach and seems to adopt a stricter approach than most of the above discussed decisions. Therefore, T 314/20 will be reviewed separately in the following.

T 314/20 provides a very detailed and critical review of T 116/18

Board 3.3.04 in T 314/20 provides the first very detailed review of the interpretation of Headnote II of G 2/21 by Board 3.3.02 in T 116/18 and seems to challenge certain assumptions made by said Board.Reason 6.13 of T 314/20 While the decision in T 314/20 was announced at the oral proceedings of 16 November 2023, the written decision was published more than a year later on 20 December 2024, which indicates that the Board took the time to write a very detailed and reflective analysis in this complex environment.

T 314/20 concerns a combination product comprising empagliflozin and linagliptin. The difference to the closest prior art was this specific compound combinationReasons 3 and 4 of T 314/20. The G 2/21 discussion applying G 2/21 centered on the purported technical effect of an increase in glucagon-like peptide-1 (GLP-1) levels (resulting in an improved diabetes treatment) obtained with this combination and as allegedly shown in post-filed dataPoint X and reasons 5 and 6.5 of T 314/20. The Board found that “the increase in active GLP-1 levels… forms the basis of an invention which has not been made or disclosed at the filing date of the application. Consequently, this technical effect cannot be considered for formulating the objective technical problem”Reason 6.28, in consideration of reasons 6.14 to 6.27, of T 314/20.

When coming to its decision, the Board seems to have rejected the later-shown improvement in the effect for the claimed combination because a preference or superiority over other disclosed combinations was lacking, although the effect as such was originally disclosed. This conclusion seems to be generally different from decisions as, e.g., T 840/22 accepting an improved effect for a specific embodiment despite a lack of preference in the original application and the improvement in the effect being shown only later (supra).

Furthermore, the Board stated that it generally “faces two issues with the interpretation of decision G 2/21” in the earlier case T 116/18Reasons 6.13.4 and 6.13.5 of T 314/20:

• First, although “the relevant Board [Board 3.3.02 in T 116/18] stated… that in decision G 2/21 the Enlarged Board did not refer to any ‘plausibility’ standards… but instead adopted new requirements” and “that when deciding whether a patent applicant or proprietor may rely on a purported technical effect for inventive step, it was the requirement(s) defined by the Enlarged Board in point 2 of the order that had to be applied, rather than simply using any rationale developed in the previous plausibility case law”Reason 6.13.6 of T 314/20, “the Board [again, Board 3.3.02 in T 116/18] effectively adopted what it defined in the referral T 116/18 as the ‘ab initio implausibility’ standard”Reason 6.13.8 of T 314/20; see a similar conclusion by the authors in their article in epi Information, issue 1, 2024. According to the Board [Board 3.3.04], this is not stated in G 2/21Reason 6.13.9 of T 314/20.

• Second, while the Board in T 116/18 “concluded that a positive verbal statement of the technical effect… in the patent application as originally filed was not necessary” for the requirements of G 2/21 to be met, “such a statement is also not to be found in decision G 2/21”Reason 6.13.9 of T 314/20. Board 3.3.04 considered that “the passages of decision G 2/21 referred to by decision T 116/18 do not lead to the conclusion that the standard of disclosure that applies to Article 87 EPC…, Article 123(2) EPC or Article 54(1) EPC… is excluded” when assessing whether the requirements of G 2/21 are fulfilled. “While this may be a valid approach, it does not follow from the text of decision G 2/21”Reason 6.13.11 of T 314/20.

• The proprietor in T 314/20 argued emerging different interpretations of G 2/21 and requested a new referral to the EBA for further clarification which was, however, denied. Next to other reasons, Board 3.3.04 considered it uncertain whether the EBA would in fact further specify Headnote II, since it could have done so already in G 2/21 but refrained.Reasons 12 to 20 of T 314/20,Central Division Munich, for further details see the authors article in Managing IP of 24 January 2025 (“After two years of G 2/21 case law, is the interpretation now settling?”)

