One year after G 2/21: where do we stand today?

It is now a year since the Enlarged Board of Appeal had to decide on the reliability of a technical effect shown only by post-filing data in the assessment of inventive step. The wording of the answer is reminiscent of answers usually given by the CJEU. It was not much of a surprise that the Enlarged Board held that post-filing data cannot be disregarded simply because they are filed after the filing date. The first referral question had, however, asked whether a technical effect that was shown exclusively by post-filing data should be disregarded under the principle of free evaluation of evidence. The word “exclusively” did not appear in the Enlarged Board’s answer, but the phrase “evidence, on which the effect rests” implied that such evidence was meant.

The standard set by G 2/21

In reviewing case law of the Boards of Appeal, the Enlarged Board identified a core issue which it considered to be common ground to all decisions: What does the skilled person, with the common general knowledge in mind, understand at the filing date from the application as originally filed to be the technical teaching of the claimed invention?See point 71 of the reasons of G 2/21 The technical effect shown by post-filing data had to be derivable by the person skilled in the art from the technical teaching of the application as filed.See point 72 of the reasons of G 2/21

At the end of the decision, the Enlarge Board summarised two aspects of “the relevant standard for the reliance on a purported technical effect”,See points 93 and 94 of the reasons i.e. a technical effect demonstrated by post-filing data: (i) the effect must be encompassed by the technical teaching of the original application, and (ii) it must be embodied by the same originally disclosed invention. As stated above, the assessment is of course to be made from the point of view of the person skilled in the art.

The main question that arises in the light of the reasons of the decision is what standards the words “encompassed” and “embodied” set for the disclosure of the application as filed. The Enlarged Board gave no guidance in this respect. It will be of interest to applicants what follows from the fact that the Enlarged Board said that it has identified the above first criterion as common ground after reviewing the case law of the last two decades: Will Technical Boards of Appeal take this as confirmation to continue as before, or will they change their approach in the light of G 2/21 when considering post-filing data?

Examples of evidence considered/not considered prior to G 2/21

Two decisions in which Boards took rather contradictory approaches illustrate where the shoe really pinched prior to the decision G 2/21.

In decision T 939/92 (AgrEvo), the claim at issue covered a large number of compounds in a product claim, and the Board did not consider it credible that all compounds had the purported technical effect. The applicant had “been given ample opportunity … to provide further evidence, either by test results or by other means” to prove the technical effect.See point 2.6.7 of the reasons of T 939/92 The Board would therefore have accepted additional data, which was, however, not submitted.

A similar situation arose in decision T 488/16: a group of chemical compounds, defined in a product claim, was ascertained to have a technical effect, but the application as filed lacked data for any of the compounds covered by the claims. The Board disregarded the post-filing data, because it did not consider it plausible to the person skilled in the art that every compound disclosed in the application as filed would have the purported technical effect.See point 4.9 of the reasons of T 488/16

In applying criterion (i) of G 2/21, the effect shown by the post-filing data submitted in T 488/16 was apparently encompassed by the general teaching of the application as filed. However, in both T 939/92 and T 488/16, the Boards had taken the view that the person skilled in the art would not consider the purported effect to be clearly applicable to all disclosed compounds. It was only in T 488/16 that the Board held that there was a need for appropriate teaching in the original application. In T 939/92, the Board was apparently prepared to accept and take into account post-filing data whenever filed.

As regards the second criterion of G 2/21, it seems difficult to dissect the claims in either of the two decisions to the effect that for some compounds the purported effect was not embodied by the same invention.

The case of G 2/21 finalised and interpreted

The case underlying decision G 2/21, namely T 116/18, has in the meantime been decided in favour of the patentee. Claim 1 at issue was a product claim, namely an insecticidal composition comprising a specific compound and a further compound defined by a Markush formula. This effect had already been disclosed in the application as filed. Data in the application as filed showed a synergistic effect on two insect species. During opposition proceedings, the patentee had submitted further data showing a synergistic effect for a third insect species. The opponent had filed data with the grounds of appeal showing that the synergistic effect did not occur over the whole range claimed for the first two insect species, i.e. the insect species for which the opposed patent contained data. Here, the problem to be solved could be limited to the third insect species.

