The Intellectual Property High Court ruled that Amgen's anti-PCSK9 antibody patent was in violation of the support requirement because the patent specification did not clearly state (1) that an antibody competing with a reference antibody for binding to PCSK9 has functional characteristics as a binding neutralizing antibody and (2) the mechanism by which the reference antibody neutralizes the binding. The court reversed the trial decision which dismissed the invalidation trial filed by Regeneron (IP High Court judgment, January 26, 2023 (Case number: 2021 (Gyo-ke) No. 10093).
Outline of the case
Defendant, Amgen Incorporated (hereinafter referred to as "Amgen"), obtained a patent registration for patent No. 5705288 (hereinafter referred to as "the Patent") on March 6, 2015.
Sanofi, which marketed Praluent®, a drug containing an anti-PCSK9 antibody, requested an invalidation trial against the Patent, but Amgen filed a request for amendment, and the JPO approved the amendment and ruled that the invalidation trial was unsuccessful (Case No. 2016-800004). Sanofi filed a suit for revocation of the trial decision, but the suit was dismissed and became final (IP High Court judgment, December 27, 2018 (Case number: 2017 (Gyo-ke) No. 10225) "Invalidation Case by Sanofi"). In this other case, the existence of the grounds for revocation (error in judgment of inventive step, error in judgment of support requirement, and error in judgment of enablement requirement) claimed by Sanofi were all denied.
The Invention 1 (claim 1 after the above amendment) and the Invention 9 (claim 9 after the above amendment) are shown below. The "antibody containing a heavy chain including a heavy chain variable region comprising the amino acid sequence of sequence number 49 and a light chain including a light chain variable region comprising the amino acid sequence of sequence number 23" is referred to as the "21B12 Antibody" and is also referred to as the "Reference Antibody").
［Claim 1］An isolated monoclonal antibody that can neutralize the binding of PCSK9 to LDLR protein and competes with an antibody containing a heavy chain including a heavy chain variable region comprising the amino acid sequence of sequence number 49 and a light chain including a light chain variable region comprising the amino acid sequence of sequence number 23 with respect to binding to PCSK9.
［Claim 9］A pharmaceutical composition comprising an isolated monoclonal antibody of claim 1.
Amgen filed an infringement lawsuit against Sanofi and this lawsuit resulted in an injunction against Praluent®, the drug containing anti-PCSK9 antibody, which Sanofi was selling (IP High Court judgment, October 30, 2019 (Case number: 2019 (Ne) No. 10014), and Praluent® sales have been suspended. After this infringement action against Sanofi, Regeneron Pharmaceuticals Incorporated ("Regeneron"), which co-developed Praluent® with Sanofi overseas, filed an invalidation trial against the Patent, and the JPO rendered a trial decision that the Patent should not be invalidated (the "Trial Decision"). The plaintiff filed this lawsuit to seek revocation of the Trial Decision.
As for the issues, as in the Invalidation Case by Sanofi, Regeneron has argued multiple grounds for revocation, such as inventive step and enablement requirements, etc. However, we will only discuss support requirements in this article, noting that only a judgment on support requirements is presented in the judgment.
Judgement
The court first specified the general criteria for support requirements as follows;
In order to judge whether a claim meets the support requirement, it is reasonable to conclude that a judgment should be made by contrasting the description of the claims with the detailed description of the invention and by examining whether or not the invention is within the scope in which a person skilled in the art can recognize that the subject matter of the invention can be solved by the description in the specifications, or, if the specifications do not provide enough descriptions, by examining whether or not the invention is within the scope in which a person skilled in the art can recognize that the subject matter of the invention can be solved in light of technical common sense at the time of application. |
Then, the court identified the Invention as follows.
About the claims for this invention, claim 1 of the Invention has the elements, (1) "being able to neutralize the binding of PCSK9 to LDLR protein," (2) "competing" with "an antibody containing a heavy chain including a heavy chain variable region comprising the amino acid sequence of sequence number 49 and a light chain including a light chain variable region comprising the amino acid sequence of sequence number 23" (21B12 antibody) (reference antibody) with regard to binding to PCSK9, and (3) "an isolated monoclonal antibody". Elements (1) and (2) are understood to determine the properties of the monoclonal antibody in (3). |
Next, the court clarified the meaning of "neutralization" and "competition" in the claims for the Invention as follows.
