Patent practice: recent developments in life sciences in China

Inventiveness has become an increasingly tough issue for patent practitioners in the chemical and biological fields in China. More and more weight is being given to the experimental data described in the original specification. When seeking to establish inventiveness, the requirements for post-filing data are rigid and strict. However, in light of recent cases, it seems that the People’s Court and the China National Intellectual Property Administration (CNIPA) tend to reconsider the weight of non-obviousness and technical effect in inventiveness evaluation; and the standard regarding acceptance of post-filing data is now becoming clear.

When evaluating the inventiveness of a chemical or biological application, both non-obviousness and technical effect are considered. If the technical solution has an unexpected technical effect, it possesses inventiveness; if it is non-obvious, a beneficial technical effect is sufficient (in which case, it is unnecessary to reach an unexpected level).

However, in practice there is a trend whereby more weight is placed on the technical effect than on non-obviousness, especially in life sciences. It is often taken for granted that it is obvious for a chemical molecule to be obtained by conventional modification or replacement, without considering any significant difference between the invention and the prior art, while it is typical for the unexpected technical effect brought by the chemical molecule to be required. In such a situation, either the original experimental data or the post-filing data must be relied on. However, there are no clear criteria for acceptance of the post-filing data.

Fortunately, this unfavourable situation is gradually changing. The revised version of the Guidelines for Patent Examination, effective as of 1 April 2017, opens doors and sets a legal basis for the acceptance of post-filing data. The following cases illustrate recent developments in patent practice in life sciences in China, especially regarding how to determine teaching from the prior art and the criteria of acceptance of post-filing data.

Teaching should be read from prior art as a whole

In AstraZeneca v Patent Re-examination Board (2018 Beijing Admin Final 6345), the Beijing High Court opined that any teaching should be read from the prior art as a whole, and an embodiment should not be interpreted separately. After construing the prior art as a whole, the Beijing High Court reversed the previous decision and judgment by relying on the difference of structure alone, which is different from the overwhelming manner of stressing the unexpected technical effect applied by the substantive examination division and re-examination division.

Ticagrelor is a platelet aggregation inhibitor developed by AstraZeneca. The basic compound patent (ZL 99815926.3, hereafter the ‘926’ patent’) of Ticagrelor will expire on 2 December 2019. A local generic company, Shenzhen Salubris Pharmaceutical Co Ltd, challenged the patent before the CNIPA’s Patent Re-examination Board (PRB). After examination, the PRB issued Invalidation Decision 33591, in which all claims of the patent were invalidated. During appeal, the Beijing IP Court upheld Decision 33591. In both Decision 33591 and the judgment made by the Beijing IP Court, the manner of evaluating inventiveness caused widespread controversy. Fortunately, the Beijing High Court issued a final judgment withdrawing both Decision 33591 and the judgment made by the Beijing IP Court in December 2018.

In Decision 33591 and the Beijing IP Court judgment, without considering the significant structural difference between the claimed compound and the prior art, both the PRB and the court took the position that it was obvious for the skilled artisan to obtain the claimed compound through conventional replacement and with limited experiments.

Specifically, the claimed compound, Ticagrelor (see Figure 1), differs from the closest compound of prior art in substituents on phenyl ring (3-F, 4-F v. 4-Cl) and R group on the core (-OCH2CH2OH v -C(O)NH2).

The reasoning of the PRB and the court at the first instance was as follows: regarding the distinguishing technical feature of substituents on phenyl ring, replacing F with Cl was a conventional substitution in the art. Meanwhile, the prior art also suggested that the phenyl ring could be substituted by one or more halo atoms. Regarding the distinguishing technical feature R group on the core, both -OCH2CH2OH and -C(O)NH2 belonged to the conventional modification to the core. No evidence was available to demonstrate that the claimed compound possess unexpected technical effect over the prior art compound. Based on such determinations, the 926’ patent was invalidated and the decision was further upheld by the first-instance court.

