I. Introduction

Patent concerned in this case is a patent for invention named “Pyrrole Substituted 2-Indolinone Protein Kinase Inhibitors” (patent number: ZL01807269.0). The patent, which is a core patent of Sunitinib compound, was granted on August 1, 2007 and will expire in February 2021. Sunitinib (trade name: “Sutent”) is a small-molecule, multiple target receptor inhibitor of tyrosine kinases (RTK). It plays multiple roles in the inhibition of tumor angiogenesis and the prevention of generation and metastases of tumor cells. It was approved and launched on US and European market in 2006 and came into Chinese market in 2008. Although Sunitinib’s sales revenue worldwide witnessed a downward trend year by year, it achieved $1.049 billion sales revenue globally in 2018.

In May 2019, invalidation petitioner CSPC OuYi Pharmaceutical Co., Ltd. filed a request for invalidation. OuYi Pharmaceutical Co., Ltd. developed Sunitinib malate capsules and received drug registration documents issued by National Medical Products Administration on January 2, 2020. The capsules became the first generic drug of this category that had been approved in domestic market.

As the outcome of the invalidation proceedings, the patent was upheld by the Chinese Patent Office on the finding that all claims are novel and inventive. Extracts from the invalidation decision (herein referred to as Invalidation Decision) are quoted in this article.

II. Brief of the Case

Patent concerned relates to pyrrole substituted 2-indolinone compounds of formula (I). They exhibit protein kinase (PK, such as PDGF and VEGF) modulating ability and could be used to treat conditions relating to abnormal PK activity, for instance, neuroglioma. The patent concerned specifically elaborates the cell activity of the compounds on PDGF and VEGF as well as the in vivo activity tests relating to the two kinases.

The petitioner filed for invalidation based on the following reasons:

Claim 1 does not possess novelty over evidence 3 or evidence 6 (priority document of evidence 3) for the evidence discloses compound “dimethyl sunitinib”. Further, claim 1 does not possess inventiveness over compound 13 of evidence 6.

III. Amendments to Markush Claims

Compound of formula (I) in claim 1 is drafted in a Markush format:

“1. Compound of formula (I) or pharmaceutically acceptable salt thereof ：

Wherein:

R¹ is selected from the group consisting of hydrogen, alkyl and –C(O)NR⁸R⁹;

……

R¹² is –NR¹³R¹⁴, hydroxy, -C(O)R¹⁵, aryl and heteroaryl;

……

n is 1, 2, 3, or 4…….”

In reply to the observations of the petitioner, the patentee restricted the radical R¹² in claim 1 to “-NR¹³R¹⁴”. The amendments involved deletion of “carboxyl, -C(O)R¹⁵, aryl and heteroxyl” from the original definition. After judgment, the panel deemed that the amendments did not belong to deletion of parallel technical solutions and were not acceptable based on the Chinese Guidelines for Examination. Thus, the amendments were not accepted.

In fact, as Markush claim has high generality, interpretation on its scope and acceptability of amendments are always subjected to strict criteria. The criteria is mainly stemmed from the precedent case “Sankyo Pharmaceuticals Co., Ltd v. Beijing Wansheng Pharmaceuticals Co., Ltd.” (concerning Patent No. 97126347.7) heard by the Supreme People’s Court. In accordance with this case, claims of compounds drafted in Markush format are a generalized technical solution rather than parallel technical solutions, nor are collections of a number of compounds. The principle for making amendments to Markush claims should be: a class of compounds or a single compound possessing new property(ies) and function(s) must not be generated due to the amendments. Moreover, any option of any variable in Markush claims is not allowed to be discretionarily deleted. As Markush claim per se is considered as not being a group of “parallel technical solutions”, small wonder that the deletion of “options” thereof is not deemed as “a deletion of parallel technical solutions” prescribed in the Chinese Guidelines for Examination.

The reason why strict criteria as such is adopted in judicial practice is that there are uncertainties over whether new protection scope will emerge provided that patentees are allowed to delete any option of any variable. For example, if a specific compound that is never disclosed in the patent could be obtained via deletion, the public may not maintain stable expectations and this will undermine the stability of the patent validity system. In this regard, the author believes that patentees would do their best to retain those critical elements, as both patentees and invalidation petitioners know that the critical elements of the core solutions covered in a Markush claim usually represent the drugs that are being launched and the two parties often center their disputes on these critical elements. For example, in this case, R¹² being -NR¹³R¹⁴ is precisely a critical element of Sunitinib. The patentee strategically retained this critical element by deleting the rest definition of R¹². Thus, at least in this case, deleting non-critical elements is in fact an amendment that could have been expected by the public. The new Markush claim obtained after the amendment still comprises the compounds that share identical properties and functions. From this perspective, the amendment to the Markush claim in this case conforms to the principle, “a class of compounds or a single compound possessing new property(ies) and function(s) must not be generated due to the amendments”, required by the Supreme People’s Court. The amendment does not go beyond the stable expectation of the public, either. Therefore, it seems that contemplation should better be made over the practice that absolutely refutes to delete any option(s) from Markush claims.

