Novelty, added subject-matter, and the ‘gold standard’

P. de Lange (NL)

1. Introduction

According to the landmark decision T 201/83, r.3, the rule for examining amended claims under Art. 123(2) is the same as for novelty, namely whether the document at issue (the application as filed or the prior art document) provides ‘sufficient information so that the person skilled in the art could derive the subject-matter in question from it directly and unambiguously, including any features implicit therein.’^{All references to legal provisions are to the European Patent Convention (EPC).} The Board reasoned that ‘the test for compliance with Art. 123(2) is basically a novelty test.’^{Guidelines C-IV, 7.2 (first edition 1978) provided that ‘a document takes away the novelty of any claimed subject-matter derivable directly and unambiguously from that document’ (currently in G-VI, 2)./sup> The phrase ‘directly and unambiguously’ as used in T 201/83 still applies and is nowadays part of the ‘gold standard’ (G 2/10, r.4.3): the subject-matter disclosed by a document (implicitly or explicitly) is the subject-matter that ‘a skilled person would derive directly and unambiguously, using common general knowledge, and seen objectively and relative to the [relevant date], from the whole of [the document].’Case Law of the Boards of Appeal, 9th edition (CLBA), II.E.1.3.1.}

Although G 2/10 firmly established the ‘gold standard’, the precise relation between the tests for novelty and added subject-matter remains unclear and the case law of the Boards appears somewhat inconsistent. For instance, according to T 1525/15, the ‘novelty test’ is no longer used for Art. 123(2). In contrast, T 906/12 held that ‘the test for compliance with Article 123(2) is essentially a novelty test’ and this ‘novelty test’ is still discussed in the Guidelines (H-V 3.2; 2019) for amendments which involve adding features. Indeed, G 2/10 confirmed that an amendment ‘may not create novel subject-matter’ (r.4.6).

In this article, I aim to shed some light on the relation between the two tests. I start from the principle that the European patent system uses the ‘gold standard’ as a ‘uniform concept of disclosure’ which is (by definition) exactly the same for Art. 54 and 123(2) (G 2/10, r.4.6; G 1/03, r.2.2.2). I propose a framework wherein each test consists of three steps, with the ‘gold standard’ forming the second step of both tests. The first step (which is also identical for both tests) is claim interpretation. The third step involves comparing the disclosed subject-matter and the claimed subject-matter. Importantly, the precise manner of comparing differs between Art. 54 and Art. 123(2). The third step entails assessing whether the disclosed subject-matter falls within the claimed subject-matter for novelty, whereas the question is whether the application as filed discloses subject-matter that is identical to the claimed subject-matter (neither narrower nor broader) for Art. 123(2). I will discuss the three steps in more detail hereinafter.

Such a three-step framework has been suggested earlier by Keukenschrijver for novelty.^{Keukenschrijver in Busse/Keukenschrijver, PatG, 8th ed., §3 note 83 (7th ed.: §3 note 78). See also Melullis in Benkard, EPÜ, 3rd ed., Art. 54 note 23; Lindner in Singer/Stauder/Luginbühl, EPÜ, 8th ed., Art. 54 note 83 and 85 in fine.} Here I propose that the framework is also useful for Art. 123(2). In particular, the ‘gold standard’ becomes more clearly an independent concept by using the three-step framework, as it is only one of the three steps of the Art. 123(2) test. In this way, the differences and similarities with the novelty test can be more easily accounted for.

The framework can be schematically illustrated as follows (Table 1):