The referring Board concluded that the patentee’s post-filing data should be taken into account, and that on this basis an inventive step could be acknowledged.

In its reasoning, the Board found that the two criteria for considering the technical effect, namely in (i) the “technical teaching”, and in (ii) the “same originally disclosed invention”, were in fact one and the same issue.See point 11.9 of the reasons of T 116/18. The Board was of the view that “G 2/21 seeks to prevent speculative inventions.”Often termed armchair inventions, see point 11.8 of the reasons of T 116/18. Therefore, the assessment had to be made on the basis of the broadest technical teaching of the application as filed.See points 11.8 and 11.9 of the reasons. For criterion (i) above, this meant that the technical effect relied on did not have to be disclosed literally.See point 11.10 of the reasons. For criterion (ii) above, this meant that there was no requirement for the application as filed to contain experimental data.See point 11.12 of the reasons, referring to points 93 and 60 of the reasons of G 2/21; see also point 11.14 of the reasons. Moreover, for criterion (ii) above, the test would be whether the skilled person would have legitimate reason to doubt that the purported technical effect could be achieved with the claimed subject-matter.See points 11.11 and 11.14 of the reasons.

Further interpretation

A further decision that provided a general interpretation regarding a new technical effect was T 2465/19, although it did not deal with the submission of additional data. Similarly to T 116/18, the Board quoted the second answer of decision G 2/21 to the effect that the technical effect relied on did not have to be explicitly stated in the application. It only had to be derivable from the original application.See point 5.3.1 of the reasons of T 2465/19.

In this respect, therefore, both decisions read G 2/21 as applying a reality-based approach, which is consistent with the Board’s observation in pre-G 2/21 decision T 2371/13 that it would be common to rely on a technical effect that was not explicitly mentioned in the application as filed.See point 6.1.2 of the reasons of T 2371/13.

Irrespective of whether the Enlarged Board intended its second criterion to have this meaning, the following overview shows that all Boards seem to have adopted this approach. It is only at first glance that one of the Boards appears to be out of line.See decision T 258/21. A closer look at the circumstances tells a different story, see below.

The Boards on post-filing data

In addition to T 116/18, a total of 12 post-G 2/21 decisions were issued in 2023, in which post-filing data had been submitted to be used for the assessment of inventive step. In two decisions, T 2911/19 and T 573/21, such data were not admitted as late filed in the absence of exceptional circumstances. These two decisions will not be discussed further here. Of the remaining 10 decisions, 8 cited the Enlarged Board in G 2/21. In three of these 10 decisions, the Board did not consider post-filing data. As explained below, one of these three cases appears to involve sufficiency aspects, and is therefore best left out of the overall picture. In one of the 10 decisions, T 364/20, the patentee relied on post-filing data for more than one technical effect. The Board took an interesting approach: the post-filing data were taken into account only to the extent that the purported technical effect was found to be derivable from the application as filed.

The remaining 6 decisions where post-filing data were successfully filed will be briefly discussed first.

Decisions where post-filing data were taken into account

In T 873/21, claim 1 related to a composition comprising two active ingredients for the treatment of a metabolic disorder in a horse. The application as filed disclosed an improved effect compared to monotherapy with one of the two active ingredients.See point 3.3.2 of the reasons of T 873/21. Post-filing data showing a synergistic effect were taken into account by the Board. The effect was regarded as derivable from the original application. As for the above criterion (ii), the effect was embodied by an originally disclosed combination, as this was the preferred combination in the original application.See points 3.3.2 and 3.3.3 of the reasons of T 2735/19.

