European patent law considers a selection from the prior art as being novel if: a) the end points of a claimed sub-range or any point falling within the claimed sub-range are not disclosed in the prior art and b) the end-points of the claimed range are "sufficiently removed" from the individual elements disclosed in the prior art. For many years, the EPO Guidelines for Examination (the “guidelines”) stated an additional requirement that the selection must also be non-arbitrary or “purposive” (T 0198/84). Similarly, the guidelines state that a range overlapping with a range disclosed in prior art is only novel if the overlapping range would not be “seriously contemplated” by a skilled person. The “purposive selection” and "seriously contemplated" criteria may be particularly influential in cases of novelty-only prior art on the basis of which an inventive step attack cannot be raised. 

However, the “purposive selection” criteria for novelty has recently lost favour with some (but apparently not all) of the Boards of Appeal. In T 1233/05, for example, the Board of Appeal was of the opinion that the purposive selection requirement was more akin to an assessment of inventive step as opposed to novelty. The 2019 EPO guidelines were subsequently updated to remove the “purposive selection” criteria (G-VI-8, and G-VII-12) (IPKat: 2019 updates to the EPO Guidelines for Examination - the highlights). However, notably, the criteria that an overlapping range should not be “seriously contemplated” has not been similarly updated (G-VI-9 iii). As pointed out in a recent article by Roel van Woudenberg, this apparently puts the guidelines for assessing the novelty of sub-ranges at odds with the guidelines for assessing the novelty of overlapping sub-ranges. 

Interestingly, the UK courts have also previous expressed dissatisfaction with the EPO’s approach to the novelty of sub-ranges with regards to the purposive selection criteria. On the other hand, UK judges do seem willing to apply the criteria that a novel overlapping sub-range must not be “seriously contemplated” (IPKat: Court of Appeal grapples with novelty of overlapping ranges in Jushi v OCV).

According to the European Patent Institute (EPI), the law on the novelty of selection inventions is  unclear, given the diverging opinions of the Boards of Appeal. In view of this purported lack of clarity, the EPI president (Francis Leyder) wrote a letter to the President of the EPO requesting the President to refer the issue to the EBA. The President of EPI cited a number of recent Boards of Appeal decisions that did not apply the “purposive selection" criteria (including T 1233/05, T 1131/06, T 230/07, T 1130/09, T 2041/09, T 378/12 and T 1404/14) and a two recent decisions that did (T 66/12 and T 673/12). In the view of EPI “[i]t is clear from the above that there are clearly two incompatible lines of Board of Appeal decision as to what are the criteria to be used in deciding whether a sub-range is novel”. The President of EPI therefore suggested a referral from the EPO President to the EBA, that might ask the following question:

“What are the criteria to be used in assessing the novelty of a claim where the allegedly distinguishing feature of the claim relative to a prior art document is a sub-range of a broader range disclosed in that prior art document”.

EPO Director and lawyer Heli Pihlajamaaresponded to EPI on behalf of the EPO President. The EPO President agreed that the decisions of the Boards of Appeal “currently go in two directions”. However, in the EPO President’s view, there is a “clear trend towards abandoning the assessment of the third criterion as part of the novelty assessment”. The EPO President views this trend as a development as opposed to a contradiction in the case law. The EPO President also believed clarity has now been provided by the update to the guidelines.

However, the guidelines do not have the same legal force as boards of appeal case law. The guidelines (as the name suggest) are guidelines for first instance departments. Furthermore, Boards of Appeal decisions are only part of the picture. It is this Kat’s understanding that there are diverging approaches within the Examining and Opposition Divisions on this issue. Does this chime with readers’ experiences?

Despite the EPO President’s refusal to contemplate a referral in the case of selection inventions, he cannot be accused of being shy with respect to referrals under Article 112(2) EPC in general. Last year, following pressure from the Administrative Council, the EPO President controversially referred the question of whether plant products produced by natural processes should be patentable. The opinion of the majority of practitioners is that the President’s referral is clearly inadmissible, and resulted more from the political pressure than legal necessity (IPKat: Pepper gets spicy: The EPO President's Referral to the EBA).

It seems that proponents of a referral with respect to selection inventions can now only hope for a referral by a Board of Appeal (Article 112(1) EPC). Indeed, EPI urges parties in appeal proceedings who find themselves involved in a case where the use or not of the “purposive selection” criteria is influential to the case, to request the Board to refer the issue to the EBA. The opposing view is that a referral is not necessary, as the Boards of Appeal do seem to be following the approach now set out by the guidelines. Indeed, Roel van Woudenberg could be said to be pushing the issue merely because he disagrees with the removal of the purposive selection criteria (as argued his article cited by EPI).

Selection inventions have always been a hotly debated issue. In another articlecited by EPI, authored by Derk Visser, it is argued that all three criteria listed in the guidelines are inconsistent with recent case law on novelty. As with second medical use inventions (another form of selection invention), critics are keen that an inventor should not “get something for nothing” by salami slicing (or as some-would have it, “sashimi slicing”) the prior art. The boarder question surrounding the EPI recent request for a referral on sub-ranges is whether the EPO is currently striking the right balance on selection inventions. What do readers think?