Table 1: Three-step framework for novelty and added subject-matter

2. Overview of the three-step framework

The first step of the three-step framework is interpreting the claim under examination.^{I refrain from discussing the case wherein the test under Art. 123(2) is applied to amendments of the description.} In this way, the subject-matter defined by the claim is abstracted from the precise claim wording, such that the subject-matter is not affected by a mere rephrasing of the claim, e.g. by using synonyms. The basic rule to be applied in the first step is that expressions in claims are given their broadest technically sensible meaning in EPO proceedings.^{E.g. T 79/96, r.2.1.3; CLBA I.C.4.1.} On the other hand, the EPO generally does not assess the scope of protection of claims.^{T 223/05, r.3.5: ‘the extent of the protection of a patent is examined by the EPO in the opposition proceedings only within the framework of Article 123(3) EPC.’} Hence, the EPO does not interpret claims as embracing equivalent embodiments in the claimed subject-matter. Accordingly, for the assessment of novelty, it is not relevant whether a prior art document discloses something which falls under the scope of protection of the claim by way of equivalency.^{This rule about the interpretation of the claim under examination is distinct from the rule that ‘the disclosure of a prior document does not include equivalents of the features which are explicitly or implicitly disclosed’ (CLBA I.C.4.5) which pertains to the second step.}

The second step involves identifying the subject-matter that is disclosed by the source document (a prior art document^{Public prior use (G 1/92) is not considered in this article.} or the application as filed). The disclosed subject-matter is hence abstracted from the precise wording of the source document, similarly as in the first step. In this step, the subject-matter that is disclosed by the document is identified by applying the specific test of G 2/10, namely the ‘gold standard’ test: a document discloses subject-matter (implicitly or explicitly) if the skilled person directly and unambiguously derives that subject-matter from the document, using common general knowledge and seen objectively and relative to the relevant date.^{See G 1/16, r.51 describing the gold standard as the ‘relevant disclosure test’ (see also CLBA II.E.1.3.1). In principle, the Boards could also have set a different threshold for disclosure, such as ‘fairly based’ or ‘expressly, implicitly, or inherently supported’. See also S. Adams, Evaluation of Claim Amendments, epi Information 3/2016.} In other words, “the disclosure content of [a document] is (…) the information that the skilled person derives - explicitly or implicitly - directly and unambiguously from the document as a whole” (T 2003/08, r.2).

The first and second step can also be carried out in reverse order. These steps together provide that a mere difference in wording between a claim and a prior art document is insufficient to establish novelty (T  114/86 hn.2), similarly a mere rephrasing is not problematic under Article 123(2) (e.g. T 119/91).

The third step involves a comparison of the claimed subject-matter and the disclosed subject-matter (these having been established in the first and second step). The precise manner of comparing differs between the tests for novelty and added subject-matter.

To be novelty destroying, the subject-matter disclosed in the prior art document must be something ‘falling within’ the ambit of the claimed subject-matter (e.g. T 890/17, r.1.2).^{See also T 508/91, r.2.4; T 793/93, r.2.1; T 137/15, r.6.1; T 308/17, r.14; T 641/99 r.4; and further Guidelines G-VII, 4 (about Art. 56).} The phrase ‘falling within’ reflects the rule that a specific disclosure takes away the novelty of a generic claim embracing that disclosure (Guidelines G-VI, 5). It also reflects the principle that all embodiments within the claims must be examined for novelty in order to exclude from patenting subject-matter which is not novel (G 1/98, r.3.3).

For assessing compliance with Art. 123(2), the application as filed must disclose subject-matter that is identical to the claimed subject-matter, i.e. neither broader nor narrower (T 201/83, r.2).^{The phrases that amended claims must have ‘basis in the application as filed’, may not involve ‘added subject-matter’ and ‘may not extend beyond the content of the application as filed’ can be used to refer to the entire three-step test of Art. 123(2); the terms ‘disclosed’ and ‘gold standard’ on the other hand can be understood as referring specifically to the second step of the proposed three-step framework.} The test is hence met neither by a more general disclosure nor by a more specific disclosure. This ‘identical’ test may seem trivial, but expressly discussing it as the third step of the Art. 123(2) test appears to be useful in order to allow for a precise comparison with the third step of the Art. 54 test. For instance, as a result of the differences between the two tests in their third step, an (intermediate) generalization does not provide for novelty but may still cause an amended claim to involve added subject-matter^{This is the main reason why the ‘novelty test’ is no longer used, see CLBA II.E.1.3.7 and T 194/84).}.

