[Article / European Patents] Decisions Relating to Plausibility After G 2/21
更新日：3 日前
Article information
Article type: Publication review
Categories: Europe
Issues: Inventive step, Enablement
Published on website: 2026-10-08
Display and classification checked: 2026-10-08
[Article / European Patents] Decisions Relating to Plausibility After G 2/21
Review of Decisions Relating to Plausibility After Decision G 2/21 (Yoshinori Shimizu, Eisan Go, Rie Kawai, Patent, November 2025)
1. Core of G 2/21: Two Hurdles
Decision G 2/21 did not take the approach of immediately disregarding post-published evidence; instead, it indicated that such evidence may be taken into account in assessing inventive step if the following two conditions are satisfied (Headnote II).
(1) Encompassed: The technical effect is encompassed by the technical teaching of the application as originally filed.
(2) Embodied: The technical effect is embodied by the same invention disclosed in the application as originally filed.
2. Cases That Divided Outcomes in Practice: Inventive Step and Sufficiency of Disclosure
In decisions following G 2/21, a clear difference has emerged between inventive step (Art. 56) and sufficiency of disclosure (Art. 83) regarding the permissible scope of post-published evidence.
A. Inventive Step (Relatively Flexible)
Where the specification as originally filed contains a plausible indication, there is a tendency for post-published data to be accepted to reinforce that effect.
Accepted Example (T 2716/19 - Gilead): Although the application as filed lacked individual data for a specific base (magnesium isopropoxide), it stated that a base having the same effect could be substituted. Post-published evidence (D24) was therefore accepted, and inventive step was acknowledged.
Rejected Example (T 0467/22): The data in the application as filed were themselves contradictory (for example, the impurity level did not change even above the boiling point), and post-published evidence failed to dispel the resulting doubt. Inventive step was therefore denied.
B. Sufficiency of Disclosure (Extremely Strict)
If proof of therapeutic effect is missing from the application as filed, post-published evidence cannot fill that gap.
Rejected Example (T 2790/17): The original data lacked scientific validity, for example by showing an adverse effect at low doses. This lack of credibility could not be remedied even by later high-quality publications (including D17).
Rejected Example (T 1779/21): The application as filed did not show any effect for monotherapy, and the effect was also difficult to infer from the prior art. The post-published data were therefore disregarded.
3. Japan-Europe Comparison and Summary of Measures
In many of the cases surveyed, patenting succeeded in Japan, while applicants struggled in Europe (or the patents were revoked).
Original article:
Originally posted on LinkedIn: https://www.linkedin.com/feed/update/urn:li:activity:7513392964388261889/
Related articles
Consultation guide by purpose: https://www.takaishihideki.com/en/single-post/consultation-guide-en
Update history
The update history records only changes actually published, with dates.
2026-10-08: First publication of the LinkedIn original; display, classification, and consultation link set.
