Who bears the burden to show that an objective technical problem has been credibly solved?

S. Kutik (IT), J. Renken (DE), German and European Patent Attorneys

Under the EPO’s problem solution approach, the subject matter defined in a patent claim satisfies the requirements of inventive step (Art. 56 EPC) when an objective technical problem vis-à-vis the closest prior art is credibly solved in a non-obvious manner. Debates whether an application contains enough data to show that the problem has been credibly solved over the entire claim scope are particularly prevalent in the life sciences. In such cases, concepts such as the “burden of proof” and “benefit of the doubt” are important. Who needs to prove “what”, and “when” are, therefore, the topics of the present article.


It is common practice at the European Patent Office (EPO) to apply the problem solution approach when determining whether the subject matter of a patent claim, starting from the closest prior art, involves an inventive step. According to this approach, it must first be “credible” (e.g. T 939/92, OJ 1996, 309) that the claimed invention solves a proposed technical problem over the prior art; and second, the claimed solution to that problem must be non-obvious over the prior art.

Whether or not a technical problem has been credibly solved can become a point of debate both in examination proceedings and in post-grant oppositions. Such a debate is a frequent occurrence particularly in the life sciences. One reason is that clinical trials involve the mandatory publication of the details of a trial at an early stage of clinical development (Kallenbach and Vallazza, epi information 04/2017, p. 36-43). This commonly results in the filing of applications before the first clinical trials begin. By filing so early, the Applicant ensures that the publication of the details of the clinical trial will not anticipate the patent application. However, at such an early stage of development, there may not be enough evidence of a plausible therapeutic effect.

As to the quality of the evidence required, “absolute proof” that the effect is achieved is not required. The effect merely has to be rendered plausible (T 716/08). However, because the “EPC requires no experimental proof for patentability” (T 578/06), in some cases the question whether it is “credible” or “plausible” that a technical problem has been solved will need to be answered in the absence of data. But, whose burden is it to show whether an effect is, or is not, plausible?

The equivalent test under Art 83 EPC

In contrast to the assessment of inventive step, the question of burden of proof is regularly addressed in the assessment of sufficiency of disclosure. For sufficiency of disclosure, the default position is that the burden to demonstrate a lack of sufficiency lies with the Opponent or Examiner who must substantiate any doubts with verifiable facts (T 19/90, T 694/92). However, in cases where a patent does not give any information on how a feature of the invention can be put into practice, the Opponent can discharge his burden by arguing that even common general knowledge would not enable the skilled person to put this feature into practice. In such a case, the burden of proof is shifted to the Patent Proprietor who then must show that common general knowledge would indeed enable the skilled person to carry out the invention (T 63/06).

Both the assessment of inventive step and the assessment of sufficiency of disclosure shall avoid the grant of purely speculative patent applications. In other words, the purpose is to avoid the grant of a monopoly to an Applicant who has left the technical field unexplored (T 1188/00; T 1329/04, T 1164/11).

Moreover, whether the presence of a technical effect is assessed under inventive step or sufficiency of disclosure is merely a question of claim drafting. If an effect is recited in the claim, this is an issue falling under Art. 83 EPC. If not, it is one falling under Art. 56 EPC (G 1/03).

Thus, it could be justified to apply the established rules regarding the burden of proof in disputes concerning sufficiency of disclosure to disputes concerning the solution of the technical problem under Art. 56. But are these rules applied at the EPO? And, is it always justifiable to apply the same rules?

Benefit of the doubt vs. substantiated doubts

By default, a technical problem set out in a patent, i.e. the original technical problem proposed by the Applicant, is considered to be credibly solved by a claimed invention “if there exists no reasons to assume the contrary” (T 1797/09); and “the benefit of doubt is given to the Patent Proprietor” (T 1797/09). This benefit is not limited to granted patents, but has been found to also be applicable to pending patent applications undergoing examination (T 578/06). Thus, in summary, when doubts remain whether the proposed technical problem has been solved, the Applicant or Patentee may be given the benefit of the doubt.

In T 602/05 the Board of Appeal found it justifiable to apply the same criteria commonly applied for assessing sufficiency of disclosure as per T 19/90 for assessing whether the objective technical problem had been solved, i.e., that doubts have to be substantiated by verifiable facts in order to establish that the technical problem has not been solved. Also in T 1707/06, the Board required the Opponent to present “convincing evidence that an effect is not obtained“; and in T 578/06, the Board found, with regard to the dispute on whether or not the objective technical problem had been solved, that “if an examining division raises an objection, it must appropriately be substantiated“. In all those cases, the Applicant or Patentee was given the benefit of the doubt. In the absence of any convincing evidence to the contrary, the technical problem was deemed to have been solved.