I note that a prior art document must not merely disclose the subject-matter in order to be novelty destroying, but must do so in an enabling way. I will discuss this in more detail hereinafter in §Enablement: novelty and priority.

I will first apply the three-step framework to a few topics that frequently turn up in the case law of the Boards, namely inherent features, selection inventions, and disclaimers.

3. Inherent features

Features which are inherent in a prior art disclosure are relevant for novelty, because it is established case law that the ‘falling within’ test for lack of novelty can be satisfied based on inherent features of the prior art subject-matter.^{T 12/81 hn.1, see e.g. also T 680/00, r.3. The exceptions are, of course, the special rules for first and second medical uses of Art. 54(4) and (5) EPC 2000, and the special rule for second non-medical use claims of G 2/88; however, these exceptions could also be seen as special rules for claim interpretation (cf. T 308/17, r.7). The remark in G 2/88, r.10.1 that ‘the question of “inherency” does not arise as such under Article 54 EPC’ is probably restricted to use claims as considered therein.} Such inherent features are, however, not relevant under Art. 123(2). For example, if a prior art document D1 provides an enabling disclosure of a compound X and a method of preparing it, and this compound X inherently has a property Y, then a claim directed to ‘a compound with property Y’ is not novel (even if document D1 is silent about property Y and even if the compound X could not be predicted to have the property Y). On the other hand, the fact that compound X has property Y is technical information and may therefore not be added to a patent application by way of amendment (G 2/10, r.4.5.1). Hence, a claim directed to ‘a substance with property Y, wherein the substance is compound X’ likely lacks basis under Art. 123(2) if the application as filed merely discloses compound X without giving any information (neither implicitly nor explicitly) about whether it has property Y.^{See e.g. T 1487/16, r.1.4.2, last paragraph. The question under Art. 54(3) was whether Example II of D9 benefited from the priority date of D9a. Example II of D9a gave specific values for two properties of the material prepared therein (BET value of 116 and Sears number of 22). Example II of D9 included a sentence mentioning that the material had a ratio Sears/BET of 0.19, this sentence was added compared to D9a. The Board found that this added information about the ratio made the priority invalid because the corresponding amendment would contravene Art. 123(2) (which is the relevant point for the present article), such that Example II of D9 could not be cited under Art. 54(3).}

4. Selection inventions: selecting from lists and from ranges

The case law about selection inventions in fact deals with two topics: selections of items from lists and claims specifying ranges which are sub-ranges of broad ranges in the prior art.

According to established case law, novelty of a claimed sub-range over prior art teaching a broader range requires that the sub-range is both narrow compared to the broad range and sufficiently far removed from the end-points of the broad range (Guidelines G-VI, 8.ii and T 279/89). This case law logically pertains to the third step since it is a special manner of comparing the claimed subject-matter and the subject-matter disclosed in the prior art.^{The same applies to the (possibly still valid) third requirement that the narrow range must involve a purposive selection in order to be novel (T 279/89).} Because this rule pertains to the third step, the factors of ‘narrowness’ and ‘sufficiently far removed’ do not pertain to the second step and hence are irrelevant for the question whether or not the prior art document discloses the narrow range in the sense of G 2/10. Accordingly, there is no need for arguing that a prior art document teaching a broad range somehow also discloses the (non-novel) narrow range in order for T 279/89 to be consistent with G 2/10, let alone for arguing that the skilled person would derive the narrow range (directly and) unambiguously merely from the mentioned broad range. In other words, even though a claimed range of 2-99 is not novel over a prior art document mentioning a broad range of 1-100, an amended claim specifying the same range of 2-99 still lacks basis under Art. 123(2) if the application as filed only mentions a range of 1-100.