In the second instance, the Beijing High Court revoked the previous judgment, holding that it was non-obvious for the skilled artisan to obtain the claimed compound in light of the prior art as a whole.

The Beijing High Court opined how to construe a Markus compound and how to determine the teaching from a Markus compound as a prior art.

A Markus compound usually includes a core moiety and a Markus element. A core moiety is essential and unchangeable for its activity, and the Markus element is exchangeable while remaining active. When the prior art relates to a Markus compound, it means that the core moiety cannot be modified if the same or similar activity is desired to be obtained. Therefore, the motivation obtained by the skilled artisan from such a prior art is to modify the Markus elements rather than the parent core, in order to obtain the same or similar activity.

In this case, the -C(O)NH2 moiety is disclosed as part of the core moiety of the Markus compound in the prior art, which means that the -C(O)NH2 moiety is not subject to modification if the prior art is taken as a whole. In other words, the prior art does not provide any teaching for the skilled artisan to replace -C(O)NH2 in the core moiety with -OCH2CH2OH while the obtained compound still has same or similar activity.

The embodiment should not be interpreted separately from the prior art and any teaching should be read from the prior art as a whole. After the distinguishing technical features are determined, the content of the prior art should be considered as a whole, rather than alone as a specific embodiment.

Based on the nature of a Markus compound, a skilled artisan would not modify the core moiety of the prior art when developing a compound with similar or same activity. In this specific case, when considering the prior art as a whole, the -C(O)NH2 moiety is an essential part of the core moiety of the Markus compound, the skilled artisan would have no motivation to modify any part of the core moiety, including the -C(O)NH2 moiety, if the same or similar activity was desired to be achieved. Decision 33591 and the first-instance judgment did not consider the prior art as a whole.

In AstraZeneca v PRB (2018) the Beijing High Court reversed the previous decision and judgment by relying on the difference of structure alone, which is different from the overwhelming manner of stressing the unexpected technical effect applied by the substantive examination division and the re-examination division. This is a positive signal for changing the current strict and rigid manner adopted by the examination and re-examination divisions. Although China is not a case-law jurisdiction, such a judgment still provides some direction for patent practitioners. In evaluating inventiveness, the three-step method prescribed in the guidelines should prevail, and unexpected technical effect is used only as an auxiliary means. Any teaching should be read from the prior art as a whole and an embodiment should not be interpreted separately.

Clarification of criteria of acceptance for post-filing data

In Gilead v PRB (2017 Beijing Admin Final 1806), the Beijing High Court accepted the comparative data generated after the filing date of the patent at issue, although the compared data of the claimed compound was not described in the original specification. The original data and supplemental data for the claimed compound belonged to a different type, while proving the same technical effect disclosed in the original specification.

Tenofovir disoproxil (Viread, Bis(poc)PMPA) is a medication developed by Gilead which is useful in treating chronic hepatitis B and preventing and treating HIV and AIDS. The basic compound patent (ZL 97197460.8, hereafter ‘460’ patent’) of Viread expired on 25 July 2017. Since 2011 but before its expiration, several generic companies challenged the 460’ patent before the PRB. The PRB issued Invalidation Decision 22284, in which all claims were invalidated. At the appeal stage, the Beijing IP Court upheld Decision 22284.

The original specification disclosed the antiviral effect of Viread with CC50 over PMPA. In the prior art, it disclosed the antiviral effect of Bis(pom)PMPA with ED50 over PMPA. The patented Viread could not be compared with the prior art compound Bis(pom)PMPA as their effects were expressed in different types.

In the invalidation stage, post-filing data was submitted to show that the antiviral effect of Viread is better than that of Bis(pom)PMPA in terms of ED50. However, in evaluating the inventiveness of Viread, the PRB and the first-instance court did not consider the post-filing data.