“Neutralization" in the Invention refers to interfering with, blocking, reducing, or modulating the interaction between PCSK9 and LDLR protein, and is not limited to directly blocking the PCSK9-LDLR protein binding site, but also includes changing the binding ability of PCSK9 to LDLR protein through indirect means (such as structural or energy changes in the ligand).
|
The court found that the extent and manner of "competition" of the monoclonal antibody with the 21B12 antibody in this case include a variety of factors as follows.
A monoclonal antibody of the Invention that competes with the Reference Antibody and that, to varying degrees, prevents or inhibits (e.g. reduces) the specific binding of the Reference Antibody, is not necessarily limited to monoclonal antibodies that bind to the same sites on PCSK9 where the Reference Antibody binds to PCSK9 and have the property of preventing or inhibiting (e.g. reducing) the specific binding of the Reference Antibody. It is recognized to include a monoclonal antibody that binds to a site on PCSK9 that overlaps with the site on PCSK9 where the Reference Antibody binds to PCSK9 and has the property of preventing or inhibiting (e.g. reducing) the specific binding of the Reference Antibody, and a monoclonal antibody that binds to PCSK9 in a manner that sterically interferes with the binding of the Reference Antibody to PCSK9 and prevent or inhibit (e.g., reduce) the specific binding of the Reference Antibody to PCSK9. |
The court also found that the significance of the Invention lies in the fact that it shows that an antibody that competes with the 21B12 Antibody will have the same mechanism to neutralize the binding between PCSK9 and the LDLR protein as a 21B12 Antibody.
In the first place, the subject matter of the Invention is to provide an antibody that neutralizes the binding between PCSK9 and LDLR protein, which decreases the amount of LDLR protein in the subject and increases the amount of LDL by binding to LDLR protein, or a pharmaceutical composition containing this antibody. In relation to the solution of such a problem, competing with a Reference Antibody itself has no original meaning. From this viewpoint, as described above, the technical significance of the Invention should be that it specifies that any antibody that competes with the 21B12 Antibody has functional characteristics as a binding-neutralizing antibody by the same mechanism as the 21B12 Antibody. |
The court then went on to say that, as for the description of the specification, the antibody competing with 21B12 Antibody to neutralize the LDLR protein includes a great variety of antibodies, and since no specific explanation is found for each of them, it is not necessarily true that they compete and neutralize the binding by the same mechanism as 21B12 Antibody.
In the specification, there is no specific description of the position where the antibody identified as one with neutralizing activity binds to PCSK9 among the antibodies identified as a competing antibody above. It is highly probable that antibodies with amino acid sequences that are highly identical to those of the 21B12 Antibody bind to PCSK9 at the same position as the 21B12 Antibody However, the position where several groups of antibodies with other amino acid sequences binds to PCSK9 is not clear even if they were evaluated as competing antibodies in assays such as epitope binning. |
In addition, it is obvious that antibodies having the property of "competing with a Reference Antibody with respect to binding to PCSK9" include a great variety of antibodies other than the several groups of antibodies specifically described in the detailed description of the Invention in the above specification.... |
The specification states that "Antigen binding proteins and fragments that compete with or bind to the same epitopes as the exemplified antigen binding proteins are expected to exhibit similar functional properties." ([0269]), but as stated above, "competing with the 21B12 Antibody for binding to PCSK9” does not specify that it binds to PCSK9 at the same position as the 21B12 antibody. Therefore, an antibody that competes with the 21B12 Antibody does not mean that it is an antigen-binding protein (antibody) that competes or binds to the same epitope as the 21B12 Antibody. Since there is no specific explanation of the mechanism by which such antibodies in general exhibit similar functional characteristics to the 21B12 Antibody, it cannot be said that "antibodies competing with the 21B12 Antibody for binding to PCSK9" in the Invention "exhibit similar functional characteristics" to the 21B12 Antibody. |
As described above, the court found that the technical significance of the Invention is that it clarifies that an antibody competing with 21B12 Antibody can directly sequester the binding site of LDLR protein and neutralize the binding between PCSK9 and LDLR protein by the same mechanism as 21B12 Antibody, but according to the specification, the position of the binding site of the Reference Antibody and the competing antibody to PCSK9 is not clear.
