With respect to the “realistic starting point” the decisions stated: “A starting point is realistic if the teaching thereof would have been of interest to a skilled person who, at the relevant date, wishes to solve the objective problem. This may for instance be the case if the relevant piece of prior art already discloses several features similar to those relevant to the invention as claimed and/or addresses the same or a similar underlying problem as that of the claimed invention. There can be more than one realistic starting point and the claimed invention must be inventive starting from each of them.”

The breadth of “realistic starting point” had, however, yet to be tested.

Two recent first instance orders from the Munich Central Division, TCL EUROPE SAS v Corning Incorporated (UPC_CFI_337/2025) and WIRPLAST – Więcek Spółka Jawna v VILPE Oy (UPC_CFI_280/2025), shed light on this issue.

TCL EUROPE SAS

TCL EUROPE SAS concerned a patent (EP3296274) to a method for producing alkali-free glass sheets by a downdraw process. The process comprising inter alia fining batch materials so that the glass comprised specified compounds at a specified mole percent on an oxide basis. Critically, the fining step is performed using neither arsenic nor antimony. According to the patent’s description, the resulting alkali-free glass sheets exhibited desirable physical and chemical properties for use as substrates in flat panel display devices.

A revocation action was brought by TCL EUROPE SAS (the claimant) including inventive step attacks starting from D19 (US 2002/0082158A1; in particular, Example 27), D20 (JP 2004-189535 A; in particular, Examples 15 and 10), and D27 (JP 2001-348247 A; in particular, Example 10).

Both D19 and D20 were considered to be realistic starting points by the court, as they both related to alkali-free glasses. D19 provided aluminosilicate glasses exhibiting desirable physical and chemical properties for substrates in flat panel display devices suitable to be prepared via a downdraw process, and D20 related to the use of alkali-free glass for making displays which requires a good etch resistance, low thermal shrinkage and a high strain point (similar to the patent). The general description of D20 was further noted to disclose downdraw as a suitable process for making the glass, albeit as one of several suitable methods. The skilled person was therefore noted to be “interested” in each of D19 and D20.  

However, the court considered it doubtful that D27 was a realistic starting point for the skilled person. It was noted that D27 related only to displays and photographs which do not substantially contain an alkali metal oxide, have a high strain point, a high Young’s modulus, a low density and a low expansion coefficient, and are excellent in BHF resistance and acid resistance. Additionally, D27 focused on float methods as a preferred manufacturing process, rather than a downdraw process. The court noted that a “high viscosity, which is required for a downdraw process” was also not mentioned in D27 (paragraph 159).

Notably, the court took a dim view of the claimant relying on specific examples in each of the starting prior art documents. This led to the court concluding that Example 27 of D19 and Example 15 of D20 were not realistic starting points. Example 27 of D19, for instance, included 0.3 mole percent of antimony (III) oxide whereas Examples 22-25 contained neither arsenic nor antimony.

The Munich Central Division summarised its approach to determining a realistic starting point in the headnotes of the decision: “A realistic starting point is typically a prior art disclosure as a whole. Absent a specific reason or pointer in the disclosure itself (or based on common general knowledge) to do so, the selection of a particular example composition as a “starting point”, merely because it happens to come “closest” to the claimed subject matter in terms of structural components, bears the risk that such selection itself already involves hindsight.”

This appears to contrast with established case law before the European Patent Office (EPO), which considers that each embodiment constitutes an element of the prior art that could equally serve as a starting point for the assessment of inventive step (Case Law of the Boards of Appeal (CLBA) I.D.3.1, citing T 1654/22, and I.D.3.7.1, citing T 787/17).

WIRPLAST

WIRPLAST concerns a patent (EP2649380) to an exhaust ventilation pipe with an exhaust pipe of a ventilation channel extending through the roof of a building provided inter alia with a hat, the upper part of the hat being provided with a conical cup preventing rain water from entering the ventilation channel, further comprising a circular spirit level integrated on the top of the conical cup in the hat.

A revocation action was filed by WIRPLAST (the claimant) including inventive step attacks starting from D1 (DE10118226C1), in combination with the common general knowledge, or any of D2 (JP2002106855A), D3 (US20070167130A1), or D4 (CN2255012Y).

D1 concerned a roof fan with a fan housing which comprised inter alia an upper conical cover. The aim of the invention of D1 was to provide a roof fan which is rainproof, snowproof, and protected against wind influences.