Reversal of the burden of proof

Although not frequently cited, T 1797/09 is of particular interest because it specifically discusses the reversal of the burden of proof when the solution of the technical problem under Art. 56 EPC is under debate. Although the Board of T 1797/09 generally acknowledged that a technical problem set out in a patent is deemed to be credibly solved in the absence of any reasons to the contrary and, therefore, the Proprietor can be given the benefit of the doubt, the Board also established that “if the Opponent succeeds to cast reasonable doubt on the alleged effect, the burden to prove its allegations is shifted to the Patent Proprietor”. It is therefore one of few decisions that explicitly deal with the reversal of the burden of proof, and with the condition required for the reversal.

In that regard, T 1797/09 appears to mirror T 63/06, which concerns the shift of the burden of proof when assessing sufficiency of disclosure.

In fact, in specific scenarios where there are (A) inventions which challenge the common general knowledge in an unsubstantiated manner and/or (B) broad claims, the burden is commonly shifted:

  1. if there is no evidence that the problem has been solved in the application as filed, and the knowledge available in the art renders it, prima facie, unlikely that the claimed invention indeed solves the problem it purports to solve, the Applicant or Patentee is not given the benefit of the doubt. For example, in T 1329/04 a specific polypeptide was claimed and the application as filed explicitly stated that said polypeptide provided, as a technical effect, a specific biological growth differentiation activity. But data supporting such an effect were not present in the patent application. Further, the claimed polypeptide lacked certain characteristic structural features that were otherwise common to known polypeptides with such growth differentiation activity. Thus, there existed doubts that the proposed technical problem had indeed been solved by the claimed subject matter. Although the claimed polypeptide was shown to indeed solve the originally proposed technical problem by means of post-filing data, the Board found “that there is not enough evidence in the application to make at least plausible that a solution was found to the problem which was purportedly solved” (emphasis added). In the Board’s view, the solution to the technical problem has to be made at least plausible by the application as filed;

  2. in the life sciences, it is not uncommon for a patent claim to cover, not only the specific entities (e.g. chemical or biological compounds) for which a technical effect is experimentally demonstrated in a patent application, but also a group of often structurally similar compounds for which no experimental validation is present in the patent application as filed.

    It is commonly accepted by the Boards of Appeal that, in the field of drug design, even “small structural modifications may cause major differences in biological activity” (T 939/92) so that in the absence of additional technical information, any structural modification of a compound is “expected to disturb the pharmaceutical activity of the initial structure” (T 548/91). Consequently, in cases where supporting evidence is limited to a few compounds, it is not considered technically plausible that all compounds under a broader product claim credibly solve the same technical problem as the ones tested. In such cases, the Applicant or Patentee is not normally given the benefit of the doubt. In the case discussed in the groundbreaking decision T 939/92, a group of chemical compounds was claimed of which only some were shown in the patent application to have herbicidal activity. The Board of Appeal considered that it was unreasonable to assume that essentially all the claimed compounds had such activity and thus solved the respective problem over the entire claim scope in the absence of further evidence. The Applicant then bore the burden to show the presence of the alleged technical effect across the entire claim scope. Equally, in opposition proceedings, when the credibility that a technical effect is achieved by substantially all claimed compounds is at issue and in a situation where it is, prima facie, unlikely that this is the case, it is not the Opponent, but the Patentee who has the burden of proving that the effect is indeed achieved over the claim scope (T 975/14).

    Can the Applicant or Patentee escape from such an objection by limiting the claim to embodiment(s) that do factually solve the problem? The patent underlying decision T 488/16 claimed a group of compounds that were allegedly protein tyrosine kinases (PTK) inhibitors. The Opponents, referring to T 939/92, argued that it was not plausible that all the claimed compounds had the desired inhibitory activity. The Patentee then restricted the claims to a specific compound, dasatinib, for which the desired effect was clearly established by means of post-published evidence. The Board nevertheless held that the Opponents had provided “technically sound and persuasive arguments as to why the alleged effect had not been made plausible” in the application as filed, and thus concluded that the burden to “counter-argue and to rebut these doubts” rests with the Patentee.

In these scenarios, the patent application per se does not show the presence of a technical effect or a technical effect over the entire claim scope. Further, there are arguments that the technical effect is (A) not plausible at all; or (B) not plausible over the entire claim scope. In both scenarios, the party raising the objection does not bear the burden to provide evidence. These two scenarios thus resemble the established reversal of the burden of proof under Art. 83 EPC in situations where the patent application itself does not demonstrate how a claim feature can be put into practice and the party raising the objection plausibly argues that common general knowledge would not enable the skilled person to put this feature into practice (see T 63/06).

