Date:20 April 2023

【Volume 108】

Taiwan Intellectual Property Court (IP Court) recently ruled in favor of the generic drug company in a declaratory judgment for a P4 declaration filed by Celltrion Healthcare Taiwan (the generic drug company) against F. Hoffmann-La Roche AG (the brand drug company). This is the second case where the generic drug company obtained a market exclusivity following Bayer v. Synmosa Biopharma Corp¹. after the patent linkage system was officially implemented in Taiwan on August 20, 2019. In addition, in this case, the IP court provided clear judging criteria as to whether an erroneous prior art disclosure has legal validity, and this is also the first time the IP court provided a definitive legal opinion on this issue.

Case Facts

Celltrion (appellee, plaintiff) claimed that its pharmaceutical product “Truxima” was approved by Taiwan Food and Drug Administration (TFDA) of the Ministry of Health and Welfare under “Wei Bu Jyun Yi Shu Zi No. 001094 drug permit license”, and its indications originally included Rheumatoid Arthritis (RA), which was deleted because Roche’s (appellant, defendant, patentee) attorney informed Celltrion that such indication is likely to infringe its invention patent “Method for Treating Joint Damage” (Certificate Number: I380826) (hereinafter ‘826 patent). Since Roche did not file a complaint against Celltrion within 45 days after the next day to its receipt of the notification of the P4 declaration (Paragraph 1 of Article 48-13 of the Pharmaceutical Affairs Act), Celltrion thus initiated an action for a declaratory judgment to dictate that Roche has no right to claim based on ‘826 patent and also asserted that the ‘826 patent was subject to various grounds for revocation. The judgment of the first instance² ruled in favor of Celltrion. Roche disagreed and thus appealed to the second instance. However, the IP court of the second instance upheld the decision of the original judgment and dismissed the appeal³.

Main Issues of the Case

1. Whether an erroneous prior art disclosure may support obviousness
2. Whether Celltrion Exhibit 9 is sufficient to prove that post-grant amended claim 1 of the ‘826 patent does not have an inventive step

Arguments of the Patentee

1. Celltrion Exhibit 9 was filed to prove the ‘826 patent obvious. The patentee challenged many contradictions described in Celltrion Exhibit 9, such as skipped numbers and typographical errors in the description about the references, and the erroneous description of administration dose. In view of the fact that there is no regulation about correction of mistakes and typographical errors in the disclosure of citation in Taiwan Patent Examination Guidelines, the judgment of patentability of the ‘826 patent should be based on the disclosure of Celltrion Exhibit 9 in the form, i.e., “an intravenous dose of 100 mg Rituximab”.
2. The patentee claimed that Celltrion Exhibit 9 did not disclose any radiographic data to prove that the treatment and retreatment regimen of Rituximab in combination with methotrexate was effective in inhibiting “the progression of structural damage and erosions” in a subject (who has exhibited an inadequate response to one or more anti-tumor necrosis factor (TNF) inhibitors) as defined in the ‘826 patent; and that there was no relation between “the progression of structural damage and erosions” and “inflammation”.

The opinion of the Intellectual Property and Commercial Court

1. The erroneous disclosure in the citation still has legal validity:

   Base on the regulations in Taiwan Patent Examination Guidelines,⁴ “when a citation is cited to judge the novelty of an invention, the technical contents disclosed in the citation shall be based upon. Said technical contents include both those which have been formally and explicitly stated and those substantially implied but not formally stated therein. The disclosure of a citation must be such that a person ordinarily skilled in the art can reproduce and use the claimed invention. The term ‘substantially implied’ refers to the contents that a person ordinarily skilled in the art can directly and unambiguously derive from the citation on the basis of common general knowledge at the time the citation was published. (Common general knowledge at the time of filing for determining inventive step is explicated in 3.2.4 “Prior Art” of this Chapter.). The prior art explicitly mentioned in a citation shall be considered part of the citation.” The court considered that the reference cited in Celltrion Exhibit 9 was, of course, the prior art explicitly described in Celltrion Exhibit 9 and belonged to a part of the disclosure of Celltrion Exhibit 9. Consequently, the description “administration dose of 100 mg” should be corrected to 1000 mg according to Reference 3 (Celltrion Exhibit 11) listed in Celltrion Exhibit 9. Thus, the administration dose after correction would be the same as the technical means of the ‘826 patent.