In a welcome update, the 2019 guidelines have been updated to reflect the recent Boards of Appeal case law on selection inventions. European patent law considers a selection from the prior art as novel under Article 54EPC if: a) the end points of a claimed sub-range or any point falling within the claimed sub-range are not be disclosed in the prior art and b) the end-points of the claimed range are "sufficiently removed" from the individual elements disclosed in the prior art.  Historically, the EPO also required the selection to be non-arbitrary or “purposive”. In other words, the selection had to be linked to a technical effect (T 0198/84 (Thiochloroformates) (see r. 7) and T 0279/89 (r. 4.1)).

Reviewing the case law

However, recent Boards of Appeal decisions have departed from the requirement of a selection to be purposive (e.g. T 261/15). In T 1233/05, the Board of Appeal observed that the requirement for a purposive selection of a claimed sub-range "appears to fall back upon considerations which should be taken into account in the assessment of inventive step rather than in that of novelty" (r. 4.4). The 2019 EPO guidelines have been updated to reflect this position (G-VI-8, and G-VII-12).

Notably however, the updated EPO guidelines do not incorporate the recent Boards of Appeal case law on the novelty of purity selection inventions. For many years the EPO considered an increase in the purity of a compound a special case with respect to novelty. The burden of proof was placed on the applicant to show that the claimed level of purity was never achieved in the prior art. However, in a decision last year (T 1085/13), a Board of Appeal found that previous decision of the Boards were contrary to the principle that novelty can only destroyed by a clear and unambiguous disclosure in the prior art (IPKat post here). With regards to purity inventions, the 2019 guidelines still only indicate that “[a] known compound is not rendered novel merely because it is available with a different degree of purity if the purity can be achieved by conventional means (see T 360/07)” (G-VI-7).

The EPC provides for the patenting of the first or second medical use of a known “substance or composition”. The provision is intended to exclude the patenting of the second medical use of a device. However, "substance or composition" is not defined in the EPC. Determining whether or not something is a substance or composition is not always easy (IPKat: Substance or device - a distinction without a difference?).

The updated guidelines states that guidance on products which can be seen as "substances or compositions" can be found in T 1758/15. The Board in T 1758/15 defined a "substance or composition" as a product that achieves a direct therapeutic effect by a chemical, as opposed to physical, interaction with the body. However, the current case law does not provide guidance on how a physical versus chemical interaction should be determined, for example with respect to products such as tissue scaffolds.

A whole new section has been added to the guidelines on obviousness in the biotechnology field (G-VII-13 Inventive step assessment in the field of biotechnology). The section in particular imports Boards of Appeal case law from T 296/93 on the meaning of “reasonable expectation of success” in the context of an obviousness analysis. The updated guidelines emphasize that a reasonable expectation of success “is not to be confused with the ‘hope to succeed’. If researchers are aware when embarking on their research that, in order to reach a technical solution, they will need not only technical skill but also the ability to make the right non-trivial decisions along the way, this cannot be regarded as a "reasonable expectation of success"’ (see also Case Law of the Boards of Appeal, I.D.7.1).

A new section confirms that a claim may contain both product and process features (F-IV-4.12.1 Product claim with process features). The guidelines also clarify that the same criteria of novelty should be applied to such claims as are applied to product-by-process claims. Particularly, the process features of a product claim with process features are only novel if the product produced by the process is novel [Merpel: is this sentence in contention for the patent tongue-twister of the year?].

The European patent community is eagerly awaiting the response of the Enlarged Board of Appeal (EBA) to the EPO president’s referral in G 3/19 (Pepper) (IPKat post: Pepper gets spicy: The EPO President's Referral to the EBA). The nub of the issue is whether the Administrative Council’s amendment of Rule 28(2) EPC (excluding products produced by essentially biological processes from patentability) was lawful given the contradiction of the new Rule with the interpretation of Article 53(b) EPC by the EBA in G 2/12 (Broccoli/Tomato II).  Whilst the President’s referral is pending, the guidance that Rule 28(2) EPC excludes products produced by essentially biological processes from patentability remains unchanged in the 2019 guidelines (G-II.5.4). However, as of 4 April 2019, all proceedings before the EPO potentially impacted by the outcome of the referral have been stayed ex officiountil the EBA’s decision on the referral (EPO Notice).

In a minor update, the guidelines on this topic have been updated to explicitly specify that selfing (i.e. self-fertilization) of a transgenic plant should be considered an “essentially biological process”.

The patentability of software is another hot topic at the moment, and subject to its own referral to the EBA (IPKat: The patentability of computer simulated methods - another referral to the Enlarged Board of Appeal). The 2019 guidelines include some updates to the section relating the patentability of artificial intelligence and machine learning. In particular, the guidelines now clarify that “[t]erms such as ‘support vector machine’, ‘reasoning engine’ or ‘neural network’ may, depending on the context, merely refer to abstract models or algorithms and thus do not, on their own, necessarily imply the use of a technical means. This has to be taken into account when examining whether the claimed subject-matter has a technical character as a whole (Art. 52(1), (2) and (3))”.

Attentive readers may have noticed that many of the updates this year are in the form of citations of Boards of Appeal decisions. Case law references from the Boards have in the past been confined to the Case Law of the Boards of Appeal. The Guidelines have previously mostly refrained from referencing Boards of Appeal decisions on which the guidance was based. Why the change? Do readers have any insights?