First impressions from the UPC

Since its launch on 1 June 2023, the UPC has rendered several decisions/orders on inventive step. The majority comes from the Court of First Instance (CFI), including the frequently cited Sanofi v. AmgenUPC_CFI_1/2023, decision of 16 July 2024 decision from the Munich Central Division that also refers to the Court of Appeal (CoA)’s inventive step assessment in NanoString Technologies v. 10x GenomicsUPC_CoA_335/2023, order of 26 February 2024, at 5.b.cc. The Court in Sanofi v. Amgen appears to follow two main steps:

1. Selection of a realistic starting point without the need to select the most promising starting point.Sanofi v. Amgen Headnote 3 and at 8.6
2. Analysis of whether it would be obvious for the skilled person, starting from a realistic prior art disclosure, in view of the underlying problem, to arrive at the claimed solution. In general, the solution is obvious if the skilled person would be motivated, i.e., have an incentive, to consider the claimed solution and to implement it as a next step in developing the prior art.Sanofi v. Amgen Headnote 4 and at 8.7 and 8.8 A technical effect or advantage achieved by the claimed subject matter compared to the prior art may be an indication for inventive step.Sanofi v. Amgen Headnote 6 and at 8.9

The underlying problem that the patented invention aims to solve was defined on the basis of the disclosure in the patent upfront to the claim interpretation and the entire validity assessment and was not specifically defined in the context of the inventive step analysis as in the EPO’s problem-solution approach.Sanofi v. Amgen at 5.9-5.16; NanoString Technologies v. 10x Genomics at 4.a and b A specific technical effect associated with the difference between the starting point in the prior art and the claimed invention seems rather optional in the UPC’s inventive step assessment. In contrast, as outlined above, in the problem-solution approach, the specific effect resulting from the difference to the “closest prior art” guides the determination of the specific “objective technical problem” and therefore plays the central role for the EPO’s inventive step assessment. Since G 2/21 concerns the possibility “to rely on this specific (and often “new”) effect and corresponding objective technical problem, the UPC inventive step assessment, at least if it consolidates in the above-outlined approach, may not give rise to the G 2/21 topic as much as we see it at the EPO.

Currently, there is no clear common UPC approach for inventive step assessment yet. Seoul v. expertUPC_CFI_363/2023, Local Division Düsseldorf, decision of 10 October 2024, at B.I.1, C.II.1, Hand Held Products v. ScanditUPC_CFI_74/2024, Local Division Munich, order of 27 August 2024, at II.1, III.3.d.aa, NanoString Technologies v. President and Fellows of Harvard CollegeUPC_CFI_252/2023, Central Division Munich, decision of 17 October 2024, at 6.5-6.7 and 11.10-11.19, and NJOY v. VMR ProductsUPC_CFI_311/2023, Central Division Paris, decision of 21 January 2025, at 55-84, and Headnote, for example, appear to generally follow the above approach.See footnotes 55-59 Also the Court in Plant-e v. Arkyne TechnologiesUPC_CFI_239/2023, Local Division The Hague, decision of 22 November 2024, at 35, 52-56 generally seems to follow this approach, though it states that it “will follow the problem and solution approach… as also used by the EPO”. The Court in Advanced Bionics v. Med-EL Elektromedizinische GeräteUPC_CFI_338/2023 and 410/2023, Central Division Paris, decision of 26 December 2024, at 85-113 in contrast seems to apply the problem-solution approach as conducted by the EPO including assessment of the effect achieved by the distinguishing features and formulating the objective technical problem based thereon. Further, in Meril v. Edwards LifesciencesUPC_CFI_255/2023 and 15/2023, Central Division Paris, decision of 19 July 2024, at 153-155, Huawei Technologies v. NetgearUPC_CFI_9/2023, Local Division Munich, decision of 18 December 2024, at III.3.a.aa, and Avago Technologies International Sales v. TeslaUPC_CFI_54/2023, Local Division Hamburg, decision of 26 August 2024, at C.VII.1-3 the Courts concluded that application of the EPO’s problem-solution approach would not lead to a different outcome on inventive step, although not performing a detailed analysis in this respect.