Claim 1 as assessed in T 2735/19 defined cancer treatment in the form of a second medical use claim, limited to a dosage regimen. While the description as filed referred to a “cancer therapeutic drug”, data in the application as filed only showed an effect on breast cancer. Post-filing data showed an effect on further cancer types. The Board concurred with the patentee that, based on the common general knowledge of the mode of action of the active ingredient, the skilled person would expect a general effect on tumours, and not just on breast cancer.See points 5.2 and 7.2.3 of the reasons of T 2735/19. Criteria (i) and (ii) of G 2/21 were therefore found to be met, and the data were considered.See point 5.2 of the reasons of T 2735/19. The purported technical effect was regarded as credible for the scope of “cancer”.See point 7.2.3 of the reasons of T 2735/19.

T 885/21 concerned a claim 1 that defined a first medical use of an antibody-cytotoxin conjugate. The description disclosed an advantageous homogenicity in terms of site specificity and stoichiometry, as well as improved stability. Pharmacological effects were shown in examples of the patent as filed. Post-filing data related to features identified as different from the prior art. They showed improved homogenicity and pharmacological effects as disclosed in the application. The effect shown by these data was thus regarded as encompassed by the technical teaching and embodied by the disclosed invention.See point 6.4.1 of the reasons of T 885/21.

In T 1329/21, claim 1 related to a cosmetic formulation comprising, inter alia, cellulose particles of a certain size. The application as filed had given a reduced stickiness of formulations as a purpose of the invention. Inter alia an improved absorption capacity and better spreadability were described. Post-filing data in the form of a table showed, under the item “sensory properties”, evaluations of properties, such as oiliness and absorption. The original application had provided similar evaluations. The Board found that the technical effect shown was clearly encompassed by the technical teaching of the original application.See point 3.2.4 of the reasons of T 1329/21.

Claim 1 at issue in T 728/21 defined a medical use of a tablet formulation for cystic fibrosis in certain patients. Post-filing data included a diagram of the dissolution properties of the components of the formulation, from which it could be concluded that the tablet formulation as claimed had an improved dissolution compared to the prior art. The original application had already in the summary section several times named dissolution values and contained a figure showing dissolution profiles. The Board therefore found that the application as filed “addressed dissolution of tablets … as an aspect of the disclosed invention … and specifically described the claimed tablet composition as an embodiment of the disclosed invention”.See point 2.2 of the reasons of T 728/21. The data could therefore be considered.

T 591/21 concerned a dishwashing composition. Data in the application as filed showed better cleaning properties compared to the closest prior art. The Opposition Division had, however, questioned that the conditions used for obtaining the data were comparable.See point 3.3.1 of the reasons of T 591/21. In response, the patentee had filed new data. Since the technical effect was unchanged, the conditions of G 2/21 were apparently not an issue, and the decision does not even cite G 2/21.

Noteworthy, in none of these 6 decisions did a Board express doubts about the plausibility of the technical effect for which post-filing data had been submitted. Any general conclusion based on these 6 decisions is therefore somewhat limited.

Decisions where post-filing data were not taken into account

The decision referred to above, in which an effect shown in post-filing data was only partially taken into account, is T 364/20. Claim 1 at issue related to a composition comprising different paraffins. The patentee had included two tables of data into a letter filed during opposition proceedings. These tables apparently resembled tables included in the application as filed, and related to further compositions falling within the claimed subject matter. The property to be relied upon was the presence of less volatile organic compounds, thereby inter alia imposing a reduced health impact. The Board held that the presence of less volatile organic compounds was not a technical effect, but that the purported effect on health was. However, there was no mention of a positive impact on health in the application as filed.See point 11.5.3 of the reasons of T 364/20. Although the Board did study the post-filing data, this technical effect was not taken into account.

Although T 364/20 does not mention decision G 2/21, the Board de facto found that criterion (i) established by the Enlarged Board was not met. While the outcome was thus not in favour of the patentee, the approach taken appears to seamlessly join the approach of the above decisions.

The decisions in which the Boards did not take post-filing data into account at all are T 887/21, T 258/21 and T 681/21.