In the second instance, the Beijing High Court revoked the previous judgment, holding that the comparative data should be considered to prove the unexpected technical effect of Viread over the prior compound Bis(pom)PMPA.

On one hand, it was confirmed that the post-filing data was published in an academic journal after the filing date of the 460’ patent, but before the date of invalidation. More importantly, the post-filing data was relevant with the 460’ patent at issue and was directed to the claimed compound and the prior art compound (ie, Viread v Bis(pom)PMPA). The technical effect demonstrated by the post-filing data was also expressly described in the original specification, although in a different form (CC50 in the original specification; ED50 in post-filing data). Therefore, the post-filing data objectively reflected the contribution of the 460’ patent, although it was generated after the filing date. In this regard, the Beijing High Court accepted the post-filing data and withdrew Decision 22284 and the first-instance judgment.

Gilead v PRB provided some directions regarding post-filing data. The technical effect of the claimed compound proved by the post-filing data should be disclosed and validated by the experimental data in the original specification, although the data type might be different from the post-filing data.

Another case also shows the criteria of acceptance of post-filing data. In Re-examination Decision 74723, the post-filing data proving the unexpected technical effect of the claimed compound over the prior art was accepted by the PRB, although no activity data for the claimed compound was disclosed in the original specification.

Specifically, a prior art (compound B) was cited as the closest prior art to reject the inventiveness of the specific compound (compound A) at issue. The original specification disclosed that a structurally similar compound (compound A’) is more potent than the prior art compound B in terms of comparative data. During the prosecution stage, the applicant submitted the post-filing data between compound A and prior art compound B, showing the unexpected technical effect seeking to establish inventiveness. Unfortunately, the data was denied. The examiner took the position that the data of the claimed compound A was not disclosed in the original specification, and the skilled artisan was unable to obtain such effect from the specification directly and unambiguously.

At the re-examination stage, the PRB accepted such post-filing data. The reasons included that:

• the claimed compound A is structurally very similar to compound A’;
• the skilled artisan could predict that the two compounds share a similar activity that is more potent than the prior art compound B; and
• the post-filing data demonstrated that compound A and compound A’ have similar activity, further demonstrating that the claimed compound A is more potent than the prior art compound B, proving the above prediction.

This case demonstrates that post-filing data may still be accepted even if there is no data for the claimed compound, so long as the original specification discloses the effect data for other similar compounds. The compounds with a similar structure might act as a bridge to introduce the newly produced data for the claimed compound.

However, in AstraZeneca v PRB the Beijing High Court declined to accept the post-filing data proving the unexpected technical effect of the claimed compound over the prior art compound in terms of metabolic stability. The effect of high stability is mentioned only in the background section, without corresponding experiment on stability in the original specification, let alone any data regarding the effect. The Beijing High Court opined that the skilled artisan was unable to obtain a stability effect from the specification directly and unambiguously and the post-filing data was rejected. This case demonstrates that post-filing data is difficult to accept if only an ‘assertion’ on the technical effect is described in the original specification.

Comment

In evaluating inventiveness, the three-step method prescribed in the guidelines should prevail and unexpected technical effect be used only as an auxiliary means. Any teaching should be read from the prior art as a whole and an embodiment should not be interpreted separately.

As for a Markus compound, it usually includes a core moiety and a Markus element. When the prior art relates to a Markus compound, it means that the core moiety cannot be modified if the same or similar activity is desired. Therefore, the motivation obtained by the skilled artisan from such a prior art is to modify the Markus elements, rather than the parent core, in order to obtain the same or similar activity.

The criteria of acceptance of post-filing data now becomes clear. In order to ensure that the post-filing data be accepted, the original specification should disclose at least the technical effect and such effect should be validated by the data for the claimed or similar compound. In drafting, it is advisable that the original specification disclose technical effect with as much data as possible. During the prosecution, re-examination and invalidation stages, it is advisable to establish the link between the technical effect proved by the post-filing data and the facts disclosed in the original specification.