In case of such a shift of the burden of proof, it thus rests with the Patentee to rebut doubts as to the effect. But the Patentee’s possibilities to do so are more limited than in the assessment of sufficiency of disclosure: while a Patentee could rely on common general knowledge to defend his case under Art. 83, this would not help him under Art. 56 because common general knowledge – while possibly supporting the plausibility of the effect – would, at the same time, render the effect obvious.

Patentees have, therefore, been relying on post-published evidence. However, both in T 1329/04 and in T 488/16, the Boards held that such post-published evidence may not serve as the sole basis for establishing the presence of the effect. Post-published evidence is considered only if the effect is at least plausible from the application text. Considering that Patentee’s possibilities of rebutting doubts are very limited when assessing plausibility of the effect, this approach appears unbalanced. While the Patentee in T 488/16 requested a referral to the Enlarged Board of Appeal regarding the assessment of the plausibility of a claimed solution to solve a technical problem under Art. 56 EPC and the burden of proof in such scenarios, the request was not granted. From the perspective of legal certainty, practitioners would have welcomed a referral to the Enlarged Board of Appeal (see Exner and Hüttermann, GRUR Int., 2018/2, p. 97-102).

The new closest prior art

Since the technical problem in the analysis of inventive step is defined in an objective manner, the advantage actually achieved (if any) over the particular prior art document which is closest to the claimed invention is decisive (T 815/16). If during proceedings before the EPO, a prior art document is cited by an Examiner or Opponent that comes closer than the prior art originally proposed as “closest” in the patent application, case law requires that the effect supporting the inventive step is made credible in comparison to this new closest prior art. The identification of a new “closest prior art” in proceedings before the EPO can put an Applicant or Patentee into a difficult situation, because comparative data regarding the solution of a technical problem in comparison to that newly identified document are, frequently, not present in a patent application as filed.

In such a case, Examiners and Opponents usually argue that, in the absence of comparative data with this new closest prior art, the objective problem should be considered as the provision of a mere alternative to the new closest prior art. Considering the principles of e.g. T 602/05, one could, however, take the position that it is the Examiner’s or Opponent’s burden to provide “verifiable facts” in order to make the non-existence of the alleged effect credible, and that the benefit of the doubt, in absence of verifiable facts, should be given to the Applicant or Patentee.

Reformulation of the technical problem

If a newly identified closest prior art is found to also solve the problem underlying the invention, the objective technical problem needs to be reformulated. This reformulated problem is usually considered to be the provision of an alternative. However, the Patentee normally attempts to refute any allegation that the technical problem is the provision of a mere alternative solution because the skilled person is considered to always be motivated to find technical alternatives to the state of the art (e.g. T 284/00) under EPO practice. Thus, an alternative solution is rarely considered inventive.

In order to refute the allegation that the solution is the provision of an alternative, the Patentee will have to make credible that (a) the claimed invention solves another problem which is not solved by the closest prior art; and (b) this problem must be derivable from the application as filed (T 1188/00). Both conditions must be fulfilled.

In this case, the Patentee bears the burden of proof to demonstrate that this problem has been solved (T 1188/00). This is normally achievable when the application as filed contains comparative data which make the solution of the technical problem over the closest prior art plausible (e.g. T 815/16). In case the application as filed does not contain data to make credible that a reformulated problem was in fact solved, it has generally been accepted by the Examining and Opposition Divisions that new data could be relied upon to do so. However, it appears that, more recently, the Boards of Appeal sometimes require that the solution to the reformulated problem is already made plausible by the application as filed. For example, in T 1196/12, the Board concluded “that the effects relied on by the appellant in its formulation of the problem are not derivable from the application as filed”. The Board then referred to T 1329/04 to justify its decision that post-published evidence cannot be accepted to demonstrate the effect and thus denied the presence of an inventive step. Similarly, in T 1285/13, the Board, citing T1329/04, held that “the verification of whether or not the claimed solution actually solves the problem, i.e. whether the claimed subject-matter actually provides the desired effect, must be based on the data in the application”.

This new trend can put Applicants in an extremely difficult, and sometimes inescapable situation. The Applicant would need to react to newly cited prior art which he may not have been aware of at the drafting stage while also being deprived of the possibility of relying on other advantages over the prior art to establish an inventive step, even if such advantages are, in fact, achieved in comparison with that newly cited document. This appears to be at odds with the general principle according to which the Applicant is given the benefit of the doubt when it comes to providing a solution to the objective technical problem (T 578/06; T 1797/09; T 602/05). Moreover, this trend appears to ignore that T 1329/04 refers to the plausibility of the problem the application purports to solve, rather than the (possibly reformulated) objective technical problem.