T 887/21 concerned a second medical use of a nutritional composition. Claim 1 defined the use as the prevention of secondary infections following a viral infection characterised by a certain enzyme activity. It may be helpful to know that the second auxiliary request limited the secondary infections to infections of the respiratory tract by pathogenic bacteria. The background section of the application as filed began with the words “Infections of the respiratory tract”, and the end of the summary section referred to influenza as the viral infection. The Board found that there was one single instance where the application as filed named infections of the gastrointestinal tract as a secondary infection. For the sake of fairness, it should be added that this paragraph seems to be the definition of the secondary infections to be prevented. No data were provided in the application as filed. The patentee filed post-filing data, in which a strain of Salmonella typhimurium was used at various concentrations, and referred to the generic claim wording. The Board found that the cells used in these data were not the cells on which the application had based a concept of action. Most of all, Salmonella typhimurium is known to cause gastroenteritis, and was not mentioned in the application as filed. The post-filing data would therefore provide entirely new information.See point 2.15.3 of the reasons of T 887/21. The skilled person would thus not find the technical effect of the post-filing data “as being encompassed by the technical teaching and embodied by the same originally disclosed invention”.See point 2.15.4 of the reasons of T 887/21.

After all, decision T 887/21 therefore appears to be largely in line with the above decisions that were in favour of the applicant/proprietor. The negative outcome in this case was based on the interpretation of the scope of the invention vis-à-vis the disclosure of the original application as a whole, and the conclusion that the effect shown by the post-filing data was beyond the underlying technical teaching.

Claim 1 at issue in T 258/21 again related to a second medical use, namely reducing ischemic stroke damage. The application as filed did not provide any data or scientific explanation of the medical use. In essence, the information provided seemed to be a statement at the beginning of the section ‘Detailed Description’: “The present invention is based on the discovery that … is effective in reducing stroke damage …”. This was followed by a paragraph with explanations on stroke, and a subsequent paragraph provided embodiments of the cause of the stroke. Two post-published abstracts were submitted as annexes, purporting to demonstrate improvements in the treatment over two cited documents, including a higher activity and reduced side effects.See point 1.3.2 of the reasons of T 258/21. The description of the application as filed named the ability to rapidly reduce blood pressure, as well as safety, including the absence of side effects, as characteristics of the active ingredient used.See page 9, second paragraph, of PCT application WO 2012/135617. However, according to the Board, improved activity and reduced side-effects were “neither contemplated nor even suggested in the original application”.27 Nevertheless, the Board then provided an assessment of the content of the annexes, beginning with the words “Moreover, even if said technical effect would have been derivable from the original application …”.

At first glance, the approach taken in T 258/21 appears to be at odds with the approach taken in T 728/21 or T 1329/21. Compared to these other two decisions, the original application in T 258/21 may have had very limited disclosure relating to the technical effect subsequently relied on. However, the purported technical effect does not seem to be a completely new effect, as in T 887/21.

What may have played an important role in T 258/21 is, firstly, that the claim defined a second medical use. Secondly, the technical effect discussed in terms of inventive step appears to be identical to the definition of the medical use in the claim at issue. Thus, it appears that the Board was actually confronted with a sufficiency issue.See the frequently cited distinction under point 2.5.2 of the reasons of G 1/03. See also point 74 of the reasons of G 2/21. As the Enlarged Board held in G 2/21See point 74 of the reasons of G 2/21., “it is necessary that the patent at the date of its filing renders it credible that the … therapeutic agent … is suitable for the claimed therapeutic application.” However, in T 258/21 the application at issue did not contain any data. This aspect of sufficiency seems to take decision T 258/21 out of a pure inventive step assessment. Therefore, it seems advisable to exclude this decision from a comparison that is intended to be limited to circumstances where a purported technical effect is relied upon when defining the problem to be solved.