As of today, we are aware of one UPC case that discussed “plausibility” and conducted the G 2/21 test. Syngenta v. Sumi AgroUPC_CFI_201/2024, Local Division Munich, order of 27 August 2024 concerned a herbicidal composition in which the herbicides were said to be chemically stabilized.Ground B.I The Court initially applied the above-outlined inventive step test referring to NanoString Technologies v. 10x Genomics and Sanofi v. AmgenGround B.I and C.II.4.b.aa and concluded that inventive step was not called into question.Ground C.II.4.b.bb.3 and 4 As a secondary consideration and in the context of an argument by the defendant that the originally-disclosed stabilizing effect is not achieved over the whole scope of the claim in view of the allegedly limited examples in the application as filed,Ground C.II.4.b.bb.1 the Court considered G 2/21 and outlined that even if the test in G 2/21 is applied, the conclusion would not change. It reasoned that the skilled person at the filing date would have derived the purported stabilizing effect over the whole scope of the claim from the application as originally filed and considered that this effect is supported by both the patent examples and post-filed data relating to the defendant’s herbicide. Therefore, the Court concluded that it “can leave open the question of whether this test [i.e., the test in G 2/21] should be applied.”Ground C.II.4.c.ff The order granting provisional measures was appealed.UPC_CoA_523/2024

Very recently, in its order of 3 March 2025, the CoA concurred with the findings by the first instance regarding inventive step. According to the CoA, Sumi Agro has not substantiated that the skilled person, despite the explicit disclosure of the stabilizing effect “would have reason to believe that this technical effect is not achieved over the whole claimed range.” Using the words of the EBA in Headnote II of G 2/21, however, not expressly mentioning G 2/21, the CoA concluded:UPC_CoA_523/2024, order of 3 March 2025, at 84-89

“This technical effect must therefore be considered encompassed by the technical teaching and embodied by the originally disclosed invention and may thus be relied upon by Syngenta when considering inventive step.”

Proceedings on the merits are ongoing.UPC_CFI_566/2024

Overall conclusion

Many decisions by the EPO’s Boards of Appeal create an appearance that the G 2/21 case law is settling towards a more lenient approach regarding acceptance of a technical effect, including a “new” effect not literally part of the original disclosure and determined following the problem-solution approach. T 314/20, however, seems to question this. There is doubt whether the detailed test outlined in T 116/18 or similar approaches will be the general standard. Furthermore, not all Boards have yet contributed equally to the above summarized case law and the current picture of a dominating more lenient approach in favour of the applicant/proprietor may be incomplete.

The G 2/21 topic is linked to the central role of the effect determined when conducting the EPO’s problem-solution approach, often resulting in a “new” effect being the main cause of the critical disputes. As the problem-solution approach may not be the common future approach of the UPC, the more critical G 2/21 topics focusing on a “new” effect may not arise. The first UPC decision touching on “plausibility” and G 2/21 (Syngenta v. Sumi Agro) did not apply .the problem-solution approach and did not deal with a “new” effect. As a secondary consideration, the question whether the originally-disclosed effect is present over the whole claim scope was discussed. The CFI applied the G 2/21 test and accepted post-filed data. Since the application of this test did not alter the outcome on inventive step, the CFI left open whether the G 2/21 test must be applied by the UPC.

The CoA seemingly applied the G 2/21 test and accepted the technical effect. As this effect was clearly originally disclosed, the outcome of this UPC case seems in line with the EPO approach in such a situation. However, in view of the current emerging differences regarding the inventive step analysis between the UPC and the EPO, it remains to be seen whether there will ever be a UPC case having to deal with the more critical situation of a “new” effect and how this would then be decided.

Comments