The court also found that the specification did not indicate that "an antibody competing with a 21B12 antibody with respect to binding to PCSK9" exhibits similar functional characteristics to the 21B12 antibody. The court held that the invention did not meet the support requirements and revoked the trial decision.
The plaintiff claimed that Sanofi's Praluent is one of the "EGFa mimic antibody" and the claims of the Invention include an “EGFa mimic antibody,” but the method described in the specification fails to provide an "EGFa mimic antibody," and therefore this should be also a reason for violation of the support requirement. The court avoided making a clear judgment on this point, but suggested that there was room for this argument to be accepted.
In addition, among the plaintiff's claims, the point concerning "EGFa mimic antibody" contains something worthy of affirmation, and we believe that it raises doubts about the defendant's claim that the requirements of support are satisfied. However, we refrain from making a further judgment because neither of the inventions 1 and 9 can be found to conform to the support requirements as described above. |
Comments
This decision is noteworthy because the court denied the patentability of the Patent, although the previous courts found that the Patent is valid in other patent infringement lawsuits and lawsuits for revocation of trial decisions regarding the Patent.
In the Trial Decision, the Japan Patent Office stated as follows.
The results of the demonstration experiment in Dr. Frenzel's affidavit (1) (A2-1) and Dr. Reichmann's affidavit (1) do not affect the conformity of this case with the support requirements. Rather, the results of the above demonstration experiment support the conformity of this case to the support requirement. In other words, the above demonstration experiment showed that 3 out of 13 antibodies (23%) competing with 21B12 antibodies screened from anti-PCSK9 monoclonal antibodies produced by the same method as described in the specification neutralized the binding between PCSK9 and LDLR. This indicates that even if the order of screening for "being able to neutralize the binding between PCSK9 and LDLR" and screening for "competing with 21B12 antibody" is reversed from the example in this specification, it is possible to obtain a number of antibodies of the invention with a sufficiently high probability. |
Based on the evidence from empirical experiments and affidavits and other evidence in response, the JPO concluded that a sufficiently high probability of obtaining a number of antibodies of the patented invention could be achieved because 3 of the 13 antibodies (23%) competing with the 21B12 antibody have binding neutralizing properties.
On the contrary, in this lawsuit, Plaintiff, claimed that
the affidavit of Dr. [A] (A2-1) shows the result of obtaining a number of monoclonal antibodies competing with PCSK9 obtained by using well-known technology at the time of the priority date of this case, conducting competition tests with 21B12 Antibody and PCSK9-LDLR binding neutralization tests. According to the results, 80% of the antibodies (10 out of 13 antibodies), the majority of which were competing antibodies, were unable to neutralize the binding of 21B12 antibodies. |
The Plaintiff alleged that the majority of antibodies were ones which could not bind and neutralize. As a result, the court only stated that "the results of the demonstration experiment submitted as evidence are reliable" with regard to the affidavit of Dr. [A] (Dr. Frenzel) (A2-1), but it is interesting to note the difference in evaluation of the same evidence (fact) between the JPO and the IP High Court.
The Patent was a functional claim that did not specify the structure of the antibody itself. Claims such as "an inhibitor that competes with ~" have the advantage that if they are allowed, the applicant can obtain a broad scope of protection. However, if all such claims were allowed, the applicant would also have a monopoly on inventions based on mechanisms completely different from those identified by the inventor. It is anticipated that a person who has obtained a protein inhibitor X may seek to obtain a patent right with a functional claim that competes with X, and in such a case, the alleged infringer would be required to actually obtain said protein inhibitor X for an infringement clearance search to see whether it competes with its own product.
Therefore, the decision to find the Patent in violation of the support requirement seems to be reasonable .
On the same day, a judgment (2021 (Gyo-ke) No. 10094) was issued to revoke a trial decision which found another Amgen's antibody patent (patent No. 5906333) valid, for the same reasons as this decision. This other Amgen patent was also found to be infringed together with the Patent in the infringement suit against Sanofi, and differs only in Reference Antibody from the Patent.
In Europe, a corresponding European patent that was substantially the same as the invention was found by the Court of Appeals for Opposition and Interlocutory Appeal to be invalid for lack of inventive step in 2020, and in the U.S., the Supreme Court ruled invalid on a corresponding U.S. patent, affirming the decision of the CAFC in 2023.
Written by: Ms. Wakana Furuhashi (Attorney at Law)