G 1/23 in practice: Inventive step based on non-reproducible prior art

We now have the first substantive decision from a Board of Appeal applying the Enlarged Board of Appeal’s decision in G 1/23. In T 1044/23, the Board of Appeal have grappled with the principles set out in G 1/23, and whether non-reproducible prior art can be used for inventive step analysis.

Background
Many readers will be familiar by now with the Enlarged Board of Appeal’s decision in G 1/23, which clarified the question of whether a product that is on the market, but whose internal structure or preparation process is undisclosed, or even undiscoverable, can form part of the state of the art under Article 54(2) EPC. In its headnote, G 1/23 held that:

"A product put on the market before the date of filing of a European patent application cannot be excluded from the state of the art within the meaning of Article 54(2) EPC for the sole reason that its composition or internal structure could not be analysed and reproduced by the skilled person before that date."

Now with T 1044/23, we have the first decision of the Board of Appeal applying G 1/23, and in particular with respect to inventive step.

Key facts of T 1044/23
The patent in T 1044/23 relates to a polyethylene composition defined by a combination of density, melt index, short-chain branching distribution and various other parameters.

The Opponent had cited as prior art two polyethylene resins, namely DMDA-8904 NT7, and DMDA-8907 NT7, which they had sold commercially before the priority date. The resins were not described in any published document, and the details of their manufacturing process (e.g. catalysts, process conditions) were not in the public domain. The dispute related to whether these marketed, but non-reproducible, resins destroyed the novelty of the claims and/or rendered them obvious.

At first instance, the Opposition Division reached their decision before the referral in G 1/23, and applying the previous case law of G 1/92, found that since the resins were non-reproducible, they did not form part of the state of the art at all. The opposition was therefore rejected.

Applying G 1/23 to non-reproducible products
Following the decision of G 1/23, the Board reversed the decision of the Opposition Division and accepted that the resins could be considered as prior art even though they were unreproducible without the detailed knowledge of how to prepare them. This followed directly from the Enlarged Board’s holding that the reproducibility requirement is no longer a valid criterion for deciding whether a product forms part of the state of the art.

The Board also accepted the Opponent’s evidence, namely invoices and a certificate of analysis, showing that the resins had been sold, and so made available to the public, before the priority date of the patent. That lead the Board to the conclusion that claim 1 of the patent lacked novelty.

Inventive step starting from a non-reproducible product
The Patentee then defended the patent on the basis of an auxiliary request, where the density range had been narrowed in order to establish novelty over the 8904 and 8907 resins.

The Board of Appeal selected the resins as the closest prior art for the considering inventive step. The Patentee had argued that this was not correct, as conceptually the skilled person would disregard a product that they could not reproduce. However, the Board rejected this argument, comparing it to the Enlarged Board’s example of adding lemon juice to Coca-Cola to make the taste less sweet. The Board reasoned that since the resins could be modified or analysed, they were a realistic starting point for arriving at the claimed invention, and could not be disregarded simply because the resin, like the Coca-Cola, cannot be reproduced.

From here the Board then considered whether the invention was obvious to the skilled person when starting from the closest prior art. Since it was accepted by both parties that increasing the density of the composition actually worsens its properties, the Board formulated the problem to be solved as the provision of an alternative composition.

The Opponent then argued that it would have been within the skilled person’s common general knowledge to modify the density of the polymer composition, even despite there being no pointer to do so, and that therefore the claimed composition was obvious. However, the Board was not persuaded, finding that while the skilled person could increase the density of a polyethylene composition, there was not enough evidence to conclude that all of the other properties of the composition could be maintained in doing so. The Board could not accept therefore that the claimed composition could be arrived at without an undue amount of experiment work, and so inventive effort.

The Board therefore acknowledged the inventiveness of the auxiliary request and maintained the patent in amended form.

What T 1044/23 tells us about applying G 1/23
A non-reproducible product can be the closest prior art
As we noted previously, G 1/23 simply left question of whether an unreproducible product could be the closest prior art as one that depends on the facts of the case. In this decision, we’ve seen the Board of Appeal very firmed accepting that it can. In effect, G 1/23 moves the question of reproducibility towards the later stages of the problem-solution approach. Reproducibility is now a factor in deciding what the skilled person would do when starting from a prior art product, not whether the product is prior art in the first place.

Trade secrets
While the Opponent was ultimately successful in their appeal, it does seem that their decision not to reveal how their product is manufactured ultimately made their position more difficult. Had the method of making the resins been known to the skilled person, they would clearly have been reproducible. However, since this wouldn’t have ultimately changed the decision, it doesn’t seem to provide much incentive for patentees to disclose their valuable manufacturing processes.

The question of modifiability
The Board’s comparison to the Enlarged Board’s Coca-Cola example does seem to highlight the limits of G 1/23. There is no question that the skilled person can simply add lemon juice to a glass of Coca-Cola and so the fact that the Coca-Cola cannot be reproduced is not relevant. In contrast, it is usually not possible to directly change the properties of a polymer by simply adding a component to it. Instead, its method of manufacturing must be changed and that may not be known to the skilled person. Therefore, it seems there is a distinction between products that can be simply bought and modified, like Coca-Cola, and products which need to be manufactured from scratch to alter their properties, such as polymers.

Final thoughts
The way G 1/23 is being applied, in our opinion, allows for a more realistic assessment of inventive step. It will often be the case in real life that the most realistic starting point is indeed a commercially available product, and it seems unrealistic to exclude this product completely. That could be, for example, a competitor product that is taking a share of the market. That product may have no published synthesis, and someone hoping to modify it may not know if is reproducible until trying to actually reproduce it. The Board’s decision in T 1044/23 seems to acknowledge that reality.

Ultimately, it seems that G 1/23 is being interpreted pragmatically and is allowing the Boards of Appeal to deal with the question of reproducibility as one of inventive step, without the ‘legal fiction’ of a marketed product simply not existing to the skilled person. Time will tell whether other Boards will follow the same approach.

Make sure to follow us on LinkedIn for the latest news and developments. For more information, please contact Chris Lindsay, Jim Denness or your usual Abel + Imray advisor.