The patentee/defendant (VILPE OY) did not dispute that D1 was a realistic starting point. Instead, it argued that the roof fan of D1 was essentially a different invention than that of the claimed patent because the roof fan of D1 was to be placed on a roof with a base plate, whereas the claimed device comprises a pipe extending through the roof.

The court agreed with the patentee. It further commented that D1 as a whole would have been of interest to the skilled person and represented a realistic starting point. This was because it related to the same technical field as the patent (“ventilation devices”) and had several technical features in common with the claimed invention (paragraph 70).

The Munich Central Division summarised its approach in the headnote of the decision: “Even if a certain prior art document in the same technical field as the patent in suit is considered to be a realistic starting point, it can be relevant for the assessment of inventive step that it relates to a different kind of device and solves a different problem than the invention of the patent.”

Comments

In proceedings before the EPO, the closest prior art document is normally a document which: discloses subject matter for the same purpose as the invention; aims to achieve the same objective as the invention; and/or has the most relevant technical features in common with the invention (Guidelines for Examination G.VII.5.1).

The CLBA also indicates that there is no requirement that the prior art contains a pointer to a particular embodiment, nor does the opponent need to justify why the skilled person would select a particular embodiment, for said embodiment to qualify as a starting point in the problem-solution approach. Instead, every embodiment constitutes an element of the prior art that could equally serve as a starting point (CLBA I.D.3.1, citing T 1654/22, and I.D.3.7.1, citing T 787/17).

There has been a lot of commentary on “realistic starting point” v “closest prior art” but what is interesting from the decisions discussed from the Munich Central Divisions is the UPC’s reluctance to start from a specific example without a “pointer” thereto in the document itself.

Such an approach represents further divergence from the EPO in its assessment of inventive step.

In some ways, the UPC’s assessment of inventive step may allow a broader approach to the selection of realistic starting points, as long as the prior art is from the same technical field as the invention and has several technical features in common. However, it may be difficult for a party to argue starting from a specific embodiment (in the absence of pointers or justification that the skilled person would have done so) unlike at the EPO. Consequently, inventive step attacks may require significantly more signposts in the prior art to convince the UPC First Instance Divisions that an invention is obvious, than to convince the Opposition Division or the Boards of Appeal of the EPO.

Given the consistent messaging that the EPO wishes to harmonise practice at a European level (for example, at the UPC), such as noted at G 1/24, reason 16, it will be interesting to see how the Boards of Appeal respond to further divergence at the UPC. For parties using both systems, a takeaway message from these two decisions is that there is additional scope for the two jurisdictions to come to contradictory decisions on inventive step.

Case details at a glance

Decision level: Munich Central Division
Parties: TCL Europe SAS (claimant) and Corning Incorporated (defendant)
Citation: UPC_CFI_337/2025 
Date: 24 February 2026 
Decision (PDF): dycip.com/upc-cfi-337-2025 

Decision level: Munich Central Division
Parties: WIRPLAST – Więcek Spółka Jawna (claimant) and VILPE Oy (defendant)
Citation: UPC_CFI_280/2025
Date: 08 April 2026 
Decision (PDF): dycip.com/upc-cfi-280-2025

Decision level: Court of Appeal 
Parties: Amgen Inc v Sanofi-Aventis Deutschland GmbH, Sanofi-Aventis Groupe and Sanofi Winthrop Industrie SA
Citation: UPC_CoA_528/2024
Date: 25 November 2025
Decision: dycip.com/upc-coa-528-2024

Decision level: Court of Appeal 
Parties: Meril v Edwards Lifesciences Corporation
Citation: [UPC_CoA_] 464/2024, 457/2024,   458/2024, 530/2024,  532/2024, 533/2024, 21/2025, 27/2025
Date: 25 November 2025
Decision: dycip.com/upc-coa-meril-edwards

Related article

Inventive step at the UPC: Court of Appeal sets definitive test, 10 December 2025.

Useful links

• EPO Guidelines for Examination, 5.1 Determination of the closest prior art: dycip.com/epo-guidelines-prior-art 
• EPO Case Law of the Boards of Appeal, 3.1 Determination of closest prior art in general: dycip.com/epo-appeal-prior-art 
• EPO Case Law of the Boards of Appeal, 3.7.1 General: dycip.com/epo-appeal-general