Deletion of options is usually allowed in European and US patent practice. The author thinks that the criteria of amendments to Markush claims in China might tend to be closer to the European and US practice in future and, in order to pursue a better balance between the interest of patentees and that of the public, might be somewhat loosened on case-by-case basis.

[Extracts from Invalidation Decision]

The panel holds the opinion that the compounds of general formula in the patent concerned is drafted in Markush format. The formula comprises seven variables from R¹ to R⁷, wherein R⁶ further comprises variable R¹⁰, and R¹⁰ further comprises variables R¹¹ and R¹², namely the seven variables involving three levels of variations. The patentee deleted the radicals from R¹² selected among multiple variables in multi-levels. Nevertheless, such deletion of some of the radicals in the Markush claim does not belong to a deletion of parallel technical solutions and thus is not in conformity with the provision concerning amendments in invalidation procedure prescribed in Chinese Guidelines for Examination.

IV. Judgement of the Disclosure of a Specific Compound in the Prior Art

One dispute between both parties is whether the prior art discloses specifically the compound “dimethyl sunitinib”. The petitioner believed that evidence 3 or evidence 6 discloses the compound 2-indolinons of the general formula, which could be obtained via the reaction of hydroxy-indoles of general formula 2 and aldehydes of general formula 3. The hydroxy-indoles of general formula 2 can include, among others, 5-fluorohydroxyl indole, and the aldehydes of general formula 3 can include, among others, 5-formyl-2,4-dimethyl-1H-pyrrole-3-carboxylic acid (2-dimethylaminoethyl)amide. On this basis, those skilled in the art could have determined that the two starting materials may be reacted to form compound 5-(5-fluoro-2-oxo-1,2-indoline-3-denemethyl)-2,4-dimethyl-1H-pyrrole-3-carboxylic acid (namely “dimethyl sunitinib”), falling within the claimed scope. On the other hand, the patentee argued that neither of evidence 3 and evidence 6 discloses every specific compound of 3-pyrrolidinyl-2-indolinones that are obtainable by a reaction of every recited hydroxyl-indole and every recited aldehyde; therefore, dimethyl sunitinib is not disclosed. As a result, the panel supported the argument of the patentee.

Judgment of novelty follows the principle of comparison against single piece of prior art. Thus, in a Markush claim, the compound which is obtained by random combination of radical groups of the general formula disclosed in the prior art reference may not be deemed as a compound specifically disclosed in the reference. Similar to the interpretation on Markush claims, in this case, the evidence only generally discloses the preparation method of the compounds of the general formula and, only in a very comprehensive list, the starting materials for preparing the compounds of the patent are mentioned. Neither a specific method for preparing the compounds of the patent is disclosed nor a specific compound of the patent is finally obtained in the evidence. Thus, the evidence cannot destroy the novelty.

In this regard, readers of the Invalidation Decision might have been thinking whether perhaps the petitioner made the novelty objection based on “a presumption of novelty” as specified in the Guidelines for Examination, i.e., it is presumed that evidence 3 and 6 “refer to” dimethyl sunitinib so that claim 1 has no novelty. It is also noted by the Patent Office on the “typicality” of this case (http://reexam. Cnipa.gov.cn/alzx/fswxsdaj/22204.htm) that:

As prescribed in the Chinese Guidelines for Examination, for a compound claimed in an application, if it has been “referred to” in a reference document, it is presumed that the compound does not possess novelty. This case affords us a lesson for conducting examination over whether a certain compound is “referred to”; that is, how to identify the the content disclosed in the prior art.

The author tends to believe that the “presumption of novelty” is not applicable to this case. Regarding “presumption of novelty”, the Chinese Guidelines for Examination prescribes that (see part II, chapter 10, section 5.1 of the Guidelines for Examination):

For a compound claimed in an application, if it has been referred to in a reference document, it is presumed that the compound does not possess novelty, unless the applicant can provide evidence to verify that the compound is not available before the date of filing.

As discussed previously, in accordance with the principle of comparison against single piece of prior art, evidence 3 and evidence 6 do not even specifically “refer to” dimethyl sunitinib. Thus, the premise on which the presumption of novelty is established did not exist. Accordingly, judgement of novelty in this case is not necessarily made based on the examination criteria “presumption of novelty”. In contrast, judgement could be simply made in the sense of substantial law, that is to say whether the prior art actually obtains the compound.

[Extracts from Invalidation Decision]:

Those skilled in the art would have had reason to believe that 5-formyl-2,4-dimethyl-1H-pyrrole-3-carboxylic acid (2-dimethylaminoethyl)amide is merely mentioned in a list of the starting materials of formula 3 and involved in a general description of the synthesis method. No proof shows that in evidence 3, each of the specific starting materials of formula 2 and of formula 3 was used and a specific product of general formula (I) was finally synthesized. The petitioner picked out two specific compounds from 41 starting materials of formula 2 as well as 3 starting materials of formula 3 in evidence 3, respectively. From these two specific compounds the petitioner deduced that the product “dimethyl sunitinib” was obtained. Nevertheless, the product is not a compound actually synthesized in evidence 3. Therefore, “dimethyl sunitinib” is not disclosed in evidence 3.