   There should be only three cited references throughout Celltrion Exhibit 9. However, there were five numbers in the bibliography of references. Also, the only content of Number 2 was a blank line and the only content of Number 4 was one word, “Abstract.”, which were obvious mistakes in the bibliography of references. A person ordinarily skilled in the art would appreciate that there were skipped numbers and typographical errors in the bibliography of references, and therefore would analyze and compare substantially the content cited in the context of Celltrion Exhibit 9 and those references when understanding the technical disclosure of Celltrion Exhibit 9 as a whole and reasonably deduce the correct correspondence.

   Besides, Celltrion Exhibit 9 disclosed that part of the subjects received the second or third course of Rituximab treatment with two 100 mg intravenous (IV) doses given two weeks apart, which was different from the administration “dose” of retreatment claimed in the ‘826 patent (two 1000 mg IV doses of Rituximab given 14 days apart). However, according to the content of Celltrion Exhibit 9, the first course of Rituximab treatment included a low dosage group with two 500 mg IV doses given 14 days apart, or a high dosage group with two 1000 mg IV doses given 14 days apart. If the second or third course of Rituximab treatment was two 100 mg IV doses given 14 days apart based on Celltrion Exhibit 9, a person ordinarily skilled in the art could easily notice that the dose of retreatment is 1/5 or 1/10 of the first course of treatment, which was apparently not able to achieve the purpose of retreatment, which was to maintain the efficacy of the first treatment. This should be a typographical error.

   Celltrion Exhibit 9 had already illustrated that the details of Rituximab retreatment cite Celltrion Exhibit 11, which disclosed the follow-up course of treatment included two 1000 mg IV doses of Rituximab given two weeks apart. Therefore, a person ordinarily skilled in the art would appreciate that a dose of 100 mg of follow-up treatment disclosed in Celltrion Exhibit 9 should be 1000 mg in typographical error based on the dose of follow-up treatment disclosed in Celltrion Exhibit 11. It should be interpreted as two 1000 mg IV doses given two weeks apart.

2. Celltrion Exhibit 9 is sufficient to prove that the post-grant amended claim 1 of the ‘826 patent lacks an inventive step:

   The difference between Celltrion Exhibit 9 and post-grant amended Claim 1 of the ‘826 patent lies in the fact that: (1) there is no direct and unambiguous disclosure of the subject as claimed (who has exhibited an inadequate response to one or more anti-tumor necrosis factor (TNF) inhibitors) in Celltrion Exhibit 9; and (2) there is no direct and unambiguous disclosure of the efficacy of the treatment to inhibit “structural joint damage and erosions”.

   However, Celltrion Exhibit 9 had revealed that approximately 27-32% of patients recruited had an inadequate response to anti-TNF inhibitors, and therefore the technical concept of initial Rituximab treatment to patients with an inadequate response to anti-TNF inhibitors had been substantially revealed. In addition, considering relevant common general knowledge and the supplementary evidence of the technical level provided by the two parties, a person ordinarily skilled in the art can understand that joint inflammation (synovitis) may cause structural damage to the joint, which is also supported by the prior art stated in the ‘826 patent specification. Hence, under the premise that the treatment with Rituximab has been shown to produce the therapeutic efficacy on RA, a person ordinarily skilled in the art would expect the aforementioned treatment to treat or relieve chronic inflammation in the synovial membrane of joints (i.e., synovitis), thereby preventing or alleviating structural or erosive joint damages, and the relevant efficacy could be confirmed by means of routine test. Thus, Celltrion Exhibit 9 is sufficient to prove the post-grant amended claim 1 of the ‘826 patent lacks an inventive step.

3. Roche’s request to summon an expert witness and consultation with TFDA of the Ministry of Health and Welfare to determine the technical levels of a person ordinarily skilled in the art is obviously unnecessary and should not be granted:

   Roche would like to request to summon an expert witness, Dr. Breedveld. However, the court found that Dr. Breedveld had provided several declarations to understand his opinions, so it was not necessary to summon him.

   Regarding consultation with TFDA of the Ministry of Health and Welfare, the court held that the application for registration and market approval of a new drug or registration of supplements or amendments to drug indications and patent examination are two separate matters. Moreover, the patent is mainly for the purpose of promoting industrial development, so the technical levels of a person ordinarily skilled in the art is determined by the technical contents revealed by the prior art of the relevant industry, which is obviously different from the way that drug registration is examined according to safe and direct experimental data without necessarily considering the technical levels of a person ordinarily skilled in the art. As a result, regardless of what the TFDA’s reply is, it cannot be used as a basis for determining the technical level of the ‘826 patent before its priority date, not to mention the judgment of whether the ‘826 patent has an inventive step.