How post-filing data rescued Novo Nordisk’s Ozempic patent in China

The Beijing IP court’s decision to revive a key semaglutide patent based on experimental data has far-reaching legal and economic implications, says Jennifer Che of Eagle IP.

Recently, all eyes have been on China as the fundamental patent covering semaglutide, the active ingredient in Ozempic and Wegovy, will expire on March 20, 2026. During the period leading up to the expiration, Novo Nordisk’s patent has been subject to multiple attacks from parties trying to invalidate the patent before this key date.

Novo Nordisk’s patent in China

On September 5, 2022, the China National Intellectual Property Association (China’s patent administrative office, hereinafter CNIPA) declared one of Novo Nordisk’s key semaglutide patents in China—ZL 200680006674.6—to be invalid (invalidation decision number 57950).

In response, Novo Nordisk submitted significant evidence in the form of post-filing data showing that semaglutide had increased half-life and a longer duration of action when compared with liraglutide, the closest prior art. Nevertheless, despite an amended claim scope and post-filing data, CNIPA declared the patent completely invalid.

Novo Nordisk appealed to the Beijing IP Court (the Court), which reversed CNIPA’s invalidation and upheld the patent.

Post-filing data

One of the biggest concerns for biopharma patent holders in China has been post-filing data. China is notoriously strict about experimental data in patents, especially in ‘unpredictable’ fields such as biology and chemistry. Patent applicants typically can only obtain a tight scope of protection around aspects of their invention that they have ‘proven’ through working examples.

Contrast this with the US and Europe, where patent applicants often obtain broader claim scopes based on fewer working examples and/or less post-filing data. Most patentees get much narrower patents in China compared to the US and Europe, at least in these ‘unpredictable’ fields.

More recently, China updated its Examination Guidelines on this issue. The current Guidelines stipulate that examiners shall consider post-filing supplemental data:

• when considering inventive step and sufficiency

• if the technical effect demonstrated by the supplemental data could undoubtedly be obtained by a skilled person in the art from the disclosure as originally filed.

What was in the originally filed disclosure?

The patent disclosure described a genus of compounds that were effective as GLP-1 receptor agonists. Notably, 22 compounds were specifically described with their preparation methods and characterisation data, including semaglutide. The patent disclosure described screening studies using db/db mice and minipigs. However, it did not specify which GLP-1 compound(s) were used in these screening studies.

CNIPA rejection

During the invalidation, CNIPA rejected all claims for lack of inventive step in view of liraglutide, a molecule with a similar structure and mechanism of action.

Novo Nordisk responded by arguing semaglutide had surprising technical effects that were markedly improved over liraglutide, pointing to post-filing data showing semaglutide’s significantly improved half-life (60-70 hours in minipigs) and long duration of action (48 hours in db/db mice) compared to liraglutide (24 hours).

As the original specification did not specify which compounds possessed the above-mentioned surprising effects (and thus no mention of semaglutide specifically having such technical effects), CNIPA opined that the effects demonstrated by the supplemental data “could not be undoubtedly obtained by a skilled person in the art from the disclosure as originally filed”.

The Court’s reasoning

However, the Court reversed, stating that the patent disclosure did possess sufficient support for the idea that semaglutide had a long duration of action. Specifically, the Court pointed to paragraph [0534] in the specification:

[0534] In one aspect of the invention, the GLP-1 agonist has a duration of action of at least 24 hours after administration to db/db mice at a dose of 30 nmol/kg.

According to the Court, the statement ‘GLP-1 agonist’ was referring to the entire genus of compounds, and thus the paragraph was asserting that all compounds (or at least the 22 examples in the specification) had a duration of action at least 24 hours after administration.

In essence, the general statement in paragraph [0534] was strong enough to indicate that the technical effect could undoubtedly be obtained by a skilled person in the art. The Court emphasised that if a patentee has already demonstrated that a general formula has a particular effect, then one can presume that all compounds within the general formula have this effect.

In this case, the patentee should have the right to submit post-filing data to confirm the effects of a specific compound within the general formula. Otherwise, if this was not allowed, the patentee would need to recite the results of each specific compound in the original specification, which would not be reasonable nor practical.

The Court contrasted the above case with the other study using minipigs on prolonged plasma half-life.

[0543] One aspect of the present invention is the preparation of GLP-1 analogues/derivatives with prolonged plasma half-life suitable for weekly administration. Pharmacokinetic properties can be assessed in minipigs, or domestic pigs as described below.

[0550] A second part of the pharmacokinetic screening was conducted on those compounds with an initial terminal half-life of 60-70 hours or more.

The Court argued that in the minipigs study, the specification did not clearly indicate which GLP-1 analogues had the technical effect of having a longer half-life. Instead, the property of having “an initial terminal half-life of 60-70 hours or more” was recited as the conditions required for a second screening rather than recited as technical effects in paragraph [0550].

The judges argued that, based on paragraph [0550], a skilled person in the art would not be able to infer that semaglutide could be suitable for the second part of the screening, as there was no way to know undoubtedly that semaglutide would have a half-life of 60-70 hours or more. As a result, the Court did not accept the post-filing data for increased half-life.

Nevertheless, the Court still upheld the patent based on the post-filing data for a long duration of action.

Conclusions

This is a huge case for so many reasons. The importance of the product, the economic and legal impact of the decision, and the fine line the Court ultimately drew to clarify China’s position on post-filing supplemental data make this a fascinating case to study.

At a minimum, this case broadens what kinds of statements could be sufficient to demonstrate that an idea can be “undoubtedly obtained” from the patent specification as filed.

Importantly, in this case semaglutide was never specifically called out as having significantly good PK properties. Instead, the specification held a general position that the compounds (implicitly all the compounds) had a > 24-hour duration of action.

The ability to have this additional ‘hook’ based on generic language could be a lifesaver in a lot of situations. Astute patent drafters should carefully consider what types of general statements assert technical effects they should add.

A general statement that’s not entirely true (and unsupported by data) could be fatal, while a general statement that is true (and can be further confirmed with post-filing data), could save the life of a patent—as it did in this case. Be careful making statements that imply only a subset of compounds have a certain technical effect, unless it is true and supported by data.

This case isn’t completely over yet. The petitioner who filed the invalidation has appealed to the Supreme People’s IP Court. We expect to have a decision before the March 20, 2026 expiration date.

Jennifer Che is president and managing director of Eagle IP. She can be contacted at: jenniferche@eipgroup.asia

Audrey Cheung is a qualified Chinese patent attorney at Eagle IP.

Yolanda Wang is a principal, Chinese patent attorney, and Chinese patent litigator at Eagle IP.