V. Rules and Considerations for Allocation of Burden of Proof

Another point of dispute between the two parties is whether compound 13 in evidence 6 has comparable modulating ability on kinases PDGF and VEGF with the compounds of the patent concerned.

In accordance with the Guidelines for Examination, allocation of burden of proof in the invalidation procedure should still abide by the general principle “who claim, who quote”. The party shall bear the burden of proof to provide evidence for what it alleges or what it refutes during the invalidation. The party that bears the burden of proof shall take the unfavorable consequence where no evidence is provided or the evidence provided is not sufficient to support its allegations.

In the invalidation procedure, petitioner, as the party that claims for the invalidity of the patent, shall bear the initial burden of proof. The evidence submitted by petitioner should sufficiently support its allegations, or else its invalidation request will not be approved. In this case, it was the petitioner’s allegation that the patent compounds have comparable effects to that of compound 13 of evidence 6. In support of the allegation, the petitioner submitted evidence 3 and evidence 6. However, after checking the evidence submitted by the petitioner, it can be found that although evidence 6 discloses the IC₅₀ of compound 13 on PDGF, its test method is different from that of the patent concerned so that the two sets of PGDF active data cannot be compared directly. Moreover, evidence 6 fails to mention the effect of compound 13 on VEGF in in vitro test, on PDGF and VEGF in cell and in vivo tests and results thereof. As a result, the evidence cannot prove the effects of compound 13 are comparable to those of the patent compounds in the above aspects, either. Clearly, the existing evidence submitted by the petitioner is insufficient to validate its allegations and the petitioner thus has to take the unfavorable consequences for failing to provide convincing evidence. Generally speaking, in the field of pharmaceutical compounds, it is usually difficult to compare the activities between the prior art and the patent directly, unless their test methods are identical.

On the other hand, the in vitro biochemical activity on protein kinases such as PDGF and VEGF, and the in vitro and in vivo activities on colon carcinoma, lung carcinoma, neuroglioma and epidermoid carcinoma of the relevant compounds have been sufficiently disclosed by the patentee in the description of the patent. Provided that no counter evidence is provided, it shall be acknowledged that the patent has proved the relevant activities of the patent compounds. Thus, the patentee does not need to provide further proof.

Accordingly, in the present case, the final unfavorable consequences are taken by the petitioner.

In the pharmaceutical field, an effective manner to compare the activities of compounds is to provide comparative data. It would also be beneficial if acceptable comparative data can be provided during the invalidation procedure (with the degree of difficulty for providing such data not being discussed herein). However, the author believes that there is certain rationality behind the requirements that burden of proof for submitting the data be first borne by petitioner rather than patentee. For one thing, petitioner should undertake the responsibility following the principle “who claim, who quote”. For another, petitioner, as the party that requests for invalidation, should have fully understood the experimental methods of the patent concerned, and also have had sufficient time to carry out comparative experiments. On the contrary, the patentee, especially if the patent is of much interest, often receives more than one invalidation requests from different generics, in which different evidence may be involved. If patentee were required to provide comparative data in response to each of the invalidation requests, the burden of proof would be excessively transferred to patentee. This is unfair to patentee from the perspective of procedural justice, timeframe for submitting proof, and cost.

[Extracts of Invalidation Decision]:

Regarding evidence 3 and evidence 6, first, they only disclose the biochemical experiments and results directed to PDGF, and the results cannot be compared with those of the patent as the experiment methods are different, as admitted by the petitioner. Correspondingly, those skilled in the art cannot determine that the biochemical activities of the patent compounds and the compound of evidence 6 are comparable. Further, none of these pieces of evidence discloses cell experiment and results directed to PDGF or VEGF. Furthermore, as for the in vivo experiment, evidence 3 only discloses the result of compound 5, which cannot represent that of compound 13; therefore, even if those skilled in the art accept that compound 13 of evidence 6 has modulating ability over PDGF in biochemical experiment, they could not envisage that it also has modulating ability over PDGF and VEGF under the circumstances of cell experiment and in vivo experiment. In sum, with evidence 6 and evidence 3, the petitioner cannot prove that patent compounds and compound 13 of evidence 6 have comparable technical effect.

VI. Conclusion

It is well known that innovation cost is extremely high for researching new drugs. More often than not, it takes 10 to 15 years or more and spends an average $2.6 billion costs on development to ultimately obtain a new drug which brings benefit to the public. In order to maintain the legitimate interests for drug R&D enterprises and to facilitate innovations on drugs, it is required that petitioner bear the burden of proof in invalidation procedure to prove that the patent is not in conformity with relevant provisions of the Patent Law, and especially when the allegation of the petitioner needs to be supported by comparative tests, the burden of proof on providing the tests should be borne by the petitioner. Further, although there are indeed some issues which are unique of this special technical field in evaluating the actual disclosure of the reference document, the principle of comparison against single piece of prior art shall always be strictly followed, like in other fields. Besides, in order to best balance the interests between patentee and the public, strict restrictions have been made on the amendments to Markush claims.