Finally, T 681/21 concerned a laundry detergent; claim 1 defined a fabric treatment composition which differed from the closest prior art in that it contained a specific cationic polymer. The patentee argued in favour of improved softness, in particular a synergy resulting from the presence of the polymer and another component of the composition, silicone.See point 1.2.1 of the reasons of T 681/21. The question arose as to whether this effect could be relied upon. A document in the form of a technical report was available as post-filing data. The first paragraph of the application as filed already named improved softening, and two paragraphs later improved softening was named as the problem to be solved. The paragraph between these passages acknowledged that silicone is conventionally added to provide softness to fabrics. In one further instance the description referred to “fabric softening silicone”. The Board concluded that the formulation of the technical effect relied on by the patentee was different from that identified in the patent.See point 1.2.2 of the reasons of T 681/21. A synergistic effect of the cationic polymer and silicone was neither disclosed nor suggested in the original application. The technical effect would therefore not have been considered by the skilled person as being encompassed by the technical teaching of the application as filed.See point 1.2.3 of the reasons of G 2/21.

Again, at first glance, the approach taken in T 681/21 does not seem consistent with that taken in T 728/21 or T 1329/21. It is also striking that in T 873/21 it was found that a synergistic effect was derivable from the application as filed, whereas this was not the case in T 681/21.

However, two aspects of the case of T 681/21 should be noted: (a) silicone is obviously a commonly used softener, and (b) the skilled person apparently also knows the cationic polymer as a softener.See point 1.2.4 of the reasons of G 2/21. The technical effect therefore boils down to the alleged synergistic effect, of which, indeed, there does not appear to be any suggestion in the application as filed. In the application underlying T 873/21 statements such as the following taught an improvement when using a combination: “The combination therapy … advantageously leads to improved insulin sensitivity where monotherapy with one or more dopamine receptor agonist is insufficient”.See the third paragraph on page 4 of WO 2016/046150. In contrast, the application underlying T 681/21 taught: “We have now found that if instead of addition as part of the laundry detergent, the silicone is provided as part of a separate composition, then the softening performance is improved”.See paragraph [0006] of EP 3 221 438. What is and what is not a new effect ultimately depends on the circumstances of each case.

Conclusion

The EBA has distilled the real problem on the basis of the keywords and additional questions provided to it. It has provided the Boards and applicants with two criteria as an answer. On the whole, an almost homogeneous approach can be deduced from the above 11 decisions. As soon as post-filing data did more than confirm data in the application, G 2/21 was consulted and at least the Enlarged Board’s criterion (i) – is the effect encompassed by the technical teaching of the original application – was assessed through the eyes of the skilled person. The Boards have breathed life into this criterion and appear to have applied it consistently. In general, the first criterion seems to be sufficient to decide whether a technical effect shown by post-filing data can be used for the assessment of inventive step or not.

As regards criterion (ii) – is the effect embodied by the same originally disclosed invention – it appears that the Boards may not have a uniform interpretation. T 116/18 even seems to provide a through ball for a further referral to the Enlarged Board. In two decisions post-filing data failed condition (i) of G 2/21 because patentees tried to rely on a completely new technical effect.

Which case has now prevailed, T 939/92 (AgrEvo) or T 488/16? In other words, is it possible to overcome the “we don’t buy it” objection with post-filing data? Firstly, it seems that G 2/21 has led to a more precise wording: instead of a problem being solved, all decisions assess the technical effect. Furthermore, the bar is obviously set too high in T 488/16 by requiring that “the asserted activity has been made sufficiently plausible” by the application as filed.See point 4.2 of the reasons of T 488/16. If this were the standard, even in T 591/21 post-filing data could not have been submitted to address the Opposition Division’s doubts about comparability of conditions with those of a cited document. Consequently, neither T 939/92 nor T 488/16 appear to be fully in line with the G 2/21 standard. As acknowledged in T 116/18See point 11.3.2 of the reasons of T 116/18., “any rationale developed in the previous plausibility case law” is no longer decisive.

According to decision T 116/18, a “we don’t buy it” objection can now be raised under the above criterion (ii). Be that as it may, situations where the person skilled in the art does not consider a purported effect for part of the claimed subject matter to be credible in the light of the application as filed will not disappearSee the situation of T 258/21, which can be viewed in the same light.. It is fairly safe to say that this situation directly translates into criterion (i) of G 2/21: the effect would not be encompassed by the technical teaching of the original application for that subject matter. The result would therefore in all likelihood be the same as in T 488/16, and not as in T 939/92.

Comments