On the other hand, the rule that a selection from two or more lists basically provides for novelty of a claim (T 12/81, r.13, Guidelines G-VI, 8.i) belongs to the ‘disclosure’ question (i.e. the second step) and is (hence) applied equally in the context of Art. 123(2).^{See e.g. CLBA II.E.1.6.2 and T 1621/16. The title ‘Novelty’ of the chapter G-VI in the Guidelines where this rule can be found therefore does not fully reflect the scope of the rule.}

5. Undisclosed and disclosed disclaimers

Similarly, case law concerning undisclosed disclaimers (e.g. disclaimers added to restore novelty over Art. 54(3) prior rights; G 2/03) can be brought under the third step of the Art. 123(2) test. In other words, the third step for Art. 123(2) involves not only the ‘identical test’, but also an additional rule that undisclosed disclaimers are disregarded in the third step of the Art. 123(2) test if they are allowable under G 2/03.^{The undisclosed disclaimer is taken into account for novelty, as it is used precisely to make the claim novel. Because of this different treatment of the disclaimer under Art. 54 and Art. 123(2), the rules about undisclosed disclaimers pertain to the third step of the three-step framework.} On the other hand, in order to determine whether a claim amended by the introduction of a ‘disclosed disclaimer’^{A ‘disclosed disclaimer’ is a disclaimer that excludes embodiments from the claim that are disclosed in the application as filed (see G 2/10, hn.1a). G 2/10 clarified that a claim amended with such a disclaimer may still infringe Art. 123(2) if ‘the subject-matter remaining in the claim after the introduction of the disclaimer is not […] directly and unambiguously disclosed […] in the application as filed’ (hn.1a). In the presently proposed framework, the phrase ‘would derive directly and unambiguously’ is used as the threshold for whether subject-matter is disclosed (cf. G 3/89, hn.1).} has basis in the application as filed, the ‘gold standard’ must be applied to ‘the subject-matter remaining in the claim’ (G 2/10, hn.1a; G 1/16, r.51). Accordingly, G 2/10 pertains to the second step. G 2/03 (about undisclosed disclaimers) and G 2/10 (about disclosed disclaimers) accordingly pertain to different steps of the three-step framework (namely, respectively the third and second step) and there is hence no contradiction between these two decisions (cf. G 1/16).

6. Enablement: novelty and priority

The third step can also be used to accommodate the rule that for lack of novelty, the prior art document must not merely provide a disclosure of something falling within the claimed subject-matter, but must do so in an enabling way (T 206/83, r.12). The case law under Art. 123(2) does not appear to require that the application as filed provides an enabling disclosure for subject-matter to have basis of the application as filed. The additional enablement requirement for lack of novelty hence appears to pertain to the third step of the three-step framework.

Similarly as for lack of novelty, validly claiming priority requires that the priority document provides an enabling disclosure (at the priority date; CLBA II.D.3.1.6). Hence, if the three-step framework is used for the requirement of ‘the same invention’ for priority (Art. 87), the third step is in principle the same as the ‘identical’ test for Art. 123(2), in view of G 2/98, but with the additional requirement that the priority document discloses the subject-matter in an enabling way.

7. Inventive step

Finally, the three-step framework can also be used for the ‘novelty analysis’ that is comprised in the problem-solution approach (PSA) for inventive step (Art. 56). The fact that an inventive step attack involves a novelty analysis might be familiar to practitioners but is not described in the Guidelines (G-VII, 5.2). A brief explanation therefore seems useful. A novelty analysis is part of the PSA, because the PSA involves three steps after the initial stage of identifying the closest prior art (CPA) document (or starting point document): 1) identifying the distinguishing features of the claim over the CPA document, 2) formulating the objective technical problem solved by those features in view of the CPA document and 3) assessing whether the distinguishing features are obvious as a solution to that objective technical problem (Guidelines G-VII, 5.2). The first step of identifying the distinguishing features normally involves showing that the other features of the claim do not provide for novelty in view of the CPA document. In other words, it is shown that the claim under examination modified by omitting the distinguishing features is not novel over the CPA document, i.e. a novelty analysis for a modified claim. This novelty analysis under Art. 56 should be done in exactly the same way as a normal novelty attack.^{N. Blokhuis and C. Mulder, ‘Smart in C’, 3rd ed. 2019, Helze B.V, p.68, §13.5. See e.g. also T 973/15, r.3.1, last para; T 1698/07 r.2.1.} The proposed three-step framework is therefore also relevant for Art. 56.

Comments